The signaling lymphocytic activation molecule family 3 (SLAMF3) is a member of the immunoglobulin superfamily expressed on T, B, and natural killer cells and modulates the activation and cytotoxicity of these cells. SLAMF3 is also expressed on plasma cells from patients with multiple myeloma (MM), although its role in MM pathogenesis remains unclear. This study found that SLAMF3 is highly and constitutively expressed on MM cells regardless of disease stage and that SLAMF3 knockdown/knockout suppresses proliferative potential and increases drug-induced apoptosis with decreased levels of phosphorylated ERK protein in MM cells. SLAMF3-overexpressing MM cells promote aggressive myeloma behavior in comparison with cytoplasmic domain-truncated SLAMF3 (ΔSLAMF3) cells. SLAMF3 interacts directly with adaptor proteins SH2 domain-containing phosphatase 2 (SHP2) and growth factor receptor bound 2 (GRB2), which also interact with each other. SLAMF3 knockdown, knockout, ΔSLAMF3, and SHP2 inhibitor-treated MM cells decreased phosphorylated ERK protein levels. Finally, serum soluble SLAMF3 (sSLAMF3) levels were markedly increased in advanced MM. Patients with high levels of sSLAMF3 progressed to the advanced stage significantly more often and had shorter progression-free survival times than those with low levels. This study revealed that SLAMF3 molecules consistently expressed on MM cells transmit MAPK/ERK signals mediated via the complex of SHP2 and GRB2 by self-ligand interaction between MM cells and induce a high malignant potential in MM. Furthermore, high levels of serum sSLAMF3 may reflect MM disease progression and be a useful prognostic factor.

Implications:

SLAMF3 may be a new therapeutic target for immunotherapy and novel agents such as small-molecule inhibitors.

Multiple myeloma (MM) is a malignancy of monoclonal plasma cells (PC), which are transformed from normal PCs by multistep genetic and bone marrow (BM) microenvironmental changes (1). In the past several years, the treatment of MM has changed dramatically with the emergence of novel agents: immunomodulatory drugs (lenalidomide and pomalidomide); proteasome inhibitors (bortezomib, carfilzomib, and ixazomib); and monoclonal antibodies (elotuzumab and daratumumab; refs. 2–5). These treatments have markedly improved survival outcome (1, 4). Nevertheless, some patients with refractory/relapsed MM remain incurable with repeated relapse.

The cell growth, drug resistance, and survival of MM cells are promoted by the interactions among MM, stromal cells, and other cell components in the BM microenvironment (6, 7). MM cells highly express various immune-associated antigens, e.g., programmed death ligand 1 (PD-L1, B7-H1; refs. 8, 9), B-cell maturation antigen (BCMA; ref. 10), CD28 (11), and B7 homology 2 (B7-H2; ref. 12). The cell-to-cell interactions via the binding of these immune-associated antigens with each ligand transmit positive signals to MM cells, induce their growth, and confer survival advantages (8–12). Thus, an understanding of the mechanisms of cell–cell interactions via immune-associated antigens and their ligands in the BM microenvironment is needed for the development of novel therapeutic targets.

The signaling lymphocytic activation molecule family 3 (SLAMF3; also known as CD229 or Ly9) was identified as the phosphorylated immunoreceptor with the greatest expression in the MM cell line MOLP-8 using the phosphoimmunoreceptor assay (13). SLAMF3 is composed of two immunoglobulin (Ig)-like V-type and two Ig-like C2-type domains in the extracellular region and two immunoreceptor tyrosine-based switch motifs (ITSM) in the cytoplasmic region (14, 15). SLAMF3 acts as a self-ligand and transmits a potent positive signal to T cells mediated via the adaptor SLAM-associated protein (SAP; refs. 15–17). T cells derived from SLAMF3-deficient mice proliferate poorly and produce significantly less IL2; moreover T helper 2 (Th2) polarization is suppressed in those mice (16, 18). These results underscore the crucial role played by SLAMF3 in the T-cell activation process. SLAMF3 is highly expressed on PCs from patients with MM and monoclonal gammopathy of undetermined significance (MGUS; refs. 13, 19), but its function in MM pathogenesis is unclear.

To determine whether SLAMF3 is useful as a new therapeutic target in advanced or refractory/relapsed MM, this study was conducted to assess the expression of SLAMF3 in MM patients in detail. Furthermore, we defined the biological functions of SLAMF3 in MM cell proliferative advantage and sensitivity to anti-MM agents.

Patient samples

BM and peripheral blood (PB) samples were obtained from patients with MM and MGUS for diagnostic purposes at six clinical institutions after informed consent had been obtained according to each Institutional Review Board–approved protocol. The diagnoses were made according to International Myeloma Working Group criteria (20). International Scoring System (ISS), revised-ISS (R-ISS), and Durie-Salmon (DS) staging were used to classify MM patients as in previous reports (21–23). Mononuclear cells (MNC) were separated from BM samples with Histopaque (Sigma-Aldrich) density centrifugation. Serum samples were obtained by centrifugation of heparinized PB and stored at −20°C until use.

Human MM cell lines

The human MM cell lines were cultured in RPMI-1640 medium (Wako Chemical Industries) containing 10% fetal bovine serum, 100 U/mL of penicillin, and 100 mg/mL of streptomycin (Thermo Fisher Scientific) at 37°C under 5% CO2. KMS18, KMS20, KMS27, KMS28-PE, KMS28-BM, and KMS34 were kindly provided by Dr. Takemi Otsuki in 2004 (Kawasaki Medical School, Okayama, Japan). U266 and RPMI8226 were obtained from the American Type Culture Collection in 2005. The human MM cell lines MOSTI-1, -2, -4, -6, and -40 were established using BM samples from MM patients in our laboratory (obtained in 2012; ref. 8). Mycoplasma was tested and was negative in all cells using e-Myco Mycoplasma PCR Detection Kit (iNtRON Biotechnology). These cell lines were not examined by short tandem repeat analysis. The cells were cultured to passages 3 to 5, frozen in the cell-freezing medium TC-protector (DS Pharma Biomedical), and used within 20 passages in all experiments.

Flow cytometry (FCM)

Flow-cytometric analysis was performed as previously described in detail (8, 9). In brief, after blocking with human gamma Ig (MP Biomedicals), cells were stained with antibody against SLAMF3 (PE, #FAB1898P; R&D Systems), CD138 (FITC, #K0108-4; MBL), CD38 (Brilliant Violet 421, #303526; BioLegend). Data acquisition was performed in an LSRFortess X-20 flow cytometer with FACSDiva software version 8.0.1 (BD Biosciences) and analyzed using FlowJo software (TreeStar).

