Réponse du grain de blé à la nutrition azotée et soufrée :

étude intégrative des mécanismes moléculaires mis en

jeu au cours du développement du grain par des
analyses -omiques
Titouan Bonnot

To cite this version:

Titouan Bonnot. Réponse du grain de blé à la nutrition azotée et soufrée : étude intégrative
des mécanismes moléculaires mis en jeu au cours du développement du grain par des analyses
-omiques. Sciences agricoles. Université Blaise Pascal - Clermont-Ferrand II, 2016. Français. ¡
NNT : 2016CLF22767 ¿.

HAL Id: tel-01544587

[Link]
Submitted on 21 Jun 2017

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.
UNIVERSITE BLAISE PASCAL UNIVERSITE D’AUVERGNE
N°DU : 2767 ANNEE : 2016

ECOLE DOCTORALE SCIENCES DE LA VIE,

SANTE, AGRONOMIE, ENVIRONNEMENT
N° d’ordre : 706

These:
Présentée à l’Université Blaise Pascal
pour l’obtention du grade de

DOCTEUR D’UNIVERSITE

(SPECIALITE : PHYSIOLOGIE ET GENETIQUE MOLECULAIRES)

soutenue le : 9 décembre 2016

TITOUAN BONNOT

Réponse du grain de blé à la nutrition azotée et soufrée :

Etude intégrative des mécanismes moléculaires mis en jeu au cours du
développement du grain par des analyses -omiques

TATOUT Christophe Professeur Université Blaise Pascal, Clermont-Ferrand Président

BUITINK Julia Directeur de Recherches INRA, Angers Rapporteur
RAJJOU Loïc Maître de Conférences AgroParisTech, Paris Rapporteur
GALLARDO Karine Directeur de Recherches INRA, Dijon Examinateur
ZIVY Michel Directeur de Recherches CNRS, Gif-sur-Yvette Examinateur
BANCEL Emmanuelle Ingénieur d'Etudes INRA, Clermont-Ferrand Co-encadrante
MARTRE Pierre Directeur de Recherches INRA, Montpellier Co-encadrant
RAVEL Catherine Ingénieur de Recherches INRA, Clermont-Ferrand Directrice de thèse

UMR 1095 INRA / UBP « Génétique, Diversité et Ecophysiologie des Céréales »

5, chemin de Beaulieu - 63039 Clermont-Ferrand Cedex 02 – France
Réponse du grain de blé à la nutrition azotée et soufrée : Etude intégrative des
mécanismes moléculaires mis en jeu au cours du développement du grain par
des analyses -omiques

RÉSUMÉ
L’augmentation des rendements est un enjeu majeur chez les céréales. Dans cet objectif, il est
nécessaire de maintenir la qualité du grain de blé, qui est principalement déterminée par sa teneur
et sa composition en protéines de réserve. En effet, une forte relation négative existe entre le
rendement et la teneur en protéines. Par ailleurs, la qualité du grain est fortement influencée par la
disponibilité en azote et en soufre dans le sol. La limitation des apports d’intrants azotés à la culture
et la carence en soufre récemment observée dans les sols représentent ainsi des difficultés
supplémentaires pour maitriser cette qualité. Une meilleure connaissance des mécanismes
moléculaires impliqués dans le contrôle du développement du grain et la mise en place de ses
réserves protéiques en réponse à la nutrition azotée et soufrée est donc primordiale. L’objectif de
cette thèse a ainsi été d’apporter de nouveaux éléments à la compréhension de ces processus de
régulation, aujourd’hui peu connus. Pour cela, les approches -omiques sont apparues comme une
stratégie de choix pour identifier les acteurs moléculaires mis en jeu. Le protéome nucléaire a été
une cible importante dans les travaux menés. L’étude de ces protéines nucléaires a révélé certains
régulateurs transcriptionnels qui pourraient être impliqués dans le contrôle de la mise en place des
réserves du grain. Dans une approche combinant des données de protéomique, transcriptomique et
métabolomique, une vision intégrative de la réponse du grain à la nutrition azotée et soufrée a été
obtenue. L’importance d’un apport de soufre dans le contrôle de la balance azote/soufre du grain,
déterminante pour la composition du grain en protéines de réserve, a été clairement vérifiée. Parmi
les changements observés au niveau du métabolisme cellulaire, certains des gènes affectés par la
modification de cette balance pourraient orchestrer l’ajustement de la composition du grain face à
des situations de carences nutritionnelles. Ces nouvelles connaissances devraient permettre de
mieux maitriser la qualité du grain de blé dans un contexte d’agriculture durable.

Mots clés : Blé, grain, protéines de réserve, azote, soufre, protéines nucléaires, données
omiques, réseaux biologiques
Wheat grain response to nitrogen and sulfur supply : Integrative study of
molecular mechanisms involved during the grain development using -omics
analyses

ABSTRACT
Improving the yield potential of cereals represents a major challenge. In this context, wheat grain
quality has to be maintained. Indeed, grain quality is mainly determined by the content and the
composition of storage proteins, but there is a strongly negative correlation between yield and grain
protein concentration. In addition, grain quality is strongly influenced by the availability of nitrogen
and sulfur in soils. Nowadays, the limitation of nitrogen inputs, and also the sulfur deficiency
recently observed in soils represent major difficulties to control the quality. Therefore,
understanding of molecular mechanisms controlling grain development and accumulation of
storage proteins in response to nitrogen and sulfur supply is a major issue. The objective of this
thesis was to create knowledge on the comprehension of these regulatory mechanisms. For this
purpose, the best strategy to identify molecular actors involved in these processes consisted of -
omics approaches. In our studies, the nuclear proteome was an important target. Among these
proteins, we revealed some transcriptional regulators likely to be involved in the control of the
accumulation of grain storage compounds. Using an approach combining proteomic,
transcriptomic and metabolomic data, the characterization of the integrative grain response to the
nitrogen and sulfur supply was obtained. Besides, our studies clearly confirmed the major influence
of sulfur in the control of the nitrogen/sulfur balance that determines the grain storage protein
composition. Among the changes observed in the cell metabolism, some genes were disturbed by
the modification of this balance. Thus these genes could coordinate the adjustment of grain
composition in response to nutritional deficiencies. These new results contribute in facing the
challenge of maintaining wheat grain quality with sustainable agriculture.

Keywords: Wheat, grain, storage proteins, nitrogen, sulfur, nuclear protein, -omics data,
biological network
Université Blaise Pascal Université d’Auvergne

Ecole Doctorale Sciences de la Vie, Santé, Agronomie, Environnement

Thèse de doctorat

Présentée à l’Université Blaise Pascal

pour l’obtention du grade de

DOCTEUR D’UNIVERSITÉ
Spécialité : Physiologie et génétique moléculaires

Réponse du grain de blé à la nutrition azotée et soufrée :

Etude intégrative des mécanismes moléculaires mis en jeu au cours du
développement du grain par des analyses -omiques

Par Titouan Bonnot

Soutenue le 9 décembre 2016 devant un jury composé de :

TATOUT Christophe Professeur Université Blaise Pascal, Clermont-Ferrand Président

BUITINK Julia Directeur de Recherches INRA, Angers Rapporteur
RAJJOU Loïc Maître de Conférences AgroParisTech, Paris Rapporteur
GALLARDO Karine Directeur de Recherches INRA, Dijon Examinateur
ZIVY Michel Directeur de Recherches CNRS, Gif-sur-Yvette Examinateur
BANCEL Emmanuelle Ingénieur d'Etudes INRA, Clermont-Ferrand Co-encadrante
MARTRE Pierre Directeur de Recherches INRA, Montpellier Co-encadrant
RAVEL Catherine Ingénieur de Recherches INRA, Clermont-Ferrand Directrice de thèse
 REMERCIEMENTS
Avant de débuter ce manuscrit, je tiens à remercier sincèrement toutes les personnes qui ont contribué
de près ou d’un peu plus loin à l’aboutissement de ce projet de thèse.
Tout d’abord je remercie les membres du jury pour avoir accepté de lire et juger mes travaux. Merci à
Julia Buitink, Karine Gallardo, Loïc Rajjou, Michel Zivy et Christophe Tatout (et bien sûr Pierre,
Catherine et Emmanuelle, également membres du jury, que je remercierai plus bas).
Merci aux membres de mes deux comités de thèse, qui m’ont donné de précieux conseils pour avancer
ce projet. Je remercie donc à nouveau Karine Gallardo qui a participé à ces deux comités, puis Colette
Larré, Ludovic Bonhomme et Christophe Chambon.
A présent je souhaite adresser un grand merci à toutes les personnes avec qui j’ai eu l’occasion de
travailler, en commençant par celles et ceux qui ont été à l’origine de ce projet et qui ont participé à
mon encadrement. Et priorité aux personnes qui se sont succédées à la direction de ma thèse. Si l’on
suit l’ordre chronologique, merci tout d’abord à Gérard Branlard. J’ai été très honoré d’avoir été l’un
de ses derniers thésards. Sa vision de la recherche et ses innombrables connaissances sont pour moi
un exemple à suivre. Merci à Pierre Martre, de qui j’ai beaucoup appris. Ta rigueur scientifique et ta
capacité d’étendre à l’infini les interprétations des résultats biologiques sont pour moi encore
aujourd’hui un mystère et en même temps une grande source d’inspiration. Merci ensuite à Catherine
Ravel, qui a fait preuve d’une très grande capacité d’adaptation suite au départ de Pierre et qui a très
bien repris la direction de ce projet. Toujours de très bon conseil, je retiendrai ton énergie débordante,
ta grande honnêteté et ta capacité de prendre rapidement les bonnes décisions. A présent j’adresse
un très grand merci à Emmanuelle Bancel, qui a été au plus près de mon encadrement. Tu es sans
aucun doute celle qui a été la plus compréhensive et à l’écoute. Merci pour ta très grande disponibilité,
tes nombreux conseils et pour m’avoir toujours proposé ton aide. Et bien sûr merci de m’avoir formé
dans le domaine de la protéomique. Pour terminer avec les personnes qui m’ont encadré, merci à Julie
Boudet. J’ai beaucoup apprécié tes conseils (je retiendrai ton aide précieuse dans ma préparation pour
le concours de l’école doctorale), pour avoir toujours su trouver un moment pour suivre ce projet et
m’aider dans son avancement, et pour ton accompagnement lorsque j’ai voulu plonger dans le bassin
de l’enseignement.
Je tiens à adresser un très grand merci à l’ensemble des personnes de l’équipe BIG, qui ont participé
au climat chaleureux dans lequel s’est déroulée cette thèse. J’ai beaucoup apprécié la période de
culture de Triticum monococcum qui a mobilisé tout le monde et qui pour moi a été le symbole de
l’unité de cette équipe (c’est peut-être un mauvais souvenir pour vous ! Je pense notamment aux
notations de floraison le samedi et dimanche matin, ou encore le planning de prélèvement des grains
qui était un véritable casse-tête pour toi Annie et toi David ! Un peu speed mais ça fait des bons
souvenirs). Donc merci à ceux qui n’ont pas été cités plus haut : Annie, Marielle, Mireille, Sibille,
Isabelle et David. David, merci aussi pour les sérieuses parties de fléchettes (si, à ce niveau c’est
sérieux !) et les discussions rugby pour déconnecter le temps de midi. Je n’oublie pas bien sûr les filles
qui ont été de passage : Ivonne, Céline et Selver, merci pour votre bonne humeur quotidienne.
Merci aux autres personnes, du GDEC et d’ailleurs qui m’ont aidé dans mon travail, et particulièrement
Marie Pailloux pour sa disponibilité et sa pédagogie pour expliquer simplement le fonctionnement de
RulNet à un novice comme moi ; Christophe Chambon, Didier Viala, Marlène Davanture, Thierry
Balliau, Olivier Langella, Michel Zivy pour leur aide et leur expertise en protéomique, Philippe Leroy
pour l’annotation fonctionnelle des gènes, sans oublier Richard, Stéphane, Christian, Raphaël et Noël
pour la gestion des plantes en serre, et puis Karine pour la gestion des déplacements et Valérie pour
la gestion du personnel non permanent. Merci à toutes les autres personnes du GDEC et du site de
Crouël en général, avec qui j’ai eu l’occasion de passer un moment, à la cantine (un grand merci au
personnel de la cantine en passant !), aux fléchettes, au ping pong…
Merci à toi Victor pour ton aide sur la fin de ma thèse. Tu as été le bon stagiaire pour aider un thésard
de troisième année : autonome et efficace ! Impec !
Merci aux thésards qui ont partagé le même bateau que moi et qui ont souvent été là pendant ces
trois ans. Honneur à miss Delphine qui a soutenu en premier, merci pour ta bonne humeur à toute
épreuve, ton soutien et les échanges qu’on a pu avoir (de vannes, de ping pong, de volley, de stress… !).
Merci Julio, on aura bien rigolé pendant tes séjours à Clermont (je sais que tu es souvent revenu juste
pour ça), même si la défaite contre le Portugal est encore dure à digérer (je sais tu y es pour rien mais
bon je garde en mémoire ton appel de Lisbonne 5 min après la fin du match…). Merci Robin pour avoir
supporté mes vannes (pas drôles ?), pour ton écoute, les moments à la salle (ça a pas duré
longtemps ;)) à la machine à caf ou autre (cette fois je crois qu’on va définitivement bosser à deux
endroits différents, fallait bien que ça arrive depuis le stage de M1!). Florent merci à toi sur tous les
fronts : dans le boulot, aux pauses café, dans les moments de détente en dehors… (et je ne te remercie
pas pour le nombre incalculable de roustes que j’ai pu prendre à cause de toi au ping). Merci Bianca
(oui je t’intègre dans le paragraphe thésards parce que je fais ce que je veux et je ne peux pas te
dissocier de Flo) pour avoir ramené un peu le soleil d’Italie ! Merci à toi Seb, tu as été tellement
important pour les moments de déconnexion et de craquage à la salle, au ping-squash, à Marcombes
ou autre ! Merci à toi Romain pour les mêmes raisons, le seul regret c’est que l’on n’ait pas débuté la
thèse ensemble… Merci Emilie, tu m’auras bien eu en ne voulant jamais parler de la thèse et en
m’annonçant deux semaines avant que tu allais rendre le manuscrit ! Bien joué ! Bien joué à toi aussi
Auré, tu as bien géré sur la fin, entre Clémence qui s’est mise à la peinture et au tricot à Limoges (dur
pour toi d’aller là-bas le week end…) et l’écriture d’un manuscrit de thèse de Physique ! Machine… Je
plaisante Clem, merci à toi aussi !
Merci à tous les autres : Thomas, GC et son fillot Antho, Floriane, Estelle, Gaétan et Agathe, Emma et
Kiki, Boubou, Yann le PDG, Vélo, Boutet, Jeff. Merci les amis pour les bons moments passés ensemble,
à Clermont, Orange, Sauternes, au FreeMusic…
Merci à toute la famille Gillard, avec qui j’ai passé de super week ends et de super vacances. Merci à
Christine, Malo, Tristan, Maël, Sylvie, Eric, Nadine, Marion, Tanguy et Romain.
Merci à mes parents. Sans votre soutien et votre compréhension je crois que cela n’aurait jamais été
possible, j’en suis sûr même. Vous m’avez toujours fait confiance et encouragé sans jamais me mettre
la pression. Merci pour tout. Merci à ma sœur Charlotte. Merci d’avoir expérimenté avant moi les 100
meilleures façons d’être privé de sortie. Je plaisante (à moitié), merci à toi Cha, la vraie grande sœur
protectrice toujours là pour moi. C’est bien sûr aussi grâce à toi si j’en suis là. Seb je ne t’oublie pas
merci beaucoup à toi, avec Cha vous avez toujours été là, merci pour les concerts, festivals, soirées,
séances travaux… Des moments tellement importants pour se défouler. Merci à ma grand-mère
Christiane et mon oncle Régis. Merci pour votre accueil sur la région parisienne lors de mes séjours à
Gif, et pour les très bons week ends passés ensemble là-bas. A ma famille, un immense merci pour
votre soutien.
40,000. Ce chiffre te parlera sûrement puisque c’est toi qui l’as estimé. Pour moi ce manuscrit
représente une étape importante sur le plan professionnel, mais marque surtout la fin d’une difficulté
sûrement plus grande que la thèse en elle-même. Merci à toi Morgane d’avoir été là. Promis, les
prochains 40,000 on les parcourra ensemble… « Crack the code, ….»
Mes derniers remerciements vont vers toi Yann, merci pour tout…
 VALORISATION DES TRAVAUX DE THÈSE

Publications
Bancel, E.*, Bonnot, T.*, Davanture, M., Branlard, G., Zivy, M., & Martre, P. (2015).
Proteomic approach to identify nuclear proteins in wheat grain. Journal of Proteome
Research, 14, 4432–4439. *Co-premiers auteurs

Bonnot, T., Bancel, E., Chambon, C., Boudet, J., Branlard, G., & Martre, P. (2015).
Changes in the nuclear proteome of developing wheat (Triticum aestivum) grain. Frontiers
in Plant Science, 6.

Bonnot, T., Bancel, E., Alvarez, D., Davanture, M., Boudet, J., Pailloux, M., Zivy, M.,
Ravel, C. and Martre P. Grain subproteome responses to nitrogen and sulfur supply in
diploid wheat Triticum monococcum ssp. monococcum. Plant Journal. (Soumis)

Bonnot, T., Hatte, V., Bancel, E., Dardevet, M., Leroy, P., Boudet, J., Moing, A., Gibbon,
Y., Pailloux, M., Martre, P. and Ravel, C. Integrative response of the einkorn (Triticum
monococcum) grain to nitrogen and sulfur nutrition. (En préparation)

Participation à des congrès

Bonnot, T., Bancel, E., Boudet, J., Chambon, C., Branlard, G and Martre, P. (2013)
Analyse des protéines nucléaires du grain de blé tendre. Communication orale au 4ème colloque
national Graines, Dijon, France.

Bonnot, T., Bancel, E., Boudet, J., Chambon, C., Branlard, G. and Martre, P. (2015)
Changes in nuclear proteome of developing wheat grain. Communication orale au IV
International Conference on Analytical Proteomics, Lisbonne, Portugal.

Branlard, G., Bancel, E., Bonnot, T., Boudet, J., Dardevet, M., Merlino, M., Nadaud, I.,
Ravel, C. (2015) Fifteen years of proteomics of developmental wheat grain. Communication
orale au IV International Conference on Analytical Proteomics, Lisbonne, Portugal.

Bonnot, T., Bancel E., Alvarez, D., Dardevet, M., Boudet, J., Davanture, M., Moing, A.,
Gibon, Y., Zivy, M., Pailloux, M., Martre, P. and Ravel, C. (2016) Integration of “omics”
data provides tools for understanding effects of variable nitrogen and sulfur supply during
wheat grain development. Pésentation sous forme de poster au COST Action FA1306- Diving
into integrative cell phenotyping through “omics”, Versailles, France.
 LISTE DES ABBRÉVIATIONS ET SIGLES

°Cj Degrés jours

2D gel Two-dimensionnal gel electrophoresis
ACN Acétonitrile
ADN Acide désoxyribonucléique
ADP Adénosine diphosphate
alg Albumines-globulines
ANOVA Analysis of variance
APS Adenosine 5’-phosphosulfate
ARN Acide ribonucléique
ARNm Acide ribonucléique messager
Asn Asparagine
Asp Aspartate
ATP Adénosine triphosphate
BIG Biologie Intégrative de la composition du Grain
BLAST Basic Local Alignment Search Tool
BSA Bovin serum albumin
CBB Coomassie Brilliant Blue
CPM Count Per Million
CS Cystéine Synthase
DM Dry Mass
DNA Deoxyribonucleic acid
DTT Dithiothréitol
ECL Enhanced chemiluminescence
ESI Electrospray Ionization
FACT Facilitate chromatin transcription
FAO Food and Agriculture Organization of the United Nations
FM Fresh Mass
FT Facteur de Transcription
GAPDH Glyceraldehyde-3-phosphate deshydrogenase
GDC Glycine decarboxylase complex
GDEC Génétique Diversité et Ecophysiologie des Céréales
GLM GCN4-Like Motif
Gln Glutamine
Glu Glutamate
GO Gene Ontology
GOGAT Glutamine OxoGlutarate AminoTransferase
GPD Grain Protein Deviation
GS Glutamine Synthétase
GSH Glutathion réduit
GSP Grain Storage Protein
H2S Sulfure d’hydrogène
HATS High-Affinity Transport System
HDAC Histone désacétylase
HMG High-Mobility Group
HMW-GS High-Molecular-Weight Glutenin Subunit
HPLC High Performance Liquid Chromatography
HRP Horseradish peroxidase
HSD Honestly Significant Different
HSP Heat Shock Protein
Ile Isoleucine
IPG Immobilized pH gradient
IT Ion Trap
LATS Low-Affinity Transport System
LC Liquid Chromatography
LMW-GS Low-Molecular-Weight Glutenin Subunit
LTQ Linear Trap Quadrupole
Lys Lysine
Ma Million d’années
Mal Malate
MBD Methyl-CpG DNA Binding Domain
Met Methionine
MS/MS Tandem Mass Spectrometry
Mt Million de tonnes
N Azote
N2 Diazote
NAD Nicotinamide adénine dinucléotide
NH4+ Ammonium
NLS Nuclear Localization Signal
NO2- Nitrite
NO3- Nitrate
NPC Nuclear Pore Complex
PB Prolamine Box
PBF Prolamine-box Binding Factor
PCA Principal Component Analysis
PHD Plant Homeodomain
PMG Poids de Mille Grains
PMSF Fluorure de phénylméthylsulfonyle
PR Protéines de Réserve
q/ha Quintaux par hectare
RNA Ribonucleic acid
RNA-Seq Ribonucleic acid Sequencing
S Soufre
SDS Sodium dodécylsulfate
SDS-PAGE Sodium dodécylsulfate polyacrylamide gel electrophoresis
SO2 Dioxyde de soufre
SO42- Sulfate
TFA Trifluoroacetic acid
Trp Tryptophane
UGP UDP-Glucose pyrophosphorylase
USDA United States Department of Agriculture
XIC Extracted Ion Chromatogram
ZALS Zwitterionic Acid Labile Surfactant
 SOMMAIRE

INTRODUCTION GENERALE ...................................................................................... 1

CHAPITRE 1 : Synthèse bibliographique ...................................................3
1. LE BLÉ : GÉNERALITÉS ........................................................................................... 5
1.1 Origine des blés cultivés ............................................................................................5
1.2 Utilisation des blés.....................................................................................................6
1.3 La Production ............................................................................................................7
1.4 Conduite culturale et développement de la plante.......................................................8
1.5 Rendement versus quantité de protéines .....................................................................9
2. LE GRAIN : STRUCTURE ET COMPOSITION .................................................... 11
2.1 Phases du développement du grain ...........................................................................11
2.2 Structure du grain de blé .......................................................................................... 12
2.3 Composition du grain de blé .................................................................................... 12
2.4 Les protéines du grain .............................................................................................. 13
2.5 Influence de l’environnement sur le développement et la composition du grain ........ 15
3. L’AZOTE ET LE SOUFRE : DE LA NUTRITION DE LA PLANTE AUX
PROTÉINES DE RÉSERVE DU GRAIN ..................................................................... 16
3.1 Métabolisme de l’azote et du soufre dans la plante ................................................... 16
- Transport et assimilation de l’azote ..........................................................................16
- Transport et assimilation du soufre ..........................................................................17
- Interactions entre l’azote et le soufre ........................................................................ 18
3.2 Remobilisation de l’azote et du soufre vers le grain ................................................. 19
3.3 L’effet de l’azote et du soufre sur les protéines de réserve du grain .......................... 19
3.4 La qualité du grain dans un contexte de diminution des intrants ............................... 20
4. LES PROTÉINES NUCLÉAIRES AU CENTRE DES MÉCANISMES DE
RÉGULATION ............................................................................................................... 22
4.1 La régulation transcriptionnelle de la synthèse des protéines de réserve ................... 22
4.2 Le noyau : un compartiment cellulaire très dynamique ............................................. 23
4.3 L’extraction et l’analyse des protéines nucléaires ..................................................... 24
4.4 L’étude des protéines nucléaires chez les plantes ..................................................... 25
5. L’INTÉGRATION DE DONNÉES -OMIQUES POUR VISUALISER LA
RÉPONSE BIOLOGIQUE ............................................................................................. 27
5.1 Biologie des systèmes .............................................................................................. 27
5.2 Les différents types de réseaux biologiques ............................................................. 28
5.3 RulNet : l’outil utilisé au cours de la thèse ............................................................... 30
 6. CONCLUSIONS DE LA SYNTHÈSE ET STRATÉGIE D’ÉTUDE ....................... 32
RÉFÉRENCES ............................................................................................................... 34

CHAPITRE 2 : Validation de la méthode d’extraction des

protéines nucléaires à partir de grains de blé ...........................................43
1. INTRODUCTION ....................................................................................................... 45
2. ARTICLE 1 : Proteomic approach to identify nuclear proteins in wheat grain ...... 46
3. CONCLUSIONS ......................................................................................................... 56

CHAPITRE 3 : Le protéome nucléaire du grain en développement

de blé tendre ................................................................................................................59
1. INTRODUCTION ....................................................................................................... 61
2. ARTICLE 2 : Changes in the nuclear proteome of developing wheat (Triticum
aestivum L.) grain ............................................................................................................ 62
3. CONCLUSIONS ......................................................................................................... 77

CHAPITRE 4 : La réponse protéomique du grain de Triticum

monococcum à la nutrition azotée et soufrée .............................................79
1. INTRODUCTION ....................................................................................................... 81
2. ARTICLE 3 : Grain subproteome responses to nitrogen and sulfur supply in
diploid wheat Triticum monococcum ssp. monococcum ................................................. 82
3. CONCLUSIONS ....................................................................................................... 110

CHAPITRE 5 : La réponse intégrative du grain de Triticum

monococcum à la nutrition azotée et soufrée ........................................... 113
1. INTRODUCTION ..................................................................................................... 115
2. ARTICLE 4 : Integrative response of the einkorn (Triticum monococcum) grain to
nitrogen and sulphur nutrition ..................................................................................... 116
3. CONCLUSIONS ....................................................................................................... 148

CHAPITRE 6 : Conclusions générales et perspectives de la thèse

........................................................................................................................................... 151
1. CONCLUSIONS ....................................................................................................... 153
2. PERSPECTIVES ...................................................................................................... 156

RÉSUMÉ ....................................................................................................................... 160

ABSTRACT .................................................................................................................. 160
 RÉSUMÉ DES FIGURES

Figure 1. Schéma de l’origine des génomes portés par les différents blés. ............................................5
Figure 2. L’origine géographique des blés cultivés. .............................................................................6
Figure 3. Principales zones et pays de production de blé dans le monde. ..............................................7
Figure 4. Schéma simplifié du développement d’une plante de blé tendre d’hiver et de la mise en place
du rendement en grains .......................................................................................................................8
Figure 5. Relation négative entre la concentration en protéines et le rendement en grains ....................9
Figure 6. Variations des masses de matière fraiche (vert), matière sèche (orange) et d’eau (bleu) au cours
du développement du grain ............................................................................................................... 11
Figure 7. Anatomie du grain de blé tendre ......................................................................................... 12
Figure 8. Composition du grain de blé et accumulation des réserves (amidon et protéines) au cours du
développement .................................................................................................................................. 13
Figure 9. Classification des protéines de la farine du grain de blé : rapprochement entre les classifications
d’Osborne et Shewry......................................................................................................................... 13
Figure 10. Représentation schématique des principaux loci des gènes de protéines de réserve chez le blé.
D’après Shewry and Halford (2003) .................................................................................................. 14
Figure 11. Schéma de l’assimilation de l’azote chez les plantes ......................................................... 16
Figure 12. Schéma des voies d’assimilation du soufre dans la plante ................................................. 17
Figure 13. Principales interactions entre les métabolismes du C, N et S chez les plantes .................... 18
Figure 14. Relations entre la quantité de LMW-GS (a et b) et HMW-GS (c et d) et la quantité d’azote (a
et c) et de soufre (b et d) par grain ..................................................................................................... 20
Figure 15. Corrélations entre la vitesse d’accumulation des LMW-GS et HMW-GS et le niveau
d’expression des gènes correspondants .............................................................................................. 22
Figure 16. Représentation schématique d’un noyau de cellule végétale .............................................. 24
Figure 17. Chronologie des analyses du protéome nucléaire réalisées chez les plantes ....................... 25
Figure 18. Les différents niveaux de données -omiques et des exemples de ressources disponibles à
chaque niveau chez Arabidopsis thaliana .......................................................................................... 27
Figure 19. Représentation schématique de différents types de réseaux biologiques ............................ 29
Figure 20. Schéma situant les objectifs du projet de thèse en lien avec les espèces de blé étudiées ..... 32
Figure 21. Principales étapes de la validation de la méthode d’extraction des protéines nucléaires ..... 56
Figure 22. Analyse du protéome nucléaire du grain de blé tendre en développement .......................... 61
Figure 23. Schéma expérimental de l’étude de la réponse du grain de Triticum monococcum à la nutrition
azotée et soufrée ............................................................................................................................... 81
Figure 24. Représentation schématique de l’objectif du chapitre 5 : intégrer les réponses du grain de T.
monococcum à l’apport de N et S aux différentes échelles mesurées (transcrits, protéines, métabolites)
pour faire le lien entre la nutrition et la composition en protéines de réserve .................................... 116
Figure 25. Effets de la nutrition azotée et soufrée sur le rapport N/S du grain (a) et sur la composition
du grain mature chez T. monococcum (b) ........................................................................................ 154
 Introduction générale

INTRODUCTION GÉNÉRALE
Première source de calories et de protéines dans l’alimentation humaine, le blé est une
plante d’intérêt agronomique indéniable, au centre d’enjeux économiques importants. Pour
répondre à une demande toujours croissante, l’augmentation des rendements en blé est depuis
plusieurs décennies l’objectif majeur des programmes de sélection variétale. Depuis le milieu
du XXème siècle, des progrès conséquents ont été réalisés tant d’un point de vue génétique, que
d’un point de vue agronomique avec des changements importants au niveau de la conduite
culturale, le développement de la mécanisation, l’utilisation importante de fertilisants
notamment azotés et l’emploi de produits phytosanitaires. En France par exemple, ces progrès
ont conduit à l’intensification du mode de production responsable d’une remarquable
augmentation du rendement en blé tendre qui est passé d’environ 20 quintaux/hectare (q/ha) en
1950 à environ 70 q/ha pendant les années 1990. Cependant, l’utilisation excessive d’engrais
azotés, entraînant sa non-assimilation par la plante et son rejet dans l’atmosphère, les rivières
et les océans, est à l’origine de nombreuses perturbations aujourd’hui observées au sein des
écosystèmes avoisinant les cultures. Face à cette problématique, des mesures ont été prises, par
exemple à l’échelle européenne avec la Directive Nitrates qui a pour but de raisonner les apports
d’azote et limiter leur impact sur l’environnement. De telles mesures peuvent expliquer
l’évolution nettement plus lente des rendements observée depuis quelques années, ce malgré
les nombreux progrès génétiques encore réalisés. A cela s’ajoutent la réduction de la diversité
des cultures au sein des rotations, avec notamment une réduction de la part des légumineuses
avant la culture de blé, les effets du changement climatique, ainsi que la réduction de l’emploi
de produits phytosanitaires avec par exemple en France la mise en place du plan Ecophyto. Ces
raisons seraient en grande partie responsables du phénomène de stagnation des rendements.
Ces difficultés d’ordre environnemental s’accompagnent d’une forte relation négative
existant entre le rendement et la teneur en protéines du grain. Or, la teneur en protéines du grain
est un critère important à la base de sa qualité nutritionnelle et technologique. Le maintien d’une
quantité de protéines optimale sans impacter le rendement constitue donc un enjeu essentiel,
notamment pour la France qui est le 5ème producteur mondial. Outre la teneur, la composition
du grain en protéines notamment en protéines de réserve (molécules riches en azote et en soufre)
qui sont responsables des propriétés rhéologiques de la farine, influence fortement la qualité
d’utilisation du grain. On pourrait ainsi espérer compenser une diminution de la teneur en
protéines induite par l’augmentation des rendements et l’évolution des pratiques culturales vers
des méthodes plus respectueuses de l’environnement par une composition en protéines de
réserve adaptée aux différents usages. Autrement dit, il convient de trouver les solutions pour
produire « plus » et « mieux ».
Dans ce contexte, les travaux de l’équipe Biologie Intégrative de la composition du Grain
(BIG) ont pour but, entre autres, d’étudier les mécanismes moléculaires impliqués dans le
déterminisme de la composition du grain de blé en protéines de réserve et d’identifier des gènes
ou régions chromosomiques associés à ces phénomènes de régulation. C’est dans cet axe de
recherche que s’est intégré mon projet de thèse. Il avait pour but d’apporter de nouveaux
éléments à la compréhension des processus de régulation du développement du grain et de
l’accumulation des protéines de réserves en réponse à la nutrition azotée et soufrée. Pour cela
j’ai produit ou participé à l’acquisition de différents types de données –omiques sur des grains
1
 Introduction générale

en cours de développement produits dans différentes conditions de culture. J’ai en particulier

étudié le protéome nucléaire, la synthèse des protéines de réserve étant soumise à une forte
régulation transcriptionnelle. Puis en analysant les différents niveaux de données –omiques
produites, j’ai étudié la réponse intégrative du grain à la nutrition azotée et soufrée. J’ai débuté
ce projet lors de mon stage de Master 2, que j’ai ensuite poursuivi au cours de ces trois années
de thèse.
Ce manuscrit débute par une synthèse bibliographique qui fait le point sur les
connaissances à l’origine de ce projet, dont la stratégie d’étude et les objectifs sont définis en
fin de ce premier chapitre. Les chapitres suivants correspondent aux articles scientifiques
publiés au cours de la thèse, soumis ou en préparation. Pour faciliter la lecture, chaque article
est accompagné d’une introduction et une conclusion en français. Enfin, une conclusion
générale de la thèse ainsi que des éléments de perspectives viennent clore ce manuscrit.

2
 CHAPITRE 1 :
Synthèse bibliographique

3
4
 Chapitre 1 : Synthèse bibliographique

1. LE BLÉ : GÉNERALITÉS
1.1 Origine des blés cultivés
Les blés aujourd’hui cultivés appartiennent au genre Triticum et possèdent différents
niveaux de ploïdie, qui témoignent des événements de polyploïdisation survenus au cours de
leur l’histoire évolutive, avant domestication. On distingue ainsi des espèces diploïdes telles
que l’engrain ou petit épeautre (Triticum monococcum), des espèces tétraploïdes comme
l’amidonnier (T. turgidum) et des espèces hexaploïdes (T. aestivum, Figure 1a). Les blés
tétraploïdes se sont différenciés en plusieurs sous-espèces comme l’amidonnier domestiqué ou
le blé dur (T. turgidum spp dicoccoïdes ou durum, respectivement). Il en est de même pour les
blés hexaploïdes où l’on distingue deux sous-espèces cultivées, le blé tendre (T. aestivum spp
aestivum) et le grand épeautre (T. aestivum spp spelta). Le génome des blés hexaploïdes est
ainsi constitué de 3 génomes différents : A, B et D, chacun constitué de 7 paires de
chromosomes, soit un total de 42 chromosomes (Figure 1b).

Figure 1. Schéma de l’origine des génomes portés par les différents blés. (a) Catégorisation de
différentes espèces de blés en fonction de leur niveau de ploïdie, portant les génomes A, B et/ou D. (b)
Schéma du caryotype du blé tendre (Triticum aestivum spp aestivum) hexaploïde. (c) Schéma de
l’histoire évolutive des blés, les dates estimées des événements étant exprimées en millions d’années,
d’après Marcussen et al. (2014).

La comparaison de ces 3 génomes a permis d’estimer leur date de divergence et de

reconstruire l’histoire phylogénétique du blé (Marcussen et al., 2014). Les génomes diploïdes
(2n=14) A et B, portés respectivement par T. urartu et Aegylops speltoides ont divergé il y
environ 6,5 Ma puis un premier événement d’hybridation a été à l’origine du génome D (2n=14)
il y a 5,5 Ma (Figure 1c). Le génome hexaploïde du blé tendre est alors apparu suite à deux
événements majeurs de polyploïdisation. Le premier correspond à une hybridation il y a environ
0,8 Ma entre T. urartu (AA) et une espèce proche d’Ae. speltoides (BB) qui a permis
l’apparition du blé dur sauvage (T. turgidum) au génome AABB à l’origine du blé dur actuel.
Le second événement a eu lieu il y a environ 0,4 Ma et correspond à un croisement entre T.

5
 Chapitre 1 : Synthèse bibliographique

turgidum (AABB) et le diploïde Ae. tauschii (DD), qui a donné naissance à l’ancêtre hexaploïde
du blé tendre (AABBDD).
Les premières cultures de blé sont apparues il y a environ 10 000 ans en Mésopotamie au
moment de la révolution néolithique (Shewry, 2009). Pour l’Homme, cette période a marqué la
transition entre un mode de vie basé sur la chasse et la cueillette à un mode de vie basé sur
l’agriculture et l’élevage, entraînant ainsi sa sédentarisation. Sur la base d’éléments
archéologiques, génétiques et botaniques, l’origine des premiers blés cultivés que sont l’engrain
diploïde (T. monococcum) et l’amidonnier tétraploïde (T. turgidum spp dicoccum) a été estimée
à une zone limitée du croissant fertile située à l’amont du Tigre et de l’Euphrate et qui
correspond à des territoires de l’actuelle Syrie et de la Turquie (Figure 2, Heun, 1997 ; Lev-
Yadun et al., 2000 ; Dubcovsky and Dvorak, 2007). L’origine géographique du blé tendre
semble quant à elle être localisée au Nord-Ouest de l’Iran et/ou au Nord-Est de la Turquie.

Figure 2. L’origine géographique

des blés cultivés. La distribution
approximative des formes sauvages
d’amidonnier (T. turgidum spp
dicoccoïdes) et d’Aegilops tauschii
est représentée par des points et celle
d’engrain est représentée par la zone
jaune. Les zones cerclées
correspondent aux régions putatives
d’origine des formes cultivées
d’amidonnier, de blé dur, d’engrain
et de blé tendre. Adapté de
Dubcovsky and Dvorak (2007).

La diffusion du blé en Europe a débuté vers environ -8000 ans avant JC à partir du bassin
anatolien (Bonjean, 1994). La culture du blé s’est également étendue à l’Est du croissant fertile
pour atteindre la Chine vers environ -5000, et à l’Afrique par l’Egypte vers -6000. Plus
récemment, le blé fut introduit au XVIe siècle en Amérique par les Espagnols, puis en Australie
par les Anglais au XVIIIe siècle, à partir de pools génétiques européens (Shewry, 2009).

1.2 Utilisation des blés

Le blé tendre (T. aestivum) est l’espèce la plus cultivée et représente à elle seule 95% de
la production mondiale de blé. La domination de cette espèce au sein des cultures s’explique
par son fort potentiel de rendement et sa grande plasticité génomique et phénotypique, qui lui
permettent d’être cultivée dans la plupart des régions agricoles mondiales (Shewry, 2009).
L’utilisation majoritaire du blé tendre provient de son grain, dont la farine panifiable sert à la
confection du pain, aliment qui a été à la base de nombreux régimes. Une étude britannique
montre que si aujourd’hui les produits céréaliers représentent environ 31% et 23% des apports
6
 Chapitre 1 : Synthèse bibliographique

journaliers d’énergie et de protéines, respectivement, le pain à lui seul apporte environ 11% des
apports pour ces deux catégories (Bates et al., 2011; Shewry and Hey, 2015). La paille qui
correspond à la tige coupée pendant la moisson fait également l’objet de diverses utilisations.
Elle peut être enfouie après la récolte dans le but de restituer au sol le carbone et les éléments
minéraux prélevés par la culture, ou si elle est récoltée, être utilisée comme litière pour les
animaux ou servir de fourrage, de combustible, ou d’élément de base pour la fabrication de
fumier utilisé comme fertilisant. La seconde espèce de blé la plus produite est le blé dur (T.
durum), plus adapté au climat sec Méditerranéen et utilisé principalement dans la fabrication
de pâtes alimentaires et de semoules. D’autres espèces de blés sont cultivées dans des zones
géographiques et sur des surfaces plus restreintes. C’est le cas de l’engrain (T. monococcum),
l’amidonnier (T. turgidum spp dicoccoïdes) et le grand épeautre (T. spelta). Les cultures
d’engrain sont limitées à des zones essentiellement à proximité de la région méditerranéenne
(Turquie, pays des Balkans, sud de l’Italie et de la France, Espagne, Maroc). Cette espèce
diploïde est très appréciée pour ses excellentes propriétés nutritionnelles et ses forts taux de
protéines et de caroténoïdes (Corbellini et al., 1999 ; Hidalgo et al., 2006 ; Hidalgo and
Brandolini, 2014).

1.3 La Production
Aujourd’hui le blé est la seconde céréale la plus produite dans le monde et représente
33% de la production céréalière totale, derrière le maïs (46%) et devant le riz blanchi (21%,
USDA, 2016). Au cours de la campagne 2014/2015, une production record a été réalisée avec
730 millions de tonnes (Mt) produites (FAO, 2016), entraînant une augmentation du stock
mondial estimé aujourd’hui à 200 Mt. La France représente environ 6,2% de la production
mondiale de blé et se situe à la première place des producteurs au niveau européen et à la 5ème
place au niveau mondial, les premiers étant la Chine, l’Inde, les Etats-Unis et la Russie (Figure
3).

Figure 3. Principales zones et pays de production de blé dans le monde. D’après (Monfreda et al.,
2008). La part des 8 principaux pays exportateurs dans la production mondiale a été indiquée. Source :
FAOSTAT ([Link]

7
 Chapitre 1 : Synthèse bibliographique

En France, les surfaces arables allouées à la culture de blé sont de 5,3 Mha, soit 2,3% des
surfaces mondiales consacrées à cette culture. Ces chiffres rendent compte des deux types
d’agriculture pratiqués dans le monde. En France comme en Europe, l’agriculture privilégiée
est de type intensive, synonyme de fort rendement (avoisinant les 60 q/ha) en moyenne en
Europe et 75 q/ha en France) sur des surfaces cultivées réduites, qui diffère de l’agriculture
extensive pratiquée par exemple aux Etats Unis où les surfaces allouées sont beaucoup plus
importantes mais où les rendements sont deux fois moins élevés, avec environ 29 q/ha
(FranceAgrimer, 2015).

1.4 Conduite culturale et développement de la plante

Pour atteindre un haut niveau de production, différents traitements et apports à la culture
doivent être réalisés pour subvenir aux besoins de la plante tout au long de son cycle de
développement et ainsi obtenir le rendement et la qualité du grain espérés. On distingue
couramment deux grandes phases dans le développement d’une plante de blé : la phase
végétative au cours de laquelle il y a production de tiges, feuilles et racines et la phase
reproductive qui correspond à l’apparition d’un épi portant des fleurs puis à la formation des
grains après fécondation. La transition entre ces deux phases diffère entre les blés dits
« d’hiver » qui sont semés à l’automne et qui nécessitent une période de froid, appelée
vernalisation, pour initier la formation de leur appareil reproducteur, et les blés dits « de
printemps » qui ne nécessitent pas cette période de vernalisation (Acevedo et al., 2002).
Après germination et levée de la graine, l’étape de tallage consiste en l’apparition de tiges
secondaires appelés talles (Figure 4). Pendant la phase de montaison, il y a une croissance rapide
des entre-nœuds. L’épi monte progressivement dans la gaine, provoquant son gonflement, puis
c’est l’épiaison généralement au mois de juin. Une fois la formation des organes floraux
achevée, l’étape de floraison marque le début du développement du grain, qui est ensuite récolté
au mois d’août lorsqu’il a atteint la maturité.

Figure 4. Schéma simplifié du développement d’une plante de blé tendre d’hiver et de la mise en
place du rendement en grains (Arvalis-Institut du végétal, 2016).

8
 Chapitre 1 : Synthèse bibliographique

Les étapes clés de tallage et de montaison sont très sensibles à l’environnement et, dans
le cas d’une agriculture conventionnelle, s’accompagnent de plusieurs traitements
phytosanitaires pour lutter contre les adventices, maladies fongiques et autres ravageurs
nuisibles à la culture (Figure 4). Afin de subvenir aux besoins azotés de la plante tout au long
de son cycle, trois apports sont généralement pratiqués : au stade de tallage, au stade épi 1 cm
et enfin, entre le stade 2 nœuds et le stade de gonflement. Un apport tardif après épiaison peut
également être pratiqué et favorise une teneur en protéines des grains plus élevée. Le
fractionnement des apports est jugé le plus efficace pour obtenir de hauts rendements. En effet,
le rendement se met en place tout au long du cycle de production, à travers différentes
composantes (Figure 4). A l’issue du tallage, le nombre de pied par m2 est défini et constitue la
première composante. Les autres composantes sont le nombre d’épis par plante, déterminé
pendant l’épiaison, le nombre de grains par épi qui dépend du nombre d’épillet par épi, du
nombre de fleurs par épillet et du pourcentage de fertilité et le poids de mille grains (PMG).

1.5 Rendement versus quantité de protéines

Chez le blé, les programmes d’amélioration variétale ciblent principalement un fort
potentiel de rendement. Malgré ces efforts, on assiste depuis la fin des années 1990 à leur
stagnation qui ne peut être expliquée par une absence des progrès génétiques (Graybosch and
Peterson, 2010; Brisson et al., 2010). En fait, cette stagnation du rendement serait due au
changement climatique d’une part et à la diminution de l’application de fertilisants azotés
d’autre part (Brisson et al., 2010). Face à cette problématique, associée à une diminution des
terres arables et une croissance constante de la population mondiale, l’augmentation de la
production de blé représente un défi de premier plan. Une seconde difficulté demeure dans la
forte relation négative existant entre le rendement et la teneur en protéines du grain (Simmonds,
1995 ; Feil, 1997 ; Brancourt-Hulmel et al., 2003 ; Bogard et al., 2010 ; Figure 5).

Figure 5. Relation négative entre la concentration

en protéines et le rendement en grains. Les
données sont les moyennes pour 27 génotypes
cultivés dans 27 environnements différents. La
droite noire représente la régression linéaire du taux
de protéines en fonction des valeurs de rendement.
Les lignes en pointillés bleus représentent
l’intervalle de confiance à 95% de la régression. Les
lignes en pointillés noirs représentent les courbes
iso-protéines-rendement en g/m2. D’après Bogard et
al. (2010).

Cette relation négative est soumise à de forts effets génétiques et environnementaux ainsi
qu’à des interactions entre génotypes et environnements. Certains génotypes présentent un écart

9
 Chapitre 1 : Synthèse bibliographique

à la relation teneur en protéines-rendement en grains (ou GPD, Grain Protein Deviation ; Figure
5). Autrement dit, le GPD permet d’identifier les génotypes ayant une concentration en
protéines supérieure à celle prédite par cette relation (Monaghan et al., 2001). Ceux associant
un fort potentiel de rendement et une forte concentration en protéines représentent ainsi des blés
très compétitifs sur le marché de l’exportation. Ainsi pour favoriser leur inscription au catalogue
des blés, de telles variétés se voient attribuer un bonus par les instances décisionnelles.
Pour l’export, un minimum de 11,5% de protéines par grain est exigé pour entre autres,
répondre au mieux aux exigences des industriels de la transformation, le blé étant en effet
consommé principalement après transformations. Néanmoins, la quantité de protéines n’est pas
le seul critère d’évaluation des grains de blé. En effet, la première transformation que subit le
blé est la mouture. Or la valeur meunière repose largement sur la dureté du grain qui détermine
l’énergie pour réduire le grain en farine, la granulométrie des farines obtenues ou encore le
pourcentage d’amidon endommagé. La qualité technologique du grain est donc appréciée par
plusieurs critères (mesure de la dureté, composition, ou humidité). Elle permet de départager
les blés dans la forte diversité de l’offre mondiale et représente donc un critère important à
prendre en compte dans les programmes d’amélioration variétale.

10
 Chapitre 1 : Synthèse bibliographique

2. LE GRAIN : STRUCTURE ET COMPOSITION

2.1 Phases du développement du grain
Le développement du grain de blé débute à l’issue de la double fécondation (soit
l’anthèse) de l’ovule. Un des deux gamètes mâles fusionne avec les deux noyaux polaires du
sac embryonnaire pour former l’albumen (3n) et le second gamète mâle féconde l’oosphère
pour former l’embryon (2n). Les stades de développement peuvent être estimés à partir du
calcul du nombre de jours après anthèse, ou en calculant le temps thermique exprimé en degré-
jours (°Cj) après anthèse, qui correspond à la somme des températures moyennes journalières
accumulées par le grain. Ce dernier mode de calcul, qui tient compte des variations de
température pouvant survenir au cours du développement du grain a été choisi dans cette thèse
pour estimer de manière précise les différents stades. Le développement du grain est
classiquement décomposé en trois phases principales : la phase de cellularisation, de
remplissage et de maturation (Figure 6, Sabelli and Larkins, 2009).

Figure 6. Variations des masses

de matière fraiche (vert), matière
sèche (orange) et d’eau (bleu) au
cours du développement du
grain. Le développement du grain
est divisé en trois phases :
cellularisation, remplissage et
maturation.

Après la double fécondation, l’albumen subit des divisions successives sans cytokinèse
qui aboutissent à environ 70°Cj à la formation d’un coenocyte à plus de 2000 noyaux répartis
à la périphérie de celui-ci, autour d’une large vacuole centrale (Mares et al., 1975). S’ensuit
une étape de cellularisation et de divisions cellulaires jusqu’à environ 220-250°Cj (Chojecki et
al., 1986), caractérisée par une forte accumulation d’eau (Figure 6). Le grain acquiert alors sa
taille finale. Pendant la phase de remplissage qui suit, le grain accumule rapidement ses
composés de réserve, essentiellement de l’amidon et des protéines. La vitesse d’accumulation
des réserves diminue vers 550°Cj pour devenir nulle généralement vers 650-700°Cj, lorsque la
concentration en eau du grain est proche de 45g pour 100g de matière fraîche (Schnyder and
Baum, 1992). La dernière phase appelée maturation ou dessiccation correspond à une perte
d’eau rapide, entraînant une réduction des activités métaboliques (Young and Gallie, 2000). Le
grain contient alors une quantité de masse sèche maximale et atteint l’état dit de maturité
physiologique.

11
 Chapitre 1 : Synthèse bibliographique

2.2 Structure du grain de blé

Le grain de blé est un fruit sec indéhiscent appelé caryopse, constitué de trois parties :
l’embryon, l’albumen et les couches périphériques qui représentent environ 3%, 83% et 14%
de la masse sèche du grain, respectivement (Figure 7, Pomeranz, 1988; Barron et al., 2007).

Figure 7. Anatomie du grain de blé tendre. Le grain de blé est constitué de trois parties : l’embryon,
l’albumen et les couches périphériques. D’après Surget and Barron (2005).

L’embryon est constitué de l’axe embryonnaire qui donnera la tigelle, le mésocotyle et la

radicule de la future plantule ainsi que du scutellum ou cotylédon qui constitue une zone
d’échange entre l’embryon et l’albumen (Evers and Millar, 2002). L’albumen correspond au
tissu de réserve et est constitué de l’albumen amylacé et de la couche à aleurone. La couche à
aleurone est composée généralement d’une couche de cellules et assure un rôle nourricier et
protecteur. Après imbibition du grain, les gènes d’enzymes protéolytiques et hydrolytiques sont
activés dans ces cellules pour digérer les parois cellulaires de l’albumen et permettre la
mobilisation des réserves pour la croissance de l’embryon (Sabelli and Larkins, 2009). Les
couches périphériques sont constituées de l’intérieur vers l’extérieur par la bande hyaline, la
testa, les cellules tubulaires, les cellules croisées et le péricarpe externe. Les cellules tubulaires
et cellules croisées forment le péricarpe interne. Pendant le développement du grain, le
péricarpe évite les pertes d’eau sans pour autant empêcher sa pénétration (Evers et al., 1999).

2.3 Composition du grain de blé

A maturité, le grain de blé est essentiellement constitué d’amidon, qui représente 65-70%
de la masse sèche totale (Figure 8a). Cet amidon est localisé au sein de l’albumen amylacé, où
il s’accumule dans les amyloplastes. Il constitue la principale ressource énergétique de la
plantule pendant le processus de germination. Les protéines représentent 10 à 13% de la matière
sèche et se retrouvent quant à elles dans tous les tissus, avec une concentration plus importante
dans l’embryon et la couche à aleurone. Amidon et protéines s’accumulent pendant la phase de
remplissage, selon un profil similaire dans le temps (Figure 8b, Laudencia-Chingcuanco et al.,
2007; Shewry et al., 2012). Les pentosanes (polysaccharides non amylacés) sont les principaux
constituants des parois cellulaires de l’albumen. Ces fibres représentent environ 2 à 3% de la

12
 Chapitre 1 : Synthèse bibliographique

masse sèche. Enfin, les lipides, la cellulose, les minéraux et les vitamines sont des constituants
mineurs du grain.

Figure 8. Composition du grain de blé et accumulation des réserves (amidon et protéines) au cours
du développement. (a) Représentation de la part des différents composants du grain mature, exprimés
en pourcentage de la masse sèche du grain. (b) Cinétique d’accumulation de l’amidon (noir) et des
protéines (rouge). D’après Laudencia-Chingcuanco et al. (2007).

2.4 Les protéines du grain

L’étude des protéines du grain de blé remonte à 250 ans, lorsque des protéines du gluten
ont été isolées pour la première fois en 1745. En 1924, Osborne a proposé une classification
des protéines en fonction de leur solubilité : les albumines (solubles dans l’eau), les globulines
(solubles dans les solutions salines), les gliadines (solubles dans de l’alcool 70%) et les
gluténines (solubles dans une base ou un acide ou des détergents en présence d’un réducteur ;
Osborne, 1907; Osborne, 1909). Cette classification a été revue par Shewry et ses collaborateurs
(1986) qui ont proposé deux groupes principaux : les protéines métaboliques et les protéines de
réserve (PR ; Figure 9).

Figure 9. Classification des protéines de la farine du grain de blé : rapprochement entre les
classifications d’Osborne et Shewry. La part de chaque classe de protéine est exprimée en
pourcentage par rapport aux protéines totales. D’après Feillet (2000).

13
 Chapitre 1 : Synthèse bibliographique

Les protéines métaboliques correspondent aux albumines et globulines et représentent 15

à 20% des protéines totales. Elles sont constituées d’un grand nombre de protéines ayant des
fonctions diverses, telles que des protéases, hydrolases, oxydo-réductases ou des inhibiteurs
d’enzymes (Feillet, 2000). Elles participent à la formation du grain et à l’accumulation des
réserves dans l’albumen (Vensel et al., 2005).
Les PR, aussi connues sous le nom de prolamines de par leur richesse en proline et
glutamine, représentent 80 à 85% des protéines totales de la farine. Elles sont divisées en deux
groupes, les gliadines et les gluténines et sont les constituants du gluten, à l’origine des
propriétés rhéologiques de la farine. Les gluténines forment des polymères et sont responsables
des propriétés viscoélastiques du gluten alors que les gliadines sont des protéines
monomériques qui confèrent l’extensibilité de la pâte (Branlard et al., 2001). Les gluténines
sont divisées en deux fractions, à savoir les gluténines de faible poids moléculaire (LMW-GS)
et de haut poids moléculaire (HMW-GS). Les gliadines sont quant à elles divisées en différentes
classes : α/β-, γ- et ω-gliadines.
Les PR sont produites par des loci relativement bien connus, exprimés spécifiquement
dans l’albumen (Figure 10). Les gènes codant les HMW-GS sont situés aux loci homéologues
Glu1 sur le bras long des chromosomes du groupe 1 (Payne et al., 1980). Chaque locus est
constitué de deux gènes étroitement liés. La synthèse de D’Ovidio and Masci (2004) rappelle
que les LMW-GS sont codées par les trois loci homéologues Glu3 sur le bras court des
chromosomes du groupe 1. Des loci supplémentaires ont cependant été détectés. On estime
ainsi que le génome du blé compte 30 à 40 gènes codant ces PR. Les gliadines sont produites
par des loci complexes situés sur les bras courts des chromosomes des groupes 1 et 6 (Payne et
al., 1984). Les α/β- et quelques γ-gliadines sont codés par les loci Gli-2 des chromosomes 6.
Les ω-gliadines et la majorité des γ-gliadines sont produites par des loci des chromosomes 1.
Plus précisément, trois loci Gli-1 codent les γ- et ω-gliadines. Ces dernières sont aussi produites
par Gli-A3 et Gli-B3. A côté de ces principaux loci, d’autres plus rares ont été mis en évidence
comme Gli-A5 et Gli-A6 (Metakovsky et al., 1996). Comme les LMW-GS, les gliadines sont
codées par des loci formés de clusters de gènes avec des séquences proches ou identiques. On
estime qu’il y a par génome, en incluant les pseudo-gènes, 75 à 450 gènes codant les α/β-
gliadines, 45 à 120 gènes codant les γ-gliadines et enfin 45 à 60 codant les ω-gliadines. Les
gènes de ces clusters sont transmis en blocs alléliques.

Figure 10. Représentation

schématique des principaux
loci des gènes de protéines
de réserve chez le blé.
D’après Shewry and
Halford (2003).

14
 Chapitre 1 : Synthèse bibliographique

2.5 Influence de l’environnement sur le développement et la composition du grain

L’environnement de la plante de blé influence la durée du développement du grain, sa
teneur et sa composition en protéines, mais aussi la polymérisation des protéines et le rendement
(Nuttall et al., 2016). La température, l’accès à l’eau et la fertilisation sont les facteurs
environnementaux qui impactent le plus ces paramètres (Malik, 2009). Les effets de fortes
températures, particulièrement pendant le remplissage du grain, sont assez bien caractérisés
(Branlard et al., 2015 ; Asseng et al., 2015). Ce type de contrainte entraîne une augmentation
des enzymes de la glycolyse et une diminution de l’ADP glucose pyrophosphorylase, l’enzyme
qui catalyse la première étape de la voie de biosynthèse de l’amidon. En conséquence, la
réduction d’accumulation d’amidon est responsable d’une diminution du rendement. Par
ailleurs, de nombreuses protéines chaperonnes essentiellement des HSP, ainsi que des protéines
de défense sont fortement produites en réponse à de fortes températures (Skylas et al., 2002;
Majoul-Haddad et al., 2013).
Les PR sont quant à elles moins affectées par le stress thermique que les protéines
métaboliques (Majoul-Haddad et al., 2013). Toutefois en condition de fortes températures, la
proportion de gliadines par rapport aux gluténines est augmentée (Dupont and Altenbach,
2003). Il a également été montré que l’expression des gènes de PR est initiée plus tôt, mais la
durée d’accumulation des protéines correspondantes est raccourcie (Altenbach et al., 2002),
comme pour l’accumulation de l’amidon. Ces observations sont en réalité dues à un
développement du grain accéléré en condition de fortes températures. Des résultats similaires
sont observés dans le cas d’un stress hydrique, qui entraîne une taille et un poids de grain
réduits. Les effets sur le grain sont accentués en cas de combinaison de ces deux types de stress,
hydrique et thermique (Nicolas et al., 1984).
La nutrition de la plante joue également un rôle très important dans le développement du
grain et la mise en place de ses réserves. Outre leur catégorisation basée sur les classifications
d’Osborne et de Shewry, les PR sont également classées selon leur composition en acides
aminés soufrés, cystéine et méthionine. Ainsi on distingue les PR riches en soufre (α/β- et γ-
gliadines, LMW-GS) de celles pauvres en soufre (ω-gliadines et HMW-GS, Shewry et al.,
1997; Shewry et al., 2001). La synthèse de ces protéines est ainsi fortement influencée à la fois
par la disponibilité en azote (N) et en soufre (S) dans le sol.

