Multiple myeloma cells recruit tumor-supportive macrophages through the CXCR4/CXCL12 axis and promote their polarization toward the M2 phenotype

Abstract

Multiple myeloma (MM) cells specifically attract peripheral-blood monocytes, while interaction of MM with bone marrow stromal cells (BMSCs) significantly increased monocyte recruitment (p<0.01). The CXCL12 chemokine, produced by both the MM and BMSCs, was found to be a critical regulator of monocyte migration. CXCL12 production was up-regulated under MM-BMSCs co-culture conditions, whereas blockage with anti-CXCR4 antibodies significantly abrogated monocyte recruitment toward a MM-derived conditioned medium (p<0.01). Furthermore, elevated levels of CXCL12 were detected in MM, but not in normal BM samples, whereas malignant MM cells often represented the source of increased CXCL12 in the BM. Blood-derived macrophages effectively supported MM cells proliferation and protected them from chemotherapy-induced apoptosis. Importantly, MM cells affected macrophage polarization, elevating the expression of M2-related scavenger receptor CD206 in macrophages and blocking LPS-induced TNFα secretion (a hallmark of M1 response). Of note, MM-educated macrophages suppressed T-cell proliferation and IFNγ production in response to activation. Finally, increased numbers of CXCR4-expressing CD163+CD206+ macrophages were detected in the BM of MM patients (n=25) in comparison to MGUS (n=11) and normal specimens (n=8). Taken together, these results identify macrophages as important players in MM tumorogenicity, and recognize the CXCR4/CXCL12 axis as a critical regulator of MM-stroma interactions and microenvironment formation.

Figure 1. Interaction between MM and BMSCs induces the ability of MM cells to selectively attract peripheral blood monocytes

(A) Trans-well migration of PBMCs toward conditioned medium (CM) produced by MM cells ARH77 and RPMI8226 cultured alone or in the presence of BMSCs. Migration of CD14-positive monocytes and CD3-positive lymphocytes was enumerated by FACS and percent of migrating cells out of total input cells was calculated. Data is presented as mean of triplicates ±STDEV (**p<0.01). Each experiment was repeated three times. (B) Migration of CD14+ monocytes in response to CM produced by MM cells incubated in either direct contact with BMSCs or separated with 0.4 μm membrane during the co-culture. Data is presented as mean of triplicates ±STDEV (**p<0.01). Each experiment was repeated three times. (C) Chemokine receptor expression repertoire on CD14+ monocytes or CD3+ lymphocytes evaluated by cell-surface co-staining of PBMCs and FACS analysis.

Figure 2. Both MM and BMSC cells express CXCL12. Interaction with BMSCs elevates CXCL12 expression in MM cell lines

(A) Secretion of CXCL12 by MM cell lines RPMI8226 and ARH77, incubated in the absence or presence of BMSCs, in direct contact or separated by 0.4 μm-transwells for 48 hours, evaluated by ELISA. (B) Expression of CXCL12 mRNA in MM cells RPMI8226 and ARH77, incubated in the absence or presence of BMSCs for 48 hours, evaluated by quantitative PCR. (C) Expression of CXCL12 mRNA in human BMSCs, untreated or treated with RPMI8226-produced CM, evaluated by quantitative PCR. Data is presented as mean of triplicates ±STDEV (**p<0.01). (D) Expression of CXCL12 in four MM (I, II, III and IV) and two normal (V and VI) BM samples evaluated by immunohistochemical staining. Original magnification of x400 is shown.

Figure 3. Peripheral blood monocytes respond in a CXCR4-dependetnt manner to MM-induced migratory signals and demonstrate dose-dependent migratory ability in response to CXCL12

PBMCs were allowed to migrate for 4 hours in response to CM produced by MM cell lines (ARH77 and RPMI8226), incubated in 1% FCS-containing medium in the absence or presence of BMSCs. (A) Quantification of migrated CD14+ CXCR4+ monocytes toward control medium or RPMI8226-produced CM was evaluated by FACS. Percent of CXCR4-positive cells out of total migrating CD14+ monocytes is presented. (B) PBMCs were allowed to migrate for 4 hours in response to elevated doses of recombinant human CXCL12. Quantification of migrated CD14+ monocytes was performed by FACS. Percent of migrated cells was calculated. Data is presented as mean of triplicates ±STDEV (**p<0.01). (C) To inhibit CXCR4, neutralizing anti-CXCR4 monoclonal antibody (20 μg/ml) was added to PBMCs in the upper chamber. Quantification of migrated CD14+ monocytes or CD3+ lymphocytes was performed using specific staining and FACS evaluation. Percent of migrated cells was calculated. Data is presented as mean of triplicates ±STDEV (**p<0.01). Each experiment was repeated twice.