Cell proliferation and apoptosis assay

Cell proliferation was determined in the 2-(2-methoxy-4-nitrophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium monosodium salt (WST-8) assay using the Cell Counting Kit-8 (CCK-8; ref. 8). Bromodeoxyuridine (BrdUrd) and Ki-67 analyses were performed using FCM as described previously (8, 9). To assess drug-induced apoptosis, cells were exposed to melphalan and bortezomib for 24 hours at concentrations optimal for inducing apoptosis, and then the cells were stained with annexin–FITC (BD Biosciences) and propidium Iodide (PI; Wako Chemical Industries) for the analysis using FCM.

mRNA expression analysis

Total RNA was extracted from CD138+ PCs isolated from BM MNCs of MM patients and MM cell lines. cDNA from total RNA was synthesized, and then quantification of mRNA using real-time PCR was performed (8, 9). The primer sequences are shown in Supplementary Table S1.

shRNA-mediated knockdown and CRISPR-mediated knockout

For small hairpin RNA (shRNA)-mediated knockdown, the cells were transfected with pLKO.1 MISSION lentiviral transduction particles containing shRNA that targets SLAMF3 (clone ID: TRCN0000157350; Sigma-Aldrich). MISSION pLKO.1-puro Non-Mammalian shRNA Control containing part of the turbo GFP (tGFP) sequence (#SHC002V; Sigma-Aldrich) was used as a negative control. To establish stable knockdown cell lines, the cells were cultured with complete medium containing 1 μg/mL of puromycin. For CRISPR/Cas9-mediated knockout, the cells were transfected with Edit-R Lentiviral hCMV-Blast-Cas9 Nuclease Particles (#VCAS10124; Horizon Discovery) and cultured with 5 μg/mL blasticidin-supplemented medium. Then Cas9-expressing cells were transfected with Edit-R CRISPR Human SLAMF3 Lentiviral single-guide RNA (sgRNA) particles (clone ID: VSGHSM_26650008; Horizon Discovery), and SLAMF3 knockout cells were selected by puromycin. Cas9-expressing cells transfected with Edit-R all-in-one lentiviral sgRNA nontargeting control particles (#VSGC11954; Horizon Discovery) were used as a control.

SLAMF3-overexpressing MM cells

Two cDNA fragments of full-length SLAMF3 and its truncated form of the cytoplasmic domain (ΔSLAMF3) were amplified from human MM cell line–derived cDNA using the primer shown in Supplementary Table S1. The cDNA was inserted into the pEF1/V5-His vector containing a geneticin-resistance gene (Thermo Fisher Scientific). KMS34 cells lacking SLAMF3 expression (Fig. 1A) were transfected with plasmid DNA using Lipofectamine LTX together with Plus reagent (Thermo Fisher Scientific), and stable clones of transfected cells were established (9). KMS34 cells transfected with a pEF1/V5-His vector (Mock) were used as negative controls.

Figure 1.

The difference in cell proliferative potential and drug sensitivity between SLAMF3high and SLAMF3low MM cells. A, Cell-surface SLAMF3 expression in nine MM cell lines analyzed using FCM. Solid line, staining with PE-conjugated antibody to SLAMF3; filled area, staining with isotype-matched control Ig. The numerical values are represented by the percentage of SLAMF3-positive cells in each MM cell line. B, Exemplary dot blot of U266 cells blotted against SLAMF3. BrdUrd incorporation (C) and Ki-67 expression (D) in SLAMF3low and SLAMF3high cell fractions in MM cell lines. Relative mean fluorescence intensity (MFI) in the ratio between the MFI of antibody staining and the MFI of control IgG staining. E, The melphalan and bortezomib sensitivity in SLAMF3low and SLAMF3high cell fractions in MM cell lines. The cells were exposed to melphalan or bortezomib overnight, and the annexin V+ apoptotic cells in each cell fraction were evaluated using FCM. After 2-day cultivation with 20 μg/mL of control Ig or anti-SLAMF3 antibody, BrdUrd incorporation (F) and melphalan sensitivity (G) in MM cell lines were examined. Data, mean ± standard deviation of triplicate experiments. *, P < 0.05; **, P < 0.01.

Figure 1.

The difference in cell proliferative potential and drug sensitivity between SLAMF3high and SLAMF3low MM cells. A, Cell-surface SLAMF3 expression in nine MM cell lines analyzed using FCM. Solid line, staining with PE-conjugated antibody to SLAMF3; filled area, staining with isotype-matched control Ig. The numerical values are represented by the percentage of SLAMF3-positive cells in each MM cell line. B, Exemplary dot blot of U266 cells blotted against SLAMF3. BrdUrd incorporation (C) and Ki-67 expression (D) in SLAMF3low and SLAMF3high cell fractions in MM cell lines. Relative mean fluorescence intensity (MFI) in the ratio between the MFI of antibody staining and the MFI of control IgG staining. E, The melphalan and bortezomib sensitivity in SLAMF3low and SLAMF3high cell fractions in MM cell lines. The cells were exposed to melphalan or bortezomib overnight, and the annexin V+ apoptotic cells in each cell fraction were evaluated using FCM. After 2-day cultivation with 20 μg/mL of control Ig or anti-SLAMF3 antibody, BrdUrd incorporation (F) and melphalan sensitivity (G) in MM cell lines were examined. Data, mean ± standard deviation of triplicate experiments. *, P < 0.05; **, P < 0.01.

Close modal

Analysis of differential gene expression

Human Genome U133 Plus 2.0 arrays (Affymetrix), which comprise 54,675 probes, were used for mRNA expression profiling of SLAMF3.KMS34 and ΔSLAMF3.KMS34 cells. Of 54,675 probes, 26,449 were filtrated with at least one present flag in two cell clones. These probes were analyzed using gene set enrichment analysis (GSEA) performed with the GSEA software (http://www.broadinstitute.org/gsea; refs. 24, 25).

Xenograft mouse model

Seven-week-old NOD/Shi-scid, IL2RγKOJic (NOG) female mice (CLEA Japan) were used for the in vivo assays. NOG mice were inoculated subcutaneously in the right flank with 1 × 107 SLAMF3-expressing KMS34 cells in 100 μL of PBS. Tumor sizes were measured with calipers every 3 to 4 days, and volumes were calculated using the formula α2 × β × 0.5, where α is the shortest diameter and β is the diameter perpendicular to α. The mice were euthanized when the tumor volume reached 2,000 mm3. Survival was evaluated from the first day of inoculation with MM cells until death. Animal studies were approved by the Institutional Animal Ethical Committee of Nippon Medical School.

Western blotting and antibodies

Cells were lysed using RIPA lysis buffer (Merck Millipore) plus protease inhibitor and phosphatase inhibitor (Roche Diagnostics). Twenty micrograms of protein were electrophoresed on an SDS–12.5% polyacrylamide gel and then transferred to a PVDF membrane. The target protein was detected with the ECL prime Western Blotting Detection kit (GE Healthcare; ref. 8). The primary antibodies were 1,000-fold diluted rabbit antibody to SAP (#2778), SHP1 (#3759), SHP2 (#3752), phospho-SHP2 (#3751), Csk (#4980), SHIP1 (#2726), p44/p42 MAPK (ERK1/2; #9102), phospho-ERK1/2 (#4370), Akt (#9272), phospho-Akt (Ser473; #4060), BCL2 (#15071), and β-actin (#4967; Cell Signaling Technology); EAT2 (#ab95270) and GRB2 (#ab32037; Abcam); and SLAMF3 (#326104; BioLegend). The secondary antibody was HRP-conjugated anti-rabbit and anti-mouse IgG (Cell Signaling Technology).

