Seminars in Cancer Biology 25 (2014) 15–22

Contents lists available at ScienceDirect

Seminars in Cancer Biology

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

Review

TGF-beta in CAF-mediated tumor growth and metastasis

A. Calon a,∗ , D.V.F. Tauriello a , E. Batlle a,b,∗
a
Oncology Department, Institute for Research in Biomedicine, 08028 Barcelona, Spain
b
Institució Catalana de Recerca i Estudis Avançats (ICREA), 08010 Barcelona, Spain

a r t i c l e i n f o a b s t r a c t

Article history: TGF-beta signaling is one of the major pathways controlling cell and tissue behavior not only in homeo-
Received 23 July 2013 stasis but also in disease. During tumorigenesis TGF-beta orchestrated processes are key due to its dual
Received in revised form role as tumor suppressor and tumor promoter. Important functions of this pathway have been described
19 December 2013
in a context-dependent manner both in epithelial cancer cells and in the tumor microenvironment during
Accepted 30 December 2013
tumor progression. Carcinoma-associated fibroblasts (CAFs) are one of the most abundant stromal cell
types in virtually all solid tumors. CAFs favor malignant progression by providing cancer cells with prolife-
Keywords:
rative, migratory, survival and invasive capacities. A complex network of signaling pathways underlying
TGF-beta
Tumor microenvironment their tumor-promoting properties is beginning to take shape. In this review, we examine current evidence
Cancer-associated fibroblast on the emerging mechanisms involving TGF-beta in CAF-mediated cancer progression, and discuss their
Metastasis potential as therapeutic targets.
Stroma © 2014 Elsevier Ltd. All rights reserved.

1. Introduction in various cell types, during different stages of tissue homeosta-

sis and disease. Nevertheless, many paradoxes, such as the role for
TGF-beta signaling emerged in early animal species and con- TGF-beta signaling both as a tumor suppressor and as a driver of
trols a variety of key functions in tissue development, homeostasis cancer progression, remain the subject of intensive study [6]. Here,
and regeneration [1]. Underlining its importance, dysfunction of we highlight some of the recent literature that focuses on the role
the pathway is linked to severe diseases. Although first described of TGF-beta in tumor progression and metastasis, noting a key role
as a secreted polypeptide hormone capable of rendering recipient for this pathway in landscaping the tumor microenvironment dur-
cells malignant, transforming growth factor-beta was subsequently ing these processes, with an emphasis on its impact over stromal
found to be able to inhibit cell growth, affect differentiation, and fibroblasts.
induce cell death as well. In fact, TGF-beta can bring about disparate
responses depending on the cell type or on the cellular context 2. Metastasis and the tumor microenvironment
[2,3]. In spite of this high level of pleiotropy, TGF-beta signaling
can be simplified as a linear pathway from a ligand-receptor The schematized route from a well-vascularized primary tumor
complex that phosphorylates an intracellular receptor-regulated to metastatic regrowth follows a cascade of discrete events,
effector (R-SMAD proteins SMAD2 or SMAD3), which then com- starting with local invasion of tumor cells through immediate
plexes with a co-SMAD (SMAD4) to enter the nucleus and regulate barriers towards blood or lymphatic vessels. There, cancer cells
target gene transcription with the help of cofactors [2,4,5]. The intravasate and survive the voyage long enough to adhere to and
diversity of outcomes depends on the regulation of these cellular extravasate from those vessels into a distant tissue. Encountering
components and on the prevailing transcriptional cofactors that a foreign microenvironment, invading cells need to colonize this
interact with SMADs, conditions that are subject to various other new location. Initially, micrometastatic nodules are formed, which
signaling pathways. Three decades of study have shed light on the somehow evade host defense. These then grow out into aggressive,
main parameters that modify the outcome of TGF-beta signaling macroscopic tumors [7,8]. The often non-random pattern of organs
targeted for distant colonization by primary tumors has long ago
been acknowledged and led to the famous ‘seed and soil’ hypothe-
sis in 1889, the proposal that metastasis depends on the crosstalk
∗ Corresponding authors at: Institute for Research in Biomedicine (IRB-Barcelona),
between tumor cells and specific organ microenvironments [9]. The
c/Baldiri Reixac 10, 08028 Barcelona, Spain. Tel.: +34 934039008;
fax: +34 934039960.
tumor microenvironment (TME), comprising tumor non-epithelial
E-mail addresses: [email protected] (A. Calon), stroma cells including fibroblasts, endothelial cells and immune
[email protected] (E. Batlle). cells, as well as the extracellular matrix (ECM), is not the same as

1044-579X/$ – see front matter © 2014 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.semcancer.2013.12.008
16 A. Calon et al. / Seminars in Cancer Biology 25 (2014) 15–22

normal tissue stroma. TGF-beta-mediated tumor–stroma interac- properties of cancer cells [35,36], and therefore may constitute a
tions are thought to provide factors that help accomplish the steps niche for this important cell type.
required for metastasis [10–12].
4. TGF-beta signaling in cancer