15
 Chapitre 1 : Synthèse bibliographique

3. L’AZOTE ET LE SOUFRE : DE LA NUTRITION DE LA PLANTE AUX

PROTÉINES DE RÉSERVE DU GRAIN
3.1 Métabolisme de l’azote et du soufre dans la plante
- Transport et assimilation de l’azote
La nutrition azotée des plantes correspond majoritairement à une absorption racinaire d’N
minéral sous forme de nitrate (NO3-) et d’ammonium (NH4+). Dans certains cas de symbioses,
la plante, peut également utiliser l’N gazeux (N2). Parmi ces symbioses, la plus connue est celle
observée chez les Fabacées mettant en jeu des bactéries du genre Rhizobium. Chez le blé, il a
été observé des associations avec des bactéries fixatrices d’N atmosphérique du genre Frankia.
Les plantes peuvent encore utiliser l’N organique présent dans le sol (acides aminés, courtes
chaines peptidiques).
La plante de blé absorbe préférentiellement l’N sous forme de nitrate (Hirel et al., 2007).
Cette absorption met en jeu deux types de transporteurs : des transporteurs de haute affinité
(Hig-affinity transport system, HATS) et les transporteurs de faible affinité (Low-affinity
transport system, LATS). Les LATS appartiennent à la famille des NRT1, qui compte 53
membres connus chez Arabidopsis thaliana (Masclaux-Daubresse et al., 2010) et 16 gènes ont
récemment été identifiés chez le blé (Buchner and Hawkesford, 2014). Le premier gène
caractérisé au sein de cette famille est le gène NRT1.1, qui est à la fois transporteur et senseur
de nitrates. Sa mutation inhibe l’élongation des racines de la plante dans les zones de sol riches
en nitrates, entraînant une réduction considérable de l’absorption d’N (Remans et al., 2006).
Les HATS sont quant à eux des transporteurs de la famille des NRT2 qui compte sept membres
chez A. thaliana, dont l’expression est inhibée en présence d’une concentration élevée en
nitrates dans le sol (Girin et al., 2007). Ce système de transporteurs de nitrates de haute ou
faible affinité permet à la plante de modifier rapidement sa capacité d’absorption des nitrates
en fonction de ses besoins et de la disponibilité en N dans la solution du sol.
Les nitrates absorbés sont ensuite soit stockés dans la vacuole, soit assimilés dans les
racines ou les feuilles après transport dans le xylème. L’assimilation de l’N consiste en deux
étapes successives de réduction des nitrates, conduisant à la formation de nitrites (NO 2-) grâce
à l’action de la nitrate réductase dans le cytosol, puis d’ammonium sous l’action de la nitrite
réductase dans le chloroplaste (Figure 11).

Figure 11. Schéma de l’assimilation de

l’azote chez les plantes. NR : Nitrate
Réductase ; AS : Asparagine Synthétase ; NiR :
Nitrite Réductase ; GS : Glutamine Synthase ;
GOGAT : Glutamate Synthase ; GDH :
Glutamate Déshydrogénase ; CPSase :
Carbamoylphosphate Synthétase. D’après
Miura (2013).

16
 Chapitre 1 : Synthèse bibliographique

L’ammonium est ensuite rapidement incorporé à des acides organiques pour former des
acides aminés. Pour cela, la principale voie est le cycle GS/GOGAT, qui correspond à la
transformation de l’ammonium en glutamine sous l’action de la glutamine synthétase (GS) puis
la formation de glutamate par la glutamate synthase (GOGAT, Miflin and Habash (2002)).
Glutamine, glutamate et asparagine sont les principales formes de transport de l’N dans la
plante.

- Transport et assimilation du soufre

Le S existe dans le sol sous la forme minérale à l’état de sulfates (SO42-) et de dihydrogène
de S (H2S), et sous la forme organique, qui peut représenter jusqu’à 99% du S total disponible.
Le sulfate du sol est la forme majoritaire de S absorbée par la plante, bien qu’elle soit également
capable d’absorber au niveau foliaire le dioxyde de S (SO2) provenant des rejets industriels. La
quantité de sulfate disponible dépend du reliquat non lessivé en sortie d’hiver, de la
minéralisation de la matière organique en début de printemps et de la retombée atmosphérique
de dioxyde de S. Le sulfate, absorbé au niveau racinaire par la plante, est transporté jusqu’aux
chloroplastes des feuilles où il est assimilé. La famille des transporteurs de sulfate comprend
14 membres chez A. thaliana et 10 ont pu être caractérisés chez le blé (Hawkesford, 2003 ;
Buchner et al., 2010). Comme l’N, le S peut également être stocké transitoirement dans la
vacuole des cellules racinaires et foliaires sous forme de sulfate pour une utilisation ultérieure
(Kataoka, 2004).
Pour être assimilé, le sulfate est dans un premier temps activé par l’ATP sulfurylase pour
former de l’adénosine 5’-phosphosulfate (APS ; Hirase and Molin, 2003 ; Figure 12). Sous
l’action de l’APS réductase, l’APS est convertie en sulfite (SO32-) puis en ion sulfure (S2-). La
biosynthèse de cystéine est ensuite divisée en deux étapes : la sérine acétyltransférase catalyse
l’acétylation de la sérine, puis la cystéine synthase (CS) utilise l’O-acétylsérine ainsi formé et
l’ion sulfure S2- pour synthétiser la cystéine. La CS joue donc un rôle majeur puisqu’elle est
responsable de la formation de la cystéine, qui tient une place centrale dans la voie
d’assimilation du S (Figure 12). En effet, la méthionine et le glutathion (GSH) sont tous deux
formés à partir de cet acide aminé.

Figure 12. Schéma des voies

d’assimilation du soufre dans la plante.
APS : Adenosine 5’-Phosphosulfate ; GSH :
Glutathion ; OAS : O-acétylsérine ; O-PHS :
O-phosphohomosérine. D’après Hirase and
Molin (2003).

17
 Chapitre 1 : Synthèse bibliographique

- Interactions entre l’azote et le soufre

Du fait du rôle central de l’N et du S dans la synthèse des acides aminés et donc des
protéines, il existe de fortes interactions entre les assimilations du sulfate et du nitrate dans la
plante, mises en évidence depuis de nombreuses années (Figure 13, Reuveny et al., 1980 ; Zhao
et al., 1993 ; Kopriva and Rennenberg, 2004 ; Jamal et al., 2010). Une insuffisance de S dans
le sol limite l’efficacité de l’apport d’N à la culture (Fazili et al., 2008). Il a également été
démontré chez A. thaliana qu’une déficience en N affecte le niveau d’expression des gènes
impliqués dans l’assimilation du sulfate, suggérant une régulation de cette voie par l’N au
niveau transcriptionnel (Yamaguchi et al., 1999 ; Koprivova et al., 2000). Du fait de ces
interactions, l’ajout d’N et de S de manière concomitante est donc nécessaire pour maximiser
l’assimilation et l’accumulation de N et de S dans la plante et dans le grain.
L’N et le S sont également tous deux impliqués dans des mécanismes de réponse aux
stress biotiques et abiotiques (Giordano and Raven, 2014). Ces réponses impliquent par
exemple des composés comme le glutathion, dont l’abondance dépend à la fois d’N et de S, et
qui joue un rôle central dans la défense de la cellule contre les stress oxydatifs (Noctor et al.,
2002). Des interactions existent également avec le métabolisme du carbone (C). En effet,
l’absorption des nitrates et du sulfate au niveau racinaire est un processus actif qui nécessite des
apports énergétiques issus du métabolisme du carbone (Figure 13). La disponibilité en carbone
est donc un facteur de régulation des transporteurs d’N et de S.

Figure 13. Principales interactions entre les métabolismes du C, N et S chez les plantes. Les flèches
pleines indiquent les voies connues, les flèches hachurées indiquent celles qui ne sont pas encore
totalement caractérisées et les flèches en pointillés représentent la diffusion du CO 2, O2 et NH3. D’après
Giordano and Raven (2014).

18
 Chapitre 1 : Synthèse bibliographique

3.2 Remobilisation de l’azote et du soufre vers le grain

La majeure partie de l’N du grain de blé est issue de la remobilisation de l’N assimilé plus
tôt dans les parties végétatives (Austin et al., 1977). Il a été estimé que cette source représente
en moyenne 70% de l’N total du grain mature, ce chiffre pouvant varier de 60 à 95% suivant le
génotype et les conditions de culture (Palta and Fillery, 1995 ; Kichey et al., 2007). La
remobilisation d’N des organes sources vers les organes puits est un mécanisme qui intervient
pendant le stade végétatif, des feuilles en sénescence vers les feuilles en développement, ou
pendant le stade reproductif, des feuilles sénescentes vers les grains (Malagoli et al., 2005). Il
a été montré que ce deuxième type de remobilisation est principalement régulé par la
disponibilité en N dans les organes sources (Martre et al., 2003). Les feuilles constituent une
source importante car elles sont riches en protéines, notamment la Rubisco qui est présente en
abondance dans les plastes. La remobilisation correspond à une mort cellulaire programmée au
cours de laquelle ces protéines sont dégradées. Les enzymes intervenant dans le métabolisme
des acides aminés permettent alors de convertir les acides aminés issus de cette dégradation des
protéines en glutamine et asparagine, qui représentent les deux principales formes azotées
transportées vers le grain. La seconde source d’N pour le grain correspond à une absorption et
une assimilation directement pendant le remplissage du grain, d’où l’intérêt d’un apport tardif
en N. Après floraison de l’épi, il a été montré que la taille et la quantité d’N du grain peuvent
être significativement réduits en cas d’une condition de déficience en N (Dupont and Altenbach,
2003).
Dans le grain mature, le S est présent majoritairement sous forme de glutathion, de
cystéine et de méthionine. Contrairement à l’N qui est principalement issu de la remobilisation
des feuilles vers le grain, le S total du grain de blé est issu à moins de 50% de la remobilisation,
les formes principales de S remobilisées étant le glutathion et ses dérivés ainsi que le sulfate
(Monaghan et al., 1999). Ce résultat démontre l’importance de l’apport de S aussi bien avant la
floraison que pendant le développement du grain.

3.3 L’effet de l’azote et du soufre sur les protéines de réserve du grain

Comme l’N et le S sont tous deux nécessaires à la formation des acides aminés et donc à
la synthèse des protéines, le rapport N/S influence la capacité de la plante à synthétiser des
protéines plus ou moins riches en S. De nombreuses études ont en effet montré l’impact de la
nutrition azotée et soufrée sur la quantité et la composition en protéines du grain de blé. Il est
connu qu’un apport d’N à la culture accroît la quantité de PR du grain en augmentant leur durée
et leur vitesse d’accumulation (Shewry et al., 2001 ; Triboï et al., 2003 ; Chope et al., 2014).
En condition de carence en S, il a été montré une concentration plus faible par grain de cystéine
et de méthionine (Wrigley et al., 1980), entraînant une concentration plus faible en PR riches
en S, compensée par une concentration plus importante des PR pauvres en S, notamment les
HMW-GS (Zhao et al., 1999 ; Wieser et al., 2004). Un apport excédentaire d’N augmente le
rapport N/S et entraîne des résultats similaires à une carence en S (Wieser and Seilmeier, 1998
; Zörb et al., 2010). Il a récemment été démontré avec l’apport de différents traitements N et S
à une culture de blé que la quantité des PR pauvres en S dépend directement de la disponibilité

19
 Chapitre 1 : Synthèse bibliographique

en N du sol, quel que soit le traitement (i.e. la quantité d’N et S apportée), alors que la quantité
en PR riches en S dépend de la disponibilité en S du sol (Figure 14 ; Dai et al., 2015).
Le ratio N/S est donc très important pour la mise en place de la qualité du grain et est
souvent utilisé pour diagnostiquer le statut soufré dans les cultures (Reussi et al., 2011). Dans
le grain de blé, ce ratio varie généralement entre 12 et 25 en fonction des apports de N et S
(Randall et al., 1981) et il a été admis qu’un rapport N/S ≥ 17 est synonyme d’une carence en
S (Zhao et al., 1999).

Figure 14. Relations entre la quantité

de LMW-GS (a et b) et HMW-GS (c
et d) et la quantité d’azote (a et c) et
de soufre (b et d) par grain. Les points
représentent les moyennes ± erreur
standard pour n = 3 répétitions
biologiques, pour quatre conditions de
nutrition : NS (3 mM N et 2 mM S),
NS– (3 mM N et 0.02 mM S), N+S– (15
mM N et 0.02 mM S) et N+S–/N+S (en
milieu de remplissage du grain, passage
d’une condition de carence en S (N+S–
) par une condition de nutrition avec 15
mM N et 2 mM S). D’après Dai et al.
(2015).

3.4 La qualité du grain dans un contexte de diminution des intrants

Afin d’augmenter les rendements et la production agricole mondiale, l’emploi d’N a
considérablement augmenté depuis le milieu du XXème siècle (Cassman et al., 2003 ; Hirel et
al., 2011). La capacité de la plante à capturer l’N du sol dépend du type de sol, de
l’environnement et de l’espèce, mais il a été estimé que 50 à 70% de l’N fourni au sol n’est pas
fixé par la plante (Hodge et al., 2000). Une partie de cet N est fixée par les microorganismes
du sol, mais le surplus peut être lessivé, ce qui entraîne une eutrophisation des eaux présentes
dans les écosystèmes à proximité (Giles, 2005 ; Michael Beman et al., 2005). Ceci a pour
conséquence un déséquilibre important dans le fonctionnement de ces écosystèmes (Matson et
al., 2002 ; Hirel et al., 2007). Face à cette problématique, des mesures ont été prises afin de
raisonner les apports d’N à la culture et ainsi limiter leur impact sur l’environnement. Dans ce
sens a été instaurée la directive européenne « Directive Nitrates » en 1991. Cette directive s’est
traduite par la définition de « zones vulnérables » où sont imposées des pratiques agricoles
particulières, avec par exemple l’implantation de bandes enherbées en bord de cours d’eau afin
de limiter les risques de ruissellement d’N, ou encore avec la définition de périodes
d’interdiction d’épandage. En plus de ce problème environnemental, l’apport d’N à la culture
et la fertilisation de manière plus générale représentent un coût élevé. Ainsi, pour la production

20
 Chapitre 1 : Synthèse bibliographique

de blé tendre, le poste fertilisation représente entre 40 et 50% des charges opérationnelles (Hirel,
2013).
La disponibilité en S du sol est également au centre de préoccupations pour la production
agricole. Depuis plusieurs années, des progrès considérables ont été réalisés afin de réduire les
émissions industrielles de S, conduisant à une diminution du dépôt du S atmosphérique dans
les sols (Eriksen, 2009). En parallèle, l’augmentation des rendements s’accompagne d’un
appauvrissement des sols en nutriments, notamment en S (Scherer, 2001). Ainsi, une carence
des sols en S est aujourd’hui observée. L’emploi de fertilisants qui contiennent de moins en
moins de S accroît un peu plus l’écart entre la disponibilité en N et la disponibilité en S et donc
la carence en S, qui impacte à la fois le rendement et la composition du grain (Hawkesford,
2000 ; Hawkesford and DeKok, 2006).
Dans ce contexte de modification des conditions de nutrition de la plante de blé, les
paramètres de quantité de protéine et de composition du grain doivent être contrôlés afin de
maintenir le taux de protéines et la qualité du grain souhaités.

21
 Chapitre 1 : Synthèse bibliographique

4. LES PROTÉINES NUCLÉAIRES AU CENTRE DES MÉCANISMES DE

RÉGULATION
4.1 La régulation transcriptionnelle de la synthèse des protéines de réserve
La régulation des PR du grain par la disponibilité en N et S semble s’effectuer au niveau
du processus de transcription. La régulation transcriptionnelle de la synthèse des PR est en effet
connue, aussi bien chez les espèces cultivées du groupe des Angiospermes dicotylédones
comme le pois (Gatehouse et al., 1982; Evans et al., 1984) que chez les céréales comme l’orge
(Rahman et al., 1983; Giese and Hopp, 1984) ou le blé (Dai et al., 2015), c’est-à-dire que la
quantité de protéine dépend directement du niveau de transcrit. Chez le blé, il a été montré que
la vitesse d’accumulation des différentes classes de gliadines et sous-unités de gluténines était
fortement corrélée au niveau d’expression des gènes correspondants, et ce quelles que soient
les quantités d’N et de S appliquées à la culture (Figure 15 ; Dai et al., 2015). Ce résultat
démontre l’effet important de cette régulation transcriptionnelle.

Figure 15. Corrélations entre la vitesse d’accumulation des LMW-GS et HMW-GS et le niveau
d’expression des gènes correspondants. Les données correspondent aux moyennes ± erreur standard
pour n = 2 répétitions biologiques pour l’expression des gènes et n = 3 répétitions pour les vitesses
d’accumulation des PR, pour trois conditions de nutrition : NS (3 mM N et 2 mM S), NS– (3 mM N et
0.02 mM S) et N+S– (15 mM N et 0.02 mM S). D’après Dai et al. (2015).

Cette régulation met en jeu plusieurs facteurs de transcription (FT) qui se fixent au niveau
de courtes séquences situées dans les régions promotrices des gènes de PR aussi appelées cis-
motifs. Un des cis-motifs les plus caractérisés est l’endosperm box, un motif bi-partite constitué
de l’élément GCN4-Like Motif (GLM) et de la Prolamin Box (PB). Le premier FT intervenant
dans cette régulation a été décrit chez le maïs, il s’agit d’Opaque2, qui se fixe sur le motif GLM
(Lohmer et al., 1991). Aujourd’hui, huit FT sont connus, tous étant activateurs de la
transcription des gènes de PR. Les interactions entre ces FT et les cis-motifs ont fait l’objet de
nombreuses études, notamment chez l’orge, conduisant à proposer un schéma de régulation de
la synthèse des PR (Diaz et al., 2002; Rubio-Somoza et al., 2006a et b ; Moreno-Risueno et al.,
2008). Ce schéma semble être conservé chez les céréales (Verdier and Thompson, 2008 ; Xi
and Zheng, 2011) et la fixation de certains FT sur les cis-motifs de gènes de PR a été confirmée
chez le blé (Albani et al., 1997 ; Conlan et al., 1999 ; Ravel et al., 2014).

22
 Chapitre 1 : Synthèse bibliographique

Le noyau apparaît donc comme un lieu de régulation de l’accumulation des PR dans le

grain de blé. Il a également été mis en évidence un effet de la nutrition sur ce schéma de
régulation. En effet, le motif GLM est activateur en condition de nutrition azotée optimale et
devient répresseur en condition de carence en N (Muller and Knudsen, 1993). Malgré ces
différents éléments, les mécanismes moléculaires qui régulent l’expression des gènes de PR
sont encore peu connus, et les interactions qui surviennent in vivo entre les différents acteurs
sont probablement à un niveau de complexité plus élevé (Mehrotra et al., 2009).

4.2 Le noyau : un compartiment cellulaire très dynamique

Au sein de la cellule eucaryote, le noyau est le compartiment cellulaire qui renferme
l’essentiel de l’information génétique, également contenue au sein des mitochondries et des
chloroplastes. L’ADN nucléaire forme avec des protéines appelées histones les unités de base
de la chromatine, les nucléosomes. La particule de cœur du nucléosome est formée d’un
octamère d’histones (deux exemplaires de chacune des histones H2A, H2B, H3 et H4) autour
duquel s’enroule l’ADN. Cette structure est ensuite scellée par les histones H1. C’est le premier
niveau de compaction de la chromatine. Cette compaction n’est pas homogène tout au long de
la molécule d’ADN. Des zones faiblement condensées, l’euchromatine, accessibles à la
machinerie de transcription et des zones fortement condensées et inaccessibles correspondant à
l’hétérochromatine sont observées. Des complexes de remodelage de la chromatine sont
capables de modifier cet état de compaction et ainsi de réguler la transcription des gènes. Ce
mécanisme épigénétique intervient dans la régulation du développement de la plante et dans la
réponse de la plante aux facteurs environnementaux (Jarillo et al., 2009).
En plus de contenir la chromatine, le noyau renferme différentes structures (Figure 16 ;
Petrovská et al., 2015). La plus connue est le nucléole, dont les principales fonctions sont la
synthèse des ARN ribosomiques et l’assemblage des ribosomes (Fromont-Racine et al., 2003;
Boisvert et al., 2007). Chez A. thaliana, 217 protéines ont été identifiées au sein du nucléole,
dont beaucoup correspondent à des protéines ribosomales (Pendle et al., 2005). La traduction
ayant lieu dans le cytoplasme, il y a un échange de matériel permanent entre le noyau et le
cytoplasme, avec une sortie des ARNm et une importation de protéines qui possèdent un signal
de localisation nucléaire. Ce transport nucléo-cytoplasmique est permis par des structures qui
interrompent l’enveloppe nucléaire, appelées pores nucléaires (Nuclear Pore Complex, NPC ;
Tamura and Hara-Nishimura, 2013) et fait intervenir des protéines de transport : les importines
et exportines. L’export des ARNm à travers les NPC est très contrôlé (Erkmann and Kutay,
2004) et il existe une synchronisation entre l’import des protéines ribosomales et le processus
d’assemblage des ribosomes (Kressler et al., 2012). Le transport nucléo-cytoplasmique est donc
un mécanisme très contrôlé, ayant un rôle central dans le métabolisme cellulaire et le
développement de la plante (Meier and Brkljacic, 2009; Merkle, 2011; Tamura and Hara-
Nishimura, 2014). Ce transport est également très rapide puisqu’il a été estimé que le nombre
de translocations par NPC avoisine les 800 par seconde (Fried and Kutay, 2003). D’autres
structures existent au sein du noyau comme les corps de Cajal et les granules (speckles), qui
semblent avoir un rôle dans l’épissage des ARN pré-messagers.

23
 Chapitre 1 : Synthèse bibliographique

Figure 16. Représentation schématique d’un noyau de cellule végétale. Sur la gauche, une photo de
noyaux purifiés par cytométrie de flux, colorés au DAPI et observés en microscopie à fluorescence, chez
l’orge. D’après Petrovská et al. (2015).

Le noyau est donc une structure très dynamique, renfermant une grande diversité de
protéines, parfois présentes dans le noyau de manière très ponctuelle, et intervenant dans des
processus très diversifiés. L’ensemble de ces processus, liés de près ou de loin à l’ADN, via sa
réplication, sa transcription ou sa conformation, à la traduction ou encore au trafic
intracellulaire a un rôle majeur dans le développement de la plante. Bien que la structure et les
fonctions du noyau soient assez bien caractérisées, l’état des connaissances est largement dû
aux études menées au sein des cellules animales et l’étude du protéome nucléaire chez les
plantes en est à ses prémices.

4.3 L’extraction et l’analyse des protéines nucléaires

Chez la souris, les protéines nucléaires représentent environ 14% du protéome total (Fink
et al., 2008). Une proportion comparable a été estimée chez les plantes, entre 10 et 20% (Narula
et al., 2013), mais la diversité de ces protéines est souvent sous-représentée dans les études de
protéomique, notamment du fait de la forte proportion des protéines histones dans les extraits
nucléaires. Les protéines nucléaires intervenant dans les mécanismes de régulation sont souvent
en très faible abondance, comme par exemple les FT dont une faible quantité est souvent
suffisante pour activer ou réprimer l’expression des gènes qu’ils régulent. Pour avoir une vision
la plus représentative possible de la diversité protéique présente au sein du noyau, il est ainsi
nécessaire d’avoir recours à des étapes de purification de noyaux et/ou d’enrichissement en
protéines nucléaires.
Les récents progrès méthodologiques et technologiques en termes de fractionnement
subcellulaire et en spectrométrie de masse permettent aujourd’hui d’étudier le protéome au
niveau des différents organites de la cellule : vacuole, mitochondrie, chloroplaste, noyau et
même au niveau du nucléole comme nous l’avons vu précédemment. Diverses méthodes
existent pour isoler et purifier l’organite souhaité, dont les plus courantes sont la centrifugation
différentielle et la centrifugation en gradient de densité (Lee et al., 2010). La centrifugation

24
 Chapitre 1 : Synthèse bibliographique

différentielle consiste en une centrifugation de l’extrait cellulaire à différentes vitesses, chaque

vitesse entraînant le dépôt dans le culot d’un organite en particulier. Pour la centrifugation en
gradient de densité, l’extrait cellulaire est déposé dans un gradient continu ou discontinu, réalisé
avec des milieux de viscosité, d’osmolarité ou de densité différents. Les plus utilisés sont les
gradients de sucrose, de Ficoll ou de Percoll. Après centrifugation, l’organite se concentre dans
le gradient à l’endroit correspondant à sa propre densité. Cette méthode basée sur un gradient
de densité a été la plus utilisée pour purifier des noyaux dans les études menées chez les plantes
(Petrovská et al., 2015), et a été utilisée au cours de cette thèse. Récemment chez l’orge, une
méthode de purification de noyaux par cytométrie de flux a été développée et a permis
l’identification de plus de 800 protéines (Petrovská et al., 2014).

4.4 L’étude des protéines nucléaires chez les plantes

Même si le fonctionnement du noyau de la cellule végétale n’est pas encore totalement
élucidé et si les acteurs nucléaires ne sont pas tous identifiés et caractérisés, l’exploration du
protéome nucléaire a tout de même fait l’objet de plusieurs études (Figure 17 ; Petrovská et al.,
2015). Chez les céréales, des études chez le maïs (Ferreira et al., 2006; Casati et al., 2008; Guo
et al., 2014), le riz (Khan and Komatsu, 2004; Tan et al., 2007; Li et al., 2008; Aki and
Yanagisawa, 2009; Jaiswal et al., 2013) et récemment l’orge (Petrovská et al., 2014) ont été
réalisées. La plupart de ces études ont été effectuées sur des plantules ou à partir de feuilles, et
peu finalement se sont intéressées au protéome nucléaire du grain (Repetto et al., 2012). Deux
études principales ont analysé l’albumen de grains de riz (Li et al., 2008) et des graines entières
de luzerne (Repetto et al., 2008), permettant d’avoir une première vision globale des protéines
nucléaires mises en action pendant le développement du grain.

Figure 17. Chronologie des analyses du protéome nucléaire réalisées chez les plantes. Les études
indiquées en bleu ont été réalisées à partir de grains. Lorsque le protéome nucléaire a été analysé en
réponse à des facteurs environnementaux, la nature du stress appliqué est indiquée par un symbole
encadré en rouge. Adapté de Petrovská et al. (2015).

25
 Chapitre 1 : Synthèse bibliographique

Outre l’objectif d’obtenir des informations sur les fonctions du noyau de la cellule
végétale, plusieurs des études présentées en Figure 17 ont ciblé le protéome nucléaire afin de
mettre en évidence des protéines régulatrices impliquées dans la réponse de la plante à
l’environnement. Divers types de facteurs environnementaux ont ainsi été étudiés, notamment
le stress hydrique (Pandey et al., 2008 ; Choudhary et al., 2009 ; Abdalla et al., 2010 ; Abdalla
and Rafudeen, 2012 ; Subba et al., 2013 ; Jaiswal et al., 2013), le stress au froid (Bae et al.,
2003) ou encore le stress dû à une maladie fongique comme la rouille du soja (Cooper et al.,
2011). Chez le riz par exemple, l’application d’un stress hydrique a permis de mettre en
évidence 78 protéines nucléaires potentiellement au centre de mécanismes d’adaptation de la
plante à la sécheresse, comme des FT et des protéines de remodelage de la chromatine (Jaiswal
et al., 2013). L’étude du protéome nucléaire est donc très utile pour avoir accès aux protéines
régulatrices du développement de la plante et de son adaptation à l’environnement, qui sont
souvent diluées dans l’extrait protéique total.

26
 Chapitre 1 : Synthèse bibliographique

5. L’INTEGRATION DE DONNEES -OMIQUES POUR VISUALISER LA REPONSE

BIOLOGIQUE
5.1 Biologie des systèmes
Les évolutions récentes en termes de méthodologie et de technologie permettent
aujourd’hui d’obtenir de très nombreuses informations sur la réponse cellulaire à un stress ou
un stimulus. L’apparition des technologies de séquençage nouvelle génération a par exemple
considérablement augmenté les débits de séquençage d’ADN tout en diminuant les coûts
associés et a permis l’émergence de nouvelles approches comme le séquençage massif d’ARN,
qui permet de connaître l’intégralité des gènes exprimés dans un échantillon, que le génome de
l’espèce étudiée soit complet et disponible ou non (Conesa et al., 2016). Des progrès ont
également été réalisés en spectrométrie de masse et permettent aujourd’hui de détecter et
quantifier finement une large diversité de protéines et métabolites présents dans un échantillon.
Plusieurs échelons sont distingués parmi les données -omiques, avec notamment la
transcriptomique, la protéomique et la métabolomique qui font le lien entre le génome et le
phénotype observé (Figure 18 ; Mochida and Shinozaki, 2011). De nouveaux types de données
-omiques voient également le jour avec par exemple l’étude de l’épigénome qui correspond à
l’ensemble des marques épigénétiques du génome ou encore l’hormonome, représentant les
molécules hormonales, qui ont un intérêt important chez les plantes du fait de leur rôle dans les
mécanismes de signalisation moléculaire et de réponse à l’environnement.

Figure 18. Les différents niveaux de données -omiques et des exemples de ressources disponibles
à chaque niveau chez Arabidopsis thaliana. D’après Mochida and Shinozaki (2011)

27
 Chapitre 1 : Synthèse bibliographique

Les technologies -omiques permettent donc d’obtenir de nombreuses informations à

différents niveaux biologiques, de la séquence des gènes à l’accumulation des protéines et des
composés métaboliques. Ces données peuvent ainsi couvrir tous les mécanismes impliqués dans
les variations qui se produisent dans la cellule et qui influencent le fonctionnement du système
d’étude. Les technologies -omiques représentent une nouvelle procédure de recherche dans
laquelle la formulation et la vérification des hypothèses biologiques sont effectuées en aval de
la production des données. Elles sont donc particulièrement utiles pour l’exploration sans a
priori et la compréhension des mécanismes moléculaires impliqués dans la réponse de la plante,
par exemple à un facteur environnemental (Urano et al., 2010).
Avec l’utilisation croissante de ces approches -omiques s’est ainsi développée la biologie
des systèmes, concept introduit par Ludwig von Bertalanffy (Von Bertalanffy, 1968).
Contrairement à la biologie moléculaire traditionnelle qui se concentre sur la caractérisation de
composants individuels de la cellule (ARNs, protéines, métabolites, etc), la biologie des
systèmes est de nature intégrative et considère les différents composants comme faisant partie
d’un ensemble (Carvunis et al., 2009). Ainsi le principe repose sur la combinaison des données
-omiques obtenues à différentes échelles dans le but d’analyser et comprendre comment les
interactions observées dans la cellule à ces différents niveaux expliquent le phénotype observé.
Mais face à la multiplication des données générées, la principale difficulté réside dans la
capacité de les analyser et les intégrer afin de tirer des conclusions biologiques. En effet, les
multiples interactions entre les différents composants du système rendent souvent les
interprétations difficiles et témoignent de la réelle complexité de la réponse cellulaire (Fernie,
2012).

5.2 Les différents types de réseaux biologiques

Face à cette accumulation des données -omiques, les réseaux biologiques sont rapidement
apparus comme une approche attractive pour extraire de ces larges jeux de données
l’information nécessaire à la compréhension des mécanismes moléculaires qui interviennent
dans une réponse biologique (Usadel et al., 2009 ; Moreno-Risueno et al., 2010).
Graphiquement, les réseaux biologiques sont représentés par des sommets qui peuvent
correspondre à diverses entités moléculaires telles que des gènes, protéines ou encore des
métabolites, et par des arêtes qui relient les sommets entre eux. Il existe différents types de
réseaux en fonction de la signification des arêtes, c’est-à-dire en fonction de la nature de
l’interaction qui existe entre les sommets. Ainsi il existe des réseaux de co-expression,
d’interactions protéine-protéine, les réseaux métaboliques et les réseaux de régulation et de
signalisation (Figure 19 ; Bassel et al., 2012).

28
 Chapitre 1 : Synthèse bibliographique

Figure 19. Représentation schématique de différents types de réseaux biologiques. (A) et (B)
Réseaux de co-expression, deux gènes sont liés s’ils ont un profil d’expression similaire (A) et/ou
partagent des propriétés communes (B). (C), (D) et (E) Réseaux de régulation, liaison entre un FT et le
promoteur d’un (C) ou plusieurs (D) gène(s) qu’il régule, liaison de plusieurs FT sur le même promoteur
d’un gène (E). (F) Réseau d’interaction physique protéine-protéine démontrée expérimentalement. (G)
Réseau d’interaction métabolique représentant une réaction biochimique qui convertit un métabolite en
un autre. Les gènes sont colorés en noir, les protéines en gris et les métabolites en blanc. D’après Bassel
et al. (2012).

Les réseaux de co-expression se sont fortement développés ces dernières années avec
l’obtention de grands jeux de données dans le domaine de la transcriptomique, générés avec les
approches de puce à ADN et de RNA-Seq (Serin et al., 2016). Pour ce type de réseau, les profils
d’expression des différents gènes sont comparés et un score de similarité est calculé pour
chaque comparaison de deux gènes, par exemple via le calcul d’un coefficient de corrélation.
A partir d’un certain seuil, on considère que les gènes sont co-exprimés. Le réseau est alors
construit à partir de la liste des gènes co-exprimés. Dans ce type de réseau, le principe de « guilt
by association » est appliqué, c’est-à-dire que les gènes qui ont des profils d’expression
similaires sont susceptibles de partager des fonctions similaires ou de faire partie de la même
voie de régulation (Oliver, 2000). Ainsi, au sein des groupes de gènes fortement interconnectés
dans le réseau, appelés modules, il est possible à partir des fonctions qui y sont représentées
d’émettre des hypothèses sur les fonctions des gènes non caractérisés fonctionnellement (Aoki
et al., 2007; Silva et al., 2016). Les modules sont souvent utilisés comme point de départ dans
l’interprétation puisqu’ils réduisent considérablement la complexité du réseau (Aoki et al.,
2007). Les sommets qui apparaissent avec un fort degré de connexion sont alors identifiés
comme des « hubs » au sein des modules et sont souvent considérés comme des gènes essentiels
du réseau (Provero, 2002).
Les réseaux d’interactions protéine-protéine ou d’interactions protéine-ADN sont
construits à partir de résultats issus des méthodes de double hybride (interaction protéine-
protéine) par exemple et de simple hybride ou d’immunoprécipitation de chromatine
(interaction protéine-ADN). Dans les réseaux générés, les interactions correspondent ainsi à des
liens physiques qui ont été démontrés expérimentalement. Dans cette optique, de nombreux
réseaux sont générés pour étudier les interactions entre des FT et des séquences ADN
(Gaudinier and Brady, 2016). Chez A. thaliana par exemple, des expériences de simple hybride
ont été à l’origine d’un réseau transcriptionnel qui a permis de révéler de nouveaux régulateurs

29
 Chapitre 1 : Synthèse bibliographique

de la synthèse de la paroi secondaire des cellules racinaires, modulés par le stress salin ou la
carence en fer (Taylor-Teeples et al., 2014). Par rapport à un réseau de co-expression qui est
non orienté, le réseau de régulation transcriptionnel généré dans cette étude va à un niveau plus
loin dans l’analyse puisqu’il permet de caractériser le sens et le type d’interaction entre les deux
entités moléculaires (Figure 19).
Il est également possible d’apporter les informations d’interactions physiques ou de
régulation disponibles pour les gènes, protéines ou métabolites présents dans un réseau de co-
expression afin d’avoir une vision la plus complète possible. Pour faire le lien entre le génotype
et le phénotype et comprendre les mécanismes de régulation impliqués il apparaît en effet
nécessaire dans la mesure du possible de combiner les réseaux obtenus à différents niveaux
moléculaires (gènes, protéines, métabolites ; Gaudinier et al., 2015).

5.3 RulNet : l’outil utilisé au cours de la thèse

Plusieurs méthodes existent pour inférer des réseaux biologiques, en fonction des données
disponibles ou encore de la nature des interactions. Dans cette thèse, des données de quantités
de protéines et de métabolites et des données d’expression de gènes ont été obtenues. La
construction de réseaux de co-expression a donc pu être envisagée pour intégrer les données
générées. Plusieurs méthodes sont couramment utilisées pour générer ce type de réseaux. En
fonction de la méthode, il existe ainsi les réseaux booléens, les Relevance Networks, les réseaux
bayésiens et les réseaux basés sur la découverte de règles d’association.
Pour les réseaux booléens, les données d’expression sont binarisées, un gène est soit dans un
état actif soit inactif. L’objectif est de trouver des règles logiques permettant de déterminer
l’état d’un gène à un instant t+1 lorsque l’on connait l’état de ce gène et d’autres gènes à un
instant t. Les réseaux bayésiens se basent quant à eux sur des distributions de probabilités
conditionnelles entre les variables. Pour chaque couple de gènes (G1, G2), la probabilité d’avoir
G1 sachant G2 est calculée. Une relation de causalité existe alors si cette probabilité est
importante. Dans le cas des Relevance Networks, la corrélation entre chaque paire de gènes est
calculée, basée sur le coefficient de corrélation de Pearson. Les coefficients sont ensuite filtrés
et conservés s’ils dépassent un certain seuil.
L’outil RulNet ([Link] ; Vincent et al., 2015), utilisé au cours de cette thèse
pour générer des réseaux de co-expression, est basé sur la découverte de règles d’association.
Une règle est une expression de la forme X  Y et se lit « X implique Y ». Ainsi cela signifie
que lorsqu’une propriété, appelée prédicat est observée sur X (antécédent) alors un prédicat est
observé sur Y (conséquent), les prédicats étant identiques ou non. Il existe ensuite différentes
mesures de qualité des règles qui permettent d’évaluer leur pertinence, dont le support, la
confiance et le lift. Le support exprime la fréquence d’une règle, et correspond à la probabilité
que X et Y soient simultanément satisfaits (exemple : proportion d’échantillons dans lesquels
le gène 1 et le gène 2 sont tous les deux surexprimés). La confiance correspond à la probabilité
que Y soit satisfait sachant que X est satisfait. Lorsque la confiance est égale à 1 la règle est
dite exacte, sinon elle est approximative. Le lift mesure quant à lui la dépendance entre X et Y

30
 Chapitre 1 : Synthèse bibliographique

et correspond au rapport entre la probabilité réelle d’avoir X et Y satisfaits et la probabilité

attendue si X et Y étaient statistiquement indépendants.
La méthode de découverte de règles d’association a été pour la première fois appliquée à
l’inférence de réseaux de gènes en 1997 (Brazma et al., 1997). Avec RulNet, cette approche a
été adaptée à l’intégration de données -omiques chez le blé (Vincent et al., 2015 ; Dai et al.,
2015). Ainsi, X et Y peuvent correspondre à différentes variables comme des gènes, des ARNs,
protéines, métabolites ou encore des variables phénotypiques. Dans RulNet, les sémantiques
choisies en fonction des règles à inférer sont rédigées sous la forme de requêtes RQL (Chardin
et al., 2014). Il est possible de définir des seuils pour les différentes mesures de qualité des
règles mentionnées précédemment, afin de ne générer que les règles qui seront considérées
comme significatives. Le réseau est ensuite visualisable sous d’autres outils comme Cytoscape
([Link] ; Smoot et al., 2011).
Cette approche laisse la possibilité au biologiste de générer plusieurs réseaux ou un seul
réseau global à partir de différentes sémantiques qu’il peut choisir en fonction de ses objectifs
biologiques et des caractéristiques de ses données. L’utilisateur a ainsi le choix de la
signification biologique des liens à inférer. Il est également possible de définir des gènes
centraux, c’est-à-dire des gènes qu’il souhaite voir apparaître dans le réseau, et les règles
générées seront alors limitées à celles impliquant ces entités d’intérêt.

31
 Chapitre 1 : Synthèse bibliographique

6. CONCLUSIONS DE LA SYNTHESE ET STRATEGIE D’ETUDE

Comme nous avons pu le voir dans cette synthèse bibliographique, la qualité du grain de
blé est essentiellement déterminée par la teneur et la composition en PR. L’accumulation des
différentes classes de gliadines et sous-unités de gluténines, qui dépend fortement de la
disponibilité en N est S du sol, doit être contrôlée afin d’obtenir les caractéristiques
nutritionnelles et technologiques souhaitées. Or, le contexte actuel avec un objectif
d’augmentation du rendement en grains et des conditions de nutrition de la plante modifiées,
avec une limitation des apports d’engrais azotés et une carence en S observée dans les sols,
menace de réduire la qualité du grain. Afin de maintenir cette qualité, il est nécessaire de
comprendre les mécanismes moléculaires impliqués dans la régulation du développement du
grain et dans la régulation de la synthèse des PR en réponse à la nutrition azotée et soufrée. Il
apparaît également que les PR du grain de blé sont régulées au niveau transcriptionnel. Le noyau
apparait donc comme un compartiment cellulaire très important et une cible de choix pour
l’étude de ces mécanismes de régulation. Le but de cette thèse a ainsi été d’étudier le protéome
nucléaire du grain de blé au cours du développement et en réponse à différents apports de N et
S, puis d’obtenir une vision globale de la réponse du grain à la nutrition dans une approche
intégrative combinant différents types de données -omiques (transcrits, métabolites, protéines).
L’étude a été divisée en trois objectifs principaux (Figure 20).

Figure 20. Schéma

situant les objectifs
du projet de thèse
en lien avec les
espèces de blé
étudiées.

(1) Basé sur une méthode de purification de noyaux et d’extraction de protéines nucléaires
mise au point par Emmanuelle Bancel avant mon arrivée qui a dû être validée (Chapitre 2), le
premier objectif a consisté à identifier le protéome nucléaire du grain de blé tendre (T. aestivum)
jusqu’alors inexploré, puis étudier les variations d’abondance de ces protéines au cours du
développement du grain (Chapitre 3).
(2) Le second objectif a été d’identifier les protéines nucléaires qui pouvaient être
impactées par la nutrition azotée et soufrée pendant le remplissage du grain et de lier ces
variations avec l’accumulation des PR et des protéines métaboliques (Chapitre 4). Ces analyses

32
 Chapitre 1 : Synthèse bibliographique

protéomiques ont été réalisées chez l’engrain (T. monococcum), espèce diploïde choisie comme
modèle du blé tendre.
(3) Le troisième objectif consistait à obtenir une vision intégrative de la réponse de T.
monococcum à la nutrition azotée et soufrée (Chapitre 5). Pour cela, des analyses de
transcriptomique et de métabolites ont été réalisées. L’ensemble des données obtenues à partir
du même matériel biologique (données de transcripomique, métabolomique et enfin de
protéomique analysées dans l’objectif 2) a été intégré au sein de réseaux biologiques.

33
 Chapitre 1 : Synthèse bibliographique

REFERENCES
Abdalla, K.O., Baker, B. and Rafudeen, M.S. (2010) Proteomic analysis of nuclear proteins
during dehydration of the resurrection plant Xerophyta viscosa. Plant Growth Regul., 62,
279–292.
Abdalla, K.O. and Rafudeen, M.S. (2012) Analysis of the nuclear proteome of the
resurrection plant Xerophyta viscosa in response to dehydration stress using iTRAQ with
2DLC and tandem mass spectrometry. J. Proteomics, 75, 2361–74.
Acevedo, E., Silva, P. and Silva, H. (2002) Wheat growth and physiology. Available at:
[Link]
Aki, T. and Yanagisawa, S. (2009) Application of rice nuclear proteome analysis to the
identification of evolutionarily conserved and glucose-responsive nuclear proteins. J.
Proteome Res., 8, 3912–24.
Albani, D., Hammond-Kosack, M.C.U., Smith, C., Conlan, S., Colot, V., Holdsworth, M.
and Bevan, M.W. (1997) The wheat transcriptional activator SPA: a seed-specific bZIP
protein that recognizes the GCN4-like motif in the bifactorial endosperm box of prolamin
genes. Plant Cell, 9, 171–184.
Altenbach, S.B., Kothari, K.M. and Lieu, D. (2002) Environmental conditions during wheat
grain development alter temporal regulation of major gluten protein genes. Cereal Chem.,
79, 279–285.
Aoki, K., Ogata, Y. and Shibata, D. (2007) Approaches for extracting practical information
from gene co-expression networks in plant biology. Plant Cell Physiol., 48, 381–390.
Arvalis-Institut du végétal (2016) Comment s’élabore le rendement des céréales à paille ?
Available at: [Link]
[Link].
Asseng, S., Ewert, F., Martre, P., et al. (2015) Rising temperatures reduce global wheat
production. Nat. Clim. Chang., 5, 143–147.
Austin, R.B., Ford, M.A., Edrich, J.A. and Blackwell, R.D. (1977) The nitrogen economy
of winter wheat. J. Agric. Sci., 88, 159–167.
Bae, M.S., Cho, E.J., Choi, E.-Y. and Park, O.K. (2003) Analysis of the Arabidopsis nuclear
proteome and its response to cold stress. Plant J., 36, 652–663.
Barron, C., Surget, A. and Rouau, X. (2007) Relative amounts of tissues in mature wheat
(Triticum aestivum L.) grain and their carbohydrate and phenolic acid composition. J.
Cereal Sci., 45, 88–96.
Bassel, G.W., Gaudinier, A., Brady, S.M., Hennig, L., Rhee, S.Y. and Smet, I. De (2012)
Systems analysis of plant functional, transcriptional, physical interaction, and metabolic
networks. Plant Cell, 24, 3859–3875.
Bates, B., Lennox, A., Prentice, A., Bates, C., Page, P., Nicholson, S. and Swan, G. (2011)
National diet and nutrition survey : results from years 1-4 (combined) of the rolling
programme about public health england. Public Heal. England, London, UK, 4, 1–24.
Bertalanffy, L. Von (1968) General system theory, Georg. Braziller New York.
Bogard, M., Allard, V., Brancourt-Hulmel, M., Heumez, E., Machet, J.-M., Jeuffroy, M.-
H., Gate, P., Martre, P. and LeGouis, J. (2010) Deviation from the grain protein
concentration-grain yield negative relationship is highly correlated to post-anthesis N
uptake in winter wheat. J. Exp. Bot., 61, 4303–4312.
Boisvert, F.-M., Koningsbruggen, S. van, Navascués, J. and Lamond, A.I. (2007) The
multifunctional nucleolus. Nat. Rev. Mol. Cell Biol., 8, 574–85.
Bonjean, A. (1994) Histoire de la culture des céréales et en particulier de celle du blé tendre
(Triticum aestivum L.). Doss. l’environnement l'INRA, 21, 29–37.
Brancourt-Hulmel, M., Doussinault, G., Lecomte, C., Bérard, P., Buanec, B. Le and
Trottet, M. (2003) Genetic improvement of agronomic traits of winter wheat cultivars

34
 Chapitre 1 : Synthèse bibliographique

released in France from 1946 to 1992. Crop Sci., 43, 37–45.

Branlard, G., Dardevet, M., Saccomano, R., Lagoutte, F. and Gourdon, J. (2001) Genetic
diversity of wheat storage proteins and bread wheat quality. Euphytica, 119, 59–67.
Branlard, G., Lesage, V.S., Bancel, E., Martre, P., Méleard, B. and Rhazi, L. (2015) Coping
with wheat quality in a changing environment: proteomics evidence for stress caused by
environmental changes. In Advances in Wheat Genetics: From Genome to Field. pp. 255–
264.
Brazma, A., Vilo, J., Ukkonen, E. and Valtonen, K. (1997) Data mining for regulatory
elements in yeast genome. Proc. Int. Conf. Intell. Syst. Mol. Biol., 5, 65–74.
Brisson, N., Gate, P., Gouache, D., Charmet, G., Oury, F.-X. and Huard, F. (2010) Why
are wheat yields stagnating in Europe? A comprehensive data analysis for France. F. Crop.
Res., 119, 201–212.
Buchner, P. and Hawkesford, M.J. (2014) Complex phylogeny and gene expression patterns
of members of the nitrate transporter 1/peptide transporter family (NPF) in wheat. J. Exp.
Bot., 65, 5697–5710.
Buchner, P., Parmar, S., Kriegel, A., Carpentier, M. and Hawkesford, M.J. (2010) The
sulfate transporter family in wheat: tissue-specific gene expression in relation to nutrition.
Mol. Plant, 3, 374–389.
Carvunis, A., Gomez, E., Thierry-mieg, N., Trilling, L. and Vidal, M. (2009) Biologie
systémique. Medecine/Sciences, 25, 578–584.
Casati, P., Campi, M., Chu, F., Suzuki, N., Maltby, D., Guan, S., Burlingame, A.L. and
Walbot, V. (2008) Histone acetylation and chromatin remodeling are required for UV-B-
dependent transcriptional activation of regulated genes in maize. Plant Cell, 20, 827–42.
Cassman, K.G., Dobermann, A., Walters, D.T. and Yang, H. (2003) Meeting cereal demand
while protecting natural resources and improving environmental quality. Annu. Rev.
Environ. Resour., 28, 315–358.
Chardin, B., Coquery, E., Pailloux, M. and Petit, J.-M. (2014) RQL: a SQL-like query
language for discovering meaningful rules. Int. Conf. data Min. Work., 1, 1203–1206.
Chojecki, A.J.S., Bayliss, M.W. and Gale, M.D. (1986) Cell production and DNA
accumulation in the wheat endosperm , and their association with grain weight. Ann. Bot.,
58, 809–817.
Chope, G.A., Wan, Y., Penson, S.P., Bhandari, D.G., Powers, S.J., Shewry, P.R. and
Hawkesford, M.J. (2014) Effects of genotype, season, and nitrogen nutrition on gene
expression and protein accumulation in wheat grain. J. Agric. Food Chem., 62, 4399–4407.
Choudhary, M.K., Basu, D., Datta, A., Chakraborty, N. and Chakraborty, S. (2009)
Dehydration-responsive nuclear proteome of rice (Oryza sativa L.) illustrates protein
network, novel regulators of cellular adaptation, and evolutionary perspective. Mol. Cell.
Proteomics, 8, 1579–1598.
Conesa, A., Madrigal, P., Tarazona, S., et al. (2016) A survey of best practices for RNA-seq
data analysis. Genome Biol., 17, 1–19.
Conlan, R.S., Hammond-Kosack, M. and Bevan, M. (1999) Transcription activation
mediated by the bZIP factor SPA on the endosperm box is modulated by ESBF-1 in vitro.
Plant J., 19, 173–181.
Cooper, B., Campbell, K.B., Feng, J., Garrett, W.M. and Frederick, R. (2011) Nuclear
proteomic changes linked to soybean rust resistance. Mol. Biosyst., 7, 773–83.
Corbellini, M., Empilli, S., Vaccino, P., Brandolini, A., Borghi, B., Heun, M. and Salamini,
F. (1999) Einkorn characterization for bread and cookie production in relation to protein
subunit composition. Cereal Chem., 76, 727–733.
D’Ovidio, R. and Masci, S. (2004) The low-molecular-weight glutenin subunits of wheat
gluten. J. Cereal Sci., 39, 321–339.

35
 Chapitre 1 : Synthèse bibliographique

Dai, Z., Plessis, A., Vincent, J., et al. (2015) Transcriptional and metabolic alternations
rebalance wheat grain storage protein accumulation under variable nitrogen and sulfur
supply. Plant J., 83, 326–343.
Diaz, I., Vicente-Carbajosa, J., Abraham, Z., Martinez, M., Isabel-La Moneda, I. and
Carbonero, P. (2002) The GAMYB protein from barley interacts with the DOF
transcription factor BPBF and activates endosperm-specific genes during seed
development. Plant J., 29, 453–464.
Dubcovsky, J. and Dvorak, J. (2007) Genome plasticity a key factor in the success of
polyploid wheat under domestication. Science, 316, 1862–1866.
Dupont, F.M. and Altenbach, S.B. (2003) Molecular and biochemical impacts of
environmental factors on wheat grain development and protein synthesis. J. Cereal Sci.,
38, 133–146.
Eriksen, J. (2009) Soil sulfur cycling in temperate agricultural systems. In Advances in
Agronomy. pp. 55–89.
Erkmann, J. a and Kutay, U. (2004) Nuclear export of mRNA: from the site of transcription
to the cytoplasm. Exp. Cell Res., 296, 12–20.
Evans, I.M., Gatehouse, J.A., Croy, R.R.D. and Boulter, D. (1984) Regulation of the
transcription of storage-protein mRNA in nuclei isolated from developing pea (Pisum
sativum L.) cotyledons. Planta, 160, 559–568.
Evers, A.D., Blakeney, A.B. and O’Brien, L. (1999) Cereal structure and composition. Aust.
J. Agric. Res., 50, 629–650.
Evers, T. and Millar, S. (2002) Cereal grain structure and development: some implications for
quality. J. Cereal Sci., 36, 261–284.
FAO (2016) Crop Prospects and Food Situation. October, 1–42.
Fazili, I.S., Jamal, A., Ahmad, S., Masoodi, M., Khan, J.S. and Abdin, M.Z. (2008)
Interactive effect of sulfur and nitrogen on nitrogen accumulation and harvest in oilseed
crops differing in nitrogen assimilation potential. J. Plant Nutr., 31, 1203–1220.
Feil, B. (1997) The inverse yield-protein relationship in cereals: Possibilities and limitations
for genetically improving the grain protein yield. Trends Agron., 1, 103–119.
Feillet, P. (2000) Le grain de blé : composition et utilisation INRA., Paris.
Fernie, A.R. (2012) Grand challenges in plant systems biology: closing the circle(s). Front.
Plant Sci., 3, 35.
Ferreira, B.S., Branco, A.T., Oliveira, M.A. de, Pereira, M.G. and Souza Filho, G.A. de
(2006) Methodological improvements on extraction of nuclear proteins and its preliminary
analysis during the maize (Zea mays L.) endosperm development. Protein Pept. Lett., 13,
981–4.
Fink, J.L., Karunaratne, S., Mittal, A., et al. (2008) Towards defining the nuclear proteome.
Genome Biol., 9, R15.
FranceAgrimer (2015) Bilans céréaliers : campagne 2014/2015.
Fried, H. and Kutay, U. (2003) Nucleocytoplasmic transport: taking an inventory. Cell. Mol.
Life Sci., 60, 1659–1688.
Fromont-Racine, M., Senger, B., Saveanu, C. and Fasiolo, F. (2003) Ribosome assembly in
eukaryotes. Gene, 313, 17–42.
Gatehouse, J.A., Evans, I.M., Bown, D., Croy, R.R.D. and Boulter, D. (1982) Control of
storage-protein synthesis during seed development in pea (Pisum sativum L.). Biochem.
J., 208, 119–127.
Gaudinier, A. and Brady, S.M. (2016) Mapping transcriptional networks in plants: data-
driven discovery of novel biological mechanisms. Annu. Rev. Plant Biol., 67, 575–94.
Gaudinier, A., Tang, M. and Kliebenstein, D.J. (2015) Transcriptional networks governing
plant metabolism. Curr. Plant Biol., 3-4, 56–64.

36
 Chapitre 1 : Synthèse bibliographique

Giese, H. and Hopp, H.E. (1984) Influence of nitrogen nutrition on the amount of hordein,
protein Z and β-amylase messenger RNA in developing endosperms of barley. Carlsberg
Res. Commun., 49, 365–383.
Giles, J. (2005) Nitrogen study fertilizes fears of pollution. Nature, 433, 791–791.
Giordano, M. and Raven, J.A. (2014) Nitrogen and sulfur assimilation in plants and algae.
Aquat. Bot., 118, 45–61.
Girin, T., Lejay, L., Wirth, J., Widiez, T., Palenchar, P.M., Nazoa, P., Touraine, B., Gojon,
A. and Lepetit, M. (2007) Identification of a 150 bp cis -acting element of the AtNRT2.1
promoter involved in the regulation of gene expression by the N and C status of the plant.
Plant. Cell Environ., 30, 1366–1380.
Graybosch, R.A. and Peterson, C.J. (2010) Genetic improvement in winter wheat yields in
the great plains of north America, 1959–2008. Crop Sci., 50, 1882.
Guo, B., Chen, Y., Li, C., et al. (2014) Maize (Zea mays L.) seedling leaf nuclear proteome
and differentially expressed proteins between a hybrid and its parental lines. Proteomics,
14, 1071–87.
Hawkesford, M.J. (2000) Plant responses to sulphur deficiency and the genetic manipulation
of sulphate transporters to improve S-utilization efficiency. J. Exp. Bot., 51, 131–138.
Hawkesford, M.J. (2003) Transporter gene families in plants: the sulphate transporter gene
family - redundancy or specialization? Physiol. Plant., 117, 155–163.
Hawkesford, M.J. and DeKok, L.J. (2006) Managing sulphur metabolism in plants. Plant,
Cell Environ., 29, 382–395.
Heun, M. (1997) Site of einkorn wheat domestication identified by DNA fingerprinting.
Science, 278, 1312–1314.
Hidalgo, A. and Brandolini, A. (2014) Nutritional properties of einkorn wheat (Triticum
monococcum L.). J. Sci. Food Agric., 94, 601–612.
Hidalgo, A., Brandolini, A., Pompei, C. and Piscozzi, R. (2006) Carotenoids and tocols of
einkorn wheat (Triticum monococcum ssp. monococcum L.). J. Cereal Sci., 44, 182–193.
Hirase, K. and Molin, W.T. (2003) Sulfur assimilation in plants and weed control: Potential
targets for novel herbicides and action sites of certain safeners. Weed Biol. Manag., 3,
147–157.
Hirel, B. (2013) Améliorer l’efficacité d’utilisation de l’azote chez les plantes cultivées, INRA-
Versailles.
Hirel, B., LeGouis, J., Ney, B. and Gallais, A. (2007) The challenge of improving nitrogen
use efficiency in crop plants: towards a more central role for genetic variability and
quantitative genetics within integrated approaches. J. Exp. Bot., 58, 2369–2387.
Hirel, B., Tétu, T., Lea, P.J. and Dubois, F. (2011) Improving nitrogen use efficiency in crops
for sustainable agriculture. Sustainability, 3, 1452–1485.
Hodge, A., Robinson, D. and Fitter, A. (2000) Are microorganisms more effective than plants
at competing for nitrogen? Trends Plant Sci., 5, 304–308.
Jaiswal, D.K., Ray, D., Choudhary, M.K., et al. (2013) Comparative proteomics of
dehydration response in the rice nucleus: new insights into the molecular basis of
genotype-specific adaptation. Proteomics, 13, 3478–97.
Jamal, A., Moon, Y.S. and Abdin, M.Z. (2010) Sulphur -a general overview and interaction
with nitrogen. Aust. J. Crop Sci., 4, 523–529.
Jarillo, J.A., Pineiro, M., Cubas, P. and Martinez-Zapater, J.M. (2009) Chromatin
remodeling in plant development. Int. J. Dev. Biol., 53, 1581–1596.
Kataoka, T. (2004) Vacuolar sulfate transporters are essential determinants controlling internal
distribution of sulfate in Arabidopsis. Pant Cell, 16, 2693–2704.
Khan, M.M.K. and Komatsu, S. (2004) Rice proteomics: recent developments and analysis
of nuclear proteins. Phytochemistry, 65, 1671–81.

