Seminars in Cancer Biology 21 (2011) 99106

Contents lists available at ScienceDirect

Seminars in Cancer Biology

journal homepage: www.elsevier.com/locate/semcancer

Review

Metastatic colonization: Settlement, adaptation and propagation of tumor cells

in a foreign tissue environment
Tsukasa Shibue a,b , Robert A. Weinberg a,b,c,
a

Whitehead Institute for Biomedical Research, 9 Cambridge Center, Cambridge, MA 02142, USA
MIT Ludwig Center for Molecular Oncology, 77 Massachusetts Ave., Cambridge, MA 02139, USA
c
Department of Biology, Massachusetts Institute of Technology, 77 Massachusetts Ave., Cambridge, MA 02139, USA
b

a r t i c l e

i n f o

Keywords:
Metastasis
Extravasation
Colonization
Dormancy
Tumorhost interactions

a b s t r a c t
Disseminated tumor cells must negotiate multiple situations that challenge their viability and/or proliferative capacity before they can successfully colonize distant organ sites. Thus, the shear stress caused
by the blood ow may physically damage tumor cells during their translocation from primary tumors to
distant organs via the circulation. In addition, the tissue microenvironment of distant organs is generally unfamiliar to tumor cells, limiting their proliferation within the parenchyma of these organs. Each
of these situations involves various types of interactions between tumor cells and host components,
which either support or inhibit the establishment and subsequent progression of metastases. The initial
formation of micrometastases, as well as their subsequent growth often termed colonization therefore require complex adaptations by tumor cells to various host components, most of which are never
encountered by these cells during their growth within primary tumor sites. These difculties explain why
the colonization of distant organs by disseminated tumor cells is an extraordinarily demanding task and
thus inefcient, and suggests a number of potential targets that might be used in the future to interfere
therapeutically with this process. Studying the details of tumorhost interactions at each of the steps
leading up to successful metastatic colonization may therefore pave the way for designing therapeutic
strategies to counteract the metastatic spread of malignant tumors.
2010 Elsevier Ltd. All rights reserved.

Hematogenous metastasis is the primary cause of cancerassociated mortality. This process, often referred to as the
invasion-metastasis cascade, proceeds in a stepwise manner that
begins with the invasion of surrounding host tissues by the tumor
cells. This is followed by the penetration of the blood vessel walls by
the tumor cells and the entrance of these cells into the circulation
(intravasation). After being disseminated via the blood stream to
sites anatomically distant from the primary tumor (transport),
circulating tumor cells are arrested in the capillary beds (arrest),
invade through the microvascular walls, and enter the parenchyma
of the target organs (extravasation), in which they may survive,
proliferate and thereby establish metastatic colonies (colonization) [1].
Various lines of evidence support the notion that systemic dissemination of tumor cells can already occur at relatively early
stages of the progression of primary tumors in certain types of can-

Corresponding author at: Whitehead Institute for Biomedical Research, 9 Cambridge Center, Cambridge, MA 02142, USA. Tel.: +1 617 258 5159;
fax: +1 617 258 5213.
E-mail addresses: shibue@wi.mit.edu (T. Shibue), weinberg@wi.mit.edu
(R.A. Weinberg).
1044-579X/$ see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.semcancer.2010.12.003

cers, including those in the breast and prostate [2]. In addition, in an

epidemiological study of breast cancer, tumor cells are estimated to
disseminate from the primary site 57 years prior to the diagnosis
of the primary tumor [3]. This phenomenon indicates the need to
understand the steps of metastases that occur after dissemination
of tumor cells from the primary site and to develop strategies to
block these later steps of the invasion-metastasis cascade.
In the present review, we will provide a brief summary of the
current understanding, controversies and future prospects in the
studies of tumorhost interactions, specically those associated
with the last steps of the invasion-metastasis cascade that follow the dissemination of tumor cells from the primary site, with
special emphasis on the direct interactions between tumor cells
and the insoluble components of the host microenvironment. For
understanding the details of the individual steps of this cascade,
readers are encouraged to refer to more specic review articles
[49].
1. Transport and arrest in the capillary beds
Tumor cells that have entered the circulation are likely to be
transported quickly to distant organ sites via the blood stream and
to become lodged in the capillary beds of these organs. Thus, in mul-

100

T. Shibue, R.A. Weinberg / Seminars in Cancer Biology 21 (2011) 99106

tiple experimental models of metastasis in mice, the great majority

of the tumor cells that are injected directly into the venous circulation are arrested within minutes in the microvasculature of the rst
organ that they encounter [10]. In accord with these experimental
observations, a clinical study of breast cancer demonstrated that
the half-life of the tumor cells within the circulation was in range
of 12.4 h, a number that was calculated by taking blood samples
from metastasis-free breast cancer patients at short intervals starting immediately after the surgical removal of the primary tumor
and tracing the changes in the numbers of circulating tumor cells
within these blood samples [11].
Two distinct mechanisms are thought to underlie the arrest of
circulating tumor cells in the microvasculature: mechanical entrapment due to size restriction and adhesion of the tumor cells to the
vascular endothelium. Indeed, B16F1 mouse melanoma cells that
are injected into mice via the portal vein are arrested in the liver
near the ends of terminal portal venules due to size restriction [12].
Similarly, CT-26 mouse colon carcinoma cells injected through the
cecal vein do not adhere to the larger portal vessels but are instead
trapped mechanically at sinusoids [13]. In contrast, HT-29 human
colon carcinoma cells can easily pass through the microvessels of
the liver when injected into rats via the left ventricle; however, a
fraction of these cells adhere to the walls of microvessels whose
luminal diameters are larger than those of these cells [14]. Hence,
both the mechanical entrapment and adhesion can contribute functionally to the initial arrest of tumor cells in the capillary beds; the
relative contributions of these two mechanisms is likely to vary
depending on the conditions such as the types of tumor cells and
target organs.
The adhesion of circulating tumor cells to the capillary walls
of target organs is analogous, at least supercially, to the process
of leukocyte adhesion to the capillary endothelium [15]. Thus, in
inammatory responses, leukocytes activated by proinammatory
cytokines such as interleukin (IL) -1, -6, -8 and tumor necrosis
factor (TNF)- rst become arrested on the luminal surfaces of
endothelial cells in inammatory tissues, which also is activated
by the same set of cytokines; this arrest is mediated by low-afnity
interactions between the glycoproteins, such as P-selectin glycoprotein ligand-1 (PSGL-1), on the surface of these leukocytes and
selectins the transmembrane receptors for these glycoproteins
that are expressed on endothelial cell surface. These weak interactions, together with the propulsion provided by the blood ow,
result in the rolling of leukocytes on the luminal surface of the
endothelium. This rolling behavior is soon followed by the formation of more stable adhesions between the cell adhesion molecules
(CAMs) of the immunoglobulin (Ig) superfamily, such as ICAM-1
(Inter-Cellular Adhesion Molecule-1) and VCAM-1 (Vascular Cell
Adhesion Molecule-1), on the luminal surfaces of endothelial cells
and their ligands mainly integrins on the surface of the leukocytes [15].
Certain types of tumor cells exhibit similar rolling behavior on the endothelial surface, which is also mediated by the
selectinglycoprotein interactions (Fig. 1). For example, MDA PCa
2b human prostate carcinoma cells exhibit rolling behavior in
monolayer culture of bone marrow endothelial cells (BMECs) under
conditions of shear ow. This behavior is dependent on the associations between sialylated glycoproteins on MDA PCa 2b cells and
E-selectin on BMECs [16]. However, there also are other situations
in which tumor cells do not exhibit rolling behavior prior to their
arrest in the capillary beds [17,18]. For example, lung metastasis
formation by cells of the DU-145 human prostate carcinoma cell
line and the MDA-MB-435 human breast cancer cell line is not
inhibited by the functional blocking of selectins with neutralizing antibodies [19]. Hence, the general importance of the selectinand glycoprotein-dependent rolling behavior for the subsequent
formation of metastases remains unclear.

