The role of cancer-associated fibroblasts (CAFs) has not been previously studied in multiple myeloma (MM). Here, cytofluorimetric analysis revealed higher proportions of bone marrow (BM) CAFs in patients with active MM (both at diagnosis and relapse) compared with patients in remission or those with monoclonal gammopathy of undetermined significance or deficiency anemia (controls). CAFs from MM patients produced increased levels of transforming growth factor-β, interleukin-6, stromal cell-derived factor-1α, insulin-like growth factor-1, vascular endothelial growth factor and fibroblast growth factor-2 and displayed an activated and heterogeneous phenotype, which supported their origin from resident fibroblasts, endothelial cells and hematopoietic stem and progenitor cells via the endothelial–mesenchymal transition as well as mesenchymal stem cells via the mesenchymal transition, as both of these processes are induced by MM cells and CAFs. Active MM CAFs fostered chemotaxis, adhesion, proliferation and apoptosis resistance in MM cells through cytokine signals and cell-to-cell contact, which were inhibited by blocking CXCR4, several integrins and fibronectin. MM cells also induced the CAFs proliferation. In syngeneic 5T33MM and xenograft mouse models, MM cells induced the expansion of CAFs, which, in turn, promoted MM initiation and progression as well as angiogenesis. In BM biopsies from patients and mice, nests of CAFs were found in close contact with MM cells, suggesting a supportive niche. Therefore, the targeting of CAFs in MM patients may be envisaged as a novel therapeutic strategy.
Tumor growth results from complex interactions between stromal cells and tumor cells, in which the former cell type regulates tumor development, progression and drug resistance.1, 2 The stromal microenvironment is characterized by a modified extracellular matrix, enhanced angiogenesis and cells with an activated phenotype, including fibroblasts referred to as ‘activated myofibroblasts’ or ‘cancer-associated fibroblasts’ (CAFs).2 In breast, prostate and pancreatic carcinomas, the number of CAFs is associated with an increased malignancy grade, tumor progression and poor prognosis.3, 4, 5, 6 Similarly, in mouse xenografts, CAFs inoculated with carcinoma cells promote tumor survival, proliferation and invasive behavior.7, 8
CAFs are heterogeneous9, 10, 11 and display phenotypes similar to those of myofibroblasts derived from quiescent fibroblasts that have undergone activation during tissue remodeling in wound healing, fibrosis. CAFs express α-smooth muscle actin (αSMA), fibroblast-specific protein-1 (FSP-1), fibroblast activation protein (FAP) and the platelet-derived growth factor receptor-β (PDGFRβ); in addition, FAP and αSMA are highly expressed by myofibroblasts,10 and FSP-1 is expressed by both resting fibroblasts and myofibroblasts.11 The diverse heterogeneity of CAFs likely results from their different origins,12, 13, 14, 15, 16, 17 as CAFs can arise from resident fibroblasts, bone marrow (BM)-derived progenitor cells14, 15 and cells undergoing the endothelial–mesenchymal transition (EndMT)16 or mesenchymal transition (MT).17 However, the molecular mechanisms by which fibroblasts are recruited to the tumor and undergo subsequent induction by the microenvironment to become mature and activated fibroblasts remain unclear. Mechanical tension is potentially required for the acquisition and maintenance of the myofibroblast phenotype,18 as fibroblasts acquire contractile fibers (‘microfilament bundles’) and express αSMA in response to mechanical stress. During this process, the conversion of physical stimulation into chemical signals (‘mechanotransduction’) leads to the activation of cancer-promoting signals.19, 20 In addition, previous work by Kojma et al.21 demonstrated that transforming growth factor-β (TGFβ) and stromal cell-derived factor-1α (SDF1α) act in autocrine loops to initiate and maintain fibroblast differentiation into myofibroblasts.
In multiple myeloma (MM), plasma cells home to and expand in the BM by establishing pathophysiological interactions with BM stromal cells (BMSCs) and the extracellular matrix, which promote their survival, proliferation and resistance to drugs.22 The existence and the role of CAFs in MM have not been investigated previously. Therefore, we analyzed the phenotype and function of BM CAFs in patients with MM and monoclonal gammopathies of undetermined significance (MGUS). The reciprocal effect of CAFs on MM cell function was studied both in vitro and in vivo using syngeneic 5T33MM and MM xenograft mouse models.
Patients and methods
Fifty-six patients fulfilling the International Myeloma Working Group diagnostic criteria for MM (n=38) and MGUS (n=18) were studied. Twenty-three MM patients had active disease (‘active MM’); of these, 14 patients were newly diagnosed and symptomatic, and 9 patients suffered from relapsed MM. Of the patients with active MM, 13 were men and 10 were women. The age range of this group was 45–88 years with a median of 64.5 years, and the patients were classified according to the Durie and Salmon stage as IIB (n=2), IIIA (n=16) or IIIB (n=5). The remaining 15 patients had complete or partial remission (‘nonactive MM’); of these, nine were men and six were women. The age of these patients ranged from 42 to 81 years, with a median age of 62.3 years. Eleven of the MGUS patients were men and 7 were women. These MGUS patients were between the ages of 38 and 85 years, with a median age of 62.5 years. Eleven patients with anemia due to iron or vitamin B12 deficiency were studied as controls.23 Of the control patients, seven were men and four were women. These patients’ ages ranged from 40 to 84 years, with a median of 63.4 years. The study was approved by the University of Bari Ethics Committee, and all patients provided informed consent according to the Declaration of Helsinki.
BM mononuclear cells were isolated using Ficoll–Hypaque gradient separation from heparinized samples and processed with the appropriate separation procedure according to the cell type, as detailed in the Supplementary Methods. The cell purity (>95%) of all cell populations was determined using the FACScanto II cytofluorimetry system (Becton Dickinson-BD, San Jose, CA, USA). In functional studies, CAFs and fibroblasts were used until the fifth passage of culture.