Enzyme-linked immunosorbent assay (ELISA) for soluble SLAMF3

Soluble SLAMF3 (sSLAMF3) concentrations were measured using the human CD229 ELISA kit (RayBiotech), according to the manufacturer's instructions. Each sample and standard protein were analyzed in duplicate; the minimum detectable level of sSLAMF3 was 80 pg/mL.

Immunoprecipitation

One milligram of cell lysate or 500 μL of serum and cell culture supernatant was diluted to 1 mL with TBS and incubated with 4 μg of rabbit antibody to SHP2 (Cell Signaling Technology), GRB2 (Abcam), or SLAMF3 (BioLegend) at 4°C for 4 hours. Then, these mixtures were inoculated with 30 μL of Protein A/G PLUS-Agarose (Santa Cruz Biotechnology) at 4°C overnight with a rotator. The agarose was washed three times with TBS using the SigmaPrep spin column (Sigma-Aldrich) and then eluted with 2x loading buffer (200 mmol/L Tris-HCl, pH 6.8, 4.0% SDS, 0.004% BPB, and 20% glycerol). The immunoprecipitated proteins were resolved by SDS-PAGE and analyzed by Western blotting as described above. The sSLAMF3 immunoprecipitated from serum samples and cell culture supernatants of MM cells with mouse monoclonal anti-SLAMF3 antibody (BioLegend) was detected using rabbit polyclonal anti-SLAMF3 antibody (Bioss).

Statistical analysis

Differences between two groups of data were determined using the χ2 or Fisher exact test for categorical variables. The Student t test, Mann–Whitney U test, and Spearman rank moment correlation coefficient were used to analyze the data. Optimal cutoff values of serum sSLAMF3 levels were determined using the receiver operating characteristics (ROC) curve. Progression-free survival (PFS) was defined as the period from the date of new diagnosis to second-line treatment or to treatment cessation date. Overall survival (OS) was defined as the period from the date of new diagnosis until death or the last follow-up. PFS and OS curves were estimated using Kaplan–Meier analysis, and statistical differences between different groups were compared by the log-rank test. Survival association with prognostic factors was determined by multivariate analysis using the Cox proportional hazards model. P values of less than 0.05 were considered to represent statistically significant differences. Statistical analyses were performed using SPSS version 23 software (SPSS).

Cell-surface SLAMF3 expression in PCs from MM patients

Cell-surface SLAMF3 molecules were highly expressed on CD38high PCs from BM samples of MM patients, and the expression levels were the same as in hematologically normal controls (Supplementary Fig. S1). Next, to clarify whether the expression levels differ in various disease stages in MM patients, we investigated the expression levels of cell-surface SLAMF3 and CD138 in CD38high PCs from 230 patients with MGUS and MM (Supplementary Table S2). SLAMF3 expression levels on PCs did not differ among patients with asymptomatic and symptomatic MM and MGUS (Table 1). Further, SLAMF3 on PCs was highly expressed regardless of MM disease stage (Table 1). In contrast, CD138 expression on PCs from patients with symptomatic MM was significantly downregulated in comparison with MGUS and asymptomatic MM patients (Table 1). CD138 expression levels on PCs were markedly lower in MM patients with ISS stage II/III than in those with ISS stage I and decreased in patients who relapsed or became refractory to anti-MM chemotherapy (Table 1). Thus, cell-surface SLAMF3 was constitutively and highly expressed on PCs from MM patients at various disease stages.

Table 1.

Relationships of CD138 and SLAMF3 expression with disease progression in MM.

CD138SLAMF3
No. of patientsExpression (%)aP valueExpression (%)aP value
Symptomatic myeloma 134 62.0 ± 24.5 − 81.0 ± 18.8 − 
 vs. MGUS 47 vs. 71.4 ± 18.1 0.0295 vs. 76.1 ± 21.7 N.S. 
 vs. Asymptomatic myeloma 19 vs. 77.0 ± 14.6 0.009 vs. 86.6 ± 13.4 N.S. 
ISS II/III 101 60.3 ± 23.3 0.0005 80.3 ± 19.1 N.S. 
 vs. I 51 vs. 71.8 ± 23.8  vs. 84.2 ± 16.9  
Refractory/relapsed myeloma 30 52.2 ± 19.4 0.006 81.6 ± 14.8 N.S. 
 vs. Newly diagnosed myeloma 153 vs. 63.9 ± 24.0  vs. 81.7 ± 18.3  
CD138SLAMF3
No. of patientsExpression (%)aP valueExpression (%)aP value
Symptomatic myeloma 134 62.0 ± 24.5 − 81.0 ± 18.8 − 
 vs. MGUS 47 vs. 71.4 ± 18.1 0.0295 vs. 76.1 ± 21.7 N.S. 
 vs. Asymptomatic myeloma 19 vs. 77.0 ± 14.6 0.009 vs. 86.6 ± 13.4 N.S. 
ISS II/III 101 60.3 ± 23.3 0.0005 80.3 ± 19.1 N.S. 
 vs. I 51 vs. 71.8 ± 23.8  vs. 84.2 ± 16.9  
Refractory/relapsed myeloma 30 52.2 ± 19.4 0.006 81.6 ± 14.8 N.S. 
 vs. Newly diagnosed myeloma 153 vs. 63.9 ± 24.0  vs. 81.7 ± 18.3  

aThe percentage expression of CD138 and SLAMF3 on CD38high-expressing PCs was analyzed as compared with isotype-matched negative controls by FCM. N.S., not significant.

SLAMF3high-expressing MM cells are endowed with cell proliferation potential and drug resistance

Cell-surface SLAMF3 was expressed in almost all MM cell lines except for KMS34 cells (Fig. 1A). The expression pattern of SLAMF3 in MM cell lines was classified into two patterns: SLAMF3 high (pattern I) and middle expression (pattern II; Supplementary Fig. S2A). In particular, SLAMF3 high-expressing MM cells (pattern I), such as KMS18, KMS28-PE, KMS28-BM, MOSTI-1, and U266 cells, contained minor cell fractions of SLAMF3low cells (Fig. 1B; Supplementary Fig. S2; 2.6%, 5%, 50%, 10%, and 2.8%, respectively).

Thus, we examined whether cell proliferation potential and sensitivity to anti-MM chemotherapy were different between SLAMF3high and SLAMF3low MM cells (Fig. 1B). BrdUrd incorporation and Ki-67 expression were significantly higher in SLAMF3high than in SLAMF3low cell fractions (Fig. 1C and D). Furthermore, melphalan- and bortezomib-induced apoptosis was markedly increased in SLAMF3low cell fractions (Fig. 1E). Next, we investigated whether those cell characteristics in SLAMF3-expressing MM cells were induced by self-ligand interaction between MM cells. Cell proliferation and melphalan-induced apoptosis in MM cells were significantly suppressed and enhanced by treatment with anti-SLAMF3 antibody, respectively, but not in KMS34 cells that lacked SLAMF3 expression (Fig. 1F and G).