Regardless of the relative importance of the rolling behavior, the

subsequent stable adhesions between tumor cells and the endothelium are likely to be mediated by the binding of integrins on
the surface of tumor cells to CAMs of the Ig superfamily on the
endothelial cell surface. These stable interactions have been found
to be important for the subsequent extravasation of tumor cells
into the parenchyma of target organs and thus for the formation
of metastatic colonies. For example, studies using A375M human
melanoma cells and B16BL6 mouse melanoma cells have revealed
the essential role of interactions between tumor cell integrin 4 1
and endothelial cell VCAM-1 in the development of lung metastases
[20,21].
The interactions between tumor cells and endothelial cells are
supposed to account, in part, for the organ tropisms of metastases.
Thus, endothelial cells from different anatomical locations exhibit
distinct adhesive properties on their luminal surfaces and certain
tumor cells may selectively bind to the endothelium of organs that
are preferentially colonized by these tumor cells. Indeed, in vitro
examination of the adhesion of various human prostate cancer cell
lines, including PC-3, TSU, LNCaP and DU-145, to the endothelial
cells of various origins has revealed that these prostate cancer cells
preferentially adhere to the endothelial cells derived from the bone
marrow relative to those derived from other origins such as the
lungs or umbilical vein [22,23]. This is consistent with the patterns
of metastasis in prostate cancer, a cancer type that metastasizes
almost exclusively to the bone.
Another example of the mechanisms accounting for the organspecic metastasis formation comes from a study of Metadherin,
a transmembrane protein that was identied in a phage display
screening for proteins mediating breast cancer cell adhesion to the
lung endothelium [24]. After intracardiac injection into mice, phage
expressing Metadherin accumulated in the lungs, but not in liver,
brain or bone; this argued that Metadherin mediates selective binding of cells to the lung endothelium. Indeed, blocking the expression
or function of Metadherin by small interfering RNA (siRNA) or by
a neutralizing antibody, respectively, resulted in impaired lung
metastasis formation by 4T1 mouse mammary carcinoma cells,
indicating that Metadherin-dependent homing of tumor cells to the
lungs enables the efcient development of lung metastases [24].
Yet other mechanisms that have been proposed to contribute to
the organ tropisms of metastases include the combination of locally
produced chemokines and chemokine receptors expressed on the
surface of tumor cells. The role of chemokines and their receptors
in the chemotaxis of leukocytes during inammatory responses
is well established [15]; tumor cells may exploit these interactions, explaining their preferential metastasis to certain organs. For
example, the CXCR4 receptor expressed on the surface of MDAMB-231 human breast cancer cells has been shown to facilitate the
metastasis of these cells to the lungs, an organ rich in the CXCR4
ligand CXCL12/SDF-1 [25]. Similarly, the CCR10-CCL27 pair is implicated in melanoma metastasis to skin [26]. However, the patterns of
circulation between the primary site and the secondary site, as well
as other properties of the target organs, such as tissue architecture
and the local availability of growth factors, might also exert strong
effects on the organ specicity of metastases. Consequently, the
importance of tumor cellendothelial cell adhesion and chemokine
receptor signaling to the organ tropisms of metastases remains
unclear.
Tumor cells also interact with various components of the blood
while passing through the general circulation and becoming lodged
in the lumina of capillaries (Fig. 1). These interactions involve
platelets, polymorphonuclear leukocytes (PMNs), monocytes, lymphocytes as well as the multiple plasma proteins. Among these, the
metastasis-promoting role of the interactions between tumor cells
and components of the blood-clotting machinery, including brin clots and aggregated platelets, is well established [27]. Indeed,

T. Shibue, R.A. Weinberg / Seminars in Cancer Biology 21 (2011) 99106

101

Fig. 1. Tumorhost interactions during the transport, arrest and extravasation steps of metastasis. During transport via the blood stream, tumor cells interact with multiple
constituents of the blood; these include components of the blood clotting machinery, such as brin clots and platelets. In the subsequent processes of arrest in the capillary
beds of target organs and extravasation into the parenchyma of these organs, tumor cells closely interact with the endothelial cells. Tumor cell-derived factors (blue) and
host-derived factors (green) involved in the tumorhost interactions occurring in each of these steps are listed. During extravasation, tumor cell-derived factors induce either
retraction (VEGF, 12(S)-HETE) or death (ROS) of the endothelial cells. See main text for the details.