37
 Chapitre 1 : Synthèse bibliographique

Kichey, T., Hirel, B., Heumez, E., Dubois, F. and LeGouis, J. (2007) In winter wheat
(Triticum aestivum L.), post-anthesis nitrogen uptake and remobilisation to the grain
correlates with agronomic traits and nitrogen physiological markers. F. Crop. Res., 102,
22–32.
Kopriva, S. and Rennenberg, H. (2004) Control of sulphate assimilation and glutathione
synthesis: interaction with N and C metabolism. J. Exp. Bot., 55, 1831–1842.
Koprivova, A., Suter, M., Camp, R.O. Den, Brunold, C. and Kopriva, S. (2000) Regulation
of sulfate assimilation by nitrogen in Arabidopsis. Plant Physiol., 122, 737–746.
Kressler, D., Bange, G., Ogawa, Y., et al. (2012) Synchronizing nuclear import of ribosomal
proteins with ribosome assembly. Science, 338, 666–71.
Laudencia-Chingcuanco, D.L., Stamova, B.S., You, F.M., Lazo, G.R., Beckles, D.M. and
Anderson, O.D. (2007) Transcriptional profiling of wheat caryopsis development using
cDNA microarrays. Plant Mol. Biol., 63, 651–668.
Lee, Y.H., Tan, H.T. and Chung, M.C.M. (2010) Subcellular fractionation methods and
strategies for proteomics. Proteomics, 10, 3935–3956.
Lev-Yadun, S., Gopher, A. and Abbo, A. (2000) The cradle of agriculture. Science, 288,
1602–1603.
Li, G., Nallamilli, B.R.R., Tan, F. and Peng, Z. (2008) Removal of high-abundance proteins
for nuclear subproteome studies in rice (Oryza sativa) endosperm. Electrophoresis, 29,
604–17.
Lohmer, S., Maddaloni, M., Motto, M., Hartings, H., Salamini, F. and Thompson, R.D.
(1991) The maize regulatory locus Opaque-2 encodes a DNA-binding protein which
activates the transcription of the b-32 gene. EMBO, 10, 617–624.
Majoul-Haddad, T., Bancel, E., Martre, P., Triboi, E. and Branlard, G. (2013) Effect of
short heat shocks applied during grain development on wheat (Triticum aestivum L.) grain
proteome. J. Cereal Sci., 57, 486–495.
Malagoli, P., Laine, P., Rossato, L. and Ourry, A. (2005) Dynamics of nitrogen uptake and
mobilization in field-grown winter oilseed rape (Brassica napus) from stem extension to
harvest. II. An 15N-labelling-based simulation model of N partitioning between vegetative
and reproductive tissues. Ann. Bot., 95, 1187–1198.
Malik, a. H. (2009) Nutrient uptake, transport and translocation in cereals: influences of
environmental and farming conditions. Introd. Pap. Fac. Landsc. Planning, Hortic. Agric.
Sci., 1, 1–46.
Marcussen, T., Sandve, S.R., Heier, L., et al. (2014) Ancient hybridizations among the
ancestral genomes of bread wheat. Science, 345, 1250092–1250092.
Mares, D., Norstog, K. and Stone, B. (1975) Early stages in the development of wheat
endosperm. I. The change from free nuclear to cellular endosperm. Aust. J. Bot., 23, 311.
Martre, P., Porter, J.R., Jamieson, P.D. and Triboï, E. (2003) Modeling grain nitrogen
accumulation and protein composition to understand the sink/source regulations of
nitrogen remobilization for wheat. Plant Physiol., 133, 1959–1967.
Masclaux-Daubresse, C., Daniel-Vedele, F., Dechorgnat, J., Chardon, F., Gaufichon, L.
and Suzuki, A. (2010) Nitrogen uptake, assimilation and remobilization in plants:
challenges for sustainable and productive agriculture. Ann. Bot., 105, 1141–1157.
Matson, P., Lohse, K.A. and Hall, S.J. (2002) The globalization of nitrogen deposition:
consequences for terrestrial ecosystems. AMBIO A J. Hum. Environ., 31, 113.
Mehrotra, R., Kumar, S., Mehrotra, S. and Singh, B.D. (2009) Seed storage protein gene
regulation- A jig-saw puzzle. Indian J. Biotechnol., 8, 147–158.
Meier, I. and Brkljacic, J. (2009) The nuclear pore and plant development. Curr. Opin. Plant
Biol., 12, 87–95.
Merkle, T. (2011) Nucleo-cytoplasmic transport of proteins and RNA in plants. Plant Cell

38
 Chapitre 1 : Synthèse bibliographique

Rep., 30, 153–176.

Metakovsky, E. V., Chernakov, V.M., Upelniek, V.P., Redaelli, R., Dardevet, M.,
Branlard, G. and Pogna, N.E. (1996) Recombination mapping of minor omega-gliadin-
coding loci on chromosome 1A of common wheat: A revision. J. Genet. Breed., 50, 277–
286.
Michael Beman, J., Arrigo, K.R. and Matson, P.A. (2005) Agricultural runoff fuels large
phytoplankton blooms in vulnerable areas of the ocean. Nature, 434, 211–214.
Miflin, B.J. and Habash, D.Z. (2002) The role of glutamine synthetase and glutamate
dehydrogenase in nitrogen assimilation and possibilities for improvement in the nitrogen
utilization of crops. J. Exp. Bot., 53, 979–987.
Miura, K. (2013) Nitrogen and phosphorus nutrition under salinity stress. In Ecophysiology
and Responses of Plants under Salt Stress. New York, NY: Springer New York, pp. 425–
441.
Mochida, K. and Shinozaki, K. (2011) Advances in omics and bioinformatics tools for
systems analyses of plant functions. Plant Cell Physiol., 52, 2017–2038.
Monaghan, J.M., Scrimgeour, C.M., Stein, W.M., Zhao, F.J. and Evans, E.J. (1999)
Sulphur accumulation and redistribution in wheat (Triticum aestivum): a study using stable
sulphur isotope ratios as a tracer system. Plant, Cell Environ., 22, 831–839.
Monaghan, J.M., Snape, J.W., Chojecki, A.J.S. and Kettlewell, P.S. (2001) The use of grain
protein deviation for identifying wheat cultivars with high grain protein concentration and
yield. Euphytica, 122, 309–317.
Monfreda, C., Ramankutty, N. and Foley, J.A. (2008) Farming the planet: 2. Geographic
distribution of crop areas, yields, physiological types, and net primary production in the
year 2000. Global Biogeochem. Cycles, 22, 1–19.
Moreno-Risueno, M.A., Busch, W. and Benfey, P.N. (2010) Omics meet networks—using
systems approaches to infer regulatory networks in plants. Curr. Opin. Plant Biol., 13,
126–131.
Moreno-Risueno, M.Á., González, N., Díaz, I., Parcy, F., Carbonero, P. and Vicente-
Carbajosa, J. (2008) FUSCA3 from barley unveils a common transcriptional regulation
of seed-specific genes between cereals and Arabidopsis. Plant J., 53, 882–894.
Muller, M. and Knudsen, S. (1993) The nitrogen response of a barley C-hordein promoter is
controlled by positive and negative regulation of the GCN4 and endosperm box. Plant J.,
4, 343–355.
Narula, K., Datta, A., Chakraborty, N. and Chakraborty, S. (2013) Comparative analyses
of nuclear proteome: extending its function. Front. Plant Sci., 4, 100.
Nicolas, M.E., Gleadow, R.M. and Dalling, M.J. (1984) Effects of drought and high
temperature on grain growth in wheat. Aust. J. Plant Physiol., 11, 553.
Noctor, G., Gomez, L., Vanacker, H. and Foyer, C.H. (2002) Interactions between
biosynthesis, compartmentation and transport in the control of glutathione homeostasis
and signalling. J. Exp. Bot., 53, 1283–1304.
Nuttall, J.G., Leary, G.J.O., Panozzo, J.F., Walker, C.K., Barlow, K.M. and Fitzgerald,
G.J. (2016) Models of grain quality in wheat—A review. F. Crop. Res.
Oliver, S. (2000) Guilt-by-association goes global. Nature, 403, 601–603.
Osborne, T.B. (1907) The proteins of the wheat kernel. Carnegie Inst. Washingt., 84, 1–119.
Osborne, T.B. (1909) The vegetable proteins. Longmans, Green Co., 1, 1–125.
Palta, J. and Fillery, I. (1995) N application increases pre-anthesis contribution of dry matter
to grain yield in wheat grown on a duplex soil. Aust. J. Agric. Res., 46, 519–531.
Pandey, A., Chakraborty, S., Datta, A. and Chakraborty, N. (2008) Proteomics approach
to identify dehydration responsive nuclear proteins from chickpea (Cicer arietinum L.).
Mol. Cell. Proteomics, 7, 88–107.

39
 Chapitre 1 : Synthèse bibliographique

Payne, P.I., Holt, L.M., Jackson, E. a., Law, C.N. and Damania, a. B. (1984) Wheat storage
proteins: their genetics and their potential for manipulation by plant breeding. Philos.
Trans. R. Soc. B Biol. Sci., 304, 359–371.
Payne, P.I., Law, C.N. and Mudd, E.E. (1980) Control by homoeologous group 1
chromosomes of the high-molecular-weight subunits of glutenin, a major protein of wheat
endosperm. Theor. Appl. Genet., 58, 113–120.
Pendle, A.F., Clark, G.P., Boon, R., et al. (2005) Proteomic analysis of the Arabidopsis
nucleolus suggests novel nucleolar functions. , 16, 260–269.
Petrovská, B., Jeřábková, H., Chamrád, I., Vrána, J., Lenobel, R., Uřinovská, J., Sebela,
M. and Doležel, J. (2014) Proteomic analysis of barley cell nuclei purified by flow sorting.
Cytogenet. Genome Res., 143, 78–86.
Petrovská, B., Šebela, M. and Doležel, J. (2015) Inside a plant nucleus: discovering the
proteins. J. Exp. Bot., 66, 1627–1640.
Pomeranz, Y. (1988) Chemical composition of kernel structures. Wheat Chem. Technol., 2,
97–158.
Provero, P. (2002) Gene networks from DNA microarray data: centrality and lethality, Turin.
Rahman, S., Shewry, P.R., Forde, B.G., Kreis, M. and Miflin, B.J. (1983) Nutritional
control of storage protein synthesis in developing grain of barley (Hordeum vulgare L.).
Planta, 159, 366–372.
Randall, P., Spencer, K. and Freney, J. (1981) Sulfur and nitrogen fertilizer effects on wheat.
I. Concentrations of sulfur and nitrogen and the nitrogen to sulfur ratio in grain, in relation
to the yield response. Aust. J. Agric. Res., 32, 203.
Ravel, C., Fiquet, S., Boudet, J., Dardevet, M., Vincent, J., Merlino, M., Michard, R. and
Martre, P. (2014) Conserved cis-regulatory modules in promoters of genes encoding
wheat high-molecular-weight glutenin subunits. Front. Plant Sci., 5, 621.
Remans, T., Nacry, P., Pervent, M., Filleur, S., Diatloff, E., Mounier, E., Tillard, P., Forde,
B.G. and Gojon, A. (2006) The Arabidopsis NRT1.1 transporter participates in the
signaling pathway triggering root colonization of nitrate-rich patches. Proc. Natl. Acad.
Sci., 103, 19206–19211.
Repetto, O., Rogniaux, H., Firnhaber, C., Zuber, H., Küster, H., Larré, C., Thompson, R.
and Gallardo, K. (2008) Exploring the nuclear proteome of Medicago truncatula at the
switch towards seed filling. Plant J., 56, 398–410.
Repetto, O., Rogniaux, H., Larré, C., Thompson, R. and Gallardo, K. (2012) The seed
nuclear proteome. Front. Plant Sci., 3, 289.
Reussi, N., Echeverría, H. and Rozas, H.S. (2011) Diagnosing sulfur deficiency in spring red
wheat: plant analysis. J. Plant Nutr., 34, 573–589.
Reuveny, Z., Dougall, D.K. and Trinity, P.M. (1980) Regulatory coupling of nitrate and
sulfate assimilation pathways in cultured tobacco cells. Proc. Natl. Acad. Sci. U. S. A., 77,
6670–6672.
Rubio-Somoza, I., Martinez, M., Abraham, Z., Diaz, I. and Carbonero, P. (2006) Ternary
complex formation between HvMYBS3 and other factors involved in transcriptional
control in barley seeds. Plant J., 47, 269–281.
Rubio-Somoza, I., Martinez, M., Diaz, I. and Carbonero, P. (2006) HvMCB1, a R1MYB
transcription factor from barley with antagonistic regulatory functions during seed
development and germination. Plant J., 45, 17–30.
Sabelli, P.A. and Larkins, B.A. (2009) The development of endosperm in grasses. Plant
Physiol., 149, 14–26.
Scherer, H.W. (2001) Sulphur in crop production. Eur. J. Agron., 14, 81–111.
Schnyder, H. and Baum, U. (1992) Growth of the grain of wheat (Triticum aestivum L.). The
relationship between water content and dry matter accumulation. Eur. J. Agron., 1, 51–57.

40
 Chapitre 1 : Synthèse bibliographique

Serin, E.A.R., Nijveen, H., Hilhorst, H.W.M. and Ligterink, W. (2016) Learning from co-
expression networks: possibilities and challenges. Front. Plant Sci., 7, 444.
Shewry, P.R. (2009) Wheat. J. Exp. Bot., 60, 1537–53.
Shewry, P.R. and Halford, N.G. (2003) Genetics of wheat gluten proteins. Adv. Genet., 49,
111–184.
Shewry, P.R. and Hey, S.J. (2015) The contribution of wheat to human diet and health. Food
Energy Secur., 4, 178–202.
Shewry, P.R., Mitchell, R.A.C., Tosi, P., et al. (2012) An integrated study of grain
development of wheat (cv. Hereward). J. Cereal Sci., 56, 21–30.
Shewry, P.R., Tatham, A.S., Forde, J., Kreis, M. and Miflin, B.J. (1986) The classification
and nomenclature of wheat gluten proteins: A reassessment. J. Cereal Sci., 4, 97–106.
Shewry, P.R., Tatham, A.S. and Halford, N.G. (2001) Nutritional control of storage protein
synthesis in developing grain of wheat and barley. Plant Growth Regul., 34, 105–111.
Shewry, P.R., Tatham, A.S. and Lazzeri, P. (1997) Biotechnology of wheat quality. J. Sci.
Food Agric., 73, 397–406.
Silva, A.T., Ribone, P.A., Chan, R.L., Ligterink, W. and Hilhorst, H.W.M. (2016) A
predictive coexpression network identifies novel genes controlling the seed-to-seedling
phase transition in Arabidopsis thaliana. Plant Physiol., 170, 2218–2231.
Simmonds, N.W. (1995) The relation between yield and protein in cereal grain. J. Sci. Food
Agric., 67, 309–315.
Skylas, D.J., Cordwell, S.J., Hains, P.G., et al. (2002) Heat shock of wheat during grain
filling: proteins associated with heat-tolerance. J. Cereal Sci., 35, 175–188.
Smoot, M.E., Ono, K., Ruscheinski, J., Wang, P.-L. and Ideker, T. (2011) Cytoscape 2.8:
new features for data integration and network visualization. Bioinformatics, 27, 431–432.
Subba, P., Kumar, R., Gayali, S., Shekhar, S., Parveen, S., Pandey, A., Datta, A.,
Chakraborty, S. and Chakraborty, N. (2013) Characterisation of the nuclear proteome
of a dehydration-sensitive cultivar of chickpea and comparative proteomic analysis with a
tolerant cultivar. Proteomics, 13, 1973–92.
Surget, A. and Barron, C. (2005) Histologie du grain de bté. Ind. des céréales, 145, 3–7.
Tamura, K. and Hara-Nishimura, I. (2014) Functional insights of nucleocytoplasmic
transport in plants. Front. Plant Sci., 5, 118.
Tamura, K. and Hara-Nishimura, I. (2013) The molecular architecture of the plant nuclear
pore complex. J. Exp. Bot., 64, 823–832.
Tan, F., Li, G., Chitteti, B.R. and Peng, Z. (2007) Proteome and phosphoproteome analysis
of chromatin associated proteins in rice (Oryza sativa). Proteomics, 7, 4511–27.
Taylor-Teeples, M., Lin, L., Lucas, M. de, et al. (2014) An Arabidopsis gene regulatory
network for secondary cell wall synthesis. Nature, 517, 571–575.
Triboï, E., Martre, P. and Triboï-Blondel, A.-M. (2003) Environmentally-induced changes
in protein composition in developing grains of wheat are related to changes in total protein
content. J. Exp. Bot., 54, 1731–1742.
Urano, K., Kurihara, Y., Seki, M. and Shinozaki, K. (2010) “Omics” analyses of regulatory
networks in plant abiotic stress responses. Curr. Opin. Plant Biol., 13, 132–138.
Usadel, B., Obayashi, T., Mutwil, M., et al. (2009) Co-expression tools for plant biology:
opportunities for hypothesis generation and caveats. Plant. Cell Environ., 32, 1633–1651.
USDA (2016) Grain: World Markets and Trade TMT.
Vensel, W.H., Tanaka, C.K., Cai, N., Wong, J.H., Buchanan, B.B. and Hurkman, W.J.
(2005) Developmental changes in the metabolic protein profiles of wheat endosperm.
Proteomics, 5, 1594–611.
Verdier, J. and Thompson, R.D. (2008) Transcriptional regulation of storage protein
synthesis during dicotyledon seed filling. Plant Cell Physiol., 49, 1263–1271.

41
 Chapitre 1 : Synthèse bibliographique

Vincent, J., Martre, P., Gouriou, B., Ravel, C., Dai, Z., Petit, J.-M. and Pailloux, M. (2015)
RulNet: A web-oriented platform for regulatory network inference, application to wheat –
omics data. PLoS One, 10, e0127127.
Wieser, H., Gutser, R. and Tucher, S. von (2004) Influence of sulphur fertilisation on
quantities and proportions of gluten protein types in wheat flour. J. Cereal Sci., 40, 239–
244.
Wieser, H. and Seilmeier, W. (1998) The influence of nitrogen fertilisation on quantities and
proportions of different protein types in wheat flour. J. Sci. Food Agric., 76, 49–55.
Wrigley, C., Cros, D., Archer, M., Downie, P. and Roxburgh, C. (1980) The sulfur content
of wheat endosperm proteins and its relevance to grain quality. Aust. J. Plant Physiol., 7,
755–766.
Xi, D.-M. and Zheng, C.-C. (2011) Transcriptional regulation of seed storage protein genes in
Arabidopsis and cereals. Seed Sci. Res., 21, 247–254.
Yamaguchi, Y., Nakamura, T., Harada, E., Koizumi, N. and Sano, H. (1999) Differential
accumulation of transcripts encoding sulfur assimilation enzymes upon sulfur and/or
nitrogen deprivation in Arabidopsis thaliana. Biosci. Biotechnol. Biochem., 63, 762–766.
Young, T.E. and Gallie, D.R. (2000) Programmed cell death during animal development.
Plant Mol. Biol., 44, 283–301.
Zhao, F., Evans, E.J., Bilsborrow, P.E. and Syers, J.K. (1993) Influence of sulphur and
nitrogen on seed yield and quality of low glucosinolate oilseed rape (Brassica napus L).
J. Sci. Food Agric., 63, 29–37.
Zhao, F., Hawkesford, M. and McGrath, S. (1999) Sulphur assimilation and effects on yield
and quality of wheat. J. Cereal Sci., 30, 1–17.
Zörb, C., Grover, C., Steinfurth, D. and Hermann Mühling, K. (2010) Quantitative
proteome analysis of wheat gluten as influenced by N and S nutrition. Plant Soil, 327,
225–234.

42
 CHAPITRE 2 :
Validation de la méthode
d’extraction des protéines
nucléaires à partir de grains de blé

43
44
 Chapitre 2 : Validation de la méthode d’extraction des protéines nucléaires

1. INTRODUCTION
Pour comprendre les mécanismes impliqués dans la régulation du développement du
grain et l’accumulation de ses réserves, l’étude des protéines nucléaires est apparue nécessaire.
Pour cela, nous avons mis au point chez le blé une méthode nous permettant d’accéder à ce type
de protéines généralement sous-représentées dans les analyses de protéomique. Parmi les études
menées chez les plantes, la méthode la plus couramment utilisée repose sur la centrifugation de
l’extrait cellulaire dans un gradient de densité pour purifier les noyaux. Dans le grain, la
principale difficulté pour la purification d’organites réside dans la présence en abondance des
composés de réserves. Ceci pourrait expliquer pourquoi peu d’études se sont intéressées au
protéome nucléaire du grain (Repetto et al., 2012).
La procédure d’obtention d’extraits de protéines nucléaires utilisée au cours de cette thèse
a été mise au point par Emmanuelle Bancel avant mon arrivée au sein de l’équipe BIG. Cette
méthode est composée de deux étapes principales : 1/ la purification de noyaux à partir de grains
de blé, 2/ l’extraction des protéines des noyaux.
Pour décrire brièvement cette procédure, les grains sont tout d’abord broyés en milieu
liquide dans un tampon d’extraction, puis l’homogénat est filtré afin d’éliminer les débris
cellulaires. Les membranes cellulaires sont ensuite lysées par l’ajout d’un détergent, le Triton
X-100, et les noyaux sont purifiés sur un gradient discontinu 30%/80% de Percoll, où ils
s’accumulent à l’interface des deux solutions de densité différentes. Les protéines sont extraites
de ces noyaux par l’utilisation du TriReagent, qui est une solution monophasique de phénol et
de guanidine isothiocyanate, permettant la séparation de l’extrait nucléaire en trois phases : la
phase aqueuse contenant les ARN, l’interphase contenant l’ADN, et la phase organique qui
contient les protéines nucléaires. Les protéines nucléaires extraites peuvent alors être analysées
en spectrométrie de masse, avec ou sans séparation préalable sur gel mono- ou bi-dimensionnel.
Afin de valider la méthode, plusieurs expérimentations auxquelles j’ai participé ont été
entreprises. Ainsi, des colorations de noyaux au Hoechst®, des Western blot utilisant des
anticorps dirigés contre des protéines marqueurs des différents compartiments subcellulaires,
puis une identification des protéines par spectrométrie de masse ont été réalisés. Ces étapes de
validation ont été réalisées pour des grains de T. aestivum et [Link], les deux espèces
de blé utilisées au cours de la thèse, récoltés à deux temps thermiques, 150 et 250°Cj après
floraison. Pour l’identification des protéines, nous avons fait le choix de les analyser en
spectrométrie de masse à l’issue de leur extraction, sans séparation sur gel.

Références :
Repetto, O., Rogniaux, H., Larré, C., Thompson, R. and Gallardo, K. (2012) The seed
nuclear proteome. Front. Plant Sci., 3, 289.

45
 Chapitre 2 : Validation de la méthode d’extraction des protéines nucléaires

2. ARTICLE 1 : Proteomic approach to identify nuclear proteins in wheat grain

Emmanuelle Bancel1,2, Titouan Bonnot1,2, Marlène Davanture3, Gérard Branlard1,2,
Michel Zivy3, and Pierre Martre1,2
1
INRA, UMR1095 Génétique, Diversité et Ecophysiologie des Céréales, 5 chemin de Beaulieu,
F-63 039 Clermont-Ferrand, France
2
Université Blaise Pascal, UMR1095 Génétique, Diversité et Ecophysiologie des Céréales,
Avenue des Landais, F-63 170 Aubière, France
3
CNRS, PAPPSO, UMR 0320/8120 Génétique Quantitative et É volution - Le Moulon, F-91190
Gif-sur-Yvette, France

46
 Technical Note

[Link]/jpr

Proteomic Approach to Identify Nuclear Proteins in Wheat Grain

Emmanuelle Bancel,*,†,‡,∥ Titouan Bonnot,†,‡,∥ Marlène Davanture,§ Gérard Branlard,†,‡ Michel Zivy,§
and Pierre Martre†,‡,⊥
†
INRA, UMR1095 Genetics, Diversity and Ecophysiology of Cereals, 5 chemin de Beaulieu, F-63 039 Clermont-Ferrand, France
‡
Blaise Pascal University, UMR1095 Genetics, Diversity and Ecophysiology of Cereals, Avenue des Landais, F-63 170 Aubière, France
§
CNRS, PAPPSO, UMR 0320/8120 Génétique Quantitative et Évolution - Le Moulon, F-91190 Gif-sur-Yvette, France
*
S Supporting Information

ABSTRACT: The nuclear proteome of the grain of the two

cultivated wheat species Triticum aestivum (hexaploid wheat;
genomes A, B, and D) and T. monococcum (diploid wheat;
genome A) was analyzed in two early stages of development
using shotgun-based proteomics. A procedure was optimized
to purify nuclei, and an improved protein sample preparation
was developed to efficiently remove nonprotein substances
(starch and nucleic acids). A total of 797 proteins
corresponding to 528 unique proteins were identified, 36%
of which were classified in functional groups related to DNA
and RNA metabolism. A large number (107 proteins) of
unknown functions and hypothetical proteins were also found.
Some identified proteins may be multifunctional and may present multiple localizations. On the basis of the MS/MS analysis, 368
proteins were present in the two species, and in two stages of development, some qualitative differences between species and
stages of development were also found. All of these data illustrate the dynamic function of the grain nucleus in the early stages of
development.
KEYWORDS: cereal, grain development, bread wheat (Triticum aestivum), einkorn wheat (Triticum monococcum), LC−MS/MS,
nuclear proteome

■ INTRODUCTION
Over the past 10 years, several studies have focused on cereal
making difficult the isolation of pure nuclei, and (ii) numerous
compounds, particularly carbon (starch or oil) and nitrogen
grain proteome. For wheat, both global1−3 and targeted (storage proteins) storage components and nucleic acids, may
studies4−7 were developed; however, little is known about interfere during protein extraction and proteomic analysis.
subcellular proteome like nuclear proteome. Nuclear proteins Therefore, optimizations are necessary to obtain high-purity
are predicted to comprise ∼20% of the total cellular proteins in nuclei in all developmental stages, which differ in the number of
yeast and animals, but their amount in plants is not yet known.8 contaminants (e.g., protein bodies, starch granules).
Several nuclear proteomics investigations have been performed The purpose of the present note is first to report on an
in tissues and organs other than grains like leaves,9 cell extraction protocol allowing nuclear proteome analysis to be
suspensions10,11 or seedlings.12 Some studies have concerned performed on developing cereal grains, with the specific
whole seeds for Medicago truncatula13 or endosperm for Oryza objective of obtaining nuclei in sufficient yield and purity for
sativa14 or Zea mays.15 Nuclear proteins are often under- proteomic approach. Because our aim is to use this protocol in
represented in proteomic studies due to the frequently low further studies to analyze the response of the nuclear proteome
abundance of proteins involved in regulatory processes.16,17 for a wide range of genetic and environmental contexts, the
Knowledge of the changes in nuclear proteome that occur method has to produce a low number of protein fractions and
during grain development is needed to improve our under- to be adapted with shotgun proteomics. The second objective is
standing of the specific mechanisms and involved in endosperm to provide a qualitative overview of the nuclear proteome
cell regulation. Indeed to understand the mechanism associated revealed through gel-free mass spectrometry analysis during the
with grain development, it is of prime importance to know early phases of endosperm cells division and cells differ-
which nuclear actors are present and to develop a proteomic entiation. These early phases of grain development determine
approach allowing qualitative and quantitative studies; however, the potential final size of the grain,18 and they are particularly
the nuclear proteome of grains and of wheat, in particular, has sensitive to the growing temperature19 and to soil water
been far less studied. Two major difficulties may explain why
cereal grain nuclear proteome was less studied: (i) in the Received: May 21, 2015
endosperm, nuclei are entrapped in protein and starch matrix, Published: July 31, 2015

© 2015 American Chemical Society 4432 DOI: 10.1021/[Link].5b00446

J. Proteome Res. 2015, 14, 4432−4439

47
Journal of Proteome Research Technical Note

Figure 1. Workflow with the main steps used to purify nuclei and extract nuclear proteins from wheat grain.

deficit.20 The methodology was validated for the cultivated were grown under natural soil conditions, fertilized, watered as
hexaploid wheat species Triticum aestivum (genomes A, B, and usual, and protected against fungi. Ears were tagged at the date
D) and its domesticated diploid progenitor T. monococcum of anthesis, and only grains from the middle of the tagged ears
(genome A). were harvested. Air temperature close to the ears was recorded,

■ METHODS
Plant Material and Tissue Sampling
and daily mean air temperature was calculated, and the sum of
daily mean temperatures was used to follow the grain
development in thermal time (degree-days, °Cd). Wheat grains
were harvested at two different thermal times after anthesis
The bread wheat (Triticum aestivum) cultivar Récital and the corresponding to the endosperm cells division (150 °Cd after
einkorn wheat (Triticum monococcum) accession FR35821 - anthesis) and cell differentiation stages (250 °Cd after
K48993 were cultivated in an environmentally controlled anthesis). For each developmental stage, grains from the
growth chamber at INRA, Clermont-Ferrand, France. Plants middle third parts of the ears were quickly harvested, weighed,
4433 DOI: 10.1021/[Link].5b00446
J. Proteome Res. 2015, 14, 4432−4439

48
Journal of Proteome Research Technical Note

frozen in liquid nitrogen, and then stored at −80 °C until use. SDS-PAGE Analysis
To validate the technical and methodological approach and for A fraction (1 mL) of each supernatant (S1−S5) was
protein quantification, we used at least four biological precipitated with TCA (w/v) 10% in cold acetone and proteins
replicates. present in the pellet were solubilized in SDS-PAGE loading
Isolation of Grain Nuclei buffer (45 mM Tris-HCl, pH 6.8, 50 mM DTT, 1% SDS, 10%
A method (Figure 1) was adapted from ref 21. All steps were glycerol, 0.001% bromophenol blue) as well as nuclear proteins
performed at 4 °C. Grains (2 g) were put into ice-cold round- pellet to check the different supernatants during the washing
bottomed centrifuge tubes immersed in 6 mL of extraction steps. Proteins were loaded at constant volume and separated
buffer (20 mM Hepes-KOH (pH 7.0), 5 mM MgCl2, 10 mM 2- by SDS-PAGE in a 12.5% polyacrylamide gel. Electrophoresis
mercaptoethanol, 0.5 mM PMSF, and 0.1% (v/v) phosphatase was performed at 70 V for 0.3 h and then at 200 V for 1 h using
inhibitors cocktail (Sigma-Aldrich) added freshly). Tissues were a Mini-Protean II gel electrophoresis system (Bio-Rad). The
disrupted with a polytron homogenizer (Kinematica POLY- gels were stained with 0.2% Coomassie brilliant blue R250
TRON PT 10), and the resultant homogenate was filtered (Sigma-Aldrich) for 1 h at room temperature and subsequently
through two layers of Miracloth (Calbiochem, pore size 25 μm) destained in 10% (v/v) acetic acid and 30% (v/v) ethanol.
to remove cell debris. Membranes were then lyzed by adding Protein concentration in nuclear extracts was determined by
Triton X-100 to the homogenate (adjusted to 12 mL) at a final Bradford assay with bovine serum albumin as standard.22
concentration of 0.5% (v/v). After centrifugation (1000g for 5 Assessment of the Purity of the Nuclei Fractions
min), the resultant supernatant adjusted to 12 mL was kept For immunoblot analysis, protein samples corresponding to
(S1), and the pellet was washed four times with 1 mL of nuclear proteins and to proteins from supernatants collected
extraction buffer, followed by centrifugation (1000g for 5 min; during nuclei enrichment were separated on a 12.5% SDS−
S2−S5) to eliminate contaminants (mainly starch and storage PAGE gel and then transferred onto a nitrocellulose membrane
proteins) and obtain semipure nuclei. These nuclei were (Hybond, ECL, GE Healthcare) using a Hoefer TE77 semidry
thoroughly suspended in 1 mL of extraction buffer. transfer blotter. The blotted membrane was incubated for 1 h at
The resulting suspension was layered in a 16 mL tube on top room temperature in blocking buffer consisting of 10 mM
of a Percoll gradient consisting of 10 mL of a 30% (v/v) Percoll Tris−HCl (pH 7.6), 150 mM NaCl, 0.01% (v/v) Tween 20,
solution and 2.5 mL of a 80% (v/v) Percoll solution. After and 5% (w/v) skim milk. After blocking, the membrane was
centrifugation (930g for 30 min), nuclei were collected at the incubated 1 h at room temperature and then overnight at 4 °C
interface of the 30 and 80% Percoll layers, washed twice with with a 1:1000 dilution of antihistone H3 (Abcam) antibody,
extraction buffer, and recovered as a pellet after centrifugation antivacuolar ATPase (Agrisera AB) antibody, antiphotosystem
(3500g for 5 min). The nuclei pellet was suspended in II reaction center protein D1 (Agrisera AB), or with a 1:2500
phosphate-buffered saline, and an aliquot was stained in a 0.1 dilution of anti-UDP-glucose pyrophosphorylase (Agrisera AB)
μg mL−1 Hoechst fluorescent dye solution and observed under and anti-H protein of glycine decarboxylase complex (GDC;
fluorescence microscopy (Zeiss Axioplan) to evaluate the Agrisera AB). Antirabbit IgG conjugated with horseradish
nuclear enrichment and integrity. peroxidase (GE Healthcare) was used as the secondary
Extraction of Nuclear Proteins antibody (1:100 000). After 1 h of incubation with the
The nuclei-containing pellets were resuspended in TRI Reagent secondary antibody, signals were detected using an ECL plus
(Sigma-Aldrich) according to the manufacturer’s instruction. Western blotting detection kit (GE Healthcare) following the
This product, a mixture of guanidine thiocyanate and phenol in manufacturer’s protocol.
a monophase solution, effectively dissolves DNA, RNA, and Trypsin Digestion of Nuclear Proteins
protein. After adding 1-bromo 3-chloropropane and centrifug- For digestion, a protein sample for each species (Triticum
ing, the mixture separates into three phases: an aqueous phase aestivum and Triticum monococcum) in each developmental
containing the RNA, the interphase containing DNA, and an stage (150 and 250 °Cd after anthesis) was analyzed.
organic phase containing proteins. Each component can then Precipitated proteins were suspended in 30 μL of buffer A
be isolated after separating the phases (Figure 1). In brief, (0.1% ZALS I, 6 M urea, 2 M thiourea, 10m M DTT, 30 mM
nuclei pellet was resuspended in 500 μL of TRI Reagent and Tris-HCl pH 8.8, 5 mM NH4HCO3). Protein concentration
incubated for 5 min at room temperature. Then, 50 μL of was measured using the 2-D Quant Kit (GE Healthcare,
bromo-3-chloropropane was added, and the mixture was Piscataway, NJ) with BSA as a standard. All samples were
vortexed for 15 s and incubated for 15 min at room adjusted to a single concentration of 5 μg/μL by adding buffer
temperature. After centrifugation (12 000g for 15 min at 4 A. Then, 40 μg of protein of each sample was incubated with
°C), the upper phase corresponding to RNA was removed. To iodoacetamide (40 mM final concentration) in the dark for 1 h.
precipitate DNA, 150 μL of 95% (v/v) ethanol was added. They were then diluted 8-fold with 50 mM NH4HCO3.
After mixing by inversion and incubation for 3 min at room Samples were digested in-solution overnight at 37 °C by adding
temperature, the mixture was centrifuged (2000g for 5 min at 4 800 ng of trypsin in 50 mM NH4HCO3 (trypsin/protein ratio
°C) and the supernatant was collected. Finally, proteins were 1/50). Trypsic digestion was stopped by the addition of TFA
precipitated by adding 750 μL of 2-propanol, incubating for 10 10%.
min at room temperature and centrifuging (12 000g for 10 min
at 4 °C). Supernatant was removed and protein pellet was LC−MS/MS Analysis
washed three times in 1 mL of 0.3 M guanidine hydrochloride/ Online liquid chromatography was performed on a NanoLC
95% ethanol solution by stirring for 20 min, followed by Ultra system (Eksigent). A 1 μg sample was loaded at 7.5 μL
centrifugation (7500g for 5 min at 4 °C). The protein pellet min−1 on a precolumn (stationary phase: C18, particles of 5
was then washed with 1 mL of 95% (v/v) ethanol. The final μm; column: 100 μm i.d., 2 cm length; NanoSeparations) and
protein pellet was dried and then stored at −20 °C. desalted with 0.1% (v/v) formic acid in 2% (v/v) ACN. After 3
4434 DOI: 10.1021/[Link].5b00446
J. Proteome Res. 2015, 14, 4432−4439

49
Journal of Proteome Research Technical Note

Figure 2. Assessment of the purity of nuclei preparation and nuclear protein extract (N) in Triticum aestivum (left) and Triticum monococcum (right).
(A) Purified nuclei fraction was stained with Hoechst and visualized by fluorescence microscopy. (B) Analytical 1-D electrophoresis profile (12.5%
SDS-PAGE, Coomassie Blue staining) of protein content in different stages of nuclei enrichment (S1−S5, N). (C) (1) Immunoblot analysis with
antihistone H3, (2) antivacuolar ATPase, (3) antiphotosystem II reaction center protein D1, (4) anti-UDP-glucose pyrophosphorylase, and (5) anti-
H protein of glycine decarboxylase complex (GDC) antibodies.

min, the precolumn was connected to a separating C18 column applied to an uncoated capillary probe (10 μ i.d.; New
(stationary phase BIOSPHERE C18, particles of 3 μm; column Objective, Woburn, MA). Xcalibur 2.2 SP1 interface was used
75 μm i.d., 150 mm length; NanoSeparations). Buffers were
to monitor data-dependent acquisition of peptide ions. This
0.1% formic acid in water (solvent A) and 0.1% formic acid in
acetonitrile (solvent B). Peptide separation was achieved using included a full MS scan covering the 300 to 1400 mass-to-
a linear gradient from 5 to 30% B for 28 min at 300 nL min−1. charge ratio (m/z) with a resolution of 70 000 and a MS/MS
Including the regeneration step at 100% B and the equilibration step (normalized collision energy: 30%; resolution: 17 500).
step at 100% A, one run took 45 min. Eluted peptides were
The MS/MS step was reiterated for the eight major ions
analyzed with a Q-Exactive mass spectrometer (Thermo
Electron, Courtaboeuf, France) using a nanoelectrospray detected during the full MS scan. The dynamic exclusion was
interface. Ionization was performed with a 1.3 kV spray voltage set to 45 s.
4435 DOI: 10.1021/[Link].5b00446
J. Proteome Res. 2015, 14, 4432−4439

50
Journal of Proteome Research Technical Note

Figure 3. Graphical representation of the functional categories of proteins found in wheat grain nuclei-enriched fraction. (A) Venn diagrams analysis
of the 797 proteins identified for the two species (Triticum aestivum and Triticum monococcum) in the two developmental stages (150 and 250 °Cd
after anthesis). (B) Graphical representation of the functional categories of the 368 commun proteins for the two species (Triticum aestivum and
Triticum monococcum) in the two developmental stages (150 and 250 °Cd after anthesis) according to Gene ontology. The percentage of proteins in
each functional class is shown.

Identification of Nuclear Proteins Functional Classification of Nuclear Proteins and

Subcellular Localization
Protein identification was performed by searching peptides
against the Uniprot protein database version 2014_07 limited Functional classification was established on the basis of gene
ontology (GO) information rules23 provided by Uniprot.
to the Triticum genus using X!Tandem (version: Sledge- Subcellular localization of identified proteins was predicted by
hammer, 2013.09.01.1, [Link] in silico analysis using Multiloc2,24 NucPred,25 WolfPSort,26 Y
Enzymatic cleavage parameters were set as trypsin digestion loc,27 and LocTree28 programs. The top two hits were
with one possible miss cleavage. Cys carboxyamidomethylation considered for Multiloc and Y loc. Some proteins were also
detected and localized from previous studies in the literature.

■
was set as static modification, whereas Met oxidation, Nter
deamidation, and Nter acetylation were set as variable RESULTS AND DISCUSSION
modifications. Precursor mass tolerance was 10 ppm and
Purity of Nuclei Preparation
fragment mass tolerance was 0.02. Identified proteins were
filtered and grouped using X!TandemPipeline v3.3.1 (http:// The development of wheat grain involves three distinct phases:
cell division and differentiation, grain filling, and desiccation/
[Link]/bioinfo/xtandempipeline/). Peptide and protein
maturation. At 110 °Cd (6 days) after anthesis, nuclei are
e-value cut offs were set to 0.01 and 10−4, respectively, with at present in the central region of the endosperm that was
least two peptides per protein. Proteins with at least one occupied by the central vacuole because of cellularization of
peptide in common and similar function were defined as unique peripheral endosperm. All major cell types are differentiated
proteins. The presence of several proteins identified by the around 250 °Cd (13 days) after anthesis.29 The nuclei were
isolated from grains in two stages corresponding to this first
same sets of peptides in a group showed the redundancy due to phase of development. Different methods were used for
the approach itself giving a list of all proteins that might be assessing organelle intactness and purity. The Hoechst
present in the sample. fluorescence image obtained assesses the integrity of nuclei
4436 DOI: 10.1021/[Link].5b00446
J. Proteome Res. 2015, 14, 4432−4439

51
Journal of Proteome Research Technical Note

(Figure 2A). The nuclei had average diameter of approximately contaminants present in small amount in the nuclear extracts.
5−10 μm. Nevertheless, a nuclear protein belonging to the vicilin
The purity of isolated organelles can be compromised by superfamily was related to the protection of chromatin
copurifying subcellular fractions. Proteins from supernatants structure against desiccation during grain development in
(S1−S5) collected during nuclei purification were solubilized Pisum sativum,30 and in Medicago truncatula vicilin (7S globulin
and analyzed by SDS-PAGE. We extracted proteins from the family) and legumin (11S globulin family), storage proteins
nuclei-enriched fraction using the TRI Reagent extraction were found in grain nuclear extracts.13 The comparison of
method and in loading this nuclear extract (N) on SDS PAGE. identified proteins revealed a core of 368 proteins (46%) that
The average final yield of nuclear proteins extraction for T. were present in the four samples (Figure 3A). They were
aestivum was 160 and 110 μg g−1 fresh mater (FM) at 150 and principally related to DNA (32%), RNA (23%), and protein
250 °Cd after anthesis, respectively, and for T. monococcum was (18%) metabolisms (Figure 3B). These data assess the
120 and 150 μg g−1 FM at 150 and 250 °Cd after anthesis, robustness of the approach that enables us to detect a pool
respectively. of proteins involved in basal nuclear metabolism. Radar plots
The combination of microscopy and biochemical methods (Figure S2, Supporting Information) show similar distribution
was used to assess the purity of our nuclei and nuclear proteins of functional categories between the two stages for the two
preparations. Coomassie blue staining revealed the absence of species. For the two species, an increase in the percentage of
significant amounts of proteins in the last supernatant (S5) and proteins involved in DNA metabolism was observed at 250 °Cd
a typical nuclear protein pattern with intense bands anthesis compared with 150 °Cd after anthesis. Interestingly, a
corresponding to histones (Figure 2C1). Therefore, we tested higher proportion of proteins involved in RNA metabolism and
the presence of possible contamination from other subcellular a higher ratio of uncharacterized proteins was observed in T.
compartments by assaying the presence of the representative Monococcum compared with T. aestivum, suggesting that these
marker enzymes for those compartments (i.e., mitochondrion, proteins could be expressed at different levels or have different
chloroplast, thylakoids, and cytoplasm) by Western immunoa- regulatory processes in the two species.
nalyses. The Western blot analyses performed with the different Subcellular Localization
antibodies showed an enrichment in low-molecular-weight Five different prediction tools were employed to investigate
proteins in nuclear extracts corresponding to histone variants subcellular localization (Table S1, Supporting Information).
(Figure 2C1) and the absence of vacuolar ATPase, photo- Eighty four percent of identified unique proteins had a
system II reaction center protein D1, UDP-glucose pyrophos- predicted nuclear localization with at least one tool prediction
phorylase, and H protein of glycine decarboxylase complex and 65% with at least two tools, which suggest that these
(GDC) in these extracts, confirming no detectable level of cross proteins are nuclear actors or spend at least some time in the
contamination (Figure 2C2−C5). nucleus. Nevertheless, the proteins that were not predicted as
Our protocol as compared with the initial one included having a nuclear localization (16%) cannot be totally excluded
additional washing steps in which contaminants are removed due to the limitations of the prediction tools and the fact that
from the nuclei-enriched pellet without important loss of the subcellular localization pattern changes as the proteins can
nuclear proteins, as shown on Western blots antihistone be rapidly exported. In eukaryotic cells, up to 35% of proteins
(Figure 2C1), and also an improved method for qualitative and have multisubcellular localizations.31 Proteins may simulta-
quantitative extraction of nuclear proteins. neously locate or move between different cellular compart-
Protein Identification through Mass Spectrometry ments, such as transcription factors and signaling pathway
Shotgun proteomics, which is the analysis of a complex mixture transduction factors. Proteins may play different roles in
of peptides using LC−MS/MS without prior separation, was biological processes when they are in different subcellular
employed in this study to analyze nuclei protein fraction. The compartments. For these proteins, single subcellular local-
databank was Uniprot and the search program used was ization annotation will lose some important information.31
Sometimes proteins had an atypical nuclear localization, but the
Xtandem (Release 2013.9.1.0). The proteins were identified
fact that they were found in several nuclear proteome studies
with at least two of their peptides matching the database entry
increases the probability of a nuclear localization. This was the
with a P value <0.01. By this approach, 797 proteins were
case for some proteins discussed later.
identified with high confidence. These proteins were distributed
in 528 groups of proteins, which shared at least one common Proteins Having an Atypical Nuclear Localization
peptide and had the same function. The taxonomy was Actin-97 belonging to the actin family was identified.
Triticum, so identifications take places in Triticum aestivum or Nucleoskeletal proteins such as actin are recognized as having
in Triticum urartu, the diploid progenitor of the bread wheat A- fundamental role in nuclear structure.32,33 Actin was also found
genome (Table S1, Supporting Information). in nuclear proteomes from Arabidopsis thaliana leaves9 and
Unique proteins were classified in nine functional categories Cicer arietinum seedlings.12
according to the GO rules (Figure S1, Supporting Informa- A 14−3−3 protein was also detected and was already found
tion). The most represented categories were: RNA metabolism in A. thaliana leaves nuclear proteome.9 This large family of
(including processing, splicing, transcription, and binding ∼30 kDa acidic proteins has been implicated in the modulation
subclasses), protein metabolism (including synthesis (ribosome of distinct biological processes by binding to phosphorylated
biogenesis), translation, and modification (including degrada- proteins.34 Many nuclear proteins are putatively phosphory-
tion and folding) subclasses), and DNA metabolism (structure, lated,35 and some protein kinases can be translocated into
binding, and chromatin remodeling). Twenty percent of the nuclei where they phosphorylate transcription factors, which, in
proteins were uncharacterized. Numerous proteins were turn, modulate gene expression.36 A protein serine/threonine
involved in metabolism and miscellaneous (15%). Some storage kinase with potential nuclear localization was also present in
proteins (3%) were found and could be considered as our nuclear extracts.
4437 DOI: 10.1021/[Link].5b00446
J. Proteome Res. 2015, 14, 4432−4439

52
Journal of Proteome Research Technical Note

Five ATP synthases were identified. These ubiquitous monococcum, because of its diploid genome, should greatly
membrane enzymes play a key role in biological energy fasten our understanding of regulations mechanisms in bread
metabolism. Orthologs of mitochondrial ATP synthase have wheat.
been detected in A. thaliana nucleus and other organelles.37,31
Several metabolic enzymes have been shown to function as
transcriptional regulators. For example, Malate dehydrogenase
■
*
ASSOCIATED CONTENT
S Supporting Information
can have a nucleocytoplasmic localization and be involved in
The Supporting Information is available free of charge on the
transcription regulation.38 An enzyme involved in carbohydrate
ACS Publications website at DOI: 10.1021/[Link]-
metabolism (glyceraldehyde-3-phosphate dehydrogenase
me.5b00446.
GAPDH) identified in our samples was also found in Pisum
sativum leaf nuclei.39 Interestingly, this GAPDH acts as a Figure S1: Graphical representation of the functional
coactivator to regulate the expression of histone H2B.40 categories of proteins found in wheat grain nuclei
The existence of multifaceted roles of glycolytic proteins enriched fraction. Figure S2: Classification in functional
suggests that links between metabolic sensors and transcription categories obtained for developing grains of T. aestivum
are established directly through enzymes that participate in and T. monococcum at 150 and 250 °Cd after anthesis.
metabolism.41 Nothing is known yet about the mechanisms (PDF)
underlying the nuclear/cytoplasmic distribution of metabolic Table S1: Proteins identified in T. aestivum and T.
enzymes. monococcum grain nuclei at 150 and 250 °Cd after
Uncharacterized Proteins anthesis. (XLSX)
In this study, many proteins (20%) did not have similarity to
known proteins and are therefore “‘uncharacterized”’. The
assigned function can be inappropriate if automated transfer to
■ AUTHOR INFORMATION
Corresponding Author
a distant species is used.42 No functional annotations based on *Tel: + 33 (0)473 624 3529. E-mail: [Link]@
Blastp similarity searches were given for those proteins. [Link].
Furthermore, the number of uncharacterized proteins is close Present Address
to that reported in others nuclear proteome studies with 10− ⊥
18% of unknown proteins8 and up to 44% of hypothetical P.M.: INRA, UMR0759 Laboratoire d’Ecophysiologie des
proteins in rice nuclear endosperm.14 In our study, 70% of Plantes sous Stress Environnementaux, 2 place Viala, F-34 060
these proteins have a predicted nuclear localization with at least Montpellier, France.
two tools (Table S1, Supporting Information). Author Contributions
∥
To explain such number of unidentified proteins, we may E.B. and T.B. contributed equally to this manuscript.
hypothesis that even when employing databases of closely Notes
related species a large number of viable tandem mass spectra of
The authors declare no competing financial interest.

■
peptides might not be assigned accurately to a protein, as a
single amino acid mutation may significantly alter the peptide
mass and the resulting fragmentation pattern. ACKNOWLEDGMENTS

■
We thank Christelle Damon (Blaise Pascal University,
CONCLUSIONS Clermont-Ferrand) for her help with microscopic observation
and Dr. Karine Gallardo (INRA, Dijon) for providing some of
In this study, we report the first analysis of the nuclear the antibodies used in this work. This work was supported in
proteome in developing grains of two wheat species. The purity the context of a Ph.D. grant from the French Ministry for
of the isolated nuclei together with the improved nuclear Higher Education and Research to T.B. and funding from the
proteins extraction allows us to perform specific proteomics French Government managed by the Research National
analysis by Western blot and shotgun approach (this work) but Agency (ANR) in the framework of the Investments for the
also by 2D electrophoresis and electrophoretic mobility shift Future (ANR-10-BTBR-03).

■
assay (data not shown).
One of the most interesting observations in this study results REFERENCES
from the identification of numerous nuclear proteins and also (1) Vensel, W. H.; Tanaka, C. K.; Cai, N.; Wong, J. H.; Buchanan, B.
qualitative changes of protein pattern evaluated in two stages B.; Hurkman, W. J. Developmental changes in the metabolic protein
during early development and between two closely related profiles of wheat endosperm. Proteomics 2005, 5, 1594−1611.
species. (2) Nadaud, I.; Girousse, C.; Debiton, C.; Chambon, C.; Bouzidi, M.
This work provides the identification of stage- or species- F.; Martre, P.; Branlard, G. Proteomic and morphological analysis of
specific proteins as well as common ones, independently off early stages of wheat grain development. Proteomics 2010, 10, 2901−
protein abundance variations. This exploratory approach reveals 2910.
a number of proteins potentially involved in the regulation/ (3) Tasleem-Tahir, A.; Nadaud, I.; Chambon, C.; Branlard, G.
control of protein synthesis in the grain and also several Expression Profiling of Starchy Endosperm Metabolic Proteins at 21
proteins involved in development. Above all, the study enabled Stages of Wheat Grain Development. J. Proteome Res. 2012, 11, 2754−
2773.
the transition from qualitative to quantitative nuclear (4) Bancel, E.; Rogniaux, H.; Debiton, C.; Chambon, C.; Branlard, G.
proteomics. Hence, a major task will now be the use of Extraction and Proteome Analysis of Starch Granule-Associated
quantitative proteomic approaches to correlate variations in the Proteins in Mature Wheat Kernel (Triticum aestivum L.). J. Proteome
relative abundances and post-translational states of nuclear Res. 2010, 9, 3299−3310.
proteins in wheat grains across developmental stages including (5) Tasleem-Tahir, A.; Nadaud, I.; Girousse, C.; Martre, P.; Marion,
or not environmental and genetic fluctuations. The study of T. D.; Branlard, G. Proteomic analysis of peripheral layers during wheat

4438 DOI: 10.1021/[Link].5b00446

J. Proteome Res. 2015, 14, 4432−4439

53
Journal of Proteome Research Technical Note

(Triticum aestivum L.) grain development. Proteomics 2011, 11, 371− (26) Horton, P.; Park, K. J.; Obayashi, T.; Fujita, N.; Harada, H.;
379. Adams-Collier, C. J.; Nakai, K. WoLF PSORT: protein localization
(6) Meziani, S.; Nadaud, I.; Gaillard-Martinie, B.; Chambon, C.; predictor. Nucleic Acids Res. 2007, 35, W585−W587.
Benali, M.; Branlard, G. Proteomic analysis of the mature kernel (27) Briesemeister, S.; Rahnenfuhrer, J.; Kohlbacher, O. YLoc-an
aleurone layer in common and durum wheat. J. Cereal Sci. 2012, 55, interpretable web server for predicting subcellular localization. Nucleic
323−330. Acids Res. 2010, 38, W497−W502.
(7) Meziani, S.; Nadaud, I.; Gaillard-Martinie, B.; Chambon, C.; (28) Goldberg, T.; Hecht, M.; Hamp, T.; Karl, T.; Yachdav, G.;
Benali, M.; Branlard, G. Proteomic Comparison of the Aleurone Layer Ahmed, N.; Altermann, U.; Angerer, P.; Ansorge, S.; Balasz, K.;
in Triticum Aestivum and Triticum Monococcum Wheat Varieties. Bernhofer, M.; Betz, A.; Cizmadija, L.; Do, K. T.; Gerke, J.; Greil, R.;
Curr. Proteomics 2014, 11, 71−77. Joerdens, V.; Hastreiter, M.; Hembach, K.; Herzog, M.; Kalemanov,
(8) Narula, K.; Datta, A.; Chakraborty, N.; Chakraborty, S. M.; Kluge, M.; Meier, A.; Nasir, H.; Neumaier, U.; Prade, V.; Reeb, J.;
Comparative analyses of nuclear proteome: extending its function. Sorokoumov, A.; Troshani, I.; Vorberg, S.; Waldraff, S.; Zierer, J.;
Front. Plant Sci. 2013, 4, 100. Nielsen, H.; Rost, B. LocTree3 prediction of localization. Nucleic Acids
(9) Bae, M. S.; Cho, E. J.; Choi, E. Y.; Park, O. K. Analysis of the Res. 2014, 42, W350−W355.
Arabidopsis nuclear proteome and its response to cold stress. Plant J. (29) Drea, S.; Leader, D. J.; Arnold, B. C.; Shaw, P.; Dolan, L.;
2003, 36, 652−663. Doonan, J. H. Systematic spatial analysis of gene expression during
(10) Khan, M. M. K.; Komatsu, S. Rice proteomics: recent wheat caryopsis development. Plant Cell 2005, 17, 2172−2185.
developments and analysis of nuclear proteins. Phytochemistry 2004, (30) Castillo, J.; Rodrigo, M. I.; Marquez, J. A.; Zuniga, A.; Franco, L.
65, 1671−1681. A pea nuclear protein that is induced by dehydration belongs to the
(11) Tan, F.; Li, G. S.; Chitteti, B. R.; Peng, Z. H. Proteome and vicilin superfamily. Eur. J. Biochem. 2000, 267, 2156−2165.
phosphoproteome analysis of chromatin associated proteins in rice (31) Zhang, S.; Xia, X. F.; Shen, J. C.; Zhou, Y.; Sun, Z. R. DBMLoc:
(Oryza sativa). Proteomics 2007, 7, 4511−4527. a database of proteins with multiple subcellular localizations. BMC
(12) Pandey, A.; Choudhary, M. K.; Bhushan, D.; Chattopadhyay, A.; Bioinf. 2008, 9, 127.
Chakraborty, S.; Datta, A.; Chakraborty, N. The nuclear proteome of (32) Cruz, J. R.; Moreno Diaz de la Espina, S. M. D. Subnuclear
chickpea (Cicer arietinum L.) reveals predicted and unexpected compartmentalization and function of actin and nuclear Myosin I in
proteins. J. Proteome Res. 2006, 5, 3301−3311. plants. Chromosoma 2009, 118, 193−207.
(13) Repetto, O.; Rogniaux, H.; Firnhaber, C.; Zuber, H.; Kuster, H.; (33) Wilson, K. L.; Dawson, S. C. Functional evolution of nuclear
Larre, C.; Thompson, R.; Gallardo, K. Exploring the nuclear proteome structure. J. Cell Biol. 2011, 195, 171−181.
of Medicago truncatula at the switch towards grain filling. Plant J. 2008, (34) Denison, F. C.; Paul, A. L.; Zupanska, A. K.; Ferl, R. J. 14−3-3
56, 398−410. proteins in plant physiology. Semin. Cell Dev. Biol. 2011, 22, 720−727.
(14) Li, G. S.; Nallamilli, B. R. R.; Tan, F.; Peng, Z. H. Removal of (35) Kumar, R.; Kumar, A.; Subba, P.; Gayali, S.; Barua, P.;
high-abundance proteins for nuclear subproteome studies in rice Chakraborty, S.; Chakraborty, N. Nuclear phosphoproteome of
(Oryza sativa) endosperm. Electrophoresis 2008, 29, 604−617. developing chickpea grainlings (Cicer arietinum L.) and protein-kinase
(15) Ferreira, B. S.; Branco, A. T.; de Oliveira, M. A.; Pereira, M. G.; interaction network. J. Proteomics 2014, 105, 58−73.
de Souza Filho, G. Methodological improvements on extraction of (36) Dahan, J.; Wendehenne, D.; Ranjeva, R.; Pugin, A.; Bourque, S.
nuclear proteins and its preliminary analysis during the maize (Zea Nuclear protein kinases: still enigmatic components in plant cell
mays L.) endosperm development. Protein Pept. Lett. 2006, 13, 981− signalling. New Phytol. 2010, 185, 355−368.
984. (37) Heazlewood, J. L.; Whelan, J.; Millar, A. H. The products of the
(16) Erhardt, M.; Adamska, I.; Franco, O. L. Plant nuclear mitochondrial orf25 and orfB genes are Fo components in the plant
proteomics - inside the cell maestro. FEBS J. 2010, 277, 3295−3307. F1Fo ATP synthase. FEBS Lett. 2003, 540, 201−205.
(17) Repetto, O.; Rogniaux, H.; Larre, C.; Thompson, R.; Gallardo, (38) Lee, S. M.; Kim, J. H.; Cho, E. J.; Youn, H. D. A
K. The grain nuclear proteome. Front. Plant Sci. 2012, 3, 289. nucleocytoplasmic malate dehydrogenase regulates p53 transcriptional
(18) Brocklehurst, P. A. Factors controlling grain weight in wheat. activity in response to metabolic stress. Cell Death Differ. 2009, 16,
Nature 1977, 266, 348−349. 738−748.
(19) Ugarte, C.; Calderini, D. F.; Slafer, G. A. Grain weight and grain (39) Anderson, L. E.; Wang, X. W.; Gibbons, J. T. 3 enzymes of
number responsiveness to pre-anthesis temperature in wheat, barley carbon metabolism or their antigenic analogs in pea leaf nuclei. Plant
and triticale. Field Crops Res. 2007, 100, 240−248. Physiol. 1995, 108, 659−667.
(20) Nicolas, M. E.; Gleadow, R. M.; Dalling, M. J. Effect of post- (40) Zheng, L.; Roeder, R. G.; Luo, Y. S phase activation of the
anthesis drought on cell division and starch accumulation in histone H2B promoter by OCA-S, a coactivator complex that contains
developing wheat grains. Ann. Bot 1985, 55, 433−444. GAPDH as a key component. Cell 2003, 114, 255−266.
(21) Aki, T.; Yanagisawa, S. Application of Rice Nuclear Proteome (41) Kim, J. W.; Dang, C. V. Multifaceted roles of glycolytic enzymes.
Analysis to the Identification of Evolutionarily Conserved and Trends Biochem. Sci. 2005, 30, 142−150.
Glucose-Responsive Nuclear Proteins. J. Proteome Res. 2009, 8, (42) Agrawal, G. K.; Bourguignon, J.; Rolland, N.; Ephritikhine, G.;
Ferro, M.; Jaquinod, M.; Alexiou, K. G.; Chardot, T.; Chakraborty, N.;
3912−3924.
Jolivet, P.; Doonan, J. H.; Rakwal, R. Plant organelle proteomics:
(22) Bradford, M. M. Rapid and sensitive method for quantitation of
collaborating for optimal cell function. Mass Spectrom. Rev. 2011, 30,
microgram quantities of protein utilizing principle of protein-dye
772−853.
binding. Anal. Biochem. 1976, 72, 248−254.
(23) Ashburner, M.; Ball, C. A.; Blake, J. A.; Botstein, D.; Butler, H.;
Cherry, J. M.; Davis, A. P.; Dolinski, K.; Dwight, S. S.; Eppig, J. T.;
Harris, M. A.; Hill, D. P.; Issel-Tarver, L.; Kasarskis, A.; Lewis, S.;
Matese, J. C.; Richardson, J. E.; Ringwald, M.; Rubin, G. M.; Sherlock,
G. Gene Ontology: tool for the unification of biology. Nat. Genet.
2000, 25, 25−29.
(24) Blum, T.; Briesemeister, S.; Kohlbacher, O. MultiLoc2:
integrating phylogeny and Gene Ontology terms improves subcellular
protein localization prediction. BMC Bioinf. 2009, 10, 274.
(25) Brameier, M.; Krings, A.; MacCallum, R. M. NucPred -
Predicting nuclear localization of proteins. Bioinformatics 2007, 23,
1159−1160.