Cell phenotype analysis
BM CAFs or fibroblasts from control patients were analyzed in whole heparinized aspirates. Cells were resuspended in lysing solution (BD) to lyse erythrocytes under gentle hypotonic conditions. Next, lysates were buffered in cytofix/cytoperm permeabilization buffer (BD) and incubated with a mouse anti-human αSMA-FITC monoclonal antibody (Abcam, Cambridge, UK) and a rabbit anti-human FSP-1 antibody (Sigma-Aldrich, St Louis, MN, USA). Samples were then incubated with anti-rabbit PE- (R&D Systems, Minneapolis, MN, USA) or APC-conjugated (Thermo Scientific, Rockford, IL, USA) antisera. Cell surface antigens were detected using the monoclonal antibodies listed in Supplementary Table S1, acquired by cytofluorimetry (FACScanto II, BD) and analyzed using FACSDiva software (BD). The negative controls included isotype-matched, irrelevant antibodies.
Cell cocultures and cell proliferation
CAFs (5 × 104) were cocultured in Roswell Park Memorial Institute medium (RPMI)-1640 medium (Euroclone, Milan, Italy) with autologous CD138+ plasma cells or RPMI8226 cells at a 1:1 or 10:1 cell ratio for 2 or 5 days. Parallel cocultures were performed using control fibroblasts/RPMI8226 cells. To evaluate the proliferation of each cell population, cocultured cells were labeled according to the manufacturer’s instructions with two viable probes: carboxyfluorescein succinimidyl ester (Invitrogen Corp., Carlsbad, CA, USA) and Dye-eFluor-670 (e-Bioscience, San Diego, CA, USA). For the transwell (0.4 μm pore size; Costar, Cambridge, MA, USA) experiments, CAFs were cultured in the lower chambers of 24-well plates, and MM cells were cultured in the upper chambers. CAFs and MM cells were also cultured individually to evaluate spontaneous proliferation. Samples were analyzed using the FACScanto II system. Results are presented as the percentage of proliferating cells and the ‘relative proliferation index’ by comparing the cell proliferation in coculture with spontaneous proliferation.
Chemotaxis, adhesion and in vitro angiogenesis
Chemotaxis was assessed using the Boyden microchamber assay. Adhesion of patient plasma cells or RPMI8266 cells to fibronectin (FN) and/or CAFs as well as the induction of in vitro angiogenesis (Matrigel, BD) by CAFs were assessed.24
In vivo mouse models
The 5T33MM mouse
The mouse cell line 5T33MM was intravenously injected into 10 syngeneic C57B1/KaLwRij mice (Harlan CPB, Horst, The Netherlands) at 6–8 weeks of age. The development of MM was assessed by protein electrophoresis of the serum samples. When the serum paraprotein concentration achieved a level >10 mg ml−l, mice were killed, and BM from the femurs, tibiae and humeri was flushed. Mononuclear cells were isolated using Lympholyte-M (Cedarlane, Hornby, Canada) gradient centrifugation, resuspended in RPMI-1640 medium supplemented with penicillin/streptomycin, glutamine, minimum essential medium and sodium pyruvate (Gibco, Life Technologies, Gent, Belgium), and were assayed for phenotype using the FACScanto II system. Mice were housed according to the Institutional Animal Care and Use Committee of the University of Brussels.
MM xenograft mice
Twenty female, 6- to 8-week-old non-obese diabetic severe combined immunodeficiency mice (NOD.CB17-Prkdcscid/NCrHsd, Harlan Laboratories, Udine, Italy) were injected subcutaneously into the left flank with RPMI8226 cells (2 or 5 × 106) suspended in 100 μl Hank’s balanced salt solution (Euroclone) and 100 μl Matrigel (BD). Coinjection experiments were performed by admixing active MM CAFs with RPMI8226 cells at a 1:1 ratio. Tumor growth was measured twice weekly, and weights (mg=mm3) were calculated as the (length (mm) × width2(mm2))/2. Mice were killed when the tumor weight was ∼2.5 g. Mice were housed according to the Institutional Animal Care and Use Committee of the University of Bari Medical School.
Matrigel plug assay
A mixture containing growth factor-reduced Matrigel (500 μl) and heparin (40 U/ml) was admixed with CAF conditioned medium (CM), 2 × 105 CAFs from active MM patients resuspended in Hank’s balanced salt solution medium, and either saline solution (negative control) or pro-angiogenic growth factors (200 ng/ml of each fibroblast growth factor (FGF)-2, vascular endothelial growth factor (VEGF) and hepatocyte growth factor; positive control). Matrigel mixtures were injected subcutaneously into the abdominal midline of 6-week-old non-obese diabetic/severe combined immunodeficiency mice. Ten days after implantation, the animals were killed and the Matrigel plugs were collected. Neovascularization was evaluated by immunofluorescence staining using an anti-CD31-PE-conjugated monoclonal antibody (BD-Pharmingen).
Immunohistochemistry and immunofluorescence
A statistical analysis was performed using GraphPad Prism 5 software (La Jolla, CA, USA). The data presented represent the mean±s.d. Results were analyzed using the Wilcoxon signed-rank test. P<0.05 was considered statistically significant. Survival curves were plotted using the Kaplan–Meier method.