Cell characteristics of SLAMF3-knockdown/knockout or -overexpressing MM cells

We examined whether SLAMF3 molecules on MM cells were directly associated with cell proliferation and resistance to anti-MM agents using the lentiviral vector–mediating shRNA system (Fig. 2A). SLAMF3-knockdown MM cells proliferated more slowly and showed higher sensitivity to melphalan and bortezomib in comparison with control cells (Fig. 2B and C). Similarly, SLAMF3 knockout by the CRISPR/Cas9 system in MM cells induced slow cell growth and high sensitivity to anti-MM agents (Fig. 2DG). Next, we generated stable KMS34 cell clones producing full-length SLAMF3 and its truncated form of the cytoplasmic domain (ΔSLAMF3) in which similar levels of SLAMF3 surface expression were established (Fig. 3A). To test whether SLAMF3 transmitted antiapoptotic and proliferation signals to MM cells, we prepared ΔSLAMF3 cells to eliminate the intracellular signaling. Cell proliferation was significantly more rapid in SLAMF3 cells than in ΔSLAMF3 and Mock cells (Fig. 3B). Ki-67 expression and BrdUrd incorporation were markedly increased in SLAMF3 cells (Fig. 3C). Furthermore, the percentage of melphalan- and bortezomib-induced apoptotic cells was decreased in SLAMF3 cells compared with ΔSLAMF3 and Mock cells (Fig. 3D). Next, we confirmed the aggressive characteristics of SLAMF3 in MM in vivo using a xenograft mouse model of MM cells. Tumor sizes in SLAMF3 cell xenograft NOG mice significantly increased compared with mice with ΔSLAMF3 cell xenografts (P = 0.0077 on day 42; Fig. 3E). Moreover, the survival of mice injected with SLAMF3 cells was markedly shorter than that of mice injected with ΔSLAMF3 and Mock cells (Fig. 3F).

Figure 2.

Characteristics of SLAMF3-knockdown and -knockout MM cells. A, Cell-surface SLAMF3 expression was analyzed by FCM (top graph) and Western blot analysis (bottom photographs) in MM cells transduced with nontarget shRNA (control; shCtl.) and SLAMF3-specific shRNA (shSLAMF3). D, Cell-surface SLAMF3 expression in SLAMF3-knockout MM (SLAMF3 KO) and control cells (Ctl.) analyzed using FCM (right photographs) and Western blot analysis (left photographs). Solid line, staining with PE-conjugated antibody to SLAMF3; filled area, staining with isotype-matched control Ig. Cell proliferation (B and E), melphalan/bortezomib sensitivity (C and G), and BrdUrd incorporation (F) of SLAMF3-knockdown and -knockout MM cells were examined. Relative MFI is the ratio between the MFI of antibody staining and the MFI of control IgG staining. The data are expressed as mean ± standard deviation of triplicate experiments. *, P < 0.05; **, P < 0.01.

Figure 2.

Characteristics of SLAMF3-knockdown and -knockout MM cells. A, Cell-surface SLAMF3 expression was analyzed by FCM (top graph) and Western blot analysis (bottom photographs) in MM cells transduced with nontarget shRNA (control; shCtl.) and SLAMF3-specific shRNA (shSLAMF3). D, Cell-surface SLAMF3 expression in SLAMF3-knockout MM (SLAMF3 KO) and control cells (Ctl.) analyzed using FCM (right photographs) and Western blot analysis (left photographs). Solid line, staining with PE-conjugated antibody to SLAMF3; filled area, staining with isotype-matched control Ig. Cell proliferation (B and E), melphalan/bortezomib sensitivity (C and G), and BrdUrd incorporation (F) of SLAMF3-knockdown and -knockout MM cells were examined. Relative MFI is the ratio between the MFI of antibody staining and the MFI of control IgG staining. The data are expressed as mean ± standard deviation of triplicate experiments. *, P < 0.05; **, P < 0.01.

Close modal
Figure 3.

Characteristics of SLAMF3-overexpressing KMS34 cells. A, Construct of full-length SLAMF3 (SLAMF3) and its truncated form of the cytoplasmic domain (ΔSLAMF3; left). ED, extracellular domain; TM, transmembrane domain; CD, cytoplasmic domain. SLAMF3 and ΔSLAMF3 overexpressing KMS34 cells were analyzed by Western blot analysis (center) and FCM (right). Solid line, staining with PE-conjugated antibody to SLAMF3; filled area, staining with isotype-matched control Ig. The numerical values are represented by relative MFI. B, Cell growth in SLAMF3- and ΔSLAMF3-expressing KMS34 cells was assessed in the MTT assay. Ki-67 expression and BrdUrd incorporation (C) and melphalan/bortezomib sensitivity (D) were analyzed using FCM in SLAMF3, ΔSLAMF3, and mock.KMS34 (Mock) cells. Data, mean ± standard deviation of triplicate experiments. *, P < 0.05; **, P < 0.01. E,In vivo analysis using a murine xenograft model of SLAMF3, ΔSLAMF3, and Mock cells inoculated into the flanks of NOG mice (each n = 7). Tumor volume was calculated from caliper measurements at the indicated time points. Data, mean ± standard error. F, Kaplan–Meier survival analysis of mice described in E.

Figure 3.

Characteristics of SLAMF3-overexpressing KMS34 cells. A, Construct of full-length SLAMF3 (SLAMF3) and its truncated form of the cytoplasmic domain (ΔSLAMF3; left). ED, extracellular domain; TM, transmembrane domain; CD, cytoplasmic domain. SLAMF3 and ΔSLAMF3 overexpressing KMS34 cells were analyzed by Western blot analysis (center) and FCM (right). Solid line, staining with PE-conjugated antibody to SLAMF3; filled area, staining with isotype-matched control Ig. The numerical values are represented by relative MFI. B, Cell growth in SLAMF3- and ΔSLAMF3-expressing KMS34 cells was assessed in the MTT assay. Ki-67 expression and BrdUrd incorporation (C) and melphalan/bortezomib sensitivity (D) were analyzed using FCM in SLAMF3, ΔSLAMF3, and mock.KMS34 (Mock) cells. Data, mean ± standard deviation of triplicate experiments. *, P < 0.05; **, P < 0.01. E,In vivo analysis using a murine xenograft model of SLAMF3, ΔSLAMF3, and Mock cells inoculated into the flanks of NOG mice (each n = 7). Tumor volume was calculated from caliper measurements at the indicated time points. Data, mean ± standard error. F, Kaplan–Meier survival analysis of mice described in E.

Close modal

Interaction of SLAMF3 molecules with adaptor proteins SHP2 and GRB2

Next, to identify adaptor protein interactions with the cytoplasmic domain of SLAMF3 in MM, we investigated the expression of the following adaptor protein src homology 2 (SH2) domain contained in eight MM cell lines: SAP; Ewing sarcoma–associated transcript 2 (EAT2); SH2 domain–containing phosphatase 1 (SHP1); SHP2; SH2-domain–containing inositol-5-phoshatase 1 (SHIP1); c-src kinase (CSK); and growth factor receptor bound 2 (GRB2) in eight MM cell lines (16, 26). Of these adaptor proteins, SHP2 and GRB2 were detected in almost all MM cell lines, but not SAP or EAT2 (Fig. 4A). Interestingly, the mRNA levels of SLAMF3, SHP2, and GRB2 were significantly correlated with each other in both MM cell lines and MM patient samples, respectively (Supplementary Fig. S3). In the immunoprecipitation assay, SLAMF3 directly interacted with SHP2 or GRB2, and SHP2 also interacted with GRB2 (Fig. 4B and C). Furthermore, transient knockdown of SHP2 or GRB2 in MM cells induced greater sensitivity to melphalan (Supplementary Fig. S4A and S4B), but not decreased cell growth (data not shown). Cell proliferation and melphalan sensitivity in MM cells were decreased and increased by treatment with the SHP1/2 inhibitor NSC87877, respectively (Supplementary Fig. S4C).