3. Fibroblasts in cancer progression TGF-beta generally restrains epithelial cells, the origin for a large
number of frequently occurring cancers. Its signaling controls the
TME-resident fibroblasts have been termed with a vari- growth within normal epithelial cell layers in e.g. the skin [37,38]
ety of names including peritumoral (myo-)fibroblasts, reactive and the intestine [39–41]. To overcome this cytostatic activity,
stromal fibroblasts, cancer-associated fibroblasts (CAFs) or tumor- many tumors bear gene alterations that inactivate critical TGF-beta
associated fibroblasts (TAFs). In general, these cells, differing from signaling components [42]. In CRC for instance, TGF-beta receptor 2
normal fibroblasts [13,14], can be identified by expression of (TGFBR2) or SMAD4 are frequently mutated or inactivated by other
smooth muscle actin (alpha-SMA), fibroblast surface protein (FSP1) means, indicating a tumor suppressing role for TGF-beta signaling
and fibroblast-activated protein (FAP) [15]. Several additional CAF [43–45].
markers have been discovered, including NG2 chondroitin sul- Although TGF-beta signaling inhibits epithelial proliferation,
fate proteoglycan (NG2), PDGFR-beta [16,17] or podoplanin (PDPN) this pathway is also frequently linked to tumor progression. In
[18]. The involvement of CAFs in tumor initiation was analyzed colorectal and prostate tumors, increased TGF-beta1 expression
by comparing the effect of normal fibroblasts and CAFs isolated correlates with poor prognosis [46,47]. Indeed, positive TGF-beta
from the primary tumor site. For instance, immortalized prostate immunostaining or elevated mRNA levels in breast and CRC pre-
epithelial cells gave rise to massive tumors in the presence of CAFs, dict metastases [27,48,49]. Moreover, experimental models in the
whereas tumors did not form in the presence of normal fibro- skin, colon and breast support a biphasic function of TGF-beta; i.e. it
blasts [19]. Furthermore, CAFs in non-small lung cell carcinoma inhibits the onset of tumorigenesis in these tissues yet once tumors
(NSCLC) have a greater ability to enhance tumorigenicity of lung are formed, TGF-beta signaling enhances malignant progression
cancer cells than normal fibroblasts. Accordingly, the gene expres- and metastasis [27,50–52]. This differential activation status of
sion signature derived from protumorigenic CAFs shows prognostic TGF-beta signaling during cancer progression depends on whether
value for disease relapse both in NSCLC [14] and in colorectal can- the cell retains or loses a functional pathway. In CRC, epithelial cells
cer (CRC) [20]. This effect is in large part driven by the secretion are markedly less stained for p-SMAD2/3 compared to adjacent
of cytokines and growth factors that affect proliferation, survival, stromal cells or to the epithelial compartment of pre-malignant
adhesion, and migration of cancer cells [21,22]. Mitogenic factors tissue. This indicates a shift from primarily epithelial TGF-beta
secreted by fibroblasts include hepatocyte growth factor (HGF), signaling in normal and adenoma tissues to a stromal and CAFs
EGF family members, chemokine (C-X-C) ligand 12 (CXCL12), sev- signaling in tumors [27,53].
eral fibroblast growth factors (FGFs), and stanniocalcin-1 (STC1) TGF-beta activity is controlled at many levels including the con-
[12,23–26]. In CRC, TGF-beta activated CAFs express several known version of the latent secreted form to its active state. Together
pro-metastatic factors that include angiopoietin-like 4 (ANGPTL4), with its latency-associated peptide (LAP), TGF-beta binds to the
connective tissue growth factor (CTGF) and vascular endothelial latent TGF-beta-binding protein (LTBP), which is part of the
growth factor (VEGFA) [27]. ECM [54]. TGF-beta latency represents a crucial step for active
Perhaps the best-understood function of CAFs is to pro- TGF-beta release. In fact, mice deficient for LTBP/TGF-beta asso-
mote tumor cell invasion. Fibroblasts promote epithelial-to- ciation exhibits inflammation and tumorigenesis associated with
mesenchymal transition (EMT) by secreting either matrix metal- decreased levels of active TGF-beta and decreased TGF-beta
loproteases (MMPs) or cytokines including tumor necrosis factor signaling [55]. In CRC patients, total TGF-beta levels increase dur-
alpha (TNF-␣). TNF-␣ induces stabilization of the Snail protein ing the mucosa-adenoma-carcinoma transition whereas active
through nuclear factor NFkB and Akt signaling pathways which in TGF-beta release is only up regulated in carcinomas where it corre-
turn enhances migratory and invasive phenotype of cancer cells lates with increased alpha-smooth muscle actin (SMA) expression,
[28]. Also, CAFs facilitate the invasiveness of originally non-invasive desmoplastic reaction and poor prognosis [56].
cancer cells [29], possibly through a PAR1-dependent Ca2+ sig-
nals and MMP1 upregulation [30]. More recently, studies on gastric 5. TGF-beta signaling in the TME and during metastasis
cancer development suggested that CAFs could promote cancer
cell migration and invasion via transgelin (TAGLN) upregulation. A large number of breast cancers, melanomas and gliomas retain
TAGLN is an extracellular matrix protein which induces MMP2 pro- a functional pathway and display TGF-beta signaling in tumor
duction [31]. MMPs promote the degradation of ECM and cleave cell cells [2]. These tumors frequently develop genetic alterations in
adhesion molecules such as cadherins, thus potentiating cancer cell downstream tumor suppressor genes such as p15 (CDKN2B), which
motility [28]. bypass the cytostatic effects of TGF-beta signaling. Additional driver
Additional studies suggest a role for activated fibroblasts at the mutations can further abrogate TGF-beta-mediated inhibition of
metastatic site, similar to CAFs in the primary tumor, promoting proliferation. Moreover, residual epithelial TGF-beta signaling can
the survival and proliferation of cancer cells. Formation of liver be rewired to promote the expression of pro-metastatic factors
metastases is associated with the activation of local hepatic stellate such as JAGGED1, angiopoietin-like 4 (ANGPTL4) and interleukin-
cells (HSC), the major cell type involved in liver fibrosis in response 11 (IL11), associated with breast to bone and lung metastasis
to liver damage [32]. In this scenario, PDGF-C promotes tumor [57–59]. In addition, TGF-beta signaling is a key driver of epithelial-
growth by stimulating the proliferation of HSCs [33]. In addition, to-mesenchymal transition (EMT), which favors invasion and
tumors and metastasis derived from engrafted mouse mammary metastatic seeding of cancer cells [60–62].
carcinoma cells are significantly reduced when injected in mice A contrasting scenario involves those tumor types that evaded
deficient for the CAF marker FSP1 [34]. Co-inoculation of wild- TGF-beta-mediated growth arrest by completely abolishing the
type fibroblasts into these mice partially reverses the phenotype, pathway, such as a large proportion of colorectal and pancreatic
providing further evidence for the involvement of metastasis- cancers [42]. The apparent paradox that high levels of TGF-beta
associated fibroblasts in the metastatic process [34]. It has also are nevertheless linked to poor prognosis in these tumors can be
been suggested that CAF-produced factors enhance stem cell-like explained by the fact that the tumor microenvironment remains
 A. Calon et al. / Seminars in Cancer Biology 25 (2014) 15–22 17

responsive to TGF-beta signaling. Using an orthotopic mouse model

for metastatic CRC, we recently demonstrated that the activity of
TGF-beta on stromal cells increases the efficiency of organ colo-
nization by CRC cells whereas mice treated with a pharmacological
inhibitor of TGF-beta receptor 1 (TGFBR1) are resilient to metas-
tasis formation [27]. A complementary finding was obtained with
a model for breast cancer metastasis, in which tumor stem cells
infiltrating the lungs use TGF-beta to instruct the microenviron-
ment and create a metastatic niche [63]. In a different model of
breast cancer carcinogenesis, treatment with low doses of ion-
izing radiation of the precancerous microenvironment prior to
transplantation of Trp53 mutant epithelial cells greatly acceler-
ated development of aggressive tumors, an effect that depended on
TGF-beta signaling in the pre-cancer niche [64]. These observations
support the notion that the TGF-beta activated stroma is crucial at
early time points of the metastatic process to facilitate the colo-
nization phase of cancer cells. This idea is also strengthened by the
finding that upregulation of a TGF-beta target gene program char-
acteristic of stromal cells in CRC patients confers poor prognosis
to CRC patients whereas low stromal TGF-beta signaling robustly
predicts disease-free survival after therapy [27]. Interestingly, the
qualitative characterization of the tumor microenvironment as a
prognostic marker for cancer progression has emerged as an impor-
tant tool for stratification of patients in breast cancer as well [65,66].
Fig. 1. Known origins of cancer-associated fibroblasts (CAFs) in the tumor microen-
TGF-beta-mediated regulation of cancer immunity is a good vironment: resident fibroblasts [74], cancer epithelial cells [80,81] via epithelial to
illustration of the pleotropic effects of this signaling pathway in mesenchymal transition (EMT), endothelial cells [79] and mesenchymal stem cells
the TME. By affecting the proliferation, survival and differentiation [86,87] from the bone marrow, where transdifferentiation depends on TGF-beta
of immune cells, TGF-beta controls the extent of the inflammatory signaling; adipose tissue derived stem cells [78] that differentiate to CAFs through
cancer cell secreted factors including (but not restricted to) TGF-beta.
response [67–70]. In the intestinal mucosa, TGF-beta signaling sup-
presses tumor formation by acting on resident T-cells, as abrogation
of this pathway leads to the expression of the proinflammatory cells may give rise to CAFs through EMT [80,81]. As mentioned
cytokine IL6 that can induce transformation of the epithelium above, the TGF-beta pathway is a main inducer of EMT in these
[71,72]. In contrast, a recent study reports that genetic deletion tumor types. Several studies have also suggested a bone marrow
of Tgfbr2 specifically in myeloid cells significantly inhibits tumor origin for myofibroblasts and CAFs. The bone marrow contains mes-
metastasis, thus suggesting a protumorigenic mechanism of TGF- enchymal stem cells (MSC) that are recruited to the tumor stroma
beta in this context [73]. and are able to enhance breast cancer metastasis to lung [82].
Although high TGF-beta levels activate gene programs in a wide In vivo studies demonstrated the presence of MSCs in the tumor
range of tumor-associated stromal cell types, our in vivo data indi- microenvironment in breast and colon cancer models [83–85]. Sim-
cate that cancer-associated fibroblasts are the main contributors ilar studies using pancreatic and ovarian tumor models showed
to the association of stromal TGF-beta-driven programs with poor that MSCs have the ability to acquire a CAF-like phenotype and
clinical outcome in CRC [27]. promote tumor vascularization and growth [86]. Bone marrow-
derived MSCs can differentiate to hepatic myofibroblasts through
TGF-beta [87]. Inhibition of TGF-beta signaling will in contrast block
6. TGF-beta and the multiple cellular origin of CAFs the differentiation of human MSCs to CAFs and abolish their pro-
tumorigenic effects [88]. The idea of a bone marrow origin of CAFs
CAFs can arise from resident tissue fibroblasts (Fig. 1) where was further strengthened by the work of Quante et al. who stud-
TGF-beta is thought to promote the differentiation of fibroblasts to ied a model of inflammation-induced gastric carcinoma [89]. In
activated fibroblasts [74], a process that involves the chloride intra- this work, the authors found that 20% of CAFs present in tumors
cellular channel 4 (CLIC4) [75]. Genetic ablation of CLIC4 in primary originated from bone marrow MSCs that are recruited in a TGF-
fibroblasts decreases TGF-beta-induced expression of ␣-SMA and beta-dependent manner.
other markers including several ECM components. In turn, CAFs
secrete large amounts of TGF-beta, which amplifies the stromal 7. A TGF-beta dependent crosstalk between cancer cells and
reaction and induces an autocrine signaling loop that maintains fibroblasts
the differentiation of fibroblasts into myofibroblasts [76].
Several studies have proposed alternative cellular origins of Crosstalk between cancer cells and CAFs can occur either
CAFs (Fig. 1) [77]. CAFs may arise from differentiation of other resi- through direct cell–cell contact or mediated by soluble factors.
dent stromal cells such as adipose-derived stem cells (ASCs) during Epithelial tumor cells such as prostate cancer cells secrete sufficient
tumorigenesis. Human adipose tissue derived stem cells (hASCs) amounts of TGF-beta to activate primary fibroblasts [90]. Treatment
can give rise to CAF-like cells when cultured with conditioned with tumor cell-derived conditioned media (CM) induced TGF-beta
media (CM) from different types of breast cancer cell lines [78]. signaling in CAFs and expression of target genes, including pro-
Endothelial cells treated with TGF-beta transdifferentiate to gain teinases that mediate cell invasion. A TGFBR1 inhibitor abolished
a spindle shaped morphology associated with decreased expres- this phenomenon, indicating a strong TGF-beta dependency. Inter-
sion of the endothelial marker CD31. Concomitantly, the expression estingly, CM could induce TGF-beta signaling to a higher extent than
of the fibroblast marker FSP1 is increased suggesting a possible recombinant TGF-beta, suggesting that additional factors secreted
endothelial origin for CAFs [79]. Also, it has been proposed that by cancers cells enhance TGF-beta signaling in fibroblast [53].
in breast cancer and squamous skin carcinoma, epithelial tumor Stuelten et al. showed that the presence of MMP9, predominantly
18 A. Calon et al. / Seminars in Cancer Biology 25 (2014) 15–22

