ANRV335-PM03-09 ARI 9 December 2007 17:30

Metastatic Cancer Cell

Marina Bacac and Ivan Stamenkovic
Experimental Pathology Unit, Department of Pathology, University of Lausanne,
Switzerland; email: [Link]@[Link], [Link]@[Link]
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

Annu. Rev. Pathol. Mech. Dis. 2008. 3:221–47 Key Words

First published online as a Review in Advance on tumor-host interactions, invasion, adhesion, proteolysis
September 17, 2007

The Annual Review of Pathology: Mechanisms of Abstract

Disease is online at [Link]
Metastasis is the result of cancer cell adaptation to a tissue microen-
This article’s doi: vironment at a distance from the primary tumor. Metastatic cancer
10.1146/[Link].3.121806.151523
cells require properties that allow them not only to adapt to a for-
Copyright  c 2008 by Annual Reviews. eign microenvironment but to subvert it in a way that is conducive
All rights reserved
to their continued proliferation and survival. Recent conceptual and
1553-4006/08/0228-0221$20.00 technological advances have contributed to our understanding of
the role of the host tissue stroma in promoting tumor cell growth
and dissemination and have provided new insight into the genetic
makeup of cancers with high metastatic proclivity.

221
 ANRV335-PM03-09 ARI 9 December 2007 17:30

INTRODUCTION has been learned regarding the properties that

such a cell requires, several key questions are
The ability to metastasize is a hallmark of
still unresolved. For example, do cells display-
malignant tumors, and metastasis is the prin-
ing metastatic proclivity emerge late in tu-
cipal cause of death among cancer patients.
mor progression as a result of multiple mu-
It is the single most challenging obstacle to
tations and selection, or are they part of the
successful cancer management and may also
cell population that constitutes early malig-
be viewed as the last frontier of cancer re-
nant growth? What role does the host mi-
search. Metastasis is the process whereby can-
croenvironment play in tumor cell dissemina-
cer cells spread throughout the body, estab-
tion, and which components of tumor-stroma
lishing new colonies in organs at a distance
cross talk might provide potential therapeutic
from the one where the primary tumor orig-
targets?
inated. It is well established that metastasis is
At least five functions are required for a
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

a complex, multistep process, which although

tumor cell to successfully complete the se-
considered highly inefficient from the cellular
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

quence of events outlined above. They in-

point of view, virtually constitutes a death sen-
clude interaction with the local microenvi-
tence for the patient. Its prevention and man-
ronment, migration, invasion, resistance to
agement are therefore among the key goals in
apoptosis, and the ability to induce angio-
clinical and basic cancer research.
genesis. All five functions are regulated by
There are three major routes for tu-
adhesion and proteolysis, which together pro-
mor dissemination: lymphatic vessels, blood
vide the most fundamental molecular effector
vessels, and serosal surfaces. Tumor metas-
mechanisms upon which a metastatic cell re-
tases occurring via these three routes are re-
lies (Table 1). Adhesion and proteolysis de-
ferred to as lymphatic, hematogenous, and
termine tumor cell interaction with other cells
transcoelomic, respectively. Epithelial malig-
and with the extracellular matrix (ECM), help
nancies, or carcinomas, typically begin their
create a path for migration, promote angio-
dissemination by the lymphatic route, with
genesis, and both directly and indirectly trig-
hematogenous metastases occurring at a later
ger survival signals.
time. In contrast, bone and soft tissue tu-
Transformation results in major pheno-
mors, or sarcomas, preferentially metasta-
typic changes that affect cell surface receptor
size by the hematogenous route, whereas
expression, cytoskeletal function, growth
transcoelomic metastasis is the property of
factor and cytokine secretion, proteolytic
a relatively small group of tumors that in-
enzyme production, and the glycosyltrans-
cludes mesotheliomas and ovarian carcino-
ferase and glycosidase repertoire (Figure 2).
mas. Because of their prevalence, lymphatic
These combined changes alter the way in
and hematogenous metastases will provide the
which the transformed cell communicates
main focus of the present review.
with its microenvironment and, by the same
A tumor cell that initiates a metastatic
token, the way in which it is perceived by
colony must (a) detach from the primary
normal surrounding cells. They not only
mass, (b) invade the local host tissue stroma,
provide transformed cells with the ability to
(c) penetrate local lymphatic and blood ves-
disrupt the barriers that keep their normal
sels, (d ) survive within the circulation, (e) be-
counterparts confined to a defined tissue
come arrested in capillaries or venules of
compartment, but also with the means to
other organs, ( f ) penetrate the corresponding
subject the host tissue microenvironment to
parenchyma, ( g) adapt to the newly colonized
their rules and use its resources for their own
milieu or subvert the local microenvironment
survival, growth, and dissemination. It is in-
to suit its own needs, and (h) divide to form
creasingly clear that tumor cells depend upon
the new tumor (Figure 1). Although much

222 Bacac · Stamenkovic

 ANRV335-PM03-09 ARI 9 December 2007 17:30

Invasive carcinoma Detachment

In situ carcinoma

Normal epithelia

Basement membrane Basement membrane

degradation
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

Invasion
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

Collagen fibers
Migration
Dormant
metastases No proliferation
Intravasation

Extravasation Circulation

Proliferation, angiogenesis,
microenvironment activation

Growing metastases

Figure 1
Principal steps in metastasis. Transformation of normal epithelial cells leads to carcinoma in situ, which,
as a result of loss of adherens junctions, evolves toward the invasive carcinoma stage. Following basement
membrane degradation, tumor cells invade the surrounding stroma, migrate and intravasate into blood or
lymph vessels, and become transported until they arrest in the capillaries of a distant organ.

their microenvironment to metastasize and diverse primary and metastatic tumors, is

that they even rely on host tissue stromal cells that metastatic cells may constitute part of
to provide functions, such as a diversity of the early makeup of a malignant tumor. This
proteolytic activity, that they themselves may review highlights our current understanding
lack. Understanding tumor-host interactions of tumor cell properties and host tissue
may therefore provide a key to understanding responses whose combination culminates in
metastasis. A more recently emerging view, cancer metastasis and discusses the origin of
based on gene expression profile analysis of metastatic cells in light of recent observations.

[Link] • Metastatic Cancer Cell 223

 ANRV335-PM03-09 ARI 9 December 2007 17:30

Table 1 Summary of adhesion molecules and proteolytic enzymes discussed in the text that are implicated in tumor
metastasis on the basis of experimental evidence
Mechanism Candidate functions Reference(s)
Adhesion
Cadherins
E-cadherin Promotes cell-cell adhesion; prevents cell detachment; cleaved by MMP-3, MMP-7, (1, 2, 4, 5)
and ADAMs; tumor suppressor
N-cadherin Promotes migration and invasion; regulates activation of FGFRs (19, 20)
Integrins
α2β1, α3β1 Enhance metastasis in selected experimental models (28, 29)
αvβ3 Promotes migration and invasion; mediates tumor cell-platelet interactions (30, 123)
α6β4 Promotes migration, invasion, and proliferation; cooperates with RTKs (31, 32)
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

α4β1 Mediates tumor cell adhesion to endothelial cells (126)

Immunoglobulin superfamily
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

VCAM-1 Mediates tumor cell–endothelial cell adhesion (126)

L1 Promotes tumorigenicity and motility (55)
NrCAM Promotes tumorigencity and migration (56)
NCAM Its downregulation is associated with enhanced lymph node metastasis in some tumor (57)
models
Selectins Mediate tumor cell–endothelial interactions and tumor cell–platelet/leukocyte adhesion (119–122)
Cell surface proteoglycans
CD44 Mediates interaction with hyaluronan; interacts with RTKs; specific isoforms provide a (58–60, 62, 63)
scaffold for the assembly of molecular complexes that can promote metastasis
Proteolysis
Matrix metalloproteinases
MMP-1 Degrades collagen; promotes invasion (79, 85)
MMP-2 Activates growth factors, including TGF-β; promotes invasion; interacts with αvβ3 (79, 85)
integrin on the cell surface
MMP-3 Promotes tuumorigenesis; activates growth factors, including HB-EGF (79, 85, 93)
MMP-7 Promotes cell survival; activates HB-EGF; interacts with CD44 on the cell surface (63, 79, 85)
MMP-9 Promotes invasion; enhances angiogenesis; promotes intravasation; activates TGF-β; (62, 79, 85)
interacts with CD44 on the cell surface
MMP-14, -15, -16 Degrade native basement membrane; promote invasion (77–79, 85)
Cathepsins Promote tumor growth and invasion; may be expressed by tumor cells or exclusively by (74)
stromal cells

FUNDAMENTAL MOLECULAR formed and malignant cells that become de-

EFFECTOR MECHANISMS tached from the epithelium display loss of ad-
OF METASTASIS herens junctions, which, in epithelial cells,
are constituted primarily by E-cadherin (1,
Adhesion 2). Like other members of the cadherin fam-
Detachment, cadherins, and epithelial- ily, E-cadherin displays homophilic binding
to-mesenchymal transition. Interepithelial specificity. Its adhesive functions are stabi-
cell interactions are regulated by complex ad- lized by β-catenin, which binds to its cyto-
hesion mechanisms, including tight and ad- plasmic domain, providing a link to α-catenin
herens junctions and desmosomes (1). Trans- and the actin cytoskeleton (3). Experiments

224 Bacac · Stamenkovic

 ANRV335-PM03-09 ARI 9 December 2007 17:30

performed in vitro and in vivo have shown that

genetic and antibody-mediated inhibition of
CD44
E-cadherin function can alter the phenotype inactive
of epithelial cells from noninvasive to invasive
(4). Conversely, introduction of E-cadherin
into E-cadherin-deficient invasive carcinoma E-cadherin
cells abrogated their invasiveness (3). Down-
regulation of E-cadherin activity in vivo us- α5 β1 α3 β1
ing a dominant negative E-cadherin construct
in the transgenic Rip-Tag mouse model of
pancreatic islet cell carcinoma resulted in the BM
transition of a well-differentiated β-cell ade- Extracellular matrix
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

noma to an invasive carcinoma (5). Crossing

the Rip-Tag mice with mice that maintain
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

E-cadherin expression in transformed β-cells

arrested tumor progression in the adenoma
stage (5). Together, these observations argue
that E-cadherin functions as a tumor suppres- Selectin ligands
sor. CD44v3
N-cadherin c-Met Glycosaminoglycan
Although loss of function mutations occur
chains
in E-cadherin, they are not commonly ob-
pro-HB-EGF
served in malignant tumors (3). Mechanisms
that decrease or abrogate E-cadherin func- FGFR MMP-7
tion in carcinoma cells include transcriptional
repression, followed by promoter methyla- ErbB4
tion (6, 7), disruption of cytoskeletal connec- αv
β3 ErbB2
α6 β4 c-Met
tions, increased intracellular degradation, and
MMP-14 MMP-15
proteolytic cleavage of the extracellular do-
main by matrix metalloproteinases (MMPs)
(3). Several of the transcriptional repressors
Extracellular matrix
that control E-cadherin expression in devel-
opment, including Snail, Slug, SIP1, Twist, Figure 2
dEF1, and E12/E47 (8–12), are implicated Changes of adhesive properties in transformed cells. Normal epithelial cell
in E-cadherin downregulation in malignant communication with its microenvironment is regulated by
cells. An alternative mechanism that can in- E-cadherin-mediated cell-cell interaction and β1-integrin-mediated
adhesion to the basement membrane (BM). Transformation results in the
activate E-cadherin function in tumor cells
cadherin switch that leads to E-cadherin loss and replacement by
is disruption of the link between E-cadherin N-cadherin, which plays an important role in invasion by regulating
and the cytoskeleton. One example is pro- fibroblast growth factor receptor (FGFR) function. Carcinomas frequently
vided by mutations in β-catenin that abro- express αvβ3 and α6β4 integrins, which promote invasion and
gate its binding to α-catenin and result in a proliferation, in part through their interactions with receptor tyrosine
kinases, including ErbB2 and Met. Transformation also results in changes
nonadhesive phenotype (3). Coordinate re-
in glycosylation of cell surface proteoglycans such as CD44, which
ceptor tyrosine kinase (RTK)-integrin sig- regulates invasion by coordinating MMP-7-mediated heparin-binding
naling can also interfere with E-cadherin epidermal growth factor (HB-EGF) activation and ErbB4 signaling.
function. Activated RTKs and Src family ki- Changes in the glycosyltransferase repertoire can also result in the
nases induce tyrosine phosphorylation of the decoration of cell surface receptors by oligosaccharides that constitute
selectin ligands. Several transmembrane MMPs are expressed by tumor
E-cadherin–β-catenin complex, which is then
cells and mediate BM degradation.
recognized by the Cbl-like E3 ubiquitin lig-
ase and downregulated by endocytosis (13). In