Figure 4.

SLAMF3 transmitted MAPK/ERK pathway signals via the complex of SHP2 and GRB2 in MM cells. A, The expression of SH2 domain-containing adaptor proteins SAP, EAT2, SHP1, SHP2, SHIP1, CSK, and GRB2 in eight MM cell lines was detected by Western blotting. Jurkat and Molt-4 (T-cell lines) and NK-92MI and YNT (NK-cell lines) were used as positive controls. MM cell lysates were immunoprecipitated with anti-SHP2, anti-GRB2, and SLAMF3 antibodies. Then, membranes were immunoblotted with the respective specific antibodies (B and C). D, Western blot analysis of SHP2, ERK, and AKT phosphorylation in SLAMF3 knockdown, knockout, and overexpression, and NSC87877 (an SHP1/2 inhibitor; 50 μmol/L) treatment. Numbers under the bands of phospho-SHP2 (pSHP2) and pERK1/2 indicate the relative intensity of each protein normalized to the signal intensity of SHP2 and ERK1/2, respectively. E, Enrichment plots for the gene sets of the MAPK/ERK pathway that were significantly upregulated in SLAMF3.KMS34 cells compared with ΔSLAMF3 cells. NES, normalized enrichment score; FDR, false discovery rate. The mRNA (F) and protein (G) expression of BCL2, CCND1, and CCND2 in SLAMF3-knockdown U266 cells. **, P < 0.01.

Figure 4.

SLAMF3 transmitted MAPK/ERK pathway signals via the complex of SHP2 and GRB2 in MM cells. A, The expression of SH2 domain-containing adaptor proteins SAP, EAT2, SHP1, SHP2, SHIP1, CSK, and GRB2 in eight MM cell lines was detected by Western blotting. Jurkat and Molt-4 (T-cell lines) and NK-92MI and YNT (NK-cell lines) were used as positive controls. MM cell lysates were immunoprecipitated with anti-SHP2, anti-GRB2, and SLAMF3 antibodies. Then, membranes were immunoblotted with the respective specific antibodies (B and C). D, Western blot analysis of SHP2, ERK, and AKT phosphorylation in SLAMF3 knockdown, knockout, and overexpression, and NSC87877 (an SHP1/2 inhibitor; 50 μmol/L) treatment. Numbers under the bands of phospho-SHP2 (pSHP2) and pERK1/2 indicate the relative intensity of each protein normalized to the signal intensity of SHP2 and ERK1/2, respectively. E, Enrichment plots for the gene sets of the MAPK/ERK pathway that were significantly upregulated in SLAMF3.KMS34 cells compared with ΔSLAMF3 cells. NES, normalized enrichment score; FDR, false discovery rate. The mRNA (F) and protein (G) expression of BCL2, CCND1, and CCND2 in SLAMF3-knockdown U266 cells. **, P < 0.01.

Close modal

SLAMF3 transmitted positive signals via the MAPK/ERK signal pathway to MM cells

The phosphatidylinositol 3-kinase (PI3K)/AKT and mitogen-activated protein kinase (MAPK)/ERK signal transduction pathways play a central role in MM progression (27, 28). Thus, we explored the signal pathway related to SLAMF3 molecules in MM cells. The expression levels of phosphorylated SHP2 and ERK1/2 proteins were decreased in SLAMF3-knockdown, -knockout, ΔSLAMF3, and NSC87877-treated MM cells, but phosphorylated AKT expression levels did not change (Fig. 4D). The adaptor proteins GRB2 and SHP2 did not differ between SLAMF3-knockdown/knockout or SLAMF3/ΔSLAMF3 cells and controls (Supplementary Fig. S5A). Moreover, the protein expression levels of phosphorylated SHP2, ERK1/2, and AKT in NSC87877-treated KMS34 cells did not change in comparison with controls (Supplementary Fig. S5B). Next, to clarify in detail the difference between MM cells with and without SLAMF3 expression, we compared gene-expression profiles between SLAMF3 and ΔSLAMF3 cells. GSEA identified 168 gene sets that were significantly enriched in SLAMF3 cells, including sets containing some SLAMF3, GRB2, or SHP2 genes (Supplementary Table S3). Interestingly, seven of these gene sets were associated with the MAPK/ERK signal transduction pathway (Fig. 4E; Supplementary Tables S3 and S4). Moreover, gene sets for the cell proliferation and survival signals associated with MM progression were significantly enriched in SLAMF3 cells (Supplementary Fig. S6). We next investigated the expression levels of cell cycle–related apoptotic and antiapoptotic genes using real-time PCR (Supplementary Fig. S7). The mRNA and protein expression of B-cell CLL/lymphoma 2 (BCL2), cyclin D1 (CCND1), and CCND2 were downregulated in SLAMF3-knockdown, -knockout, and ΔSLAMF3-expressing MM cells (Fig. 4F and G; Supplementary Fig. S8).

Relationship between soluble SLAMF3 levels and clinical significance in MM patients

Several reports showed that soluble forms of various immune-associated molecules were detected in sera from patients with different types of cancer and that the levels reflected tumor progression or poor survival (29–32). Therefore, to determine whether the levels are a useful prognostic factor, we measured soluble SLAMF3 (sSLAMF3) levels in sera from 16 MGUS and 96 newly diagnosed MM patients as well as from 16 healthy controls using ELISA (Supplementary Table S5). sSLAMF3 levels were significantly increased in MM patients compared with MGUS patients and healthy controls (Fig. 5A). Among MM patients, the levels were significantly higher in symptomatic than in asymptomatic patients (Fig. 5A), and markedly increased in advanced stages (ISS stage II/III and R-ISS stage III) of MM (Fig. 5B). sSLAMF3 was detected in the cell culture supernatants of SLAMF3-expressing MM cells and increased in a time-dependent manner (Fig. 5C). In MM cell culture supernatants and serum samples from MM patients, sSLAMF3 was detected at about 40 kDa (Fig. 5D). Among secreted-type matrix metalloproteinases (MMP), the mRNA expression of MMP-9 was detected in almost all MM cell lines (Supplementary Fig. S9A). Thus, when MM cells were treated with the MMP-9 inhibitor, sSLAMF3 levels in culture supernatants were markedly decreased in a concentration-dependent manner (Supplementary Fig. S9B). Cell growth and cytotoxicity were not affected by treatment with the MMP-9 inhibitor (data not shown). On the other hand, sSLAMF3 was detected in the cell supernatant of MMP-9–negative 293T and HeLa cells transfected with the SLAMF3 gene (Supplementary Fig. S9C). sSLAMF3 in serum of xenograft–bearing NOG mice was detected when the tumor volumes reached 2,000 mm3, but not in the serum of control mice (Fig. 5E). The levels of serum sSLAMF3 had a low correlation with the percentage of BM PCs or SLAMF3-positive cells, or serum IL6 concentration (Supplementary Fig. S10). Thus, we next investigated the differences in clinical characteristics between two groups according to serum sSLAMF3 levels: high (≥3.3 ng/mL, n = 63) and low (<3.3 ng/mL, n = 33). The optimal cutoff values of serum sSLAMF3 levels for predicting the R-ISS score (I vs. II/III) were determined using the ROC plot (Supplementary Fig. S11). MM patients with high levels of sSLAMF3 progressed to the advanced stage more frequently than those with low levels and had more aggressive clinical characteristics (Table 2). Furthermore, MM patients in the high group had significantly shorter PFS times than those in the low group (Fig. 5F). Finally, the cell proliferative advantage and drug resistance in MM cells were unchanged by cocultivation with recombinant human SLAMF3 (rhSLAMF3; Supplementary Fig. S12).