secreted by fibroblasts, correlated with enhanced malignancy in 9. Reconciling the paradoxical roles of TGF-beta signaling
breast cancer [91]. They demonstrated that cancer cells instruct in fibroblasts during tumor progression
stromal fibroblasts to express MMP9 via secretion of TGF-beta. In
addition to their role in degrading basement membrane compo- As discussed throughout this review, CAFs support tumor pro-
nents, MMPs also target and activate TGF-beta itself [92]. These gression in many ways through TGF-beta-dependent mechanisms.
findings support the idea that MMP9 creates a microenvironment For example, forced expression of TGF-beta in the stroma primes
favorable for invasion and metastasis both through activation of fibroblasts to promote the outgrowth of initiated human breast
TGF-beta and by local decomposition of the ECM, adding complex- epithelial cells [107]. Compared to normal fibroblasts, CAFs have a
ity to the activation of TGF-beta pathway in cancer [93,94]. greater ability to enhance tumorigenesis in NSCLC [14]. In this set-
The release of active TGF-beta from its inactive complex not ting, differentially expressed genes are regulated by the TGF-beta
only involves MMP9 but depends also on a variety of additional signaling pathway. Also, high levels of TGF-beta in CRC impose a
factors, among them MMP1, FN1 and integrin alphavbeta6 [54]. pro-metastatic program in the tumor microenvironment [27]. TGF-
Interestingly, CAFs from breast cancers display high MMP1 levels beta activated fibroblasts participate in this process by enhancing
compared with normal fibroblasts [95]; TGF-beta activated colon STAT3 dependent survival of metastatic cells through IL11 produc-
fibroblasts produce FN1 [27]; and integrin alphavbeta6 expres- tion [27]. In sharp contrast to these evidences, a number of genetic
sion correlates significantly with CAF numbers in gastric cancers studies based on conditional inactivation of Tgfbr2 demonstrated
[96]. Taken together, these data suggest that tumor-promoting that loss of TGF-beta in fibroblasts could also promote tumori-
myofibroblasts might both participate to and arise from increased genesis, suggesting a more complex effect of TGF-beta signaling
TGF-beta activation. Furthermore mechanical stress is an addi- on CAFs. Moses and colleagues used a model where Tgfbr2flox/flox
tional way to activate latent TGF-beta when produced by cultured mice are crossed with mice expressing Cre recombinase under
myofibroblasts [97], suggesting that matrix elasticity may also link the control of the FSP1 promoter, which is active in a subset of
TGF-beta activation with CAFs. stromal fibroblasts. Fsp1-driven deletion of Tgfbr2 (Tgfbr2FspKO )
Finally, during disease progression, cancer cells are exposed to resulted in intraepithelial neoplasia in prostate and invasive squa-
CAFs present in the tumor microenvironment and subsequently mous cell carcinoma of the forestomach, both associated with an
to normal fibroblasts as they invade the underlying tissue layers increased abundance of stromal cells. Activation of paracrine HGF
and metastasize. Stuelten et al. studied the transient interac- signaling was identified as one possible mechanism for stimulation
tion occurring between tumor cells and normal fibroblasts. They of epithelial proliferation in this experimental setting [108]. Using
demonstrated that cancer cells induced secretion of TGF-beta by the Tgfbr2FspKO mouse model, a subsequent study showed that the
normal fibroblasts, in turn enhancing the malignant behavior of loss of Tgfbr2 in fibroblasts enhances HGF signaling through c-Met
tumor cells in vivo. Interestingly, even a short interaction between and Ron, which promotes scattering and invasion of mammary car-
tumor cells and normal fibroblasts was sufficient to maintain these cinoma cells [109]. These studies indicate a tumor-suppressive role
TGF-beta cell-targeted effects [98]. for TGF-beta signaling in fibroblasts, in part by suppressing HGF
signaling between mammary fibroblasts and epithelial cells. Even
the loss of one Tgfbr2 allele in fibroblasts (Fsp1-driven heterozy-
8. TGF-beta-driven metabolic reprogramming in CAFs gous deletion of Tgfbr2) was sufficient to reduce tumor latency
and induce stromal cell accumulation and mammary carcinoma
Cancer cells secrete oxidative stress factors into the microen- cell invasiveness to finally promote lung metastasis [110]. Further,
vironment, which can induce autophagy, a fundamental catabolic quantitative proteomics was used to characterize the secretome
mechanism ensuring the synthesis, degradation and recycling from Tgfbr2FspKO mammary fibroblasts. This study identified over
of cellular components through the lysosomal machinery [99]. 1000 proteins amongst which CXCL10 was shown to directly stim-
Emerging data suggest that the microenvironment of some tumors ulate breast tumor cell proliferation and migration [111].
undergo an autophagy-to-senescence transition (AST), which Overall, loss-of-function studies demonstrate that genetic abro-
involves the activation of the TGF-beta pathway in fibroblasts gation of TGF-beta response in fibroblasts triggers a protumorigenic
[100,101]. Autophagic fibroblasts show mitochondrial dysfunction, effect. Thus, the question that arises is how can we conciliate
increased secretory capacity and a senescent phenotype. Further- protumorigenic effects of both activation and loss of the TGF-
more, autophagy in CAFs enhances glycolysis, resulting in elevated beta pathway in fibroblasts? One answer may come from the
production of pyruvate, ketone bodies and l-lactate. These metabo- experiments of Franco et al. They used a xenograft mouse model
lites are utilized by cancer cells for anabolic growth and metastasis where prostate cancer cells were co-injected with a mixed stromal
[102]. Additionally, oxidative stress and autophagy in the tumor population containing normal fibroblasts and fibroblasts unre-
microenvironment are linked to the loss of stromal caveolin-1 sponsive to TGF-beta through expression of a dominant negative
(Cav1), which is associated with early tumor recurrence, metas- TGFBR2 (DNTBRII). Abrogation of TGF-beta signaling in a subset
tasis, and drug resistance in breast cancer [103]. A recent study of fibroblasts enhanced TGF-beta ligand secretion, and stromal
indicates that activation of TGF-beta signaling in fibroblasts leads heterogeneity resulted in malignant transformation of prostate
to attenuation of Cav1 expression, mediated by oxidative stress epithelial cells [112]. The authors proposed that elevated produc-
[104]. The loss of Cav1 can induce autophagy, elevate aerobic tion of TGF-beta by DNTBRII fibroblasts could impact on fibroblasts
glycolysis and enhance in vivo tumorigenesis in adjacent human with intact TGFBR2 supporting their conversion to myofibrobla-
breast carcinoma cells. Capparelli et al. brought further under- sts. In this study, overexpression of TGF-beta in benign human
standing to this process by identifying the CTGF, a TGF-beta target prostate fibroblasts induced in vivo malignant transformation of
gene, as an inducer of AST and aerobic glycolysis in CAFs [105]. prostate epithelial cells. These findings link the protumorigenic
Finally, and in line with the importance of glycolysis in CAFs, over- effect driven by Tgfbr2FspKO fibroblasts to a collateral amplification
expression of migration stimulating factor (MSF), an isoform of of the stromal TGF-beta response. An alternative answer to this
fibronectin, is sufficient to confer myofibroblast differentiation via problem comes from a very recent study by Achyut and colleagues.
increased TGF-beta signaling. This mechanism induces metabolic The authors investigated the molecular mechanisms responsible
reprogramming of CAFs toward a glycolytic phenotype depend- for the development of invasive squamous cell carcinoma (SCC) in
ing on NFkB, and these MSF-expressing fibroblasts significantly Tgfbr2FspKO mice. They observed increased expression of inflam-
enhance tumor growth [106]. matory mediators, such as inducible nitrogen oxide synthetase
 A. Calon et al. / Seminars in Cancer Biology 25 (2014) 15–22 19