[Link] • Metastatic Cancer Cell 225

 ANRV335-PM03-09 ARI 9 December 2007 17:30

v-src-transformed cells, this process is de- phospholipase C-γ, phosphatidylinositol

pendent upon integrin signaling and fo- 3-kinase, and mitogen-activated protein ki-
cal adhesion kinase (FAK) phosphorylation nase (MAPK). The combined action of these
(14, 15). Integrin signaling operates through signaling pathways promotes cell survival,
Snail/Slug to suppress E-cadherin expres- migration, and invasion (19, 20). Similar to
sion and disrupt adherens junctions. In some E-cadherin, the extracellular domain of N-
epithelial cells, this process may be medi- cadherin is susceptible to proteolytic cleavage
ated by integrin-linked kinase (16, 17). Fi- by MMPs. The N-cadherin cleavage product
nally, proteolytic cleavage of the extracellu- can block N-cadherin-mediated cell-cell ad-
lar domain of E-cadherin by MMP-3 and hesion, but can also stimulate FGFR signaling
MMP-7 disrupts its ability to promote cellular on adjacent cells in paracrine fashion (19,
interactions (18). Abrogation of E-cadherin- 20). Cleavage of N-cadherin at a site within
the transmembrane or cytoplasmic domain
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

mediated cell-cell adhesion results in detach-

ment of tumor cells from the epithelial cell by a γ-secretase-type protease results in the
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

layer and affects signaling pathways impli- translocation of its carboxy-terminal segment
cated in cell migration and growth, includ- to the nucleus, where it represses transcrip-
ing Rho GTPase-mediated modulation of the tion mediated by the CREB-binding protein
actin cytoskeleton and the canonical Wnt sig- (21).
naling pathway (3).
Loss of E-cadherin in malignant cells Interaction with the extracellular matrix
may be replaced by other cadherins, most integrins. Tumor cell interactions with the
commonly, N-cadherin (Figure 2). This ECM are mediated primarily by integrins and
process, known as the cadherin switch, is play a key role in tumor invasion and spread.
associated with a phenotypic change observed Integrins form a large family of adhesion re-
in vitro known as epithelial-to-mesenchymal ceptors, each member consisting of an α and
transition (EMT). EMT, defined as the a β transmembrane chain (22). In mammals,
conversion of epithelial cells to motile, 18 α and 8 β chains associate in various combi-
fibroblast-like cells that express mesenchymal nations to give rise to 24 integrins that recog-
rather than epithelial cell markers, is a nize distinct ECM ligands, with, nevertheless,
common event during normal embryonic some degree of overlap (22). When integrins
development and is observed as epithelial cell bind to ECM, they aggregate in the plane
progress through the stages of carcinogenesis of the cell membrane and associate with a
in vitro. It is proposed to reflect invasive and molecular complex composed of adaptor, sig-
metastatic properties of transformed epithe- naling, and cytoskeletal proteins that orches-
lial cells, but remains somewhat controversial trate the organization of actin filaments. Or-
because full EMT is difficult to prove in ganization of actin filaments into stress fibers,
vivo. Nevertheless, expression of N-cadherin in turn, promotes further integrin aggrega-
may make a critical contribution to invasion tion that results in increased ECM binding in
through both its adhesive and signaling a positive feedback loop. The outcome is an
functions. N-cadherin can mediate cell-cell integrin-mediated assembly of ECM and cy-
interactions with N-cadherin-expressing toskeletal protein clusters on each side of the
stromal cells, which may play an important cell membrane known as focal adhesions and
role in the ability of tumor cells to direct host ECM contacts (23).
responses. N-cadherin binds to and regulates Integrins trigger both mechanical and
the activation of fibroblast growth factor chemical signals that organize and remodel
receptors (FGFRs), thereby helping assemble the cytoskeleton of the cell, regulate adhesive
the FGFR-signaling complex, which triggers versus migratory interactions with the ECM,
downstream signaling pathways, including impart polarity, and control proliferation

226 Bacac · Stamenkovic

 ANRV335-PM03-09 ARI 9 December 2007 17:30

and survival. To exert their effects, integrins lial cell migration and carcinoma cell invasion
cooperate with RTKs, thereby jointly control- (32). RTKs may therefore augment the sig-
ling survival and mitogenic pathways. A re- naling functions of α6β4 at the expense of
ductionist view might be that integrins me- its ability to mediate stable adhesion. Simi-
diate cell adhesion and impose positional larly, cooperation between αvβ3 and platelet-
control on RTK activity, which together de- derived growth factor receptor (PDGFR) may
termine whether cells migrate and proliferate enhance growth and migration of tumor cells
in response to cytokines and growth factors. overexpressing PDGF (34). By altering their
Although seemingly straightforward, integrin integrin repertoire, neoplastic cells can hone
implication in cellular functions is compli- part of the molecular machinery that under-
cated by their diverse adhesive and signaling lies adhesion, migration, survival, and growth
properties that provide them the ability to to optimally serve their needs (Figure 2).
Upon clustering in focal adhesions, inte-
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

affect cell function in a variety of ways that

are often context dependent. Thus, α2β1 and grins activate several protein tyrosine kinases,
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

α3β1 integrins mediate epithelial cell adhe- central among which is FAK (35, 36). The
sion to basal lamina and maintain them in adaptor proteins paxillin and talin mediate in-
a quiescent state. These integrins are often tegrin interaction with FAK, which coordi-
downregulated in carcinomas and their re- nates many of the key events that constitute
expression in carcinoma cells can decrease or or are related to integrin signaling (36). Acti-
even revert the malignant phenotype (24–27). vation of FAK is initiated by autophosphory-
However, both of these integrins can enhance lation at Tyr397, which results in a structural
metastasis in selected experimental models motif recognized by SH2-domain-containing
(28, 29), underscoring the cell context and tu- proteins, including Src. Binding to FAK acti-
mor stage dependence of the effects of β1 in- vates Src, resulting in phosphorylation of ad-
tegrins. By contrast, αvβ3 (30) and α6β4 (31) ditional FAK residues and recruitment of sev-
integrins are frequently upregulated in car- eral signaling adaptor and effector proteins,
cinomas, where they may promote migration, guanine nucleotide exchange factors, GTPase
invasion, and proliferation. In addition to par- activating proteins, cytoskeletal adaptors, and
ticipating in hemidesmosome organization, proteolytic enzymes (23, 36, 37). The N-
the α6β4 integrin cooperates with epider- terminal domain of FAK provides a signal-
mal growth factor receptor (EGFR), ErbB2, ing linkage between integrins and RTKs,
and Met, and is likely to promote the growth especially EGFR and PDGFR (38). FAK
of carcinomas in which activating mutations thus constitutes a platform for the coordi-
of the corresponding growth factor receptor nated growth-factor-receptor–integrin signal
genes represent the oncogenic driving force exchange, regulation of Rho GTPase activ-
(31–33). In support of this view, introduction ity, and focal complex turnover. Binding of
of the β4 chain into β4-negative breast car- the GRB2 adaptor protein to FAK generates
cinoma cells activates the phosphatidylinosi- an important link to the activation of Ras
tol 3-kinase pathway, which results in activa- and the MAPK/extracellular-signal-regulated
tion of Rac and increased invasiveness of these kinase-2 (ERK2) cascade. ERK2 phospho-
cells in vitro (31). The cytoplasmic domain rylation can modulate focal contact dynam-
of β4 also acts as an adaptor and amplifier ics in motile cells in addition to promoting
of proinvasive signals induced by the hepato- both proliferation and survival. Recruitment
cyte growth factor receptor Met in cells un- of guanine nucleotide exchange factors, in-
dergoing Met-mediated transformation (33). cluding p190RhoGEF, may provide a direct
Both EGF and Met induce phosphorylation link to RhoA activation (39) and provide reg-
of β4 and enhance SHC signaling, which dis- ulation of promigratory versus adhesive inter-
rupts hemidesmosomes and increases epithe- actions with substrate. Cytoskeletal adaptors,

[Link] • Metastatic Cancer Cell 227

 ANRV335-PM03-09 ARI 9 December 2007 17:30

including ezrin, and proteolytic enzymes, with low and high colony-forming abilities
including calpain, regulate intracellular link- in the lung in experimental metastasis assays
age of focal contacts to the actin cytoskeleton identified RhoC as one of the most robustly
and focal contact turnover (40). upregulated genes in the highly metastatic
Cell migration induced by EGF or PDGF variants (49).
requires FAK association with both RTKs Although FAK is commonly associated
and integrin-containing focal complexes, con- with coordinating migration signals, there
sistent with the notion that FAK can inte- is abundant evidence to suggest that it can
grate promigratory signals from integrins and also promote invasion in both normal and
RTKs. These signals culminate in the activa- neoplastic cells (50, 51). Malignant cells fre-
tion of Rho GTPases and downstream effec- quently display elevated FAK levels and activ-
tors of the MAPK pathway, including ERK ity (50), which is associated with shape change,
podosome formation, and induction of in-
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

and Jun-amino-terminal kinase ( JNK) (23).

ERK and JNK participate in regulating cell vadopodia (52). In experimental models, the
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

migration by phosphorylating and activating invasive tumor phenotype was associated with
the myosin light chain kinase, which induces the accumulation of FAK-Src signaling com-
contraction of actomyosin fibers (41). JNK in- plexes within invadopodia, specialized cell
duces phosphorylation of paxillin, which may protrusions enriched in integrins and MMPs.
participate in cell migration by facilitating fo- Consistent with these observations, overex-
cal adhesion turnover (42). Rho GTPases ac- pression of FAK in some tumor cell types was
tivated by FAK include Rho, Rac, and Cdc42 reported to induce invasion (53, 54). Inter-
(43, 44). Cdc42 and Rac have both been estingly, the invasion-promoting property of
implicated in carcinoma invasion, as they FAK appears to be distinct from its ability to
promote actin polymerization at the leading induce migration (52).
edge and, consequently, formation of filopo-
dia and lamellipodia, respectively (43). Both Immunoglobulin superfamily adhesion
GTPases activate the ARP2/3 complex and receptors. Adhesion receptors belonging to
induce actin filament assembly coordinated by the immunoglobulin superfamily have been
the Wiskott-Aldrich syndrome protein (45). implicated in the progression of some carci-
They also activate p21-activated kinase, which nomas. The adhesion receptor L1 is highly ex-
enhances actin polymerization by activating pressed at the invasive front of colorectal can-
LIM kinase. Whereas Cdc42 and Rac pro- cers. L1 is a direct Wnt/β-catenin signaling
mote actin polymerization at the leading edge, pathway target in colorectal cancer cells, and
Rho orchestrates the assembly and contrac- in experimental model systems has been ob-
tion of actomyosin fibers, which pulls the served to enhance motility and tumorigenicity
trailing edge forward during migration. At (55).
least two Rho effector molecules, Rho kinase Similar to L1, the neuronal cell adhe-
(ROCK) and mammalian diaphanous, func- sion receptor NrCAM, which is also a tar-
tion jointly to induce the assembly of acto- get of Wnt signaling, promotes tumorigenic-
myosin fibers (46). By inhibiting myosin light ity and migration in various tumor cell types
chain phosphatase, ROCK promotes myosin (56). By contrast, downregulation of neural
light chain phosphorylation and actomyosin cell adhesion molecule is associated with en-
fiber contraction (47). Rho-ROCK signaling hanced lymph node metastasis in the Rip-Tag
appears to regulate several aspects of carci- transgenic mouse model of pancreatic islet
noma dissemination and is required for cancer cell carcinogenesis (57). Like N-cadherin,
cells to invade three-dimensional matrices by L1 and neural cell adhesion molecule bind
amoeboid movement (48). Gene expression to and activate FGFRs in neurons and tu-
profile comparison between melanoma cells mor cells, thereby participating in modulating