microscopic observations of B16F10 mouse melanoma cells lodged

within the lung microvasculature have revealed their close associations with brin clots and platelets [28]. The deposition of brin
and platelets around the tumor cells appears to serve as a barrier
that protects tumor cells from the mechanical stress of blood ow
as well as attack by immunocytes, such as natural killer (NK) cells
[27]. Moreover, brin and platelets might also facilitate the arrest
of tumor cells in small-diameter capillaries, doing so via a mechanism involving bridge formation between tumor cells that have
previously adhered to the endothelium and those owing freely in
the blood [29]. In support of these notions, pharmacological inhibitions of blood clotting by various agents, including heparin, vitamin
K antagonists and prostacyclin, all result in impaired metastasis
formation in various experimental models in mice and rats [27].
Similarly, a genetic depletion of brinogen, the precursor of brin, as well as genetic defects of platelet production, both result in
remarkable reductions in the efciency of lung metastasis formation by B16F10 mouse melanoma cells [30].
While residing in the blood, tumor cells can trigger coagulation and platelet activation through mechanisms that include
the elevated expression of tissue factor (TF) on their surface. TF
is a transmembrane glycoprotein that produces thrombin, a key
enzyme in the coagulation machinery, from its precursor prothrombin; thrombin, in turn, induces brin clot formation and
platelet activation [31]. The importance of TF-dependent clotting
machinery on the progression of metastases is supported by the
tight correlation between elevated TF expression and poor prognosis in multiple tumor types, including colorectal, breast and
non-small cell lung carcinomas [31].
The interactions between tumor cells and platelets are also
mediated, in part, via the associations between mucin-like glycoproteins on the surface of tumor cells and P-selectin expressed
by platelets [29]. The importance of these specic associations
is supported by a study in which LS180 human colon carcinoma
cells were injected intravenously into wild-type and P-selectindecient mice [32]. Within the circulation of P-selectin-decient
mice, LS180 cells did not become associated with platelets, whereas
those cells injected into wild-type mice acquired a thick coat of
platelets. In the P-selectin-decient mice, the failure by LS810 cells
to acquire a platelet coat was associated with a reduced arrest in the

lung microvasculature and an impaired lung metastasis formation

[32]. These various pieces of experimental and clinical evidence
collectively point to the importance to metastasis formation of
tumorhost interactions occurring intraluminally within the blood
vessels.
In addition to the heterotypic interactions between tumor
cells and the host components, homotypic interactions within
the tumor cell population (i.e., cell clumping) might also contribute to the progression of metastases. Thus, in a study of T241
mouse brosarcoma model, when cells were administered intravenously as clumps of 1012 cells, lung metastases developed
far more efciently than when the same number of cells were
injected as a single cell suspension [33]. Indeed, the formation of
intravascular tumor clumps in lung arteries has long been recognized clinically [34]. Further studies will be required to assess the
general importance of intravascular tumor cell clumping in metastases. For example, it will be interesting to determine whether
metastasis formation by cell clumps can be reconciled with the
largely monoclonal nature of subsequently arising metastatic
colonies [35].
2. Extravasation/intravascular growth
A fraction of tumor cells that are arrested in the capillary
beds of target organs may subsequently extravasate through the
endothelial walls and enter the parenchyma of these organs, which
is usually considered to be essential for the eventual establishment of metastatic colonies [1] (Fig. 1). The precise mechanisms
underlying tumor cell extravasation still remain elusive. Often, this
process is analogized to the better understood process of leukocyte
extravasation into inammatory tissuesthe process of diapedesis
[6,36]. Indeed, the extravasation of both tumor cells and leukocytes is preceded by, and probably requires, the formation of stable
interactions between the extravasating cells (i.e., tumor cells and
leukocytes) and endothelial cells.
Nonetheless, the differences between extravasating cancer cells
and leukocytes undergoing diapedesis are fundamental: leukocytes are biologically adapted to execute diapedesis, resulting in
its completion within minutes, while tumor cells do so only inefciently and often require a day or two to successfully extravasate

102

T. Shibue, R.A. Weinberg / Seminars in Cancer Biology 21 (2011) 99106

[37,38]. Moreover, while leukocytes usually transmigrate through

the endothelium without signicantly disrupting the endothelial
monolayer, the tumor cell extravasation is frequently associated
with retraction of endothelial cells from one another or their death
[6,36]. In sinusoidal capillaries, such as those observed in liver and
bone marrow [39], tumor cells may pass through preexisting gaps
between endothelial cells without damaging these cells.
Several distinct mechanisms have been reported by which
tumor cells induce endothelial cells to retract from one another or
endothelial death during extravasation. The involvement of tumor
cell-derived bioactive lipids in endothelial retraction has been suggested; some tumor cells can induce this process via production of
12(S)-hydroxyeicosatetraenoic acid [12(S)-HETE], a major lipoxygenase metabolite of arachidonic acid [4042]. Indeed, the addition
of 12(S)-HETE to a monolayer culture of endothelial cells induces a
reversible retraction of these cells through a mechanism involving
the redistribution of integrin v 3 on the endothelial cell surface
[41]. In a series of B16a mouse melanoma cell lines, the production of 12(S)-HETE was correlated with the activity of these cells to
induce endothelial cell retraction as well as with the ability of these
cells to form metastases [42]. More recently, 12(S)-HETE was also
identied as a factor responsible for the endothelial cell-retracting
ability within the culture medium of MCF7 human breast cancer
cells [43].
As an alternative to endothelial retraction, tumor cell extravasation might also be facilitated by mechanisms involving endothelial
injury; these include the oxidative stress arising from the interactions between tumor cells and endothelial cells. Indeed, multiple
lines of malignant melanoma cells release reactive oxygen species
(ROS) upon adherence to the conuent monolayer of endothelial cells in vitro, which subsequently cause irreversible damage
to the endothelial cells [44]. Moreover, in an experimental lung
metastasis model of rats, Walker 256 rat carcinosarcoma cells that
are arrested at the lung microvasculature exhibit H2 O2 production
along the contact sites with the lung tissue [45], suggesting the contribution of ROS to the in vivo interactions between tumor cells and
endothelial cells.
Accumulating evidence also implicates regulators of angiogenesis/vascular remodeling in tumor cell extravasation. Thus, vascular
endothelial growth factor (VEGF), a central regulator of angiogenesis, appears to play essential roles in the tumor cell extravasation
under several experimental conditions. The extravasation of CT26 colon carcinoma cells through the lung endothelium depends
on the VEGF-induced dissolution of endothelial cellcell junctions;
this response is mediated, in turn, by the VEGF-dependent activation of Src family kinases in the endothelial cells and the resulting
dissociation of -catenin from VE-cadherin, which is responsible
for forming the adherens junctions between adjacent endothelial
cells [46]. VEGF was also found to be essential for the migration of MDA-MB-231 human breast cancer cells through a layer
of human brain microvascular endothelial cells (HBMECs) in vitro
[47].
More recently, in a search for factors whose expression is correlated with the capability of breast cancer cells to metastasize to the
lungs, Massagu and colleagues identied multiple factors that are
also known to be involved in vascular remodeling; these include
epiregulin, cyclooxygenase 2, as well as matrix metalloproteinase
1 and 2 [48]. Knocking-down the expression of these factors in combination impairs the extravasation of MDA-MB-231 human breast
cancer cells into the lung parenchyma. Accordingly, these agents
promote breast cancer metastasis at least in part by facilitating
the tumor cell extravasation in this tissue. Interestingly, expression proling of clinical samples has revealed that the expression of
these factors within the primary tumor correlated with the development of metastases to the lungs but not to the bone [49]; this
is consistent with the notion that the mechanism of tumor cell