Table 1
Therapeutic outcome of clinical trials targeting TME/TGF-beta in patients.

Drug Type Target Disease application Clinical trial Results Refs

Bevacizumab Humanized VEGFA Glioblastoma, Phase II/III Slow metastatic [115]

monoclonal antibody colorectal, lung and disease progression
renal cancers
TKI (e.g. Sorafenib, Small molecule VEGFRs PDGF-Rs Renal cell carcinoma, Phase II/III No or modest response [116]
Sunitinib) inhibitors colorectal cancer, and
hepatocellular
carcinoma
MPIs (Marimastat, MMP inhibitors MMPs Pancreatic, prostate Phase II/III No survival benefit [117–119]
Prinomastat, and lung cancer
Tanomastat)
Sibrotuzumab Humanized FAP Colorectal cancer Phase I/II No result in disease [123,124]
monoclonal remission
unconjugated mAbs
LY2157299 Small molecule TGFBR1 Glioblastoma Phase I/II Antitumor effects with [130]
inhibitor durable responses in
some patients
Trabedersen Specific antisense TGF-beta2 Glioma, colorectal Phase I/II Significant beneficial [131,132]
oligo-deoxynucleotide cancer, pancreatic effects
carcinoma, malignant
melanoma
Belagenpumatucel-L Antisense TGF-beta2 Non-small-cell lung Phase II/III Survival advantage [130]
gene-modified carcinoma
allogeneic tumor cell
vaccine
Fresolimumab Humanized Pan-TGF-beta Advanced renal cell Phase I/II Shrinkage of liver [131]
pan-TGF-beta carcinoma, malignant metastases in some
monoclonal melanoma patients
neutralizing antibodies

The table summarizes clinical studies with results on therapeutic outcome for treatment targeting TGF-beta, TGF-beta activated fibroblasts and related factors discussed in
this review, and their phases, taken from clinicaltrials.gov.

(NOS2), cyclooxygenase-2 (COX2) and NFkB subunit (p65), in the growth factors are currently being targeted using small molecule
epithelium and stroma of the Tgfbr2FspKO mice. In addition, these inhibitors against their receptors including PDGF-Rs and VEGFRs
SCC tumors contained higher number of CD45+ leukocytes com- [116]. Also, various MMP inhibitors reached phase III clinical trials
pared to control mice, suggesting indirect effect on inflammation for pancreatic, prostate and lung cancer [117–119], yet no single
due to loss of Tgfbr2 in fibroblasts [113]. The infiltrating CD45+ treatment has shown efficacy so far (Table 1).
population included myeloid-derived suppressor cells (MDSCs) A different approach consists in targeting CAFs directly by
and TH17 cells. Both cell types are involved in the generation of means of specific surface marker genes. Elevated expression of
immunosuppressive environment that facilitates tumor initiation some of CAF markers predicts poor prognosis in patients and thus
and progression [67]. Epithelial tumor cells of the forestomach they are linked to disease progression. Examples are FAP [120];
in Tgfbr2FspKO mice showed COX2-dependent downregulation of PDPN [18] and PALLADIN: [121]. FAP was proposed as a potential
the expression of cell cycle inhibitors such as CDKN1A. Indeed, target in stromal cells present in human epithelial cancers [122] and
uncontrolled epithelial proliferation and CDKN1A repression in humanized monoclonal antibodies against this CAF surface marker
Tgfbr2FspKO mice were counteracted by COX2 inhibitors [113]. entered clinical trials. However, an early phase II trial based on an
Therefore, whereas studies based on gain of function experiments immunological approach using unconjugated mAbs failed to show
demonstrate cell-autonomous expression of a protumorigenic TGF- efficacy to treat metastatic CRC (Table 1) [123,124].
beta-driven program in CAFs, genetic ablation of TGF-beta signaling The TGF-beta pathway is increasingly considered as a therapeu-
in fibroblast induces secondary activation of TGF-beta or other tic target not only because its role in cancer cells but also because
inflammatory responses in neighboring wild-type cells that could of its capacity to instruct a protumorigenic program in tumor
indirectly promote cancer initiation. stromal cells [125,126]. A wide variety of therapeutic agents that
block TGF-beta signaling (Table 1) are currently available including
10. Therapeutic targeting of TGF-beta signaling in neutralizing antibodies, soluble receptors and antisense oligonu-
fibroblasts cleotides against TGF-beta ligands and small molecule inhibitors
targeting the function of TGFBRI or TGFBRII receptors [126]. One of
CAFs and the TME provide many key functions in cancer pro- these TGFBRI kinase inhibitors developed by Eli Lilly, LY2157299,
gression and metastasis and are therefore attractive therapeutic is used in phase II trials for treatment of glioblastoma multi-
targets. In addition, products secreted by these cells such as ECM forme, hepatocellular carcinoma and advanced pancreatic cancer.
components form rigid structures that can attenuate the diffusion Recently, we found that mice treated with this inhibitor are resilient
of anticancer agents. to metastasis formation by colon cancer cells due to an inhibition of
The most straightforward therapeutic approach is to inhibit stromal activation by TGF-beta, suggesting these inhibitors could
signals emanating from the TME required for tumor growth and benefit those patients at risk of developing metastatic disease.
progression. A prime example is the targeting of secreted stro- Importantly, CRC cells used in these experiments carry inactivat-
mal factors such as VEGFA to inhibit angiogenesis [114]. The ing mutations in TGF-beta pathway components implying that they
monoclonal antibody (mAb) bevacizumab has been approved for were not targeted by the inhibitor [27]. In contrast, inhibition of
first-line use in various cancers, including CRC, lung cancers, renal TGF-beta signaling in CRC cells with an intact pathway could exac-
cancers and glioblastoma multiforme, and has proven to slow erbate tumor growth by releasing cells from the cytostatic effect
metastatic disease progression [115]. Similarly, stroma-produced imposed by TGF-beta. This illustrates that in-depth knowledge of
20 A. Calon et al. / Seminars in Cancer Biology 25 (2014) 15–22