228 Bacac · Stamenkovic

 ANRV335-PM03-09 ARI 9 December 2007 17:30

integrin-mediated cell adhesion to and migra- Proteolysis, Invasion,

tion on the ECM (20). and Tumor-Host Interactions
Following several rounds of division within
Cell surface proteoglycan function
the epithelial compartment, generating a
changes. Transformation may also result in
carcinoma in situ, malignant cells disrupt the
the activation and differential glycosylation
basement membrane (BM), allowing them
of cell surface proteoglycans. One such
to come into direct contact with structural
proteoglycan relevant to tumor metastasis
and cellular components of the stroma. This
is CD44, the principal cell surface receptor
phase can be subdivided into several events
for hyaluronan. In addition to mediating cell
key for subsequent tumor dissemination,
attachment to hyaluronan-coated surfaces
including (a) expression by the tumor cells
and participating in hyaluronan metabolism,
of the proteolytic arsenal required to de-
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

CD44, by virtue of its structural polymor-

grade the BM; (b) interaction with stromal
phism and facultative decoration with a
fibroblasts and modification of their function
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

variety of glycosaminoglycans, can serve

to better serve the requirements of tumor
as a multipurpose cell surface scaffold that
cell growth, migration, and survival; (c)
orchestrates the assembly of complexes
recruitment of leukocytes that may amplify
comprising various classes of molecules (58).
the stromal reaction and further facilitate
Thus, CD44 associates with and promotes
tumor cell dissemination; (d ) angiogenesis;
oligomerization of the ErbB family of
and finally (e) intravasation (Figure 3). The
growth factor receptors (59) and Met (60).
stromal reaction to invading tumor cells is
It also localizes various MMPs and some of
variable, depending in part upon tumor cell
their substrates to the cell surface (61–63).
properties and in part upon the local stromal
Its cytoplasmic domain interacts with the
composition. Tumor cells typically produce
ezrin-radixin-moesin family of cytoskeletal
mediators that can initiate local stromal
interactors (64), including merlin, which is
cell activation, leading to ECM remodeling
implicated in tumor metastasis (65). The
and recruitment of additional stromal cell
mechanisms whereby CD44 promotes tumor
populations, which provides permissive
invasion and dissemination include enhance-
conditions for tumor growth. The combined
ment of cell migration (66), coordination of
proteolytic machinery of the tumor and acti-
proteolytic activity on the cell surface (62,
vated stromal cells degrades ECM proteins,
63), and enhancement of survival (62, 63, 67).
uncovering cryptic sites that may display pro-
Although it is expressed in most carcinomas,
migratory properties and releasing se-
CD44 promotes invasion and metastasis in
questered growth and survival factors,
a selective manner. By analogy to integrins,
including insulin-like growth factor-1, trans-
CD44 may be expressed in an inactive confor-
forming growth factor-β (TGF-β), PDGF,
mation, and its activation requires, at the very
vascular endothelial growth factor (VEGF),
least, partial desialylation of the extracellular
fibroblast growth factor, and hepatocyte
domain and a critical cell surface expression
growth factor, thereby augmenting their
density (68–71). CD44 appears to be consti-
bioavailability (73).
tutively active in fibroblasts and mesenchymal
Recent assessment of the expression pro-
progenitor cells, and to regulate migration
file of stroma associated with invasive can-
and adhesion of both cell types. Consistent
cer in a transgenic model of multistage car-
with this observation, as well as observations
cinogenesis of the prostate revealed a gene
in mouse model systems, sarcomas may
signature reminiscent of that associated with
retain CD44 function and display at least
wound healing (74). Importantly, genes that
partial dependence upon its expression for
were upregulated in the reactive stroma were
migration, invasion, and metastasis (72).

[Link] • Metastatic Cancer Cell 229

 ANRV335-PM03-09 ARI 9 December 2007 17:30

Angiogenesis
Recruitment,
proliferation,
VEGF-A
and activation
HPCs and EPCs
VEGFR1
Lymphangiogenesis
Tumor cell
VEGFR3 VEGF-C, VEGF-D
Blood vessel

Lymphatic vessel

Proliferation/
differentiation
Normal epithelia
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

TGF-β
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

BM EMT / invasion

ECM remodeling
Fibroblast Activated MMP-9, uPA,
Collagen fibroblast cathepsins
fibers CAF

Granulocyte Release of
sequestered
growth factors
Monocyte/TAM IGF-1, TGF-β, PDGF,
VEGF, bFGF, HGF/SF

SDF-1 EPC recruitment

Figure 3
Tumor-host interactions. Basement membrane (BM) degradation allows tumor cells to enter into contact
with stromal fibroblasts and alter their phenotype toward that of myofibroblasts (CAFs). Invading tumor
cells secrete numerous growth factors that stimulate angiogenesis, including VEGF-A, which binds to
VEGFR1 on endothelial hematopietic precursor cells (HPCs) and endothelial precursor cells (EPCs), as
well as on monocytes/tumor-associated macrophages (TAMs), resulting in their recruitment and
activation. Tumor cells also secrete VEGF-C and -D, which bind and activate VEGFR3 on lymphatic
endothelial cells and stimulate lymphangiogenesis. Activated CAFs, TAMs, and tumor cells secrete
numerous extracellular matrix (ECM)-degrading enzymes, including matrix metalloproteinases (MMPs),
cathepsins, and uPA, whose combined activity releases numerous ECM-sequestered growth factors that
further stimulate stromal fibroblast proliferation and tumor cell invasion.

found to have predictive value for both overall Disruption of the basement membrane.
and metastasis-free survival of prostate and The first barrier to invasion by carcinoma
breast carcinoma patients. Several of these cells is constituted by the BM. Conventional
genes were found to be upregulated in tumor- wisdom would have it that a broad range of
associated stromal remodeling in studies using MMPs can degrade the various components
different approaches to address tumor-host of the BM. However, most of the evidence
interactions (75, 76). Together these obser- supporting this view was derived from studies
vations strongly suggest that the stromal re- of tumor invasion of matrigel, which is
sponse to primary carcinoma growth may composed of denatured BM but does not
hold the key to subsequent development and recapitulate its native structure. Recent work
dissemination. has provided direct evidence that proteolytic

230 Bacac · Stamenkovic

 ANRV335-PM03-09 ARI 9 December 2007 17:30

degradation of the BM is mediated by trans- in invasion, MT-MMPs have been observed

membrane MMPs (MT-MMPs), including to colocalize with integrins to invadopodia
MT1-, MT2-, and MT3-MMP (MMP-14, (82). MMP-14 has also been shown to bind
-15, and -16, respectively), all three of which and cleave the extracellular domain of CD44,
are commonly expressed by cancer cells which helps detach tumor cells from the ECM
(77). Importantly, the ability to degrade and promotes migration (83, 84).
native BM has been shown to be restricted
to these three MMPs. Neither soluble nor
artificially membrane-tethered MMP-2 and Proteolytic events within the extracellular
MMP-9, which are commonly associated matrix. The evidence that MT-MMPs me-
with tumor invasion, were able to promote diate tumor cell disruption of the BM as well
cell penetration of the BM (77). The same as subsequent ECM invasion raises the ques-
tion as to what role secreted MMPs play in
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

held true for several other secreted MMPs,

providing the first clear evidence that native tumor metastasis, particularly because the ma-
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

BM degradation by tumor cells is dependent jority of them appear to be supplied by stro-

upon MT-MMP-mediated proteolysis (77). mal cells (79). Several lines of evidence suggest
MT-MMPs display collagenase activity that at least some of the secreted MMPs may
and are capable of degrading collagen IV, have a robust tumor-initiating effect (79, 85).
which is a major constitutent of the BM. However, secreted MMPs may also partici-
Their ability to degrade other collagens pate in tumor dissemination (79, 85). Secreted
and numerous other ECM components MMPs can interact with cell surface adhesion
suggests that MT-MMPs may be implicated receptors and proteoglycans, leading to co-
in events that occur beyond BM invasion. operation between adhesive and proteolytic
Consistent with this view, experiments using mechanisms (79, 85), by analogy to the well-
three-dimensional collagen gels showed that established functional relationship between
MT1-MMP expression provides tumor cells integrins and the urokinase receptor (86).
with a growth advantage in vitro and in vivo Hyaluronan-dependent association between
(78). The replicative advantage conferred by CD44 and MMP-9, which is cell-context de-
MT1-MMP requires pericellular ECM pro- pendent, has been shown to promote pro-
teolysis, as proliferation is abrogated in tumor teolytic activation of latent TGF-β relevant
cells suspended in protease-resistant collagen to both inflammation (87) and cancer (62).
gels. In the absence of proteolysis, tumor In a mouse model of mammary carcinoma
cells embedded in physiologically relevant dissemination, the functional CD44/MMP-
ECM matrices adopt a spherical config- 9/TGF-β complex promoted angiogenesis
uration and fail to display shape changes and survival of metastatic tumor cells (62, 67).
and cytoskeletal reorganization required for Similarly, the cell surface complex comprising
three-dimensional growth. These obser- CD44HSPG, pro–heparin-binding epider-
vations suggest that MT1-MMP regulates mal growth factor (HB-EGF), MMP-7, and
proliferation by controlling cell geometry ErbB4 promotes both normal and malignant
within the the three-dimensional ECM (78). cell survival by facilitating MMP-7-mediated
In addition to these essential roles in HB-EGF activation and its engagement of
the early steps of tumor metastasis, MT1- ErbB4 (Figure 2) (63). The sum of these ob-
MMP activity provides the principal source servations suggests that secreted MMPs teth-
of MMP-2 activation (79) and promotes an- ered to the surface of stromal or tumor cells
giogenesis (80) by degrading the fibrin ma- may indirectly participate in metastasis by ac-
trix that surrounds newly formed blood ves- tivating relevant growth factors, promoting
sels, facilitating endothelial cell penetration of angiogenesis, and probably further disrupting
tumor tissue (81). Consistent with their role structural ECM barriers to invading cells.

[Link] • Metastatic Cancer Cell 231

 ANRV335-PM03-09 ARI 9 December 2007 17:30

Interaction with fibroblasts and soluble in tumor cells resistant to its cytostatic ef-
regulators of tumor-host cross talk. Hav- fects and, as already discussed, may be a major
ing crossed the disrupted BM, tumor cells for player in the activation of normal fibroblasts
the first time find themselves in direct contact to display tumor-promoting functions (88).
with stromal fibroblasts. Tumor-derived and
degraded ECM-released growth factors, Migration. ECM remodeling and the pres-
including PDGF and TGF-β, alter the ence of activated stroma fibroblasts create
fibroblast phenotype to one reminiscent conditions favorable for tumor cell migra-
of myofibroblasts (Figure 3) (88). These tion. Although the molecular mechanisms
tumor-conditioned stromal fibroblasts are that underlie migration are reasonably well
referred to as carcinoma-associated fibrob- understood (see above), the question as to
lasts (CAFs), and their contribution to tumor how tumor cells actually migrate in a three-
dimensional structure has only recently been
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

initiation and growth is now well established

(89–91). However, the mechanisms whereby addressed. Real-time imaging of invading tu-
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