Phenotypic analysis of CAFs in MM and MGUS patients
The gated CD45− population of BM CAFs was analyzed by cytofluorimetry (Figure 1a) for the expression of FSP-1, αSMA and FAP (Figures 1b–d). On the basis of FSP-1 and αSMA expression, the following subpopulations were distinguished: FSP-1+αSMA− (P1), FSP-1+αSMA+ (P2) and FSP-1−αSMA+ (P3). Each of these subpopulations was significantly increased in patients with active MM (Figures 1b and c). As a whole, CAFs represented 75±20% of the CD45− cells in active MM patients compared with 18±7 and 33±12% in patients with nonactive MM and MGUS, respectively. In contrast, fibroblasts comprised only 9±5% of the CD45− cells in control patients. Representative dot plots are shown in Figure 1c. Analysis of FAP expression (selectively expressed on activated fibroblasts10, 26) revealed that almost the entire gated αSMA+ subpopulation (P2 plus P3) was FAP+ (70–99%) in active MM patients, whereas the gated FSP-1+ subpopulation (P1) displayed variable proportions of FAP+ cells (20–98%; Figure 1d). These data suggest that the FSP-1+ cell population includes both resting and activated fibroblasts.11 The increased frequency of αSMA+FSP-1+ CAFs in active MM patients was also demonstrated by the immunofluorescence analysis of cells selected with D7-FIB-conjugated (anti-fibroblast) microbeads27 (Figure 1e). Overall, these results suggest that MM activity is associated with the expansion and activation of CAFs in the BM milieu. Indeed, the CM of active MM CAFs contained significantly increased levels of TGFβ (a major differentiation factor for fibroblasts16); SDF1α, interleukin (IL)-6 and insulin-like growth factor (IGF)-1 (key cytokines for MM progression28); and VEGF and FGF-2 (key angiogenic cytokines22, 23) as compared with the CM of CAFs from nonactive MM, MGUS and control patients (Figure 1f). Finally, the expression of specific markers of pericytes (NG2 and PDGFRβ), endothelial cells (ECs; CD31, VEGFR2 and CD144), hematopoietic stem and progenitor cells (HSPCs; CD133) and mesenchymal stem cells (MSCs; CD146 and CD90) was analyzed for CAFs from MM and MGUS patients. CAFs from active MM patients showed phenotypic heterogeneity because of increased percentages of PDGFRβ+ (9–22%), CD31+ (5–25%), VEGFR2+ (3–15%), CD144+ (5–15%), CD133+ (1–3%) and CD90+ (8–21%) cells as compared with CAFs from the other patient groups (Supplementary Figure 1).
EndMT and MT contribute to the formation of CAFs
As active MM CAFs expressed EC markers (CD31 and VEGFR2), we investigated the possibility that these cells originate from ECs via TGFβ-induced EndMT.16 ECs from active MM patients were cultured for 5 days in CM from paired CAFs or RPMI8226 cells producing TGFβ (Figure 1f, Supplementary Figure 2). The addition of TGFβ served as a positive control, whereas cultures lacking TGFβ were used as a negative control. MM ECs treated with CAF CM, RPMI8226 cell CM or TGFβ exhibited significantly increased expression of αSMA (58±9%, 37±7% and 55±11%, respectively, vs 8±4% for the negative control) and FAP (44±6%, 24±9% and 32±8%, respectively, vs 8±5% for the negative control). In contrast, increased expression of FSP-1 (18±7%) was only evident in active MM CAF CM (Figure 2a). The increased expression of αSMA, FAP and FSP-1 was also demonstrated by immunofluorescence (Figure 2b) and real-time reverse transcriptase-PCR (Figure 2c). Similar results, albeit to a lesser extent, were obtained by culturing ECs from MGUS patients with autologous CAF CM for 10 days (Supplementary Figure 3). Next, we examined whether CM from active MM CAFs and RPMI8226 cells could influence the in vitro angiogenesis of MM ECs (Figure 2d). Using a Matrigel assay, MM ECs sprouted and aligned to form branching and anastomosed tubes with multicentric junctions, giving rise to a closely knit capillary plexus (control). However, treatment of MM ECs with TGFβ, CAF CM, and, to a lesser extent, RPMI8226 cell CM reduced the number and thickness of these junctions (Figure 2d), signifying changes in vessel differentiation. Moreover, inhibition of the TGFβ pathway by the addition of the TGFβ-receptor1 kinase inhibitor SD208 to CAF CM reduced αSMA expression (Supplementary Figure 4a) and preserved the in vitro angiogenic activity (Supplementary Figure 4b) of ECs. Overall, our data suggest that CAFs from active MM patients may drive EndMT via TGFβ.
Next, the effect of CAF and RPMI8226 cell CM on CD133+ HSPCs as precursors of ECs was analyzed. HSPCs were isolated from the peripheral blood of mobilized MM patients and grown on FN-coated plates in EC differentiation medium for 14 days.29 Morphologically, the adherent HSPCs were heterogeneous because of the presence of small and large round cells as well as large flat cells (Figure 2e, control). As demonstrated by cytofluorimetry (Figure 2f), HSPCs were CD34+CD31+CD144+VEGFR2+, indicating commitment to the EC lineage, but were also FSP-1−αSMA−FAP− (control). Exposure to active MM CAF CM and TGFβ for 7 days induced a drastic morphological transformation; cells became spindle shaped (Figure 2e) and demonstrated increased expression of FSP-1+ (31±11% for MM CAF CM and 33±9% for TGFβ) and αSMA+ (24±6% for MM CAF CM and 27±8% for TGFβ). These results indicated that CAF CM and TGFβ induced HSPC differentiation into CAF-like cells (Figure 2f), whereas CM from MGUS CAFs and control fibroblasts did not induce these changes (data not shown).
The origin of active MM CAFs from MSCs was also evaluated. BM MSCs from active MM patients were CD146+CD90+CD105+CD45− and CD31−CD34− (ECs and HSPCs markers; Figure 2g), and these cells contained marginal proportions of FSP-1+ (6±3%), αSMA+ (12±6%) and FAP+ (10±5%) cells (Figure 2g, control) that significantly increased after a 7-day culture with CM from autologous CAFs (30±10%, 36±12% and 31±11%, respectively), RPMI8226 cells (23±8%, 30±12% and 29±9%, respectively) and TGFβ (52±15%, 45±12% and 48±14%, respectively; Figure 2g). Immunofluorescence for FSP-1 and αSMA confirmed the finding that active MM CAFs and RPMI8226 cells induced MSC transdifferentiation into CAF-like cells (Figure 2h). As MSCs have a preferential tropism for tumor sites,30 we investigated the migration of MSCs from active MM patients in response to paired CAF and RPMI8226 cell CM, as well as IGF-1 (positive control), and significant migration toward both types of CM was observed (Figure 2i).