4439 DOI: 10.1021/[Link].5b00446

J. Proteome Res. 2015, 14, 4432−4439

54
 Chapitre 2 : Validation de la méthode d’extraction des protéines nucléaires

Supplementary information

Table S1: Proteins identified in T. aestivum and T. monococcum grain nuclei at 150 and 250°Cd after
anthesis. Accession corresponding to Uniprot database. The proteins were sorted according to functional
categories. Five predicting tools were used to determine nuclear localization (N), the number 1 to 5
correspond to the number of tools giving a nuclear localization.
(You can see this table online at: [Link]

Figure S1: Graphical representation of the

functional categories of proteins found in wheat
grain nuclei enriched fraction. The 528 unique
identified proteins were classified in 9 functional
classes according to Gene ontology. The
percentage of proteins in each functional class is
shown.

Figure S2. Classification in functional categories obtained for developing grains of T. aestivum and T.
monococcum at 150 and 250C°Cdays after anthesis. The radar plots show the percentage of proteins in
each functional class.

55
 Chapitre 2 : Validation de la méthode d’extraction des protéines nucléaires

3. CONCLUSIONS
Dans cet article, les analyses réalisées ont permis de répondre à deux objectifs
principaux : la validation de la méthode d’extraction des protéines nucléaires, et pour la
première fois l’exploration du protéome nucléaire du grain de blé (Figure 21). Les colorations
de noyaux ont révélé leur intégrité à l’issue de l’étape de purification sur gradient de densité.
Les analyses par western blot ont ensuite mis en évidence l’enrichissement des extraits
nucléaires en protéines histones, et l’absence dans ces extraits des protéines marqueurs des
autres compartiments subcellulaires (vacuole, chloroplaste, mitochondrie, cytoplasme). A
l’issue de l’analyse des protéines nucléaires en spectrométrie de masse, 797 protéines ont été
identifiées. Parmi elles, 368 (46%) étaient communes aux différents échantillons, c’est-à-dire
aux deux espèces de blé (T. aestivum et T. monococcum) et aux deux stades de développement
étudiés (150 et 250°Cj après floraison). Sur cet aspect, nous avons pu voir que malgré les
différences observées entre échantillons, les proportions des différentes fonctions retrouvées
étaient comparables, notamment entre les deux stades de développement du grain d’une même
espèce. Nous avons estimé grâce à des outils de localisation subcellulaire que parmi l’ensemble
des protéines identifiées, 64 à 85% sont prédites comme ayant une localisation nucléaire. La
méthode utilisée est donc robuste et permet d’obtenir des extraits protéiques enrichis en
protéines nucléaires, pour les deux espèces de blé considérées.

Figure 21. Principales étapes de la validation de la méthode d’extraction des protéines nucléaires.

En plus de valider la méthode, l’identification des protéines a permis de révéler l’identité

d’un grand nombre d’acteurs nucléaires du grain de blé. Ces protéines sont impliquées
principalement dans le métabolisme des ARN, des protéines et de l’ADN avec des fonctions
associées aux processus d’assemblage des ribosomes, de la transcription, la maturation des
ARNm ou encore la conformation de l’ADN.

56
 Chapitre 2 : Validation de la méthode d’extraction des protéines nucléaires

Les différences observées entre les temps thermiques après floraison (150 et 250°Cj) d’un
point de vue qualitatif ont laissé entrevoir de probables variations quantitatives au cours du
développement du grain au niveau de ce sous-protéome. De plus, un rendement d’extraction de
l’ordre de 110 à 160 µg de protéines extraites par gramme de grains frais, suivant l’espèce de
blé et le stade de développement testés ainsi que la bonne reproductibilité de la méthode nous
laissent la possibilité d’étudier ces variations quantitatives.

57
58
 CHAPITRE 3 :
Le protéome nucléaire du grain en
développement de blé tendre

59
60
 Chapitre 3 : Le protéome nucléaire du grain en développement de blé tendre

1. INTRODUCTION
La compréhension des mécanismes qui contrôlent le développement du grain de blé est
importante pour améliorer les paramètres de rendement et maintenir la qualité du grain. Avant
d’atteindre la maturité physiologique, le grain passe par trois grandes phases : cellularisation,
remplissage où les réserves sont accumulées et maturation, où le grain se déshydrate (Shewry
et al., 2012). Au niveau moléculaire, le développement du grain est caractérisé par la variation
d’expression de nombreux gènes (Wan et al., 2008; Capron et al., 2012; Yu et al., 2016).
Dans cette étude, nous avons émis l’hypothèse que les changements morphologiques et
physiologiques du grain de blé observés au cours de son développement impliquent des
variations d’abondance de protéines nucléaires à des moments spécifiques. L’objectif a ainsi
été d’analyser le protéome nucléaire à différents temps thermiques après floraison pour
visualiser ces variations quantitatives et identifier les protéines concernées.
Pour cela, nous avons choisi d’analyser le protéome nucléaire du grain de blé tendre à six
temps thermiques après floraison, qui couvrent les trois phases du développement du grain et
leurs transitions (Figure 22). Des grains ont ainsi été prélevés pendant la phase de cellularisation
(150°Cj), à la transition entre les phases de cellularisation et de remplissage (250°Cj), pendant
le remplissage (350 et 450°Cj), à la transition entre les phases de remplissage et de maturation
(600°Cj) et pendant la maturation (750°Cj). Les protéines nucléaires ont été extraites de ces
grains avec la méthode décrite dans le chapitre 2. Nous avons alors fait le choix de séparer ces
protéines sur gel bidimensionnel. Afin de maximiser la résolution de la séparation, deux
gradients de point isoélectrique (4-7 et 6-11) ont été analysés. Afin de révéler l’identité des
protéines nucléaires variant ou non au cours du développement du grain, un maximum de spots
protéiques a été analysé sans a priori par spectrométrie de masse.

Figure 22. Analyse du

protéome nucléaire du
grain de blé tendre en
développement. Les
protéines nucléaires
extraites des grains à six
temps thermiques après
floraison ont été analysées
sur gel bidimensionnel (2D)
dans deux gradients de point
isoélectrique (4-7 et 6-11).

Références
Capron, D., Mouzeyar, S., Boulaflous, A., Girousse, C., Rustenholz, C., Laugier, C., Paux,
E. and Bouzidi, M. (2012) Transcriptional profile analysis of E3 ligase and hormone-
related genes expressed during wheat grain development. BMC Plant Biol., 12, 35.
Shewry, P.R., Mitchell, R. a. C., Tosi, P., et al. (2012) An integrated study of grain
61
 Chapitre 3 : Le protéome nucléaire du grain en développement de blé tendre

development of wheat (cv. Hereward). J. Cereal Sci., 56, 21–30.

Wan, Y., Poole, R.L., Huttly, A.K., et al. (2008) Transcriptome analysis of grain development
in hexaploid wheat. BMC Genomics, 9, 121.
Yu, Y., Zhu, D., Ma, C., Cao, H., Wang, Y., Xu, Y., Zhang, W. and Yan, Y. (2016)
Transcriptome analysis reveals key differentially expressed genes involved in wheat grain
development. Crop J., 4, 92–106.

2. ARTICLE 2 : Changes in the nuclear proteome of developing wheat (Triticum aestivum

L.) grain
Titouan Bonnot1, 2, Emmanuelle Bancel1, 2, Christophe Chambon3, Julie Boudet1, 2, Gérard
Branlard1, 2 and Pierre Martre1, 2
1
INRA, UMR1095 Génétique, Diversité et Ecophysiologie des Céréales, 5 chemin de Beaulieu,
F-63 039 Clermont-Ferrand, France
2
Université Blaise Pascal, UMR1095 Génétique, Diversité et Ecophysiologie des Céréales,
Avenue des Landais, F-63 170 Aubière, France
3
Plateforme d’Exploration du Métabolisme, INRA, Saint-Genès Champanelle, France

62
 ORIGINAL RESEARCH
published: 28 October 2015
doi: 10.3389/fpls.2015.00905

Changes in the nuclear proteome of

developing wheat (Triticum aestivum
L.) grain
Titouan Bonnot 1, 2 , Emmanuelle Bancel 1, 2*, Christophe Chambon 3 , Julie Boudet 1, 2 ,
Gérard Branlard 1, 2 and Pierre Martre 1, 2 †
1
UMR1095 Genetics, Diversity and Ecophysiology of Cereals, Institut National de la Recherche Agronomique,
Clermont-Ferrand, France, 2 UMR1095 Genetics, Diversity and Ecophysiology of Cereals, Blaise Pascal University, Aubière,
France, 3 Metabolism Exploration Platform Proteomic Component, Institut National de la Recherche Agronomique,
Saint-Genès Champanelle, France

Wheat grain end-use value is determined by complex molecular interactions that

occur during grain development, including those in the cell nucleus. However, our
Edited by:
knowledge of how the nuclear proteome changes during grain development is limited.
Dominique Job,
Centre National de la Recherche Here, we analyzed nuclear proteins of developing wheat grains collected during the
Scientifique, France cellularization, effective grain-filling, and maturation phases of development, respectively.
Reviewed by: Nuclear proteins were extracted and separated by two-dimensional gel electrophoresis.
Ombretta Repetto,
Centro di Riferimento Oncologico di
Image analysis revealed 371 and 299 reproducible spots in gels with first dimension
Aviano—SOC FSC, Italy separation along pH 4–7 and pH 6–11 isoelectric gradients, respectively. The relative
Beata Petrovska,
abundance of 464 (67%) protein spots changed during grain development. Abundance
Centre of Plant Structural and
Functional Genomics, Czech Republic profiles of these proteins clustered in six groups associated with the major phases and
*Correspondence: phase transitions of grain development. Using nano liquid chromatography-tandem
Emmanuelle Bancel mass spectrometry to analyse 387 variant and non-variant protein spots, 114 different
[Link]@[Link]
†
proteins were identified that were classified into 16 functional classes. We noted that
Present Address:
Pierre Martre,
some proteins involved in the regulation of transcription, like HMG1/2-like protein
UMR759 Laboratoire and histone deacetylase HDAC2, were most abundant before the phase transition
d’Ecophysiologie des Plantes sous
from cellularization to grain-filling, suggesting that major transcriptional changes occur
Stress Environnementaux, Institut
National de la Recherche during this key developmental phase. The maturation period was characterized by high
Agronomique, Montpellier, France relative abundance of proteins involved in ribosome biogenesis. Data are available via
ProteomeXchange with identifier PXD002999.
Specialty section:
This article was submitted to Keywords: wheat, developing grain, nuclear proteins, 2D gel electrophoresis, LC-MS/MS
Plant Proteomics,
a section of the journal
Frontiers in Plant Science
INTRODUCTION
Received: 01 August 2015
Accepted: 10 October 2015 Wheat (Triticum aestivum L.) grain is a major staple crop in many parts of the world. The end-
Published: 28 October 2015
use value is determined by complex molecular interactions that occur during grain development.
Citation: Development of wheat grain is typical of grass seeds and is commonly subdivided into three
Bonnot T, Bancel E, Chambon C, developmental phases that overlap (Sabelli and Larkins, 2009). After double fertilization, the
Boudet J, Branlard G and Martre P
triploid endosperm divides successively without cytokinesis leading at 70◦ Cd after anthesis (i.e.,
(2015) Changes in the nuclear
proteome of developing wheat
3–4 days after anthesis at an average daily temperature of 20◦ C) to the formation of a coenocyte
(Triticum aestivum L.) grain. whose nuclei are distributed throughout the endosperm (Mares et al., 1975). Cellularization follows
Front. Plant Sci. 6:905. which is a phase of cell division and differentiation until 220◦ Cd (11 days) after anthesis (Chojecki
doi: 10.3389/fpls.2015.00905 et al., 1986). The effective grain-filling phase follows when storage compounds, mainly starch

Frontiers in Plant Science | [Link] 1 October 2015 | Volume 6 | Article 905

63
Bonnot et al. Nuclear proteome of wheat grain

and proteins, rapidly accumulate (Shewry et al., 2012). The MATERIAL AND METHODS
rate of accumulation of starch and proteins slows down at
around 550◦ Cd (27 days) after anthesis, when endosperm nuclei Plant Material
and protein bodies become compressed by starch granules Plants of hexaploid winter wheat (T. aestivum L.) cv Recital were
(Hoshikawa, 1962; Ferreira et al., 2012) and progressively used in this study. Seeds were sown in plug trays filled with a peat
disintegrate. Accumulation stops at 650–700◦ Cd (32–35 days) moss mixture and were kept in a greenhouse until the ligule of
after anthesis when the concentration of water in grain is the third leaf appeared. Air temperatures in the greenhouse were
close to 45 g per 100 g of fresh mass (Schnyder and Baum, maintained at 18/10◦ C (light/dark) and air relative humidity at
1992). Grains then start a phase of rapid desiccation and 70/50% (light/dark). Plants were then vernalized for 8 weeks in
maturation during which desiccation tolerance is acquired. a growth chamber where the air temperature was maintained
Because of the importance of wheat grain in the human at 4 ± 1◦ C, the air relative humidity at 40% and the mean
diet, much research has focused on identifying processes daily photosynthetic photon flux density (PPFD) at the top of
which regulate these different phases of development in the plants at 43 mmol m−2 d−1 during the 8-h photoperiod.
order to optimize grain yield and its quality (Shewry et al., After vernalization, the plants were transplanted into 293-mL
2012). plastic pots filled with a mixture of soil-pozzolan (2:1, w/w) and
The regulation of most of these processes involves transferred to a walk-in growth chamber. The conditions in the
transcriptional regulation and the nucleus plays a key role growth chamber were 20/15◦ C (light/dark), 55/75% air relative
in the regulation of grain development and storage compound humidity (light/dark), with an average PPFD of 550 µmol m−2
accumulation. In plants, the nuclear proteome of leaves or s−1 at top of the plants during the 16-h photoperiod. Plants were
whole seedlings has been studied for several species (Erhardt irrigated twice a day with a commercial nutrient solution.
et al., 2010; Petrovská et al., 2015) including cereals like Oryza Air temperature at the top of the plants was measured
sativa (Khan and Komatsu, 2004; Tan et al., 2007; Aki and continuously. Main stems were tagged when the anthers of the
Yanagisawa, 2009; Choudhary et al., 2009; Jaiswal et al., 2013), central florets emerged (anthesis date). The sum of mean daily
Hordeum vulgare (Petrovská et al., 2014), and Zea mays (Ferreira air temperature after anthesis was calculated to follow grain
et al., 2006; Guo et al., 2014). However, there have been few development in thermal time in degree-days (◦ Cd) above 0◦ C
such studies on seeds (Repetto et al., 2012). In O. sativa, 468 after anthesis. Grains were harvested at 150, 250, 350, 450, 600,
nuclear proteins were identified from endosperm at 9 days after and 750◦ Cd after anthesis and stored at −80◦ C. Only grains of
pollination (dap) (Li et al., 2008) and in Medicago truncatula the first floret from the central part of the ears were collected
143 different nuclear proteins were identified from whole (approximately 10 grains per ear). Four independent replicates
seeds harvested at 12 dap (Repetto et al., 2008). A study of were used.
Z. mays showed that some nuclear proteins extracted from
endosperm isolated from grains harvested between 8 and Nuclei Isolation and Nuclear Protein
35 dap, analyzed on one-dimensional (1D) gels, were more Extraction from Wheat Grains
abundant at certain times of development (Ferreira et al., The method used in the present study to purify nuclei from wheat
2006), but these proteins remain to be identified. No proteomic grains and to extract nuclear proteins was recently validated
study has analyzed the temporal changes in abundance by Bancel et al. (2015). The authors verified the absence of
of nuclear proteins during grain development. However, contamination from non-nuclear proteins by western blots with
identifying and quantifying nuclear proteins is an important primary antibodies which detect protein markers of the different
step in characterizing some of the numerous regulatory sub-cellular compartments.
mechanisms that take place during the dynamic phases of Briefly, nuclei were isolated from 2 g of grains. Grains
grain development. We hypothesized that the developmental were ground in extraction buffer [20 mM Hepes-KOH, pH
physiology and morphology of the wheat grain requires changes 7, 5 mM MgCl2 , 10 mM 2-ME, 0.5 mM PMSF, 0.1% (v/v)
in abundance of several nuclear proteins at specific times of grain phosphatase inhibitor cocktail (Sigma-Aldrich)] with a Polytron
development. homogenizer (Kinematica POLYTRON R PT 10) during 1 min.
The aim of the present study was to analyze the nuclear The homogenate was filtered through two layers of Miracloth
proteome of the developing wheat grain in order to obtain a (Calbiochem) to remove cell debris and cells were lysed by adding
first overview of which nuclear proteins vary in abundance 0.5% (v/v) Triton X-100. After incubation for 15 min at 4◦ C,
during grain development. Nuclear proteins were extracted the resulting lysate was centrifuged at 1000 × g for 10 min at
from wheat (T. aestivum L.) grains collected during the 4◦ C. Each pellet was then washed four times by resuspension in
cellularization, effective grain-filling and maturation phases 1 mL of extraction buffer followed by centrifugation at 1000 ×
of development, and analyzed using two-dimensional (2D) g for 10 min at 4◦ C. Nuclei were purified from the pellet by
gel electrophoresis and electrospray ionization ion trap centrifugation at 930 × g for 30 min at 4◦ C through a stepwise
mass spectrometry (ESI-IT-MS/MS). This allowed us to Percoll density gradient, 30–80% Percoll prepared in extraction
show that some nuclear proteins involved in signaling, buffer. Nuclei floating at the interface were collected and washed
proteolysis, transcription regulation or ribosome biogenesis twice with 3 mL of extraction buffer followed by centrifugation at
were more abundant at specific developmental phases or phase 3500 × g for 5 min at 4◦ C. To verify the purity of isolated nuclei,
transitions. nuclei pellets were washed in 500 µL of PBS (137 mM NaCl,

Frontiers in Plant Science | [Link] 2 October 2015 | Volume 6 | Article 905

64
Bonnot et al. Nuclear proteome of wheat grain

2.7 mM KCl, 4.3 mM Na2 HPO4 , 1.47 mM KH2 PO4 , pH 7.4) and analysis step, isoelectric focusing was performed with 240 µg of
centrifuged at 3500 × g for 5 min at 4◦ C. The supernatant proteins (40 µg of nuclear proteins from each of the six thermal
was removed and nuclei were stained in phosphate buffered times after anthesis). Isoelectric focusing was performed for a
saline (PBS) solution containing 0.1 µg/mL Hoechst for 5 min total of 60,000 voltage hours (Vhr) on a IPGphor II apparatus
in the dark. After two washes in 50 µL of PBS, 5 µL aliquots (GE Healthcare). Focused proteins on strips were then reduced
were observed under fluorescence microscopy (Zeiss Axioplan with 2% (w/v) dithiothreitol in 0.1 M Tris-HCl buffer (pH 8.8)
2 microscope). To verify absence of pigments in nuclei pellets, containing 6 M urea, 30% (v/v) glycerol, and 2% (w/v) SDS for
a chlorophyll assay was performed according to (Pandey et al., 15 min, followed by alkylation with 2.5% (w/v) iodoacetamide
2006). Briefly, 100 µL of sample was mixed with 100 µL of in the same buffer for 15 min. The strips were then loaded
water and 800 µL of acetone. After centrifugation at 1000 × onto 12.5% polyacrylamide gels for SDS-PAGE separation in
g for 5 min, the optical density was measured at 652 nm. The the second dimension. The migration conditions were 10 mA
amount of chlorophyll was observed as µg per µL by calculating per gel for the first 30 min, then 35 mA per gel for 2.5 h. Gels
optical density/34.5 (at 652 nm chlorophyll a and b intersect, 34.5 were stained using Coomassie Brilliant Blue G250 (CBB, Sigma-
is the specific absorption coefficient for both pigments at this Aldrich) (Neuhoff et al., 1985). To improve detection of low
wavelength). abundance protein spots and allow their collection, gels were
Nuclear proteins were prepared using TRI Reagent R (Sigma- destained overnight in a solution containing 40% (v/v) ethanol
Aldrich) according to the manufacturer’s instructions. The final and 10% (v/v) acetic acid and silver-stained following a mass
protein pellet was dried under ambient conditions for 1 h and spectrometry compatible method (Shevchenko et al., 1996).
then stored at −20◦ C.
Image and Statistical Analyses
SDS-PAGE Analysis and Immunoblotting Images (300 dpi, 16-bit greyscale pixel depth) of two-dimensional
gels stained with CBB were acquired with a GS-800 (Biorad)
Using Anti-histone H3 Antibody scanner and analyzed using SameSpots v4.5 (TotalLab) 2D gel
Nuclear protein pellets and proteins from supernatants (S1–
image analysis software. Statistical analyses were performed
S5) collected during nuclei purification were solubilized in
on normalized protein spot volume values. Differences in
100 µL of solubilization buffer (45 mM Tris-HCl, pH 6.8,
normalized protein spot volume due to grain development were
50 mM DTT, 1% (w/v) SDS, 10% (v/v) glycerol, 0.001% (w/v)
analyzed using One-way ANOVA. P-values and adjusted P-
bromophenol blue). A fixed volume (35 µL of each supernatant
values (q-value) were calculated using SameSpots procedures.
or 25 µL of final nuclear protein extract) was loaded onto
The abundance of a protein spot was considered to have changed
12.5% SDS-polyacrylamide gels as described in (Bancel et al.,
during grain development when its P-value and q-value were
2015). For immunoblotting analysis, proteins were transferred
both <0.05. In this case, the protein spot was considered as
from the 1D gels to nitrocellulose membranes (Hybond ECL,
“variant” and others as “non variant.” Principal component
GE Healthcare) during 1 h in a semidry unit apparatus (GE
analysis was performed using the FactoMineR (Husson et al.,
Healthcare). Membranes were incubated with an anti-histone
2014) package for R v3.0.1 (R Core Team, 2013) statistical
H3 antibody (Abcam) diluted at 1:1000. Membranes were
software on the set of spots detected on 2D gels and hierarchical
then incubated with anti-rabbit secondary antibody coupled
clustering on principal components was computed on significant
to horseradish peroxidase (HRP, GE Healthcare) diluted at
spots to build protein abundance profiles.
1:5000. The chemiluminescence was developed according to the
manufacturer’s instructions (ECL Western Blotting, SuperSignal
West Pico Chemiluminescent Substrate kit, Amersham).
Protein Identification
Protein spots were excised manually from 2D gels. For 2D
gels stained with CBB, protein spots were destained once with
Two-dimensional Electrophoresis of 25 mM NH4 HCO3 containing 5% (v/v) acetonitrile (ACN) for
Nuclear Proteins 30 min and twice with 25 mM NH4 HCO3 , 50% (v/v) ACN for
Dried pellets containing nuclear proteins were dissolved in 30 min. Spots were then dehydrated in 100% ACN for 10 min and
50 µL of 2D gel sample buffer [7 M urea, 2 M thiourea, 4% dried for 15 min under an extraction hood at room temperature.
(w/v) CHAPS, 70 mM dithiothreitol, 1% (v/v) immobilized pH For 2D gels stained with silver nitrate, protein spots were first
gradient (IPG) buffer (either for the pH 4–7 range or the pH 6– destained with 30 mM K3 Fe(CN)6 , 100 mM Na2 S2 O3 for 1–
11 range), and 0.34% (v/v) protease inhibitor (Sigma-Aldrich)] 2 min, then washed twice in water for 15 min, before following
for 1 h at room temperature with constant agitation. An aliquot the CBB destaining steps as above. Proteins were digested
(5 µL) of each sample was used to quantify protein content overnight at 37◦ C by adding 120 ng of trypsin (Promega). After
(Bradford, 1976) using bovine serum albumin as standard. extracting peptides with ACN, 8 µL of hydrolysate were injected
Isoelectric focusing was carried out with 150 µg of proteins, into an Ultimate R 3000 HPLC system (Dionex) coupled to
made up to 250 µL with the 2D buffer containing 0.05% an electrospray ionization ion trap mass spectrometer (ESI-IT-
(w/v) bromophenol blue and used to passively rehydrate 13- MS/MS; LTQ Velos, ThermoScientific).
cm immobilized pH gradient strips (pH range 4–7 or 6–11 Protein identity was sought by using Mascot v2.3 (Matrix
Immobilin Dry Strips, GE Healthcare) overnight at 20◦ C. For Science) software against a custom database containing 249,032
the comigration gels, made to be a reference for the digital sequences from Aegilops tauschii, A. thaliana, Brachypodium

Frontiers in Plant Science | [Link] 3 October 2015 | Volume 6 | Article 905

65
Bonnot et al. Nuclear proteome of wheat grain

distachyon, H. vulgare, O. sativa, T. aestivum, and T. urartu grain of nuclear proteins extracted at various thermal times after
and sequences from the wheat transcription factor database anthesis were similar, except at 750◦ Cd after anthesis when the
wDBFT (Romeuf et al., 2010). Proteins were considered to be variability was too high to evaluate this (Figure 1D).
identified if at least two non-redundant peptides were found
to match a single reference in the databases. A cut-off was Two-dimensional Electrophoresis of Wheat
applied for individual peptide ion scores according to the
significance threshold of the MASCOT program (P < 0.05).
Grain Nuclear Proteins
To maximize resolution of protein spots in 2D electrophoresis,
Curated protein sequences which had no functional information
two different pH gradients were used in the first dimension of
were submitted as BLASTP searches against the National
isoelectric focusing. Comigration gels are shown in Figure 2 in
Center for Biotechnology Information non-redundant database
which samples from all 6 stages of wheat grain development were
([Link] Proteins were then classified in
combined. With the pH 4–7 gradient, 371 spots were detected
functional classes according to the KEGG PATHWAY database
and with the pH 6–11 gradient 299 spots were detected. From this
(Kanehisa et al., 2014) and gene ontology (Ashburner et al., 2000).
total of 690 protein spots, 213 were manually excised from the pH
Subcellular localization of identified proteins was predicted by
4–7 gradient gel and 174 from the pH 6–11 gradient gel, without
in silico analysis using Multiloc2 (Blum et al., 2009), WolfPSort
any a priori (Figure 2). From the total of 387 excised spots,
(Horton et al., 2007), Y loc (Briesemeister et al., 2010), and
343 (88.6%) polypeptides were identified by LC-ESI-MS/MS,
LocTree (Goldberg et al., 2014) programs. The top two hits
which correspond to 114 different proteins (Supplementary
were considered for Multiloc and Y loc. The mass spectrometry
Tables S1, S2). In many cases, the same protein was found in
proteomics data have been deposited to the ProteomeXchange
multiple spots, indicative of isoforms and/or post-translational
Consortium (Vizcaíno et al., 2014) via the PRIDE partner
modifications.
repository with the dataset identifier PXD002999.
The identified proteins were organized into 16 functional
classes (Figure 3 and Supplementary Table S1). Among the
RESULTS proteins with known functions, a large proportion (15%)
corresponded to ribosomal proteins involved in ribosome
Nuclei Purified and Nuclear Proteins biogenesis. The second largest category (11%) comprised
proteins involved in transcription and transcription regulation.
Extracted from Developing Wheat Grains For example, DNA-directed RNA polymerases I, II, and III
Here we analyzed the nuclear proteome of developing wheat
subunit rpabc3 is involved in RNA synthesis, RNA helicases,
grain. For this, grain was harvested at six thermal times
arginine/serine-rich splicing factor, and Mago nashi-like protein
corresponding to the cellularization phase (150◦ Cd after
are all involved in mRNA processing. Transcriptional regulators
anthesis), at the transition between cellularization and grain-
were also identified such as a MADS-box transcription factor
filling phases (250◦ Cd after anthesis), during the grain-filling
and a HMG1/2-like protein. Three spots corresponded to
phase (350 and 450◦ Cd after anthesis), at the transition between
histone deacetylases, which mediate the deacetylation of lysine
grain-filling and maturation phases (600◦ Cd after anthesis), and
residues on the N-terminal part of core histones and three
during the maturation phase (750◦ Cd after anthesis). Nuclei
others matched the FACT complex subunit SSRP1-B which
were isolated from wheat grains using Percoll density gradient
regulates transcription by modifying nucleosomal structure. In
purification. The integrity of isolated nuclei was verified by
the same category, WD repeat-containing protein RBAP1 was
Hoechst staining. For all stages of grain development studied,
also identified. Proteins involved in protein folding represented
Hoechst staining showed that nuclei with an average diameter of
the third largest category (9%) among characterized proteins.
approximately 20 µm had been purified (Supplementary Figure
Other proteins were related to translation (6%) including
1A). Chlorophyll assay was also performed which showed that
initiation and elongation factors. In the nucleosome assembly
nuclear extracts were free of these pigments (Supplementary
category (6%) histones H1, H2A, H2B, and H4 were identified.
Figure 1B). Nuclear proteins and proteins from supernatants
Finally, some proteins with storage functions were identified in
collected during nuclei purification were analyzed by SDS-PAGE
the nutrient reservoir activity category (6%) and are thought
and western blot analysis with an anti-histone H3 antibody
to be contaminants from the purification method. The function
(Figures 1A,B). For each stage of grain development, 1D protein
of 17 (16%) proteins was not known so they were classified as
profiles of the nuclear fractions were different from those of
uncharacterized.
supernatants (Figure 1A). After blotting, histone H3 proteins
were clearly detected in the nuclear fraction, confirming that the
protein extracts were enriched in nuclear proteins (Figure 1B). The Relative Abundance of Many Nuclear
Histone H3 was less abundant in supernatant S1 from protein Proteins Varied During Grain Development
samples from 250, 350, 450, and 600◦ Cd after anthesis indicating For the six thermal times after anthesis, CBB stained gels
that a few nuclei were lost during purification. from four biological replicates were analyzed by digital
The 1D gels show that the quantity of nuclear proteins imaging. Principal component analysis was performed with
extracted from an equal mass of grain decreased during grain the normalized volumes of the 690 protein spots detected by
development. This result was confirmed by protein assays of the image analysis (Section Two-dimensional Electrophoresis of
nuclear protein fraction (Figure 1C). However, the amount per Wheat Grain Nuclear Proteins). The four replicates for a given

Frontiers in Plant Science | [Link] 4 October 2015 | Volume 6 | Article 905

66
Bonnot et al. Nuclear proteome of wheat grain

FIGURE 1 | Nuclear proteins extracted from developing wheat grains. Proteins from supernatants (S1–S5) and nuclear-enriched fractions (N) for each thermal
time after anthesis (150–750◦ Cd) were analyzed on SDS-PAGE (A) then western blot analysis was performed using antibody directed against histone H3 (B). M,
molecular weight protein standards. Amount of proteins in the nuclear-enriched fraction per fresh mass (FM) of grain (C) or per grain (D) vs. thermal time after
anthesis. In (C,D) boxes show the 25th to 75th percentile range, horizontal lines in boxes show medians, and error bars outside boxes show the 10th to 90th
percentile range for n ≥ 10 independent extractions.

time point segregated away from those of other time points highly represented in the variant protein group (five different
(Supplementary Figure 2). From the 690 protein spots detected, proteins from each class) than in the non-variant protein group
226 (33%) had a constant relative abundance during grain (1 and 3 different proteins in each class, respectively).
development, and 464 varied significantly (67%). Protein spots The clustering analysis was first performed separately on the
were excised without any a priori. In this way, among the 387 pH gradients. Since the two pH ranges gave similar clusters,
protein spots analyzed by LC-MS/MS, 153 (40%), corresponding the clustering analysis presented here was performed using data
to 69 different proteins, did not vary in relative abundance during from the two pH ranges. The 464 variant protein spots detected
grain development, and are qualified as non-variant, whereas 234 by image analysis were grouped into six profiles according to
(60%), corresponding to 72 different proteins, were variant. An their relative abundance at different stages of grain development
interesting initial conclusion is that a protein can be identified (Figure 5). Profile 1 included 37 spots with a maximum
in multiple spots, some of which vary during grain development, normalized volume at 150◦ Cd after anthesis that decreased
and others which do not. to a minimum value at 450◦ Cd after anthesis (Figure 5A).
Proteins involved in functional classes ribosome Profile 2 grouped 58 spots whose normalized volume peaked
biogenesis (12 proteins), uncharacterized (9 proteins) and at 250◦ Cd after anthesis (Figure 5B). The 54 spots that defined
transcription/transcription regulation (8 proteins) were the profile 3 had a maximum normalized volume between 150 and
most numerous among the non-variant proteins (Figure 4). 450◦ Cd after anthesis (Figure 5C). Profile 4 grouped 24 spots
The ribosome biogenesis functional class (10 proteins) was also with normalized volume that peaked at 450◦ Cd after anthesis
highly represented among the variant proteins (Figure 4). The (Figure 5D). Profile 5 grouped 184 spots whose normalized
relative abundance of seven histones varied significantly. Only volume increased throughout grain development (Figure 5E)
50% (6/12) of the identified proteins related to transcription and profile 6 included 107 spots whose normalized volume
regulation varied during grain development. Conversely, the decreased from 150 to 450◦ Cd after anthesis and then increased
functional classes of proteolysis and plant defense were more until 750◦ Cd after anthesis (Figure 5F).

Frontiers in Plant Science | [Link] 5 October 2015 | Volume 6 | Article 905

67
Bonnot et al. Nuclear proteome of wheat grain

FIGURE 3 | Functional classification of nuclear proteins. The 114

different proteins identified were grouped into 16 functional classes according
to the KEGG pathway database and gene ontology (GO). The percentage of
proteins in each functional class is shown.

and therefore were more abundant as a class during the latter

part of the grain filling phase and during maturation. Proteins
with profile 6, increasing steadily in abundance during the late
grain-filling phase, included three serpins, which are known
to negatively regulate endopeptidase activity, four proteins
involved in protein folding and two proteins involved in plant
defense.

Focus to Variation in Abundance for Some

Identified Protein
In some cases, only one spot matched a unique protein and the
volume of this spot varied significantly during seed development,
e.g., the Mago nashi-like protein and the Do-like 9 protease
with profile 1 (Figures 6A,B). In 25% of the cases (28 different
proteins), several protein spots corresponding to the same
FIGURE 2 | Map of wheat grain nuclear protein spots. Comigration in 2-D
electrophoresis of nuclear proteins extracted from grains harvested at six
protein did not vary significantly in the same way. For example,
thermal time after anthesis. Proteins were loaded onto 13-cm IPG strips in the two protein spots corresponded to a histone deacetylase HDAC2.
pH range 4–7 (A) and 6–11 (B) for isoelectric focusing in the first dimension One spot varied as in profile 1, but the other did not vary during
and SDS-PAGE in the second dimension. M, molecular weight protein grain development (Figure 6C). This was also the case for a 60S
standards. Protein spots that were excised and analyzed by mass acidic ribosomal protein P0 protein. One protein spot peaked
spectrometry are indicated.
in relative abundance at 600◦ Cd after anthesis, while the other
spot remained constant (Figure 6D). In a few cases (16%), several
Each of the 72 identified variant proteins had at least one protein spots corresponding to the same protein were variant but
of the six profiles (Supplementary Table S1). There were 12 did not have the same relative abundance profile. For example,
proteins with profile 1, 19 with profile 2, 15 with profile 3, an HMG1/2-like protein was present in eight protein spots. Of
6 with profile 4, 21 with profile 5, and 21 with profile 6. these, two varied with profile 1 dynamics and four with profile 2
In some cases therefore the same protein, perhaps different dynamics (Figure 6E). Nevertheless overall this protein was more
isoforms, had different abundance profiles. Profiles 1 and 2 abundant during the cellularization phase. Finally in some cases,
were characterized by proteins involved in transcription and a protein was identified in several spots which all showed the
transcription regulation. For example, HMG1/2-like protein, same abundance dynamics. For example, 40S ribosomal protein
Mago nashi-like protein and a histone deacetylase HDAC2 S12 was identified in three spots, which were all classified as
were more abundant during the cellularization phase of profile 5 (Figure 6F), all increasing in abundance during grain
grain development (profile 1). Two helicase proteins and a development.
HMG1/2-like protein peaked in relative abundance at the end
of the cellularization phase (profile 2). Seven histones had DISCUSSION
profile 3 indicating that they were more abundant during the
cellularization phase and the beginning of the grain-filling phase. DNA replication, repair and modification, RNA transcription,
Several variant ribosomal proteins had profiles 4, 5 and 6, and ribosome biogenesis occur in the nucleus. Continuous flux

Frontiers in Plant Science | [Link] 6 October 2015 | Volume 6 | Article 905

68
Bonnot et al. Nuclear proteome of wheat grain

FIGURE 4 | Functional distribution of non-variant and variant nuclear proteins. Proteins identified showing no statistically significant change are represented
on the left and the variant proteins are presented on the right.

FIGURE 5 | Abundance profiles of variant protein spots. Six profiles (A–F) were established from the normalized volumes of the 464 variant spots using
hierarchical clustering on principal components. Data are means of n = 4 replicates. In each profile, the black line represents the mean. The number of protein spots in
each profile is given in parenthesis after the profile number.

of RNAs and proteins across the nuclear envelope make the electrophoresis and LC-ESI-MS/MS, 114 different proteins were
nucleus a very dynamic structure and the site of developmental identified. Changes in the relative abundance of proteins and
regulation. We analyzed the nuclear proteins that were extracted their isoforms during grain development are highlighted. Certain
from nuclei at key stages of wheat grain development. Using 2D proteins which could have a potential role in the regulation of

Frontiers in Plant Science | [Link] 7 October 2015 | Volume 6 | Article 905

69
Bonnot et al. Nuclear proteome of wheat grain

FIGURE 6 | Changes in volumes of several nuclear protein spots. Mago nashi-like protein (A) and protease Do-like 9 (B) were each identified in only one spot
whose abundance varied significantly during grain development. Histone deacetylase HDAC2 (C) and 60S acidic ribosomal protein P0 (D) were each found in two
spots including one which varied significantly (open circles) and one which did not vary significantly during grain development (closed circles). HMG1/2-like protein
was identified in eight spots including six that varied, presented graphically in (E). The 40S ribosomal protein S12 (F) was found in three spots whose volumes varied
significantly during grain development. Data are means ± 1 s.d. for n = 4 independent replicates.

grain development are discussed, and could indeed be the object nuclear proteomes, such as the histone deacetylases (Li et al.,
of further studies. 2008; Repetto et al., 2008). However, the HMG1/2-like protein,
identified here in eight protein spots, had not previously been
found in the nuclear proteome of grain from other species. The
Nuclear Proteins Identified Provide presence of histone proteins was expected as they are involved
Information on the Functions of Wheat in nucleosome assembly, the first level of DNA compaction, and
Grain Nuclei seven were indeed identified.
After translation in the cytoplasm, ribosomal proteins are The presence of other proteins in the nucleus was harder
transported to the nucleolus, where ribosome assembly begins to predict. Four different tools were used to predict subcellular
(Fromont-Racine et al., 2003; Boisvert et al., 2007). Among the localization. Seventy one percent of identified protein spots were
114 identified proteins, 18 (15%) are ribosomal proteins. By predicted to correspond to a nuclear protein with at least two
comparison, in a study of the M. truncatula nuclear proteome, 32 tools (N2, Supplementary Table S1), which suggest that these
ribosomal proteins (22%) were identified (Repetto et al., 2008). proteins correspond to nuclear actors or spend some time in the
The second largest category of proteins (12 proteins) identified nucleus. However, proteins not predicted to be nuclear can’t be
were those involved in transcription/transcription regulation. excluded due to the limitation of the prediction tools. Indeed,
Some of them have been identified in previous studies of grain several proteins known to spend some time in the nucleus were

Frontiers in Plant Science | [Link] 8 October 2015 | Volume 6 | Article 905

70
Bonnot et al. Nuclear proteome of wheat grain

not predicted to be nuclear with the tools employed here, like similar to M. truncatula seed in which transcripts encoding
for example some ribosomal proteins, involved in the ribosome genes involved in ribosome biogenesis were more abundant at
biogenesis that occur in the nucleolus. While it is probable the end of seed development (Repetto et al., 2008). This result
that some of proteins identified are contaminants from the suggests that there may be a ribosome pool formed at the end
purification process, it has been estimated that 35% of total of the wheat grain development, which is necessary for grain
proteins have multiple subcellular locations (Zhang et al., 2008), germination. This is in accordance with results in Arabidopsis
so some are worth discussing. For example, luminal binding thaliana which showed that potential for seed germination is
proteins and HSP70 act mainly in the endoplasmic reticulum or largely programmed during the seed maturation phase (Rajjou
in the cytosol to contribute to the formation of three-dimensional et al., 2004).
structures of proteins or protein complexes. These two types of
protein were found in wheat grain nuclei. HSP70 has also been Potential Regulators of the Beginning and
identified in the nucleus of other plant species (Calikowski et al., End of the Filling Phase
2003; Pandey et al., 2006; Repetto et al., 2008) and some HSP RNA helicases 2 and 34 and Mago nashi-like proteins, all
are translocated to the nuclei of hamster fibroblasts following involved in mRNA maturation, were most abundant at 250
heat stress (Nollen et al., 2001). Two guanine nucleotide binding and 150◦ Cd after anthesis, respectively. Similarly, HMG1/2-like
proteins were identified. They are commonly associated with the protein was most abundant during the cellularization phase.
plasmalemma acting in multiple signal transduction pathways. HMG are abundant DNA–binding chromosomal non-histone
In eukaryotes, guanine nucleotide binding proteins can also be proteins which are not essential for chromatin organization but
associated with endomembranes, nucleus and the cytoskeleton act with transcription factors in transcriptional control (Calogero
(Willard and Crouch, 2000). Four serpin proteins were identified et al., 1999; Jerzmanowski et al., 2000). They are probably also
in this study. In mammals, several members of the serpin family architectural factors in the assembly of certain nucleoprotein
have been found to localize in the nucleus and some have a complexes (Jerzmanowski et al., 2000). Staining of the 2D gels
nuclear localization signal (Silverman et al., 2001). with Pro-Q Diamond R (Invitrogen) revealed that one spot
Translation mostly takes place in the cytosol. However, of this protein was phosphorylated at 150◦ Cd (Supplementary
initiation and elongation factors involved in translation have Figure 3). The histone deacetylase HDAC2 was most abundant
been identified in the nuclear proteomes of M. truncatula, O. at 150◦ Cd after anthesis. In A. thaliana this protein is one of 18
sativa, and here in T. aestivum (Repetto et al., 2008; Choudhary histone deacetylases involved in the repression of gene expression
et al., 2009). Several studies have raised the possibility of nuclear in multiple developmental processes by causing chromatin
translation (Dahlberg et al., 2003), which could take place in compaction (Hollender and Liu, 2008; Liu et al., 2014). Staining
the nucleoplasm and the nucleolus in mammalian cells (David of the 2DE gels with Pro-Q Diamond R showed that the two spots
et al., 2012). Indeed the idea of a translasome has been described, corresponding to this protein were phosphorylated at 150◦ Cd
a super-complex identified in the nucleus of yeast cells which after anthesis and one was still phosphorylated at 250◦ Cd after
would consist of an assembly of ribosomal proteins, elongation anthesis (Supplementary Figure 3). In mammals, many HDACs
factors, proteasome, chaperones and tRNA synthetases (Sha et al., were found to be phosphorylated both in vitro and in vivo
2009). (Sengupta and Seto, 2004). This post-translational modification
Proteins classified in the nutrient storage activity group were could affect the activity of this protein. These results suggest that
identified in many protein spots. They are very abundant at the some changes in transcriptional regulation occur at the transition
end of the grain-filling phase and probably are contaminants. between the cellularization and the grain-filling phases and that
Their presence may even explain the variability in the amount a number of genes are repressed in early grain development. In
of proteins extracted from grains harvested at 750◦ Cd after barley grain, a massive transcriptional reprogramming occurs
anthesis. during this developmental transition (Sreenivasulu et al., 2004,
2010) and the proteins identified in the present study may be
potential regulators of this key stage.
Constant Need for Ribosome Biogenesis Identified histones had a higher relative abundance between
Proteins Increasing at the End of Grain 150 and 450◦ Cd after anthesis than later during grain
Development development and more variant types were also detected during
Among the 18 ribosomal proteins identified, six were present in this period. This was surprising as some histones have previously
multiple spots and for 10 at least one of their protein spots varied been shown to remain constant throughout wheat grain
in relative abundance during grain development. These variant development with the synthesis of histones at 3 dap ending
proteins had abundance profiles 3, 4, 5, or 6. Thus, ribosome around 16 dap, approximately 320◦ Cd after anthesis (Spiker et al.,
formation seems to occur in each phase of grain development. 1987).
However, more different proteins were present in profiles 4, 5, Several proteins accumulated in the nucleus at the end of
and 6 (nine different proteins in total) than in the other three grain development. For example, arginine/serine rich splicing
profiles (one protein) and were thus more abundant as a class factor which is known to be localized in nuclear specks and
after 450◦ Cd after anthesis. Possibly there is a high demand to be part of the spliceosome (Tillemans et al., 2006). Serpin
for ribosome synthesis during the second half of the effective proteins, which negatively regulate proteases, were also more
grain-filling and early maturation phases. This is somewhat abundant in late development, as previously observed in the

Frontiers in Plant Science | [Link] 9 October 2015 | Volume 6 | Article 905

71
Bonnot et al. Nuclear proteome of wheat grain

FIGURE 7 | Schematic view of some highlighted nuclear proteins accumulated during early and late grain development. These proteins were extracted
from wheat grain at six thermal times after anthesis, from 150 to 750◦ Cd and analyzed on 2-D gel electrophoresis. The color code for normalized volume of
correspondent protein spots is indicated at the bottom left.

endosperm of developing wheat grain (Tasleem-Tahir et al., have an overview of some of the quantitative changes occurring
2012). Interestingly, proteins involved in proteolysis were more in 2D nuclear proteome of the developing wheat grain. This study
abundant between 150 and 450◦ Cd after anthesis, before serpins revealed that the dynamics of the nuclear proteome of wheat
accumulate. Proteins with a role in mRNA maturation or in grain seems to be divided into two periods (Figure 7). The first
protection against degradation might regulate processes at the phase corresponds to the cellularization and early effective grain-
end of the grain-filling phase. filling phases, during which a change in transcription regulation
occurs with a high abundance of proteins involved in mRNA
Do Non-variant Nuclear Proteins have a processing. The second phase corresponds to the end of the
Regulatory Role During Grain effective grain-filling phase and the early maturation phase, when
Development? there is an activation of ribosome synthesis and an increase
Proteins which did not vary are likely to be essential throughout in proteins inhibiting protease action. This study opens the
grain development and may still regulate grain development. way for more precise research into the regulatory mechanisms
An example is the FACT complex subunit SSRP1-B, which like that govern the accumulation of starch, storage proteins, and
HMG1-2 facilitates the formation of nucleoprotein structures micronutrients that determine the processing and health value of
(Röttgers et al., 2000). It may also act in protein complexes to cereal grains.
control transcription mechanisms modulating the properties of
chromatin. Another example is the histone deacetylase HDT2, ACKNOWLEDGMENTS
which could repress transcription in the same manner as Z. mays
HDAC2, by forming a complex of three polypeptides (Hollender The authors thank Didier Viala (INRA, Clermont-Ferrand)
and Liu, 2008). These proteins probably play an important for help with MS analyses and protein identification, Joelle
role during grain development even though we didn’t see any Henry-Berger and Christelle Damon (Blaise Pascal University,
change in their abundance. Interestingly, the histone deacetylase Clermont-Ferrand) for help with microscopic observation, and
HDT2 protein spot was stained with both CBB and Pro-Q Dr. Philippe Leroy (INRA, Clermont-Ferrand) for compiling
Diamond R at 150 and 250◦ Cd after anthesis, indicating that the database for functional annotation of proteins. This work
this protein was phosphorylated at the end of the cellularization was supported by a Ph.D. grant from the French Ministry for
phase (Supplementary Figure 3). Perhaps post-translational Higher Education and Research to TB and funding from the
modification such as phosphorylation regulates these proteins French Government managed by the Research National Agency
during grain development. (ANR) in the framework of Investments for the Future (ANR-10-
BTBR-03), France AgriMer and the French Fund to support Plant
CONCLUDING REMARKS Breeding (FSOV).

Some nuclear proteins are central actors in biological processes SUPPLEMENTARY MATERIAL
that regulate seed development. This study identified 114
different wheat proteins with various functions and dynamics, The Supplementary Material for this article can be found
some of which had been found in previous studies of nuclear online at: [Link]
proteomes of other plant species and organs. For the first time, we 00905

Frontiers in Plant Science | [Link] 10 October 2015 | Volume 6 | Article 905

72
Bonnot et al. Nuclear proteome of wheat grain

REFERENCES Horton, P., Park, K.-J., Obayashi, T., Fujita, N., Harada, H., Adams-Collier, C. J.,
et al. (2007). WoLF PSORT: protein localization predictor. Nucleic Acids Res.
Aki, T., and Yanagisawa, S. (2009). Application of rice nuclear proteome analysis to 35, W585–W587. doi: 10.1093/nar/gkm259
the identification of evolutionarily conserved and glucose-responsive nuclear Hoshikawa, K. (1962). Studies on the development of endosperm in rice. 11.
proteins research articles. J. Proteome Res. 8, 3912–3924. doi: 10.1021/ Development of starch granules in endosperm tissue. Proc. Crop Sci. Soc. Jpn
pr900187e 37, 207–216. doi: 10.1626/jcs.37.207
Ashburner, M., Ball, C. A., Blake, J. A., Botstein, D., Butler, H., Cherry, J. M., et al. Husson, F., Josse, J., Le, S., and Mazet, J. (2014). Package “FactoMineR.” Available
(2000). Gene ontology: tool for the unification of biology. The gene ontology online at: [Link]
consortium. Nat. Genet. 25, 25–29. doi: 10.1038/75556 Jaiswal, D. K., Ray, D., Choudhary, M. K., Subba, P., Kumar, A., Verma,
Bancel, E., Bonnot, T., Davanture, M., Branlard, G., Zivy, M., and Martre, P. (2015). J., et al. (2013). Comparative proteomics of dehydration response in
Proteomic approach to identify nuclear proteins in wheat grain. J. Proteome Res. the rice nucleus: new insights into the molecular basis of genotype-
14, 4432–4439. doi: 10.1021/[Link].5b00446 specific adaptation. Proteomics 13, 3478–3497. doi: 10.1002/pmic.
Blum, T., Briesemeister, S., and Kohlbacher, O. (2009). MultiLoc2: integrating 201300284
phylogeny and Gene Ontology terms improves subcellular protein localization Jerzmanowski, A., Przewłoka, M., and Grasser, K. D. (2000). Linker Histones and
prediction. BMC Bioinformatics 10:274. doi: 10.1186/1471-2105-10-274 HMG1 proteins of higher plants. Plant Biol. 2, 586–597. doi: 10.1055/s-2000-
Boisvert, F.-M., van Koningsbruggen, S., Navascués, J., and Lamond, A. I. (2007). 16648
The multifunctional nucleolus. Nat. Rev. Mol. Cell Biol. 8, 574–585. doi: Kanehisa, M., Goto, S., Sato, Y., Kawashima, M., Furumichi, M., and Tanabe, M.
10.1038/nrm2184 (2014). Data, information, knowledge and principle: back to metabolism in
Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of KEGG. Nucleic Acids Res. 42, D199–D205. doi: 10.1093/nar/gkt1076
microgram quantities of protein utilizing the principle of protein-dye binding. Khan, M. M. K., and Komatsu, S. (2004). Rice proteomics: recent developments
Anal. Biochem. 72, 248–254. doi: 10.1016/0003-2697(76)90527-3 and analysis of nuclear proteins. Phytochemistry 65, 1671–1681. doi:
Briesemeister, S., Rahnenführer, J., and Kohlbacher, O. (2010). YLoc–an 10.1016/[Link].2004.04.012
interpretable web server for predicting subcellular localization. Nucleic Acids Li, G., Nallamilli, B. R. R., Tan, F., and Peng, Z. (2008). Removal of high-abundance
Res. 38, W497–W502. doi: 10.1093/nar/gkq477 proteins for nuclear subproteome studies in rice (Oryza sativa) endosperm.
Calikowski, T. T., Meulia, T., and Meier, I. (2003). A proteomic study Electrophoresis 29, 604–617. doi: 10.1002/elps.200700412
of the arabidopsis nuclear matrix. J. Cell. Biochem. 90, 361–378. doi: Liu, X., Yang, S., Zhao, M., Luo, M., Yu, C.-W., Chen, C.-Y., et al. (2014).
10.1002/jcb.10624 Transcriptional repression by histone deacetylases in plants. Mol. Plant 7,
Calogero, S., Grassi, F., Aguzzi, A., Voigtländer, T., Ferrier, P., Ferrari, S., et al. 764–772. doi: 10.1093/mp/ssu033
(1999). The lack of chromosomal protein Hmg1 does not disrupt cell growth Mares, D., Norstog, K., and Stone, B. (1975). Early stages in the development of
but causes lethal hypoglycaemia in newborn mice. Nat. Genet. 22, 276–280. doi: wheat endosperm. I. The change from free nuclear to cellular endosperm. Aust.
10.1038/10338 J. Bot. 23, 311–326. doi: 10.1071/BT9750311
Chojecki, A. J. S., Bayliss, M. W., and Gale, M. D. (1986). Cell production and DNA Neuhoff, V., Stamm, R., and Eibl, H. (1985). Clear background and highly sensitive
accumulation in the wheat endosperm, and their association with grain weight. protein staining with Coomassie Blue dyes in polyacrylamide gels: a systematic
Ann. Bot. 58, 809–817. analysis. Electrophoresis 6, 427–448. doi: 10.1002/elps.1150060905
Choudhary, M. K., Basu, D., Datta, A., Chakraborty, N., and Chakraborty, Nollen, E. A. A., Salomons, F. A., Brunsting, J. F., van der Want, J. J. L.,
S. (2009). Dehydration-responsive nuclear proteome of rice (Oryza sativa Sibon, O. C. M., and Kampinga, H. H. (2001). Dynamic changes in the
L.) illustrates protein network, novel regulators of cellular adaptation, localization of thermally unfolded nuclear proteins associated with chaperone-
and evolutionary perspective. Mol. Cell. Proteomics 8, 1579–1598. doi: dependent protection. Proc. Natl. Acad. Sci. U.S.A. 98, 12038–12043. doi:
10.1074/mcp.M800601-MCP200 10.1073/pnas.201112398
Dahlberg, J. E., Lund, E., and Goodwin, E. B. (2003). Nuclear translation: what is Pandey, A., Choudhary, M. K., Bhushan, D., Chattopadhyay, A., Chakraborty, S.,
the evidence? RNA 9, 1–8. doi: 10.1261/rna.2121703 Datta, A., et al. (2006). The nuclear proteome of chickpea (Cicer arietinum L.)
David, A., Dolan, B. P., Hickman, H. D., Knowlton, J. J., Clavarino, G., Pierre, P., reveals predicted and unexpected proteins. J. Proteome Res. 5, 3301–3311. doi:
et al. (2012). Nuclear translation visualized by ribosome-bound nascent chain 10.1021/pr060147a
puromycylation. J. Cell Biol. 197, 45–57. doi: 10.1083/jcb.201112145 Petrovská, B., Jeřábková, H., Chamrád, I., Vrána, J., Lenobel, R., Uřinovská, J.,
Erhardt, M., Adamska, I., and Franco, O. L. (2010). Plant nuclear proteomics– et al. (2014). Proteomic analysis of barley cell nuclei purified by flow sorting.
inside the cell maestro. FEBS J. 277, 3295–3307. doi: 10.1111/j.1742- Cytogenet. Genome Res. 143, 78–86. doi: 10.1159/000365311
4658.2010.07748.x Petrovská, B., Šebela, M., and Doležel, J. (2015). Inside a plant nucleus: discovering
Ferreira, B. S., Branco, A. T., de Oliveira, M. A., Pereira, M. G., and de the proteins. J. Exp. Bot. 66, 1627–1640. doi: 10.1093/jxb/erv041
Souza Filho, G. A. (2006). Methodological improvements on extraction Rajjou, L., Gallardo, K., Debeaujon, I., Vandekerckhove, J., Job, C., and Job, D.
of nuclear proteins and its preliminary analysis during the maize (Zea (2004). The Effect of α-Amanitin on the Arabidopsis seed proteome highlights
mays L.) endosperm development. Protein Pept. Lett. 13, 981–984. doi: the distinct roles of stored and neosynthesized mRNAs during germination.
10.2174/092986606778777470 Plant Physiol. 134, 1598–1613. doi: 10.1104/pp.103.036293
Ferreira, M. S. L., Martre, P., Mangavel, C., Girousse, C., Rosa, N. N., Samson, R Core Team (2013). R: A Language and Environment for Statistical Computing. R
M.-F., et al. (2012). Physicochemical control of durum wheat grain filling and Foundation for Statistical Computing, Vienna.
glutenin polymer assembly under different temperature regimes. J. Cereal Sci. Repetto, O., Rogniaux, H., Firnhaber, C., Zuber, H., Küster, H., Larré, C.,
56, 58–66. doi: 10.1016/[Link].2011.11.001 et al. (2008). Exploring the nuclear proteome of Medicago truncatula at
Fromont-Racine, M., Senger, B., Saveanu, C., and Fasiolo, F. (2003). Ribosome the switch towards seed filling. Plant J. 56, 398–410. doi: 10.1111/j.1365-
assembly in eukaryotes. Gene 313, 17–42. doi: 10.1016/S0378-1119(03)00629-2 313X.2008.03610.x
Goldberg, T., Hecht, M., Hamp, T., Karl, T., Yachdav, G., Ahmed, N., et al. (2014). Repetto, O., Rogniaux, H., Larré, C., Thompson, R., and Gallardo, K. (2012). The
LocTree3 prediction of localization. Nucleic Acids Res. 42, W350–W355. doi: seed nuclear proteome. Front. Plant Sci. 3:289. doi: 10.3389/fpls.2012.00289
10.1093/nar/gku396 Romeuf, I., Tessier, D., Dardevet, M., Branlard, G., Charmet, G., and Ravel,
Guo, B., Chen, Y., Li, C., Wang, T., Wang, R., Wang, B., et al. (2014). Maize C. (2010). wDBTF: an integrated database resource for studying wheat
(Zea mays L.) seedling leaf nuclear proteome and differentially expressed transcription factor families. BMC Genomics 11:185. doi: 10.1186/1471-2164-
proteins between a hybrid and its parental lines. Proteomics 14, 1071–1087. doi: 11-185
10.1002/pmic.201300147 Röttgers, K., Krohn, N. M., Lichota, J., Stemmer, C., Merkle, T., and Grasser, K. D.
Hollender, C., and Liu, Z. (2008). Histone deacetylase genes in Arabidopsis (2000). DNA-interactions and nuclear localisation of the chromosomal HMG
development. J. Integr. Plant Biol. 50, 875–885. doi: 10.1111/j.1744- domain protein SSRP1 from maize. Plant J. 23, 395–405. doi: 10.1046/j.1365-
7909.2008.00704.x 313x.2000.00801.x

Frontiers in Plant Science | [Link] 11 October 2015 | Volume 6 | Article 905

73
Bonnot et al. Nuclear proteome of wheat grain

Sabelli, P. A, and Larkins, B. A (2009). The development of endosperm in grasses. Tan, F., Li, G., Chitteti, B. R., and Peng, Z. (2007). Proteome and phosphoproteome
Plant Physiol. 149, 14–26. doi: 10.1104/pp.108.129437 analysis of chromatin associated proteins in rice (Oryza sativa). Proteomics 7,
Schnyder, H., and Baum, U. (1992). Growth of the grain of wheat (Triticum 4511–4527. doi: 10.1002/pmic.200700580
aestivum L.). The relationship between water content and dry matter Tasleem-Tahir, A., Nadaud, I., Chambon, C., and Branlard, G. (2012).
accumulation. Eur. J. Agron. 1, 51–57. doi: 10.1016/S1161-0301(14)80001-4 Expression profiling of starchy endosperm metabolic proteins at 21 stages
Sengupta, N., and Seto, E. (2004). Regulation of histone deacetylase activities. of wheat grain development. J. Proteome Res. 11, 2754–2773. doi: 10.1021/
J. Cell. Biochem. 93, 57–67. doi: 10.1002/jcb.20179 pr201110d
Sha, Z., Brill, L. M., Cabrera, R., Kleifeld, O., Scheliga, J. S., Glickman, M. H., Tillemans, V., Leponce, I., Rausin, G., Dispa, L., and Motte, P. (2006).
et al. a (2009). The eIF3 interactome reveals the translasome, a supercomplex Insights into nuclear organization in plants as revealed by the dynamic
linking protein synthesis and degradation machineries. Mol. Cell 36, 141–152. distribution of Arabidopsis SR splicing factors. Plant Cell 18, 3218–3234. doi:
doi: 10.1016/[Link].2009.09.026 10.1105/tpc.106.044529
Shevchenko, A., Wilm, M., Vorm, O., and Mann, M. (1996). Mass spectrometric Vizcaíno, J. A., Deutsch, E. W., Wang, R., Csordas, A., Reisinger, F., Ríos, D.,
sequencing of proteins from silver-stained polyacrylamide gels. Anal. Chem. et al. (2014). ProteomeXchange provides globally coordinated proteomics
68, 850–858. doi: 10.1021/ac950914h data submission and dissemination. Nat. Biotechnol. 32, 223–226. doi:
Shewry, P. R., Mitchell, R. A. C., Tosi, P., Wan, Y., Underwood, C., Lovegrove, A., 10.1038/nbt.2839
et al. (2012). An integrated study of grain development of wheat (cv. Hereward). Willard, F. S., and Crouch, M. F. (2000). Nuclear and cytoskeletal translocation and
J. Cereal Sci. 56, 21–30. doi: 10.1016/[Link].2011.11.007 localization of heterotrimeric G-proteins. Immunol. Cell Biol. 78, 387–394. doi:
Silverman, G. A., Bird, P. I., Carrell, R. W., Church, F. C., Coughlin, P. B., Gettins, 10.1046/j.1440-1711.2000.00927.x
P. G., et al. (2001). The serpins are an expanding superfamily of structurally Zhang, S., Xia, X., Shen, J., Zhou, Y., and Sun, Z. (2008). DBMLoc: a Database of
similar but functionally diverse proteins. Evolution, mechanism of inhibition, proteins with multiple subcellular localizations. BMC Bioinformatics 9:127. doi:
novel functions, and a revised nomenclature. J. Biol. Chem. 276, 33293–33296. 10.1186/1471-2105-9-127
doi: 10.1074/jbc.R100016200
Spiker, S., Hopkins, R., Fischer, R., and Quatrano, R. S. (1987). Synthesis Conflict of Interest Statement: The authors declare that the research was
of nucleosomal histone variants during wheat grain development. Biochim. conducted in the absence of any commercial or financial relationships that could
Biophys. Acta Gene Struct. Expr. 910, 157–162. doi: 10.1016/0167-4781(87) be construed as a potential conflict of interest.
90068-6
Sreenivasulu, N., Altschmied, L., Radchuk, V., Gubatz, S., Wobus, U., and Copyright © 2015 Bonnot, Bancel, Chambon, Boudet, Branlard and Martre. This
Weschke, W. (2004). Transcript profiles and deduced changes of metabolic is an open-access article distributed under the terms of the Creative Commons
pathways in maternal and filial tissues of developing barley grains. Plant J. 37, Attribution License (CC BY). The use, distribution or reproduction in other forums
539–553. doi: 10.1046/j.1365-313X.2003.01981.x is permitted, provided the original author(s) or licensor are credited and that the
Sreenivasulu, N., Borisjuk, L., Junker, B. H., Mock, H.-P., Rolletschek, H., Seiffert, original publication in this journal is cited, in accordance with accepted academic
U., et al. (2010). Barley grain development toward an integrative view. Int. Rev. practice. No use, distribution or reproduction is permitted which does not comply
Cell Mol. Biol. 281, 49–89. doi: 10.1016/S1937-6448(10)81002-0 with these terms.