Figure 5.

Serum-soluble SLAMF3 (sSLAMF3) levels in MM patients. A and B, sSLAMF3 levels in serum samples from 96 MM and 16 MGUS patients and 16 healthy controls. **, P < 0.01. C, Detection of sSLAMF3 in culture supernatant of MM cell lines and SLAMF3-overexpressing KMS34 cells (SLAMF3). D, Cell culture supernatants in SLAMF3-expressing MM cells were immunoprecipitated by anti-SLAMF3 antibody (lanes 1 and 3) and the isotype control (lanes 2 and 4). Serum samples in sSLAMF3-positive (lane 5) and -negative (lane 6) MM patients were immunoprecipitated with anti-SLAMF3 antibody. sSLAMF3 bands were detected at about 40 kDa (arrows). E, Serum sSLAMF3 in SLAMF3 cell xenograft–bearing NOG mice (n = 5). Serum was collected when tumor volumes reached 2,000 mm3. The serum of NOG mice that rejected tumors after inoculation with ΔSLAMF3 cells was used as a control (n = 5). F, Kaplan–Meier estimates of PFS and OS in MM patients with high and low sSLAMF3 levels. HR, hazard ratio; 95% CI, 95% confidence interval.

Figure 5.

Serum-soluble SLAMF3 (sSLAMF3) levels in MM patients. A and B, sSLAMF3 levels in serum samples from 96 MM and 16 MGUS patients and 16 healthy controls. **, P < 0.01. C, Detection of sSLAMF3 in culture supernatant of MM cell lines and SLAMF3-overexpressing KMS34 cells (SLAMF3). D, Cell culture supernatants in SLAMF3-expressing MM cells were immunoprecipitated by anti-SLAMF3 antibody (lanes 1 and 3) and the isotype control (lanes 2 and 4). Serum samples in sSLAMF3-positive (lane 5) and -negative (lane 6) MM patients were immunoprecipitated with anti-SLAMF3 antibody. sSLAMF3 bands were detected at about 40 kDa (arrows). E, Serum sSLAMF3 in SLAMF3 cell xenograft–bearing NOG mice (n = 5). Serum was collected when tumor volumes reached 2,000 mm3. The serum of NOG mice that rejected tumors after inoculation with ΔSLAMF3 cells was used as a control (n = 5). F, Kaplan–Meier estimates of PFS and OS in MM patients with high and low sSLAMF3 levels. HR, hazard ratio; 95% CI, 95% confidence interval.

Close modal
Table 2.

Differences in clinical features between MM patients with high (≥3.3 ng/mL) and low (<3.3 ng/mL) sSLAMF3 levels.

sSLAMF3
Low (n = 33)High (n = 63)P value
Characteristics 
 Age Median (range) 69 (32–88) 71 (43–88) N.S. 
 Gender Male/female 17/16 29/34 N.S. 
 Diagnosis Asymptomatic/symptomatic 12/21 5/58 0.0005 
 ISS l/ll/lll 15/11/7 6/28/29 0.0002 
 R-ISS l/ll/lll/unknown 8/19/3/3 4/36/20/3 0.0058 
 DS l/ll/lll/unknown 8/13/8/4 5/10/48/0 <0.0001 
 M protein IgG/IgA/IgM/BJP/unknown 24/3/0/5/1 40/10/0/8/5 N.S. 
 Ig type κ/λ/unknown 19/9/5 27/31/5 0.0634 
 DS type A/B/unknown 26/3/4 51/12/0 N.S. 
 Bone lesions 0/1/2/3/unknown 13/5/6/6/3 16/6/5/34/2 0.0130 
Laboratory data 
 BM plasma cells 21.4 ± 18.2 35.0 ± 22.6 0.0045 
 White blood cell count (/μL) 5,231 ± 2,582 5,754 ± 2,433 N.S. 
 Hemoglobin (g/dL) 11.1 ± 2.08 9.45 ± 2.03 0.0007 
 Platelets (× 104/μL) 25.3 ± 27.8 19.5 ± 7.11 N.S. 
 LDH (IU/L) 180 ± 62.2 219 ± 81.3 0.0117 
 Creatinine (mg/dL) 1.09 ± 0.873 1.95 ± 3.28 N.S. 
 eGFR (mL/min) 64.4 ± 30.4 55.2 ± 30.6 N.S. 
 Corrected calcium (mg/dl) 9.55 ± 1.05 10.2 ± 1.43 0.0108 
 Albumin (g/dL) 3.54 ± 0.607 3.26 ± 0.797 0.0857 
 C-reactive protein (mg/dL) 0.338 ± 0.769 0.84 ± 1.97 0.0328 
 β2-microglobulin (μg/mL) 4.34 ± 4.28 8.32 ± 8.21 <0.0001 
 IL6 (pg/mL) 5.13 ± 7.05 12.9 ± 20.1 0.0010 
Cytogenetic abnormalities 
 t(4;14) +/–/unknown 6/24/3 6/50/7 N.S. 
 t(14;16) +/–/unknown 1/24/8 0/38/25 N.S. 
 del(17) +/–/unknown 0/29/4 5/48/10 0.0878 
sSLAMF3
Low (n = 33)High (n = 63)P value
Characteristics 
 Age Median (range) 69 (32–88) 71 (43–88) N.S. 
 Gender Male/female 17/16 29/34 N.S. 
 Diagnosis Asymptomatic/symptomatic 12/21 5/58 0.0005 
 ISS l/ll/lll 15/11/7 6/28/29 0.0002 
 R-ISS l/ll/lll/unknown 8/19/3/3 4/36/20/3 0.0058 
 DS l/ll/lll/unknown 8/13/8/4 5/10/48/0 <0.0001 
 M protein IgG/IgA/IgM/BJP/unknown 24/3/0/5/1 40/10/0/8/5 N.S. 
 Ig type κ/λ/unknown 19/9/5 27/31/5 0.0634 
 DS type A/B/unknown 26/3/4 51/12/0 N.S. 
 Bone lesions 0/1/2/3/unknown 13/5/6/6/3 16/6/5/34/2 0.0130 
Laboratory data 
 BM plasma cells 21.4 ± 18.2 35.0 ± 22.6 0.0045 
 White blood cell count (/μL) 5,231 ± 2,582 5,754 ± 2,433 N.S. 
 Hemoglobin (g/dL) 11.1 ± 2.08 9.45 ± 2.03 0.0007 
 Platelets (× 104/μL) 25.3 ± 27.8 19.5 ± 7.11 N.S. 
 LDH (IU/L) 180 ± 62.2 219 ± 81.3 0.0117 
 Creatinine (mg/dL) 1.09 ± 0.873 1.95 ± 3.28 N.S. 
 eGFR (mL/min) 64.4 ± 30.4 55.2 ± 30.6 N.S. 
 Corrected calcium (mg/dl) 9.55 ± 1.05 10.2 ± 1.43 0.0108 
 Albumin (g/dL) 3.54 ± 0.607 3.26 ± 0.797 0.0857 
 C-reactive protein (mg/dL) 0.338 ± 0.769 0.84 ± 1.97 0.0328 
 β2-microglobulin (μg/mL) 4.34 ± 4.28 8.32 ± 8.21 <0.0001 
 IL6 (pg/mL) 5.13 ± 7.05 12.9 ± 20.1 0.0010 
Cytogenetic abnormalities 
 t(4;14) +/–/unknown 6/24/3 6/50/7 N.S. 
 t(14;16) +/–/unknown 1/24/8 0/38/25 N.S. 
 del(17) +/–/unknown 0/29/4 5/48/10 0.0878 