Overall, these results suggest that active MM CAFs and MM cells induce ECs and HSPCs (via EndMT) as well as MSCs (via MT) to differentiate into CAFs.
MM cells induce the proliferation of CAFs
Dye-eFluor-labeled RPMI8226 cells were cocultured with carboxyfluorescein succinimidyl ester (CFSE)-labeled CAFs from MM and MGUS patients or fibroblasts from control patients at a ratio of 1:10 or 1:1, either with or without a permeable transwell (Figures 3a, c and d, ‘proliferation in coculture’). CAFs or fibroblasts were also cultured alone (Figure 3b, ‘spontaneous proliferation’). CAF proliferation (Q1+Q3) and FAP expression (Q1+Q2) were evaluated after 5 days (Figures 3b and c). CAFs from active MM patients displayed significantly increased spontaneous proliferation (Figure 3b, Q1+Q3=25±6%; Figure 3d) compared with those from patients with nonactive MM (10±3%) or MGUS (7±4%) as well as fibroblasts from control patients (5±2%). CAFs from active MM patients also exhibited significantly increased percentages of FAP+ cells (Q1+Q2=23±8% vs 11±5% (nonactive MM), 7±3% (MGUS) and <1% (control); Figures 3c and e). RPMI8226 cells enhanced both CAF proliferation and FAP expression (Figures 3c–e), and the relative proliferation index (Figure 3e) showed that the ability of RPMI8226 cells to induce CAF proliferation was independent of the disease state or FAP expression but was dependent on the number of MM cells (higher at 1:1 cell ratio) and cell-to-cell contact (higher without transwell). However, physical separation of CAFs from MM cells by the addition of the transwell did not affect FAP expression.
Active MM CAFs induce MM cell proliferation and prevent apoptosis
Dye-eFluor-labeled CAFs from MM and MGUS patients or fibroblasts from control patients were cocultured with (CFSE)-labeled RPMI8226 cells or CD138+ plasma cells from active MM patients at 1:1 or 10:1 ratio, either with or without a permeable transwell, and proliferation was evaluated after 2 days. CAFs from active MM patients induced robust RPMI8226 cell proliferation, whereas CD138+ plasma cell proliferation was observed only at the 10:1 cell ratio, possibly because these cells were only minimally proliferative in vitro (Figure 4a). Moreover, this stimulatory effect was significantly reduced with the addition of the transwell (Figure 4a), implying a requirement for cell-to-cell contact.
The antiapoptotic effect of CAFs was analyzed in the 1:1 and 10:1 cocultures with autologous CD138+ plasma cells, given that these stroma-dependent cells are more prone to spontaneous apoptosis than stroma-independent RPMI8226 cells. CD138+ cells were also cultured in serum-free medium alone or in the presence of IGF-1 plus IL-6 to prevent apoptosis.28 The percentage of cells undergoing spontaneous apoptosis was 48±8% in serum-free medium, whereas it was only 8±3% in cultures with IGF-1 plus IL-6 (Figure 4b). CAFs from active MM patients displayed an overlapping antiapoptotic effect at both 1:1 (12±8% without transwell; 25±11% with) and 10:1 cell ratios (9±5% without; 15±9% with). The results are expressed as the percentage of apoptosis observed in control cultures (Figure 4c) and demonstrated that CAFs from active MM significantly reduced spontaneous apoptosis. As the antiapoptotic effect was observed both with and without the transwell, cell contact-dependent and cytokine-dependent mechanisms may be involved. Overall, these data indicate that active MM CAFs prevent cell apoptosis, as demonstrated by a reduction in Annexin-V expression (Figure 4b), and drive MM cell proliferation, as measured by (CFSE) (Figure 4a).
SDF1α and integrins mediate cross talk between CAFs and MM cells
The role of SDF1α and its receptor CXCR4 in mediating cross talk between CAFs and MM cells was investigated using AMD3100 (a specific CXCR4 inhibitor). The expression of SDF1α and CXCR4 was previously shown to be elevated in CAFs in breast cancer7 and MM cells, respectively.31 As shown in Figure 1f, significantly increased levels of SDF1α were observed in the CAF CM from active MM patients compared with nonactive MM and MGUS patients. Moreover, CXCR4+ RPMI8226 cells and CD138+ plasma cells from active MM patients migrated toward the CM from paired CAFs and SDF1α cells (positive control, Figure 5a) and adhered to CAFs and FN (Figure 5b), whereas treatment of RPMI8226 cells and CD138+ plasma cells with AMD3100 reduced CXCR4 expression (Figure 5c). As specifically shown for RPMI8226 cells cocultured with CAFs, treatment with AMD3100 also inhibited migration (Figure 5d), adhesion (Figure 5e) and the relative proliferation index (Figure 5f).
As integrins are known to govern interactions between BMSCs and MM cells,32 the role of β1, β3, β7, VLA4, VLA5 and αVβ3 integrins (major FN receptors) as well as cellular FN was analyzed (Figure 6). FN expression was significantly increased in CAFs from active MM patients (70±12%) compared with nonactive MM and MGUS patients (44±7% and 31±9%, respectively; Figure 6a). FN was also highly expressed in CD138+ plasma cells from active MM patients (85±12%) and RPMI8226 cells (55%; Figure 6a). Moreover, treatment of MM cells with anti-integrin and anti-FN blocking antibodies reduced their adhesion to CAFs. Only anti-β3, -β7 and -FN treatment of CAFs inhibited their adhesion to MM cells (Figure 6c), whereas treatment of MM cells with anti-β3, -β7, -VLA4, -αVβ3 and -FN produced inhibition of their adhesion to CAFs (Figure 6b). Overall, these data suggest that the interaction of MM cells with CAFs is mediated by several integrins as well as cell surface FN.