CAFs promote tumor progression are only mor cells in three-dimensional collagen gels
beginning to emerge. CAFs are an abundant has given rise to some surprising observations.
source of proteolytic enzymes, including Single tumor cells that had detached from
MMPs and cathepsins (74, 92, 93), which the original tumor mass were shown to dis-
may stimulate tumor cell growth and inva- play two possible migration patterns. Mes-
sion at both primary and secondary sites. enchymal cells, or malignant epithelial cells
Increased deposition of collagen I and III that had undergone EMT, migrate along a
as well as de novo expression of tenascin C classical scheme that includes protrusion of
may provide additional signals that facilitate the leading edge, formation of focal contacts
tumor invasion and metastasis; by secreting with the ECM, recruitment of surface pro-
chemokines such as monocyte chemotactic teases to ECM contacts resulting in localized
protein-1 and cytokines such as interleukin-1 proteolysis, Rho-mediated contraction of ac-
(IL-1), CAFs participate in regulating the tomyosin leading to cell contraction, and fi-
inflammatory response to tumor invasion. In nally detachment of the trailing edge. How-
addition, CAF-derived stromal-cell-derived ever, other malignant cells display amoeboid
factor-1 (SDF-1) has been shown to mediate movement through collagen gels. This type
bone marrow–derived endothelial cell pre- of movement relies on cell deformability and
cursor recruitment and to directly increase relatively weak interactions with the ECM.
tumor cell proliferation (94). Movement is generated by cortical filamen-
Among the soluble factors implicated in tous actin, whereas focal contacts, stress fibers,
coordinating tumor-host cross talk, TGF-β and localized proteolysis at cell-ECM con-
plays a leading role. Although TGF-β in- tacts are lacking (97). Cells displaying amoe-
hibits proliferation of normal epithelial cells boid movement typically circumscribe but do
and carcinoma cells at early stages of progres- not degrade collagen fibers (97). Although
sion, it stimulates fibroblast growth and ECM this type of movement characterizes lymphoid
secretion and promotes late-stage carcinoma cells, carcinoma cells can adopt it as well
invasiveness (88). Both transgenic models (97).
(95) and experimental metastasis assays have Tumor cells can also migrate in groups
shown that TGF-β can enhance dissemina- or aggregates (Figure 2). Here again two
tion of at least some carcinomas (67). Accord- types of movement have been described. One
ingly, soluble TGF-receptor fusion proteins consists of chain migration, where cells follow
were observed to reduce metastatic growth each other in single file and form a chain-like
of tumor cells injected into immunocompro- image. It is displayed by neural crest cells
mised mice (96). TGF-β can induce EMT (98) and normal myoblasts (99), but can also

232 Bacac · Stamenkovic

 ANRV335-PM03-09 ARI 9 December 2007 17:30

be observed in melanomas. Lobular invasive to CAFs, tumor cells themselves may recruit
breast carcinoma as well as ovarian carcinoma hematopoietic precursors and leukocytes.
cells often display a chain-type arrangement. Myeloid precursors as well as monocytes
The second type of group movement of express receptors for several growth fac-
malignant cells is referred to as collective tors/cytokines secreted by tumor cells, includ-
migration and invasion and mimics a well- ing VEGFR1, which binds tumor-derived
described phenomenon that occurs during VEGF-A and placental growth factor PIGF,
development. Following neural tube closure, facilitating their recruitment to the tumor mi-
cells in the blastoderm or ectoderm migrate in croenvironment where they can differentiate
sheets (100), and similar migration is observed into TAMs and promote tumor growth and
during branching morphogenesis of mam- dissemination. In a model of skin carcinogen-
mary glands and ducts (101). Malignant cells esis in the mouse, TAM-derived MMP-9 was
shown to play a key role in promoting tu-
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

also have the ability to aggregate and migrate

as a functional unit (97, 102). In contrast to mor angiogenesis (107). Macrophages were
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

individual migrating cells, cell-cell adhesion also found to play an essential role in tumor
that occurs in cell aggregates leads to a specific intravasation (108), whereas macrophage de-
form of cortical actin filament assembly along pletion has been observed to repress late-stage
cell junctions that allows the formation of a tumor progression and metastasis but not pri-
larger-sized, multicellular contractile body mary tumor growth (109, 110). Together with
(97, 102). The cells at the front of the body are CAFs, TAMs and possibly other leukocytes
designated path-finding cells and are the ones may supply tumor cells with proinvasive fac-
that generate traction via pseudopod activity tors that facilitate metastasis (111).
and expression of clusters of integrins and
MMPs within the corresponding invadopodia Angiogenesis. One of the prerequisites for
(102, 103). Cells in the inner and trailing metastatic tumor growth is the induction
regions are passively pulled along. Recent of angiogenesis (112, 113). Angiogenesis is
work has shown that the type-1 mucin-like frequently induced by transforming events
cell surface receptor podoplanin is upregu- that promote tumor progression and aug-
lated at the outer edge of growing tumors ment expression of angiogenic factors. Thus,
and may promote collective tumor cell mi- VEGF-A expression is induced by the MAPK
gration (104). Collective movement has been signaling pathway and hypoxia that accom-
observed in several types of carcinoma (97). panies rapid primary tumor growth. Hypoxia
induces and stabilizes expression of hypoxia-
Inflammation. Recruitment of leukocytes inducible-factor-1α, which drives VEGF-A
may have different effects in different tu- transcription. VEGF-A is among growth
mor types. Thus, accumulation of myeloid factors deposited in the ECM and whose
cells, including neutrophils monocytes and bioavailability, in addition to that of other
macrophages, is associated with indolent evo- proangiogenic factors including basic fibrob-
lution in some cancers, but bears a much more last growth factor and TGF-β, is increased as
somber prognosis in others (105). Infiltrat- a result of tissue remodeling (114).
ing tumor-associated macrophages (TAMs) Studies in cancer patients as well as in
present antigen and secrete cytokines that mouse models provide evidence that lym-
support an adaptive antitumor immune re- phangiogenesis, defined as the outgrowth
sponse. On the other hand, if tumor cells re- of new lymphatic vessels from preexisting
sist the immune reponse, which most solid ones, promotes metastasis to regional drain-
tumors appear to do succesfully, the TAM- ing lymph nodes of a tumor (115, 116). Lym-
derived chemokine/cytokine repertoire may phangiogenesis is induced by tumor-derived
promote tumor progression (106). In addition VEGF-C and -D members of the VEGF

[Link] • Metastatic Cancer Cell 233

 ANRV335-PM03-09 ARI 9 December 2007 17:30

family, which bind to VEGFR3 on the sur- Tumor Dissemination

face of lymphatic endothelial cells. VEGF-C
Survival in lymph and blood and interac-
expression is regulated, at least in part, by in-
tions with endothelium. Once tumor cells
flammatory responses triggered by IL-1β and
have penetrated the blood circulation, they
tumor necrosis factor α signaling. VEGF-D
are exposed to shear stress and to interac-
is the product of an immediate early, Fos-
tions with leukocytes that may lead to their
regulated gene, and its expression may there-
destruction. It would appear, however, that tu-
fore be regulated by oncogenic signaling path-
mor cells are capable of resisting shear stress,
ways (115, 116).
possibly aided by platelets and leukocytes, and
that they may rely on some of the mechanisms
Intravasation. It would seem logical that used by leukocytes to adhere to endothelium.
an increased lymphatic and vascular network Transformed cells often express an altered
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

may facilitate penetration of the vascular lu- glycosyltranferase repertoire with respect to
men by invading tumor cells. Carcinomas ini- normal counterparts. Glycosyl- and sialyl-
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

tially form metastasis in local lymph nodes and transferases expressed in many carcinoma
only at a later stage in other organs. There types decorate cell surface receptors with
is an ongoing and as yet unresolved debate oligosaccharide structures that correspond to
as to whether distal hematogenous metastases ligands of selectins, a C-type lectin class of
in carcinomas develop as a result of vascular cell surface adhesion molecules that regulate
invasion and penetration at the primary tu- leukocyte endothelial interactions and leuko-
mor site or whether they are derived from cyte trafficking (118). P-selectin/CD62P is
cells that have colonized local lymph nodes. expressed on the surface of activated platelets
In the first case, tumor cells would need to and endothelial cells, and E-selectin/CD62E
degrade vascular BM and irrupt into the cir- is predominantly induced in activated en-
culation. An argument favoring this possibil- dothelium. L-selectin/CD62L is expressed on
ity is that vascular invasion and penetration the surface of a broad range of leukocyte
by tumor cells is observed by microscopic ex- subpopulations.
amination of tissue sections. This is further A variety of potentially metastatic tu-
supported by experimental approaches using mor cells, particularly those that are mucin
an in vivo assay (117), where intravasation de- rich—often derived from colon—express se-
pended upon MMP-9 activity and constituted lectin ligands (119–121). Circulating tumor
a rate-limiting step in metastasis. Tumor cell cells that express selectin ligands can become
dissemination from lymph nodes could occur coated with platelets and leukocytes, creat-
by migration of the cells to efferent lymph ing a microembolus that may obstruct cap-
vessels and transport to the vena cava from illaries of various organs (119); they may also
where hematogenous spread would be possi- adhere to activated endothelial cells. In an ex-
ble. Currently, it would seem plausible that perimental metastasis model, B16 melanoma
both mechanisms might be operational and cells engineered to synthesize oligosaccha-
that the relative ease of lymph vessel inva- rides that constitute selectin ligands displayed
sion and penetration might explain that lymph an altered pattern of organ colonization
nodes are usually the first metastatic site in compared with their parental counterparts
carcinomas. It is also possible, however, that (122).
colonization of both lymphoid and nonlym- Platelets and leukocytes can also inter-
phoid organs occurs within a comparable time act with tumor cells via αvβ3-dependent
frame, but owing to local conditions tumor adhesion (123). Experimental models sug-
growth proceeds more rapidly in lymphoid gest that tumor-cell–platelet/leukocyte inter-
tissues. actions may favor tumor metastasis (120, 121,

234 Bacac · Stamenkovic

 ANRV335-PM03-09 ARI 9 December 2007 17:30

124, 125), but whether such a mechanism is sive candidate mechanisms proposed thus
important in human cancer metastasis is sub- far are chemokine-receptor-mediated chemo-
ject to debate. taxis, the establishment of a metastatic niche,
Experimental evidence suggests that inte- and a tumor cell genetic program that facili-
grins and immunoglobulin superfamily adhe- tates adaptation to a particular microenviron-
sion molecules are also implicated in tumor ment.
cell adhesion to endothelium. The α4β1 in-
tegrin, associated primarily with lymphocytes,
Chemokine-receptor-mediated chemo-
is expressed on a variety of tumor cell types,
taxis. Chemokines are believed to cooperate
and its ligand VCAM-1 was shown to support
with adhesion receptors in determining where
melanoma cell adhesion to endothelial cells
tumor cells arrest and extravasate. Tumor
(126).
cells can express a variety of chemokine
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

Leukocyte adhesion to activated en-

receptors, including CXCR4, that serves as
dothelium typically relies on three sets of
a receptor for CXCL12/SDF. Secretion of
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

events: selectin-mediated low-affinity inter-

SDF by host tissue stromal fibroblasts is
actions responsible for leukocyte rolling on
suggested to promote chemotaxis of tumor
the endothelium; endothelial cell-derived
cells expressing CXCR4 and to determine, at
chemokine-mediated leukocyte activation
least in part, the localization of metastases of
that changes leukocyte β2 integrin confor-
certain tumor types (128).
mation from low to high affinity; and high-
affinity interaction between leukocyte β2
integrins and endothelial ICAM-1, which is Preparation of the metastatic microenvi-
required for the leukocyte arrest that precedes ronment. Recent studies have suggested that
and is necessary for diapedesis/extravasation by virtue of their cytokine and chemokine
(118). Despite being larger and having dif- repertoire, tumors may have the ability to pre-
ferent morphological properties than circu- pare the microenvironment of distant organs
lating leukocytes, carcinoma cells can make to receive disseminating cells and allow their
use of at least some of the adhesive mecha- proliferation. Tumor and associated stromal
nisms that govern leukocyte trafficking to in- cell–derived chemokines can recruit endothe-
teract with vascular endothelium. However, lial and hematopoietic progenitor cells (EPCs
it remains to be demonstrated whether these and HPCs, respectively) to the relevant or-
adhesive events are relevant to human cancer gans prior to tumor cell arrival, which, to-
metastasis. gether with tumor cell–derived deposition of
fibronectin, appear to precondition the local
Random arrest or programmed organ- microenvironment (129, 130). Targeted inhi-
specific homing. Intravital microscopy im- bition of these cells using anti-VEGFR1 neu-
ages argue that tumor cells tend to obstruct tralizing antibodies suggests that the proposed
capillaries, particularly if they are aggregated preconditioning is necessary for metastatic
or bound to leukocytes and platelets. Follow- spread. The mechanisms that govern HPC
ing arrest in the capillary bed, they proliferate recruitment to potential metastatic sites are
locally and disrupt the capillary wall, whereby still unclear, as is the manner in which HPCs
they penetrate the local parenchyma (127). render any given site permissive for metastatic
It would therefore seem that tumor cell ar- tumor growth. It is possible that recruitment
rest in capillaries and subsequent extravasa- is initiated by the very first tumor cells that
tion are primarily mechanical processes that arrive at a distant site and that local HPC
could occur in all organs. Organ-specific tu- activity then alters the local microenviron-
mor cell homing would then require spe- ment, amplifying tumor cell recruitment and
cific mechanims. Three nonmutually exclu- allowing subsequent division. However, the