the response of different tumors to TGF-beta is vital to fine-tune [5] Chaffer CL, Weinberg RA. A perspective on cancer cell metastasis. Science
personalized treatments based on inhibition of this pathway. 2011;331(6024):1559–64.
[6] Ikushima H, Miyazono K. TGFbeta signalling: a complex web in cancer pro-
Looking at the bigger picture, exploration of the crosstalk gression. Nat Rev Cancer 2010;10(6):415–24.
between cancer cells and CAFs through TGF-beta could suggest [7] Talmadge JE, Fidler IJ. AACR centennial series: the biology of cancer metasta-
additional therapeutic targets. A good example is IL11, a megakary- sis: historical perspective. Cancer Res 2010;70(14):5649–69.
[8] Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell
opoietic cytokine [127,128]. Our recent investigation highlighted a 2011;144(5):646–74.
pro-metastatic effect of IL11, which is secreted by TGF-beta stim- [9] Paget S. The distribution of secondary growths in cancer of the breast. Lancet
ulated CAFs in CRC. We showed that IL11 promotes the survival of 1889;133:3.
[10] Fidler IJ. The pathogenesis of cancer metastasis: the ‘seed and soil’ hypothesis
tumor cells at the sites of metastatic colonization [27]. These find-
revisited. Nat Rev Cancer 2003;3(6):453–8.
ings should prompt reconsideration of the current use of IL11 to [11] Stover DG, Bierie B, Moses HL. A delicate balance: TGF-beta and the tumor
counteract thrombocytopenia caused by chemotherapy and eval- microenvironment. J Cell Biochem 2007;101(4):851–61.
[12] Hanahan D, Coussens LM. Accessories to the crime: functions of cells recruited
uation of anti-IL11 therapies against metastatic CRCs. Blockade of
to the tumor microenvironment. Cancer Cell 2012;21(3):309–22.
several other factors which expression is induced by TGF-beta in [13] Bauer M, Su G, Casper C, He R, Rehrauer W, Friedl A. Heterogeneity of gene
CAFs during metastatic colonization could also be beneficial to pre- expression in stromal fibroblasts of human breast carcinomas and normal
vent dissemination of the disease. breast. Oncogene 2010;29(12):1732–40.
[14] Navab R, Strumpf D, Bandarchi B, Zhu CQ, Pintilie M, Ramnarine VR,
et al. Prognostic gene-expression signature of carcinoma-associated fibro-
blasts in non-small cell lung cancer. Proc Natl Acad Sci USA 2011;108(17):
11. Conclusion and outlook
7160–5.
[15] Ostman A, Augsten M. Cancer-associated fibroblasts and tumor growth
TGF-beta is considered a central player during the process of – bystanders turning into key players. Curr Opin Genet Dev 2009;19(1):
tumorigenesis as it exerts various context-dependent effects on all 67–73.
[16] Sugimoto H, Mundel TM, Kieran MW, Kalluri R. Identification of fibro-
cells that retain a functional pathway in the tumor. A special role is blast heterogeneity in the tumor microenvironment. Cancer Biol Ther
recognized for TGF-beta-activated CAFs one of the most abundant 2006;5(12):1640–6.
stromal cell types in virtually all solid tumors. TGF-beta activated [17] Brentnall TA, Lai LA, Coleman J, Bronner MP, Pan S, Chen R. Arousal of
cancer-associated stroma: overexpression of palladin activates fibroblasts to
CAFs are emerging as a new target for biological therapies in various promote tumor invasion. PLoS One 2012;7(1):e30219.
cancer settings. [18] Schoppmann SF, Jesch B, Riegler MF, Maroske F, Schwameis K, Jomrich G,
While our understanding of the complex and sometimes para- et al. Podoplanin expressing cancer associated fibroblasts are associated with
unfavourable prognosis in adenocarcinoma of the esophagus. Clin Exp Metas-
doxical mechanisms involving TGF-beta in the TME continues to tasis 2013;30(4):441–6.
grow, it is vital that we develop preclinical animal models that [19] Olumi AF, Grossfeld GD, Hayward SW, Carroll PR, Tlsty TD, Cunha GR.
recapitulate the stromal reactions observed in human cancers, and Carcinoma-associated fibroblasts direct tumor progression of initiated
human prostatic epithelium. Cancer Res 1999;59(19):5002–11.
that allow genetic analysis. One recent example that takes a step
[20] Herrera M, Islam AB, Herrera A, Martin P, Garcia V, Silva J, et al. Func-
in this direction is a transgenic mouse in which FAP-expressing tional heterogeneity of cancer-associated fibroblasts from human colon
cells can be genetically targeted using inducible activity of the Cre- tumors shows specific prognostic gene expression signature. Clin Cancer Res
2013;19(21):5914–26.
ERt2 recombinase [129]. This model allows for the tracing, isolation,
[21] Desmouliere A, Guyot C, Gabbiani G. The stroma reaction myofibrob-
characterization and ablation of CAFs, as well as the deletion of spe- last: a key player in the control of tumor cell behavior. Int J Dev Biol
cific genes in FAP+ fibroblasts at different stages of tumorigenesis. 2004;48(5–6):509–17.
An alternative ideal model of the future would allow the transplan- [22] De Wever O, Demetter P, Mareel M, Bracke M. Stromal myofibroblasts are
drivers of invasive cancer growth. Int J Cancer 2008;123(10):2229–38.
tation of patient-specific primary cancer cells in order to screen for [23] Matsumoto K, Nakamura T. Hepatocyte growth factor and the Met system as
therapies that successfully target and destroy the whole tumor: a mediator of tumor-stromal interactions. Int J Cancer 2006;119(3):477–83.
both the cancer cells and the stromal cells. [24] Pena C, Cespedes MV, Lindh MB, Kiflemariam S, Mezheyeuski A, Edqvist PH,
et al. STC1 expression by cancer-associated fibroblasts drives metastasis of
colorectal cancer. Cancer Res 2013;73(4):1287–97.
Conflict of interest statement [25] Orimo A, Gupta PB, Sgroi DC, Arenzana-Seisdedos F, Delaunay T, Naeem
R, et al. Stromal fibroblasts present in invasive human breast carcinomas
promote tumor growth and angiogenesis through elevated SDF-1/CXCL12
The authors declare that there are no conflicts of interest. secretion. Cell 2005;121(3):335–48.
[26] Pietras K, Ostman A. Hallmarks of cancer: interactions with the tumor stroma.
Exp Cell Res 2010;316(8):1324–31.
Acknowledgments [27] Calon A, Espinet E, Palomo-Ponce S, Tauriello DV, Iglesias M, Cespedes
MV, et al. Dependency of colorectal cancer on a TGF-beta-driven pro-
gram in stromal cells for metastasis initiation. Cancer Cell 2012;22(5):
We thank Elena Sancho for carefully reading the manuscript 571–84.
and providing excellent comments, Iris Joval for artwork and all [28] Cirri P, Chiarugi P. Cancer-associated-fibroblasts and tumour cells:
members of the Batlle laboratory for support and discussions. This a diabolic liaison driving cancer progression. Cancer Metastasis Rev
2012;31(1–2):195–208.
work has been supported by grants to E.B. from the European [29] Dimanche-Boitrel MT, Vakaet Jr L, Pujuguet P, Chauffert B, Martin MS, Ham-
Research Council (Starting Grant no. 208488) and Consolider pro- mann A, et al. In vivo and in vitro invasiveness of a rat colon-cancer cell line
grams (MICINN), and to A.C. by the Spanish Ministry of Science and maintaining E-cadherin expression: an enhancing role of tumor-associated
myofibroblasts. Int J Cancer 1994;56(4):512–21.
Innovation. D.V.F.T holds a Juan de la Cierva postdoctoral fellow-
[30] Boire A, Covic L, Agarwal A, Jacques S, Sherifi S, Kuliopulos A. PAR1 is a
ship. matrix metalloprotease-1 receptor that promotes invasion and tumorigenesis
of breast cancer cells. Cell 2005;120(3):303–13.
[31] Yu B, Chen X, Li J, Qu Y, Su L, Peng Y, et al. Stromal fibroblasts in the microen-
References vironment of gastric carcinomas promote tumor metastasis via upregulating
TAGLN expression. BMC Cell Biol 2013;14:17.
[1] Huminiecki L, Goldovsky L, Freilich S, Moustakas A, Ouzounis C, Heldin CH. [32] Olaso E, Santisteban A, Bidaurrazaga J, Gressner AM, Rosenbaum J, Vidal-
Emergence, development and diversification of the TGF-beta signalling path- Vanaclocha F. Tumor-dependent activation of rodent hepatic stellate cells
way within the animal kingdom. BMC Evol Biol 2009;9:28. during experimental melanoma metastasis. Hepatology 1997;26(3):634–42.
[2] Massague J. TGFbeta signalling in context. Nat Rev Mol Cell Biol [33] Bandapalli OR, Macher-Goeppinger S, Schirmacher P, Brand K. Paracrine
2012;13(10):616–30. signalling in colorectal liver metastases involving tumor cell-derived
[3] Roberts AB, Anzano MA, Wakefield LM, Roche NS, Stern DF, Sporn MB. Type PDGF-C and hepatic stellate cell-derived PAK-2. Clin Exp Metastasis
beta transforming growth factor: a bifunctional regulator of cellular growth. 2012;29(5):409–17.
Proc Natl Acad Sci USA 1985;82(1):119–23. [34] Grum-Schwensen B, Klingelhofer J, Berg CH, El-Naaman C, Grigorian M,
[4] Massague J. How cells read TGF-beta signals. Nat Rev Mol Cell Biol Lukanidin E, et al. Suppression of tumor development and metastasis forma-
2000;1(3):169–78. tion in mice lacking the S100A4(mts1) gene. Cancer Res 2005;65(9):3772–80.
 A. Calon et al. / Seminars in Cancer Biology 25 (2014) 15–22 21