CAFs and MM cells reciprocally sustain their proliferation in vivo
We next investigated whether MM cells support CAF expansion as well as whether CAFs from active MM patients support MM tumor growth in syngeneic 5T33MM mice and tumor initiation in mice bearing MM xenografts. C57BlKaLwRijHsd mice received 0.5 × 106 5T33MM cells intravenously, and after 3 weeks, BM cells from 5T33MM-treated and naive mice were analyzed for CAF content. The 5T33MM mice showed significantly increased proportions of FSP-1+αSMA+CD45− cells compared with naive mice (Figures 7a and b), suggesting that MM cells induced substantial CAFs expansion in vivo.
In addition, non-obese diabetic/severe combined immunodeficiency mice were injected subcutaneously with RPMI8226 cells, CAFs isolated from active MM patients or a 1:1 mixture of 5 × 106 MM cells and CAFs. The coinjection led to accelerated tumor growth, which began 35 days after transplantation (Figure 7c). The tumor weight at the 50th day was significantly higher for mice receiving the coinjection as compared with the RPMI8226 plasmocytoma alone. The role of CAFs in MM tumor initiation was studied in mice injected with 2 × 106 RPMI8226 cells alone or in combination with CAFs at a 1:1 ratio, and the addition of CAFs significantly increased the development of plasmocytomas (Figure 7d). Specifically, four of six (67%) coinjected mice developed plasmocytomas compared with only one of six (17%) mice injected with MM cells alone, and these data led us to hypothesize that CAFs sustain MM tumor growth and initiation.
Sections from RPMI8226 plasmocytoma xenografts both with and without CAFs were stained for CD31 to evaluate the angiogenic effect of CAFs (Figures 7e and f). Enhanced angiogenesis was observed in CAF-coinjected plasmocytomas (Figure 7e), which displayed a microvascular density approximately three times greater than plasmocytomas injected alone (Figure 7f). This effect may be related to increased tumor burden, although experiments with Matrigel plugs revealed that CAFs from active MM patients and their CM were able to attract mouse CD31+ ECs into the plug and induce angiogenesis (Figure 7g). In addition, the CAF CM behaved as a strong chemoattractant in vitro for MM ECs (Figure 7h). Overall, our data suggest that CAFs have a direct angiogenic effect, thereby potentially supporting MM tumor growth, at least partly, by promoting neovascularization.
Immunohistochemistry of CAFs in the BM of MM and MGUS patients and mice
BM sections from active MM patients displayed an increased frequency and more intense staining for FSP-1 (Figure 8a) and αSMA (Figure 8b) cells as compared with those from MGUS patients. Similar images were obtained for the femur biopsies of 5T33MM mice vs naive mice (Figures 8d and e). MM cells (brown cells) were consistently found in close contact with FSP-1+ and αSMA+ CAFs (pink cells) in both MM patients and mice (Figure 8, inserts). In MGUS biopsies, CAFs coexisted in close contact with plasma cells (Figure 8c, left), and no FSP-1- and αSMA-stained cells were observed in the absence of myeloma cells (Figure 8c, right). This finding is in agreement with our in vitro results showing the important role of cell-to-cell contact for MM cells and CAF expansion.
CAFs have been widely studied in solid tumors where they are associated with high-grade histology and poor prognosis,5, 6, 7, 8, 33, 34 but not in hematological tumors such as MM. Here, CAFs were identified according to FSP-1, αSMA and FAP expression in BM samples from patients with MM or MGUS; fibroblasts from patients with deficiency anemia were studied as controls. The highest proportions of CAFs were found in active MM patients, suggesting that the MM activity phase involves the expansion of CAFs. No correlation between the percentage of CAFs and the Durie and Salmon stage was observed, which may be attributed to the following reasons. First, MM cell proliferation, and hence tumor burden, is induced not only by CAFs but also by other BMSCs, such as ECs and osteoclasts.23 Second, CAFs demonstrate an elongated shape, which allows for contact with numerous MM cells (Figure 8), and a single CAF may therefore support the cell contact-induced proliferation of many MM cells. Similar results were obtained in 5T33MM mice, in which the increase in CAFs corresponded with MM development. Further, immunohistochemical analysis of BM from patients and mice confirmed that high levels of CAFs coexist with MM cells.