Figure 4. Macrophages support MM cell survival and proliferation, protect MM cells from chemotherapy-induced apoptosis and elevate expression of pro-angiogenic factors in MM cells

(A) CFSE-labeled MM cells ARH77 and RPMI8226 were co-cul. tured under serum-reduced (1% FCS) or full-serum (10%) conditions in the absence or presence of peripheral blood-derived macrophages for 24, 48 and 72 hours. Viable MM cell number was determined by FACS using PI exclusion. Proliferation index was calculated. Data is presented as mean of triplicates ±STDEV (**p<0.01). (B) BM samples (n=3) from MM patients containing CD138+ cells were cultured in 10% FCS medium in the absence or presence of macrophages for 5 days and percent of viable CD138+ PI-negative plasma cells was detected by FACS. Data is presented as mean of triplicates ±STDEV (**p<0.01).(C) Viability of MM cells ARH77 and RPMI8226, treated with melphalan (5 μM), bortezomib (5nM) or lenalidomide (10 μM) in the absence or presence of peripheral blood-derived macrophages for 48 hours, evaluated by the XTT method. Data is presented as mean of triplicates ±STDEV (**p<0.01). (D, E) RPMI8226 cells were cultured under serum-reduced (1% FCS) conditions, with or without melphalan (5 μM), in the absence or presence of peripheral blood-derived macrophages. Cell cycle distribution was evaluated and percent of apoptotic (sub G0/G1) cells was detected. (D) Representative images demonstrating cell cycle distribution. (E) Quantification of apoptotic DNA fragmentation (percent of sub G0/G1 population). Data is presented as mean of triplicates ±STDEV (**p<0.01).(F) Expression of IL-1β, IL-8, CCL2 and CCL5 mRNA in MM cells RPMI8226 and ARH77, incubated in the absence or presence of peripheral blood-derived macrophages for 48 hours, evaluated by quantitative PCR.

Figure 5. MM cells affect macrophage polarization, promote M2 phenotype acquisition and block M1 response in LPS-treated macrophages

(A) MM cells ARH77 and RPMI8226 were incubated in the absence or presence of BMSCs for 48 hours and IL-10 secretion to culture medium was measured using ELISA kit. Data is presented as mean of triplicates ±STDEV (**p<0.01). (B) Peripheral-blood generated macrophages were incubated in the absence or presence of MM cells ARH77 and RPMI8226 for 48 hours, and surface expression of CD206 on CD14-expressing macrophage cell population was evaluated using flow cytometry analysis. (C, D) Effect of MM education on M1 stimuli response (LPS treatment) in macrophages. Peripheral-blood generated macrophages were pre- incubated in the absence or presence of MM cells ARH77 and RPMI8226 for 24 hours; non-adherent MM cells were removed and cells were stimulated with LPS (100 ng/ml) for an additional 48 hours. Medium was collected and levels of TNFα (C) and IL-10 (D) were measured using ELISA commercial kits.

Figure 6. MM-educated macrophages suppress proliferation and IFNγ secretion of autologous T cells in response to polyclonal stimulation

Peripheral-blood derived macrophages were pre-cultured in the absence or presence of RPMI8226 cells for 48 hours, and excess of myeloma was removed by pipetting. Frozen lymphocyte-enriched autologous PBMCs were thawed, labeled with CFSE (5 μM), plated in the absence or presence of macrophages and stimulated with anti-CD3 (OKT3) (10 μg/ml) and anti-CD28 (1 μg/ml) antibodies for 5 days. Cell division was monitored by FACS. Number of cell divisions and relative number of proliferated cells was measured based on the reduction in CFSE intensity. (A) Representative histogram plots showing decrease in the proliferation rate of stimulated T cells in the presence of MM-educated macrophages. (B) Numbers of proliferating T cells were quantified. Data is presented as mean of triplicates ±STDEV (**p<0.01). (C) Proliferation index of polyclonally-stimulated T cells, calculated as described in methods section. (D) IFNγ secretion to the culture medium, collected at the end of proliferation (day 5) and measured using ELISA. Data is presented as mean of triplicates ±STDEV (**p<0.01).

Figure 7. Elevated numbers of CXCR4-expressing M2 macrophages in BM samples of MM patients

The presence of CD163+ CD206+ macrophages was assessed in BM samples from patients with MM, smoldering myeloma, MGUS and normal controls using FACS analysis. (A) Numbers of CD163+ CD206 + macrophages assessed in 25 patients with MM, 11 patients with MGUS and smoldering MM, and 8 normal controls (out of 20,000 cells acquired by FACS). (B) Representative plots demonstrating the M2 population on two MM samples and one normal control. Histograms represent CXCR4 surface levels expressed by CD163+ CD206+ macrophages.