[35] Weiland A, Roswall P, Hatzihristidis TC, Pietras K, Ostman A, Strell C. [65] Roman-Perez E, Casbas-Hernandez P, Pirone JR, Rein J, Carey LA, Lubet
Fibroblast-dependent regulation of the stem cell properties of cancer cells. RA, et al. Gene expression in extratumoral microenvironment predicts
Neoplasma 2012;59(6):719–27. clinical outcome in breast cancer patients. Breast Cancer Res 2012;14(2):
[36] Vermeulen L, De Sousa EMF, van der Heijden M, Cameron K, de Jong JH, R51.
Borovski T, et al. Wnt activity defines colon cancer stem cells and is regulated [66] Galon J, Costes A, Sanchez-Cabo F, Kirilovsky A, Mlecnik B, Lagorce-Pages C,
by the microenvironment. Nat Cell Biol 2010;12(5):468–76. et al. Type: density, and location of immune cells within human colorectal
[37] Wang XJ, Liefer KM, Tsai S, O’Malley BW, Roop DR. Development of gene- tumors predict clinical outcome. Science 2006;313(5795):1960–4.
switch transgenic mice that inducibly express transforming growth factor [67] Yang L, Pang Y, Moses HL. TGF-beta and immune cells: an important regula-
beta1 in the epidermis. Proc Natl Acad Sci USA 1999;96(15):8483–8. tory axis in the tumor microenvironment and progression. Trends Immunol
[38] Shipley GD, Pittelkow MR, Wille Jr JJ, Scott RE, Moses HL. Reversible inhibition 2010;31(6):220–7.
of normal human prokeratinocyte proliferation by type beta transforming [68] Li MO, Wan YY, Sanjabi S, Robertson AK, Flavell RA. Transforming
growth factor-growth inhibitor in serum-free medium. Cancer Res 1986;46(4 growth factor-beta regulation of immune responses. Annu Rev Immunol
Pt 2):2068–71. 2006;24:99–146.
[39] Barnard JA, Beauchamp RD, Coffey RJ, Moses HL. Regulation of intestinal [69] Liao D, Luo Y, Markowitz D, Xiang R, Reisfeld RA. Cancer associated fibro-
epithelial cell growth by transforming growth factor type beta. Proc Natl Acad blasts promote tumor growth and metastasis by modulating the tumor
Sci USA 1989;86(5):1578–82. immune microenvironment in a 4T1 murine breast cancer model. PLoS One
[40] Moses HL, Yang EY, Pietenpol JA. TGF-beta stimulation and inhibition of cell 2009;4(11):e7965.
proliferation: new mechanistic insights. Cell 1990;63(2):245–7. [70] Grivennikov SI, Greten FR, Karin M. Immunity: inflammation, and cancer. Cell
[41] Siegel PM, Massague J. Cytostatic and apoptotic actions of TGF-beta in homeo- 2010;140(6):883–99.
stasis and cancer. Nat Rev Cancer 2003;3(11):807–21. [71] Becker C, Fantini MC, Schramm C, Lehr HA, Wirtz S, Nikolaev A, et al. TGF-beta
[42] Jakowlew SB. Transforming growth factor-beta in cancer and metastasis. Can- suppresses tumor progression in colon cancer by inhibition of IL-6 trans-
cer Metastasis Rev 2006;25(3):435–57. signaling. Immunity 2004;21(4):491–501.
[43] Markowitz S, Wang J, Myeroff L, Parsons R, Sun L, Lutterbaugh J, et al. Inactiva- [72] Kim BG, Li C, Qiao W, Mamura M, Kasprzak B, Anver M, et al. Smad4
tion of the type II TGF-beta receptor in colon cancer cells with microsatellite signalling in T cells is required for suppression of gastrointestinal cancer.
instability. Science 1995;268(5215):1336–8. Nature 2006;441(7096):1015–9.
[44] Lu SL, Zhang WC, Akiyama Y, Nomizu T, Yuasa Y. Genomic structure of [73] Pang Y, Gara SK, Achyut BR, Li Z, Yan HH, Day CP, et al. TGF-beta
the transforming growth factor beta type II receptor gene and its muta- signaling in myeloid cells is required for tumor metastasis. Cancer discovery
tions in hereditary nonpolyposis colorectal cancers. Cancer Res 1996;56(20): 2013;3:936–51.
4595–8. [74] Evans RA, Tian YC, Steadman R, Phillips AO. TGF-beta1-mediated fibroblast-
[45] Grady WM, Markowitz SD. Genetic and epigenetic alterations in colon cancer. myofibroblast terminal differentiation-the role of Smad proteins. Exp Cell Res
Annu Rev Genomics Hum Genet 2002;3:101–28. 2003;282(2):90–100.
[46] Tsushima H, Kawata S, Tamura S, Ito N, Shirai Y, Kiso S, et al. High levels of [75] Shukla A, Edwards R, Yang Y, Hahn A, Folkers K, Ding J, et al. CLIC4 regu-
transforming growth factor beta 1 in patients with colorectal cancer: associ- lates TGF-beta-dependent myofibroblast differentiation to produce a cancer
ation with disease progression. Gastroenterology 1996;110(2):375–82. stroma. Oncogene 2013. Advanced Online Publication (AOP):18 February,
[47] Wikstrom P, Stattin P, Franck-Lissbrant I, Damber JE, Bergh A. Transforming 2013.
growth factor beta1 is associated with angiogenesis: metastasis, and poor [76] Kojima Y, Acar A, Eaton EN, Mellody KT, Scheel C, Ben-Porath I, et al. Autocrine
clinical outcome in prostate cancer. Prostate 1998;37(1):19–29. TGF-beta and stromal cell-derived factor-1 (SDF-1) signaling drives the evo-
[48] Walker RA, Dearing SJ. Transforming growth factor beta 1 in ductal lution of tumor-promoting mammary stromal myofibroblasts. Proc Natl Acad
carcinoma in situ and invasive carcinomas of the breast. Eur J Cancer Sci USA 2010;107(46):20009–14.
1992;28(2–3):641–4. [77] McDonald LT, LaRue AC. Hematopoietic stem cell derived carcinoma-
[49] Friedman E, Gold LI, Klimstra D, Zeng ZS, Winawer S, Cohen A. High levels associated fibroblasts: a novel origin. Int J Clin Exp Pathol 2012;5(9):
of transforming growth factor beta 1 correlate with disease progression in 863–73.
human colon cancer. Cancer Epidemiol Biomarkers Prev 1995;4(5):549–54. [78] Jotzu C, Alt E, Welte G, Li J, Hennessy BT, Devarajan E, et al. Adipose tissue-
[50] Cui W, Fowlis DJ, Bryson S, Duffie E, Ireland H, Balmain A, et al. TGFbeta1 derived stem cells differentiate into carcinoma-associated fibroblast-like