[Link] • Metastatic Cancer Cell 235

 ANRV335-PM03-09 ARI 9 December 2007 17:30

existence of metastatic niches remains to be membrane or secreted molecules that were in

confirmed. some way relevant to the bone microenviron-
ment. Expression of any one of these genes in
Gene expression patterns that determine parental MDA-MB231 cells failed to augment
organ-specific homing of tumor cells. Pi- their intrinsic metastatic potential to bone.
oneering work using an experimental metasta- However, the combination of three to four of
sis approach has shown that repeated cycles of the genes augmented the metastatic activity
in vivo passage of tumor cells that develop few of parental cells to levels comparable to those
lung tumors following tail vein injection in displayed by the most aggressive cell lines ex-
mice result in selection of cells that preferen- pressing the entire bone metastasis gene set
tially develop lung metastasis (131). As men- (133). These observations strongly suggest
tioned above, the gene expression profile anal- that cooperation among several genes that en-
code proteins with complementary functions
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

ysis of B16 melanoma cells selected for lung

colonization revealed upregulation of numer- underlies the metastatic phenotype. Cells ex-
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

ous genes including RhoC (49), and forced pressing the required genes were identified in
expression if RhoC in parental B16 cells re- the initiating tumor cell line, suggesting that
sulted in the preferential homing to the lung breast cancer cells that display a gene expres-
(49). sion signature associated with bone or lung
To address this issue in human tumors, metastatic proclivity exist in the parental tu-
one study focused on the human breast mor cell population.
carcinoma MDA-MB231 cell line derived
from the pleural effusion of a patient with
widespread metastasis. Repeated rounds of in Establishment of new colonies. Extrava-
vivo passage and selection allowed isolation of sation was long thought to be a rate-
different sublines of MDA-MB231 cells that limiting step in metastasis. However, RAS-
predictably formed tumors in given organs transformed NIH3T3 cells and parental
following intravenous injection (132). Gene counterparts were found to extravasate into
expression profiling of these sublines shows the liver at a comparable rate following in-
that they express a specific set of genes that jection into the portal vein, whereas only the
correlates with general metastatic proclivity. RAS-transformed cells could form metastatic
However, the selected cell lines expressed growth (134). These observations provide
additional gene signatures that correlated convincing evidence that the rate-limiting
with the organ to which they metastasized step is not extravasation, but rather the ability
(132). Thus, a set of 54 genes was identi- of the cells to establish themselves in the new
fied that distinguished cell lines displaying host tissue microenvironment.
lung tropism (132). Although none of the Once tumor cells have penetrated the
identified genes alone could recapitulate the parenchyma of an organ other than the one
metastatic phenotype of selected cell lines, where they originated, they must create a mi-
combinations of the genes could induce the croenvironment conducive to their survival
poorly metastatic parental MDA-MB231 and proliferation. The situation is analogous
cells to colonize the organ from which the to that of invading cells of the primary tu-
metastatic cell variants were retrieved. mor penetrating its BM and entering into di-
Prior to this study, the same cell line had rect physical contact for the first time with the
been used to identify genes that may be in- stroma. However, the stromal composition of
volved in promoting bone metastasis (133). the secondary site may differ from that sur-
Most of the highly overexpressed genes in rounding the primary tumor, and the ability of
cell lines derived from and selected for their the tumor cells to subvert the local microenvi-
ability to induce bone metastases encoded cell ronment will most likely determine their fate.

236 Bacac · Stamenkovic

 ANRV335-PM03-09 ARI 9 December 2007 17:30

Dormancy. Metastases from some human documented correlation between primary tu-
cancers occur as many as 20 years follow- mor size and risk of metastasis. However, the
ing removal of the primary tumor. These association of clinical features such as tumor
metastatic lesions are believed to be dormant grade with metastatic proclivity and the oc-
for an extended period of time and to become currence of bone micrometastases early in the
proliferative once local conditions become ap- evolution of cancer are inconsistent with a
propriately permissive. strictly stochastic model.
Dormancy is a concealed state, and as such The use of DNA microarray studies to
does not readily lend itself to direct study. Ex- identify transcriptional signatures that may
perimental work, however, has provided ev- distinguish metastatic from primary tumor
idence for the existence of micrometastases growth has further challenged this view, sug-
that fail to induce angiogenesis and in which gesting that relevant signatures may be de-
tected in early-stage primary tumors that are
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

cell proliferation is balanced by cell death be-

cause of inadequate blood supply (135). These destined to metastasize. This second view
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

small metastatic lesions may therefore consti- is supported by observations from several
tute one source of tumor dormancy. Another independent and distinctly designed studies
possible source of tumor dormancy are iso- (Figure 4). The first of these studies showed
lated tumor cells that arrive at secondary sites that the clinical outcome of breast cancer pa-
where they may persist for long periods of tients can be predicted by a poor progno-
time without being able to divide (127). Inter- sis gene expression signature present in the
estingly, recovery of isolated dormant mam- majority of early-stage primary tumors (137,
mary carcinoma cells from liver tissue showed 138). This gave rise to the notion that certain
that they retain tumorigenicity when injected tumors may have the properties required for
into immunocompromised mice (127). These metastasis from the very beginning. The size
observations suggest that the host tissue may of the primary tumor would then be irrele-
provide permissive or restrictive cues that de- vant in terms of the risk of developing metas-
termine whether metastatic cells may prolif- tases, and even small tumors may be expected
erate and generate secondary tumors. Recent to contain cells with metastatic potential.
evidence suggests that changes in the ECM In a second study, comparison of the gene
that lead to a shift in the equilibrium between expression profile of metastatic adenocarci-
natural stimulators and inhibitors of angio- noma lesions to unmatched primary carcino-
genesis, many of which are ECM degradation mas revealed a 17-gene expression signature
products, may allow dormant metastatic le- that distinguished primary from metastatic tu-
sions to induce an angiogenic switch and de- mors (139). Subsequent analysis revealed that
velop into full-blown secondary tumors (136). numerous early-stage primary solid tumors of
diverse histotypes harbored the same gene ex-
pression signature, suggesting that these tu-
EARLY EMERGENCE OR mors were most likely associated with metas-
LATE-STAGE SELECTION tases and poor clinical outcome.
OF METASTATIC CELLS? Because the bone marrow is a major hom-
Until recently, the prevailing reasoning was ing site for breast cancer, a third study un-
that metastasis arises from rare tumor cells dertook the task of analyzing genomic differ-
that emerge relatively late in tumor progres- ences between single bone marrow–derived
sion. The genetic makeup of these cells was micrometastatic cells and the primary tumor
believed to be the consequence of stochastic by comparative genomic hybridization (140).
accumulation of mutations that provide them Single viable disseminated breast cancer cells
with all of the properties required for dissem- had an abundance of chromosomal copy num-
ination. This view is supported by the well- ber changes in their genome, with significant

[Link] • Metastatic Cancer Cell 237

 ANRV335-PM03-09 ARI 9 December 2007 17:30

Primary tumors vs. metastatic nodules

17-gene
signature

Early-stage primary
invasive breast tumors

70-gene
signature
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

No mets

Mets
Good
Poor

Good Lung Breast Prostate

Breast
Poor The metastatic potential is encoded in the
bulk of primary tumors
(Ramaswamy et al. 2003)
Metastatic proclivity is present in early tumors
(van’t Veer et al. 2002)
Bone metastases vs. primary tumor
Metastasis
Tumor microenvironment is an
early event CGH
in tumor
progression Short -term
culture

Dissemination occurs early during tumorigenesis

(Gangnus et al. 2004)
In situ Invasive

Stem cell–ness and cancer

LCM stroma
Metastases vs. 11-gene
Mouse-human Bmi +/+ primary signature
translational tumor
Wound- genomics Bmi –/–
Good
response
genes Poor Good
Prostate Breast Poor
Lung Breast Prostate
Wound response signature in primary tumors
predicts increased risk of metastasis Tumors with stem cell– like signatures are
and poor outcome likely to have poor prognosis
(Bacac et al. 2006, Chang et al. 2005) (Glinsky et al. 2005)

238 Bacac · Stamenkovic

 ANRV335-PM03-09 ARI 9 December 2007 17:30

intercellular heterogeneity—but more impor- tion of stem cell–resembling malignant cells

tantly, they displayed numerous differences in that constitute the driving force of the tumor
comparison to the matching primary tumors. and are essential for progression and metasta-
These observations led to the conclusion that sis. In support of this view, eight out of nine
metastasis is an early event in malignant tu- tumor samples that served as a source for the
mors and that disseminated cells evolve inde- identification of breast cancer–initiating cells
pendently of the primary tumor. were derived from metastatic lesions (146).
A fourth study addressed the possibility These cells may share attributes of normal
that cancers with poor prognosis and high stem cells that are relevant to their natural be-
metastatic proclivity may display some prop- havior and resistance to chemotherapy. Nor-
erties that characterize normal stem cells. mal stem cells express multidrug-resistance
Comparison of primary and metastatic mouse genes, which, coupled to their slow prolifer-
ative rate, may play a key role in their ability
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

prostate cancers with normal stem cells that

had retained or lost their self-renewing po- to withstand cytotoxic drugs and repopulate
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

tential identified a common 11-gene signature tissues that had been depleted of their nor-
that was then used to probe human cancers mal cells by chemotherapy. Metastatic cancer
(141–143). Expression of this gene signature lesions are notorious for resistance to con-
in 11 distinct types of primary cancers was ventional chemotherapy, at least in part be-
consistently a powerful predictor of a short cause of multidrug-resistance gene expres-
interval to disease recurrence, distant metas- sion. Whether or not the stem cell connection
tasis, and death following therapy. These ob- turns out to be correct, the lack of respon-
servations are consistent with the possibility siveness of metastatic lesions to conventional
that cancer cells with high metastatic procliv- therapy and the increasingly clear evidence
ity display features that are at least in part that the stroma is implicated in tumor metas-
reminiscent of those of normal stem cells. In- tasis warrant a closer look at the molecular
terestingly, a fraction of each of the gene ex- mechanisms that govern tumor-host interac-
pression signatures identified in these stud- tions at both primary and metastatic sites.
ies comprised stromal cell transcripts, some
of which were part of the recently described
predictive stromal gene set (74). This further SUMMARY AND FUTURE
underscores the notion that the stroma may DIRECTIONS
actively participate in determining the ability Recent studies strongly support the view that
of cancer to metastasize (74, 75, 144, 145). the capacity of a tumor to disseminate is ac-
The emerging, and still controversial, con- quired at early steps during the multistep pro-
cept of cancer stem cells suggests that among cess of tumorigenesis. They also suggest that
the heterogeneous populations of cells that specific genes uniquely responsible for cancer
compose primary tumors is a small popula- cell dissemination and metastatic growth are

←−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−
Figure 4
Overview of recent microarray and CGH studies that led to the notion that metastasis is an early event in
tumorigenesis. Validation of signatures identified by Ramaswamy et al. (139) on patient cohorts with
early-stage tumors showed that metastatic proclivity is present early in tumor evolution. Studies
performed by Gangnus et al. (140) on bone micrometastasis confirmed this notion. More recently,
Glinsky et al. (141) proposed that tumors with a stem cell expression signature are likely to have a poor
prognosis. Most of these signatures contained stroma-associated genes, consistent with the notion that
the tumor microenvironment plays an important role in dissemination. This was also suggested by recent
studies focusing on stromal reactions to tumor invasion and injury that showed that a wound-response
signature in primary tumors predicts increased risk of metastasis.

[Link] • Metastatic Cancer Cell 239

 ANRV335-PM03-09 ARI 9 December 2007 17:30

unlikely to be discovered. Rather, the com- to meet with success are those that simul-
bined effect of oncogene signaling and tu- taneously target several mechanisms upon
mor suppressor gene loss in the appropriate which tumor cell dissemination depends, in-
cellular environment is likely to determine cluding angiogenesis, proteolysis, and growth
whether a cancer cell has the potential to col- factor signaling. Much more challenging
onize distant organs. Success of secondary tu- is the prospect of reversing already estab-
mor growth is then determined by the nature lished metastatic growth. The formidable
of the host response and the tumor cells’ abil- ability of metastatic lesions to evade cyto-
ity or inability to subvert it. toxic drug effects, at least in part due to
The realization that early cancer har- multidrug resistance family gene expression,
bors metastatic potential should warrant pre- indicates that we need to look elsewhere
ventive treatment of metastatic disease. Our for therapeutic solutions. As metastatic tu-
mors require local stroma support for growth,
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

extensive understanding of the mechanisms

whereby cancer cells spread should allow the identifying targetable tumor-host interac-
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

development of strategies that can effectively tion mechanisms appears to be an appealing

block tumor dissemination. Approaches likely pursuit.