Note: Gender and cytogenetic abnormalities were analyzed using the Fisher exact probability test. Diagnosis, ISS, R-ISS, DS, M protein, Ig type, DS type, and bone lesions were analyzed using the χ2 test. BM PCs, white blood cell count, hemoglobin, platelets, LDH, creatinine, eGFR, corrected calcium, albumin, C-reactive protein, β2-microglobulin, and IL6 were analyzed using the Mann–Whitney U test. N.S., not significant.

Our study showed that SLAMF3 is constitutively and highly expressed on PCs from patients with MM regardless of disease progression. Furthermore, SLAMF3 transmits positive signals to MM cells via the ERK signal transduction pathway mediated by the adaptor proteins GRB2 and SHP2. These results suggest that SLAMF3 expression on PCs plays a role in MM pathogenesis (Supplementary Fig. S13).

Other researchers also reported that SLAMF3 is highly expressed on PCs from patients with MGUS, smoldering MM, MM, and PC leukemia (13, 14, 19). However, the mechanism of SLAMF3 upregulation on PCs remains unknown. When MM cell lines were incubated with anti-MM agents (e.g., melphalan, bortezomib, and immunomodulatory drugs) and some cytokines (e.g., IFNα, IFNγ, TNFα, and IL6), or cocultivated with stromal HS-5 cells, SLAMF3 expression was not induced (data not shown). This suggests that SLAMF3 expression on MM cells is not strongly influenced by the BM microenvironment and anti-MM therapy and its levels may remain high as MM and related diseases develop.

Surprisingly, some MM cell lines that expressed high SLAMF3 levels contained minor cell fractions. SLAMF3low MM cells had the phenotypic characteristics of CD38 high expression but not of CD138 expression (Supplementary Fig. S14A). In cells from primary MM patients, there were minor cell fractions of SLAMF3low cells in SLAMF3 high-expressing MM samples (pattern I), which were absent in SLAMF3low cell fractions (pattern III; Supplementary Fig. S2B). Similar to MM cell lines, CD138 expression on SLAMF3low cells was decreased in comparison with that on SLAMF3high cells (Supplementary Fig. S14B). Other researchers reported SLAMF3 expression levels on pre–PC-propagating cells (CD19CD38highCD319+CD138 cells) were as high as on PCs (CD19CD38highCD319+CD138+ cells) in MM patients (19). pre-PCs in MM patients exhibited “cancer stem cell like-features,” which were relatively more cell-cycle quiescent and up to 300-fold more drug resistant in comparison with PCs (33). In our study, SLAMF3low cells were converted into SLAMF3high MM cells after SLAMF3low and SLAMF3high MM cells were subjected to cell sorting and then cultured for 1 week (Supplementary Fig. S15A). SLAMF3low MM cell fractions had increased G0–G1 cell populations in comparison with SLAMF3high cells (Supplementary Fig. S15B). Furthermore, the mRNA expression of retinoic acid receptor alpha 2 (RARα2) was higher in SLAMF3low cell fractions than in SLAMF3high cell fractions, and the gene expression of BCL2, c-myc, and CCND1, which are among RARα2-targeted genes, was markedly increased in SLAMF3low cell fractions (Supplementary Fig. S15C). CD138+ MM cells showed significantly higher cell growth compared with CD138 cells and were resistant to anti-MM agents such as lenalidomide, dexamethasone, and bortezomib (34, 35). CD138 MM cells from MM cell lines and patients cloud engraft and propagate MM progression in immunodeficient mice (34, 36). RARα2 levels are significantly higher in CD138 MM cells than in CD138+ cells and play an important role in maintaining myeloma stem cell “stemness” through activating Wnt and Hedgehog signaling (34, 37). Our data suggest that CD138SLAMF3low MM cells might have “myeloma stem cell–like” characteristics. However, our data also demonstrate that SLAMF3low MM cells show lower cell growth but higher sensitivity to bortezomib and melphalan in comparison with SLAMF3high MM cells.

SLAMF3 knockdown and knockout decreased the cell proliferative advantage and canceled drug resistance; conversely, SLAMF3 overexpression induced aggressive characteristics in MM cells. Atanackovic and colleagues also showed that downregulation of SLAMF3 by transfection with siRNA significantly decreased survival and antitumor activity in MM cells (13). Those findings suggest that SLAMF3 expression on MM cells can directly induce aggressive characteristics by self-ligand interaction between MM cells. Further, it is assumed that SLAMF3 on MM cells may interact not only with MM cells but also with SLAMF3-expressing lymphocytes in the BM microenvironment. SLAMF3 expressed on T cells recruits the adaptor protein SAP at phosphotyrosine (Y) of ITSM-1 and ITSM-2 (T-x-Y-x-x-V/I; refs. 38, 39), and SAP bridges SLAM molecules to the Src family kinase Fyn (40). This SLAM–SAP–Fyn complex mediates protein kinase Cθ-dependent NF-κB activation and induces Th2 differentiation (41). SLAMF3 expressed on B cells and macrophages interacts with EAT2 at tyrosine residues of ITSM-1/2 (42). However, all MM cells expressed SHP2 and GRB2, but not SAP and EAT2. Some MM cell lines also expressed CSK and SHP1, which are negative regulators of src family tyrosine kinases and JAK/STAT signaling, respectively (17). Therefore, we focused on SHP2 and GRB2, which act not only as negative but also as positive regulators in immune cells (17). SHP2 binding sites in SLAMF3 are the same as SAP binding sites, and it binds to ITSM-1/2 in the absence of SAP (38). Furthermore, the site at which GRB2 binds to SLAMF3 is different from the SAP binding sites, and GRB2 interacts directly with the C-terminal end of SLAMF3 (motif, YENE; ref. 43). SHP2 at phosphorylated Tyr580, which is the main binding site of GRB2, recruits GRB2, and this interaction induces activation of the MAPK/ERK signal transduction pathway (44). In addition, GRB2 controls SHP2 phosphatase activity (44). Phosphorylated ERK1/2 was decreased in SLAMF3-knockdown, -knockout, and ΔSLAMF3 cells. Furthermore, GSEA also showed that the MAPK/ERK signal cascades were significantly upregulated in SLAMF3 cells compared with ΔSLAMF3 cells. The MAPK/ERK signal pathway is essential for cell proliferation, cycling, survival, and drug resistance in MM (7, 45, 46). BCL2 and CCND1/2, which were downregulated in SLAMF3-knockdown, knockout, and ΔSLAMF3 cells, are a downstream target of the ERK signal pathway (47, 48). Our findings suggest that the interaction among SLAMF3, SHP2, and GRB2 may transmit the MAPK/ERK signal transduction pathway to MM cells and induce aggressive MM behaviors.