extravasation may differ depending on the identity of the target

organ.
There are also cases in which tumor cells initially proliferate
intraluminally within the vasculature and begin to form metastases without undergoing extravasation; as these metastases grow
in size, they will inevitably rupture the microvascular wall and
thereby invade into the tissue parenchyma. Several cancer cell
lines, including HT1080 human brosarcoma cells, 2.10.10 rat
embryo broblast-derived cells [50] and PC-3 human prostate cancer cells [51] do not effectively extravasate into the parenchyma
of the lungs and liver when injected into mice. Instead, these
cells start proliferating while still attached to the luminal surfaces
of the endothelium. Analyses of metastases from the spontaneous mammary tumor formed in C3H/He mice have also revealed
the intravascular origin of lung metastases [52]. Accordingly,
metastatic colonization may result either from the proliferation of
tumor cells within the organ parenchyma following extravasation
or from their initial proliferation within the vascular lumen prior to
extravasation. The relative contributions of these alternative processes may well vary depending on the types of tumor cells and
target organs.
3. Colonization
Independent on the mode of extravasation, a rapidly growing
body of evidence indicates the key role of tumorhost interactions
within the parenchyma of target organs as a primary determinant
of the subsequent success of metastatic colonizationthe process
involving the growth of micrometastases into macroscopic tumors.
Using the technique of intravital imaging and various mouse mammary carcinoma and melanoma cell lines, Chambers and colleagues
have demonstrated that the inability to actively proliferate following extravasation into the parenchyma of a target organ is a major
obstacle in metastatic colonization [12,53,54]. In fact, untransformed mammary epithelial cells can survive in the parenchyma of
the lungs for up to 4 months after intravenous injection, indicating
that even these untransformed cells can pass through the process of
hematogenous transport, arrest in microvasculature and extravasation, albeit in the absence of vigorous proliferation within the
lung parenchyma [55]. In the sections that follow, we will focus on
the interactions between extravasated tumor cells and components
within the host parenchyma that either foster or hinder subsequent
tumor cell proliferation.
In general, extravasated tumor cells will follow one of the three
alternative coursescell death, dormancy (i.e., survival without
apparent increase in cell number), or colony formation by continuous proliferation in the absence of counterbalancing cell death,
permitting net increases in cell number [56]. The determinants
of these alternative fates include the interactions of tumor cells
with the constituents of the target organ parenchyma, such as ECM
components and various host stromal cells (Fig. 2). In addition, vascularization and immune surveillance appear to have signicant
effects on the fate of extravasated tumor cells (Fig. 2).
3.1. Interactions with the ECM components of the target organ
parenchyma
The important role of cellECM interactions in the proliferation and survival of tumor cells has been well established [9,57].
Thus, binding of tumor cell integrins to ECM ligands activates
a form of intracellular signaling that promotes cell proliferation
and/or survival; this type of integrin-dependent intracellular signal
transduction is generally referred to as outside-in signaling [58].
Integrins can directly activate signaling proteins, such as focal adhesion kinase (FAK). In addition, they can also crosstalk with other

T. Shibue, R.A. Weinberg / Seminars in Cancer Biology 21 (2011) 99106

103

Fig. 2. Tumorhost interactions in the post-extravasation process of metastasis. The interactions between tumor cells and host components also play key roles in determining
the fate of tumor cells after extravasating into the parenchyma of target organs. The host components interacting with tumor cells in this post-extravasation process of
metastasis include extracellular matrix (ECM), immunocytes (such as macrophages, CD8+ T cells, and NK cells), blood vessels (i.e., endothelial cells), and other organ-specic
cells (such as osteoclasts and osteoblasts in the bone and hepatocytes in the liver). As in Fig. 1, tumor cell-derived factors and host-derived factors involved in these
interactions are listed in blue and green text, respectively. (*) Different types of tumor cells interact with the organ-specic cell types in distinct manners. For example, breast
cancer metastases to the bone often result in the preferential activation of osteoclasts and are therefore osteolytic. In contrast, prostate cancer metastases to the bone are
predominantly osteoblastic because of the selective activation of osteoblasts by the prostate cancer cells. See main text for the details.

signaling pathways, such as those governed by growth factor receptors, to augment the effects of ligand-activated signaling [57]. Given
the differences that are likely to distinguish the organization of ECM
in the primary tumor site from that present in the ECM of target
organs, it is likely that certain integrin-dependent signals that permitted tumor cell proliferation and/or survival within the primary
site are no longer available for tumor cells after their extravasation
into the parenchyma of target organs. This thinking points to the
potential importance of integrin-mediated cellECM interactions
in determining the fate of extravasated tumor cells (Fig. 2).
For example, recent studies by others and ourselves have
revealed the essential contribution of integrin 1 -mediated
cellECM interactions and the resulting activation of FAK, a central
mediator of integrin-dependent signaling, in enabling the proliferation of tumor cells following their extravasation into the
lung parenchyma [59,60]. Thus, post-extravasation proliferation
in the lung parenchyma of colonization-competent D2A1 mouse
mammary carcinoma cells is diminished by knocking down the
expression of either integrin 1 or FAK in these cells [59].
Our subsequent analyses show that the related, D2.0R and D2.1
colonization-decient mouse mammary carcinoma cells exhibit
lower levels of FAK activation after entering the lung parenchyma,
although these cells display levels of integrin 1 and FAK expression
comparable to those observed in the colonization-competent D2A1
cells. Moreover, the patterns of integrin 1 distribution are different between colonization-competent D2A1 cells and the other
two colonization-decient cells; thus, only D2A1 cells, but not the

other two cell types, developed abundant integrin 1 -containing

adhesion plaques of elongated morphology within the parenchyma
of the lungs (T.S. and R.A.W., manuscript in preparation). This,
together with the functional connection between elongated adhesion plaque formation and FAK activation, suggested that the
mechanism enabling the assembly of these adhesion plaques is also
likely to govern the proliferation of cancer cells that have already
extravasated into the parenchyma of target organs, doing so by
regulating the activation of FAK.
Cell-surface receptors of ECM other than integrins, such as CD44
a receptor of the ECM component hyaluronan also play important roles in determining the fate of extravasated tumor cells. Thus,
the overexpression of a dominant-inhibitory, soluble isoform of
CD44 (sCD44) in extravasated TA3/St mouse mammary carcinoma
cells resulted in a remarkable increase in their rate of apoptosis
within the lung parenchyma [61]. Interestingly, the overexpression
of sCD44 does not inhibit the adhesion of these cells to the lung
endothelium or their penetration to the parenchyma of the lungs;
this indicated the specic role of the CD44hyaluronan interactions
in promoting the post-extravasation survival of these tumor cells.
The ECM also functions as a reservoir of multiple growth
factors, such as transforming growth factor beta (TGF-), bone morphogenic proteins (BMPs) and vascular endothelial growth factor
(VEGF) [62]. These growth factors, some of which become available to tumor cells upon ECM processing by proteases like matrix
metalloproteinases (MMPs), may also contribute to the proliferation and/or survival of extravasated tumor cells. These diverse