DISCLOSURE STATEMENT
The authors are not aware of any biases that might be perceived as affecting the objectivity of
this review.

ACKNOWLEDGMENTS
This work was supported by the Fonds de la Recherche Scientifique grant number 3100A0-
105833 and by the National Center of Competence in Research Molecular Oncology.

LITERATURE CITED
1. Perez-Moreno M, Jamora C, Fuchs E. 2003. Sticky business: orchestrating cellular signals
at adherens junctions. Cell 112:535–48
2. Conacci-Sorrell M, Zhurinsky J, Ben-Ze’ev A. 2002. The cadherin-catenin adhesion
system in signaling and cancer. J. Clin. Invest. 109:987–91
3. Wheelock MJ, Johnson KR. 2003. Cadherins as modulators of cellular phenotype. Annu.
Rev. Cell Dev. Biol. 19:207–35
4. Vleminckx K, Vakaet LJ, Mareel M, Fiers W, van Roy F. 1991. Genetic manipulation of
E-cadherin expression by epithelial tumor cells reveals an invasion suppressor role. Cell
66:107–19
5. Perl AK, Wilgenbus P, Dahl U, Semb H, Christofori G. 1998. A causal role for E-cadherin
in the transition from adenoma to carcinoma. Nature 392:190–93
6. Graff JR, Gabrielson E, Fujii H, Baylin SB, Herman JG. 2000. Methylation patterns of
the E-cadherin 5 CpG island are unstable and reflect the dynamic, heterogeneous loss
of E-cadherin expression during metastatic progression. J. Biol. Chem. 275:2727–32
7. Nass SJ, Herman JG, Gabrielson E, Iversen PW, Parl FF, et al. 2000. Aberrant methy-
lation of the estrogen receptor and E-cadherin 5 CpG islands increases with malignant
progression in human breast cancer. Cancer Res. 60:4346–48
8. Bolos V, Peinado H, Perez-Moreno MA, Fraga MF, Esteller M, Cano A. 2003. The
transcription factor Slug represses E-cadherin expression and induces epithelial to mes-
enchymal transitions: a comparison with Snail and E47 repressors. J. Cell Sci. 116:499–511

240 Bacac · Stamenkovic

 ANRV335-PM03-09 ARI 9 December 2007 17:30

9. Cano A, Perez-Moreno MA, Rodrigo I, Locascio A, Blanco MJ, et al. 2000. The transcrip-
tion factor Snail controls epithelial-mesenchymal transitions by repressing E-cadherin
expression. Nat. Cell Biol. 2:76–83
10. Peinado H, Ballestar E, Esteller M, Cano A. 2004. Snail mediates E-cadherin repression
by the recruitment of the Sin3A/histone deacetylase 1 (HDAC1)/HDAC2 complex. Mol.
Cell. Biol. 24:306–19
11. Perez-Moreno MA, Locascio A, Rodrigo I, Dhondt G, Portillo F, et al. 2001. A new
role for E12/E47 in the repression of E-cadherin expression and epithelial-mesenchymal
transitions. J. Biol. Chem. 276:27424–31
12. Yang J, Mani SA, Donaher JL, Ramaswamy S, Itzykson RA, et al. 2004. Twist, a master
regulator of morphogenesis, plays an essential role in tumor metastasis. Cell 117:927–39
13. Fujita Y, Krause G, Scheffner M, Zechner D, Leddy HE, et al. 2002. Hakai, a c-Cbl-like
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

protein, ubiquitinates and induces endocytosis of the E-cadherin complex. Nat. Cell Biol.
4:222–31
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

14. Avizienyte E, Frame MC. 2005. Src and FAK signalling controls adhesion fate and the
epithelial-to-mesenchymal transition. Curr. Opin. Cell Biol. 17:542–47
15. Avizienyte E, Wyke AW, Jones RJ, McLean GW, Westhoff MA, et al. 2002. Src-induced
de-regulation of E-cadherin in colon cancer cells requires integrin signalling. Nat. Cell
Biol. 4:632–38
16. Novak A, Hsu SC, Leung-Hagesteijn C, Radeva G, Papkoff J, et al. 1998. Cell adhesion
and the integrin-linked kinase regulate the LEF-1 and β-catenin signaling pathways. Proc.
Natl. Acad. Sci. USA 95:4374–79
17. Tan C, Costello P, Sanghera J, Dominguez D, Baulida J, et al. 2001. Inhibition of integrin
linked kinase (ILK) suppresses β-catenin-Lef/Tcf-dependent transcription and expres-
sion of the E-cadherin repressor, Snail, in APC−/− human colon carcinoma cells. Oncogene
20:133–40
18. Noe V, Fingleton B, Jacobs K, Crawford HC, Vermeulen S, et al. 2001. Release of
an invasion promoter E-cadherin fragment by matrilysin and stromelysin-1. J. Cell Sci.
114:111–18
19. Cavallaro U, Christofori G. 2004. Cell adhesion and signalling by cadherins and Ig-CAMs
in cancer. Nat. Rev. Cancer 4:118–32
20. Cavallaro U, Niedermeyer J, Fuxa M, Christofori G. 2001. N-CAM modulates tumour-
cell adhesion to matrix by inducing FGF-receptor signalling. Nat. Cell Biol. 3:650–57
21. Marambaud P, Wen PH, Dutt A, Shioi J, Takashima A, et al. 2003. A CBP binding
transcriptional repressor produced by the PS1/ε-cleavage of N-cadherin is inhibited by
PS1 FAD mutations. Cell 114:635–45
22. Hynes RO. 2002. Integrins: Bidirectional, allosteric signaling machines. Cell 110:673–87
23. Giancotti FG, Ruoslahti E. 1999. Integrin signaling. Science 285:1028–32
24. Ivaska J, Nissinen L, Immonen N, Eriksson JE, Kahari VM, Heino J. 2002. Integrin
α2β1 promotes activation of protein phosphatase 2A and dephosphorylation of Akt and
glycogen synthase kinase 3 β. Mol. Cell. Biol. 22:1352–59
25. Ivaska J, Reunanen H, Westermarck J, Koivisto L, Kahari VM, Heino J. 1999. Inte-
grin α2β1 mediates isoform-specific activation of p38 and upregulation of collagen gene
transcription by a mechanism involving the α2 cytoplasmic tail. J. Cell Biol. 147:401–16
26. Owens DM, Watt FM. 2001. Influence of β 1 integrins on epidermal squamous cell
carcinoma formation in a transgenic mouse model: α3β1, but not α2β1, suppresses
malignant conversion. Cancer Res. 61:5248–54

[Link] • Metastatic Cancer Cell 241

 ANRV335-PM03-09 ARI 9 December 2007 17:30

27. Zutter MM, Santoro SA, Staatz WD, Tsung YL. 1995. Re-expression of the α2β1 integrin
abrogates the malignant phenotype of breast carcinoma cells. Proc. Natl. Acad. Sci. USA
92:7411–15
28. Chan BM, Matsuura N, Takada Y, Zetter BR, Hemler ME. 1991. In vitro and in vivo
consequences of VLA-2 expression on rhabdomyosarcoma cells. Science 251:1600–2
29. Wang H, Fu W, Im JH, Zhou Z, Santoro SA, et al. 2004. Tumor cell α3β1 integrin and
vascular laminin-5 mediate pulmonary arrest and metastasis. J. Cell Biol. 164:935–41
30. Gladson CL, Cheresh DA. 1991. Glioblastoma expression of vitronectin and the αvβ3
integrin. Adhesion mechanism for transformed glial cells. J. Clin. Invest. 88:1924–32
31. Mercurio AM, Rabinovitz I. 2001. Towards a mechanistic understanding of tumor
invasion—lessons from the α6β4 integrin. Semin. Cancer Biol. 11:129–41
32. Mariotti A, Kedeshian PA, Dans M, Curatola AM, Gagnoux-Palacios L, Giancotti FG.
2001. EGF-R signaling through Fyn kinase disrupts the function of integrin α6β4 at
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

hemidesmosomes: role in epithelial cell migration and carcinoma invasion. J. Cell Biol.
155:447–58
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

33. Trusolino L, Bertotti A, Comoglio PM. 2001. A signaling adapter function for α6β4
integrin in the control of HGF-dependent invasive growth. Cell 107:643–54
34. Schneller M, Vuori K, Ruoslahti E. 1997. αvβ3 integrin associates with activated in-
sulin and PDGFβ receptors and potentiates the biological activity of PDGF. EMBO J.
16:5600–7
35. Hannigan G, Troussard AA, Dedhar S. 2005. Integrin-linked kinase: a cancer therapeutic
target unique among its ILK. Nat. Rev. Cancer 5:51–63
36. Schlaepfer DD, Mitra SK. 2004. Multiple connections link FAK to cell motility and
invasion. Curr. Opin. Genet. Dev. 14:92–101
37. Mitra SK, Schlaepfer DD. 2006. Integrin-regulated FAK-Src signaling in normal and
cancer cells. Curr. Opin. Cell Biol. 18:516–23
38. Sieg DJ, Hauck CR, Ilic D, Klingbeil CK, Schaefer E, et al. 2000. FAK integrates growth-
factor and integrin signals to promote cell migration. Nat. Cell Biol. 2:249–56
39. Zhai J, Lin H, Nie Z, Wu J, Canete-Soler R, et al. 2003. Direct interaction of focal
adhesion kinase with p190RhoGEF. J. Biol. Chem. 278:24865–73
40. Carragher NO, Westhoff MA, Fincham VJ, Schaller MD, Frame MC. 2003. A novel role
for FAK as a protease-targeting adaptor protein: regulation by p42 ERK and Src. Curr.
Biol. 13:1442–50
41. Klemke RL, Cai S, Giannini AL, Gallagher PJ, de Lanerolle P, Cheresh DA. 1997.
Regulation of cell motility by mitogen-activated protein kinase. J. Cell Biol. 137:481–92
42. Huang C, Rajfur Z, Borchers C, Schaller MD, Jacobson K. 2003. JNK phosphorylates
paxillin and regulates cell migration. Nature 424:219–23
43. Keely PJ, Westwick JK, Whitehead IP, Der CJ, Parise LV. 1997. Cdc42 and Rac1 induce
integrin-mediated cell motility and invasiveness through PI(3)K. Nature 390:632–36
44. Raftopoulou M, Hall A. 2004. Cell migration: Rho GTPases lead the way. Dev. Biol.
265:23–32
45. Pollard TD, Borisy GG. 2003. Cellular motility driven by assembly and disassembly of
actin filaments. Cell 112:453–65
46. Watanabe N, Kato T, Fujita A, Ishizaki T, Narumiya S. 1999. Cooperation between
mDia1 and ROCK in Rho-induced actin reorganization. Nat. Cell Biol. 1:136–43
47. Kimura K, Ito M, Amano M, Chihara K, Fukata Y, et al. 1996. Regulation of myosin
phosphatase by Rho and Rho-associated kinase (Rho-kinase). Science 273:245–48
48. Sahai E, Marshall CJ. 2003. Differing modes of tumour cell invasion have distinct require-
ments for Rho/ROCK signalling and extracellular proteolysis. Nat. Cell Biol. 5:711–19

242 Bacac · Stamenkovic

 ANRV335-PM03-09 ARI 9 December 2007 17:30

49. Clark EA, Golub TR, Lander ES, Hynes RO. 2000. Genomic analysis of metastasis
reveals an essential role for RhoC. Nature 406:532–35
50. Gabarra-Niecko V, Schaller MD, Dunty JM. 2003. FAK regulates biological processes
important for the pathogenesis of cancer. Cancer Metastasis Rev. 22:359–74
51. Ilic D, Genbacev O, Jin F, Caceres E, Almeida EA, et al. 2001. Plasma membrane-
associated pY397FAK is a marker of cytotrophoblast invasion in vivo and in vitro. Am. J.
Pathol. 159:93–108
52. Hsia DA, Mitra SK, Hauck CR, Streblow DN, Nelson JA, et al. 2003. Differential reg-
ulation of cell motility and invasion by FAK. J. Cell Biol. 160:753–67
53. Hauck CR, Sieg DJ, Hsia DA, Loftus JC, Gaarde WA, et al. 2001. Inhibition of focal ad-
hesion kinase expression or activity disrupts epidermal growth factor-stimulated signaling
promoting the migration of invasive human carcinoma cells. Cancer Res. 61:7079–90
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