inhibits the formation of benign skin tumors: but enhances progression to cells under the influence of tumor-derived factors. Anal Cell Pathol (Amst)
invasive spindle carcinomas in transgenic mice. Cell 1996;86(4):531–42. 2010;33(2):61–79.
[51] Tang B, Vu M, Booker T, Santner SJ, Miller FR, Anver MR, et al. TGF-beta [79] Zeisberg EM, Potenta S, Xie L, Zeisberg M, Kalluri R. Discovery of endothelial
switches from tumor suppressor to prometastatic factor in a model of breast to mesenchymal transition as a source for carcinoma-associated fibroblasts.
cancer progression. J Clin Invest 2003;112(7):1116–24. Cancer Res 2007;67(21):10123–8.
[52] Glick AB, Lee MM, Darwiche N, Kulkarni AB, Karlsson S, Yuspa SH. Targeted [80] Petersen OW, Nielsen HL, Gudjonsson T, Villadsen R, Rank F, Niebuhr E, et al.
deletion of the TGF-beta 1 gene causes rapid progression to squamous cell Epithelial to mesenchymal transition in human breast cancer can provide a
carcinoma. Genes Dev 1994;8(20):2429–40. nonmalignant stroma. Am J Pathol 2003;162(2):391–402.
[53] Hawinkels LJ, Paauwe M, Verspaget HW, Wiercinska E, van der Zon JM, van [81] Oft M, Akhurst RJ, Balmain A. Metastasis is driven by sequential elevation of
der Ploeg K, et al. Interaction with colon cancer cells hyperactivates TGF-beta H-ras and Smad2 levels. Nat Cell Biol 2002;4(7):487–94.
signaling in cancer-associated fibroblasts. Oncogene 2014;33:97–107. [82] Phillips RJ, Burdick MD, Hong K, Lutz MA, Murray LA, Xue YY, et al. Circulating
[54] Horiguchi M, Ota M, Rifkin DB. Matrix control of transforming growth factor- fibrocytes traffic to the lungs in response to CXCL12 and mediate fibrosis. J
beta function. J Biochem 2012;152(4):321–9. Clin Invest 2004;114(3):438–46.
[55] Shibahara K, Ota M, Horiguchi M, Yoshinaga K, Melamed J, Rifkin DB. Pro- [83] Wang H, Cao F, De A, Cao Y, Contag C, Gambhir SS, et al. Trafficking mes-
duction of gastrointestinal tumors in mice by modulating latent TGF-beta1 enchymal stem cell engraftment and differentiation in tumor-bearing mice
activation. Cancer Res 2013;73(1):459–68. by bioluminescence imaging. Stem Cells 2009;27(7):1548–58.
[56] Hawinkels LJ, Verspaget HW, van der Reijden JJ, van der Zon JM, Verhei- [84] Kidd S, Spaeth E, Dembinski JL, Dietrich M, Watson K, Klopp A, et al.
jen JH, Hommes DW, et al. Active TGF-beta1 correlates with myofibroblasts Direct evidence of mesenchymal stem cell tropism for tumor and wound-
and malignancy in the colorectal adenoma-carcinoma sequence. Cancer Sci ing microenvironments using in vivo bioluminescent imaging. Stem Cells
2009;100(4):663–70. 2009;27(10):2614–23.
[57] Kang Y, He W, Tulley S, Gupta GP, Serganova I, Chen CR, et al. Breast cancer [85] Shinagawa K, Kitadai Y, Tanaka M, Sumida T, Kodama M, Higashi Y, et al.
bone metastasis mediated by the Smad tumor suppressor pathway. Proc Natl Mesenchymal stem cells enhance growth and metastasis of colon cancer. Int
Acad Sci USA 2005;102(39):13909–14. J Cancer 2010;127(10):2323–33.
[58] Padua D, Zhang XH, Wang Q, Nadal C, Gerald WL, Gomis RR, et al. TGFbeta [86] Spaeth EL, Dembinski JL, Sasser AK, Watson K, Klopp A, Hall B, et al.
primes breast tumors for lung metastasis seeding through angiopoietin-like Mesenchymal stem cell transition to tumor-associated fibroblasts con-
4. Cell 2008;133(1):66–77. tributes to fibrovascular network expansion and tumor progression. PLoS One
[59] Sethi N, Dai X, Winter CG, Kang Y. Tumor-derived JAGGED1 promotes oste- 2009;4(4):e4992.
olytic bone metastasis of breast cancer by engaging notch signaling in bone [87] Yang L, Chang N, Liu X, Han Z, Zhu T, Li C, et al. Bone marrow-
cells. Cancer Cell 2011;19(2):192–205. derived mesenchymal stem cells differentiate to hepatic myofibroblasts by
[60] Xu J, Lamouille S, Derynck R. TGF-beta-induced epithelial to mesenchymal transforming growth factor-beta1 via sphingosine kinase/sphingosine 1-
transition. Cell Res 2009;19(2):156–72. phosphate (S1P)/S1P receptor axis. Am J Pathol 2012;181(1):85–97.
[61] Scheel C, Eaton EN, Li SH, Chaffer CL, Reinhardt F, Kah KJ, et al. Paracrine and [88] Shangguan L, Ti X, Krause U, Hai B, Zhao Y, Yang Z, et al. Inhibition of TGF-
autocrine signals induce and maintain mesenchymal and stem cell states in beta/Smad signaling by BAMBI blocks differentiation of human mesenchymal
the breast. Cell 2011;145(6):926–40. stem cells to carcinoma-associated fibroblasts and abolishes their protumor
[62] Massague J. TGFbeta in Cancer. Cell 2008;134(2):215–30. effects. Stem Cells 2012;30(12):2810–9.
[63] Malanchi I, Santamaria-Martinez A, Susanto E, Peng H, Lehr HA, Delaloye JF, [89] Quante M, Tu SP, Tomita H, Gonda T, Wang SS, Takashi S, et al. Bone marrow-
et al. Interactions between cancer stem cells and their niche govern metastatic derived myofibroblasts contribute to the mesenchymal stem cell niche and
colonization. Nature 2012;481(7379):85–9. promote tumor growth. Cancer Cell 2011;19(2):257–72.
[64] Nguyen DH, Oketch-Rabah HA, Illa-Bochaca I, Geyer FC, Reis-Filho JS, Mao JH, [90] Grubisha MJ, Cifuentes ME, Hammes SR, Defranco DB. A local paracrine and
et al. Radiation acts on the microenvironment to affect breast carcinogenesis endocrine network involving TGFbeta: Cox-2, ROS, and estrogen receptor beta
by distinct mechanisms that decrease cancer latency and affect tumor type. influences reactive stromal cell regulation of prostate cancer cell motility. Mol
Cancer Cell 2011;19(5):640–51. Endocrinol 2012;26(6):940–54.
22 A. Calon et al. / Seminars in Cancer Biology 25 (2014) 15–22