104

T. Shibue, R.A. Weinberg / Seminars in Cancer Biology 21 (2011) 99106

lines of evidence illustrate the critical role of the ECM within the
parenchyma of target organs in controlling the fate of extravasated
tumor cells.
3.2. Interactions with the host cells residing in the target organ
parenchyma
In addition complex ECM interactions, as detailed above, disseminated tumor cells must also cope with a variety of cells that
reside within these sites (Fig. 2). In particular, cells of the immune
system have both supportive and inhibitory effects on the process
of colonization [63]. Early clinical experience with organ transplantation revealed cases of the outgrowth of donor-derived tumor cells
within the bodies of recipients shortly after the transplantation
[64]. In most of these cases, the donors had a history of potentially
metastatic cancer in other organs; the formation of donor-derived
tumor was therefore attributable to the minimal metastatic disease within the donated organs whose outgrowth was permitted by
the immune suppression of the transplant recipients. These cases
reveal the role of immune surveillance in suppressing the formation of macroscopic metastases after tumor cells have established
residence within distant organ sites.
In fact, several distinct types of immune cells exert inhibitory
effects on metastatic colonization in experimental models. For
example, Killion and Fidler demonstrated that macroscopic metastasis formation by B16BL6 mouse melanoma cells in the lungs was
efciently blocked by systemically activating macrophages via the
intravenous administration of liposome-encapsulated macrophage
activators; in these experiments, the liposomes were introduced
after the melanoma cells had taken up residence within the lung
parenchyma [65], indicating that activated macrophages blocked
the post-extravasation metastatic progression. In the RET.AAD
model of mouse melanoma, the depletion of CD8+ cytotoxic T cells
resulted in the more rapid outgrowth of metastases in multiple visceral organs, including lungs, liver and bladder, without affecting
the efciency of initial melanoma cell seeding to these organs [66].
In the liver metastasis model of mouse brosarcoma L929 cells,
the cytotoxic effect of NK cells eliminated the formation of macroscopic metastases via a mechanism involving TRAIL (tumor necrosis
factor-related apoptosis-inducing ligand) expressed on the surface
of NK cells [67]. In contrast to these, certain types of immune cells,
such as macrophages, may also have supportive effects on the process of colonization [68], possibly by remodeling the environment
of the target organ parenchyma, as discussed below.
Endothelial cells represent yet another important cellular component of the host tissue that can inuence metastatic colonization.
Thus, access to sufcient supplies of oxygen and nutrients is essential for tumors to grow continuously. Since oxygen can only diffuse
a distance of 150200 m from the capillaries, the formation of
new blood vessels (i.e., angiogenesis) is critical for tumors to grow
beyond a certain size (generally 12 mm3 ); this applies to both primary tumors and secondary tumors (metastases). A study using
human liposarcoma cells has revealed that cancer cells within a
primary tumor exhibit heterogeneity in their angiogenic activities
[69]. This, taken together with the largely monoclonal origin of
metastases [35], indicates that tumor angiogenesis, a process that
is mediated by direct and indirect interactions between tumor cells
and endothelial cells or their precursors [70], can be rate-limiting
for the outgrowth of metastases.
Other organ-specic cell types can also exert considerable
inuence on the fate of extravasated tumor cells. This is well documented in the case of breast cancer metastasis to the bone. Thus,
these metastases to bone are usually osteolytic, a response that is
mediated by the production of osteoclast-activating factors, such
as parathyroid hormone-related protein (PTHrP), interleukin-1, 6
and 11, and granulocyte-macrophage colony stimulating factor

(GM-CSF), by the tumor cells [7173]. Activated osteoclasts, in

turn, facilitate the outgrowth of metastases by releasing bonederived growth factors, such as transforming growth factor beta
(TGF-) and insulin-like growth factor 1 (IGF-1) [74,75]. In contrast, prostate cancer metastases to the bone usually stimulate
bone formation, which is mediated by the production by the tumor
cells of osteoblast-activating factors, such as endothelin-1, bone
morphogenic proteins (BMPs), and platelet-derived growth factor
(PDGF). Activated osteoblasts may, in turn, support the proliferation and survival of tumor cells by releasing growth factors such
as IGF-1 [76]. Hence, multiple distinct host cell types have a major
impact on the fate of extravasated tumor cells.
3.3. Processing the microenvironment of target organ
parenchyma by tumor cells
In addition to interacting with pre-existing components of the
target organ parenchyma, active remodeling by tumor cells of the
microenvironment within the target organs appears to play a key
role in their post-extravasation behavior. In some cases, tumor
cells growing in the primary site secrete factors that can directly
or indirectly alter the microenvironments of distant sites prior to
the dissemination of tumor cells, thereby preparing these sites
for the subsequent arrival of tumor cells. Such altered microenvironments of the distant sites are refereed to as pre-metastatic
niches. Thus, mice bearing subcutaneous Lewis lung carcinoma
(LLC) tumors or B16 melanoma tumors exhibit accumulations of
bone-marrow derived cells in the parenchyma of distant organs,
notably the lungs, prior to the appearance of disseminated tumor
cells in these sites [77,78]. The types of cells reported to accumulate to these distant sites include VEGFR1+ hematopoietic precursor
cells and Mac-1+ myeloid cells, and experiments have shown that
these cells facilitate the subsequent development of metastases at
these sites [77,78].
Another mechanism reported to enable the microenvironmental remodeling of target organs includes the secretion of lysyl
oxidase (LOX) into the circulation by orthotopoically implanted
MDA-MB-231 breast cancer cells; this LOX supports metastasis
development by crosslinking collagen IV in the basement membrane of the lungs and subsequently recruiting Mac-1+ myeloid
cells to the lung tissue prior to the dissemination of tumor cells
[79]. Moreover, a recent study in our group showed that a secreted
glycoprotein osteopontin released from certain primary tumors
can contribute to the outgrowth of metastases through a mechanism involving the enhanced mobilization of bone marrow cells;
in this case, the mobilized bone marrow cells appear to instigate
the growth of already-established but otherwise-indolent tumors
at distant sites rather than priming the microenvironment prior to
the arrival of disseminated tumor cells [80].
4. Concluding remarks
Since the proposal of seed and soil hypothesis by Stephen
Paget in 1889 [81], it has been recognized that interactions between
tumor cells and host components have a major impact on the progression of metastases. Similarly, the idea of invasion-metastasis
cascade, i.e., the stepwise progression of metastases, was proposed
more than 30 years ago [82]. However, the molecular mechanisms
underlying the tumorhost interactions associated with each steps
of the invasion-metastasis cascade still remain largely elusive.
Attempts to understand tumor cell dissemination and metastasis formation at the molecular level have been thwarted, in part,
by the difculties in experimentally accessing and manipulating
the in vivo conditions in which metastasis occur as well as the
lack of in vitro model systems that closely recapitulate the com-