Similar to solid tumors,10, 11, 12 the CAF population in MM was heterogeneous and expressed cell markers of ECs, HSPCs and MSCs, suggesting their derivation from multiple cell lineages. Although the origin of CAFs is not well understood, one hypothesis suggests that they may derive from resident fibroblasts that differentiate into myofibroblasts by upregulating αSMA in response to signals from tumor cells.35 In our study, MM cells were able to activate existing fibroblasts and recruit them via the secretion of TGFβ. In this regard, MGUS or MM BM sections showed that FSP-1+ and αSMA+ CAFs coexist only with MGUS or MM plasma cells. Previous work has shown that human breast fibroblasts are converted into CAFs in vivo in step with tumor progression via the production of small amounts of TGFβ that prime the TGFβ autocrine signaling pathway.21 Moreover, plasma cells and BMSCs from MM patients secrete high levels of TGFβ,36 and TGFβ blockade suppresses MM cell growth and adhesion to BMSCs,36, 37 implying an important role for TGFβ in MM pathogenesis. CAFs, in turn, transform BM stroma by producing collagen and FN and secreting growth factors (TGFβ, hepatocyte growth factor and IGF-1), cytokines (IL-1 and IL-6) and chemokines (SDF1α).4 In our study, CAFs from patients with active MM produced increased levels of TGFβ, IGF-1, IL-6 and SDF1α compared with CAFs from nonactive MM, MGUS and control patients. Moreover, in vitro experiments using a coculture system have shown that MM cells stimulate CAFs to produce cytokines through chemokine- and cell adhesion-mediated processes (manuscript in preparation). In addition, active MM CAFs have been shown to recruit ECs, HSPCs and MSCs and induce them to undergo EndMT15, 16 and MT.14 Cell transdifferentiation may also increase the CAF component in the MM BM stroma. EndMT has been described in heart development,38 fibrosis and cancer,16, 39 and B16F10 melanomas grown in transgenic mice were shown to induce 30% of ECs to transdifferentiate into CAFs coexpressing CD31 and FSP-1 or αSMA.16 In our study, TGFβ and CM obtained from RPMI8226 cells and active CAFs from MM patients were able to convert ECs and HSPCs into CAF-like cells, in a manner similar to that observed in solid tumors.40 Further, BM MSCs may contribute to the generation of CAFs.41 Indeed, the MSCs of active MM patients migrated toward the CM of RPMI8226 cells and active CAFs from MM patients and transdifferentiated into FSP-1+αSMA+FAP+ cells. Finally, MM cells induced the in vitro activation (via cytokines) and proliferation (via cell-to-cell contact) of CAFs. Overall, our data suggest that MM cells trigger changes in the BM microenvironment in terms of CAF number and activity.
CAFs from active MM patients support MM cell growth by the induction of proliferation by cell-to-cell contact and the prevention of apoptosis by both contact-dependent and -independent (that is, cytokine-mediated) pathways. These data are in agreement with other studies showing that the binding of MM cells to whole BMSCs triggers the secretion of cytokines promoting MM cell growth and survival.42, 43
Studies in animal models have reported that CAFs promote cancer initiation and progression.7, 8, 21 Here, using an in vivo xenograft MM mouse model, animals coinjected with active MM CAFs and RPMI8226 cells showed accelerated tumor growth compared with those injected with RPMI8226 cells alone. Active MM CAFs also supported MM initiation, as mice coinjected with low number of RPMI8226 cells and active MM CAFs developed plasmocytomas, which was likely a consequence of both the antiapoptotic and proliferative effects of CAFs; this response was not observed in mice injected with a low number of RPMI8226 cells alone. Immunohistochemistry of plasmocytoma xenografts and Matrigel plugs showed an angiogenic effect of active MM CAFs; indeed, CAFs may favor angiogenesis through the production of SDF1α, VEGF and FGF-2. Therefore, MM CAFs may contribute to the angiogenic switch and the subsequent angiogenic phase that parallels the transition of MGUS into MM.44
Recently, Ghobrial28 emphasized the role of BM niches in MM initiation and diffusion by introducing the concept of ‘premetastatic’ and ‘metastatic’ niches. It is also plausible that MM cells use systems of cell dissemination similar to those regulating the cell trafficking of hematopoietic stem cells or the metastasis of carcinoma cells, a process that includes cell invasion, blood vessel entry, homing to distant foci as micrometastasis and finally growth to macrometastasis. Metastasis is mediated by growth factors, cytokines and/or exosomes of MM cells that upregulate FN production by resident fibroblasts to promote the metastatic niche. Tentatively, we suggest an initial role for MM cells in the induction of BM micrometastatic niches through a contact-independent, cytokine-mediated conversion of resident fibroblasts into CAFs, thereby allowing CAFs to prepare the metastatic vascular niche and undergo cross talk with MM cells. We found that inhibition of the SDF1α/CXCR4 axis affected MM cell migration, adhesion and proliferation, indicating that MM CAFs recruit CXCR4+ MM cells by SDF1α secretion to promote CAFs-to-MM cell adhesion and the reciprocal enhancement of cell proliferation and survival.
Integrins have a role in the pathophysiology of MM cells by mediating MM cell survival, proliferation, homing into and egress from the BM and cell adhesion-mediated drug resistance.45, 46, 47 Although the intracellular mechanisms mediating these processes have not been fully elucidated, MM cell interactions with BMSCs and extracellular matrix induce cytoskeletal reorganization in MM cells to form ‘adhesion plaques’ via integrin clustering and the phosphorylation of FAK tyrosine kinase, which triggers the RAS/MAP kinase signal transduction cascade.48 Here, using anti-integrin blocking monoclonal antibodies, we found that MM cell adhesion to CAFs is mediated through the reciprocal involvement of β3, β7, VLA4, VLA5 and αVβ3 integrins expressed on MM cells and β3 and β7 integrins expressed on CAFs. Whether these integrins contribute simultaneously or individually through synchronized temporal-spatial expression to trigger the reciprocal activities of MM cells and CAFs remains to be determined. Further, the blockade of FN on MM cells and CAFs inhibited adhesion by 70% and 30%, respectively, suggesting that cellular FN may act as an integrin receptor for the cross talk between MM cells and CAFs. Previous studies have also shown that MM plasma cells produce FN,49, 50 which suggests that bidirectional interactions between CAFs and MM cells mediated by surface FN and β3, β7, VLA4, VLA5 and αVβ3 integrins may occur.
In conclusion, our results highlight an important interplay between CAFs and plasma cells during MM initiation and progression. Cancer cells induce and maintain the CAF-activated phenotype, which, in turn, supports tumor progression by promoting extracellular matrix remodeling, cell proliferation, apoptosis resistance and angiogenesis. As these effects are mediated by cell adhesion molecules, growth factors and chemokines released by both MM cells and CAFs, targeting CAFs in MM can be envisaged as a novel and potentially effective therapeutic approach.
Tlsty TD, Hein PW . Know thy neighbor: stromal cells can contribute oncogenic signals. Curr Opin Genet Dev 2001; 11: 54–59.
Li H, Fan X, Houghton J . Tumor microenvironment: the role of the tumor stroma in cancer. J Cell Biochem 2007; 101: 805–815.