[91] Stuelten CH, DaCosta Byfield S, Arany PR, Karpova TS, Stetler-Stevenson WG, [110] Fang WB, Jokar I, Chytil A, Moses HL, Abel T, Cheng N. Loss of one Tgfbr2
Roberts AB. Breast cancer cells induce stromal fibroblasts to express MMP-9 allele in fibroblasts promotes metastasis in MMTV: polyoma middle T trans-
via secretion of TNF-alpha and TGF-beta. J Cell Sci 2005;118(Pt 10):2143–53. genic and transplant mouse models of mammary tumor progression. Clin Exp
[92] Yu Q, Stamenkovic I. Cell surface-localized matrix metalloproteinase-9 prote- Metastasis 2011;28(4):351–66.
olytically activates TGF-beta and promotes tumor invasion and angiogenesis. [111] Xu BJ, Yan W, Jovanovic B, An AQ, Cheng N, Aakre ME, et al. Quantitative anal-
Genes Dev 2000;14(2):163–76. ysis of the secretome of TGF-beta signaling-deficient mammary fibroblasts.
[93] Lynch CC, Matrisian LM. Matrix metalloproteinases in tumor-host cell com- Proteomics 2010;10(13):2458–70.
munication. Differentiation 2002;70(9–10):561–73. [112] Franco OE, Jiang M, Strand DW, Peacock J, Fernandez S, Jackson 2nd RS, et al.
[94] Sternlicht MD, Werb Z. How matrix metalloproteinases regulate cell behavior. Altered TGF-beta signaling in a subpopulation of human stromal cells pro-
Annu Rev Cell Dev Biol 2001;17:463–516. motes prostatic carcinogenesis. Cancer Res 2011;71(4):1272–81.
[95] Eck SM, Cote AL, Winkelman WD, Brinckerhoff CE. CXCR4 and matrix [113] Achyut BR, Bader DA, Robles AI, Wangsa D, Harris CC, Ried T, et al.
metalloproteinase-1 are elevated in breast carcinoma-associated fibroblasts Inflammation-mediated genetic and epigenetic alterations drive cancer
and in normal mammary fibroblasts exposed to factors secreted by breast development in the neighboring epithelium upon stromal abrogation of TGF-
cancer cells. Mol Cancer Res 2009;7(7):1033–44. beta signaling. PLoS Genet 2013;9(2):e1003251.
[96] Zhuang Z, Zhou R, Xu X, Tian T, Liu Y, Lian P, et al. Clinical significance of integ- [114] Carmeliet P, Jain RK. Molecular mechanisms and clinical applications of angio-
rin alphavbeta6 expression effects on gastric carcinoma invasiveness and genesis. Nature 2011;473(7347):298–307.
progression via cancer-associated fibroblasts. Med Oncol 2013;30(3):580. [115] Lambrechts D, Lenz HJ, de Haas S, Carmeliet P, Scherer SJ. Markers of response
[97] Wipff PJ, Rifkin DB, Meister JJ, Hinz B. Myofibroblast contraction for the antiangiogenic agent bevacizumab. J Clin Oncol 2013;31(9):1219–30.
activates latent TGF-beta1 from the extracellular matrix. J Cell Biol [116] Berz D, Wanebo H. Targeting the growth factors and angiogenesis pathways:
2007;179(6):1311–23. small molecules in solid tumors. J Surg Oncol 2011;103(6):574–86.
[98] Stuelten CH, Busch JI, Tang B, Flanders KC, Oshima A, Sutton E, et al. Transient [117] Coussens LM, Fingleton B, Matrisian LM. Matrix metalloproteinase inhibitors
tumor-fibroblast interactions increase tumor cell malignancy by a TGF-Beta and cancer: trials and tribulations. Science 2002;295(5564):2387–92.
mediated mechanism in a mouse xenograft model of breast cancer. PLoS One [118] Rasanen K, Vaheri A. Activation of fibroblasts in cancer stroma. Exp Cell Res
2010;5(3):e9832. 2010;316(17):2713–22.
[99] Martinez-Outschoorn UE, Whitaker-Menezes D, Pavlides S, Chiavarina B, [119] Dorman G, Cseh S, Hajdu I, Barna L, Konya D, Kupai K, et al. Matrix metallo-
Bonuccelli G, Casey T, et al. The autophagic tumor stroma model of cancer or proteinase inhibitors: a critical appraisal of design principles and proposed
battery-operated tumor growth: a simple solution to the autophagy paradox. therapeutic utility. Drugs 2010;70(8):949–64.
Cell Cycle 2010;9(21):4297–306. [120] Wikberg ML, Edin S, Lundberg IV, Van Guelpen B, Dahlin AM, Rutegard J,
[100] Capparelli C, Chiavarina B, Whitaker-Menezes D, Pestell TG, Pestell RG, et al. High intratumoral expression of fibroblast activation protein (FAP)
Hulit J, et al. CDK inhibitors (p16/p19/p21) induce senescence and in colon cancer is associated with poorer patient prognosis. Tumour Biol
autophagy in cancer-associated fibroblasts: fueling tumor growth via 2013;34(2):1013–20.
paracrine interactions, without an increase in neo-angiogenesis. Cell Cycle [121] Gupta V, Bassi DE, Simons JD, Devarajan K, Al-Saleem T, Uzzo RG, et al. Elevated
2012;11(19):3599–610. expression of stromal palladin predicts poor clinical outcome in renal cell
[101] Capparelli C, Guido C, Whitaker-Menezes D, Bonuccelli G, Balliet R, Pestell TG, carcinoma. PLoS One 2011;6(6):e21494.
et al. Autophagy and senescence in cancer-associated fibroblasts metabol- [122] Garin-Chesa P, Old LJ, Rettig WJ. Cell surface glycoprotein of reactive stromal
ically supports tumor growth and metastasis via glycolysis and ketone fibroblasts as a potential antibody target in human epithelial cancers. Proc
production. Cell Cycle 2012;11(12):2285–302. Natl Acad Sci USA 1990;87(18):7235–9.
[102] Zhao X, He Y, Chen H. Autophagic tumor stroma: mechanisms and roles in [123] Scott AM, Wiseman G, Welt S, Adjei A, Lee FT, Hopkins W, et al. A
tumor growth and progression. Int J Cancer 2013;132(1):1–8. Phase I dose-escalation study of sibrotuzumab in patients with advanced
[103] Sotgia F, Martinez-Outschoorn UE, Howell A, Pestell RG, Pavlides S, Lisanti MP. or metastatic fibroblast activation protein-positive cancer. Clin Cancer Res
Caveolin-1 and cancer metabolism in the tumor microenvironment: markers, 2003;9(5):1639–47.
models, and mechanisms. Annu Rev Pathol 2012;7:423–67. [124] Hofheinz RD, al-Batran SE, Hartmann F, Hartung G, Jager D, Renner C, et al.
[104] Guido C, Whitaker-Menezes D, Capparelli C, Balliet R, Lin Z, Pestell Stromal antigen targeting by a humanised monoclonal antibody: an early
RG, et al. Metabolic reprogramming of cancer-associated fibroblasts by phase II trial of sibrotuzumab in patients with metastatic colorectal cancer.
TGF-beta drives tumor growth: connecting TGF-beta signaling with Warburg- Onkologie 2003;26(1):44–8.
like cancer metabolism and L-lactate production. Cell Cycle 2012;11(16): [125] Lonning S, Mannick J, McPherson JM. Antibody targeting of TGF-beta in cancer
3019–35. patients. Curr Pharm Biotechnol 2011;12(12):2176–89.
[105] Capparelli C, Whitaker-Menezes D, Guido C, Balliet R, Pestell TG, Howell A, [126] Hawinkels LJ, Ten Dijke P. Exploring anti-TGF-beta therapies in cancer and
et al. CTGF drives autophagy: glycolysis and senescence in cancer-associated fibrosis. Growth Factors 2011;29(4):140–52.
fibroblasts via HIF1 activation, metabolically promoting tumor growth. Cell [127] Musashi M, Yang YC, Paul SR, Clark SC, Sudo T, Ogawa M. Direct and synergistic
Cycle 2012;11(12):2272–84. effects of interleukin 11 on murine hemopoiesis in culture. Proc Natl Acad Sci
[106] Carito V, Bonuccelli G, Martinez-Outschoorn UE, Whitaker-Menezes D, Car- USA 1991;88(3):765–9.
oleo MC, Cione E, et al. Metabolic remodeling of the tumor microenvironment: [128] Teramura M, Kobayashi S, Hoshino S, Oshimi K, Mizoguchi H. Interleukin-11
migration stimulating factor (MSF) reprograms myofibroblasts toward lactate enhances human megakaryocytopoiesis in vitro. Blood 1992;79(2):327–31.
production, fueling anabolic tumor growth. Cell Cycle 2012;11(18):3403–14. [129] Kraman M, Bambrough PJ, Arnold JN, Roberts EW, Magiera L, Jones JO, et al.
[107] Kuperwasser C, Chavarria T, Wu M, Magrane G, Gray JW, Carey L, et al. Recon- Suppression of antitumor immunity by stromal cells expressing fibroblast
struction of functionally normal and malignant human breast tissues in mice. activation protein-alpha. Science 2010;330(6005):827–30.
Proc Natl Acad Sci USA 2004;101(14):4966–71. [130] Akhurst RJ, Hata A. Targeting the TGFbeta signalling pathway in disease. Nat
[108] Bhowmick NA, Chytil A, Plieth D, Gorska AE, Dumont N, Shappell S, et al. TGF- Rev Drug Discov 2012;11(10):790–811.
beta signaling in fibroblasts modulates the oncogenic potential of adjacent [131] Connolly EC, Freimuth J, Akhurst RJ. Complexities of TGF-beta targeted cancer
epithelia. Science 2004;303(5659):848–51. therapy. Int J Biol Sci 2012;8(7):964–78.
[109] Cheng N, Chytil A, Shyr Y, Joly A, Moses HL. Transforming growth factor-beta [132] Bogdahn U, Hau P, Stockhammer G, Venkataramana NK, Mahapatra AK, Suri
signaling-deficient fibroblasts enhance hepatocyte growth factor signaling in A, et al. Targeted therapy for high-grade glioma with the TGF-beta2 inhibitor
mammary carcinoma cells to promote scattering and invasion. Mol Cancer trabedersen: results of a randomized and controlled phase IIb study. Neuro
Res 2008;6(10):1521–33. Oncol 2011;13(1):132–42.