Frontiers in Plant Science | [Link] 12 October 2015 | Volume 6 | Article 905

74
 Chapitre 3 : Le protéome nucléaire du grain en développement de blé tendre

Supplementary information

A.

B.
Supplementary Figure 1. Validation of nuclei purification. Nuclei were isolated from wheat grains
harvested between 150°C and 750°Cd after anthesis. A. Nuclei were observed under fluorescence
microscopy (Zeiss Axioplan 2 microscope). Scale bar represents 50 µm sizes.
B. Chlorophyll assay was performed on supernatants (S) collected during washing steps of nuclei and
on nuclei extracts (N).

Supplementary Figure 2. Principal component analysis of protein spots detected by image

analysis. Normalized volumes of protein spots detected in the four replicates were collected. 1-4:
replicates at 150°Cd; 5-8: replicates at 250°Cd; 9-12: replicates at 350°Cd; 13-16: replicates at 450°Cd;
17-20: replicates at 600°Cd; 21-24: replicates at 750°Cd.

75
 Chapitre 3 : Le protéome nucléaire du grain en développement de blé tendre

Supplementary Figure 3. Some nuclear protein spots phosphorylated during division /

differentiation phase of wheat grain development. 2D gels were stained with Pro-Q Diamond (Pro-
Q, Invitrogen) according to the manufacturer’s instructions and adapted (Agrawal and Thelen, 2005)
then stained with Coomassie Brilliant Blue G250 (CBB). Proteins spots stained both with Pro-Q and
CBB were considered as phosphorylated. a: HMG1/2-like protein (spot 57) was phosphorylated at
150°Cd after anthesis. b: Histone deacetylase HDAC2 was phosphorylated at 150°Cd (spot 135, 136)
and 250°Cd after anthesis (spot 135); Histone deacetylase HDT2 was phosphorylated at 150 and 250°Cd
(spot 137) after anthesis.

Reference:
Agrawal, G. K., and Thelen, J. J. (2005). Development of a simplified, economical polyacrylamide gel
staining protocol for phosphoproteins. Proteomics 5, 4684–8. doi:10.1002/pmic.200500021.

Supplementary table 1 and table 2 can be viewed online at:

[Link]

76
 Chapitre 3 : Le protéome nucléaire du grain en développement de blé tendre

3. CONCLUSIONS
Cette étude a permis pour la première fois d’analyser les variations d’abondance des
protéines nucléaires au cours du développement du grain de blé. D’un point de vue qualitatif,
les fonctions représentées au sein des 114 protéines identifiées sont comparables à celles
retrouvées dans le chapitre précédent. En effet, nous avons retrouvé de nombreuses protéines
ribosomales ainsi que plusieurs protéines histones, ce qui était attendu. Plusieurs protéines liées
à la conformation de l’ADN, à la transcription et sa régulation, ou à la maturation des ARNm
ont également été identifiées.
La principale difficulté rencontrée dans cette étude a été la présence de protéines de
réserve en quantité non négligeable dans les extraits nucléaires, représentant ainsi des
contaminants. Du fait de leur présence sous plusieurs isoformes, ces protéines ont été identifiées
dans de nombreux spots protéiques prélevés sur les gels bidimensionnels. Ces spots étaient
particulièrement abondants à 600 et 750°Cj après floraison. Les protéines de réserve, en très
grande abondance dans le grain en fin de phase de remplissage posent donc un problème pour
l’analyse du protéome nucléaire. L’hypothèse que nous émettons est que les corpuscules
protéiques dans lesquels s’accumulent les protéines de réserve du grain ont pu être co-purifiés
avec les noyaux sur le gradient de densité car ils auraient à ces stades une densité proche de
celle des noyaux cellulaires. De plus, à cette période du développement du grain les noyaux
sont probablement plus difficilement purifiables du fait de leur dégradation progressive dans la
mort cellulaire programmée, orchestrée dans l’albumen au cours de la phase de maturation
(Ferreira et al., 2012 ; Dominguez and Cejudo, 2014). Nous avons ainsi fait le choix pour les
prochaines analyses du protéome nucléaire de n’extraire les noyaux que pour des stades
antérieurs à 600°Cj après floraison.
Malgré cette difficulté, nous avons pu voir que les quantités de protéines nucléaires
extraites étaient comparables pour un même nombre de grains entre 150 et 600°Cj après
floraison. L’analyse des images de gels et les analyses statistiques ont démontré qu’une grande
majorité des spots protéiques révélés sur les gels d’électrophorèse a présenté des variations
d’abondance sur la cinétique considérée. Le stade de développement du grain a donc un effet
très important sur la dynamique du protéome nucléaire. Au vu de leur profil d’accumulation,
certaines protéines liées à la conformation de l’ADN (protéine HMG, histone déacétylase)
pourraient jouer un rôle important dans la transition entre la phase de cellularisation et la phase
de remplissage.
Les fortes variations quantitatives observées au niveau des protéines nucléaires pendant
le développement du grain nous ont ainsi laissé entrevoir la possibilité d’étudier la réponse de
ce sous-protéome à d’autres facteurs tels que la nutrition de la plante.

Références :
Dominguez, F. and Cejudo, F.J. (2014) Programmed cell death (PCD): an essential process
of cereal seed development and germination. Front. Plant Sci., 5, 366.
Ferreira, M.S.L., Martre, P., Mangavel, C., Girousse, C., Rosa, N.N., Samson, M.-F. and
Morel, M.-H. (2012) Physicochemical control of durum wheat grain filling and glutenin
polymer assembly under different temperature regimes. J. Cereal Sci., 56, 58–66.

77
78
 CHAPITRE 4 :
La réponse protéomique du grain
de Triticum monococcum à la
nutrition azotée et soufrée

79
80
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

1. INTRODUCTION
Les effets de l’N et du S sur la quantité totale de PR et sur la proportion des différentes
classes et sous-unités de PR du grain de blé ont largement été démontrés. En revanche, les
mécanismes moléculaires qui interviennent entre la nutrition du grain et l’ajustement de la
synthèse des PR ont été peu caractérisés. Les approches -omiques nous sont apparues comme
les plus adaptées pour apporter de nouveaux éléments à la compréhension de ces mécanismes.
Notre hypothèse de travail a été la suivante : les quantités d’N et de S disponibles pour le grain
lors du remplissage affectent l’abondance de nombreux transcrits, protéines et métabolites
interconnectés, qui agiraient en amont de l’expression des gènes de PR.
Afin de visualiser la réponse du grain à la nutrition azotée et soufrée, nous avons conduit
une expérimentation chez Triticum monococcum. Cette espèce de blé est un bon modèle d’étude
du fait de son génome diploïde, plus simple que le génome hexaploîde du blé tendre et en même
temps très proche de son sous-génome A. De plus, les PR présentes chez cette espèce sont les
mêmes que celles du blé tendre. Dans cette expérimentation, quatre nutritions contenant des
quantités différentes en N et S ont été appliquées à la culture à partir de 200°Cj après floraison,
c’est-à-dire en fin de phase de cellularisation du grain (Figure 23). La première nutrition ne
contenait ni apport d’N ni de S et a été appelée N-S-, la seconde, nommée N+S-, correspondait
à une nutrition azotée dépourvue de soufre (6mM N, 0mM S), la troisième, N-S+, était une
nutrition soufrée (0,5mM N, 2mM S) et la quatrième nutrition, N+S+, combinait des fortes
doses des deux éléments (6mM N, 2mM S).
Les grains ont été récoltés tous les 100°Cj après floraison dans le but de réaliser
différentes analyses omiques (protéomique, transcriptomique, métabolomique). Nous avons fait
le choix de traiter ces données dans deux études différentes :
- La première étude correspond à la réponse protéomique du grain à la nutrition et sera décrite
dans ce chapitre 4.
- La seconde correspond à l’analyse du transcriptome et des métabolites. Afin d’obtenir une
vision intégrative de la réponse du grain à la nutrition, ces données ont été intégrées aux données
de protéomique obtenues dans le chapitre 4. Ce travail est traité dans le chapitre 5.

Figure 23. Schéma

expérimental de l’étude
de la réponse du grain de
Triticum monococcum à
la nutrition azotée et
soufrée.

81
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

Dans ce chapitre 4 qui concerne la réponse protéomique du grain à la nutrition azotée et

soufrée, nous avons ciblé différents sous-protéomes : les PR, les albumines globulines qui
correspondent majoritairement à des enzymes ou inhibiteurs d’enzymes, et les protéines
nucléaires. Nous avons ainsi voulu visualiser suite à une modification de la nutrition de la plante
les changements qui s’opèrent au niveau du métabolisme cellulaire et des régulateurs
transcriptionnels, changements qui pourraient influencer la synthèse des PR. Les trois sous-
protéomes ont tous été analysés entre 300 et 500°Cj, c’est-à-dire pendant le remplissage du
grain et après l’application des quatre traitements. Les PR ont elles été quantifiées jusqu’à
maturité du grain pour suivre leurs profils d’accumulation et observer la composition du grain
mature. D’autres stades ont été analysés pour obtenir un maximum d’information sur
l’identification et la quantification des protéines (100 et 200°Cj après floraison pour les
protéines nucléaires, 200 et 600°Cj après floraison pour les albumines globulines). Comme
nous l’avons évoqué, nous avons fait le choix de ne pas extraire les protéines nucléaires des
grains au-delà de 500°Cj pour des raisons techniques (cf conclusion chapitre 3). Pour identifier
et quantifier les protéines nucléaires et albumines globulines, nous avons opté pour une analyse
des échantillons protéiques en spectrométrie de masse, sans passer au préalable par une étape
de séparation sur gel d’électrophorèse (gel free).

2. ARTICLE 3 : Grain subproteome responses to nitrogen and sulfur supply in diploid

wheat Triticum monococcum ssp. monococcum
Titouan Bonnot1, Emmanuelle Bancel1, David Alvarez1, Marlène Davanture2, Julie
Boudet1, Marie Pailloux3, Michel Zivy2, Catherine Ravel1 and Pierre Martre1
1
UMR GDEC, INRA, Université Blaise Pascal, 5 chemin de Beaulieu, 63 177 Aubière, France
2
UMR GQE, INRA, Université Paris-Sud, CNRS, AgroParisTech, Université Paris-Saclay,
91 190 Gif-sur-Yvette, France
3
LIMOS, CNRS, Université Blaise Pascal, 63 173 Aubière, France
†
Present address: UMR LEPSE, INRA, Montpellier SupAgro, 34 060 Montpellier, France
(Article soumis à Plant Journal)

82
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

Grain subproteome responses to nitrogen and sulfur supply in diploid wheat

Triticum monococcum ssp. monococcum
Titouan Bonnot1, Emmanuelle Bancel1, David Alvarez1, Marlène Davanture2, Julie Boudet1,
Marie Pailloux3, Michel Zivy2, Catherine Ravel1,* and Pierre Martre1,†
1
UMR GDEC, INRA, Université Blaise Pascal, 5 chemin de Beaulieu, 63 177 Aubière, France
3
UMR GQE, INRA, Université Paris-Sud, CNRS, AgroParisTech, Université Paris-Saclay, 91 190 Gif-sur-
Yvette, France
4
LIMOS, CNRS, Université Blaise Pascal, 63 173 Aubière, France
†
Present address: UMR LEPSE, INRA, Montpellier SupAgro, 34060 Montpellier, France

SUMMARY
Wheat grain storage proteins (GSPs) make up most of the protein content of grain and determine flour end-
use value. The synthesis and accumulation of GSPs depend highly on nitrogen (N) and sulfur (S) availability
and it is important to understand the underlying control mechanisms. Here we studied how the einkorn
(Triticum monococcum ssp. monococcum) grain proteome responds to different amounts of N and S supply
during grain development. GSP composition at grain maturity was clearly impacted by nutrition treatments,
due to early changes in the rate of GSP accumulation during grain filling. Large-scale analysis of the nuclear
and albumin-globulin sub-proteomes during this key developmental phase revealed that the abundance of
203 proteins was significantly modified by the nutrition treatments. Our results showed that the grain
proteome was highly impacted by perturbation in the N:S balance. S supply strongly increased the rate of
accumulation of S-rich α/β-gliadin and γ-gliadin, and of several other proteins involved in glutathione
metabolism. Post-anthesis N supply resulted in the activation of amino acid metabolism at the expense of
carbohydrate metabolism and the activation of transport processes including nucleocytoplasmic transit.
Protein accumulation networks were analyzed. Several central actors in the response were identified whose
variation in abundance was related to variation in the amounts of many other proteins and are thus potentially
important for GSP accumulation. This detailed analysis of grain sub-proteomes provides information on
how wheat GSP composition can possibly be controlled in low-level fertilization condition.
Keywords: einkorn, grain, nitrogen, sulfur, storage proteins, nuclear proteins, albumin-globulin, network.

INTRODUCTION and gliadins are monomeric proteins which confer

extensibility (Branlard et al., 2001). The GSP
Bread wheat (Triticum aestivum L.) is the main
composition of wheat grain thus determines its use-
stable crop in many regions of the world and
end value. Commonly, gliadins are subdivided into
globally provides about 20% of calories and
ω1,2-, ω5-, α/β- and γ-gliadin classes and glutenins
proteins in the human diet. In mature grain, 10-15%
are classified as high-molecular-weight subunits
of the dry mass is protein. GSPs make up 60-80%
(HMW-GS) or low-molecular-weight subunits
of the total protein in grain and metabolic proteins
(LMW-GS; Wieser, 2007). These different GSP
(the albumins and globulins) 15-20% (Shewry and
subclasses and subunits differ in the proportions of
Halford, 2002). GSPs accumulate during the
cysteine and methionine residues they contain, and
effective filling phase of grain development
are thus classified as being S-rich (α/β- and γ-
(Shewry et al., 2012). The two main GSP fractions
gliadins, LMW-GS) or S-poor (ω1,2- and ω5-
are glutenin and gliadin that when mixed together
gliadins, HMW-GS; Shewry et al., 1997; Shewry et
with water form gluten and are thus the main
al., 2001).
determinants of the rheological and bread-making
The availability of N and S in the soil highly
properties of wheat dough. More precisely,
influences the GSP composition of grain. N supply
glutenins form polymers that are mainly
increases the rate and duration of protein
responsible for the viscoelastic properties of dough

83
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

accumulation and so increases the proportion of S- quality determinants were expected to be nuclear
poor GSPs in mature grain (Shewry et al., 2001; localized (Verdier and Thompson, 2008). The
Triboï et al., 2003; Chope et al., 2014). In S- transcriptional regulation of GSP synthesis was
deficient conditions, GSP synthesis and confirmed by Dai et al. (2015) who reported
accumulation occurs earlier because the significantly linear relationships between
cellularization phase is shortened (Castle and regulatory gene expression and GSP accumulation
Randall, 1987). Mature grain contains low rates. Several nuclear proteome studies in other
concentrations of cysteine and methionine when S plant species have identified protein actors central
is deficient (Wrigley et al., 1980), causing S-rich to various stress responses (Yin and Komatsu,
GSPs to accumulate more slowly, which is 2015; Petrovská et al., 2015). Albumin-globulin
compensated by an increased rate of accumulation and nuclear proteins are thus two subproteomes that
of S-poor GSPs, particularly HMW-GS (Zhao et can be targeted as being involved in GSP synthesis
al., 1999; Wieser et al., 2004; Zörb et al., 2010). affected by N and S nutrition.
The molecular mechanisms that control GSP Einkorn (Triticum monococcum) is well tolerant
synthesis and composition in response to N and S of low input cropping and is appreciated for its
supply are not yet well understood (Dai et al., excellent nutritional properties due to its high
2015). A priority for wheat breeders is to improve protein, high carotenoid and high tocol contents
grain yield while reducing the need for fertilizers, (Corbellini et al., 1999; Hidalgo et al., 2006;
especially N input. There is a strongly negative Hidalgo and Brandolini, 2014). It is an ancestral
correlation between grain yield and grain protein wheat whose diploid genome is a sister to bread
concentration traits, so increasing grain yield is wheat (Triticum aestivum) genome A (Marcussen et
generally detrimental to grain protein concentration al., 2014). Genomic resources are now available for
and hence grain quality (Simmonds, 1995; Oury this species. Its genomic sequence is available at
and Godin, 2007). In addition, S deficiency is more [Link]
frequently observed in soils nowadays, partly C_WGS_assemblies_of_other_wheat_species/ and
because of reductions in industrial SO2 emissions its transcriptome was analyzed and annotated by
(Eriksen, 2009), hindering the synthesis of S-rich Fox et al. (2014). Therefore, einkorn appears to be
GSPs in wheat. Analyzing responses to N and S a suitable species in which to explore proteome
nutrition will reveal key actors that can be targeted responses to N and S nutrition.
in developing plants that produce high quality grain Our eventual aim is to identify key factors able to
in low-level fertilization conditions. ensure high quality grain (high protein content,
Large-scale proteomics and other omics adequate GSP composition) under low fertilization
approaches are commonly used to gain an overview levels. In this study we produced an overview of the
of changes taking place in an organism or tissue by T. monococcum grain proteome response to N and
identifying the proteins and genes involved in a S nutrition. Four post-anthesis N and S treatments
given biological response (Das et al., 2015). When were applied to a T. monococcum genotype grown
proteomics is used to study the grain response to in a glasshouse and grains were harvested at
nutrition, bioactive proteins are expected to be different developmental stages. GSPs were
found either in the albumin and globulin fraction or quantified and N and/or S supply was found to have
among regulatory proteins. Albumins and globulins significant effects on the amounts of gliadin classes
are mainly enzymes or enzyme inhibitors involved and glutenin subunits. In mature grains, α/β- and γ-
in cell metabolism and development and they may gliadins were increased by S supply, resulting in an
thus influence the rheological properties of wheat increased gliadin-to-glutenin ratio, while HMW-
flour (Hill et al., 2008). Albumin-globulin synthesis GS and ω1,2-gliadins were increased with N
depends on the nutritional status of the plant. supply. By focusing on nuclear and albumin-
Although the synthesis of such proteins is less globulin fractions during the effective grain filling
affected by N than that of GSPs (Wieser and phase, 203 proteins were highlighted as being
Seilmeier, 1998; Martre et al., 2003), N and S significantly affected by nutrition. Protein data
nutrition has been reported to affect several integration grouped proteins with similar
enzymes in wheat grain. For example, a accumulation behaviours, which may potentially
glyceraldehyde-3-phosphate dehydrogenase and a act synergistically in the grain. A few of these
serpin were both increased in conditions combining nuclear and albumin-globulin proteins are
high N with low S (Flæte et al., 2005). As GSP described in more detail as they potentially play a
synthesis is controlled at a transcriptional level and central role in GSP synthesis control in response to
many regulatory factors are nuclear proteins, N and S nutrition.

84
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

RESULTS 1b,d; Table S2). Mature grain contained more N

following high N treatment (N+S- and N+S+)
Grain storage protein composition is modified
compared with the low N and low S treatment (N-
by N and S supply
S-), mainly because N accumulation lasted longer
To study the response of nuclear, albumin- when more N was supplied (Figure 1b; Tables S1,
globulin and storage protein subproteomes of S2). Similarly, the higher quantity of S per grain at
developing einkorn wheat grains to N and S supply, maturity for the high S treatments (N-S+ and N+S+)
four treatments were applied to plants during the compared with the low S treatments (N+S- and N-
effective grain filling period. Plants were grown in S-) was mainly explained by the longer duration of
a greenhouse in conditions resembling those in the accumulation of grain S with high S supply (Figure
field. The four treatments were N-S-, N+S-, N-S+, 1d; Table S2). The duration of dry mass
N+S+ according to whether N and/or S were accumulation increased when more N and S was
applied. Dry mass and protein, N and S content of supplied, but lower rate of accumulation
grain harvested from ears every 100°Cd were compensated for this, so the final quantity of dry
measured (Figure 1). Temporal data were fitted mass per grain at maturity was not modified by the
with logistic function equations (eqn. 1) to illustrate treatments (Figure 1f; Table S2). Therefore, at
the rate and duration of compound accumulation maturity concentrations of grain protein and S
(Figure 1 b, d, f) which determine the total quantity differed between the treatments and reflected the
of storage compounds at maturity. changes in the quantity of N and S per grain (Figure
The treatments altering N and S availability 1a, c, e; Table S1). In N+S- grain, the N-to-S ratio
strongly modified the duration during which N and was higher than 30, an indication of severe S
S accumulated in grains, but the rates of deficiency (Figure 1e).
accumulation were much less modified (Figure

Figure 1. Effects of N and S supply

on the accumulation of total grain N,
S and dry mass (DM) in einkorn.
Total protein concentration (a), mass
of N per grain (b), S concentration
(c), mass of S per grain (d), N-to-S
ratio (e) and grain dry ass (f) were
measured. The four treatments N-S-,
N+S-, N-S+ and N+S+ were applied
from 200 to 700 °Cd after anthesis.
Data are means ± 1 s.e. for n = 4
independent replicates. Lines in (b),
(d) and (f) were fitted to the data
using eqn. (1).

85
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

The treatments also had significant effects on the (+106%), which more than compensated for the
rate and duration of GSP accumulation (Figure 2, shorter duration of accumulation (-21%) for N+S-
Table S2) and the final quantity in mature grains (Figure 2b; Table S2). No significant difference
(Figure S1). At maturity, there was 67% more α/β- between the treatments was measured at grain
and 163% more γ-gliadin per grain for the high S maturity for the S-poor ω5-gliadin and the S-rich
treatments (N-S+ and N+S+) compared with the LMW-GS. For LMW-GS, the effects of N and S on
low S treatments (N+S- and N-S-). These S-rich their rate of accumulation was compensated by
proteins accumulated at higher rates (+29% and opposing effects on the duration of their
+53%) and for longer durations (+30% and +13%) accumulation, which could explain this result
under the higher S supply (Figure 2c, e; Table S2). (Figure 2a; Table S2). Therefore, except for ω5-
The quantity per grain of the S-poor ω1,2-gliadin gliadin and LMW-GS, the variations in S-rich and
was 45% higher for N+S+ compared with N+S- S-poor GSP due to the different treatments reflected
because accumulation lasted longer (+242°Cd, ~ grain N-to-S ratios at maturity. The gliadin-to-
+11.7 d), while the rate of accumulation was similar glutenin ratio was different between treatments and
(Figure 2d; Table S2). For S-poor HMW-GS, there increased when grain N-to-S ratio decreased
was 52% more per grain for N+S- than for the other (Figure S1), modifying the GSP composition at
treatments because of a higher rate of accumulation maturity (Figure S1d).

Figure 2. Effects of N and S supply on the accumulation of LMW-GS and HMW-GS and α/β-, γ-, ω5- and ω1,2-
gliadin in einkorn grain. The four treatments N-S-, N+S-, N-S+ and N+S+ were applied from 200 to 700 °Cd after
anthesis. Data are means ± 1 s.e. for n = 4 independent replicates. Lines were fitted to the data using eqn. (1).

86
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

Analyzing nuclear and albumin-globulin protein process. In the albumin-globulin fraction, several
fractions gives high coverage of the grain proteins were involved in responses to stimulus
proteome (12%), protein metabolic processes (11%) and
Some of the mechanisms by which N and S carbohydrate metabolic processes (8%). In nuclear
supply influence GSP accumulation during grain extracts, many proteins were either ribosomal
filling must be governed by proteins. To investigate proteins or histones and were thus classified as
the grain proteome response to nutrition, two having functions in protein metabolic processes
subproteomes from the same biological material (17%) and cellular component organization (10%),
were analyzed by shotgun mass spectrometry: respectively. Several proteins involved in transport
nuclear proteins from 100 to 500°Cd after anthesis processes were also quantified (5% of the proteins
and albumin-globulin proteins from 200 to 600°Cd from nuclear and 6% from albumin-globulin
after anthesis. In total 1677 and 2475 proteins were subproteomes). A clear difference in GO term
identified in nuclear and albumin-globulin protein enrichment for molecular functions was found
extracts, respectively. These proteins were between albumin-globulin and nuclear proteins.
classified as participating in 24 biological processes Many albumin-globulin proteins possess catalytic
and 20 molecular functions according to their gene activity (20%), hydrolase activity (9%) or
ontology (GO) information. A total of 978 (24%) nucleotide binding activity (12%). In the nuclear
different proteins were accurately quantified (466 extract, an enrichment of proteins with DNA
nuclear proteins and 512 albumin-globulin binding (10%), nucleic acid binding (13%) and
proteins; Figure 3a, b). Overall 47% and 33% of all nucleotide binding (15%) activities was observed as
quantified nuclear proteins and albumin-globulin expected. These results highlighted the value of
respectively were classified as “uncharacterized” as independently analyzing the two subproteomes
they could not be assigned to a particular biological from the same biological material as the
information obtained was not redundant.

Figure 3. Functional classification of quantified nuclear and albumin-globulin proteins in einkorn grain. Proteins were
grouped as being involved in 24 biological processes (a) and 20 molecular functions (b) according to gene ontology
annotation. In (a) and (b) nuclear proteins and albumin-globulin are represented in the inner and outer circles,
respectively. (c) Venn diagram showing the numbers of quantified proteins in nuclear and albumin-globulin extracts.
Biological processes (d) and molecular functions (e) of the 69 proteins found in both nuclear and albumin-globulin
protein fractions.
87
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

Figure 4. Functions and response to nutrition profiles of proteins showing significant nutrition effect. (a) Percentage
of proteins in each biological process and molecular function classes for the 978 quantified proteins (open bars) and
for the 209 proteins (155 albumin-globulin proteins, 48 nuclear proteins, and 6 grain storage proteins) which differed
in abundance between the treatments (filled bars). (b) Boxplots of the normalized quantity of proteins in the four
clusters for the four treatments. Boxes show the 25th to 75th percentile range, horizontal lines in boxes show medians,
error bars outside boxes show the 10th to 90th percentile range, dots are outliers. The number of proteins in each cluster
is indicated at the top of each panel. alg, albumin-globulin, nucl, nuclear protein. (c) Heatmaps of the number of
proteins in different biological process and molecular function classes for the four clusters shown in (b).

88
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

Only 69 (7%) of the 978 quantified proteins were were highly abundant in the N+S- treated grain and
present in both fractions, showing again the very included the S-poor HMW-GS and ω1,2-gliadin.
low redundancy between the albumin-globulin and Cluster 3 grouped 53 proteins that were more
nuclear protein analyses (Figure 3c). Of these 69 abundant in response to N-S+ and included the S-
proteins, 20 (29%) were ribosomal proteins, rich γ-gliadin. Clusters 2 and 3 indicate a clear
classified as functioning in protein metabolic difference between N and S imbalances on the grain
processes and structural molecule activity (Figure proteome, as the 99 proteins grouped in these two
3d, e). Other proteins are involved in response to clusters behaved similarly in response to N-S- and
stimulus (11%) or transport processes (8%), with N+S+.
for example two importins (Figure 3d), and many Generally, proteins with similar molecular
possess nucleotide binding (12%) or RNA binding functions or acting in the same biological process
(9%) activities (Figure 3e). were found in several clusters. However, some
processes/functions were preferentially detected in
one, two or three clusters (Figure 4c). For example,
Various functions and response patterns in
proteins acting in carbohydrate metabolic processes
proteins impacted by N and S supply
or in response to stimulus were poorly represented
The abundance of 209 proteins (i.e. 21% of the in cluster 2, few of them being increased by high N
quantified proteins) was significantly affected by supply, unlike proteins involved in transport
the nutrition: the six GSPs, 155 albumin-globulin processes. Proteins acting in the generation of
proteins and 48 nuclear proteins; Table S5). A precursor metabolites and energy, also classified as
difference in GO term enrichment was observed acting in carbohydrate metabolic processes, were
between the 978 quantified proteins and the 209 mainly members of cluster 1 and thus decreased
nutrition-responsive proteins. Proteins acting in with high N and S supply. Nucleotide binding,
carbohydrate metabolic processes, generation of RNA binding and DNA binding proteins were
precursor metabolites and energy, or storage mostly found in cluster 4, and were nuclear proteins
functions were more represented among the that mostly increased by high N and S supply.
proteins whose abundance was modified by N
and/or S supply (Figure 4a). Conversely, protein
Network analysis reveals central actors in the
metabolic process (e.g. many ribosomal proteins)
grain proteome response to nutrition
and cellular component organization process (e.g.
many histones), which made up a high proportion To provide an integrative view of the grain
of the quantified proteome (22.7% and 11.8%, response to N and S supply and highlight proteins
respectively) were less well represented among the that co-accumulate during grain filling, a directed
proteins whose abundance was modified by N multilevel network was constructed using
and/or S supply (11.1% and 4.9%, respectively). proteomic data for the 209 proteins whose
Albumin-globulin proteins were proportionally abundance was modified by N and/or S supply from
more affected by N and S supply (30.2%) than 300 to 500°Cd after anthesis. We used a data
nuclear proteins (10.3%), with enrichment of GO mining method based on the discovery of
terms for catalytic, transferase, hydrolase or association rules between attributes (Agier et al.,
enzyme regulator activities (Figure 4a). Consistent 2007; Vincent et al., 2015). Associations between
with this result, DNA and RNA binding activities proteins and comparisons between treatments were
were underrepresented among the proteins whose built according to individual protein amounts in the
abundance was nutrition-responsive. different samples. The resulting network comprised
To analyze how the abundance of the 209 proteins 206 significant linkages (edges) and 111 proteins
responded to N and/or S supply, clustering analysis (nodes), i.e. about 50% of the 209 input proteins,
was performed and 4 clusters were established and included the six GSPs, 25 nuclear proteins, and
(Figure 4b). Clusters 1 and 4 grouped proteins with 80 albumin-globulin proteins (Figure 5a; Table S5).
opposite response profiles. Cluster 1 included 45 Networked proteins showed significant and high
proteins that were much less abundant in response differences in abundance between at least one pair
to N+S+, none of which were GSPs. On the of treatments. The comparison between N+S- and
contrary, cluster 4, the largest cluster with 65 N-S+ treatments created the highest number of
proteins, grouped proteins which became more edges (61, 29.6%) suggesting that the effect of these
abundant in response to N+S+ and included α/β- treatments on the grain proteome differed the most.
gliadin, ω5-gliadin, LMW-GS and 35 (73%)
nuclear proteins. Cluster 2 grouped 46 proteins that

89
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

Figure 5. Network analysis of einkorn grain subproteome response to N and S supply. (a) Co-abundance network for
proteins which differed in abundance between treatments and linkages with N and S treatments defined as central
attributes. (b) to (e) Co-abundance networks for proteins in cluster 1 (b), cluster 2 (c), cluster 3 (d) and cluster 4 (e) as
shown in Figure 4. Node shapes reflect attribute categories: squares, comparison between two treatments; circles, grain
storage proteins (GSPs); hexagons, albumin-globulin proteins; and diamonds, nuclear protein. Node colours indicate
which cluster the protein belongs to: purple, cluster 1; brown, cluster 2; red, cluster 3; green, cluster 4. In (b) to (e) the
intensity of the node colour reflects the number of edges (connectivity). Edge colour indicates the biological
significance between source and target: in (a) red when the protein was more abundant in the first treatment used in
the comparison and blue when the protein was less abundant in the first treatment; in (b, c, d, e), red shows the
abundance of proteins was high for the two nodes and blue when the abundance of proteins was low for the two nodes.
Nodes with thick borders were found both in (a) and a corresponding cluster network in (b), (c), (d), or (e). Albumin-
globulin and nuclear proteins are identified by their Uniprot accession reference (for details, see Table S6). Accessions
written in red correspond to proteins highlighted in Figure 6.
90
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

The comparison between N+S+ and N+S- was death protein 4. The network connected an importin
involved in 49 rules, highlighting the importance of subunit beta-1-like and a tyrosine-protein
N × S interactions on the grain proteome compared phosphatase with two edges (abundant in similar
to N alone. Only 27 proteins differed in abundance conditions and not abundant in other similar
between N+S+ and N-S-. This result demonstrates conditions), showing that these two proteins
the importance of the balance between N and S strongly co-accumulate. The co-accumulation
supply. N+S-/N-S-, N-S+/N-S- and N+S+/N-S+ network from cluster 3 contains 15 (28%) nodes,
pairwise comparisons linked 35, 12 and 22 proteins, including two vicilin-like antimicrobial peptides 2-
respectively. 2, a cationic peroxidase SPC4-like, an ankyrin
In this network, many groups of proteins were repeat domain-containing protein and a glutathione
formed. Grouped proteins had the same response to transferase, which were related to the accumulation
N and S supply and were mostly part of the same of other proteins (Figure 5d). The co-accumulation
cluster, an indicator of the robustness of the network with cluster 4 proteins is the largest,
network analysis. For example, the quantity of meaning that proteins that increased with N+S+ co-
LMW-GS was higher for N+S+ compared with N- accumulated in tight synchrony (Figure 5e). This
S- and four nuclear proteins and three albumin- network was composed of two modules. One
globulin proteins were also grouped with LMW- module contained 19 nodes (i.e. 29% of the proteins
GS: an RNA helicase 27, an alanine of cluster 4) and 70 edges, the other one contained
aminotransferase, a betaine dehydrogenase, a 24 nodes (37%) and 85 edges. Some proteins in the
guanine nucleotide-binding 3 homolog, a first module were highly connected to others,
monothiol glutaredoxin-S11, a putative NADP- including a RNA helicase 27, a L-ascorbate
dependent oxidoreductase P1 and a chitinase 1. peroxidase, a phosphoglycerate kinase, a RNA
Some proteins were linked because they had very polymerase and a nucleolar complex protein 3. The
different quantities in three or four treatment second module comprised LMW-GS, ω5- and α/β-
comparisons and were therefore strongly affected gliadin and several other co-accumulating proteins:
by N and S supply. For example, HMW-GS was a betaine aldehyde dehydrogenase, a probable aldo-
abundant with N+S- treatment compared with the keto reductase 1, a phosphorylase and a
three other treatments. Several albumin-globulin multiprotein-bridging factor 1a.
with similar responses to N and S supply were thus In these two network approaches several proteins
grouped with HMW-GS: a zeta-carotene were revealed as being strongly affected by N
desaturase, a cysteine synthase, a glutathionyl- and/or S supply and co-accumulating with other
hydroquinone reductase, an isoflavone reductase, a proteins (nodes with thick border line in Figure 5;
methylthioribose kinase and a beta-D-glucan Table S5). These proteins thus represent central
exohydrolase. Other proteins were highly actors of the proteome response to nutrition.
connected in this network, for example, γ- and
ω1,2-gliadin and several albumin-globulin proteins,
Focus on central actors of the einkorn grain
like an importin subunit alpha, which was
proteome response to N and S supply
connected with four edges. This importin subunit
was more abundant for N+S- and N+S+ than with Among proteins revealed in the network analysis,
N-S- and N-S+, showing that this protein is a DEAD-box ATP-dependent RNA helicase 27
increased by high N availability independently of S (M7ZPU9) was found in cluster 4 (Figure 6a). This
supply. protein is known to be involved in many processes
Co-accumulation of proteins was also including RNA transcription, pre-mRNA splicing,
investigated in other networks, built within cluster mRNA export, ribosome biogenesis and translation
(Figure 5b-e). In cluster 1 co-accumulation initiation (Linder, 2006). This RNA helicase
network, there were 20 (44%) proteins (nodes) responded like LMW-GS to N and S supply, and
including four involved in carbohydrate metabolic was connected in the co-accumulation network to
process, two acting in response to stress, two 11 other proteins, including a trihelix transcription
ribosomal proteins, a serpin-ZX and a purple acid factor (Figure 6b) and a peter-pan like protein
phosphatase (Figure 5b). In cluster 2 co- (Figure 6c). The trihelix transcription factor
accumulation network, there were 13 (28%) contains a SANT/Myb domain, like those found in
proteins including HMW-GS and ω1,2-gliadins chromatin-remodeling complexes (Boyer et al.,
(Figure 5c). According to the network, the 2004), and its quantity was linked in network with
abundance of these GSPs was connected to that of that of four proteins (Figure 5e).
two other storage proteins and a programmed cell

91
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

Figure 6. Effects of N and S supply on the quantity of selected nuclear and albumin-globulin proteins during einkorn
grain filling. (a) DEAD-box ATP-dependent RNA helicase 27, (b) trihelix transcription factor, (c) peter pan-like
protein, (d) purple acid phosphatase, (e) serpin-ZX, (f) glutathione transferase. Data are scaled and are means ± 1 s.e.
for n = 3 independent replicates.

The peter-pan like protein possesses a Brix domain, value. There is currently a concerted effort to
found in proteins involved in ribosomal RNA reduce the use of N-based fertilizers and soil S
processing (Weis et al., 2015) and was networked availability is decreasing in many wheat growing
with six proteins (Figure 5e). These three central regions. It is in this context that we describe grain
actors co-accumulated and all increased with N and GSP, nuclear and albumin-globulin subproteome
S supply. Two other central actors were part of responses to N and S supply in a diploid species
cluster 1 and thus decreased with N and S supply: a closely related to bread wheat and durum wheat (T.
purple acid phosphatase (Figure 6d) acting in turgidum L. ssp. durum (Desf.) Husn.). The 203
protein post-translational modification and a albumin-globulin and nuclear proteins whose
serpin-ZX (Figure 6e) that negatively regulates abundance was modified by N and/or S supply have
endopeptidase activity. The amounts of these two diverse functions. Clustering and network analyses
proteins were associated with those of seven and allowed us to group proteins that had similar
four other proteins, respectively, including a abundance profiles in response to N and/or S supply
probable sarcosine oxidase. Like γ-gliadin, the during grain filling. We found that S supply highly
abundance of two glutathione transferases impacted GSP accumulation while N and S supply
increased markedly with high S, one of which was had different effects on the grain proteome. Several
central in the network being connected to nine other proteins central to the proteome response and
proteins (Figures 5d and 6f). potentially involved in the regulation of GSP
synthesis in response to nutrition were highlighted.
This information is summarized in a scheme of
DISCUSSION grain metabolism illustrating which molecular
pathways are affected by N and S supply during
In wheat, N and S supply highly influences GSP grain filling (Figure 7).
accumulation, and is thus expected to impact the
viscoelastic properties of gluten and flour end-use

92
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

Figure 7. Schematic overview of einkorn grain subproteome responses to N and S supply. Proteins represented are
some of the 209 proteins showing significant nutrition effects during grain filling. Relative protein quantities are scaled
and means for n = 3 independent replicates are represented by a color scale on a 1 × 4 grid, each square representing
one treatment. Functional classification was based on gene ontology annotation and subcellular localization was
determined by prediction tools. Uniprot accessions corresponding to protein descriptions are given in parentheses.
Accessions in red correspond to proteins highlighted in Figure 6.

The strong influence of S on the quantity and is in accordance with previous reports that the
composition of grain storage protein and other highest amount of grain gliadin was obtained with
grain subproteomes moderate N fertilization and high S supply (Zörb et
al., 2010).
S deficiency in soil leads to decreases in the
In the present study, the abundance of three
amount of GSP in wheat grain (Shewry et al.,
glutathione transferases was higher in response to
2001). We obtained similar results with more total
N-S+ and N+S+ treatments than to N-S-. One of
GSP per grain at maturity in the two S-containing
them was also identified as a central actor and had
treatments than in the low S treatments. This was
similar nutrition response profiles to those of γ-
due to an early increase in the rate of accumulation
gliadin. This result suggests that glutathione may
and longer duration of accumulation of gliadin,
increase with S supply, as occurs in wheat (Dai et
especially the S-rich classes α/β- and γ-gliadin. This

93
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

al., 2015). Glutathione is one of the major pools of cysteine synthase gene transcripts were observed in
non-protein S in wheat grain (Rhazi et al., 2003; S starvation and non-limiting N conditions
Steinfurth et al., 2012). In S-deficient conditions, (Nakamura et al., 1999). The latter authors
lesser amounts of total cysteine and glutathione concluded that activation of this gene regulated the
were found in wheat flour, which affected GSP balance between S-containing amino acids
composition and hence technological flour (methionine and cysteine) and other amino acids.
properties (Reinbold et al., 2008). Glutathione is These results broadly indicate that amino acid
involved in the glutathione-ascorbate cycle metabolism is activated at the expense of sugar
controlling the concentration of active oxygen in metabolism in response to N nutrition.
the cell (Noctor and Foyer, 1998). In this Intracellular transport was likely influenced by
mechanism, ascorbate peroxidase is involved in the nutrition. Indeed, transport was one of the few
detoxification of H2O2 (Caverzan et al., 2012). One biological processes strongly enriched in protein
ascorbate peroxidase was also more abundant in cluster 2. Proteins participating in transport were
einkorn grain treated with S. It is therefore possible more abundant with high N supply and less
that glutathione metabolism increases when S is abundant with high S supply. Among transporter
freely available. proteins significantly affected by nutrition were
In network analysis, N+S- and N-S+ had the most three importins, two of which were present in both
contrasting effects on the grain proteome, since the nuclear and albumin-globulin protein extracts.
three GSPs (HMW-GS, ω1,2- and γ-gliadin) and Importins are involved in the transport of NLS-
many nuclear and albumin-globulin proteins containing proteins from the cytoplasm to the
differentially accumulated between these two nucleus through nuclear pore complexes localized
treatments. N+S+ and N+S- also had very different in the nuclear envelope (Bednenko et al., 2003). In
effects, again demonstrating the strong effect of S plants, communication between these two cell
on the proteome. Interestingly, even though many compartments is important for growth,
proteins were increased with N+S+, the network development and responses to environmental
analysis of the N+S+/N-S- comparison connected stimuli (Merkle, 2011; Tamura and Hara-
fewer proteins than the N+S-/N-S+ comparison (27 Nishimura, 2013). Nucleocytoplasmic transport
and 61 proteins, respectively). Taken together, may be a regulatory mechanism activated by grain
these results suggest that the grain proteome N/S status to modify transcription by activating
responded more to the N/S balance than to the import of transcriptional regulators.
availability of either N or S individually. These
results are in good agreement with previous studies
Nutrition impact on DNA binding proteins
showing the importance of N-to-S ratio in
suggests transcriptional reprogramming
determining the rheological properties of wheat
dough and GSP composition (Zhao et al., 1999; In wheat, GSP synthesis is regulated at the
Zörb et al., 2010; Dai et al., 2015). transcriptional level (Dai et al., 2015). This
regulation, first described in barley and conserved
in other cereals, is governed by a network involving
Grain metabolism was influenced by N and S
cis-regulatory elements located in promoters of
supply
GSP genes and their interacting transcription
Carbohydrate metabolism was the biological factors (Rubio-Somoza et al., 2006; Verdier and
process most represented in proteins with Thompson, 2008). The eight transcription factors
significant nutrition effects. Such proteins were known to act in this regulation were not quantified
found in all four of the clusters established, but a in our experiment, which didn’t allow us to say if
few of them were in cluster 2, that is they didn’t their abundance was disturbed by N and S supply.
increase with N only. Conversely, proteins involved Among the 48 nuclear proteins showing significant
in amino acid metabolism were highly abundant nutrition effects, 10 (21%) have a DNA binding
with high N supply. This is consistent with previous function. Proportionally then nutrition had a strong
results where an increased concentration of free effect on proteins with DNA binding domains as
amino acids in wheat flour was measured in high N- this molecular function represented only 10% of the
to-S ratio conditions (Granvogl et al., 2007). We quantified nuclear proteins. Among them, two
noted that the abundance of a cysteine synthase DNA-directed RNA polymerases, responsible for
increased considerably with high N supply, which RNA synthesis, were increased by N and S supply.
tallies with a previous observation in rice (Oryza A similar response was observed for a multiprotein-
sativa L.) shoots and roots where high levels of bridging factor 1a. In Arabidopsis thaliana, this

94
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

type of factor is known to be a transcriptional co- potential effects on chromatin compaction and
activator (Tsuda et al., 2004). These results suggest accessibility governed by epigenetic mechanisms.
that transcription is activated under conditions of In summary, major changes occur in the proteome
high N and S availabilities. As might be expected if of developing einkorn wheat grain when the
more transcripts are being produced, several nutritional status is modified. Post-flowering N and
proteins involved in mRNA splicing also S nutrition clearly influenced the N-to-S ratio in the
accumulated when N and/or S were provided. grain, leading to significant changes in the GSP
In the DNA-binding domain class were three composition in mature grain, by modification of the
histone H1 proteins. Histones are responsible for rate and duration of GSP accumulation, especially
DNA condensation, organization and regulation in of HMW-GS, α/β- and γ-gliadin. Changes in N-to-
the nucleus. In eukaryotes, assembly of two copies S ratios were also related to changes in the nuclear
of each of the core histone, H2A, H2B, H3 and H4, and albumin-globulin proteomes. The balance of
results in an octamer protein that binds to cell functions was probably disturbed by altering N
superhelical DNA to form the basic structure of and S supply, as albumin-globulin proteins
DNA compaction called the nucleosome. Histone involved in carbohydrate and amino acid
H1 molecules are linkers acting to maintain metabolism and transport were differentially
chromatin structure (Harshman et al., 2013). Three affected. Central to these major changes, several
variants of histone H1 were part of clusters 1 and 2, DNA-binding nuclear proteins could have roles in
i.e. they were less abundant when S was supplied. the early grain response. Several proteins
In A. thaliana, histone H1 mutations resulted in highlighted in the present study will be targeted to
modifications of DNA methylation patterns, with investigate how grain composition is controlled.
more methylation found in some promoter
sequences in endosperm (Wierzbicki and
Jerzmanowski, 2005; Rea et al., 2012). In plants as EXPERIMENTAL PROCEDURES
in animals, DNA methylation affects the binding of
specific proteins to DNA and therefore the Plant material and growth conditions
formation of the transcription machinery (Finnegan Triticum monococcum ssp. monococcum
et al., 1998; Vanyushin and Ashapkin, 2011). (accession ERGE 35821) seeds were germinated (5
Interestingly in our study, there was one protein February) at room temperature on filter paper
with a methyl-CpG DNA binding domain (MBD) moistened with demineralized water in Petri dishes.
in cluster 3, i.e. it increased with high S supply. When the radicles were 0.5 cm to 1 cm long,
MBD proteins potentially act in transcriptional seedlings were transplanted to 50 mL PVC columns
repression by recruiting chromatin remodeling (7.5 cm internal diameter (i.d.) × 50 cm deep; 2
factors, histone methyltransferases and histone seedlings per column) filled with a 2:1 (v/v)
deacetylases, leading to chromatin compaction mixture of washed perlite and river sand. The
(Grafi et al., 2007). These results suggest that an columns were arranged in a greenhouse to form a
epigenetic response to S supply occurs in einkorn homogenous stand with a population density of 512
grain. A trihelix transcription factor that increased plants m-2. The experimental design was a
in the high N and high S condition was predicted to randomized complete block design with four blocks
be a central actor in the grain proteome response. and four treatments. The high plant density
This uncharacterized protein contains a SANT/Myb inhibited the development of axillary tillers which
domain, characteristic of some nuclear receptor co- favored the synchronous development of the plants
repressors and in many chromatin-remodeling within and between containers. In the greenhouse,
complexes (Boyer et al., 2004). In barley, it has air temperature at the top of the plant stand was
been reported that methylation of the promoters of maintained at 20°C/15°C (16 h light/8 h dark) and
storage protein genes could repress storage protein air relative humidity at 55%/75%. During light
synthesis and control grain development (Sørensen periods plants received a mean total daily
et al., 1996; Radchuk et al., 2005). Thus the photosynthetic photon flux (PPF) of 152 and 161
methylation state of GSP promoters could be a µmol m-2 d-1 between transplantation and anthesis
regulatory mechanism, modifying the DNA- and between anthesis and grain ripeness maturity,
binding capacity of transcription factors, as in respectively. Air temperature and relative humidity
maize Opaque 2 transcription (Rossi et al., 1997; were measured in miniature forced-draft flues
Sturaro and Viotti, 2001; Locatelli et al., 2009). placed in each block in the center of the plant stands
Taken together, these different elements indicate at ear height. Air temperature, relative humidity,
that N and S alter transcription dynamics, with and PPF were measured every min and 15-min

95
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

averages were recorded with a data logger. Thermal Storage protein extraction and quantification
time was calculated as cumulative degree-days
Gliadin and glutenin proteins were sequentially
(0°C base temperature) using average daily air
extracted from 100 mg of wholemeal flour milled
temperature.
from each replicate of grain sampled between
Until anthesis, each PVC column received 167
300°Cd and 1000°Cd after anthesis, as described by
mL d-1 of a modified Hoagland’s nutrient solution
Plessis et al. (2013). Gliadin classes and glutenin
(Castle and Randall, 1987) containing 3 mM N and
subunits were separated and quantified by reverse
0.1 mM S prepared with demineralized water (1
phase high performance liquid chromatography
mM KH2PO4, 1 mM KNO3, 0.5 mM Ca(NO3)2, 0.5
(RP-HPLC) using an Agilent 1290 Infinity LC
mM NH4NO3, 0.1 mM MgSO4, 1.9 mM MgCl2, 3.5
system (Agilent Technologies,
mM CaCl2, 4 mM KCl, 10 µM H3BO3, 0.7 µM
[Link] as described by Dai et al.
ZnCl2, 0.4 µM CuCl2, 4.5 µM MnCl2, 0.22 µM
(2015). Briefly, gliadin and glutenin extracts were
MoO3 and 50 µM EDFS-Fe). At anthesis when N
filtered through regenerated cellulose syringe filters
and S demand is low, the nutrient solution was
(0.45 µm pore diameter, UptiDisc, Interchim,
replaced with demineralized water to avoid excess
[Link] then 4 µL (for
build-up of N and S compounds in plants or potting
gliadins) or 2 µL (for glutenins) of protein extract
substrate. Then from 200°Cd to 700°Cd after
were injected into a C8 reversed-phase ZORBAX
anthesis four combinations of N and S were
300StableBound column (2.1 × 100 mm, 3.5 µm,
supplied: N-S-, nutrient solution with no N or S;
300 Å; Agilent Technologies) maintained at 50°C.
N+S-, 6 mM N with no S; N-S+, low N (0.5 mM)
Proteins were separated at a flow rate of 1 mL min-
and high S (2 mM); or N+S+, high N (6 mM) and 1
by using linear solvent gradients from 24% to 50%
high S (2 mM). Nitrogen was not totally excluded
acetonitrile containing 0.1% (v/v) trifluoroacetic
from the N-S+ nutrient solution as previous
acid over 13 min for gliadins, and from 23% to 42%
experiments have shown that S uptake and/or
over 25 min for glutenins. Proteins were detected
metabolism are inhibited when N is absent from the
by UV absorbance at 214 nm. Chromatograms were
nutrient solution (P. Martre, unpublished data). The
processed with ChemStation 10.1 software (Agilent
nutrient solutions were modified as described in Dai
Technologies) and the HPLC peaks corresponding
et al. (2015).
to each of the four gliadin classes and the two
glutenin subunits were identified following the
Grain sampling and processing observations of Wieser et al. (1998; Figure S2).
Main-stem ears were tagged when the anthers of
the central florets appeared. Grains were collected Nuclear and albumin-globulin protein
from the central portion of ears every 100°Cd extraction, digestion and desalting
(approximately every 5 d) from 300°Cd after
Nuclei were extracted and purified from 2 g of
anthesis to ripeness. For each treatment and
whole grains ground in extraction buffer (20 mM
sampling date, four to ten main-stem ears were
Hepes-KOH pH 7, 5 mM MgCl2, 10 mM 2-
sampled per replicate, depending on the grain
mercaptoethanol, 0.5 mM PMSF, 0.1% (v/v)
developmental stage. Except for those used to
phosphatase inhibitor cocktail (Sigma-Aldrich))
determine grain dry mass, grains were frozen in
with a Polytron homogenizer (Kinematica
liquid N2 just after harvesting then stored at -80°C.
POLYTRON® PT 10) according to Bancel et al.
Four biological replicates of whole grain were
(2015). Nuclear proteins were then prepared using
analyzed unless indicated otherwise.
TRI Reagent® (Sigma-Aldrich) according to the
manufacturer’s instructions. Albumins and
Determination of total grain S and N globulins were extracted from whole grains
concentration according to Marion et al. (1994) with the following
modifications. Proteins were extracted for 2 h at
Aliquots of 5 mg of wholemeal flour were
4°C in 10 mM sodium phosphate, 10 mM NaCl, pH
weighed in tin capsules and the total N and S
7.8. After centrifugation at 8000 × g for 20 min at
concentrations were determined with the Dumas
4°C, proteins in the supernatant were precipitated
combustion method using a FlashEA 1112 NC
with ice-cold acetone for 2 h at -20°C. After
Analyzer (Thermo Electron) following the
centrifugation at 10 000 × g for 5 min at 4°C, the
manufacturer’s recommendations. Grain protein
resulting pellets were washed three times in ice-
content was calculated by multiplying grain N
cold acetone then dried at room temperature.
concentration by 5.62 (Mossé et al., 1985).