54. Schneider GB, Kurago Z, Zaharias R, Gruman LM, Schaller MD, Hendrix MJ. 2002. Ele-
vated focal adhesion kinase expression facilitates oral tumor cell invasion. Cancer 95:2508–
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

15
55. Gavert N, Conacci-Sorrell M, Gast D, Schneider A, Altevogt P, et al. 2005. L1, a novel
target of β-catenin signaling, transforms cells and is expressed at the invasive front of
colon cancers. J. Cell Biol. 168:633–42
56. Conacci-Sorrell ME, Ben-Yedidia T, Shtutman M, Feinstein E, Einat P, Ben-Ze’ev A.
2002. Nr-CAM is a target gene of the β-catenin/LEF-1 pathway in melanoma and
colon cancer and its expression enhances motility and confers tumorigenesis. Genes Dev.
16:2058–72
57. Perl AK, Dahl U, Wilgenbus P, Cremer H, Semb H, Christofori G. 1999. Reduced
expression of neural cell adhesion molecule induces metastatic dissemination of pancreatic
β tumor cells. Nat. Med. 5:286–91
58. Ponta H, Sherman L, Herrlich PA. 2003. CD44: from adhesion molecules to signalling
regulators. Nat. Rev. Mol. Cell Biol. 4:33–45
59. Sherman LS, Rizvi TA, Karyala S, Ratner N. 2000. CD44 enhances neuregulin signaling
by Schwann cells. J. Cell Biol. 150:1071–84
60. Orian-Rousseau V, Chen L, Sleeman JP, Herrlich P, Ponta H. 2002. CD44 is required
for two consecutive steps in HGF/c-Met signaling. Genes Dev. 16:3074–86
61. Yu Q, Stamenkovic I. 1999. Localization of matrix metalloproteinase 9 to the cell surface
provides a mechanism for CD44-mediated tumor invasion. Genes Dev. 13:35–48
62. Yu Q, Stamenkovic I. 2000. Cell surface-localized matrix metalloproteinase-9 prote-
olytically activates TGF-β and promotes tumor invasion and angiogenesis. Genes Dev.
14:163–76
63. Yu WH, Woessner JFJ, McNeish JD, Stamenkovic I. 2002. CD44 anchors the assembly of
matrilysin/MMP-7 with heparin-binding epidermal growth factor precursor and ErbB4
and regulates female reproductive organ remodeling. Genes Dev. 16:307–23
64. Tsukita S, Oishi K, Sato N, Sagara J, Kawai A, Tsukita S. 1994. ERM family members as
molecular linkers between the cell surface glycoprotein CD44 and actin-based cytoskele-
tons. J. Cell Biol. 126:391–401
65. McClatchey AI. 2003. Merlin and ERM proteins: unappreciated roles in cancer develop-
ment? Nat. Rev. Cancer 3:877–83
66. Thomas L, Byers HR, Vink J, Stamenkovic I. 1992. CD44H regulates tumor cell migra-
tion on hyaluronate-coated substrate. J. Cell Biol. 118:971–77
67. Yu Q, Stamenkovic I. 2004. Transforming growth factor-β facilitates breast carcinoma
metastasis by promoting tumor cell survival. Clin. Exp. Metastasis 21:235–42

[Link] • Metastatic Cancer Cell 243

 ANRV335-PM03-09 ARI 9 December 2007 17:30

68. English NM, Lesley JF, Hyman R. 1998. Site-specific de-N-glycosylation of CD44 can
activate hyaluronan binding, and CD44 activation states show distinct threshold densities
for hyaluronan binding. Cancer Res. 58:3736–42
69. Lesley J, Hascall VC, Tammi M, Hyman R. 2000. Hyaluronan binding by cell surface
CD44. J. Biol. Chem. 275:26967–75
70. Skelton TP, Zeng C, Nocks A, Stamenkovic I. 1998. Glycosylation provides both stim-
ulatory and inhibitory effects on cell surface and soluble CD44 binding to hyaluronan.
J. Cell Biol. 140:431–46
71. Sleeman J, Rudy W, Hofmann M, Moll J, Herrlich P, Ponta H. 1996. Regulated clus-
tering of variant CD44 proteins increases their hyaluronate binding capacity. J. Cell Biol.
135:1139–50
72. Weber GF, Bronson RT, Ilagan J, Cantor H, Schmits R, Mak TW. 2002. Absence of the
CD44 gene prevents sarcoma metastasis. Cancer Res. 62:2281–86
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

73. Liotta LA, Kohn EC. 2001. The microenvironment of the tumour-host interface. Nature
411:375–79
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

74. Bacac M, Provero P, Mayran N, Stehle JC, Fusco C, Stamenkovic I. 2006. A mouse
stromal response to tumor invasion predicts prostate and breast cancer patient survival.
PLoS One 1:e32
75. Chang HY, Sneddon JB, Alizadeh AA, Sood R, West RB, et al. 2004. Gene expression
signature of fibroblast serum response predicts human cancer progression: similarities
between tumors and wounds. PLoS Biol. 2:e7
76. Allinen M, Beroukhim R, Cai L, Brennan C, Lahti-Domenici J, et al. 2004. Molecular
characterization of the tumor microenvironment in breast cancer. Cancer Cell 6:17–32
77. Hotary K, Li XY, Allen E, Stevens SL, Weiss SJ. 2006. A cancer cell metalloprotease
triad regulates the basement membrane transmigration program. Genes Dev. 20:2673–86
78. Hotary KB, Allen ED, Brooks PC, Datta NS, Long MW, Weiss SJ. 2003. Membrane
type I matrix metalloproteinase usurps tumor growth control imposed by the three-
dimensional extracellular matrix. Cell 114:33–45
79. Egeblad M, Werb Z. 2002. New functions for the matrix metalloproteinases in cancer
progression. Nat. Rev. Cancer 2:161–74
80. Zhou Z, Apte SS, Soininen R, Cao R, Baaklini GY, et al. 2000. Impaired endochondral os-
sification and angiogenesis in mice deficient in membrane-type matrix metalloproteinase
I. Proc. Natl. Acad. Sci. USA 97:4052–57
81. Hiraoka N, Allen E, Apel IJ, Gyetko MR, Weiss SJ. 1998. Matrix metalloproteinases
regulate neovascularization by acting as pericellular fibrinolysins. Cell 95:365–77
82. Nakahara H, Howard L, Thompson EW, Sato H, Seiki M, et al. 1997. Transmem-
brane/cytoplasmic domain-mediated membrane type 1-matrix metalloprotease docking
to invadopodia is required for cell invasion. Proc. Natl. Acad. Sci. USA 94:7959–64
83. Kajita M, Itoh Y, Chiba T, Mori H, Okada A, et al. 2001. Membrane-type 1 matrix
metalloproteinase cleaves CD44 and promotes cell migration. J. Cell Biol. 153:893–904
84. Mori H, Tomari T, Koshikawa N, Kajita M, Itoh Y, et al. 2002. CD44 directs membrane-
type 1 matrix metalloproteinase to lamellipodia by associating with its hemopexin-like
domain. EMBO J. 21:3949–59
85. Stamenkovic I. 2003. Extracellular matrix remodelling: the role of matrix metallopro-
teinases. J. Pathol. 200:448–64
86. Wei Y, Lukashev M, Simon DI, Bodary SC, Rosenberg S, et al. 1996. Regulation of
integrin function by the urokinase receptor. Science 273:1551–55
87. Teder P, Vandivier RW, Jiang D, Liang J, Cohn L, et al. 2002. Resolution of lung inflam-
mation by CD44. Science 296:155–58

244 Bacac · Stamenkovic

 ANRV335-PM03-09 ARI 9 December 2007 17:30

88. Bierie B, Moses HL. 2006. Tumour microenvironment: TGFβ: the molecular Jekyll and
Hyde of cancer. Nat. Rev. Cancer 6:506–20
89. Bhowmick NA, Chytil A, Plieth D, Gorska AE, Dumont N, et al. 2004. TGF-β signaling
in fibroblasts modulates the oncogenic potential of adjacent epithelia. Science 303:848–51
90. Bhowmick NA, Neilson EG, Moses HL. 2004. Stromal fibroblasts in cancer initiation
and progression. Nature 432:332–37
91. Olumi AF, Grossfeld GD, Hayward SW, Carroll PR, Tlsty TD, Cunha GR. 1999.
Carcinoma-associated fibroblasts direct tumor progression of initiated human prostatic
epithelium. Cancer Res. 59:5002–11
92. Boire A, Covic L, Agarwal A, Jacques S, Sherifi S, Kuliopulos A. 2005. PAR1 is a matrix
metalloprotease-1 receptor that promotes invasion and tumorigenesis of breast cancer
cells. Cell 120:303–13
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

93. Sternlicht MD, Lochter A, Sympson CJ, Huey B, Rougier JP, et al. 1999. The stromal
proteinase MMP3/stromelysin-1 promotes mammary carcinogenesis. Cell 98:137–46
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

94. Orimo A, Gupta PB, Sgroi DC, Arenzana-Seisdedos F, Delaunay T, et al. 2005. Stro-
mal fibroblasts present in invasive human breast carcinomas promote tumor growth and
angiogenesis through elevated SDF-1/CXCL12 secretion. Cell 121:335–48
95. Muraoka RS, Koh Y, Roebuck LR, Sanders ME, Brantley-Sieders D, et al. 2003. Increased
malignancy of Neu-induced mammary tumors overexpressing active transforming growth
factor β1. Mol. Cell. Biol. 23:8691–703
96. Muraoka RS, Dumont N, Ritter CA, Dugger TC, Brantley DM, et al. 2002. Blockade of
TGF-β inhibits mammary tumor cell viability, migration, and metastases. J. Clin. Invest.
109:1551–59
97. Friedl P, Wolf K. 2003. Tumour-cell invasion and migration: diversity and escape mech-
anisms. Nat. Rev. Cancer 3:362–74
98. Jacques TS, Relvas JB, Nishimura S, Pytela R, Edwards GM, et al. 1998. Neural precursor
cell chain migration and division are regulated through different β1 integrins. Development
125:3167–77
99. El Fahime E, Torrente Y, Caron NJ, Bresolin MD, Tremblay JP. 2000. In vivo migra-
tion of transplanted myoblasts requires matrix metalloproteinase activity. Exp. Cell Res.
258:279–87
100. Davidson LA, Keller RE. 1999. Neural tube closure in Xenopus laevis involves medial
migration, directed protrusive activity, cell intercalation and convergent extension. De-
velopment 126:4547–56
101. Simian M, Hirai Y, Navre M, Werb Z, Lochter A, Bissell MJ. 2001. The interplay of
matrix metalloproteinases, morphogens and growth factors is necessary for branching of
mammary epithelial cells. Development 128:3117–31
102. Friedl P, Hegerfeldt Y, Tusch M. 2004. Collective cell migration in morphogenesis and
cancer. Int. J. Dev. Biol. 48:441–49
103. Hegerfeldt Y, Tusch M, Brocker EB, Friedl P. 2002. Collective cell movement in primary
melanoma explants: plasticity of cell-cell interaction, β1-integrin function, and migration
strategies. Cancer Res. 62:2125–30
104. Wicki A, Lehembre F, Wick N, Hantusch B, Kerjaschki D, Christofori G. 2006. Tu-
mor invasion in the absence of epithelial-mesenchymal transition: podoplanin-mediated
remodeling of the actin cytoskeleton. Cancer Cell 9:261–72
105. Coussens LM, Werb Z. 2002. Inflammation and cancer. Nature 420:860–67
106. Dranoff G. 2004. Cytokines in cancer pathogenesis and cancer therapy. Nat. Rev. Cancer
4:11–22