T. Shibue, R.A. Weinberg / Seminars in Cancer Biology 21 (2011) 99106

plex conditions within the microenvironments of living tissues. In

light of recent technical progresses in both in vitro culture, such as
the development of three-dimensional tissue culture systems [83],
and in vivo experimental systems, such as advances in intravital
microscopy [84], we anticipate that our understanding of metastasis will progress rapidly in the near future. With such information
in hand, we will nally begin to understand in detail the pathogenic
mechanisms that are responsible for 90% of cancer-associated mortality.
Funding
Breast Cancer Research Foundation (R.A.W.).
Conict of interests
The authors declare that there is no conict of interest.
Acknowledgements
We thank present and former members of Weinberg Laboratory for the productive discussions. T.S. is a recipient of Long-Term
Fellowship from Human Frontier Science Program, Postdoctoral
Fellowship from Japan Society for the Promotion of Science, and
Postdoctoral Fellowship from the Ludwig Fund for Cancer Research.
R.A.W. is an American Cancer Society research professor and a
Daniel K. Ludwig Foundation cancer research professor. This work
was funded by Breast Cancer Research Foundation.
References
[1] Talmadge JE, Fidler IJ. AACR centennial series: the biology of cancer metastasis:
historical perspective. Cancer Res 2010;70:564969.
[2] Klein CA. Parallel progression of primary tumours and metastases. Nat Rev
Cancer 2009;9:30212.
[3] Engel J, Eckel R, Kerr J, Schmidt M, Furstenberger G, Richter R, et al. The process
of metastasisation for breast cancer. Eur J Cancer 2003;39:1794806.
[4] Orr FW, Wang HH, Lafrenie RM, Scherbarth S, Nance DM. Interactions between
cancer cells and the endothelium in metastasis. J Pathol 2000;190:31029.
[5] Gassmann P, Haier J. The tumor cellhost organ interface in the early onset of
metastatic organ colonisation. Clin Exp Metastasis 2008;25:17181.
[6] Miles FL, Pruitt FL, van Golen KL, Cooper CR. Stepping out of the ow: capillary
extravasation in cancer metastasis. Clin Exp Metastasis 2008;25:30524.
[7] Pantel K, Alix-Panabieres C, Riethdorf S. Cancer micrometastases. Nat Rev Clin
Oncol 2009;6:33951.
[8] Hedley BD, Chambers AF. Tumor dormancy and metastasis. Adv Cancer Res
2009;102:67101.
[9] Desgrosellier JS, Cheresh DA. Integrins in cancer: biological implications and
therapeutic opportunities. Nat Rev Cancer 2010;10:922.
[10] Weiss L, Orr FW, Honn KV. Interactions of cancer cells with the microvasculature during metastasis. FASEB J 1988;2:1221.
[11] Meng S, Tripathy D, Frenkel EP, Shete S, Naftalis EZ, Huth JF, et al. Circulating tumor cells in patients with breast cancer dormancy. Clin Cancer Res
2004;10:815262.
[12] Luzzi KJ, MacDonald IC, Schmidt EE, Kerkvliet N, Morris VL, Chambers AF, et al.
Multistep nature of metastatic inefciency: dormancy of solitary cells after
successful extravasation and limited survival of early micrometastases. Am J
Pathol 1998;153:86573.
[13] Steinbauer M, Guba M, Cernaianu G, Kohl G, Cetto M, Kunz-Schughart LA,
et al. GFP-transfected tumor cells are useful in examining early metastasis in
vivo, but immune reaction precludes long-term tumor development studies in
immunocompetent mice. Clin Exp Metastasis 2003;20:13541.
[14] Schluter K, Gassmann P, Enns A, Korb T, Hemping-Bovenkerk A, Holzen J, et al.
Organ-specic metastatic tumor cell adhesion and extravasation of colon carcinoma cells with different metastatic potential. Am J Pathol 2006;169:106473.
[15] Simon SI, Green CE. Molecular mechanics and dynamics of leukocyte recruitment during inammation. Annu Rev Biomed Eng 2005;7:15185.
[16] Dimitroff CJ, Lechpammer M, Long-Woodward D, Kutok JL. Rolling of human
bone-metastatic prostate tumor cells on human bone marrow endothelium
under shear ow is mediated by E-selectin. Cancer Res 2004;64:52619.
[17] Haier J, Korb T, Hotz B, Spiegel HU, Senninger N. An intravital model to monitor
steps of metastatic tumor cell adhesion within the hepatic microcirculation. J
Gastrointest Surg 2003;7:50714 [discussion 5145].
[18] Ito S, Nakanishi H, Ikehara Y, Kato T, Kasai Y, Ito K, et al. Real-time observation of
micrometastasis formation in the living mouse liver using a green uorescent
protein gene-tagged rat tongue carcinoma cell line. Int J Cancer 2001;93:2127.