Orimo A, Weinberg RA . Stromal fibroblasts in cancer: a novel tumor-promoting cell type. Cell Cycle 2006; 5: 1597–1601.
Kalluri R, Zeisberg M . Fibroblasts in cancer. Nat Rev Cancer 2006; 6: 392–401.
Franco OE, Shaw AK, Strand DW, Hayward SW . Cancer associated fibroblasts in cancer pathogenesis. Semin Cell Dev Biol 2010; 21: 33–39.
Shimoda M, Mellody KT, Orimo A . Carcinoma-associated fibroblasts are a rate-limiting determinant for tumor progression. Semin Cell Dev Biol 2010; 21: 19–25.
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 Secretion. Cell 2005; 121: 335–348.
Erez N, Truitt M, Olson P, Arron ST, Hanahan D . Cancer-associated fibroblasts are activated in incipient neoplasia to orchestrate tumor-promoting inflammation in an NF-kappaB-dependent manner. Cancer Cell 2010; 17: 135–147.
De Wever O, Demetter P, Mareel M, Bracke M . Stromal myofibroblasts are drivers of invasive cancer growth. Int J Cancer 2008; 123: 2229–2238.
Orimo A, Weinberg RA . Heterogeneity of stromal fibroblasts in tumors. Cancer Biol Ther 2007; 6: 618–619.
Sugimoto H, Mundel TM, Kieran MW, Kalluri R . Identification of fibroblast heterogeneity in the tumor microenvironment. Cancer Biol Ther 2006; 5: 1640–1646.
Anderberg C, Pietras K . On the origin of cancer-associated fibroblasts. Cell Cycle 2009; 8: 1461–1462.
Hinz B, Phan SH, Thannickal VJ, Galli A, Bochaton-Piallat ML, Gabbiani G . The myofibroblasts: one function, multiple origins. Am J Pathol 2007; 170: 1807–1816.
Mishra PJ, Mishra PJ, Humeniuk R, Medina DJ, Alexe G, Mesirov JP et al. Carcinoma-associated fibroblasts-like differentiation of human mesenchymal stem cells. Cancer Res 2008; 68: 4331–4339.
Direkze NC, Hodivala-Dilke K, Jeffery R, Hunt T, Poulsom R, Oukrif D et al. Bone marrow contribution to tumor-associated myofibroblasts and fibroblasts. Cancer Res 2004; 64: 8492–8495.
Zeisberg EM, Potenta S, Xie L, Zeisberg M, Kalluri R . Discovery of endothelial to mesenchymal transition as a source for carcinoma-associated fibroblasts. Cancer Res 2007; 67: 10123–10128.
Kalluri R, Weinberg RA . The basics of epithelial-mesenchymal transition. J Clin Invest 2009; 119: 1420–1428.
Hinz B, Gabbiani G . Mechanisms of force generation and transmission by myofibroblasts. Curr Opin Biotechnol 2003; 14: 538–546.
Tschumperlin DJ, Dai G, Maly IV, Kikuchi T, Laiho LH, McVittie AK et al. Mechanotransduction through growth-factor shedding into the extracellular space. Nature 2004; 429: 83–86.
Kumar S, Weaver VM . Mechanics, malignancy, and metastasis: the force journey of a tumor cell. Cancer Metastasis Rev 2009; 28: 113–127.
Kojima Y, Acar A, Eaton EN, Mellody KT, Scheel C, Ben-Porath I et al. Autocrine TGF-beta and stromal cell-derived factor-1 (SDF-1) signaling drives the evolution of tumor-promoting mammary stromal myofibroblasts. Proc Natl Acad Sci USA 2010; 107: 20009–20014.
Ria R, Reale A, De Luisi A, Ferrucci A, Moschetta M, Vacca A . Bone marrow angiogenesis and progression in multiple myeloma. Am J Blood Res 2011; 1: 76–89.
Vacca A, Ribatti D . Bone marrow angiogenesis in multiple myeloma. Leukemia 2006; 20: 193–199.
Basile A, Moschetta M, Ditonno P, Ria R, Marech I, De Luisi A et al. Pentraxin 3 (PTX3) inhibits plasma cell/stromal cell cross-talk in the bone marrow of multiple myeloma patients. J Pathol 2013; 229: 87–98.
Berardi S, Caivano A, Ria R, Nico B, Savino R, Terracciano R et al. Four proteins governing overangiogenic endothelial cell phenotype in patients with multiple myeloma are plausible therapeutic targets. Oncogene 2012; 31: 2258–2269.
Santos AM, Jung J, Aziz N, Kissil JL, Purè E . Targeting fibroblast activation protein inhibits tumor stromagenesis and growth in mice. J Clin Invest 2009; 119: 3613–3625.
Jones EA, Kinsey SE, English A, Jones RA, Straszynski L, Meredith DM et al. Isolation and characterization of bone marrow multipotential mesenchymal progenitor cells. Arthritis Rheum 2002; 46: 3349–3360.
Ghobrial IM . Myeloma as a model for the process of metastasis: implications for therapy. Blood 2012; 120: 20–30.
Ria R, Piccoli C, Cirulli T, Falzetti F, Mangialardi G, Guidolin D et al. Endothelial differentiation of hematopoietic stem and progenitor cells from patients with multiple myeloma. Clin Cancer Res 2008; 14: 1678–1685.
Xu S, Menu E, De Becker A, Van Camp B, Vanderkerken K, Van Riet I . Bone marrow-derived mesenchymal stromal cells are attracted by multiple myeloma cell-produced chemokine CCL25 and favor myeloma cell growth in vitro and in vivo. Stem Cells 2012; 30: 266–279.
Möller C, Strömberg T, Juremalm M, Nilsson K, Nilsson G . Expression and function of chemokine receptors in human multiple myeloma. Leukemia 2003; 17: 203–210.