[Link] • Metastatic Cancer Cell 245

 ANRV335-PM03-09 ARI 9 December 2007 17:30

107. Coussens LM, Tinkle CL, Hanahan D, Werb Z. 2000. MMP-9 supplied by bone marrow-
derived cells contributes to skin carcinogenesis. Cell 103:481–90
108. Condeelis J, Pollard JW. 2006. Macrophages: obligate partners for tumor cell migration,
invasion, and metastasis. Cell 124:263–66
109. Aharinejad S, Abraham D, Paulus P, Abri H, Hofmann M, et al. 2002. Colony-stimulating
factor-1 antisense treatment suppresses growth of human tumor xenografts in mice. Cancer
Res. 62:5317–24
110. Lin EY, Nguyen AV, Russell RG, Pollard JW. 2001. Colony-stimulating factor 1 pro-
motes progression of mammary tumors to malignancy. J. Exp. Med. 193:727–40
111. Lin EY, Pollard JW. 2004. Role of infiltrated leucocytes in tumour growth and spread.
Br. J. Cancer 90:2053–58
112. Bergers G, Benjamin LE. 2003. Tumorigenesis and the angiogenic switch. Nat. Rev.
Cancer 3:401–10
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

113. Hanahan D, Folkman J. 1996. Patterns and emerging mechanisms of the angiogenic
switch during tumorigenesis. Cell 86:353–64
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

114. Coussens LM, Raymond WW, Bergers G, Laig-Webster M, Behrendtsen O, et al. 1999.
Inflammatory mast cells up-regulate angiogenesis during squamous epithelial carcino-
genesis. Genes Dev. 13:1382–97
115. Achen MG, McColl BK, Stacker SA. 2005. Focus on lymphangiogenesis in tumor metas-
tasis. Cancer Cell 7:121–27
116. Saharinen P, Tammela T, Karkkainen MJ, Alitalo K. 2004. Lymphatic vasculature: de-
velopment, molecular regulation and role in tumor metastasis and inflammation. Trends
Immunol. 25:387–95
117. Kim J, Yu W, Kovalski K, Ossowski L. 1998. Requirement for specific proteases in cancer
cell intravasation as revealed by a novel semiquantitative PCR-based assay. Cell 94:353–62
118. Sackstein R. 2005. The lymphocyte homing receptors: gatekeepers of the multistep
paradigm. Curr. Opin. Hematol. 12:444–50
119. Rosen SD. 2004. Ligands for L-selectin: homing, inflammation, and beyond. Annu. Rev.
Immunol. 22:129–56
120. Kim YJ, Borsig L, Han HL, Varki NM, Varki A. 1999. Distinct selectin ligands on colon
carcinoma mucins can mediate pathological interactions among platelets, leukocytes, and
endothelium. Am. J. Pathol. 155:461–72
121. Kim YJ, Borsig L, Varki NM, Varki A. 1998. P-selectin deficiency attenuates tumor
growth and metastasis. Proc. Natl. Acad. Sci. USA 95:9325–30
122. Biancone L, Araki M, Araki K, Vassalli P, Stamenkovic I. 1996. Redirection of tumor
metastasis by expression of E-selectin in vivo. J. Exp. Med. 183:581–87
123. Felding-Habermann B, O’Toole TE, Smith JW, Fransvea E, Ruggeri ZM, et al. 2001.
Integrin activation controls metastasis in human breast cancer. Proc. Natl. Acad. Sci. USA
98:1853–58
124. Borsig L, Wong R, Feramisco J, Nadeau DR, Varki NM, et al. 2001. Heparin and cancer
revisited: mechanistic connections involving platelets, P-selectin, carcinoma mucins, and
tumor metastasis. Proc. Natl. Acad. Sci. USA 98:3352–57
125. Laubli H, Stevenson JL, Varki A, Varki NM, Borsig L. 2006. L-selectin facilitation of
metastasis involves temporal induction of Fut7-dependent ligands at sites of tumor cell
arrest. Cancer Res. 66:1536–42
126. Taichman DB, Cybulsky MI, Djaffar I, Longenecker BM, Teixido J, et al. 1991. Tumor
cell surface α4β1 integrin mediates adhesion to vascular endothelium: demonstration
of an interaction with the N-terminal domains of INCAM-110/VCAM-1. Cell Regul.
2:347–55

246 Bacac · Stamenkovic

 ANRV335-PM03-09 ARI 9 December 2007 17:30

127. Chambers AF, Groom AC, MacDonald IC. 2002. Dissemination and growth of cancer
cells in metastatic sites. Nat. Rev. Cancer 2:563–72
128. Muller A, Homey B, Soto H, Ge N, Catron D, et al. 2001. Involvement of chemokine
receptors in breast cancer metastasis. Nature 410:50–56
129. Kaplan RN, Rafii S, Lyden D. 2006. Preparing the “soil”: the premetastatic niche. Cancer
Res. 66:11089–93
130. Kaplan RN, Riba RD, Zacharoulis S, Bramley AH, Vincent L, et al. 2005. VEGFR1-
positive haematopoietic bone marrow progenitors initiate the pre-metastatic niche.
Nature 438:820–27
131. Fidler IJ. 2003. The pathogenesis of cancer metastasis: the ‘seed and soil’ hypothesis
revisited. Nat. Rev. Cancer 3:453–58
132. Minn AJ, Gupta GP, Siegel PM, Bos PD, Shu W, et al. 2005. Genes that mediate breast
cancer metastasis to lung. Nature 436:518–24
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

133. Kang Y, Siegel PM, Shu W, Drobnjak M, Kakonen SM, et al. 2003. A multigenic program
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

mediating breast cancer metastasis to bone. Cancer Cell 3:537–49

134. Varghese HJ, Davidson MT, MacDonald IC, Wilson SM, Nadkarni KV, et al. 2002.
Activated ras regulates the proliferation/apoptosis balance and early survival of developing
micrometastases. Cancer Res. 62:887–91
135. Holmgren L, O’Reilly MS, Folkman J. 1995. Dormancy of micrometastases: balanced
proliferation and apoptosis in the presence of angiogenesis suppression. Nat. Med. 1:149–
53
136. Nyberg P, Xie L, Kalluri R. 2005. Endogenous inhibitors of angiogenesis. Cancer Res.
65:3967–79
137. van de Vijver MJ, He YD, van’t Veer LJ, Dai H, Hart AA, et al. 2002. A gene-expression
signature as a predictor of survival in breast cancer. N. Engl. J. Med. 347:1999–2009
138. van’t Veer LJ, Dai H, van de Vijver MJ, He YD, Hart AA, et al. 2002. Gene expression
profiling predicts clinical outcome of breast cancer. Nature 415:530–36
139. Ramaswamy S, Ross KN, Lander ES, Golub TR. 2003. A molecular signature of metas-
tasis in primary solid tumors. Nat. Genet. 33:49–54
140. Gangnus R, Langer S, Breit E, Pantel K, Speicher MR. 2004. Genomic profiling of viable
and proliferative micrometastatic cells from early-stage breast cancer patients. Clin. Cancer
Res. 10:3457–64
141. Glinsky GV, Berezovska O, Glinskii AB. 2005. Microarray analysis identifies a death-
from-cancer signature predicting therapy failure in patients with multiple types of cancer.
J. Clin. Invest. 115:1503–21
142. Glinsky GV. 2005. Death-from-cancer signatures and stem cell contribution to metastatic
cancer. Cell Cycle 4:1171–75
143. Glinsky GV. 2006. Genomic models of metastatic cancer: functional analysis of death-
from-cancer signature genes reveals aneuploid, anoikis-resistant, metastasis-enabling
phenotype with altered cell cycle control and activated Polycomb Group (PcG) protein
chromatin silencing pathway. Cell Cycle 5:1208–16
144. Chang HY, Nuyten DS, Sneddon JB, Hastie T, Tibshirani R, et al. 2005. Robustness,
scalability, and integration of a wound-response gene expression signature in predicting
breast cancer survival. Proc. Natl. Acad. Sci. USA 102:3738–43
145. West RB, Nuyten DS, Subramanian S, Nielsen TO, Corless CL, et al. 2005. Determi-
nation of stromal signatures in breast carcinoma. PLoS Biol. 3:e187
146. Al-Hajj M, Wicha MS, Benito-Hernandez A, Morrison SJ, Clarke MF. 2003. Prospective
identification of tumorigenic breast cancer cells. Proc. Natl. Acad. Sci. USA 100:3983–88

[Link] • Metastatic Cancer Cell 247

 AR335-FM ARI 10 December 2007 18:20

Annual Review
of Pathology:
Mechanisms
of Disease

Contents Volume 3, 2008

Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]
by University of Massachusetts - Amherst on 09/30/12. For personal use only.

The Relevance of Research on Red Cell Membranes to the

Understanding of Complex Human Disease: A Personal Perspective
Vincent T. Marchesi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p1
Molecular Mechanisms of Prion Pathogenesis
Adriano Aguzzi, Christina Sigurdson, and Mathias Heikenwalder p p p p p p p p p p p p p p p p p p p p 11
The Aging Brain
Bruce A. Yankner, Tao Lu, and Patrick Loerch p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 41
Gene Expression Profiling of Breast Cancer
Maggie C.U. Cheang, Matt van de Rijn, and Torsten O. Nielsen p p p p p p p p p p p p p p p p p p p p p p 67
The Inflammatory Response to Cell Death
Kenneth L. Rock and Hajime Kono p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 99
Molecular Biology and Pathogenesis of Viral Myocarditis
Mitra Esfandiarei and Bruce M. McManus p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p127
Pancreatic Cancer
Anirban Maitra and Ralph H. Hruban p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p157
Kidney Transplantation: Mechanisms of Rejection and Acceptance
Lynn D. Cornell, R. Neal Smith, and Robert B. Colvin p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p189
Metastatic Cancer Cell
Marina Bacac and Ivan Stamenkovic p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p221
Pathogenesis of Thrombotic Microangiopathies
X. Long Zheng and J. Evan Sadler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p249
Anti-Inflammatory and Proresolving Lipid Mediators
Charles N. Serhan, Stephanie Yacoubian, and Rong Yang p p p p p p p p p p p p p p p p p p p p p p p p p p p p p279
Modeling Morphogenesis and Oncogenesis in Three-Dimensional
Breast Epithelial Cultures
Christy Hebner, Valerie M. Weaver, and Jayanta Debnath p p p p p p p p p p p p p p p p p p p p p p p p p p p p313

v
 AR335-FM ARI 10 December 2007 18:20

The Origins of Medulloblastoma Subtypes

Richard J. Gilbertson and David W. Ellison p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p341
Molecular Biology and Pathology of Lymphangiogenesis
Terhi Karpanen and Kari Alitalo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p367
Endoplasmic Reticulum Stress in Disease Pathogenesis
Jonathan H. Lin, Peter Walter, and T.S. Benedict Yen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p399
Autophagy: Basic Principles and Relevance to Disease
Mondira Kundu and Craig B. Thompson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p427
The Osteoclast: Friend or Foe?
Annu. Rev. Pathol. Mech. Dis. 2008.3:221-247. Downloaded from [Link]

Deborah V. Novack and Steven L. Teitelbaum p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p457

by University of Massachusetts - Amherst on 09/30/12. For personal use only.

Applications of Proteomics to Lab Diagnosis

Raghothama Chaerkady and Akhilesh Pandey p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p485
The Pathology of Influenza Virus Infections
Jeffrey K. Taubenberger and David M. Morens p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p499
Airway Smooth Muscle in Asthma
Marc B. Hershenson, Melanie Brown, Blanca Camoretti-Mercado,
and Julian Solway p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p523
Molecular Pathobiology of Gastrointestinal Stromal Sarcomas
Christopher L. Corless and Michael C. Heinrich p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p557
Notch Signaling in Leukemia
Jon C. Aster, Warren S. Pear, and Stephen C. Blacklow p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p587
The Role of Hypoxia in Vascular Injury and Repair
Tony E. Walshe and Patricia A. D’Amore p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p615

Indexes

Cumulative Index of Contributing Authors, Volumes 1–3 p p p p p p p p p p p p p p p p p p p p p p p p p p p645

Cumulative Index of Chapter Titles, Volumes 1–3 p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p647

Errata

An online log of corrections to Annual Review of Pathology: Mechanisms of Disease

articles may be found at [Link]

vi Contents