105

[19] Glinskii OV, Huxley VH, Glinsky GV, Pienta KJ, Raz A, Glinsky VV. Mechanical
entrapment is insufcient and intercellular adhesion is essential for metastatic
cell arrest in distant organs. Neoplasia 2005;7:5227.
[20] Garofalo A, Chirivi RG, Foglieni C, Pigott R, Mortarini R, Martin-Padura I, et al.
Involvement of the very late antigen 4 integrin on melanoma in interleukin
1-augmented experimental metastases. Cancer Res 1995;55:4149.
[21] Okahara H, Yagita H, Miyake K, Okumura K. Involvement of very late activation antigen 4 (VLA-4) and vascular cell adhesion molecule 1 (VCAM-1) in
tumor necrosis factor alpha enhancement of experimental metastasis. Cancer
Res 1994;54:32336.
[22] Lehr JE, Pienta KJ. Preferential adhesion of prostate cancer cells to a human
bone marrow endothelial cell line. J Natl Cancer Inst 1998;90:11823.
[23] Scott LJ, Clarke NW, George NJ, Shanks JH, Testa NG, Lang SH. Interactions of
human prostatic epithelial cells with bone marrow endothelium: binding and
invasion. Br J Cancer 2001;84:141723.
[24] Brown DM, Ruoslahti E. Metadherin, a cell surface protein in breast tumors that
mediates lung metastasis. Cancer Cell 2004;5:36574.
[25] Muller A, Homey B, Soto H, Ge N, Catron D, Buchanan ME, et al. Involvement of chemokine receptors in breast cancer metastasis. Nature 2001;410:
506.
[26] Murakami T, Cardones AR, Finkelstein SE, Restifo NP, Klaunberg BA, Nestle FO,
et al. Immune evasion by murine melanoma mediated through CC chemokine
receptor-10. J Exp Med 2003;198:133747.
[27] Hejna M, Raderer M, Zielinski CC. Inhibition of metastases by anticoagulants. J
Natl Cancer Inst 1999;91:2236.
[28] Im JH, Fu W, Wang H, Bhatia SK, Hammer DA, Kowalska MA, et al. Coagulation facilitates tumor cell spreading in the pulmonary vasculature during early
metastatic colony formation. Cancer Res 2004;64:86139.
[29] Konstantopoulos K, Thomas SN. Cancer cells in transit: the vascular interactions
of tumor cells. Annu Rev Biomed Eng 2009;11:177202.
[30] Camerer E, Qazi AA, Duong DN, Cornelissen I, Advincula R, Coughlin SR.
Platelets, protease-activated receptors, and brinogen in hematogenous
metastasis. Blood 2004;104:397401.
[31] Rickles FR, Patierno S, Fernandez PM. Tissue factor, thrombin, and cancer. Chest
2003;124:58S68S.
[32] Kim YJ, Borsig L, Varki NM, Varki A. P-selectin deciency attenuates tumor
growth and metastasis. Proc Natl Acad Sci USA 1998;95:932530.
[33] Liotta LA, Saidel MG, Kleinerman J. The signicance of hematogenous tumor
cell clumps in the metastatic process. Cancer Res 1976;36:88994.
[34] Roberts KE, Hamele-Bena D, Saqi A, Stein CA, Cole RP. Pulmonary tumor
embolism: a review of the literature. Am J Med 2003;115:22832.
[35] Talmadge JE, Wolman SR, Fidler IJ. Evidence for the clonal origin of spontaneous
metastases. Science 1982;217:3613.
[36] Strell C, Entschladen F. Extravasation of leukocytes in comparison to tumor
cells. Cell Commun Signal 2008;6:10.
[37] Muller WA, Randolph GJ. Migration of leukocytes across endothelium and
beyond: molecules involved in the transmigration and fate of monocytes. J
Leukoc Biol 1999;66:698704.
[38] Kramer RH, Nicolson GL. Interactions of tumor cells with vascular endothelial cell monolayers: a model for metastatic invasion. Proc Natl Acad Sci USA
1979;76:57048.
[39] Aird WC. Phenotypic heterogeneity of the endothelium. I. Structure, function,
and mechanisms. Circ Res 2007;100:15873.
[40] Honn KV, Grossi IM, Diglio CA, Wojtukiewicz M, Taylor JD. Enhanced tumor
cell adhesion to the subendothelial matrix resulting from 12(S)-HETE-induced
endothelial cell retraction. FASEB J 1989;3:228593.
[41] Tang DG, Chen YQ, Diglio CA, Honn KV. Protein kinase C-dependent effects of
12(S)-HETE on endothelial cell vitronectin receptor and bronectin receptor. J
Cell Biol 1993;121:689704.
[42] Honn KV, Tang DG, Grossi I, Duniec ZM, Timar J, Renaud C, et al. Tumor cellderived 12(S)-hydroxyeicosatetraenoic acid induces microvascular endothelial
cell retraction. Cancer Res 1994;54:56574.
[43] Uchide K, Sakon M, Ariyoshi H, Nakamori S, Tokunaga M, Monden M. Cancer
cells cause vascular endothelial cell (vEC) retraction via 12(S)HETE secretion; the possible role of cancer cell derived microparticle. Ann Surg Oncol
2007;14:8628.
[44] Offner FA, Schiefer J, Wirtz HC, Bigalke I, Pavelka M, Hollweg G, et al. Tumourcell-endothelial interactions: free radicals are mediators of melanoma-induced
endothelial cell damage. Virchows Arch 1996;428:99106.
[45] Soares FA, Shaughnessy SG, MacLarkey WR, Orr FW. Quantication and morphologic demonstration of reactive oxygen species produced by Walker 256
tumor cells in vitro and during metastasis in vivo. Lab Invest 1994;71:
4809.
[46] Weis S, Cui J, Barnes L, Cheresh D. Endothelial barrier disruption by VEGFmediated Src activity potentiates tumor cell extravasation and metastasis. J
Cell Biol 2004;167:2239.
[47] Lee TH, Avraham HK, Jiang S, Avraham S. Vascular endothelial growth factor
modulates the transendothelial migration of MDA-MB-231 breast cancer cells
through regulation of brain microvascular endothelial cell permeability. J Biol
Chem 2003;278:527784.
[48] Gupta GP, Nguyen DX, Chiang AC, Bos PD, Kim JY, Nadal C, et al. Mediators of
vascular remodelling co-opted for sequential steps in lung metastasis. Nature
2007;446:76570.
[49] Minn AJ, Kang Y, Serganova I, Gupta GP, Giri DD, Doubrovin M, et al. Distinct organ-specic metastatic potential of individual breast cancer cells and
primary tumors. J Clin Invest 2005;115:4455.