96
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

Nuclear and albumin-globulin protein pellets 17,500). The MS/MS step was reiterated for the
were solubilized in 30 µL of a buffer containing eight major ions detected during the full MS scan.
0.1% (v/v) ZALS I, 6 M urea, 2 M thiourea, 10 mM Dynamic exclusion was set to 40 s.
DTT, 30 mM Tris-HCl pH 8.8, and 5 mM
NH4HCO3. Protein concentration was measured
Peptide identification and quantification
using the 2-D Quant Kit (GE Healthcare) with
bovine serum albumin as standard. For each Nuclear and albumin-globulin proteins were
sample, 40 µg of protein was alkylated with 58 mM identified by matching peptides against the Uniprot
iodoacetamide for 50 min in the dark at room protein database version 2014_07 limited to the
temperature, then diluted 10-fold with 50 mM Triticum genus using X!Tandem version
NH4HCO3 and digested in solution overnight at Sledgehammer 2013.09.01.1
37°C with 800 ng of trypsin in 50 mM NH4HCO3. ([Link]/tandem). Enzymatic cleavage
Digestion was stopped by adding 10% TFA. parameters were set as trypsin digestion with one
The protein samples were desalted using a Strata- possible missed cleavage. Cys
XL SPE column (Phenomenex) regenerated with carboxyamidomethylation was set as static
500 µL of 100% CH3CN and equilibrated three modification, whereas Met oxidation, N-terminal
times with 500 µL of buffer A (3% CH3CN, 0.06% deamidation, and N-terminal acetylation were set as
acetic acid). Digested protein (15 µg) mixed with variable modifications. Precursor mass tolerance
buffer A was loaded onto the column. After three was 10 ppm and fragment mass tolerance was 0.02.
washes with 500 µL of buffer B (40% CH3CN, Identified proteins were filtered and grouped using
0.06% acetic acid), peptides were eluted twice with X!TandemPipeline version 3.3.4 (http://
300 µL of buffer B, then dried and stored at -20°C. [Link]/bioinfo/xtandempipeline/). Peptide
and protein e-value cut offs were set to 0.01 and
10−4, respectively, with at least two peptides per
LC-MS/MS analysis of nuclear and albumin-
protein.
globulin proteins
Relative quantification of all identified peptides
Peptide samples were solubilized in a buffer of was done using MassChroQ software version 2.1.3
3% CH3CN, 0.05% TFA and 0.05% formic acid. (Valot et al., 2011) and extracting ion
Liquid chromatography was performed on a chromatograms (XICs) for all identified peptides
NanoLC Ultra system (Eksigent). Samples (1 μg) and integrating the area of XIC peaks at the
were loaded at 7.5 μL min−1 on a C18 precolumn (5 corresponding retention time. Non repeated
μm, 100 μm i.d. × 2 cm length; NanoSeparations) peptides were then removed. For the albumin-
connected to a separating BIOSPHERE C18 globulin fraction, these corresponded to peptides
column (3 μm, 75 μm i.d. × 150 mm length or 300 present in less than 90% of the samples. For nuclear
mm length for nuclear and albumin-globulin proteins, this threshold was set at 30%, due to the
proteins, respectively; NanoSeparations). Solvent high effect of the developmental stage on this
A was 0.1% formic acid in water and solvent B was subproteome, as previously reported on wheat grain
0.1% formic acid in CH3CN. Peptide separation (Bonnot et al., 2015). To quantify proteins, only
was achieved using a linear gradient from 5% to peptides not shared by other proteins and correlated
35% solvent B for 28 min (for nuclear proteins) or peptides were used. The quantity of the
60 min (for albumin-globulin) at 300 nL min−1. corresponding protein was calculated by summing
Including the regeneration and the equilibration the quantities of specific peptides for albumin-
steps, a single run took 45 min for nuclear proteins globulin, and by averaging the quantities of specific
and 87 min for albumin-globulin proteins. Eluted peptides for nuclear proteins.
peptides were analyzed with a Q-Exactive mass
spectrometer (Thermo Electron) using a
Functional classification and subcellular
nanoelectrospray interface. Ionization was
localization
performed with a 1.3 kV spray voltage applied to an
uncoated capillary probe (10 μm i.d., New Identified nuclear and albumin-globulin proteins
Objective). The Xcalibur 2.3 SP1 interface was were classified according to gene ontology (GO;
used to monitor data-dependent acquisition of Ashburner et al., 2000). Proteins that differed in
peptide ions. This included a full MS scan covering quantity between treatments but did not have any
a mass-to-charge ratio (m/z) of 400 to 1,400 with a functional annotation or GO information were
resolution of 70,000 and an MS/MS step analyzed with Blast2GO tool version 3.2 (Conesa et
(normalized collision energy, 27%; resolution, al., 2005) in order to assign potential function.

97
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

Predicted subcellular localization using Proteomic network inference and analysis

MultiLoc2 (Blum et al., 2009), WoLF PSORT
Relationships among GSP, albumin-globulin and
(Horton et al., 2007) and LocTree3 (Goldberg et al.,
nuclear protein dynamics were analyzed using the
2014) programs were used to validate nuclear
RulNet platform for network inference
protein identification. Results were collected using
([Link] Vincent et al., 2015). RulNet
the integrative tool PSI (Liu et al., 2013) which
uses an algorithm describing rule semantics
combines prediction results of different programs.
between attributes (Agier et al., 2007) to generate
The top three hits were considered for WoLF
rules among data. Protein data were scaled and
PSORT and MultiLoc2. Nuclear proteins were used
semantics were written to find rules between
for statistical analysis if they were predicted to be
proteins whose accumulation was significantly
nuclear with at least one tool.
affected by nutrition and the four N/S treatments
according to the relative abundance of proteins in
Statistical analysis the different samples (semantic 1, Method S1). Two
quality measures were used to select the best rules
Data were analyzed using the statistical software
(confidence and lift). Rules with a confidence > 0.6
program R v3.2.2 (R Core Team, 2015).
and a lift > 1.5 were selected. A second semantic
Differences in grain dry mass per grain and per ear,
was used to find rules between co-accumulated
total quantity of N and S per grain, and the quantity
proteins (semantic 2, Method S1). For this second
of GSP per grain at maturity were analyzed using a
semantic, thresholds for support, confidence and lift
one-way ANOVA with the factor nutrition (four
were set at 0.3, 0.9 and 1.5, respectively. Validated
levels). For the quantity of GSP, albumin-globulin
rules were visualized using Cytoscape software
and nuclear proteins in immature grains, a two-way
version 3.3.0 (Smoot et al., 2011).
ANOVA was performed, using the factors of
nutrition and developmental stage (i.e. the thermal
time after anthesis). Significant differences (α =
ACKNOWLEDGEMENTS
0.05) were then analyzed by Tukey’s honestly
significantly different (HSD) post-hoc test. We thank Richard Blanc (UMR GDEC, INRA,
To determine the rate and duration of Blaise Pascal University) for help in the greenhouse
accumulation of grain dry mass, total N and S and and Mireille Dardevet, Annie Faye, Marielle
protein fractions, data were fitted with a 3- Merlino, Sibille Perrochon and Isabelle Nadaud for
parameter logistic function equation (Triboï et al., help collecting grains (UMR GDEC). This work
2003): was supported by a Ph.D. grant from the French
Ministry for Higher Education and Research to TB
Q max (1) and funding from the French Government managed
Q(t ) 
 4r (t  t95 )  by the National Research Agency (ANR) in the
1  0.05exp  
 Q max  framework of Investments for the Future (ANR-10-
where Q is the quantity of dry mass, N, S, or GSP BTBR-03), France AgriMer and the French Fund
per grain, t is the thermal time after anthesis in Supporting Plant Breeding (FSOV).
degree-days, Q max is the final value of Q
approached as t → ∞, r is the maximum rate of SUPPORTING INFORMATION
accumulation defined as the derivative of the point
of inflexion, and t95 is the duration of accumulation Figure S1. Effects of N and S supply on grain
defined as the period from anthesis onwards during storage protein composition at maturity.
Figure S2. HPLC chromatograms of gliadin (a) and
which 95% of Q max is accumulated. glutenin (b) in mature einkorn grain.
Hierarchical clustering on principal components Table S1. Quantity of dry mass, N, S and GSP per
was computed on proteins with significant nutrition grain at maturity.
effects using the FactomineR package for R Table S2. Estimated maximum rate and duration of
(Husson et al., 2015), with the hcpc() function and accumulation of grain dry mass, total N, total S and
the ‘ward’ method. The number of clusters was GSP.
determined with the suggested partition, i.e. the one Table S3. Dry mass, N mass and concentration, S
with the higher relative loss of inertia. mass and concentration, N-to-S ratio, and protein
composition of developing einkorn grain.

98
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

Table S4. Quantification of nuclear proteins and Bioinformatics, 21, 3674–3676.

albumin-globulin in einkorn wheat grain. Corbellini, M., Empilli, S., Vaccino, P., Brandolini,
Table S5. Summary of the 209 proteins A., Borghi, B., Heun, M. and Salamini, F. (1999)
significantly impacted by N and S supply. Einkorn characterization for bread and cookie
production in relation to protein subunit
Method S1. Semantics used to infer regulatory
composition. Cereal Chem., 76, 727–733.
network from proteomics data. Dai, Z., Plessis, A., Vincent, J., et al. (2015)
Transcriptional and metabolic alternations
rebalance wheat grain storage protein
REFERENCES accumulation under variable nitrogen and sulfur
supply. Plant J., 83, 326–343.
Agier, M., Petit, J.M. and Suzuki, E. (2007) Unifying
Das, P., Nutan, K.K., Singla-Pareek, S.L. and Pareek,
framework for rule semantics: Application to gene
A. (2015) Understanding salinity responses and
expression data. Fundam. Informaticae, 543–559.
adopting “omics-based” approaches to generate
Ashburner, M., Ball, C.A., Blake, J.A., et al. (2000)
salinity tolerant cultivars of rice. Front. Plant Sci.,
Gene ontology: tool for the unification of biology.
6, 712.
The Gene Ontology Consortium. Nat. Genet., 25,
Eriksen, J. (2009) Chapter 2 soil sulfur cycling in
25–29.
temperate agricultural systems. In Advances in
Bancel, E., Bonnot, T., Davanture, M., Branlard, G.,
Agronomy. pp. 55–89.
Zivy, M. and Martre, P. (2015) Proteomic
Finnegan, E.J., Genger, R.K., Peacock, W.J. and
approach to identify nuclear proteins in wheat
Dennis, E.S. (1998) Dna methylation in plants.
grain. J. Proteome Res., 14, 4432–4439.
Annu. Rev. Plant Physiol. Plant Mol. Biol., 49,
Bednenko, J., Cingolani, G. and Gerace, L. (2003)
223–247.
Nucleocytoplasmic transport: navigating the
Flæte, N.E.S., Hollung, K., Ruud, L., Sogn, T.,
channel. Traffic, 4, 127–135.
Færgestad, E.M., Skarpeid, H.J., Magnus, E.M.
Blum, T., Briesemeister, S. and Kohlbacher, O.
and Uhlen, A.K. (2005) Combined nitrogen and
(2009) MultiLoc2: integrating phylogeny and
sulphur fertilisation and its effect on wheat quality
Gene Ontology terms improves subcellular protein
and protein composition measured by SE-FPLC
localization prediction. BMC Bioinformatics, 10,
and proteomics. J. Cereal Sci., 41, 357–369.
274.
Fox, S.E., Geniza, M., Hanumappa, M., et al. (2014)
Bonnot, T., Bancel, E., Chambon, C., Boudet, J.,
De novo transcriptome assembly and analyses of
Branlard, G. and Martre, P. (2015) Changes in
gene expression during photomorphogenesis in
the nuclear proteome of developing wheat
diploid wheat Triticum monococcum G. K.
(Triticum aestivum L.) grain. Front. Plant Sci., 6.
Pandey, ed. PLoS One, 9, e96855.
Boyer, L.A., Latek, R.R. and Peterson, C.L. (2004)
Goldberg, T., Hecht, M., Hamp, T., et al. (2014)
Opinion: The SANT domain: a unique histone-tail-
LocTree3 prediction of localization. Nucleic Acids
binding module? Nat. Rev. Mol. Cell Biol., 5, 158–
Res., 42, W350–W355.
163.
Grafi, G., Zemach, A. and Pitto, L. (2007) Methyl-
Branlard, G., Dardevet, M., Saccomano, R., CpG-binding domain (MBD) proteins in plants.
Lagoutte, F. and Gourdon, J. (2001) Genetic
Biochim. Biophys. Acta - Gene Struct. Expr., 1769,
diversity of wheat storage proteins and bread
287–294.
wheat quality. Euphytica, 119, 59–67.
Granvogl, M., Wieser, H., Koehler, P., Tucher, S. Von
Castle, S. and Randall, P. (1987) Effects of sulfur
and Schieberle, P. (2007) Influence of sulfur
deficiency on the synthesis and accumulation of
fertilization on the amounts of free amino acids in
proteins in the developing wheat seed. Aust. J.
wheat. Correlation with baking properties as well
Plant Physiol., 14, 503–516.
as with 3-aminopropionamide and acrylamide
Caverzan, A., Passaia, G., Rosa, S.B., Ribeiro, C.W., generation during baking. J. Agric. Food Chem.,
Lazzarotto, F. and Margis-Pinheiro, M. (2012)
55, 4271–4277.
Plant responses to stresses: role of ascorbate
Harshman, S.W., Young, N.L., Parthun, M.R. and
peroxidase in the antioxidant protection. Genet.
Freitas, M.A. (2013) H1 histones: current
Mol. Biol., 35, 1011–1019.
perspectives and challenges. Nucleic Acids Res.,
Chope, G.A., Wan, Y., Penson, S.P., Bhandari, D.G., 41, 9593–9609.
Powers, S.J., Shewry, P.R. and Hawkesford, Hidalgo, A. and Brandolini, A. (2014) Nutritional
M.J. (2014) Effects of genotype, season, and
properties of einkorn wheat (Triticum
nitrogen nutrition on gene expression and protein
monococcum L.). J. Sci. Food Agric., 94, 601–612.
accumulation in wheat grain. J. Agric. Food
Hidalgo, A., Brandolini, A., Pompei, C. and Piscozzi,
Chem., 62, 4399–4407.
R. (2006) Carotenoids and tocols of einkorn wheat
Conesa, A., Gotz, S., Garcia-Gomez, J.M., Terol, J., (Triticum monococcum ssp. monococcum L.). J.
Talon, M. and Robles, M. (2005) Blast2GO: a
Cereal Sci., 44, 182–193.
universal tool for annotation, visualization and
Hill, K., Horváth-Szanics, E., Hajós, G. and Kiss, É.
analysis in functional genomics research.

99
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

(2008) Surface and interfacial properties of water- grain protein composition reveals that gliadin and
soluble wheat proteins. Colloids Surfaces A glutenin composition are trans-regulated by
Physicochem. Eng. Asp., 319, 180–187. different chromosome regions. J. Exp. Bot., 64,
Horton, P., Park, K.-J., Obayashi, T., Fujita, N., 3627–3644.
Harada, H., Adams-Collier, C.J. and Nakai, K. R Core Team (2015) R: A language and environment
(2007) WoLF PSORT: protein localization for statistical computing. R foundation for
predictor. Nucleic Acids Res., 35, W585–W587. statistical computing, Vienna, Austria.
Husson, F., Josse, J., Le, S., Mazet, J. and & Husson, Radchuk, V. V., Sreenivasulu, N., Radchuk, R.I.,
M.F. (2015) Package “FactoMineR”. Wobus, U. and Weschke, W. (2005) The
Linder, P. (2006) Dead-box proteins: a family affair- methylation cycle and its possible functions in
active and passive players in RNP-remodeling. barley endosperm development. Plant Mol. Biol.,
Nucleic Acids Res., 34, 4168–4180. 59, 289–307.
Liu, L., Zhang, Z., Mei, Q. and Chen, M. (2013) PSI: Rea, M., Zheng, W., Chen, M., Braud, C., Bhangu, D.,
A comprehensive and integrative approach for Rognan, T.N. and Xiao, W. (2012) Histone H1
accurate plant subcellular localization prediction J. affects gene imprinting and DNA methylation in
L. Heazlewood, ed. PLoS One, 8, e75826. Arabidopsis. Plant J., 71, 776–786.
Locatelli, S., Piatti, P., Motto, M. and Rossi, V. (2009) Reinbold, J., Rychlik, M., Asam, S., Wieser, H. and
Chromatin and DNA modifications in the Koehler, P. (2008) Concentrations of total
Opaque2-mediated regulation of gene glutathione and cysteine in wheat flour as affected
transcription during maize endosperm by sulfur deficiency and correlation to quality
development. Plant Cell, 21, 1410–1427. parameters. J. Agric. Food Chem., 56, 6844–6850.
Marcussen, T., Sandve, S.R., Heier, L., et al. (2014) Rhazi, L., Cazalis, R., Lemelin, E. and Aussenac, T.
Ancient hybridizations among the ancestral (2003) Changes in the glutathione thiol–disulfide
genomes of bread wheat. Science, 345, 1250092– status during wheat grain development. Plant
1250092. Physiol. Biochem., 41, 895–902.
Marion, D., Nicolas, Y., Popineau, Y., Branlard, G. Rossi, V., Motto, M. and Pellegrini, L. (1997) Analysis
and Landry, J. (1994) A new and improved of the methylation pattern of the maize Opaque-2
sequential extraction procedure of wheat proteins. (O2) promoter and in vitro binding studies indicate
Molecular and functional aspects (p.47), ed., that the O2 B-Zip protein and other endosperm
Presented at Wheat kernel proteins, Viterbo, ITA factors can bind to methylated target sequences. J.
(1994-09-28 - 1994-09-30). Biol. Chem., 272, 13758–13765.
Martre, P., Porter, J.R., Jamieson, P.D. and Triboï Rubio-Somoza, I., Martinez, M., Abraham, Z., Diaz,
(2003) Modeling grain nitrogen accumulation and I. and Carbonero, P. (2006) Ternary complex
protein composition to understand the sink/source formation between HvMYBS3 and other factors
regulations of nitrogen remobilization for wheat. involved in transcriptional control in barley seeds.
Plant Physiol., 133, 1959–1967. Plant J., 47, 269–281.
Merkle, T. (2011) Nucleo-cytoplasmic transport of Shewry, P.R. and Halford, N.G. (2002) Cereal seed
proteins and RNA in plants. Plant Cell Rep., 30, storage proteins: structures, properties and role in
153–176. grain utilization. J. Exp. Bot., 53, 947–958.
Mossé, J., Huet, J.C. and Baudet, J. (1985) The amino Shewry, P.R., Mitchell, R. a. C., Tosi, P., et al. (2012)
acid composition of wheat grain as a function of An integrated study of grain development of wheat
nitrogen content. J. Cereal Sci., 3, 115–130. (cv. Hereward). J. Cereal Sci., 56, 21–30.
Nakamura, T., Yamaguchi, Y. and Sano, H. (1999) Shewry, P.R., Tatham, A.S. and Halford, N.G. (2001)
Four rice genes encoding cysteine synthase: Nutritional control of storage protein synthesis in
isolation and differential responses to sulfur, developing grain of wheat and barley. Plant
nitrogen and light. Gene, 229, 155–161. Growth Regul., 34, 105–111.
Noctor, G. and Foyer, C.H. (1998) Ascorbate and Shewry, P.R., Tatham, A.S. and Lazzeri, P. (1997)
Glutathione: keeping active oxygen under control. Biotechnology of wheat quality. J. Sci. Food
Annu. Rev. Plant Physiol. Plant Mol. Biol., 49, Agric., 73, 397–406.
249–279. Simmonds, N.W. (1995) The relation between yield and
Oury, F.-X. and Godin, C. (2007) Yield and grain protein in cereal grain. J. Sci. Food Agric., 67,
protein concentration in bread wheat: how to use 309–315.
the negative relationship between the two Smoot, M.E., Ono, K., Ruscheinski, J., Wang, P.-L.
characters to identify favourable genotypes? and Ideker, T. (2011) Cytoscape 2.8: new features
Euphytica, 157, 45–57. for data integration and network visualization.
Petrovská, B., Šebela, M. and Doležel, J. (2015) Inside Bioinformatics, 27, 431–432.
a plant nucleus: discovering the proteins. J. Exp. Sørensen, M.B., Müller, M., Skerritt, J. and Simpson,
Bot., 66, 1627–1640. D. (1996) Hordein promoter methylation and
Plessis, A., Ravel, C., Bordes, J., Balfourier, F. and transcriptional activity in wild-type and mutant
Martre, P. (2013) Association study of wheat barley endosperm. Mol. Gen. Genet., 250, 750–

100
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

760. Weis, B.L., Palm, D., Missbach, S., Bohnsack, M.T.

Steinfurth, D., Zörb, C., Braukmann, F. and and Schleiff, E. (2015) atBRX1-1 and atBRX1-2
Mühling, K.H. (2012) Time-dependent are involved in an alternative rRNA processing
distribution of sulphur, sulphate and glutathione in pathway in Arabidopsis thaliana. RNA, 21, 415–
wheat tissues and grain as affected by three sulphur 425.
fertilization levels and late S fertilization. J. Plant Wierzbicki, A.T. and Jerzmanowski, A. (2005)
Physiol., 169, 72–77. Suppression of histone H1 genes in Arabidopsis
Sturaro, M. and Viotti, A. (2001) Methylation of the results in heritable developmental defects and
Opaque2 box in zein genes is parent-dependent stochastic changes in DNA methylation. Genetics,
and affects O2 DNA binding activity in vitro. Plant 169, 997–1008.
Mol. Biol., 46, 549–560. Wieser, H. (2007) Chemistry of gluten proteins. Food
Tamura, K. and Hara-Nishimura, I. (2013) The Microbiol., 24, 115–119.
molecular architecture of the plant nuclear pore Wieser, H., Antes, S. and Seilmeier, W. (1998)
complex. J. Exp. Bot., 64, 823–832. Quantitative determination of gluten protein types
Triboï, E., Martre, P. and Triboï-Blondel, A.-M. in wheat flour by reversed-phase high-
(2003) Environmentally-induced changes in performance liquid chromatography. Cereal
protein composition in developing grains of wheat Chem., 75, 644–650.
are related to changes in total protein content. J. Wieser, H., Gutser, R. and Tucher, S. von (2004)
Exp. Bot., 54, 1731–1742. Influence of sulphur fertilisation on quantities and
Tsuda, K., Tsuji, T., Hirose, S. and Yamazaki, K. proportions of gluten protein types in wheat flour.
(2004) Three Arabidopsis MBF1 homologs with J. Cereal Sci., 40, 239–244.
distinct expression profiles play roles as Wieser, H. and Seilmeier, W. (1998) The influence of
transcriptional co-activators. Plant Cell Physiol., nitrogen fertilisation on quantities and proportions
45, 225–231. of different protein types in wheat flour. J. Sci.
Valot, B., Langella, O., Nano, E. and Zivy, M. (2011) Food Agric., 76, 49–55.
MassChroQ: A versatile tool for mass Wrigley, C., Cros, D., Archer, M., Downie, P. and
spectrometry quantification. Proteomics, 11, Roxburgh, C. (1980) The sulfur content of wheat
3572–3577. endosperm proteins and its relevance to grain
Vanyushin, B.F. and Ashapkin, V. V (2011) DNA quality. Aust. J. Plant Physiol., 7, 755–766.
methylation in higher plants: Past, present and Yin, X. and Komatsu, S. (2015) Quantitative
future. Biochim. Biophys. Acta, 1809, 360–368. proteomics of nuclear phosphoproteins in the root
Verdier, J. and Thompson, R.D. (2008) tip of soybean during the initial stages of flooding
Transcriptional regulation of storage protein stress. J. Proteomics, 119, 183–195.
synthesis during dicotyledon seed filling. Plant Zhao, F., Hawkesford, M. and McGrath, S. (1999)
Cell Physiol., 49, 1263–1271. Sulphur assimilation and effects on yield and
Vincent, J., Martre, P., Gouriou, B., Ravel, C., Dai, quality of wheat. J. Cereal Sci., 30, 1–17.
Z., Petit, J.-M. and Pailloux, M. (2015) RulNet: Zörb, C., Grover, C., Steinfurth, D. and Hermann
A web-oriented platform for regulatory network Mühling, K. (2010) Quantitative proteome
inference, application to wheat –omics data. PLoS analysis of wheat gluten as influenced by N and S
One, 10, e0127127. nutrition. Plant Soil, 327, 225–234.

101
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

Supplementary information

Method S1. Semantics used to infer regulatory network from proteomics data.

Query 1: Discover rules between nutrition comparisons and relative abundance of proteins.

FINDRULES
SCOPE t1 IN Data_proteomique_RulNet_scaled
HAVING high: [Link] > 0.5 OVER ALL MINUS Stage, code_treatment, replicate, treatment,
ST_treat
AND low: [Link] < -0.5 OVER ALL MINUS Stage, code_treatment, replicate, treatment,
ST_treat
AND Nitrogen: ([Link] = 'T1' OR [Link] = 'T3') OVER code_treatment
AND NoNitrogen: ([Link] = 'C' OR [Link] = 'T2') OVER code_treatment
AND NoSulfur: ([Link] = 'C' OR [Link] = 'T1') OVER code_treatment
AND Sulfur: ([Link] = 'T2' OR [Link] = 'T3') OVER code_treatment
AND 300: [Link] = 300 OVER Stage
AND 400: [Link] = 400 OVER Stage
AND 500: [Link] = 500 OVER Stage

Query 2: Discover rules between co-accumulated proteins

FINDRULES
SCOPE t1 IN Data_proteomique_RulNet_scaled, t2 IN Data_proteomique_RulNet_scaled
WHERE [Link] = [Link] AND t1.code_treatment < t2.code_treatment
HAVING higher: [Link] > 0.5 OVER ALL MINUS Stage, code_treatment, replicate,
treatment, ST_treat
AND lower: [Link] > 0.5 OVER ALL MINUS Stage, code_treatment, replicate,
treatment, ST_treat
AND C_vs_T1: [Link]='C' AND [Link] = 'T1' OVER code_treatment
AND C_vs_T2: [Link]='C' AND [Link] = 'T2' OVER code_treatment
AND C_vs_T3: [Link]='C' AND [Link] = 'T3' OVER code_treatment
AND T1_vs_T2: [Link]='T1' AND [Link] = 'T2' OVER code_treatment
AND T1_vs_T3: [Link]='T1' AND [Link] = 'T3' OVER code_treatment
AND T2_vs_T3: [Link]='T2' AND [Link] = 'T3' OVER code_treatment

102
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

Figure S1. Effects of N and S supply on einkorn grain storage protein (GSP) composition at maturity.
The four treatments were N-S- (0 mM N and 0 mM S), N+S- (6 mM N and 0 mM S), N-S+ (0.5 mM N
and 2 mM S) and N+S+ (6 mM N and 2 mM S). (a) Quantity per grain of total gliadins, α/β-, γ-, ω1,2-,
and ω5-gliadin classes. (b) Quantity per grain of total glutenins, LMW-GS, and HMW-GS. (c) Gliadin-
to-glutenin ratio. (d) Grain storage protein composition expressed as percentage of grain dry mass. In
(a), (b) and (c) data are mean ± 1 s.d. for n = 4 independent replicates and different letters above the
error bars indicate statistical differences at P < 0.05.

103
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

Figure S2. HPLC chromatograms of gliadin (a) and glutenin (b) in mature einkorn grain. Red dotted
lines indicate separations between ω5-, ω1,2-, α/β- and γ-gliadin classes (a) and between HMW-GS and
LMW-GS subunits (b).

Table S1. Quantity of dry mass, N, S and GSP per grain at maturity. Different letters within a row
indicate significant differences (P < 0.05) between treatments. Data are mean ± 1 s.e. for n = 4
independent replicates.
Treatments P-value
Variables N-S- N+S- N-S+ N+S+
Grain dry mass (mg DM grain-1) 41.8 ± 1.17 a 38.3 ± 5.07a 43.9 ± 1.15a 42.2 ± 0.72a 0.523
Grain protein concentration (%) 19.3 ± 1.22b 27.0 ± 1.96a 23.6 ± 1.28ab 27.4 ± 0.45a 0.004
Grain S concentration (%) 0.19 ± 0.02b 0.19 ± 0.02b 0.29 ± 0.02a 0.29 ± 0.01a < 0.001
N- to to-S ratio (-) 18.1 ± 0.78b 24.8 ± 1.07a 14.1 ± 0.23c 16.5 ± 0.29bc < 0.001
Grain N (mg N grain-1) 1.42 ± 0.08b 1.76 ± 0.15ab 1.82 ± 0.10a 2.03 ± 0.02a 0.006
Grain S (mg S grain -1) 0.08 ± 0.01b 0.07 ± 0.01b 0.13 ± 0.01a 0.12 ± 0.01a < 0.001

104
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

Table S2. Estimated maximum rate and duration of accumulation of grain dry mass, total N, total S and
GSPs. Data shown in Figures 1 and 2 were fitted against thermal time using eqn. (1). Data are mean ±
1 s.e. P-values of the estimated maximum rate and duration are given in square brackets.
Duration Maximum rate
Variables Treatments (°Cd) (mg grain-1 [°Cd]-1)
-11
Dry mass N-S- 590 ± 49 [7 × 10 ] (1.37 ± 0.25) × 10-1 [2 × 10-6]
N+S- 600 ± 63 [1 × 10-9] (1.12 ± 0.24) × 10-1 [7 × 10-6]
N- S+ 633 ± 58 [1 × 10-10] (1.2 ± 0.21) × 10-1 [1 × 10-6]
N+S+ 710 ± 77 [8 × 10-10] (0.92 ± 0.17) × 10-1 [1 × 10-6]
N N-S- 550 ± 46 [2 × 10-10] (5.87 ± 1.19) × 10-3 [1 × 10-5]
N+S- 702 ± 67 [3 × 10-10] (4.13 ± 0.68) × 10-3 [7 × 10-7]
N- S+ 688 ± 63 [2 × 10-10] (4.42 ± 0.71) × 10-3 [4 × 10-7]
N+S+ 691 ± 53 [2 × 10-11] (5.13 ± 0.71) × 10-3 [8 × 10-8]
S N-S- 502 ± 79 [1 × 10-6] (3.65 ± 1.5) × 10-4 [6 × 10-3]
N+S- 2015 ± 3674 [8 × 10-4] (0.55 ± 0.21) × 10-4 [3 × 10-2]
N- S+ 798 ± 128 [2 × 10-7] (2.4 ± 0.56) × 10-4 [4 × 10-5]
N+S+ 753 ± 119 [2 × 10-7] (2.5 ± 0.61) × 10-4 [8 × 10-5]
LMW-GS N-S- 595 ± 58 [3 × 10-7] (2.15 ± 0.54) × 10-3 [2 × 10-3]
N+S- 725 ± 118 [5 × 10-5] (1.11 ± 0.31) × 10-3 [3 × 10-3]
N- S+ 705 ± 76 [8 × 10-7] (1.63 ± 0.33) × 10-3 [4 × 10-4]
N+S+ 717 ± 74 [5 × 10-7] (1.68 ± 0.32) × 10-3 [2 × 10-4]
HMW-GS N-S- 615 ± 44 [8 × 10-9] (6.02 ± 1.36) × 10-4 [8 × 10-4]
N+S- 624 ± 31 [1 × 10-10] (8.39 ± 1.2) × 10-4 [1 × 10-5]
N- S+ 946 ± 132 [1 × 10-5] (2.32 ± 0.47) × 10-4 [4 × 10-4]
N+S+ 818 ± 67 [4 × 10-8] (3.91 ± 0.6) × 10-4 [3 × 10-5]
α/β-gliadin N-S- 730 ± 150 [4 × 10-4] (5.5 ± 1.87) × 10-4 [1 × 10-2]
N+S- 772 ± 266 [1 × 10-2] (3.42 ± 1.74) × 10-4 [7 × 10-2]
N- S+ 1049 ± 237 [8 × 10-4] (4.91 ± 1.12) × 10-4 [9 × 10-4]
N+S+ 898 ± 136 [3 × 10-5] (6.55 ± 1.34) × 10-4 [4 × 10-4]
γ-gliadin N-S- 928 ± 421 [5 × 10-2] (1.33 ± 0.69) × 10-4 [8 × 10-2]
N+S- 1666 ± 14680 [9 × 10-1] (0.21 ± 1.13) × 10-4 [9 × 10-1]
N- S+ 1101 ± 250 [9 × 10-4] (2.49 ± 0.52) × 10-4 [5 × 10-4]
N+S+ 1000 ± 188 [2 × 10-4] (2.73 ± 0.58) × 10-4 [5 × 10-4]
ω1,2-gliadin N-S- 707 ± 109 [3 × 10-5] (4.41 ± 1.24) × 10-5 [4 × 10-3]
N+S- 684 ± 74 [8 × 10-7] (6.31 ± 1.4) × 10-5 [7 × 10-4]
N- S+ 1292 ± 383 [6 × 10-3] (2.89 ± 0.52) × 10-5 [1 × 10-4]
N+S+ 926 ± 98 [6 × 10-7] (5.56 ± 0.81) × 10-5 [2 × 10-5]
ω5-gliadin N-S- 736 ± 107 [2 × 10-5] (2.25 ± 0.58) × 10-4 [2 × 10-3]
N+S- 698 ± 78 [1 × 10-6] (2.97 ± 0.69) × 10-4 [1 × 10-3]
N- S+ 1033 ± 175 [7 × 10-5] (1.9 ± 0.35) × 10-4 [2 × 10-4]
N+S+ 808 ± 79 [3 × 10-7] (3.06 ± 0.5) × 10-4 [5 × 10-5]

105
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

Tables S3 and S4 are not necessary to understand the PhD manuscript since they correspond to quantitative
data for GSP, albumin-globulin and nuclear proteins. Thus they have not been included in this manuscript.

Table S5. Summary of the 209 proteins significantly impacted by N and S supply. Column ‘Type’ mentions
if protein is part of the Grain Storage Protein (GSP), nuclear (nucl) or albumin-globulin (alg) fraction.
Column ‘Code’ provides codes used in figures to qualify proteins, for nuclear and albumin-globulin proteins
it corresponds to their Uniprot accession number. Then cluster number and connectivity of these proteins in
network analyses are provided. In columns related to the nutrition network, sign ‘+’ or ‘-‘ indicate that the
protein was present in network, a ‘+’ indicating a higher amount in the first treatment and a ‘-‘ indicating a
lower amount in the first treatment. The column named ‘connectivity’ indicates the number of edges relied
to the protein, and thus reflects association number.
Network co-
Nutrition network
accumulation
Type Code Description Cluster
N+S- N-S+ N+S+ N-S+ N+S+ N+S+
Connectivity
N-S- N-S- N-S- N+S- N+S- N-S+
GSP HMW HMW-GS 2 + - - 5
GSP LMW LMW-GS 4 + 5
GSP α/β α/β-gliadin 4 + +
GSP γ γ-gliadin 3 + + +
GSP ω1.2 ω1.2 2 + - + 4
GSP ω5 ω5-gliadin 4 + 5
Nucl B6RRA6 Puroindoline A 3 - + +
Nucl D8L9M8 RNA recognition domain containing protein,expressed 4 + +
Nucl E2F3W4 40S Ribosomal protein S15a 4 + 2
Nucl F4Y595 Heat shock protein 9 4
Nucl M7YCG2 Importin subunit alpha 4 +
Nucl M7YGL9 Alanine aminotransferase 2 4 + 4
Nucl M7YRR2 Actin-depolymerizing factor 4 4 4
Nucl M7YRS8 Histone H1 1 -
Nucl M7YRT1 Crooked neck-like protein 1 4 +
Nucl M7YWU3 Heat shock cognate 7 kDa protein 1 4 +
Nucl M7YZD9 60S Ribosomal protein L18 1 - 4
DNA-directed RNA polymerases I, II, and III subunit
Nucl M7ZFZ3 RPABC1 4 15
Nucl M7ZGF0 dna-binding protein 2 -
Nucl M7ZJH4 14-3-3-like protein B 4 3
Nucl M7ZPU0 peter pan-like protein 4 + + 6
Nucl M7ZPU9 DEAD-box ATP-dependent RNA helicase 27 4 + 11
Nucl M7ZXC1 phosphorylase 4 18
Nucl M8A0B6 Histone H1 2 -
Nucl M8A0L6 coiled-coil domain-containing protein 12 4 2
Nucl M8A407 60S acidic Ribosomal protein P2B 3 + +
Nucl M8A553 60S Ribosomal protein L23a 4 + +
Nucl M8A571 Elongation factor 2 4 +
Nucl M8AAX5 Multiprotein-bridging factor 1a 4 12
Nucl M8AL83 40S Ribosomal protein S11 4 9
Nucl M8AVG0 Serine/arginine-rich splicing factor 4 4
Nucl P12783 phosphoglycerate kinase, cytolic 4 5
Nucl P93616 Poly(A)-binding protein 4
Nucl Q03387 Eukaryotic translation initiation factor isoform 4G-1 4 + 6
Nucl Q3S4I0 Eukaryotic translation initiation factor 5A2 4
Nucl Q9XHL9 Histone H1 WH1B.1 2
Nucl T1LTD9 tyrosine-protein phosphatase 2 3
Nucl T1MN05 fam10 family protein at4g22670-like 4
Nucl T1N8M0 probable aldo-keto reductase 1 4 14
Nucl W5AA91 Importin subunit alpha 4
Nucl W5AGD5 trihelix transcription factor 4 + 3
Nucl W5ANL4 protein rrp5 homolog 4 10
Nucl W5AP12 DNA-directed RNA polymerase 4
Nucl W5B117 guanine nucleotide-binding 3 homolog 4 +
Nucl W5BW98 protein rcc2 4 +
Nucl W5CME5 60s ribosomal protein l30 1 - 1
Nucl W5DPT1 Spastin 2 1
Nucl W5EU71 formate-tetrahydrofolate ligase 4 8

106
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

Nucl W5F910 dead-box atp-dependent rna helicase 28 4 + + 2

Nucl W5FES4 brain acid soluble protein 1 3 +
Nucl W5G441 nucleolar complex protein 3 homolog 4 15
Nucl W5G6L7 betaine aldehyde dehydrogenase 4 + 14
Nucl W5HA67 hypothetical protein F775_28656 3 1
Nucl W5HDD8 sucrose synthase 3 +
Alg A1YUM8 Limit dextrinase type starch debranching enzyme 3 - + +
Alg A9U8G4 Alcohol dehydrogenase ADH1A 3 - + + 8
Alg A9YPI3 Puroindoline A 3 - - +
Alg B2CGM5 Triticin 3 +
Alg B2ZGG4 GTPase SAR1 1
Alg B3TZE7 Subtilisin protease (Fragment) 2 +
Alg B6DU63 Zeta-carotene desaturase 2 + - -
Alg B6UKS0 Gamma-gliadin 3 +
Alg B8XU49 Gamma gliadin (Fragment) 1
Alg C0KTA6 Fructose-bisphosphate aldolase 4 3
Alg C5J3R4 Trypsin inhibitor 3 - + + 1
Alg D2KZ08 Aminotransferase 2 - 1
Alg D3KVP3 Purple acid phosphatase 1 - 7
Alg D6QY49 Omega-gliadin (Fragment) 2 - 4
Alg D8L9B3 Protein disulfide isomerase family protein 4-1 1 4
Alg D8L9M8 RNA recognition domain containing protein,expressed 2 1
Alg F6LAX4 Protein phosphatase 2A structural subunit 2
Alg G0YVZ4 High molecular weight glutenin 1Ax2.1 2 4
Alg G3CCE7 Starch branching enzyme IIa 1 -
Alg H9NAV2 Superoxide dismutase [Cu-Zn] (Fragment) 1
Alg M7Y8M9 Translocon-associated protein subunit beta 1
Alg M7YBY6 Basic endochitinase C 3 8
Alg M7YCG2 Importin subunit alpha 2 +
Alg M7YE46 Vicilin-like antimicrobial peptides 2-2 3 11
Alg M7YIZ6 Xylulose kinase 1 - - - 1
Isocitrate dehydrogenase [NAD] catalytic subunit 5,
Alg M7YM33 mitochondrial 2
Alg M7YMQ8 GrpE protein homolog 4 1
Alg M7YMT4 Putative aconitate hydratase, cytoplasmic 3 +
Alg M7YPB1 Glutaredoxin-C6 4 + + 1
Alg M7YQA5 Aminopeptidase N 4 3
Alg M7YQW7 Acyl-[acyl-carrier-protein] desaturase 2, chloroplastic 4
Alg M7YT47 zinc finger protein 1
Alg M7YUM7 Intracellular protease 1 3 - +
Alg M7YUQ6 Aspartic proteinase oryzasin-1 1 7
Alg M7YVM9 Chitinase 1 4 + 12
Alg M7Z025 Ankyrin repeat domain-containing protein 2 3 10
Alg M7Z0D7 DNA damage-inducible protein 1 1
Alg M7Z2B3 transmembrane 214-B 1 -
Alg M7Z2D6 amino acid selective channel protein 1
Alg M7Z2J3 Putative NADP-dependent oxidoreductase P1 4 + 10
Alg M7Z3I6 2-oxoglutarate dehydrogenase, mitochondrial 4 2
Alg M7Z3P0 Peroxisomal multifunctional enzyme type 2 3 -
Alg M7Z3U2 non-specific lipid transfer GPI-anchored 2-like 3 -
Alg M7Z534 Glutaredoxin-C8 3 +
Alg M7Z7R6 isoflavone reductase homolog irl-like 2 + - -
Alg M7Z8W1 Cytochrome b5 1 1
Alg M7ZA33 Serpin-ZX 1 - 4
Pyrophosphate--fructose 6-phosphate 1-
Alg M7ZA82 phosphotransferase subunit alpha 2 - -
Alg M7ZDG7 Monothiol glutaredoxin-S11 4 +
Alg M7ZF09 Endogenous alpha-amylase/subtilisin inhibitor 4 10
Alg M7ZGB1 Programmed cell death protein 4 2 + - +
Alg M7ZIQ2 Putative glutathione S-transferase GSTF1 3 + +
Alg M7ZIY2 2'-deoxymugineic-acid 2'-dioxygenase 4
Alg M7ZJJ1 Peptide methionine sulfoxide reductase A2-1 2 1
Alg M7ZKE7 glutathionyl-hydroquinone reductase -like 2 + - -
Alg M7ZMS4 Endoglucanase 11 3 +
Alg M7ZNK7 proton pump-interactor 1-like 1
Alg M7ZQM4 L-ascorbate peroxidase 1, cytosolic 4 15
Alg M7ZRM5 Programmed cell death protein 4 2 - 4
Alg M7ZT31 Gamma-hordein-3 3 - + + + -
Bifunctional aspartokinase/homoserine dehydrogenase 2,
Alg M7ZVP6 chloroplastic 3
Alg M7ZXX2 Pyruvate, phosphate dikinase 1, chloroplastic 3 - -

107
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

Alg M7ZY63 Coatomer subunit gamma 2 - - +

Alg M7ZZM5 Adenosine kinase 2 4 9
Alg M8A0T7 Sucrose synthase 2 1
Alg M8A1S2 Trypsin/alpha-amylase inhibitor CMX1/CMX3 3 - + + -
Alg M8A1U1 UDP-glucuronic acid decarboxylase 1 4 5
Alg M8A2G0 ADP,ATP carrier protein, mitochondrial 1
Alg M8A2R3 Uncharacterized protein 1
Alg M8A3G4 Cell division protein ftsZ-like protein 2-1, chloroplastic 2 - +
Alg M8A710 Clathrin heavy chain 2 - +
Alg M8A7I9 Chitinase 2 4 13
Lipoamide acyltransferase component of branched-chain
Alg M8A8D6 alpha-keto acid dehydrogenase complex, mitochondrial 2 + 1
Alg M8A8V1 hypothetical protein TRIUR3_12434 1 1
Alg M8AMK8 Actin-7 1 - - 4
Alg M8APL9 DEAD-box ATP-dependent RNA helicase 37 2 -
Alg M8AQI0 Inositol-3-phosphate synthase 4
Alg M8AQI5 Avenin-3 3 - + +
Putative bifunctional methylthioribulose-1-phosphate
Alg M8AS42 dehydratase/enolase-phosphatase E1 3 - + +
Alg M8ATK6 Protein argonaute 4A 3
Alg M8AUX2 caleosin 2 1
Alg M8B5G5 Beta-amylase 3 - + +
Alg O04074 Starch branching enzyme 1 1 7
Glucose-1-phosphate adenylyltransferase large subunit,
Alg P12299 chloroplastic/amyloplastic 1
Alg P12782 Phosphoglycerate kinase, chloroplastic 4 14
Alg P38076 Cysteine synthase 2 + - -
Alg P82901 Non-specific lipid-transfer protein 2P 3 2
Alg Q1ALB3 Disproportionating enzyme 1
Alg Q24M29 Starch branching enzyme IIb 3 - + +
Alg Q3S832 Delta-1-pyrroline-5-carboxylate dehydrogenase 3 + +
Alg Q58U44 Ribulose bisphosphate carboxylase large chain 4 12
Alg Q6QFA1 Expansin EXPB3 3 +
Alg Q6RUJ1 Glutamine synthetase 4 1
Alg Q7X9M4 Putative uncharacterized protein (Fragment) 3 - + +
Alg Q8RW02 Glutathione transferase 3 + + 9
Alg Q8RWR5 Beta-D-glucan exohydrolase 2 + - -
Alg Q8S4P7 Thaumatin-like protein 3 + 1
Alg Q8W3V6 Low-molecular-weight glutenin subunit group 6 type IV 3 - + + 1
Alg Q9FUU8 Starch branching enzyme 1 1 7
Alg Q9M4Z1 Glucose-1-phosphate adenylyltransferase 1
Alg Q9ZRB0 Tubulin beta-3 chain 3 +
Alg T1LJ90 glutathione s-transferase 3 4 1
phospho-2-dehydro-3-deoxyheptonate aldolase
Alg T1MV14 chloroplastic-like 2 + +
Alg T1N8M0 probable aldo-keto reductase 1-like 3 +
Alg T1NLZ4 Uncharacterized protein 3 +
Alg W4ZUH4 importin subunit beta-1-like 2 2
Alg W4ZXS7 12s seed storage globulin 1-like 4
Alg W4ZY98 stromal 70 kda heat shock-related chloroplastic 1
Alg W5A2X8 universal stress protein a-like protein 1 1
pyrophosphate-energized vacuolar membrane proton
Alg W5A845 pump 1-like 1
Alg W5AA91 Importin subunit alpha 2 + + - +
Alg W5AS47 clathrin light chain 2-like 2
Alg W5AUD8 glutamate--glyoxylate aminotransferase 2 2
Alg W5AUJ3 ECERIFERUM 26-like 4
Alg W5AWM0 methylthioribose kinase 2 + - -
Alg W5B0Q9 Xylose isomerase 3 +
Alg W5BD57 chaperone protein chloroplastic-like 2 + - - +
Alg W5BEK6 Uncharacterized protein 2 + - +
Alg W5BKC3 udp-glucuronic acid decarboxylase 2-like 1 2
Alg W5BLP9 Uncharacterized protein (Fragment) 2 + - -
Alg W5CCM5 carbonic anhydrase 3 - + +
Alg W5CFJ5 flower-specific gamma-thionin precursor 3 + -
Alg W5CPX1 uridine 5 -monophosphate synthase-like 4 2
Alg W5CR31 cationic peroxidase SPC4-like 3 11
Alg W5D1Z6 pectin acetylesterase 5-like 1 15
Alg W5DGP3 pyruvate decarboxylase 3 + + 8
Alg W5DLU4 glutamate-rich wd repeat-containing protein 1 4
Alg W5DP16 UDP-glucose 6-dehydrogenase 3 - +

108
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

Alg W5DR83 gibberellin 20 oxidase 2 2 + - - -

Alg W5DS49 rna recognition motif containing family protein 1 4
Alg W5DWS0 1,2-dihydroxy-3-keto-5-methylthiopentene dioxygenase 1
Alg W5DZZ4 transport SEC23-like 2 +
Alg W5E0G4 vicilin-like antimicrobial peptides 2-2 3 11
Alg W5E0G5 cop9 signalosome complex subunit 4 4
Alg W5E5E8 Alcohol dehydrogenase-like 2 3 9
Alg W5EKP0 cop9 signalosome complex subunit 1 2
Alg W5EM56 phosphoenolpyruvate carboxykinase 3 +
Alg W5EQ81 cbs domain-containing protein 3
Alg W5F7H1 probable n-succinyldiaminopimelate aminotransferase 2 - -
Alg W5F9D8 probable sarcosine oxidase 1 10
Alg W5F9U4 probable sarcosine oxidase 1 7
Alg W5FBU7 ripening-related 1 4 + + 9
Alg W5FEQ2 prohibitin- mitochondrial-like 1
alpha,alpha-trehalose-phosphate synthase [UDP-forming]
Alg W5FF37 6-like 2 - +
Alg W5FHA6 Oleosin 1 14
Alg W5FZQ1 rRNA N-glycosidase 1 5
Alg W5G958 acetolactate synthase small subunit 2, chloroplastic-like 2 - 1
Alg W5GC44 hyoscyamine 6-dioxygenase 3 + +
Alg W5GWT6 aspartic proteinase oryzasin-1-like 1
Alg W5H829 protein kinase superfamily protein 2 - -
probable gamma-aminobutyrate transaminase 3,
Alg W5HAJ8 mitochondrial 2 - -
Alg W5HFK6 Formate dehydrogenase, mitochondrial 1
Alg W5HXV0 fructokinase-2 3 - + 1
Dolichyl-diphosphooligosaccharide-- glycosyltransferase
Alg W5HY12 48 kDa subunit 1 -
Alg W5I774 sucrose synthase 4 + 8

109
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

3. CONCLUSIONS
Dans notre étude, les traitements nutritionnels appliqués à 200°Cj après floraison ont
entrainé des quantités et des durées d’accumulation différentes des éléments N et S au sein du
grain de T. monococcum. Cela a conduit à une modification du ratio N/S qui a été très impacté
par un fort apport d’N sans apport conjoint de S. Dans cette condition, ce rapport N/S avoisinait
30, synonyme d’une sévère déficience en S (Randall et al., 1981). A maturité du grain, la
composition en PR a été nettement modifiée, avec une augmentation de l’accumulation des PR
riches en S avec les traitements soufrés et une augmentation de celle des PR pauvres en S avec
les traitements azotés. Ces résultats confirment ainsi ceux rapportés dans de nombreuses études
chez le blé tendre (Wieser et al., 2004 ; Zörb et al., 2010 ; Dai et al., 2015). L’apport de S a
entrainé un taux élevé de gliadines, expliqué par une forte augmentation de la vitesse et de la
durée d’accumulation des α/β- et γ-gliadines. Au final, le rapport gliadines/gluténines qui
influence les propriétés du gluten a été clairement modifié par les différents traitements.
L’analyse des protéines nucléaires d’une part et des albumines globulines d’autre part a
permis d’identifier et de quantifier de nombreuses protéines et ainsi d’obtenir une bonne vision
du protéome du grain de T. monococcum en réponse à l’apport de N et de S. Même si cela n’a
pas été discuté dans cette étude consacrée à la réponse du protéome à la nutrition, le nombre
élevé de protéines nucléaires identifiées (1677) apporte de nouvelles informations sur les
acteurs protéiques rencontrés dans le noyau de la cellule végétale.
Si la modification du ratio N/S a été responsable des changements de la composition en
PR du grain, nous avons également supposé qu’il fût à l’origine des variations quantitatives
observées pour de nombreuses albumines globulines et protéines nucléaires au cours du
remplissage du grain. En effet, beaucoup de ces protéines ont présenté des variations
d’abondance entre les deux conditions où l’N ou le S avaient été apportés seuls à la culture.
Parmi les changements observés, l’augmentation de la quantité de protéines impliquées dans le
métabolisme des acides aminés a été observée lorsqu’un fort apport d’N a été effectué, à
l’inverse de plusieurs protéines liées au métabolisme des sucres.
Les protéines nucléaires ont été moins affectées par le changement de nutrition que les
albumines globulines. La plupart de celles qui ont été impactées l’ont été en réponse à un apport
élevé d’N et de S. Parmi elles, de nombreuses protéines impliquées dans la maturation des
ARNm ont été retrouvées, synonyme d’une activation de la transcription dans cette condition.
Les protéines de liaison à l’ADN ont quant à elle présenté des réponses très variées, suggérant
une action isolée de ces protéines et spécifique d’un état nutritionnel du grain. Parmi ces
réponses, une régulation par des mécanismes épigénétiques via la méthylation de l’ADN
semblerait être activée suite à un fort apport de S. Les résultats observés au niveau des protéines
impliquées dans le transport nucléo-cytoplasmique témoignent de l’importance des échanges
d’ARN et de protéines entre le noyau et le cytoplasme. Même si peu de protéines nucléaires ont
été impactées par la nutrition, le noyau semble occuper une place centrale dans l’ajustement de
la composition du grain suite à différents apports de N et de S.
Les réseaux biologiques générés nous ont permis de visualiser les protéines réagissant de
la même manière aux apports d’N et de S. Il est probable que ces groupes de protéines sont en
partie responsables de l’ajustement du métabolisme cellulaire, à l’origine de la modification de

110
 Chapitre 4 : La réponse protéomique du grain de Triticum monococcum à la nutrition N et S

la composition du grain en PR. L’intégration de ces données de protéomique avec d’autres types
de données -omiques semble importante pour détecter des acteurs centraux de la réponse à la
nutrition, qui pourraient faire l’objet d’études plus approfondies.

Références
Dai, Z., Plessis, A., Vincent, J., et al. (2015) Transcriptional and metabolic alternations
rebalance wheat grain storage protein accumulation under variable nitrogen and sulfur
supply. Plant J., 83, 1–18.
Randall, P., Spencer, K. and Freney, J. (1981) Sulfur and nitrogen fertilizer effects on wheat.
I. Concentrations of sulfur and nitrogen and the nitrogen to sulfur ratio in grain, in relation
to the yield response. Aust. J. Agric. Res., 32, 203.
Wieser, H., Gutser, R. and Tucher, S. von (2004) Influence of sulphur fertilisation on
quantities and proportions of gluten protein types in wheat flour. J. Cereal Sci., 40, 239–
244.
Zörb, C., Grover, C., Steinfurth, D. and Hermann Mühling, K. (2010) Quantitative
proteome analysis of wheat gluten as influenced by N and S nutrition. Plant Soil, 327,
225–234.

111
112
 CHAPITRE 5 :
La réponse intégrative du grain
de Triticum monococcum à la
nutrition azotée et soufrée

113
114
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

1. INTRODUCTION
Pour compléter l’étude des protéines nucléaires et des albumines globulines, des analyses
du transcriptome et de plusieurs métabolites à partir du même matériel biologique ont été
réalisées. Des études récentes chez le blé et le seigle montrant l’impact de la disponibilité en N
et S sur l’expression de nombreux gènes du grain liés au métabolisme cellulaire et au transport
des acides aminés ainsi que sur des régulateurs transcriptionnels ont motivé ces analyses (Dai
et al., 2015 ; Postles et al., 2016).
Pour la partie transcriptomique, les ARN ont été quantifiés par séquençage d’ARN (RNA-
Seq), entre 300 et 600°Cj après floraison pour observer les effets de la nutrition pendant le
remplissage du grain, ainsi qu’à 100 et 200°Cj. Les métabolites étudiés correspondent à des
sucres, des acides aminés, des acides organiques, le glutathion, ainsi que quelques dérivés
aminés et nucléosides. Ils ont été analysés sur l’ensemble du développement du grain, de 100 à
950°Cj après floraison.
Après avoir analysé l’impact de l’N et du S sur l’expression des gènes d’une part et sur
l’accumulation des métabolites d’autre part, nous avons intégré les données acquises à ces
différentes échelles (transcrits, métabolites) en incluant les résultats de protéomique obtenus
dans le chapitre précédent. L’objectif a ainsi été d’obtenir une vision globale de la réponse du
grain de T. monococcum à des apports contrastés en N et S. La stratégie adoptée a été de générer
des réseaux biologiques de co-expression/co-accumulation à partir de l’ensemble des variables
répondant significativement à la nutrition. Un premier réseau a été généré afin de lier les
ARN/protéines/métabolites ayant les mêmes profils d’abondance pendant le remplissage du
grain. Le but a été de visualiser des groupes de variables intervenant potentiellement dans les
mêmes voies de réponse. Puis comme dans le précédent chapitre, les variables répondant de la
même façon aux différents traitements pendant le remplissage du grain ont été groupées au sein
d’un second réseau. L’utilisation de ces deux types de réseaux, porteurs d’informations
complémentaires, nous a semblé pertinente pour identifier les variables ayant des profils
d’abondance similaires à la fois sur la cinétique étudiée et en réponse aux traitements azotés et
soufrés.
La finalité de cette étude est de faire le lien entre le facteur environnemental étudié, la
nutrition N et S et le phénotype observé, la composition du grain en PR, en considérant la
réponse cellulaire comme un système dont les parties, qui correspondent aux différents types
de données -omiques, sont fortement connectées (Figure 24). L’idée sous-jacente à ce travail
est de comprendre cette réponse afin d’identifier les leviers d’action pour maitriser la
composition protéique dans un contexte de limitation des engrais.

115
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

Figure 24. Représentation schématique de l’objectif du chapitre 5 : intégrer les réponses du grain
de T. monococcum à l’apport de N et S aux différentes échelles mesurées (transcrits, protéines,
métabolites) pour faire le lien entre la nutrition et la composition en protéines de réserve.

Références :
Dai, Z., Plessis, A., Vincent, J., et al. (2015) Transcriptional and metabolic alternations
rebalance wheat grain storage protein accumulation under variable nitrogen and sulfur
supply. Plant J., 83, 1–18.
Postles, J., Curtis, T.Y., Powers, S.J., Elmore, J.S., Mottram, D.S. and Halford, N.G.
(2016) Changes in free amino acid concentration in rye grain in response to nitrogen and
sulfur availability, and expression analysis of genes involved in asparagine metabolism.
Front. Plant Sci., 7, 1–11.

2. ARTICLE 4 : Integrative response of the einkorn (Triticum monococcum) grain to

nitrogen and sulphur nutrition
Titouan Bonnot1, Victor Hatte1, Emmanuelle Bancel1, Mireille Dardervet1, Philippe
Leroy1, Julie Boudet1, Annick Moing2, Yves Gibbon2, Marie Pailloux3, Pierre Martre1 and
Catherine Ravel1
1
UMR GDEC, INRA, Université Blaise Pascal, 5 chemin de Beaulieu, 63 177 Aubière, France
2
INRA, UMR1332 Biologie du Fruit et Pathologie, Villenave d’Ornon F-33 882, France
3
LIMOS, CNRS, Université Blaise Pascal, 63 173 Aubière, France
(Article en préparation)

116
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

Integrative response of einkorn (Triticum monococcum) grain to nitrogen and

sulphur nutrition
Titouan Bonnot1, Victor Hatte1, Emmanuelle Bancel1, Mireille Dardervet1, Philippe Leroy1,
Julie Boudet1, Annick Moing2, Yves Gibbon2, Marie Pailloux3, Catherine Ravel1,* and Pierre
Martre1, †
1
UMR GDEC, INRA, Université Blaise Pascal, 5 chemin de Beaulieu, 63 177 Aubière, France
2
INRA, UMR1332 Biologie du Fruit et Pathologie, Villenave d’Ornon F-33 882, France
3
LIMOS, CNRS, Université Blaise Pascal, 63 173 Aubière, France
†
Present address: UMR LEPSE, INRA, Montpellier SupAgro, 34 060 Montpellier, France

Abstract: In wheat, one major challenge is to maintain grain quality while increasing yield and preserving
the environment. To achieve this goal, it is essential to considerate the nitrogen (N) and sulfur (S) nutrition,
which has long been known for its impact on the grain composition and particularly grain storage protein
(GSP) composition. The identification and understanding of molecular mechanisms regulated by the
nutrition during grain development is necessary. In this study four treatments with various N and S amounts
were applied to an einkorn (Triticum monococcum) culture during grain development. Transcriptome and
metabolite analyses performed at different time points and for the different treatments have highlighted the
strong impact of S deficiency in the grain response to nutrition, conducting to the activation of amino acid
and S transport and metabolism. An integrative view of this response was obtained by supplementing
transcriptome and metabolome with proteomic data previously obtained from the same biological material.
Major changes were observed when N-to-S ratio was disturbed in grain. Several transporters and genes
involved in transcription process were revealed in network analysis and can have a central role in the grain
adaptation to such changes. Finally, the large dataset obtained provides information for the identification
of genes and pathways involved in the response to nutrition, which could allow a better control of the grain
composition and thus its quality.
Keywords: einkorn, nitrogen, sulfur, grain, transcriptome, metabolites, proteins, co-expression network

INTRODUCTION and glutenins, which form large macropolymers

during grain maturation (Don et al., 2006).
During grain development, storage compounds
Together they represent major constituents of
accumulate to form a nutrient stock used during
gluten and determine the rheological properties of
germination process and first steps of seedling
dough, with glutenins that are mainly responsible
growth before acquisition of autotrophy. Among for viscoelastic properties and gliadins that confer
storages in wheat, grain storage proteins (GSP) extensibility to dough (Branlard et al., 2001).
provide N and S resources and account for 60 to Gliadins are subdivided into ω1,2-, ω5-, α/β- and
80% of the total protein amount in mature grain
γ-gliadin classes, and low-molecular-weight-
(Shewry and Halford, 2002). In addition to their
subunits (LMW-GS) and high-molecular-weight-
essential role in seedling nutrition, quantity and
subunits (HMW-GS) subunits are distinguished
composition of GSP in mature grain determine its
among glutenins (Wieser, 2007). Proportion of
end-use value for food. They are divided into two
these different classes/subunits determines grain
fractions: gliadins, which are monomeric proteins quality.