Damiano JS, Dalton WS . Integrin-mediated drug resistance in multiple myeloma. Leuk Lymphoma 2000; 38: 71–81.
Saito RA, Micke P, Paulsson J, Augsten M, Peña C, Jönsson P et al. Forkhead box F1 regulates tumor-promoting properties of cancer-associated fibroblasts in lung cancer. Cancer Res 2010; 70: 2644–2654.
Ko SY, Barengo N, Ladanyi A, Lee JS, Marini F, Lengyel E et al. HOXA9 promotes ovarian cancer growth by stimulating cancer-associated fibroblasts. J Clin Invest 2012; 122: 3603–3617.
Östman A, Augsten M . Cancer-associated fibroblasts and tumor growth-bystanders turning into key players. Curr Opin Genet Dev 2009; 19: 67–73.
Urashima M, Ogata A, Chauhan D, Hatziyanni M, Vidriales MB, Dedera DA et al. Transforming growth factor-beta1: differential effects on multiple myeloma versus normal B cells. Blood 1996; 87: 1928–1938.
Hayashi T, Hideshima T, Nguyen AN, Munoz O, Podar K, Hamasaki M et al. Transforming growth factor beta receptor I kinase inhibitor down-regulates cytokine secretion and multiple myeloma cell growth in the bone marrow microenvironment. Clin Cancer Res 2004; 10: 7540–7546.
Markwald RR, Fitzharris TP, Smith WN . Sturctural analysis of endocardial cytodifferentiation. Dev Biol 1975; 42: 160–180.
van Meeteren LA, ten Dijke P . Regulation of endothelial cell plasticity by TGF-β. Cell Tissue Res 2012; 347: 177–186.
Moonen JR, Krenning G, Brinker MG, Koerts JA, van Luyn MJ, Harmsen MC . Endothelial progenitor cells give rise to pro-angiogenic smooth muscle-like progeny. Cardiovasc Res 2010; 86: 506–515.
Spaeth EL, Dembinski JL, Sasser AK, Watson K, Klopp A, Hall B et al. Mesenchymal stem cell transition to tumor-associated fibroblasts contributes to fibrovascular network expansion and tumor progression. PLoS One 2009; 4: e4992.
Wang LH, Yang XY, Zhang X, Farrar WL . Inhibition of adhesive interaction between multiple myeloma and bone marrow stromal cells by PPARgamma cross talk with NF-kappaB and C/EBP. Blood 2007; 110: 4373–4384.
Teoh G, Anderson KC . Interaction of tumor and host cells with adhesion and extracellular matrix molecules in the development of multiple myeloma. Hematol Oncol Clin North Am 1997; 11: 27–42.
Ribatti D, Nico B, Crivellato E, Roccaro AM, Vacca A . The history of the angiogenic switch concept. Leukemia 2007; 21: 44–52.
Damiano JS, Hazlehurst LA, Dalton WS . Cell adhesion-mediated drug resistance (CAM-DR) protects the K562 chronic mielogenous leukemia cell line from apoptosis induced by BCR/ABL inhibition, cytotoxic drugs, and gamma-irradiation. Leukemia 2001; 15: 1232–1239.
Damiano JS . Integrins as novel drug targets for overcoming innate drug resistance. Curr Cancer Drug Targets 2002; 2: 37–43.
Nefedova Y, Landowski TH, Dalton WS . Bone marrow stromal-derived soluble factors and direct cell contact contribute to de novo drug resistance of myeloma cells by distinct mechanisms. Leukemia 2003; 17: 1175–1182.
Vacca A, Ria R, Presta M, Ribatti D, Iurlaro M, Merchionne F et al. alpha(v)beta(3) integrin engagement modulates cell adhesion, proliferation, and protease secretion in human lymphoid tumor cells. Exp Hematol 2001; 29: 993–1003.
van Riet I, de Greef C, del Favero H, Demanet C, Van Camp B . Production of fibronectin and adherence to fibronectin by human myeloma cell lines. Br J Haematol 1994; 87: 258–265.
Tancred TM, Belch AR, Reiman T, Pilarski LM, Kirshner J . Altered expression of fibronectin and collagens I and IV in multiple myeloma and monoclonal gammopathy of undetermined significance. J Histochem Cytochem 2009; 57: 239–247.
We are grateful to K De Veirman (Vrije Universiteit Brussels) for help with the cytofluorimetric experiments in syngeneic 5T33 mice. This work was supported by the Associazione Italiana per la Ricerca sul Cancro (AIRC); an Investigator Grant (number 10099 to AV); the Special Program Molecular Clinical Oncology 5 per 1000 (number 9965 to AV), Milan; the European Commission’s Seventh Framework Programme (EU FPT7) OVER-MyR (number 278706 to AV and KV) and OPTATIO (number 278570 to DR); and the Ministry of Health (Progetto PRIN 2009 to RR and 2012 to AV), Rome, Italy.
The authors declare no conflict of interest.
Supplementary Information accompanies this paper on the Leukemia website
About this article
Cite this article
Frassanito, M., Rao, L., Moschetta, M. et al. Bone marrow fibroblasts parallel multiple myeloma progression in patients and mice: in vitro and in vivo studies. Leukemia 28, 904–916 (2014). https://doi.org/10.1038/leu.2013.254
- cancer-associated fibroblasts
- multiple myeloma
- tumor microenvironment
Bone marrow mesenchymal stem cells in microenvironment transform into cancer-associated fibroblasts to promote the progression of B-cell acute lymphoblastic leukemia
Biomedicine & Pharmacotherapy (2020)
Journal of Clinical Medicine (2020)
International Journal of Molecular Sciences (2020)
Targeted-Alpha-Therapy Combining Astatine-211 and anti-CD138 Antibody in a Preclinical Syngeneic Mouse Model of Multiple Myeloma Minimal Residual Disease