106

T. Shibue, R.A. Weinberg / Seminars in Cancer Biology 21 (2011) 99106

[50] Al-Mehdi AB, Tozawa K, Fisher AB, Shientag L, Lee A, Muschel RJ. Intravascular origin of metastasis from the proliferation of endothelium-attached tumor
cells: a new model for metastasis. Nat Med 2000;6:1002.
[51] Zhang Q, Yang M, Shen J, Gerhold LM, Hoffman RM, Xing HR. The role of the
intravascular microenvironment in spontaneous metastasis development. Int
J Cancer 2010;126:253441.
[52] Vaage J, Harlos JP. Spontaneous metastasis from primary C3H mouse mammary
tumors. Cancer Res 1987;47:54750.
[53] Morris VL, Koop S, MacDonald IC, Schmidt EE, Grattan M, Percy D, et al. Mammary carcinoma cell lines of high and low metastatic potential differ not in
extravasation but in subsequent migration and growth. Clin Exp Metastasis
1994;12:35767.
[54] Naumov GN, MacDonald IC, Weinmeister PM, Kerkvliet N, Nadkarni KV, Wilson
SM, et al. Persistence of solitary mammary carcinoma cells in a secondary site:
a possible contributor to dormancy. Cancer Res 2002;62:21628.
[55] Podsypanina K, Du YC, Jechlinger M, Beverly LJ, Hambardzumyan D, Varmus H.
Seeding and propagation of untransformed mouse mammary cells in the lung.
Science 2008;321:18414.
[56] Chambers AF, Groom AC, MacDonald IC. Dissemination and growth of cancer
cells in metastatic sites. Nat Rev Cancer 2002;2:56372.
[57] Guo W, Giancotti FG. Integrin signalling during tumour progression. Nat Rev
Mol Cell Biol 2004;5:81626.
[58] Giancotti FG, Ruoslahti E. Integrin signaling. Science 1999;285:102832.
[59] Shibue T, Weinberg RA. Integrin beta1-focal adhesion kinase signaling directs
the proliferation of metastatic cancer cells disseminated in the lungs. Proc Natl
Acad Sci USA 2009;106:102905.
[60] Barkan D, Kleinman H, Simmons JL, Asmussen H, Kamaraju AK, Hoenorhoff MJ,
et al. Inhibition of metastatic outgrowth from single dormant tumor cells by
targeting the cytoskeleton. Cancer Res 2008;68:624150.
[61] Yu Q, Toole BP, Stamenkovic I. Induction of apoptosis of metastatic mammary
carcinoma cells in vivo by disruption of tumor cell surface CD44 function. J Exp
Med 1997;186:198596.
[62] Hynes RO. The extracellular matrix: not just pretty brils. Science
2009;326:12169.
[63] de Visser KE, Eichten A, Coussens LM. Paradoxical roles of the immune system
during cancer development. Nat Rev Cancer 2006;6:2437.
[64] Gandhi MJ, Strong DM. Donor derived malignancy following transplantation:
a review. Cell Tissue Bank 2007;8:26786.
[65] Killion JJ, Fidler IJ. Therapy of cancer metastasis by tumoricidal activation of
tissue macrophages using liposome-encapsulated immunomodulators. Pharmacol Ther 1998;78:14154.
[66] Eyles J, Puaux AL, Wang X, Toh B, Prakash C, Hong M, et al. Tumor cells disseminate early, but immunosurveillance limits metastatic outgrowth, in a mouse
model of melanoma. J Clin Invest 2010;120:20309.
[67] Takeda K, Hayakawa Y, Smyth MJ, Kayagaki N, Yamaguchi N, Kakuta S,
et al. Involvement of tumor necrosis factor-related apoptosis-inducing lig-

[68]
[69]

[70]
[71]

[72]

[73]

[74]
[75]
[76]
[77]

[78]

[79]

[80]

[81]
[82]
[83]

[84]

and in surveillance of tumor metastasis by liver natural killer cells. Nat Med
2001;7:94100.
Qian BZ, Pollard JW. Macrophage diversity enhances tumor progression and
metastasis. Cell 2010;141:3951.
Achilles EG, Fernandez A, Allred EN, Kisker O, Udagawa T, Beecken WD,
et al. Heterogeneity of angiogenic activity in a human liposarcoma: a proposed mechanism for no take of human tumors in mice. J Natl Cancer Inst
2001;93:107581.
Folkman J. Angiogenesis. Annu Rev Med 2006;57:118.
Yin JJ, Selander K, Chirgwin JM, Dallas M, Grubbs BG, Wieser R, et al. TGF-beta
signaling blockade inhibits PTHrP secretion by breast cancer cells and bone
metastases development. J Clin Invest 1999;103:197206.
Kang Y, Siegel PM, Shu W, Drobnjak M, Kakonen SM, Cordon-Cardo C, et al. A
multigenic program mediating breast cancer metastasis to bone. Cancer Cell
2003;3:53749.
Park BK, Zhang H, Zeng Q, Dai J, Keller ET, Giordano T, et al. NF-kappaB in breast
cancer cells promotes osteolytic bone metastasis by inducing osteoclastogenesis via GM-CSF. Nat Med 2007;13:629.
Roodman GD. Mechanisms of bone metastasis. N Engl J Med
2004;350:165564.
Mundy GR. Metastasis to bone: causes, consequences and therapeutic opportunities. Nat Rev Cancer 2002;2:58493.
Ibrahim T, Flamini E, Mercatali L, Sacanna E, Serra P, Amadori D. Pathogenesis of
osteoblastic bone metastases from prostate cancer. Cancer 2010;116:140618.
Kaplan RN, Riba RD, Zacharoulis S, Bramley AH, Vincent L, Costa C, et al. VEGFR1positive haematopoietic bone marrow progenitors initiate the pre-metastatic
niche. Nature 2005;438:8207.
Hiratsuka S, Watanabe A, Aburatani H, Maru Y. Tumour-mediated upregulation of chemoattractants and recruitment of myeloid cells predetermines lung
metastasis. Nat Cell Biol 2006;8:136975.
Erler JT, Bennewith KL, Cox TR, Lang G, Bird D, Koong A, et al. Hypoxia-induced
lysyl oxidase is a critical mediator of bone marrow cell recruitment to form the
premetastatic niche. Cancer Cell 2009;15:3544.
McAllister SS, Gifford AM, Greiner AL, Kelleher SP, Saelzler MP, Ince TA, et al.
Systemic endocrine instigation of indolent tumor growth requires osteopontin.
Cell 2008;133:9941005.
Paget S. The distribution of secondary growths in cancer of the breast. Lancet
1889;1:5713.
Bross ID, Viadana E, Pickren JW. The metastatic spread of myeloma and
leukemias in men. Virchows Arch A: Pathol Anat Histol 1975;365:91101.
Weigelt B, Bissell MJ. Unraveling the microenvironmental inuences on the
normal mammary gland and breast cancer. Semin Cancer Biol 2008;18:
31121.
Weigert R, Sramkova M, Parente L, Amornphimoltham P, Masedunskas A.
Intravital microscopy: a novel tool to study cell biology in living animals. Histochem Cell Biol 2010;133:48191.