117
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

N and S are two essential elements for plant accumulation and particularly GSP accumulation.
nutrition, growth and accumulation of storages in Because physiological responses rely on complex
grain. Indeed they represent two major constituents networks of interactions at different scales, one
of proteins and other organic compounds and many strategy to meet this objective is to use large-scale
studies have previously demonstrated connexions ‘omics’ analysis, and combine information
between their respective assimilation (Kopriva and obtained at transcript, protein and metabolite levels
Rennenberg, 2004; Jamal et al., 2010; Sosa- (Fukushima et al., 2009). In addition, recent
Valencia et al., 2013). In wheat grain, N supply advances in high-throughput technologies and tool
leads to an increase GSP concentration and development for data integration allow the
duration of accumulation (Pechanek et al., 1997; inference of regulatory networks in plants
Triboï et al., 2003; Chope et al., 2014). N and S (Moreno-Risueno et al., 2010; Bassel et al., 2012).
availability in soil strongly impact GSP Nowadays, this approach is used to study plant
composition in mature grain. Indeed GSP differ in responses to stress (Urano et al., 2010), including
their composition in S-amino acids (cysteine and nutritional stresses (Hirai et al., 2004; Dai et al.,
methionine) and S-rich GSP (α/β- and γ-gliadin, 2015). In bread wheat, genes involved in transport
LMW-GS) are distinguished from S-poor GSP and metabolism were identified in the grain
(ω1,2-, ω5-gliadin and HMW-GS; Shewry et al., response to nutrition, which influenced
2001). S-deficiency decreases amount of S-amino accumulation of several amino acids, suggested to
acids (Wrigley et al., 1980), conducting to a determine GSP expression (Dai et al., 2015). We
decrease amount of S-rich GSP, while an increase assume that post flowering N and S supply disturbs
amount was observed for S-poor GSP (Zhao et al., cell metabolism and functioning of the wheat grain
1999; Wieser et al., 2004; Zörb et al., 2010). N and at different levels, by modifying several gene
S are therefore important to control GSP expression, protein and metabolite amounts,
composition and grain N-to-S ratio, generally conducting to changes in GSP composition.
varying from 12 to 25 (Randall et al., 1981), is a The aim of our study was to obtain an integrative
good indicator of grain quality (Wrigley et al., response of the grain to the N and S supply. We
1980). used einkorn (Triticum monococcum), a wheat
But since many years, S deficiency in soils is diploid species whose genome is close to the A
reported (Tisdale et al., 1986; Zhao et al., 1999). genome of breed wheat (T. aestivum, Marcussen et
This is explained by the reduction of industrial SO2 al., 2014). Four treatments containing different N
emissions and by a supply of N to the culture and S amounts were applied to an einkorn culture
without S addition, increasing N-to-S ratio and after flowering state and grains were collected at
consequently conducting to S-deficient condition. different times after nutritional change to perform
Besides impacting GSP composition, S deficiency transcriptome and metabolome analyses. Networks
also affects plant growth, decreases yield (Zhao et generated from the 386 differentially expressed
al., 1999; Hawkesford, 2000) and reduces plant genes, 14 metabolites significantly impacted by
resistance to biotic and abiotic stresses (Rausch nutrition, supplemented by proteomic data
and Wachter, 2005; Kruse et al., 2007) It makes S previously obtained (Bonnot et al., submitted),
one of the most limiting nutrients for agricultural have revealed co-expression/accumulation of gene
production (Eriksen, 2009). Since many years, products involved in storage, amino acid
excessive supply of N to the crop in order to metabolism, sulfur metabolism and transcription
increase yield causes environmental perturbations, processes. S deficiency highly perturbed gene
like water eutrophication (Hirel et al., 2007). To expression and protein accumulation. Despite
overcome this problem, directives have been connections between N and S assimilations, few
implemented to reduce N input to the crop. N and gene products were similarly impacted by these
S supply are therefore to be reconsidered and two elements (i.e. increased by N and S or
represent an agronomic issue for wheat grain decreased by N and S). Several genes involved in
quality. transcription regulation could be central to the
In this context, it is necessary to identify and network response and could influence GSP
understand regulatory mechanisms involved in the accumulation.
grain response to nutrition that influence storage

118
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

RESULTS (100, 200, 300, 400, 500, 600°Cd) four treatments

(N-S-, N+S-, N-S+, N+S+) from 300 to 600°Cd
Transcriptome analysis revealed 71,5% of
after anthesis, and three biological replicates.
expressed genes in the einkorn grain
Paired-end sequencing produced 16,386,593 reads
To investigate effects of post flowering N and S per sample in mean. Reads were then mapped
supply on the einkorn grain at transcriptional, against the T. monococcum genome at
metabolic and protein level, four treatments were [Link]
applied to a T. monococcum culture: a low N and S wheat/IWGSC/IWGSC_genePredictions_of_other
treatment called N-S-, a high N treatment called _wheat_species/, which contains 32,047
N+S-, a high S treatment called N-S+ and a fourth structurally annotated genes. In mean, 94.2% of the
treatment with high N and S amounts, called reads were successfully mapped at unique mapping
N+S+. Previous results obtained at protein level position. Then only genes with Counts Per Million
revealed that N and S modified GSP composition (CPM) value > 1 were considered as expressed
in mature grain, explained by changes in the rate (Table S1). A total of 22,910 (71.5%) genes was
and/or the duration of their accumulation (Bonnot expressed in at least one sample. From these genes,
et al., submitted). These results were associated to a large majority was expressed in all conditions and
changes observed for several nuclear proteins and 9,265 (40%) were expressed in all samples (Figure
albumin-globulin including some appearing as 1a). However, the number of expressed genes was
central actors to the proteome response. somewhat different between stages and/or nutrition
The biological material used in Bonnot et al. treatments (Figure 1b). For instance, the number of
(submitted) was used for a transcriptome analysis expressed genes at 600°Cd after anthesis was lower
based on RNA sequencing from 54 samples for all treatments than in other stages.
corresponding to six thermal time after anthesis

Figure 1. Number of expressed genes in einkorn grain and their functional classification. (a) Histogram showing
number of expressed genes per number of samples. (b) Radar plot of number of expressed genes per thermal time
after anthesis (300, 400, 500 and 600°Cd) and per nutrition (N-S-, N+S-, N-S+, N+S+). Expressed genes are classified
into biological processes (c) and molecular functions (d) according to their Gene Ontology annotation.

119
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

Gene ontology was searched for the 22,910 clusters. For example, the cluster 3 included genes
expressed genes. They are classified into 25 expressed, in average, more intensely in the N+S+
biological processes and 31 molecular functions treatment than in the N+S- treatment. Cluster 2
(Figure 1c and Figure 1d, respectively). grouped 95 genes whose expression level was
For biological processes, the biggest class clearly increased with the N+S- treatment, from
corresponded to cellular protein modification 400 to 600°Cd after anthesis. The biggest
process (30%) that includes genes involved in difference is observed for the comparison between
protein metabolic process, regulation of N+S- and N-S+ treatments.
transcription, proteolysis, mRNA processing and Gene ontology enrichment was then investigated
others. The second biggest process was electron for the 386 genes significantly impacted by the
transport chain process (26%). nutrition, compared to all expressed genes (Table
Then many genes were classified in localization 1). Thereby, significant genes were enriched in
(9%), primary metabolism (8%), vesicle-mediated seven biological processes, corresponding to plant
transport (8%) and response to stimulus (5%). defense (defense response, killing of cells of other
Protein binding was the molecular function most organism, heat acclimation), metabolism (cellular
represented among the expressed genes (37%), catabolic process, asparagine biosynthetic process,
followed by translation factor activity, nucleic acid lipid phosphorylation) and transport (organic acid
binding (20%), which mainly corresponds to genes transport, Table 1) processes. Enrichment was also
involved in DNA or nucleic acid binding. Then observed for 14 molecular functions related to
many genes possess amino acid binding (9%) and storages and degradation inhibition (nutrient
transferase (12%) activities. reservoir activity, peptidase inhibitor activity,
serine-type endopeptidase inhibitor activity), and
binding (zinc ion binding, phospholipid binding,
N and S significantly impacted 386 genes during lipid binding, chitin binding, Table 1).
grain filling
N and S supply highly impacts GSP
N and S impacted metabolite pools during
accumulation during grain filling. So we expected
einkorn grain development
changes at transcript level for genes involved in
amino acid metabolism and storage accumulation. To investigate effects of N and S on amino acid
From the 22,910 expressed genes, 386 (1.7%) were pools and carbon (C) and S metabolisms, which
differentially expressed between at least two could influence storage accumulation in grain,
treatments and at one time point during grain filling several metabolites were quantified. We measured
(from 300 to 600°Cd after anthesis, Tables S1 and from 100 to 950°Cd after anthesis 15 amino acids,
S3). Principal component analysis was performed five sugars, five organic acids, the S-containing
from expression data for these 386 genes (Figure tripeptide glutathione, amine derivatives and
2a). First axis explained 37.2% of total variation nucleosides or derivatives (Table S2). Statistical
and clearly separates the time points of the grain analysis revealed that 14 metabolites were
development (300, 400, 500 and 600°Cd). Second impacted by nutrition. They were replaced in their
axis explained 22.3% of the variation and separates respective pathway and their variations observed
the N+S- treatment from others, particularly from during grain development and with the four
N-S+ treatment. treatments are schematically represented in Figure
Hierarchical clustering analysis based on 3.
principal components summarized these different N, applied alone or with S (N+S- and N+S+
observations (Figure 2b). Three clusters were built. treatments) have clearly increased total amino acid
Clusters 1 and 3 grouped 182 and 109 genes pool in the grain, compared to the low N treatments
respectively, whose expression level was strongly (N-S- and N-S+). In details, major effects were
determined by developmental stages. Cluster 1 observed for aspartic acid (Asp), lysine (Lys) and
included genes whose expression decreased during glutamine (Gln) that showed fairly similar
grain development while opposite profile was response profiles as they were both increased with
found for genes in cluster 3. Some differences N+S- and N+S+ treatments (Figures 3 and S1).
between treatments were found in these two

120
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

Figure 2. Clustering analysis of differentially expressed genes in response to N and S supply. Principal component
analysis (a) from expression data of differential genes allowed the establishment of three expression profiles (b). The
number of genes in each cluster is indicated at the top of each panel. Line represent the mean for each nutrition.

121
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

Table 1: GO enrichment analysis of Biological process and Molecular function terms for the 386 differential
expressed genes in response to nutrition. GO terms enriched in the set of genes that were significantly impacted by
nutrition are indicated. For each term, number of genes present in all expressed genes and in significant genes lists
are indicated in ‘Annotated’ and ‘Significant’ columns, respectively. Number of genes expected in the list of
significant genes is indicated in ‘Expected’ column. Fisher’s exact tests were performed to investigate significant
difference in GO term enrichment between annotated and significant genes, and Pvalue are indicated.

Biological processes
GO ID Term Annotated Significant Expected P value
GO:0031640 killing of cells of other organism 10 4 0.19 2.3e-05
GO:0044248 cellular catabolic process 117 9 2.20 0.00031
GO:0006952 defense response 102 8 1.92 0.00060
GO:0006529 asparagine biosynthetic process 4 2 0.08 0.00205
GO:0015849 organic acid transport 49 4 0.92 0.01304
GO:0046834 lipid phosphorylation 10 2 0.19 0.01428
GO:0010286 heat acclimation 1 1 0.02 0.01881
Molecular functions
GO ID Term Annotated Significant Expected P value
GO:0045735 nutrient reservoir activity 36 9 0.59 4.4e-09
GO:0008061 chitin binding 2 2 0.03 0.00027
GO:0030414 peptidase inhibitor activity 23 4 0.38 0.00049
GO:0046912 transferase activity, transferring acyl groups 3 2 0.05 0.00080
GO:0008289 lipid binding 97 7 1.60 0.00106
GO:0004553 hydrolase activity, hydrolyzing O-glycosyl compounds 224 11 3.70 0.00115
GO:0004867 serine-type endopeptidase inhibitor activity 5 2 0.08 0.00262
GO:0005543 phospholipid binding 31 3 0.51 0.01409
GO:0004134 4-alpha-glucanotransferase activity 1 1 0.02 0.01650
GO:0004781 sulfate adenylyltransferase (ATP) activity 1 1 0.02 0.01650
GO:0008270 zinc ion binding 432 13 7.13 0.02567
GO:0003852 2-isopropylmalate synthase activity 2 1 0.03 0.03274
GO:0046524 sucrose-phosphate synthase activity 2 1 0.03 0.03274
GO:0005274 allantoin uptake transmembrane transporter activity 3 1 0.05 0.04870

Glutamic acid (Glu) and asparagine (Asn) were oxidative stress, osmoregulator role or could
more abundant in N+S+ condition during grain represent NAD reserve during grain germination
filling, with an early effect on Glu from 400°Cd (Shimizu and Mazzafera, 2000). Interestingly,
after anthesis. Surprisingly, treatments containing trigonelline was strongly decreased at grain
high level of N, S or both negatively impacted maturity in all treatments, except with N+S+. In
isoleucine (Ile) and tryptophane (Trp) amounts this condition, a very high level of trigonelline was
during grain development. Ile was impacted early observed especially from 700°Cd to maturity.
in the development while Trp was impacted later,
mostly during grain maturation. S-deficiency
caused in the N+S- condition decreased S-amino Co-expression network reveal highly
acid methionine, S-tripeptide glutathione and interconnected genes, metabolites and proteins
malic acid, an organic acid that is part of the TCA N and S supply have modified expression level
cycle, necessary in the synthesis of the amino acid for 386 genes and the amount of 14 metabolites.
precursor 2-oxoglutarate (Araujo et al., 2014). Previous results that we obtained from the same
N+S- condition also decreased saccharose biological material have underlined the impact of
amount in grain. Other quantified sugars glucose, nutrition for 209 proteins including GSP, nuclear
fructose, raffinose and trehalose were not affected proteins and albumins-globulins. All the -omic
by treatments, as previously reported on wheat data obtained allowed us to investigate relations
grain (Dai et al., 2015). Trigonelline is an amine between transcripts, metabolites and proteins
derivative and is associated to several during filling, using network analysis.
physiological roles in plants, including response to

122
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

Figure 3. Effects of N and S on several metabolite amounts during einkorn grain development. Amino acids, sugars
and organic acids written in bold were those quantified and metabolites with significant nutrition effect are indicated
in bold and red. Each is represented in its corresponding pathway. Mean of scaled mass per grain of the 14 significant
metabolites are represented on a 4 x 6 grids, with rows and columns representing the four N and S treatments and the
six thermal time after anthesis, respectively. The color of each plot represents the mean of the scaled mass of a given
metabolite according to a color scale from yellow (low scaled mass) to red (high scaled mass).

First, co-expression/co-accumulation between treatments (i.e. 386 transcripts, 209 proteins, 14

the different variables was studied. As described in metabolites), the grain N and S contents (GNC and
Bonnot et al. (submitted), a directed multilevel GSC, respectively) and grain dry weight (DW)
network, based on the discovery of association variables. Associations between these variables
rules between attributes, was constructed using all were built according to the amount of each variable
the -omic data significantly impacted by the in the different samples.

123
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

The resulting network comprised 189 variables N and S not concomitantly applied highly
(nodes) and 1634 linkages (edges), distributed into perturbed grain cell functioning during filling
five modules (Figure 4, Table S3). The biggest
module (Figure 4a) was composed of 97 variables To group genes, metabolites and proteins
highly interconnected since they were involved in strongly and similarly impacted by nutrition during
1114 linkages (68%). They mainly corresponded to filling phase, another network was built (Figure 5).
storage compounds attributes, with all GSP Transcript level or metabolite/protein amount
(LMW-GS and HMW-GS, ω1,2-, ω5- and γ- measured in the different samples from 300 to
gliadin) except α/β-gliadin, several genes and 500°Cd after anthesis were compared in the
proteins involved in carbohydrate metabolic different treatment conditions. Attributes more
process and other with defence role and expressed/accumulated in one treatment compared
degradation inhibition function. It also contained to another were linked to the corresponding
GNC, GSC and DW. Interestingly and only in this treatment comparison. The corresponding network
module, three groups can be distinguished based on comprised the most significant variables in at least
their linkage number: a core constituted of 20 one treatment comparison. It included 499 linkages
variables, mainly proteins, with high linkage (edges) and 275 variables (nodes) including four
number (44 to 64 edges each), a second group with GSP, seven metabolites, 178 transcripts and 86
37 variables including four GSP and N and S proteins (70 albumin-globulin and 16 nuclear
content, with an intermediate linkage number (21 proteins).
to 39 edges each) and a third group constituted of First, the most striking observation was the
41 variables with low linkage number (1 to 12 disparity in the linkage number for the different
edges each, Figure S2a). Two other small modules treatment comparisons (Figures 5 and S2b).
containing genes mainly involved in similar Indeed, the three comparisons containing the N+S-
functions than in module (a) (nutrient reservoir treatment were those with high linkage number,
activity, defence, degradation inhibition) were suggesting that this treatment highly differ from
constituted (Figure 4b and c). others regarding its effect on the einkorn grain. 165
Module (d) contained 22 nodes, with several variables showed significant difference between
genes involved in amino acid and S metabolisms, the N+S- and N-S+ treatments, highlighting that
with two genes of glutathione S-transferase, a these two conditions had the most different effects.
taurine catabolism dioxygenase, a homocysteine S- Interestingly, as previously observed at protein
methyltransferase, a serine level, very few attributes (44) were linked to the N-
hydroxymethyltransferase and an aminotransferase S-/N+S+ comparison. So these two treatments did
(Figure 4d). Interestingly, central to this module not have very different effects, whereas they had
was found a membrane transporter EamA-like the most contrasting N and S amounts. Similar
(G51, Table S3) that was linked to 14 genes and results were observed for N-S+/N+S+ and N-S-/N-
proteins. S+ comparisons, each linked to 33 and 22
The second largest module comprised 53 variables, respectively.
variables and 464 edges (Figure 4e). This module
(e) is very different from module (a) regarding Many attribute groups were constituted on this
functions. It comprised many genes or proteins network, which showed similar response to the
involved in DNA metabolism, transcription, RNA nutrition. Approximately half of attributes (137)
processing or protein modification. Indeed, in this was linked to only one treatment comparison,
module (e) three DNA replication genes, two RNA showing that their abundance only differs between
helicases, a DNA polymerase, several genes or two nutrition conditions. Other attributes were
proteins with DNA or nucleic acid binding associated to two or more treatment comparisons,
activities, five kinases/phosphatase were found, indicating the strong effect of N and S supply on
among others. Four proteins and a gene involved in these variables, which showed very different
carbohydrate metabolic process and the saccharose quantities in the four nutrition conditions.
were also part of this module (e).

124
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

Figure 4. Networks of co-expression/accumulation during grain development of transcripts, metabolites and proteins
impacted by N and S supply. Node shape and color reflect attribute category: white circles, transcripts; green triangles,
metabolites; orange rectangles, GSP; grey diamonds, nuclear proteins; grey hexagons, albumins-globulins; yellow
squares, N and S content (GNC and GSC, respectively) and dry weight (DW). Edge colour indicates the biological
significance between source and target, red when the quantity of RNA/metabolite/protein was high for the two nodes
and blue when the quantity was low for the two nodes. Codes were attributed for genes and proteins, for details see
Table S1. Sacch: saccharose; Glu: glutamate; Asp: aspartic acid.

125
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

In this case, one group stood out, composed of 47 increased by high level of N and low level of S. In
attributes including HMW-GS and N-to-S ratio addition, HMW-GS showed a strong and early
variables which were all increased by N+S- increased rate of accumulation in the N+S-
compared to the three other treatments. condition. In this group of 47 attributes, two sulfate
Interestingly, 16 attributes in this group were part transporters were also found, as well as an amino
of the 22 variables that composed module (d) in co- acid transporter and two cysteine synthase,
expression network (Figure 4d, Table S3). The highlighting the activation of S metabolism and
corresponding genes and proteins, mainly involved amino acid metabolism/transport in S-deficiency
in amino acid and S metabolism were thus strongly condition caused with the N+S- treatment.

Figure 5. Network analysis for gene, metabolite and protein response to N and S nutrition. Genes differentially
expressed and metabolites and proteins differentially accumulated in response to the nutrition were related to the six
treatment comparisons. Node shape and color reflect attribute category: black squares, treatment comparisons; white
circles, genes; green triangles, metabolites; yellow rectangles, GSP; grey diamonds, nuclear proteins; grey hexagons,
albumins-globulins, orange squares, N and S content (GNC and GSC, respectively) and dry weight (DW). Edge
colour indicates the biological significance between source and target, red when the quantity of
transcript/metabolite/protein was higher in the first treatment used in the comparison and blue when the quantity was
lower in the first treatment. Codes were attributed for genes and proteins, for details see Table S1. Phe: phenylalanine;
Trp: tryptophane; Ile: isoleucine; Met: methionine; Mal: malic acid; Gln: glutamine; Asp: aspartic acid.

126
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

The two generated networks (co-expression and reduction, transcription and for Asp, Mal and the
nutrition networks) allowed us to point that 46% of two GSP, LMW-GS and α/β-gliadin. In the same
co-expressed attributes (Figure 4) were also found way, Ile and genes involved in response to stress,
in the nutrition network (Figure 5). The module (d) carbohydrate metabolic process and translation
in co-expression network is present at 95% in the were biological processes both decreased by N and
nutrition network. From these 22 variables, 16 S. But a large majority of variables showed very
(72%) were in the same group, which was strongly different responses to N or S. From this
increased in the N+S- condition. observation it was shown that many genes and
Then the combined effects of N and S supply on proteins acting in S and amino acid metabolism
the einkorn grain were represented in a simplified (including module [d]), transport and transcription
scheme (Figure 6). Based on network analysis, it were increased with N and decreased with S.
summarized the antagonist effect of N and S Conversely, among those increased with S and
evidenced in present study. Indeed, it was found decreased with N, nutrient reservoir activity,
that few variables were increased or decreased both carbohydrate metabolic process and response to
with N and S. However, this was the case for stress are processes that were highly represented.
several genes involved in transport process, oxido-

Figure 6. Overview of main processes impacted by N and S supply during filling. Genes, metabolites and proteins
presented are part of the network present in Figure 5.

127
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

DISCUSSION the GSP polymerization and thus impact dough

properties (Tea et al., 2005).
In wheat, the GSP composition defines dough
More generally, N+S- treatment was shown to be
properties and thus the baking quality.
very different from others in the different analysis.
Identification of genes and molecular mechanisms
Principal component analysis and hierarchical
impacted by N and S supply in the wheat grain
clustering performed on the 386 significant genes
appears necessary to understand the importance of
highlighted in the transcriptome analysis revealed
nutrition, known for a long time for its effect on the
a clear distinction of this condition from others.
GSP composition in mature grain. The aim of
This observation was confirmed with the results
present study was to obtain an integrative view of
obtained from all -omic datasets. In network
the grain response to N and S supply and highlight
analysis, many genes and proteins and several
genes and mechanisms that could be essential in
metabolites were connected to the treatment
this regulation. To this end, we have conducted an
comparisons which included N+S- condition.
experiment in einkorn. Analysis of metabolites and
Among them, some were strongly increased in this
transcriptome, combined to results previously
condition, compared to the three others. Several
obtained from the same biological material at
were part of the same module, which showed their
protein level, allowed us to show the high impact
strong co-expression/accumulation. Their
of N supply which increased amino acid
functional annotation supposed an activation of
metabolism and also S metabolism and transport,
amino acid metabolism and transport with a high
probably to compensate for low S availability. We
level of N. This is in accordance with results
therefore observed that major changes in gene
obtained at metabolite level, with the increased
expression or metabolite/protein accumulation
amount of total amino acid pool in the high N
were explained by a non-concomitant supply of N
treatments, which was previously observed in
and S to the culture, which highly perturbed the N-
several studies (Howarth et al., 2008; Dai et al.,
to-S ratio.
2015). HMW-GS were also strongly and early
increased in this high N condition, without any S
High effect of N nutrition highlighted the strong supply. They are part of S-poor GSP, since S-
incidence of S-deficiency on the einkorn grain amino acid represent 0.8% of its amino acid total,
while they represent, for example, 4.6% in the S-
In present study, the four treatments applied to rich γ-gliadin (Shewry et al., 2001).
the einkorn culture corresponded to a low N and S Several genes involved in S metabolism and
condition (N-S-), a high N treatment (N+S-), a S transport were co-expressed and increased in the
treatment (N-S+) and a fourth treatment with a N+S- condition. This suggests an activation of S
concomitant supply of N and S (N+S+). These assimilation, probably to compensate for S
different conditions have clearly modified GSP deficiency. This plant response to S-starvation is
composition at maturity, explained by an early relatively well-understood and reported in several
modification of rate of accumulation during filling studies (Nikiforova et al., 2003; Hawkesford and
(Bonnot et al., submitted). Since GSP differ in their DeKok, 2006).
composition in S-amino acids, N-to-S ratio in grain In our study, S supply was associated to an
is responsible for the balance between S-rich and increased amount of GSP, rate of accumulation of
S-poor GSP synthesis. In our experiment, N-to-S S-rich GSP and generally to an increased
ratio was strongly increased in the N+S- condition, expression level and protein accumulation for
from 300°Cd after anthesis and it exceeded 30 at genes involved in reserve establishment and
600°Cd. This corresponded to a S-deficient carbohydrate metabolic process. Taking these
condition due to a high amount of N without results together, we hypothesized that lack of S in
concomitant S supply (Zhao et al., 1999; Eriksen, a condition of high N supply probably highly
2009). This condition leads to decrease gliadin-to- perturbed einkorn grain to limit the impact on the
glutenin ratio, known to be negatively correlated to amount of reserves available in grain that are
bread loaf volume (Dhaka and Khatkar, 2015). essential to the germination process.
Moreover, a decreased amount was also observed
for the S-tripeptide glutathione, known to modify

128
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

Several genes highly perturbed by the N to S (Figure 4d, Table S3). EamA family is not well
ratio modification could be central in functionally characterized in plant, but it is part of
controlling GSP composition the drug/metabolite transporter superfamily
(Västermark et al., 2011) involved in metabolite
In our study, results obtained in the N-S-
efflux, and particularly O-acetylserine and cysteine
condition were quite surprising and interesting.
in Escherichia coli (Franke et al., 2003). Thus this
Indeed, the lack of N and S supply with this
transporter probably was central to the grain
treatment resulted in correct amount and
response and adaptation to S-deficiency and
composition of GSP in mature grain. Even though
probably acts with the two sulfate transporters
total GSP amount was increased in the N+S+
which showed a similar response to nutrition.
condition, amount obtained with N-S- was not
Several genes with DNA binding activity and
lower than in the N+S- condition and GSP
involved in transcription process were revealed in
composition was comparable between N-S- and
present study, including some which showed very
N+S+ treatments. It is widely understood that the
different expression level or protein accumulation
major source of grain N is accumulated before
between N+S- and N-S+ conditions. A CBF
anthesis. Indeed from 50 to 95% of grain N comes
transcription factor was more expressed with high
from its remobilization from shoots and roots
level of N than S. CBF transcription factors can act
where it was stored (Palta and Fillery, 1995;
in multiple stress responses as previously reported
Kichey et al., 2007). We therefore hypothesized
(Akhtar et al., 2012; Kidokoro et al., 2015).
that in our experiment and in the N-S- condition, a
Another gene with similar response profile
sufficient accumulation of N and S before
possessed a plant homeodomain (PHD), thought to
flowering and an activation of their remobilization
be involved in chromatin-mediated transcriptional
from the vegetative organs to the grains during
regulation (Sanchez and Zhou, 2011). These
development occurred.
results suggest a control mechanism, triggered by
In our results, N-to-S ratio was almost the same
the disruption of the N-to-S ratio, involving
between N+S+ and N-S- conditions. Many genes
transcriptional regulators and transporters for
and proteins seemed to be increased by N and
regulating the balance and therefore the grain
decreased by S, or conversely, demonstrating an
composition.
antagonist effect of N and S on these genes. This
explains why major differences were observed
when only N or only S were applied to the culture.
Genes, proteins and metabolites which did not
We therefore hypothesize that these genes were
emerge from network analysis
regulated by the N-to-S ratio. In addition, network
analysis revealed that N-S- and N+S+ did not have In this study, we performed two types of network
very different effect on transcripts, metabolites and analysis to investigate co-expression and co-
proteins, since only 44 variables showed accumulation of genes, proteins and metabolites
significant changes between these two conditions. and to group those similarly impacted by N and S
Taken together, these results showed that a high N supply. Several variables were both revealed in
and S supply slightly affected the einkorn grain, these two analysis, like attributes found in module
while the modification in N-to-S ratio seemed to (d) and discussed below. But several variables
impact an important number of genes and proteins. were absent from the co-expression network or the
Among them, some could be involved in the nutrition network or both.
plant response and adaptation to a S or N In module (e), 53 highly interconnected variables
deficiency, to control GSP synthesis and mainly involved in DNA transcription, replication,
composition. As we mentioned above, activation mRNA processing, protein modification, among
of S and amino acid metabolisms was observed in other functions, were revealed. Only 10 (19%)
S-deficient condition, when N-to-S ratio was were present in the nutrition network, which
clearly increased. In network analysis, co- contained attributes presenting a clear different
expression of genes and proteins involved in this expression level or amount between at least two
response have highlighted a membrane EamA-like treatments. Thus some variables significantly
transporter, central to the corresponding module impacted by nutrition could be absent from
network analyses because of the thresholds that we

129
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

used for the parameters that define rule quality. central to the grain response, which deserve further
However, these variables have to be considered investigation.
too. Indeed, some transcripts, proteins or
metabolites could be slightly but significantly
impacted by the nutrition, at one time point during EXPERIMENTAL PROCEDURES
grain development, and yet could have a high
impact in the cell. It was previously reported the Plant material
dynamics of co-expression network across plant Triticum monococcum ssp. monococcum
development, in different cell types or in response (accession ERGE 35821) seeds were germinated (5
to a nutritional stress (Brady et al., 2007; Wei et February) at room temperature on filter paper in
al., 2013; Serin et al., 2016). Some genes present Petri dishes moistened with demineralized water.
in module (e) could correspond to key When the radicles were 0.5 cm to 1 cm long,
transcriptional regulators temporarily impacted by seedlings were transplanted to 50 mL PVC
N and S, which could conduct to major effect on
columns (7.5 cm i.d. × 50 cm deep; 2 seedlings per
the grain adaptation to nutritional changes. For
column) filled with a 2:1 (v/v) mixture of washed
example, a gene with OB-fold nucleic acid binding
perlite:river sand. The columns were arranged in a
domain was revealed and was connected to 26
other variables. greenhouse to form a homogenous stand with a
An ABC transporter was linked to 24 other genes population density of 512 plant m-2. The
or proteins in module (a) but was absent from the experimental design was a randomized complete
nutrition network. ABC transporters represent one block design with four blocks and four treatments.
of the largest protein families and are involved in The high plant density inhibited the development
such diverse processes as plant nutrition, response of axillary tillers which coordinated the
to abiotic stimulus or plant pathogen defence development of the plants within and between the
(Kang et al., 2011). Since this ABC transporter was containers. In the greenhouse air temperature at the
present in module (a), it was co-expressed with top of the plant stand was controlled at 20°C/15°C
several other genes involved in storage (day/night) and air relative humidity was
accumulation and it therefore could be responsible controlled at 55%/75% (day/night). Plant received
for adjustment in storage accumulation in response a mean total daily photosynthetic photon flux
to N and S supply.
(PPF) of 152 and 161 µmol m-2 d-1 between
Moreover, genes presenting atypical profile, not
transplanting and anthesis and between anthesis
present in co-expression network can act alone and
and grain ripeness maturity, respectively, over the
yet have an important role in the grain cell. In this
case, a multiprotein bridging factor (MBF) was light period (16-h light/8-h dark). Air temperature
impacted by nutrition but absent from networks. and relative humidity were measured in miniature
Nevertheless MBF can be central in the grain forced-draft flues placed in each block in the center
response since its role in response to abiotic stress of the plant stands at the height of the ears. Air
was previously demonstrated in cereals (Qin et al., temperature, relative humidity, and PPF were
2014). Other transcription factors like a bZIP and measured every min and 15-min averages were
MYB-related protein have to be considered too. recorded on a data logger. Thermal time was
They all could be involved in a transcriptional calculated by summing degree-days (base
reprogramming occurring after N and S supply and temperature 0°C) calculated as average daily air
modifying storage accumulation in einkorn grain temperature.
during filling phase. Until anthesis, each PVC column received 167
In present study, our results have evidenced that mL d-1 of a modified Hoagland’s nutrient solution
a non-concomitant N and S supply, which highly (Castle and Randall, 1987) prepared with
disturbed N-to-S ratio was responsible for major
demineralized water and containing 3 mM of N and
changes observed at transcript, metabolite and
0.1 mM of S (1 mM KH2PO4, 1 mM KNO3, 0.5
protein levels. Without forgetting genes slightly
impacted by the nutrition, data integration with mM Ca(NO3)2, 0.5 mM NH4NO3, 0.1 mM MgSO4,
network analysis allowed us to point several genes 1.9 mM MgCl2, 3.5 mM CaCl2, 4 mM KCl, 10 µM

130
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

1
H3BO3, 0.7 µM ZnCl2, 0.4 µM CuCl2, 4.5 µM H-NMR analyses
MnCl2, 0.22 µM MoO3, and 50 µM EDFS-Fe). At Polar metabolites were extracted from 50 mg of
anthesis, the nutrient solution was replaced by grain lyophilised powder with an ethanol–water
demineralized water to avoid any over series at 80°C (adapted from Moing et al. [2004]),
accumulation of N and S compounds in the plant Allwood et al. [2011]). The supernatants were
or in the potting substrate as during that period N combined, dried under vacuum and lyophilized.
and S plant demand is very limited. Then four Each lyophilized extract was solubilized in 500 µL
of 200 mM deuterated potassium phosphate buffer
combinations of N and S supply were applied from
solution (apparent pH 6.0), containing 3 mM
200°Cd to 700°Cd after anthesis: a low N and S ethylene diamine tetraacetic acid disodium salt
supply treatment corresponding to a nutrient (EDTA) to chelate paramagnetic cations and
solution with no N and S (treatment termed N-S-); improve spectrum resolution (especially in the
a high N and low S treatment with 6 mM N and no citrate region), titrated with KOD 1 M or DCL 0.1
S (N+S-); a low N and high S treatment with 0.5 M solutions by means of BTpH (Bruker BioSpin
mM N and 2 mM S (N-S+); and a high N and S GmbH, Rheinstetten, Germany) to apparent pH
6.00 ± 0.02, and lyophilized again. The lyophilized
treatment with 6 mM N and 2 mM S (N+S+).
titrated extracts were stored in darkness under
Nitrogen was not totally excluded from the N-S+ vacuum at room temperature, before 1H-NMR
nutrient solution as previous experiments have analysis was completed within one week.
shown that in the absence of N in the nutrient For 1H-NMR analysis, 500 µL of D2O with
solution S uptake and/or metabolism are inhibited sodium trimethylsilyl [2,2,3,3-d4] propionate
(P. Martre unpublished data). The nutrient (TSP, 0.01% mg/mL final concentration for
chemical shift calibration) were added to each
solutions were modified as described in Dai et al.
lyophilized titrated extract. The mixture was
(2015). Grains were collected every 100°Cd after centrifuged at 17,700 × g for 5 min at room
anthesis from 100°Cd to full maturity (950°Cd) for temperature. The supernatant was then transferred
metabolite analysis, then from 100 to 600°Cd for into a 5 mm NMR tube for acquisition.
transcriptome analysis. Quantitative 1H-NMR spectra were recorded at
500.162 MHz and 300 K on a Avance III
spectrometer (Bruker Biospin, Wissembourg,
Metabolite assays France) using a 5-mm ATMA broadband inverse
probe flushed with nitrogen gas and an electronic
Robotized analyses reference for quantification (Digital ERETIC,
Metabolites were extracted from about 20 mg of Bruker TopSpin 3.0). Sixty-four scans of 32k data
ground fresh weight with ethanol/water and points each were acquired with a 90 ° pulse angle,
analyzed as previously described (Biais et al., a 6000 Hz spectral width, a 2.73 s acquisition time,
and a 20 s recycle delay. The assignments of
2014; Bénard et al., 2015). Soluble sugars
metabolites in the NMR spectra were made by
including sucrose, glucose and fructose (Jelitto et comparing the proton chemical shifts with
al., 1992), malate (Nunes-Nesi et al., 2007) citrate literature (Likes et al., 2007) or database values
(Tompkins and Toffaletti, 1982) and the sum of (MeRy-B, HMDB; BMRB, MMCD), by
free amino acids were determined in the ethanolic comparison with spectra of authentic compounds
supernatant. Starch content was determined on the and by spiking the samples. For assignment
pellet as described in (Hendriks et al., 2003). purposes, 1H-1H COSY, 1H-13C HSQC spectra
Starch and amino acids contents were determined were acquired for a selected sample. For absolute
in 96-well polystyrene microplates (Sarstedt, quantification, four calibration curves (glucose and
Marnay, France) using a robotized Starlet platform fructose: 2.5 to 100 mM; glutamate and glutamine:
(Hamilton, Villebon sur Yvette, France). 2.5 to 30 mM) were prepared and analysed under
Individual amino acids were analyzed using the the same conditions. The glucose calibration was
AQC-tag method. Oxidized (GSSG) and reduced used for the quantification of all compounds, as a
function of the number of protons of selected
(GSH) glutathione were extracted and assayed
resonances except fructose, glutamate and
enzymatically with glutathione reductase and 5,50- glutamine that were quantified using their own
dithiobis-2-nitrobenzoic acid, as described by calibration curve. The concentration of each
Griffith (1980).

131
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

organic or amino acid was expressed as g of the RTA version 1.18.61 and CASAVA 1.8.4 was used
acid form per weight unit. The metabolite to generate FASTQ-files.
concentrations were calculated using AMIX
(version 3.9.14, Bruker) and Excel (Microsoft,
Redmond, WA, USA) softwares. Read alignment and expression analysis
Reads were mapped on the T. monococcum
genome available in MIPS genome database
RNA extraction and sequencing
(Mewes, 2002) reduced to scaffolds containing the
RNA were extracted from einkorn grain 32,047 annotated genes and the biggest scaffolds
according to method used by Pont et al. (2011). (337,103 scaffolds used). Mapping was performed
Three biological replicates were used. Grain using BWA-MEM (version 0.7.12-r1039,
samples (stages 100, 200, 300, 400, 500, 600°Cd [Link] Only reads with
after anthesis, treatments N-S-, N+S-, N-S+, N+S+ unique mapping position and a mapping quality
for each stage from 300 to 600°Cd after anthesis) score of at least 10 were considered for read
were ground in liquid nitrogen and 4.5 mL of counting. Paired-end reads that were mapped to the
buffer (10 mM Tris-HCl, pH 7.4, 1 mM EDTA, 0.1 same reference with about the expected insert size
M NaCl, 1% SDS) were added to approximately were counted as one read. Paired-end reads that
0.5g of powder. The supernatant was collected and were mapped to different references of with an
3 mL of phenol-chloroform-isoamyl alcohol unexpected insert size were counted as two reads.
mixture ([Link]) were added to eliminate starch If only one read of a pair was mapped, it was
and proteins. Supernatant was collected and counted as one read. Single-end reads were used
nucleic acids were precipitated by adding 0.1 vol straightforwardly. Only reads overlapping exon-
of 3 M AcNA pH 5.2 and 2 vol of 100% EtOH. features were counted. All reads mapping to
Pellet was dried under ambient condition for 1 h features with the same grouping-tag were summed.
and dissolved in RNase-free water. DNAse A Trimmed Mean of M-values (TMM)
treatment was then performed using a RNase-Free normalization was performed using the edgeR
DNase set (Qiagen) and then a RNeasy MinElute package (Robinson and Oshlack, 2010; Robinson
Cleanup kit (Qiagen). The amount and integrity of et al., 2010) to reconstruct transcripts and estimate
RNA integrity was checked by agarose gel transcript abundance in counts-per-million (CPM).
electrophoresis. Only samples with a RIN greater Genes with CPM value higher than 1 were
than 7 were used for the library construction. considered as expressed.
RNA Sequencing was done at Eurofins MWG
Operon ([Link], Ebersberg,
Germany). RNA strand-specific libraries were Statistical analysis
created using commercially available kits
according to the manufacturer’s instructions. In For RNA sequencing data, differential
brief, polyA-RNA was extracted from total RNA expression analysis was performed using the
using an oligodT-bead based method. After EdgeR package (Robinson and Oshlack, 2010;
fragmentation of the mRNA, first and dUTP based Robinson et al., 2010). Genes with a CPM superior
to 1 in at least three samples were considered.
second strand synthesis was carried out, followed
by end-repair, A-tailing, ligation of the indexed EdgeR estimates the genewise dispersions by
Illumina Adapter and digestion of the dUTP- conditional maximum likelihood, conditioning on
strand. Size selection was done using a bead based the total count for that gene. An empirical Bayes
method targeting an average insert size of 150-400 procedure was used to shrink the dispersions
base pairs. After PCR amplification the resulting towards a consensus value, effectively borrowing
fragments were cleaned up, pooled, quantified and information between genes. At each thermal time
used for cluster generation. For sequencing, pooled after anthesis from 300 to 600°Cd, differentially
libraries were loaded on the cBot (Illumina) and expressed genes among the four N and S treatments
cluster generation was performed using were identified using an exact test analogous to
manufacturer’s instructions. Paired-end Fisher's exact test, but adapted for over-dispersed
sequencing using 125 bp read length was data (excerpt from Robinson et al., 2010).
performed on a HiSeq2500 machine (HiSeq Differential expression was considered to be
Control Software 2.2.38) using HiSeq Flow Cell v4 significant at adjusted P value < 0.001 after false
and TruSeq SBS Kit v4. For processing of raw data discovery rate correction (Benjamini and
Hochberg, 1995).

132
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

Principal component analysis performed on scaled and a semantic was written which allowed
differential expressed genes was realized with to find rules between co-expressed/accumulated
FactomineR (Husson et al., 2015) package for R RNAs, metabolites and proteins (First semantic,
and hierarchical clustering on principal Method S1). Three quality measures were used to
components was performed with the ‘hcpc’ select the best rules (support, confidence and lift)
function and the ‘ward’ method to build gene and rules with a support >0.3, confidence = 1 and
expression profiles. Number of clusters was a lift >2 were selected. A second semantic was used
determined with the suggested partition, to find rules between RNAs/metabolites/proteins
corresponding to the one with the higher relative and comparisons of treatments, according to their
loss of inertia. R package ggplot2 (Wickham, quantity in the different samples (Second semantic,
2011) was used to draw plots. Method S1). For this second semantic, confidence
Metabolite data were analysed using the and lift thresholds were set at 0.7 and 1.5,
statistical software R v3.2.2 (R Core Team, 2015). respectively. Rules validated allowed to build
Two-way (developmental stage cross nutrition) network using CYTOSCAPE software version
ANOVA, followed by Tukey’s honestly 3.3.0 (Smoot et al., 2011).
significantly different (HSD) test were performed.
Statistical differences were judged at P value <
0.05. ACKNOWLEDGEMENTS
We thank Richard Blanc (UMR GDEC, INRA,
Functional annotation and gene ontology Blaise Pascal University) for help in the
enrichment analysis greenhouse and David Alvarez, Annie Faye,
Annotation of differential expressed genes in Marielle Merlino, Sibille Perrochon and Isabelle
response to nutrition was performed using Nadaud for help collecting grains (UMR GDEC).
TriAnnot (Leroy et al., 2012) pipeline. Gene We thank Florent Murat for his help to adapt the T.
ontology (GO) enrichment was then performed, as monococcum reference genome used in the
described in Choulet et al. (2014). Briefly, GO mapping step and Eurofins Genomics for
terms were searched for all expressed genes by performing RNA sequencing analyses. This work
BLAST (e-value < 1e-05), performed against the
PLAZA 2.5 database (Proost et al., 2009). Only the was supported by a Ph.D. grant from the French
five best hits with more than 50% coverage were Ministry for Higher Education and Research to TB
used for the analysis. Enrichment calculations were and funding from the French Government
then performed using the topGO R package (Alexa managed by the National Research Agency (ANR)
and Rahnenfuhrer, 2010). GO terms for expressed in the framework of Investments for the Future
genes were used as the reference comparison set (ANR-10-BTBR-03), France AgriMer and the
against differential expressed genes. Fisher’s exact
French Fund Supporting Plant Breeding (FSOV).
test were performed and statistical difference in
GO term enrichment was considered at P value <
0.05.
REFERENCES
Agier, M., Petit, J.M. and Suzuki, E. (2007) Unifying
Network analysis framework for rule semantics: Application to gene
Data set used for network analysis comprised the expression data. Fundam. Informaticae, 543–559.
386 genes and 14 metabolites which showed Akhtar, M., Jaiswal, A., Taj, G., Jaiswal, J.P.,
significant nutrition effect in present study, Qureshi, M.I. and Singh, N.K. (2012)
supplemented by 6 GSP and 203 proteins DREB1/CBF transcription factors: their structure,
previously obtained and variant in response to function and role in abiotic stress tolerance in
nutrition from the same biological material plants. J. Genet., 91, 385–395.
(Bonnot et al., in prep). Network were built using Alexa, A. and Rahnenfuhrer, J. (2010) topGO:
RulNet platform (([Link] Vincent et Enrichment analysis for Gene Ontology.
al., 2015). An algorithm describing rule semantics
Allwood, J.W., Vos, R.C.H. De, Moing, A., Deborde,
between attributes (Agier et al., 2007) is used to
C., Erban, A., Kopka, J., Goodacre, R. and
generate rules among -omic data, and make
possible to define different semantics. Data were Hall, R.D. (2011) Plant metabolomics and its

133
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

potential for systems biology research: rebalance wheat grain storage protein
Background concepts, technology, and accumulation under variable nitrogen and sulfur
methodology 1st ed., Elsevier Inc. supply. Plant J., 83, 326–343.
Araujo, W.L., Martins, A.O., Fernie, A.R. and Dhaka, V. and Khatkar, B.S. (2015) Effects of
Tohge, T. (2014) 2-Oxoglutarate: linking TCA gliadin/glutenin and HMW-GS/LMW-GS ratio
cycle function with amino acid, glucosinolate, on dough rheological properties and bread-
flavonoid, alkaloid, and gibberellin biosynthesis. making potential of wheat varieties. J. Food
Front. Plant Sci., 5, 552. Qual., 38, 71–82.
Bassel, G.W., Gaudinier, A., Brady, S.M., Hennig, Don, C., Mann, G., Bekes, F. and Hamer, R.J. (2006)
L., Rhee, S.Y. and Smet, I. De (2012) Systems HMW-GS affect the properties of glutenin
analysis of plant functional, transcriptional, particles in GMP and thus flour quality. J. Cereal
physical interaction, and metabolic networks. Sci., 44, 127–136.
Plant Cell, 24, 3859–3875. Eriksen, J. (2009) Soil sulfur cycling in temperate
Bénard, C., Bernillon, S., Biais, B., et al. (2015) agricultural systems. In Advances in Agronomy.
Metabolomic profiling in tomato reveals diel pp. 55–89.
compositional changes in fruit affected by source- Franke, I., Resch, A., Dassler, T., Maier, T. and
sink relationships. J. Exp. Bot., 66, 3391–3404. Bock, A. (2003) YfiK from Escherichia coli
Benjamini, Y. and Hochberg, Y. (1995) Controlling promotes export of O-acetylserine and cysteine. J.
the false discovery rate: a practical and powerful Bacteriol., 185, 1161–1166.
approach to multiple testing. J. R. Stat. Soc. Ser. Fukushima, A., Kusano, M., Redestig, H., Arita, M.
B, 57, 289–300. and Saito, K. (2009) Integrated omics approaches
Biais, B., Bénard, C., Beauvoit, B., et al. (2014) in plant systems biology. Curr. Opin. Chem. Biol.,
Remarkable reproducibility of enzyme activity 13, 532–538.
profiles in tomato fruits grown under contrasting Griffith, W. (1980) Determination of glutathione and
environments provides a roadmap for studies of glutathione disulfide glutathione reductase and 2-
fruit metabolism. Plant Physiol., 164, 1204–1221. vinylpyridine. Anal. Biochem., 212, 207–212.
Brady, S.M., Orlando, D.A., Lee, J.-Y., Wang, J.Y., Hawkesford, M.J. (2000) Plant responses to sulphur
Koch, J., Dinneny, J.R., Mace, D., Ohler, U. deficiency and the genetic manipulation of
and Benfey, P.N. (2007) A high-resolution root sulphate transporters to improve S-utilization
spatiotemporal map reveals dominant expression efficiency. J. Exp. Bot., 51, 131–138.
patterns. Science, 318, 801–806. Hawkesford, M.J. and DeKok, L.J. (2006) Managing
Branlard, G., Dardevet, M., Saccomano, R., sulphur metabolism in plants. Plant, Cell
Lagoutte, F. and Gourdon, J. (2001) Genetic Environ., 29, 382–395.
diversity of wheat storage proteins and bread Hendriks, J.H.M., Kolbe, A., Gibon, Y., Stitt, M. and
wheat quality. Euphytica, 119, 59–67. Geigenberger, P. (2003) ADP-glucose
Castle, S. and Randall, P. (1987) Effects of sulfur pyrophosphorylase is activated by
deficiency on the synthesis and accumulation of posttranslational redox-modification in response
proteins in the developing wheat seed. Aust. J. to light and to sugars in leaves of Arabidopsis and
Plant Physiol., 14, 503–516. other plant species. Plant Physiol., 133.
Chope, G., Wan, Y., Penson, S., Bhandari, D., Hirai, M.Y., Yano, M., Goodenowe, D.B., Kanaya,
Powers, S. and Shewry, P. (2014) Effects of S., Kimura, T., Awazuhara, M., Arita, M.,
genotype, season, and nitrogen nutrition on gene Fujiwara, T. and Saito, K. (2004) Integration of
expression and protein accumulation in wheat transcriptomics and metabolomics for
grain. J. Agric. Food Chem., 62, 4399–4407. understanding of global responses to nutritional
Choulet, F., Alberti, A., Theil, S., et al. (2014) stresses in Arabidopsis thaliana. Proc. Natl. Acad.
Structural and functional partitioning of bread Sci. U. S. A., 101, 10205–10210.
wheat chromosome 3B. Science, 345, 1249721– Hirel, B., LeGouis, J., Ney, B. and Gallais, A. (2007)
1249721. The challenge of improving nitrogen use
Dai, Z., Plessis, A., Vincent, J., et al. (2015) efficiency in crop plants: towards a more central
Transcriptional and metabolic alternations role for genetic variability and quantitative

134
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

genetics within integrated approaches. J. Exp. 1250092.

Bot., 58, 2369–2387. Mewes, H.W. (2002) MIPS: a database for genomes
Howarth, J.R., Parmar, S., Jones, J., et al. (2008) Co- and protein sequences. Nucleic Acids Res., 30,
ordinated expression of amino acid metabolism in 31–34.
response to N and S deficiency during wheat grain Moing, A., Maucourt, M., Renaud, C., et al. (2004)
filling. J. Exp. Bot., 59, 3675–3689. Quantitative metabolic profiling by 1-
Husson, F., Josse, J., Le, S., Mazet, J. and & Husson, dimensional 1H-NMR analyses: Application to
M.F. (2015) Package “FactoMineR”. plant genetics and functional genomics. Funct.
Jamal, A., Moon, Y.S. and Abdin, M.Z. (2010) Plant Biol., 31, 889–902.
Sulphur -a general overview and interaction with Moreno-Risueno, M.A., Busch, W. and Benfey, P.N.
nitrogen. Aust. J. Crop Sci., 4, 523–529. (2010) Omics meet networks—using systems
Jelitto, T., Sonnewald, U., Willmitzer, L., Hajirezeai, approaches to infer regulatory networks in plants.
M. and Stitt, M. (1992) Inorganic pyrophosphate Curr. Opin. Plant Biol., 13, 126–131.
content and metabolites in potato and tobacco Nikiforova, V., Freitag, J., Kempa, S., Adamik, M.,
plants expressing E. coli pyrophosphatase in their Hesse, H. and Hoefgen, R. (2003) Transcriptome
cytosol. Planta, 188, 238–244. analysis of sulfur depletion in Arabidopsis
Kang, J., Park, J., Choi, H., Burla, B., Kretzschmar, thaliana: interlacing of biosynthetic pathways
T., Lee, Y. and Martinoia, E. (2011) Plant ABC provides response specificity. Plant J., 33, 633–
transporters. Arab. B., 9, e0153. 50.
Kichey, T., Hirel, B., Heumez, E., Dubois, F. and Nunes-Nesi, A., Carrari, F., Gibon, Y., et al. (2007)
LeGouis, J. (2007) In winter wheat (Triticum Deficiency of mitochondrial fumarase activity in
aestivum L.), post-anthesis nitrogen uptake and tomato plants impairs photosynthesis via an effect
remobilisation to the grain correlates with on stomatal function. Plant J., 50, 1093–1106.
agronomic traits and nitrogen physiological Palta, J. and Fillery, I. (1995) N application increases
markers. F. Crop. Res., 102, 22–32. pre-anthesis contribution of dry matter to grain
Kidokoro, S., Watanabe, K., Ohori, T., et al. (2015) yield in wheat grown on a duplex soil. Aust. J.
Soybean DREB1/CBF-type transcription factors Agric. Res., 46, 519–531.
function in heat and drought as well as cold stress- Pechanek, U., Karger, A., Gröger, S., Charvat, B.,
responsive gene expression. Plant J., 81, 505– Schöggl, G. and Lelley, T. (1997) Effect of
518. nitrogen fertilization on quantity of flour protein
Kopriva, S. and Rennenberg, H. (2004) Control of components, dough properties, and breadmaking
sulphate assimilation and glutathione synthesis: quality of wheat. Cereal Chem., 74, 800–805.
interaction with N and C metabolism. J. Exp. Bot., Pont, C., Murat, F., Confolent, C., Balzergue, S. and
55, 1831–1842. Salse, J. (2011) RNA-seq in grain unveils fate of
Kruse, C., Jost, R., Lipschis, M., Kopp, B., neo- and paleopolyploidization events in bread
Hartmann, M. and Hell, R. (2007) Sulfur- wheat (Triticum aestivum L.). Genome Biol., 12,
enhanced defence: effects of sulfur metabolism, R18.
nitrogen supply, and pathogen lifestyle. Plant Proost, S., Bel, M. Van, Sterck, L., Billiau, K., Parys,
Biol., 9, 608–619. T. Van, Peer, Y. Van de and Vandepoele, K.
Leroy, P., Guilhot, N., Sakai, H., et al. (2012) (2009) PLAZA: a comparative genomics resource
TriAnnot: a versatile and high performance to study gene and genome evolution in plants.
pipeline for the automated annotation of plant Plant Cell, 21, 3718–3731.
genomes. Front. Plant Sci., 3, 1–14. Qin, D., Wang, F., Geng, X., Zhang, L., Yao, Y., Ni,
Likes, R., Madl, R.., Zeisel, S.. and Craig, S.A.. Z., Peng, H. and Sun, Q. (2014) Overexpression
(2007) The betaine and choline content of a whole of heat stress-responsive TaMBF1c, a wheat
wheat flour compared to other mill streams. J. (Triticum aestivum L.) Multiprotein Bridging
Cereal Sci., 46, 93–95. Factor, confers heat tolerance in both yeast and
Marcussen, T., Sandve, S.R., Heier, L., et al. (2014) rice. Plant Mol. Biol., 87, 31–45.
Ancient hybridizations among the ancestral R Core Team (2015) R: A language and environment
genomes of bread wheat. Science, 345, 1250092– for statistical computing. R foundation for

135
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

statistical computing, Vienna, Austria. 295–314.

Randall, P., Spencer, K. and Freney, J. (1981) Sulfur Tompkins, D. and Toffaletti, J. (1982) Enzymic
and nitrogen fertilizer effects on wheat. I. determination of citrate in serum and urine , with
Concentrations of sulfur and nitrogen and the use of the Worthington “ Ultrafree ” device. Clin.
nitrogen to sulfur ratio in grain, in relation to the Chem., 28, 2031–2033.
yield response. Aust. J. Agric. Res., 32, 203. Triboï, E., Martre, P. and Triboï-Blondel, A.-M.
Rausch, T. and Wachter, A. (2005) Sulfur (2003) Environmentally-induced changes in
metabolism: a versatile platform for launching protein composition in developing grains of wheat
defence operations. Trends Plant Sci., 10, 503– are related to changes in total protein content. J.
509. Exp. Bot., 54, 1731–1742.
Robinson, M.D., McCarthy, D.J. and Smyth, G.K. Urano, K., Kurihara, Y., Seki, M. and Shinozaki, K.
(2010) edgeR: a bioconductor package for (2010) “Omics” analyses of regulatory networks
differential expression analysis of digital gene in plant abiotic stress responses. Curr. Opin. Plant
expression data. Bioinformatics, 26, 139–140. Biol., 13, 132–138.
Robinson, M.D. and Oshlack, A. (2010) A scaling Västermark, Å., Almén, M., Simmen, M.W.,
normalization method for differential expression Fredriksson, R. and Schiöth, H.B. (2011)
analysis of RNA-seq data. Genome Biol., 11, R25. Functional specialization in nucleotide sugar
Sanchez, R. and Zhou, M.-M. (2011) The PHD finger: transporters occurred through differentiation of
a versatile epigenome reader. Trends Biochem. the gene cluster EamA (DUF6) before the
Sci., 36, 364–372. radiation of Viridiplantae. BMC Evol. Biol., 11,
Serin, E.A.R., Nijveen, H., Hilhorst, H.W.M. and 123.
Ligterink, W. (2016) Learning from co- Vincent, J., Martre, P., Gouriou, B., Ravel, C., Dai,
expression networks: possibilities and challenges. Z., Petit, J.-M. and Pailloux, M. (2015) RulNet:
Front. Plant Sci., 7, 444. A web-oriented platform for regulatory network
Shewry, P.R. and Halford, N.G. (2002) Cereal seed inference, application to wheat –omics data. PLoS
storage proteins: structures, properties and role in One, 10, e0127127.
grain utilization. J. Exp. Bot., 53, 947–958. Wei, H., Yordanov, Y.S., Georgieva, T., Li, X. and
Shewry, P.R., Tatham, A.S. and Halford, N.G. (2001) Busov, V. (2013) Nitrogen deprivation promotes
Nutritional control of storage protein synthesis in Populus root growth through global transcriptome
developing grain of wheat and barley. Plant reprogramming and activation of hierarchical
Growth Regul., 34, 105–111. genetic networks. New Phytol., 200, 483–497.
Shimizu, M.M. and Mazzafera, P. (2000) A role for Wickham, H. (2011) ggplot2. Wiley Interdiscip. Rev.
trigonelline during imbibition and germination of Comput. Stat., 3, 180–185.
coffee seeds. Plant Biol., 2, 605–611. Wieser, H. (2007) Chemistry of gluten proteins. Food
Smoot, M.E., Ono, K., Ruscheinski, J., Wang, P.-L. Microbiol., 24, 115–119.
and Ideker, T. (2011) Cytoscape 2.8: new Wieser, H., Gutser, R. and Tucher, S. von (2004)
features for data integration and network Influence of sulphur fertilisation on quantities and
visualization. Bioinformatics, 27, 431–432. proportions of gluten protein types in wheat flour.
Sosa-Valencia, G., Covarrubias, A.A. and Reyes, J. Cereal Sci., 40, 239–244.
J.L. (2013) Stress signaling in plants: genomics Wrigley, C., Cros, D., Archer, M., Downie, P. and
and proteomics perspective, Volume 1 M. Sarwat, Roxburgh, C. (1980) The sulfur content of wheat
A. Ahmad, and M. Abdin, eds., New York, NY: endosperm proteins and its relevance to grain
Springer New York. quality. Aust. J. Plant Physiol., 7, 755–766.
Tea, I., Genter, T., Violleau, F. and Kleiber, D. (2005) Zhao, F., Hawkesford, M. and McGrath, S. (1999)
Changes in the glutathione thiol-disulfide status in Sulphur assimilation and effects on yield and
wheat grain by foliar sulphur fertilization : quality of wheat. J. Cereal Sci., 30, 1–17.
consequences for the rheological properties of Zörb, C., Grover, C., Steinfurth, D. and Hermann
dough. , 41, 305–315. Mühling, K. (2010) Quantitative proteome
Tisdale, S.L., Reneau, R.B. and Platou, J.S. (1986) analysis of wheat gluten as influenced by N and S
Atlas of sulfur deficiencies. Sulfur Agric., 27, nutrition. Plant Soil, 327, 225–234.

136
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

Supplementary information

Figure S1. Mass per grain in the four N and S treatments of the 14 metabolites significantly impacted by nutrition
during einkorn grain development. The four treatments were N-S- (0mM N and 0mM S), N+S- (6mM N and 0mM
S), N-S+ (0.5mM N and 2mM S), N+S+ (6mM N and 2mM S), applied from 200 to 700°Cd after anthesis. Data
are means ± 1 SE for n = 4 biological replicates.

137
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

Figure S2. Linkage number in network analysis. (a) Number of edges associated to each attribute present in
modules 1 and 5 in co-expression network (Figure 4). (b) Number of edges that were associated to each treatment
comparison in the nutrition network (Figure 5).

Tables S1 and S2 are not necessary to understand the manuscript thesis since they correspond to
quantitative data for metabolites and to gene expression data. Thus they have not been included in this
manuscript.

138
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

Table S3. Summary of the 609 variables significantly impacted by N and S supply. Column ‘Code’ indicates
codes used in figures 4, 5 and 6 to qualify transcripts (codes that begin with ‘G’ letter), metabolites and proteins
(albumins-globulins and nuclear proteins are distinguished by codes that begin with ‘Alg’ and ‘N’,
respectively). For transcriptomic data, Gene Ontology annotation of genes uncharacterized is given in
parenthesis in column ‘Functional annotation’. Sometimes, genes had no functional annotation. If the variable
was part of the co-expression network, its module and number of associations are indicated in columns
‘Module’ and ‘Degree’, respectively. In columns related to the nutrition network, signs ‘+’ or ‘-’ indicate that
the protein was present in network, ‘+’ or ‘-’ indicating a higher or a lower amount in the first treatment,
respectively. In the last column, cluster number is indicated for transcriptomic data and refers to the clustering
analysis presented in Figure 2.
Co-expression
network
Nutrition network Cluster
Code Functional annotation
Module Degree
N+S- N-S+ N+S+ N-S+ N+S+ N+S+ (gene)
N-S- N-S- N-S- N+S- N+S- N-S+
G1 Major intrinsic protein + - - 2
G2 uncharacterized (GO: nucleic acid binding) 5 2 1
G3 Cysteine proteinase inhibitor 3
G4 5 22 1
G5 Xyloglucan endo-transglycosylase - XET - C-terminus + + 2
G6 1 -3-beta-glucan synthase component 1
G7 uncharacterized (GO: phospholipid binding) 1
G8 2OG-Fe- II - oxygenase superfamily + - 1
G9 Wound-induced protein 1
G10 Uncharacterised conserved protein - DUF2359 1
G11 uncharacterized (GO: peptidase inhibitor activity) - 2
G12 + 3
G13 Cys-rich Gliadin N-terminal 1
G14 uncharacterized (GO: hydrolase activity) 3
G15 uncharacterized (GO: zinc ion binding) + 1
G16 NAD binding domain of 6-phosphogluconate dehydrogenase 1
G17 1
G18 - + + 3
G19 Cys-rich Gliadin N-terminal 2 2 + + + 1
G20 uncharacterized (GO: transcription) 5 14 1
G21 Low-molecular-weight glutenin subunit 3 2 + + 3
G22 Serine hydroxymethyltransferase 4 5 + - - 2
G23 uncharacterized (GO: metal ion binding) 1
G24 3
G25 uncharacterized (GO: cell periphery) 1
G26 uncharacterized (GO: transferase activity) 1
G27 Phosphoethanolamine N-methyltransferase 1 + 1
G28 + - 2
G29 40S ribosomal protein S20 1
G30 Pollen proteins Ole e I like 1
G31 uncharacterized (GO: zinc ion binding) 1
G32 uncharacterized (GO: purine ribonucleoside binding) 1
G33 uncharacterized (GO: protein binding) 1
G34 uncharacterized (GO: primary metabolic process) 5 35 1
G35 Protein tyrosine kinase 3
G36 Bowman-Birk serine protease inhibitor family 3
G37 Maintenance of mitochondrial structure and function 3
G38 Gamma-gliadin 3 1 - + 1
G39 uncharacterized (GO: mRNA processing) + - - 1
G40 uncharacterized (GO: protein binding) + - - 2
G41 uncharacterized (GO: transmembrane transport) 2
G42 Asparagine synthetase (Glutamine-hydrolyzing) + 1
G43 Endosperm transfer cell specific PR60 1
G44 Basic endochitinase A 1 30 - 3
G45 2
G46 Allinase - 2
G47 GHMP kinases C terminal 2
G48 Rhodanese-like domain 1
G49 Adenylyl-sulfate kinase + - - 2
G50 Chitinase (Fragment) 1 34 - 3
G51 EamA-like transporter family 4 14 + - - 2
G52 unknown (GO: root morphogenesis) - 3
G53 Glycosyl hydrolase family 1 - 1
G54 unknown + - - 2
G55 unknown + 1
G56 uncharacterized (GO: oxidation reduction process) 1 28 + 2

139
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S
G57 1
G58 uncharacterized (GO: oxidation reduction process) + - - 2
G59 Choline -ethanolamine kinase 1
G60 unknown (GO: zinc ion binding) + - - 2
G61 Sulfate transporter N-terminal domain with GLY motif + - - 2
G62 uncharacterized (GO: metal ion binding) 1
G63 uncharacterized (GO: protein binding) 3
G64 1
G65 DNA polymerase 1 + - + 1
G66 - 1
G67 uncharacterized (GO: defense response) 3
G68 Cytochrome P450 1 27 3
G69 uncharacterized (GO: integral component of membrane ) 1 - 1
G70 uncharacterized (GO: cell wall macromolecule) - 1
G71 uncharacterized (GO: protein binding) 1
G72 unknown (GO: defense response bacterium) 1 32 - 3
G73 uncharacterized (GO: protein binding) - 1
G74 uncharacterized (GO: 4-alpha-glucanotransferase activity) 2
G75 uncharacterized (GO: binding) 1
G76 Beta purothionin - + + 1
G77 1 23 3
G78 uncharacterized (GO: RNA glycosylase activity) + 3
G79 5 26 1
G80 1
G81 Cupin 1 53 3
G82 Late embryogenesis abundant protein + 1
G83 uncharacterized (GO: protein binding) 1
G84 Trehalose-phosphatase + - 2
G85 Fatty acid hydroxylase superfamily 1 22 3
G86 Catalase isozyme 1 - + + 1
G87 uncharacterized (GO: zinc ion binding) - 3
G88 uncharacterized (GO: nucleic acid binding) 1
G89 Multiprotein bridging factor 1 1
G90 Glutathione S-transferase - N-terminal domain 4 1 + - - - 2
G91 uncharacterized (GO: integral component of membrane) 2
G92 1
G93 HVA22-like protein 1 29 + 3
G94 1
G95 Possible lysine decarboxylase 1
G96 uncharacterized (GO: transferase activity) 1
G97 uncharacterized (GO: zinc ion binding) 3
G98 Eukaryotic cytochrome b561 - 2
G99 Serine carboxypeptidase + - 2
G100 NmrA-like family 4 3 + - - 2
G101 1
G102 Class II chitinase 1 34 3
G103 Bowman-Birk type proteinase inhibitor I-2B (Fragment) + 3
G104 unknown 3
G105 uncharacterized (GO: oxidation reduction process) 1 21 3
G106 uncharacterized (GO: nutrient reservoir actitivy) 2
G107 unknown 3
G108 Tubulin C-terminal domain 1
G109 Alpha amylase - catalytic domain 1
G110 Defensin + 1
G111 Stress responsive A -B Barrel Domain 1 24 3
G112 + - + 1
G113 1
G114 Protein BREAST CANCER SUSCEPTIBILITY 1-like protein + 2
G115 Cysteine synthase + - - 2
G116 Chitin recognition protein 1 52 3
G117 Chitinase class I 1 30 3
G118 uncharacterized (GO: killing of cells of other organism) 3
G119 Endoglucanase 5 43 1
G120 TPR repeat 4 2 + - - 2
G121 Chitin recognition protein 1 25 - 3
G122 1
G123 unknown 4 1 + - - 2
G124 uncharacterized (GO: killing of cells of other organism) + + 3
G125 unknown_function 3
G126 Non-specific lipid-transfer protein - 3
G127 Rhodanese-like domain + - - 2
G128 2
G129 uncharacterized (GO: heme binding) 1
G130 Ureide permease 1 24 3
G131 Hsp20 -alpha crystallin family - 1
G132 3

140
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S
G133 - + 1
G134 unknown 5 23 1
G135 unknown 2
G136 uncharacterized (GO: zinc ion binding) - 1
G137 Haloacid dehalogenase-like hydrolase 3
G138 E2F -DP family winged-helix DNA-binding domain 1
G139 Pectinacetylesterase 1 6 3
G140 unknown_function + 2
G141 Glycosyl hydrolase family 1 1 30 - 3
G142 + - - 2
G143 60S acidic ribosomal protein P0 2
G144 Glycosyl hydrolase family 65 - N-terminal domain - + + 1
G145 Sucrose synthase 2 5 34 1
G146 unknown (GO: regulation of amino acid export) + + 2
G147 Glycosyl hydrolase family 3 C-terminal domain 3
G148 Tubulin beta-1 chain 2 2 - + + + 1
G149 Protein of unknown function - DUF506 4 3 + - - 2
G150 unknown 3
unknown (GO: regulation of transcription from RNA polymerase
G151
II promoter) 1
G152 uncharacterized (GO: oxidation reduction process) + 1
G153 uncharacterized (GO: viral process) 1
G154 uncharacterized (GO: primary metabolic process) 1 28 2
G155 Citrate synthase + + 3
G156 unknown 1
G157 3
G158 unknown 3
G159 unknown - 1
G160 Late embryogenesis abundant protein 1 2 - 1
G161 unknown_function + - - 2
G162 Aminotransferase class I and II 1
G163 uncharacterized (GO: protein binding) - - + 1
G164 Probable allantoinase + 1
G165 uncharacterized (GO: transcription) 3
G166 + + 3
G167 Macrophage migration inhibitory factor - MIF + + 1
G168 uncharacterized (GO: cellular catabolic process) 1 22 - 3
G169 Glycosyltransferase like family 2 5 1 1
G170 uncharacterized (GO: purine ribonucleoside binding) + 2
G171 uncharacterized (GO: integral component of membrane) 3
G172 5 38 1
G173 unknown (GO: lipid binding) 5 38 1
G174 RuBisCO large subunit (Fragment) 1
G175 unknown - - 2
G176 + - - 2
G177 uncharacterized (GO: zinc ion binding) - 1
G178 1
G179 1
G180 - 1
G181 uncharacterized (GO: thiamine pyrophosphate binding) 2
G182 5 35 - 1
G183 MA3 domain + - + 2
G184 Serpin - serine protease inhibitor + - - 2
G185 uncharacterized (GO: oxidation reduction process) 4 2 + 2
G186 Cys-rich Gliadin N-terminal 3 3 - + + + 1
G187 1
G188 uncharacterized (GO: phospholipid binding) 3
G189 Trypsin alpha-amylase inhibitor CMX1 CMX3 + + 1
G190 unknown 1 26 - 3
G191 + 1
G192 unknown 3
G193 - - 2
G194 uncharacterized (GO: purine ribonucleoside binding) - 3
G195 - - 1
G196 uncharacterized (GO: heat shock protein binding) 1
G197 unknown (GO: metal ion binding) 2
G198 + + 1
G199 + - + 1
G200 1
G201 Aminotransferase class-III 4 4 + - - 1
G202 uncharacterized (GO: nucleic acid binding) 1 28 - 3
G203 unknown - - 1
G204 unknown 5 12 1
G205 1
G206 uncharacterized (GO: transcription) 1
G207 uncharacterized (GO: zinc ion binding) 1

141
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S
G208 tRNA synthetase class II core domain - G - H - P - S and T + 2
G209 Probable inositol oxygenase 2
G210 unknown + - 2
G211 DNA polymerase 3
uncharacterized (GO: regulation of transcription from RNA
G212
polymerase II promoter) 1
G213 - + 1
G214 uncharacterized (GO: transcription) 1
G215 Flavin containing amine oxidoreductase 3
G216 1
G217 uncharacterized (GO: binding) 1
G218 Phosphoadenosine phosphosulfate reductase family 4 3 + - - 2
G219 1
G220 OB-fold nucleic acid binding domain 5 26 1
G221 - + 3
G222 1
G223 uncharacterized (GO: heat acclimatation) 1 1 3
G224 uncharacterized (GO: sucrose phosphate synthase activity) 1 27 3
G225 + - 3
G226 ATP-sulfurylase 4 2 + - - 2
G227 Chitinase 2 1 1 3
G228 unknown 3
G229 Gamma-thionin family + 1
G230 uncharacterized (GO: protein binding) 1
G231 unknown_function 3
G232 1
G233 Formate dehydrogenase mitochondrial 2 1 + - - 1
G234 Glycosyl hydrolases family 31 + - - 2
G235 DNA replication licensing factor MCM4 5 7 1
G236 Serine threonine-protein kinase AtPK19 + 1
G237 Peptidase family M28 1 1 - + 3
G238 uncharacterized (GO: proteolysis) 5 13 + - 1
G239 RWD domain - 2
G240 ENTH domain + + 1
G241 Taurine catabolism dioxygenase TauD - TfdA family 4 2 + - - 2
G242 uncharacterized (GO: primary metabolic process) 1
G243 - 2
G244 uncharacterized (GO: zinc ion binding) 1
G245 Malonyl-CoA decarboxylase - MCD - 1
G246 uncharacterized (GO: purine ribonucleoside binding) 1
G247 Sulfate transporter family + - - 2
G248 uncharacterized (GO: cellular catabolic process) - 3
G249 Cupin 1
G250 uncharacterized (GO: primary metabolic process) 1
G251 Diacylglycerol kinase 5 17 1
G252 5 25 1
G253 uncharacterized (GO: small GTPase mediated signal transduction) 1
G254 DNA replication licensing factor MCM6 5 22 + + - 1
G255 uncharacterized (GO: ATPase activity) 2
G256 unknown (GO : ATP binding) - - 3
G257 Xylanase inhibitor N-terminal + - - 1
G258 EF-hand domain pair - 1
G259 Sulfite exporter TauE -SafE 3
G260 Xylanase inhibitor N-terminal 5 20 1
G261 1
G262 Gamma-thionin family 1
G263 Myb-like DNA-binding domain + 3
G264 3
G265 - 1
G266 uncharacterized (GO: polysaccharide biosynthetic process) 1
G267 uncharacterized (GO: ATP binding) - 2
G268 1
G269 unknown 4 5 + - - 2
G270 Transmembrane amino acid transporter protein + - - 2
G271 unknown (GO: lipid binding) 5 17 1
G272 Fructan 6-exohydrolase 1
G273 3
G274 uncharacterized (GO: oxidation reduction process) + - - 2
G275 Putative 12-oxophytodienoate reductase 11 1
G276 - 1
G277 Betaine aldehyde dehydrogenase 2 + - 2
G278 Putative xylanase inhibitor 5 2 + 1
G279 uncharacterized (GO: alpha-amylase activity) 3
G280 + 2
G281 Eukaryotic translation initiation factor 5A-4 1
G282 Rare lipoprotein A - RlpA --like double-psi beta-barrel + 1

142
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S
G283 uncharacterized (GO: binding) 3
G284 Kelch motif 2
G285 unknown 3
G286 Serpin - serine protease inhibitor 2 1 + - - 1
G287 Alcohol dehydrogenase 1 2 - + + 3
G288 Cyclin-dependent kinase inhibitor 4 3
G289 Auxin-repressed protein 1 51 3
G290 Potassium transporter 3
G291 PHD-finger + - - 1
G292 Universal stress protein family + 2
G293 Transmembrane amino acid transporter protein 1
G294 uncharacterized (GO: hydrolase activity) 1 12 + + 3
G295 ABC transporter G family member 39 + + 3
G296 Homocysteine S-methyltransferase 4 4 + - - 2
G297 Myb-like DNA-binding domain 3
G298 uncharacterized (GO: lipid binding) 1
G299 ACT domain + - - 2
G300 uncharacterized (GO: protein binding) - 1
G301 unknown (GO: ATP binding) 1
G302 Pyridine nucleotide-disulphide oxidoreductase 2
G303 Dihydroorotate dehydrogenase (quinone) mitochondrial 3
G304 unknown + 2
G305 Glycerophosphoryl diester phosphodiesterase family 2
G306 uncharacterized (GO: protein binding) - 1
G307 uncharacterized (GO: oxidation reduction process) 1 27 3
G308 5 24 + 1
G309 uncharacterized (killing of cells of other organism) + 3
G310 unknown (GO: sulfate adenylyltransferase activity) 4 5 + - - 2
G311 Asparagine synthetase glutamine-hydrolyzing 3
G312 1
G313 uncharacterized (glucan biosynthetic process) 5 31 1
G314 uncharacterized (GO: zinc ion binding) + 3
G315 5 25 1
G316 Puroindoline-A 3 2 - + + + 3
G317 1
G318 HMGL-like 1
G319 uncharacterized (GO: zinc ion binding) 3
G320 1 24 3
G321 Raffinose synthase or seed imbibition protein Sip1 + 2
G322 uncharacterized (GO: nutrient reservoir activity) 2 3 + + - 1
G323 Avenin-like a2 2 3 + + 1
G324 Senescence regulator 2
G325 uncharacterized (GO: calcium ion binding) 3
G326 Puroindoline-B 1
G327 Glutathione S-transferase - C-terminal domain 4 5 + - - 2
G328 Non-specific serine threonine protein kinase + - - 2
G329 uncharacterized (GO: chromatin modification) 3
G330 Beta-1 3-glucanase 1 21 3
G331 3
G332 uncharacterized (GO: hydrolase activity) 3
G333 1
G334 WD domain - G-beta repeat + 1
G335 uncharacterized (GO: signal peptide processing) 2
G336 uncharacterized (GO: structural molecule activity) 1
G337 uncharacterized (GO: purine NTP-dependent helicase activity) - 3
G338 unknown (GO: ribonucleoside binding) 2 2 + - - 1
G339 unknown (GO: protein binding) 1
G340 Gamma-thionin family 5 3 1
G341 Plant PDR ABC transporter associated 1 24 2
G342 unknown_function + - - 2
G343 uncharacterized (GO: protein binding) 2
G344 1
G345 1
G346 unknown (GO: serine-type endopeptidase inhibitor activity) 3 3 - + + + 1
G347 uncharacterized (GO: protein binding) 1
G348 1
G349 + 1
G350 Nuclear transcription factor Y subunit C-2 - - 2
G351 1
G352 3
G353 Tetrapyrrole - Corrin -Porphyrin - Methylases + - - 2
G354 uncharacterized (GO: cation transport) 3
G355 Gamma-thionin family 1
G356 Glutathione S-transferase - C-terminal domain 3
G357 uncharacterized (GO: killing of cells of other organism) 3
G358 Purple acid phosphatase 1 35 3

143
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S
G359 uncharacterized (GO: mRNA splicing) 3
G360 unknown_function 2
G361 unknown (GO: transcription factor activity) 2
G362 2
G363 uncharacterized (GO: nucleoside metabolic process) 5 4 - 1
G364 uncharacterized (GO: heme binding) 3
G365 RNA dependent RNA polymerase 3
G366 uncharacterized (GO: oxidation reduction process) + 2
G367 uncharacterized (GO: oxidation reduction process) 1 2 + 3
G368 Beta-amylase 1 1 - + + 3
G369 Putative NADP-dependent oxidoreductase 4 4 + - - 2
G370 unknown + + 3
G371 uncharacterized (GO: nucleic acid binding) - 2
G372 3
G373 unknown_function 1
G374 - 1
G375 unknown_function - - 2
G376 4 4 - - 2
G377 5 23 1
G378 Peptide N-acetyl-beta-D-glucosaminyl asparaginase amidase A 2
G379 Tetratricopeptide repeat 4 4 + - - 2
G380 Defensin 3 1 + + 1
G381 uncharacterized (GO: defense response) 1 39 3
G382 unknown 2
G383 Probable lipid transfer 1
G384 uncharacterized (GO: oxidation reduction process) 1
G385 uncharacterized (GO: multicellular organism development) 1 29 3
G386 BNR repeat-like domain - 2
Asp Aspartate 2 +
Ile Isoleucine - - -
Glu Glutamate 5 12
Mal Malic acid +
Met Methionine -
Phe Phenylalanine - + -
Sacch saccharose 5 28
Trp Tryptophane - - +
Gln Glutamine +
Trig Trigoneline
Asn Asparagine
Lys Lysine
Fum Fumarate
GSH Glutathion
alpha alpha-gliadin
gamma gamma-gliadin 1 2
w1.2 omega-1,2-gliadin 1 25 + - +
w5 omega-5-gliadin 1 24 +
HMW HMW-GS 1 21 + - -
LMW LMW-GS 1 21 +
Alg1 Limit dextrinase type starch debranching enzyme 1 1 - + +
Alg2 Alcohol dehydrogenase ADH1A 1 11 - + +
Alg3 Puroindoline A - - +
Alg4 Triticin +
Alg5 GTPase SAR1
Alg6 Subtilisin protease (Fragment)
Alg7 Zeta-carotene desaturase + - -
Alg8 Gamma-gliadin +
Alg9 Gamma gliadin (Fragment) 1 1
Alg10 Fructose-bisphosphate aldolase 5 9
Alg11 Trypsin inhibitor - +
Alg12 Aminotransferase 1 2
Alg13 Purple acid phosphatase 1 6 +
Alg14 Omega-gliadin (Fragment) 1 52 -
Alg15 Protein disulfide isomerase family protein 4-1
Alg16 RNA recognition domain containing protein,expressed
Alg17 Protein phosphatase 2A structural subunit
Alg18 High molecular weight glutenin 1Ax2.1 1 52
Alg19 Starch branching enzyme IIa -
Alg20 Superoxide dismutase [Cu-Zn] (Fragment)
Alg21 Translocon-associated protein subunit beta +
Alg22 Basic endochitinase C 1 52 -
Alg23 Importin subunit alpha +
Alg24 Vicilin-like antimicrobial peptides 2-2 1 61
Alg25 Xylulose kinase - - -
Isocitrate dehydrogenase [NAD] catalytic subunit 5,
Alg26
mitochondrial
Alg27 GrpE protein homolog

144
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S
Alg28 Putative aconitate hydratase, cytoplasmic +
Alg29 Glutaredoxin-C6 1 2
Alg30 Aminopeptidase N 5 6
Alg31 Acyl-[acyl-carrier-protein] desaturase 2, chloroplastic
Alg32 zinc finger protein
Alg33 Intracellular protease 1 +
Alg34 Aspartic proteinase oryzasin-1 1 51
Alg35 Chitinase 1 1 53
Alg36 Ankyrin repeat domain-containing protein 2 1 58
Alg37 DNA damage-inducible protein 1
Alg38 transmembrane 214-B -
Alg39 amino acid selective channel protein
Alg40 Putative NADP-dependent oxidoreductase P1 1 3 +
Alg41 2-oxoglutarate dehydrogenase, mitochondrial 1 2
Alg42 Peroxisomal multifunctional enzyme type 2
Alg43 non-specific lipid transfer GPI-anchored 2-like
Alg44 Glutaredoxin-C8 5 2 +
Alg45 isoflavone reductase homolog irl-like 4 2 + - -
Alg46 Cytochrome b5 1 6
Alg47 Serpin-ZX 1 9 -
Pyrophosphate--fructose 6-phosphate 1-phosphotransferase
Alg48
subunit alpha - -
Alg49 Monothiol glutaredoxin-S11 +
Alg50 Endogenous alpha-amylase/subtilisin inhibitor 1 2
Alg51 Programmed cell death protein 4 + +
Alg52 Putative glutathione S-transferase GSTF1 +
Alg53 2'-deoxymugineic-acid 2'-dioxygenase
Alg54 Peptide methionine sulfoxide reductase A2-1
Alg55 glutathionyl-hydroquinone reductase -like - -
Alg56 Endoglucanase 11 +
Alg57 proton pump-interactor 1-like
Alg58 L-ascorbate peroxidase 1, cytosolic 5 9
Alg59 Programmed cell death protein 4 1 4 -
Alg60 Gamma-hordein-3 + + +
Bifunctional aspartokinase/homoserine dehydrogenase 2,
Alg61
chloroplastic
Alg62 Pyruvate, phosphate dikinase 1, chloroplastic -
Alg63 Coatomer subunit gamma -
Alg64 Adenosine kinase 2 5 30
Alg65 Sucrose synthase 2
Alg66 Trypsin/alpha-amylase inhibitor CMX1/CMX3 - + +
Alg67 UDP-glucuronic acid decarboxylase 1 5 6
Alg68 ADP,ATP carrier protein, mitochondrial
Alg69 Uncharacterized protein +
Alg70 Cell division protein ftsZ-like protein 2-1, chloroplastic +
Alg71 Clathrin heavy chain
Alg72 Chitinase 2 1 61
Lipoamide acyltransferase component of branched-chain alpha-
Alg73
keto acid dehydrogenase complex, mitochondrial
Alg74 hypothetical protein TRIUR3_12434 5 30
Alg75 Actin-7 1 7 - -
Alg76 DEAD-box ATP-dependent RNA helicase 37 -
Alg77 Inositol-3-phosphate synthase
Alg78 Avenin-3 - + +
Putative bifunctional methylthioribulose-1-phosphate
Alg79
dehydratase/enolase-phosphatase E1 - + +
Alg80 Protein argonaute 4A
Alg81 caleosin 2 1 7
Alg82 Beta-amylase 1 1 - + +
Alg83 Starch branching enzyme 1 1 48
Glucose-1-phosphate adenylyltransferase large subunit,
Alg84
chloroplastic/amyloplastic
Alg85 Phosphoglycerate kinase, chloroplastic 5 37
Alg86 Cysteine synthase + - -
Alg87 Non-specific lipid-transfer protein 2P 1 6
Alg88 Disproportionating enzyme
Alg89 Starch branching enzyme IIb - +
Alg90 Delta-1-pyrroline-5-carboxylate dehydrogenase +
Alg91 Ribulose bisphosphate carboxylase large chain 5 22
Alg92 Expansin EXPB3 +
Alg93 Glutamine synthetase
Alg94 uncharacterized (GO: storage protein) - + +
Alg95 Glutathione transferase 1 30 -
Alg96 Beta-D-glucan exohydrolase - -
Alg97 Thaumatin-like protein 1 1 - +
Alg98 Low-molecular-weight glutenin subunit group 6 type IV 1 2 - + +

145
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S
Alg99 Starch branching enzyme 1 1 48
Alg100 Glucose-1-phosphate adenylyltransferase
Alg101 Tubulin beta-3 chain
Alg102 glutathione s-transferase 3 5 1
Alg103 phospho-2-dehydro-3-deoxyheptonate aldolase chloroplastic-like 1 1
Alg104 probable aldo-keto reductase 1-like +
Alg105 Uncharacterized protein
Alg106 importin subunit beta-1-like 5 8
Alg107 12s seed storage globulin 1-like
Alg108 stromal 70 kda heat shock-related chloroplastic
Alg109 universal stress protein a-like protein
Alg110 pyrophosphate-energized vacuolar membrane proton pump 1-like
Alg111 Importin subunit alpha + -
Alg112 clathrin light chain 2-like
Alg113 glutamate--glyoxylate aminotransferase 2
Alg114 ECERIFERUM 26-like
Alg115 methylthioribose kinase - -
Alg116 Xylose isomerase +
Alg117 chaperone protein chloroplastic-like - +
Alg118 Uncharacterized protein +
Alg119 udp-glucuronic acid decarboxylase 2-like 5 4
Alg120 Uncharacterized protein (Fragment) + - -
Alg121 carbonic anhydrase 1 2 - + +
Alg122 flower-specific gamma-thionin precursor + -
Alg123 uridine 5 -monophosphate synthase-like
Alg124 cationic peroxidase SPC4-like 1 56 -
Alg125 pectin acetylesterase 5-like 1 26
Alg126 pyruvate decarboxylase 1 1 + +
Alg127 glutamate-rich wd repeat-containing protein 1
Alg128 UDP-glucose 6-dehydrogenase -
Alg129 gibberellin 20 oxidase 2 4 5 - - -
Alg130 rna recognition motif containing family protein 1 1
Alg131 1,2-dihydroxy-3-keto-5-methylthiopentene dioxygenase 4 2
Alg132 transport SEC23-like +
Alg133 vicilin-like antimicrobial peptides 2-2 1 58 +
Alg134 cop9 signalosome complex subunit 4
Alg135 Alcohol dehydrogenase-like 2 1 55
Alg136 cop9 signalosome complex subunit 1
Alg137 Uncharacterized protein
Alg138 cbs domain-containing protein
Alg139 probable n-succinyldiaminopimelate aminotransferase - -
Alg140 probable sarcosine oxidase 1 54
Alg141 probable sarcosine oxidase 1 44
Alg142 ripening-related 1 1 35
Alg143 prohibitin- mitochondrial-like
Alg144 alpha,alpha-trehalose-phosphate synthase [UDP-forming] 6-like +
Alg145 Oleosin 1 61
Alg146 rRNA N-glycosidase 1 6 +
Alg147 acetolactate synthase small subunit 2, chloroplastic-like
Alg148 hyoscyamine 6-dioxygenase
Alg149 aspartic proteinase oryzasin-1-like
Alg150 protein kinase superfamily protein - -
Alg151 probable gamma-aminobutyrate transaminase 3, mitochondrial - -
Alg152 Formate dehydrogenase, mitochondrial
Alg153 fructokinase-2 5 2 +
Dolichyl-diphosphooligosaccharide-- glycosyltransferase 48 kDa
Alg154
subunit -
Alg155 sucrose synthase 5 29
N1 Puroindoline A - + +
N2 RNA recognition domain containing protein,expressed 5 1
N3 4S Ribosomal protein S15a
N4 Heat shock protein 9
N5 Importin subunit alpha
N6 Alanine aminotransferase 2 +
N7 Actin-depolymerizing factor 4
N8 Histone H1 -
N9 Crooked neck-like protein 1
N10 Heat shock cognate 7 kDa protein 1
N11 6S Ribosomal protein L18 1 1 -
N12 DNA-directed RNA polymerases I, II, and III subunit RPABC1 5 25
N13 dna-binding protein -
N14 14-3-3-like protein B
N15 peter pan-like protein 5 10 +
N16 DEAD-box ATP-dependent RNA helicase 27 5 14
N17 phosphorylase 1 8
N18 Histone H1 -

146
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S
N19 coiled-coil domain-containing protein 12 5 1
N20 6S acidic Ribosomal protein P2B + +
N21 6S Ribosomal protein L23a
N22 Elongation factor 2
N23 Multiprotein-bridging factor 1a 1 10
N24 4S Ribosomal protein S11 1 4
N25 Serine/arginine-rich splicing factor 4
N26 phosphoglycerate kinase, cytolic 1 1
N27 Poly(A)-binding protein
N28 Eukaryotic translation initiation factor isoform 4G-1
N29 Eukaryotic translation initiation factor 5A2
N30 Histone H1 WH1B.1
N31 tyrosine-protein phosphatase 5 30
N32 fam10 family protein at4g22670-like
N33 probable aldo-keto reductase 1 1 4
N34 Importin subunit alpha
N35 trihelix transcription factor +
N36 protein rrp5 homolog 5 2
N37 DNA-directed RNA polymerase
N38 guanine nucleotide-binding 3 homolog +
N39 protein rcc2
N40 60s ribosomal protein l30 -
N41 Spastin
N42 formate--tetrahydrofolate ligase 1 5
N43 dead-box atp-dependent rna helicase 28 5 1 +
N44 brain acid soluble protein 1 +
N45 nucleolar complex protein 3 homolog 5 27
N46 betaine aldehyde dehydrogenase 1 4 +
N47 Uncharacterized (GO: nucleic acid binding)
N48 sucrose synthase +

147
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

3. CONCLUSIONS
L’analyse transcriptomique réalisée dans cette étude a révélé que 70% des gènes annotés
sur le génome de T. monococcum étaient exprimés pendant le développement du grain.
Récemment dans le grain de riz une proportion similaire a été observée (Biselli et al., 2015).
Parmi ces gènes exprimés, nous avons vu que 386 étaient très significativement
différentiellement exprimés entre au moins deux traitements et à au moins un stade de
développement donné pendant le remplissage du grain (entre 300 et 600°Cj). Beaucoup de ces
gènes ont des fonctions liées à la mise en place des réserves et aux mécanismes de défense.
Parmi les résultats, nous avons observé une diminution de l'expression de plusieurs gènes
intervenant dans le métabolisme des sucres lorsqu'une forte quantité d'N est disponible pour le
grain, appuyée par des résultats similaires observés sur les quantités de saccharose du grain. Le
métabolisme des acides aminés se comporte de manière inverse et semble activé dans cette
condition, comme le montrent l’expression plus forte de gènes liés à leur transport et à leur
synthèse ainsi que la quantité totale d’acides aminés du grain. Parmi ces acides aminés, la
glutamine et l’asparagine, qui sont les principaux acides aminés transportés des parties
végétatives vers le grain (Masclaux-Daubresse et al., 2010) ont été augmentés avec les
traitements contenant des forts niveaux d’N. Les résultats obtenus dans cette étude confirment
donc ceux obtenus dans le chapitre précédent (notamment le rôle du ratio N/S) et apportent des
informations supplémentaires sur les gènes impliqués dans les mécanismes de réponse décrits.
L’analyse transcriptomique a clairement révélé un fort effet de la carence en S sur le
niveau d’expression des gènes du grain de T. monococcum, carence provoquée par un apport
d’N important sans ajout de S. Le clustering a permis d’identifier de nombreux gènes
surexprimés dans cette condition, dont plusieurs sont apparus fortement co-exprimés dans les
réseaux générés, supposant leur action synergique dans le grain. Parmi eux, outre des gènes
impliqués dans le métabolisme des acides aminés mentionnés dans le paragraphe précédent,
plusieurs gènes liés au transport et au métabolisme du S ont été retrouvés. L’activation du
métabolisme du S en condition de carence est un mécanisme plusieurs fois décrit chez les
plantes (Nakamura et al., 1999 ; Lewandowska and Sirko, 2008). Certains gènes tels que des
transporteurs pourraient avoir un rôle capital dans la réponse du grain à cette situation de
carence. De plus, quelques facteurs de transcription étaient surexprimés dans cette condition et
pourraient être impliqués dans le contrôle de ce mécanisme de régulation.
Si certains gènes ont pu être mis en évidence dans cette étude, l’interprétation des résultats
demeure encore incomplète. En effet, plusieurs gènes n’ont pu être annotés fonctionnellement.
Leur rôle putatif dans la réponse biologique peut être imaginé en observant les gènes
caractérisés auxquels ils sont connectés au sein des réseaux de co-expression, sur la base du
principe de « guilt by association » (Oliver, 2000). Néanmoins, il est prévu de poursuivre la
caractérisation in silico de ces gènes afin d’obtenir le maximum d’informations sur les fonctions
impactées dans le grain par l’N et le S.
Dans cette étude, nous avons proposé une interprétation de la réponse globale du grain de
[Link] à la nutrition N et S. Toutefois, même si les interactions mises en évidence sont
parfois complexes, il est nécessaire de garder à l'esprit que les conclusions tirées sont
dépendantes des choix qui ont été faits (méthodes d'analyses, valeurs de seuils utilisées,...) et

148
 Chapitre 5 : La réponse intégrative du grain de Triticum monococcum à la nutrition N et S

une complexité beaucoup plus grande dans la réalité de la réponse cellulaire face au stress
nutritionnel peut facilement être imaginée.

Références :
Biselli, C., Bagnaresi, P., Cavalluzzo, D., et al. (2015) Deep sequencing transcriptional
fingerprinting of rice kernels for dissecting grain quality traits. BMC Genomics.
Lewandowska, M. and Sirko, A. (2008) Recent advances in understanding plant response to
sulfur-deficiency stress. , 55, 457–471.
Masclaux-Daubresse, C., Daniel-Vedele, F., Dechorgnat, J., Chardon, F., Gaufichon, L.
and Suzuki, A. (2010) Nitrogen uptake, assimilation and remobilization in plants:
challenges for sustainable and productive agriculture. Ann. Bot., 105, 1141–1157.
Nakamura, T., Yamaguchi, Y. and Sano, H. (1999) Four rice genes encoding cysteine
synthase: isolation and differential responses to sulfur, nitrogen and light. Gene, 229, 155–
161.
Oliver, S. (2000) Guilt-by-association goes global. Nature, 403, 601–603.

149
150
 CHAPITRE 6 :
Conclusions générales et
perspectives de la thèse

151
152
 Chapitre 6 : Conclusions générales et perspectives de la thèse

1. CONCLUSIONS
Cette thèse a apporté de la connaissance à la compréhension des mécanismes moléculaires
pouvant réguler le développement du grain de blé et la mise en place de ses réserves en fonction
de la nutrition. Le protéome nucléaire a occupé une place importante dans les analyses réalisées
dans ce travail. Les protéines nucléaires identifiées ont tout d’abord permis de valider une
méthode permettant leur extraction à partir de grains de blé. L’une des difficultés rencontrées
pour ce type d’extraction a été la présence de PR au sein des extraits protéiques. Toutefois, d’un
point de vue qualitatif, l’enrichissement en protéines nucléaires a été estimé entre 60 et 85%
suivant le stade de développement des grains utilisés. Outre les protéines histones et protéines
ribosomales, attendues et retrouvées en abondance au sein des extraits nucléaires, un nombre
important de protéines intervenant dans divers processus ont pu être identifiées, jusqu’à 1677
retrouvées dans le grain de T. monococcum. Cette liste de protéines apporte des informations
sur le fonctionnement du noyau, qui reste aujourd’hui peu connu chez les plantes malgré son
rôle central dans l’organisation et l’expression du génome végétal (Petrovská et al., 2015).
L’analyse de ce sous-protéome chez le blé tendre et l’engrain a permis de mettre en
évidence des protéines qui pourraient avoir un rôle important dans la régulation de la
transcription. Chez le blé tendre, des protéines liées à l’organisation de la chromatine (e.g.
complexe de remodelage de la chromatine, protéines HMG, Histone déacétylase) ont présenté
des profils d’accumulation suggérant des modifications dans la conformation de l’ADN en fin
de phase de cellularisation et au début de la phase de remplissage. Chez l’engrain nous avons
pu voir que ces mécanismes épigénétiques, via la méthylation de l’ADN, pourraient avoir un
rôle dans la réponse du grain à un changement de nutrition au cours du remplissage, notamment
suite à un apport de S. Comme il l’a été discuté dans le chapitre 4, la méthylation des promoteurs
de gènes de PR observée chez l’orge semble être impliquée dans la régulation de leur
transcription et dans le contrôle du développement du grain (Sørensen et al., 1996; Radchuk et
al., 2005). La méthylation de l’ADN pourrait également modifier la capacité de liaison des FT
à leur séquence promotrice cible, comme il l’a été observé pour le FT Opaque2 chez le maïs
(Sturaro and Viotti, 2001 ; Locatelli et al., 2009).
Dans l’ensemble des jeux de données du protéome nucléaire, peu de FT, du moins
fonctionnellement annotés de la sorte, ont été identifiés et quantifiés. De plus, parmi les
protéines répondant à la nutrition, il était attendu de retrouver les FT du modèle de régulation
transcriptionnelle établi chez l’orge et qui sont également présents chez les céréales (Rubio-
Somoza et al., 2006 ; Verdier and Thompson, 2008). Toutefois l’un d’eux, PBF, a été identifié
dans le grain de T. monococcum au stade 300°Cj après floraison. D’après les données de
transcriptomique obtenues à partir du même matériel biologique, il est apparu que ce FT était
le plus exprimé des huit FT du modèle de régulation et présentait un pic d’expression au stade
300°Cj (Résultats non présentés). Ceci suggère que l’absence de ces FT des jeux de données du
protéome nucléaire, pourtant conséquents, résulte d’une trop faible abondance de ces protéines
régulatrices. Malgré un fractionnement subcellulaire, probablement que cette faible quantité
limite techniquement leur détection au sein de mélanges peptidiques complexes, composés de
nombreux peptides issus de protéines majoritaires telles que les histones, limitant ainsi leur
identification par spectrométrie de masse.

153
 Chapitre 6 : Conclusions générales et perspectives de la thèse

Néanmoins, plusieurs protéines possédant un domaine de liaison à l’ADN ont été

significativement impactées par la nutrition (21% des protéines nucléaires répondant à la
nutrition). L’apport de N et de S impacte donc des régulateurs transcriptionnels, qui pourraient
jouer un rôle dans la régulation du développement du grain et la mise en place de ses réserves.
Les différents traitements appliqués à la culture de T. monococcum ont conduit à
différents rapports N/S du grain (Figure 25a) dès 300°Cj après floraison, au début de la période
de remplissage. Ces différents ratios N/S se traduisent à maturité par différentes compositions
en PR du grain, les PR riches en S étant plus fortement accumulées avec un apport de S et les
PR pauvres en S étant fortement accumulées en présence d’N, confirmant ainsi les résultats
précédemment obtenus chez le blé tendre (Wieser et al., 2004 ; Zörb et al., 2010). Ce rapport
N/S montre une carence en S dans deux des quatre traitements appliqués (N-S- et N+S-). Il est
très fortement augmenté avec le traitement N+S-, c’est-à-dire lorsque de l’N a été apporté à la
culture sans apport concomitant de S, conduisant à une situation de forte carence en S dans le
grain. Dans cette condition, la composition du grain en PR est clairement impactée avec une
augmentation nette des HMW-GS au détriment de certaines gliadines (Figure 25b). Ces
résultats montrent une nouvelle fois l’importance d’un apport conjoint d’N et de S. En revanche,
lorsque le S est apporté en grande quantité sans apport de N, le ratio N/S, bien qu’il soit le plus
faible dans cette condition de nutrition, n’est pas aussi nettement impacté. Il semble donc que
dans cette condition, l’N du grain soit issu en grande partie de la remobilisation de l’N accumulé
dans les parties aériennes pendant le développement végétatif de la plante (Kichey et al., 2007).
La situation de carence en S observée lorsque l’N est apporté seul traduit probablement une
faible remobilisation du S. En effet, chez le blé, le S du grain est issu pour moins de 50% de la
remobilisation (Monaghan et al., 1999).

Figure 25. Effets de la nutrition azotée et soufrée sur le rapport N/S du grain (a) et sur la
composition du grain mature chez T. monococcum (b). En (b) les données sont exprimées en
pourcentage de la quantité totale de PR mesurée en g.g-1 de matière sèche.

L’apport de S a conduit à une quantité plus importante de PR, due à une forte
accumulation des gliadines, essentiellement les sous-unités α/β- et γ- dont les vitesses et durée
d’accumulation ont été nettement augmentées. Nous avons ainsi pu observer que chez l’engrain,
un apport de S après floraison conduit à une quantité plus importante de PR, sans vraiment avoir

154
 Chapitre 6 : Conclusions générales et perspectives de la thèse

une incidence sur la composition du grain, probablement du fait d’une bonne remobilisation de
l’N. La plus grande quantité de PR a été observée lorsque de l’N et du S ont été apportés de
manière conjointe et ce résultat est sûrement lié à la plus forte abondance observée pour
plusieurs protéines nucléaires associées à la transcription et la maturation des ARNm.
Contrairement à ce que l’on aurait pu attendre, la majorité des ARNs, protéines, et
métabolites quantifiés n’a pas répondu de cette façon-là à la nutrition. L’hypothèse principale
est que ces variables ne répondent pas à la quantité totale d’N et de S dans le grain mais plutôt
à la modification du rapport N/S. En effet, les perturbations les plus importantes ont été
obtenues en réponse à un apport de N ou de S seul. Ainsi, nous avons observé une forte
expression et accumulation des gènes et protéines impliqués dans le métabolisme et le transport
du S, sans doute pour compenser la situation de carence provoquée par un fort apport de N sans
ajout de S. Dans le même temps, de nombreux ARNs et protéines liés au métabolisme des
acides aminés ont vu leur quantité augmenter face à un fort apport d’N. Ces observations ont
été confortées par des résultats similaires obtenus au niveau des acides aminés quantifiés. De
nombreuses variables liées au transport ont également été activées par l’N. Il s’agit de
transporteurs de sulfate et d’acides aminés ou encore d’importines impliquées dans le transport
nucléo-cytoplasmique. Certains de ces transporteurs pourraient avoir un rôle central dans
l’ajustement du métabolisme du grain en réponse à une modification du rapport N/S, comme
par exemple le transporteur EamA-like qui représentait un « hub » au sein du module de réponse
à l’N dans le réseau de co-expression.
En parallèle de ces changements observés au niveau du métabolisme et du transport
cellulaire, comme nous l’avons vu précédemment plusieurs gènes ou protéines possédant des
domaines de liaison à l’ADN ont montré des variations significatives en réponse à la nutrition.
Pour ces variables, il est difficile de définir un type majoritaire de réponse à la nutrition, toutes
étant retrouvées à des endroits très différents au sein des réseaux biologiques générés. Plusieurs
hypothèses peuvent être émises pour l’expliquer. Peut-être que ces protéines agissent seules,
chacune ayant un rôle associé à une situation précise de stress. Peut-être qu’au contraire ces
protéines forment des complexes, mais dont les partenaires n’ont pu être mis en évidence dans
cette étude du fait de leur quantité très faible dans la cellule. Quoi qu’il en soit ces gènes
pourraient agir comme des senseurs du rapport N/S ou de la quantité totale d’N et de S et avoir
un rôle très important dans l’adaptation du grain pour l’accumulation de ses réserves face à un
stress nutritionnel.
Si les FT connus pour être impliqués dans la régulation de la synthèse des PR n’ont pu
être détectés en protéomique, leur niveau d’expression qui a pu être quantifié par RNA-Seq n’a
pas été affecté par la nutrition. Ces gènes ne semblent donc pas régulés par l’N et le S au niveau
transcriptionnel, du moins dans les conditions d’analyse que nous avons utilisées. On peut alors
émettre l’hypothèse que ces FT pourraient être régulés par la nutrition à d’autres niveaux, par
exemple par des modifications post-traductionnelles. On sait en effet que ces modifications
peuvent affecter la conformation, l’activité ou encore la localisation des protéines. Chez le maïs
par exemple, il a été démontré que l’état de phosphorylation d’Opaque 2 influence sa capacité
de liaison aux promoteurs de gènes de PR (Ciceri et al., 1997). La phosphorylation est une des
modifications post-traductionnelles les plus caractérisées et est impliquée dans de nombreuses
réponses physiologiques (Bond et al., 2011 ; Ytterberg and Jensen, 2010). Chez A. thaliana, il

155
 Chapitre 6 : Conclusions générales et perspectives de la thèse

a été démontré que plusieurs régulateurs transcriptionnels subissent une modification de leur
état de phosphorylation suite à un apport d’N (Engelsberger and Schulze, 2012).
Nous avons pu visualiser chez le blé tendre via des colorations de gels 2D au Pro-Q
Diamond (un colorant spécifique des groupements phosphates des phosphopeptides) que
certaines protéines nucléaires étaient phosphorylées en fin de phase de cellularisation. Il s’agit
entre autres d’une protéine HMG et une histone déacétylase qui ont un rôle dans l’organisation
de la chromatine. Ces analyses n’ont pas été poursuivies. Néanmoins, elles appuient l’hypothèse
que la phosphorylation pourrait jouer un rôle majeur dans la régulation du développement du
grain. De plus, l’impact de l’N et du S sur la quantité de plusieurs kinases et phosphatases sous-
entend l’implication de mécanismes de phosphorylation/déphosphorylation dans la réponse du
grain à la nutrition.

2. PERSPECTIVES
Les travaux réalisés au cours de cette thèse ont apporté de nouvelles informations sur la
nature et les fonctions des protéines nucléaires de la cellule végétale et sur les mécanismes
moléculaires mis en jeu en réponse à la nutrition N et S au cours du développement du grain de
blé.
En intégrant différents types de données -omiques, nous avons pu voir que la réponse du
grain à la nutrition met en jeu de nombreux gènes, protéines et métabolites, présentant des
profils très variés, témoignant ainsi de la complexité de la réponse cellulaire. Bien sûr, la vision
globale obtenue ne rend pas compte de l’intégralité des niveaux de régulation impliqués dans
ce type de réponse. La cellule réagit probablement de manière bien plus complexe que ce qui a
pu être observé. En effet, comme nous l’avons vu les modifications post-traductionnelles et les
mécanismes épigénétiques pourraient tenir une place importante dans les processus de
régulation mis en jeu. Des approches d’étude du phosphoprotéome et de la méthylation de
l’ADN pourraient être envisagées pour compléter notre compréhension des effets de l’N et du
S sur le grain de blé. Ces éléments permettraient d’apporter des pièces importantes au puzzle
de la régulation de l’accumulation des PR du grain (Mehrotra et al., 2009).
Au sein des réseaux générés dans cette thèse, les liens entre attributs reflètent leur co-
expression ou co-accumulation. Les réseaux sont donc non-orientés. Pour aller plus loin dans
la compréhension des mécanismes de régulation de l’accumulation des réserves du grain, il
serait intéressant d’identifier le sens de ces interactions. Tout d’abord, les interactions protéine-
ADN et protéine-protéine qui ont été démontrées expérimentalement pourraient être ajoutées
aux réseaux obtenus. Puis dans cette optique, certains gènes apparus comme des « hub » au sein
des réseaux pourrait faire l’objet d’études plus fonctionnelles. Une façon de procéder
consisterait à inspecter la variabilité génétique existante pour ces gènes centraux. Si du
polymorphisme de séquence est détecté, la culture des variétés de blé concernées et l’analyse
des variations d’expression de ces gènes et des gènes avec qui ils sont fortement co-exprimés
dans les réseaux permettrait de détecter des locus de traits quantitatifs d’expression (eQTL ;
Kliebenstein et al., 2006 ; Kliebenstein, 2009). D’une part, en visualisant l’influence des

156
 Chapitre 6 : Conclusions générales et perspectives de la thèse

variations d’expression d’un gène central sur l’expression d’autres gènes, le sens de certaines
interactions existant dans les réseaux serait déterminé et les liens de régulation génique seraient
alors mis en évidence. D’autre part, une analyse des PR en parallèle permettrait alors de relier
les variations quantitatives des gènes étudiés au phénotype d’intérêt, à savoir la composition du
grain du blé.
Nos travaux se sont inscrits dans un sujet de recherche à connotation très fondamentale.
En effet, il faut garder à l’esprit que les résultats obtenus sont issus de culture de blé en
conditions contrôlées. Les dispositifs choisis nous ont permis de contrôler au mieux la
disponibilité de l’N et du S au sein du substrat et éviter l’effet d’autres variables
environnementales. Nos résultats auraient pu être différents si les plantes avaient été cultivées
en plein champ, où la disponibilité en éléments nutritifs est largement dépendante de la
compaction du sol ou encore de la pluviométrie. Toutefois, les résultats obtenus suggèrent des
pistes de travail utilisables en sélection variétale à moyen terme. Certains gènes mis en
évidence, notamment des transporteurs activés en situation de carence en S, ou encore des
régulateurs transcriptionnels, pourraient faire l’objet d’études fonctionnelles et pourquoi pas,
s’avérer utiles dans l’objectif de maitrise de la qualité du grain de blé dans un contexte de
limitation des intrants.

Références :
Bond, A.E., Row, P.E. and Dudley, E. (2011) Post-translation modification of proteins;
methodologies and applications in plant sciences. Phytochemistry, 72, 975–996.
Ciceri, P., Gianazza, E., Lazzari, B., Lippoli, G., Genga, A., Hoschek, G., Schmidt, R.J.
and Viotti, A. (1997) Phosphorylation of Opaque2 changes diurnally and impacts its DNA
binding activity. , 9, 97–108.
Engelsberger, W.R. and Schulze, W.X. (2012) Nitrate and ammonium lead to distinct global
dynamic phosphorylation patterns when resupplied to nitrogen-starved Arabidopsis
seedlings. Plant J., 69, 978–95.
Kichey, T., Hirel, B., Heumez, E., Dubois, F. and LeGouis, J. (2007) In winter wheat
(Triticum aestivum L.), post-anthesis nitrogen uptake and remobilisation to the grain
correlates with agronomic traits and nitrogen physiological markers. F. Crop. Res., 102,
22–32.
Kliebenstein, D. (2009) Quantitative genomics: analyzing intraspecific variation using global
gene expression polymorphisms or eQTLs. Annu. Rev. Plant Biol., 60, 93–114.
Kliebenstein, D.J., West, M. AL, Leeuwen, H. van, Loudet, O., Doerge, R. and St Clair,
D. a (2006) Identification of QTLs controlling gene expression networks defined a priori.
BMC Bioinformatics, 7, 308.
Locatelli, S., Piatti, P., Motto, M. and Rossi, V. (2009) Chromatin and DNA modifications
in the Opaque2-mediated regulation of gene transcription during maize endosperm
development. Plant Cell, 21, 1410–1427.
Mehrotra, R., Kumar, S., Mehrotra, S. and Singh, B.D. (2009) Seed storage protein gene
regulation- A jig-saw puzzle. Indian J. Biotechnol., 8, 147–158.
Monaghan, J.M., Scrimgeour, C.M., Stein, W.M., Zhao, F.J. and Evans, E.J. (1999)
Sulphur accumulation and redistribution in wheat (Triticum aestivum): a study using stable
sulphur isotope ratios as a tracer system. Plant, Cell Environ., 22, 831–839.
Petrovská, B., Šebela, M. and Doležel, J. (2015) Inside a plant nucleus: discovering the
proteins. J. Exp. Bot., 66, 1627–1640.

157
 Chapitre 6 : Conclusions générales et perspectives de la thèse

Radchuk, V. V., Sreenivasulu, N., Radchuk, R.I., Wobus, U. and Weschke, W. (2005) The
methylation cycle and its possible functions in barley endosperm development. Plant Mol.
Biol., 59, 289–307.
Rubio-Somoza, I., Martinez, M., Abraham, Z., Diaz, I. and Carbonero, P. (2006) Ternary
complex formation between HvMYBS3 and other factors involved in transcriptional
control in barley seeds. Plant J., 47, 269–281.
Sørensen, M.B., Müller, M., Skerritt, J. and Simpson, D. (1996) Hordein promoter
methylation and transcriptional activity in wild-type and mutant barley endosperm. Mol.
Gen. Genet., 250, 750–760.
Sturaro, M. and Viotti, A. (2001) Methylation of the Opaque2 box in zein genes is parent-
dependent and affects O2 DNA binding activity in vitro. Plant Mol. Biol., 46, 549–560.
Verdier, J. and Thompson, R.D. (2008) Transcriptional regulation of storage protein
synthesis during dicotyledon seed filling. Plant Cell Physiol., 49, 1263–1271.
Wieser, H., Gutser, R. and Tucher, S. von (2004) Influence of sulphur fertilisation on
quantities and proportions of gluten protein types in wheat flour. J. Cereal Sci., 40, 239–
244.
Ytterberg, a J. and Jensen, O.N. (2010) Modification-specific proteomics in plant biology.
J. Proteomics, 73, 2249–66.
Zörb, C., Grover, C., Steinfurth, D. and Hermann Mühling, K. (2010) Quantitative
proteome analysis of wheat gluten as influenced by N and S nutrition. Plant Soil, 327,
225–234.

158
RÉSUMÉ
L’augmentation des rendements est un enjeu majeur chez les céréales. Dans cet objectif, il est
nécessaire de maintenir la qualité du grain de blé, qui est principalement déterminée par sa teneur
et sa composition en protéines de réserve. En effet, une forte relation négative existe entre le
rendement et la teneur en protéines. Par ailleurs, la qualité du grain est fortement influencée par
la disponibilité en azote et en soufre dans le sol. La limitation des apports d’intrants azotés à la
culture et la carence en soufre récemment observée dans les sols représentent ainsi des difficultés
supplémentaires pour maitriser cette qualité. Une meilleure connaissance des mécanismes
moléculaires impliqués dans le contrôle du développement du grain et la mise en place de ses
réserves protéiques en réponse à la nutrition azotée et soufrée est donc primordiale. L’objectif de
cette thèse a ainsi été d’apporter de nouveaux éléments à la compréhension de ces processus de
régulation, aujourd’hui peu connus. Pour cela, les approches -omiques sont apparues comme une
stratégie de choix pour identifier les acteurs moléculaires mis en jeu. Le protéome nucléaire a été
une cible importante dans les travaux menés. L’étude de ces protéines nucléaires a révélé certains
régulateurs transcriptionnels qui pourraient être impliqués dans le contrôle de la mise en place des
réserves du grain. Dans une approche combinant des données de protéomique, transcriptomique
et métabolomique, une vision intégrative de la réponse du grain à la nutrition azotée et soufrée a
été obtenue. L’importance d’un apport de soufre dans le contrôle de la balance azote/soufre du
grain, déterminante pour la composition du grain en protéines de réserve, a été clairement vérifiée.
Parmi les changements observés au niveau du métabolisme cellulaire, certains des gènes affectés
par la modification de cette balance pourraient orchestrer l’ajustement de la composition du grain
face à des situations de carences nutritionnelles. Ces nouvelles connaissances devraient permettre
de mieux maitriser la qualité du grain de blé dans un contexte d’agriculture durable.
Mots clés : Blé, grain, protéines de réserve, azote, soufre, protéines nucléaires, données
omiques, réseaux biologiques

ABSTRACT
Improving the yield potential of cereals represents a major challenge. In this context, wheat grain
quality has to be maintained. Indeed, grain quality is mainly determined by the content and the
composition of storage proteins, but there is a strongly negative correlation between yield and
grain protein concentration. In addition, grain quality is strongly influenced by the availability of
nitrogen and sulfur in soils. Nowadays, the limitation of nitrogen inputs, and also the sulfur
deficiency recently observed in soils represent major difficulties to control the quality. Therefore,
understanding of molecular mechanisms controlling grain development and accumulation of
storage proteins in response to nitrogen and sulfur supply is a major issue. The objective of this
thesis was to create knowledge on the comprehension of these regulatory mechanisms. For this
purpose, the best strategy to identify molecular actors involved in these processes consisted of -
omics approaches. In our studies, the nuclear proteome was an important target. Among these
proteins, we revealed some transcriptional regulators likely to be involved in the control of the
accumulation of grain storage compounds. Using an approach combining proteomic,
transcriptomic and metabolomic data, the characterization of the integrative grain response to the
nitrogen and sulfur supply was obtained. Besides, our studies clearly confirmed the major
influence of sulfur in the control of the nitrogen/sulfur balance that determines the grain storage
protein composition. Among the changes observed in the cell metabolism, some genes were
disturbed by the modification of this balance. Thus these genes could coordinate the adjustment
of grain composition in response to nutritional deficiencies. These new results contribute in facing
the challenge of maintaining wheat grain quality with sustainable agriculture.
Keywords: Wheat, grain, storage proteins, nitrogen, sulfur, nuclear protein, -omics data,
biological network