Finally, we demonstrated for the first time that serum sSLAMF3 levels greater than 3.3 ng/mL are significantly associated with poor PSF and aggressive MM stage. The SLAMF3 gene does not have an alternative splicing variant that lacks the transmembrane domain. Thus, since the soluble form of SLAMF3 was detected at about 40 kDa, it could be produced via cleavage of the extracellular domain (length, 407 aa; molecular weight, 44.8 kDa) by some mechanism such as MMP-9. In practice, MM cells constitutively express and secrete MMP-9, but not BM stromal cells (49). MMP-9 recognizes the most prevalent sequence motif Pro-X-X-H-Ser/Thr (where X is any residue, and H is a hydrophobic residue) and cleavage at P3 through P2′ (50). However, this sequence motif is not located at around 40 kDa of the extracellular domain of SLAMF3, suggesting that MMP-9 might recognize other sequences of SLAMF3. Moreover, sSLAMF3 was also generated from MMP-9–negative 293T and HeLa cells transfected with the SLAMF3 gene (Supplementary Fig. S9), although the mechanism remains unknown. Thus, sSLAMF3 might be produced from MM cells by both MMP-9 cleavage and another mechanism. The sSLAMF3 concentrations correlated with the percentage of BM PCs and IL6, a growth factor for MM, suggesting that the serum levels of sSLAMF3 might reflect the number of MM cells in the BM and disease progression of MM. Therefore, sSLAMF3 levels may be a useful new prognostic factor in MM (Supplementary Fig. S13). rhSLAMF3 bound surface SLAMF3 on MM cells, although this interaction neither increased cell proliferation nor decreased the apoptosis induced by anti-MM agents in vitro. The reason for this is that the SLAMF3–SLAMF3 interaction between MM cells and sSLAMF3 produced from MM cells already induces ERK signal transduction and thus the addition of rSLAMF3 might not change the activated ERK signaling in MM cells.

In conclusion, SLAMF3 molecules in MM may be not only new cell-surface markers but also a new therapeutic target of immunotherapy and novel agents such as small-molecule inhibitors. In particular, anti-SLAMF3 blocking antibody, which we are now trying to develop, alone or in combination with chemotherapy could demonstrate strong, promising effects in MM patients. Our study will also give new insight into the mechanism of aggressive myeloma behaviors in refractory/relapsed MM patients.

H. Handa reports receiving a commercial research grant from and has received honoraria from speakers bureau of an entity. S. Ito has received honoraria from speakers bureau of Bristol-Myers Squibb, Celgene, Takeda, Ono, and Janssen. H. Tamura has received honoraria from speakers bureau of Celgene, Takeda Pharma Co. Limited, and Janssen Pharmaceutical K.K. No potential conflicts of interest were disclosed by the other authors.

Conception and design: M. Ishibashi, H. Tamura

Development of methodology: M. Ishibashi, R. Takahashi, A. Tsubota

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): M. Ishibashi, M. Sasaki, H. Handa, Y. Imai, N. Tanaka, Y. Tsukune, S. Tanosaki, S. Ito, T. Asayama, M. Sunakawa, Y. Kaito, Y. Kuribayashi-Hamada, A. Onodera, K. Moriya, N. Komatsu, J. Tanaka, K. Inokuchi, H. Tamura

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): M. Ishibashi, R. Takahashi, A. Tsubota, H. Sugimori, H. Tamura

Writing, review, and/or revision of the manuscript: M. Ishibashi, H. Tamura

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): T. Odajima

Study supervision: H. Sugimori, K. Inokuchi, H. Tamura

This work was supported by Japan Society for the Promotion of Science KAKENHI grant numbers JP26461433 (H. Tamura) and JP17K16196 (M. Ishibashi), and International Myeloma Foundation Japan's grants (H. Tamura).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Palumbo
A
,
Anderson
K
. 
Multiple myeloma
.
N Engl J Med
2011
;
364
:
1046
60
.
2.
Moreau
P
,
Richardson
PG
,
Cavo
M
,
Orlowski
RZ
,
San Miguel
JF
,
Palumbo
A
, et al
Proteasome inhibitors in multiple myeloma: 10 years later
.
Blood
2012
;
120
:
947
59
.
3.
Mimura
N
,
Hideshima
T
,
Anderson
KC
. 
Novel therapeutic strategies for multiple myeloma
.
Exp Hematol
2015
;
43
:
732
41
.
4.
Rajkumar
SV
,
Kumar
S
. 
Multiple myeloma: diagnosis and treatment
.
Mayo Clin Proc
2016
;
91
:
101
19
.
5.
Touzeau
C
,
Moreau
P
,
Dumontet
C
. 
Monoclonal antibody therapy in multiple myeloma
.
Leukemia
2017
;
31
:
1039
47
.
6.
Hideshima
T
,
Anderson
KC
. 
Molecular mechanisms of novel therapeutic approaches for multiple myeloma
.
Nat Rev Cancer
2002
;
2
:
927
37
.
7.
Hideshima
T
,
Mitsiades
C
,
Tonon
G
,
Richardson
PG
,
Anderson
KC
. 
Understanding multiple myeloma pathogenesis in the bone marrow to identify new therapeutic targets
.
Nat Rev Cancer
2007
;
7
:
585
98
.
8.
Ishibashi
M
,
Tamura
H
,
Sunakawa
M
,
Kondo-Onodera
A
,
Okuyama
N
,
Hamada
Y
, et al
Myeloma drug resistance induced by binding of myeloma B7-H1 (PD-L1) to PD-1
.
Cancer Immunol Res
2016
;
4
:
779
88
.
9.
Tamura
H
,
Ishibashi
M
,
Yamashita
T
,
Tanosaki
S
,
Okuyama
N
,
Kondo
A
, et al
Marrow stromal cells induce B7-H1 expression on myeloma cells, generating aggressive characteristics in multiple myeloma
.
Leukemia
2013
;
27
:
464
72
.
10.
Tai
YT
,
Acharya
C
,
An
G
,
Moschetta
M
,
Zhong
MY
,
Feng
X
, et al
APRIL and BCMA promote human multiple myeloma growth and immunosuppression in the bone marrow microenvironment
.
Blood
2016
;
127
:
3225
36
.
11.
Murray
ME
,
Gavile
CM
,
Nair
JR
,
Koorella
C
,
Carlson
LM
,
Buac
D
, et al
CD28-mediated pro-survival signaling induces chemotherapeutic resistance in multiple myeloma
.
Blood
2014
;
123
:
3770
9
.
12.
Yamashita
T
,
Tamura
H
,
Satoh
C
,
Shinya
E
,
Takahashi
H
,
Chen
L
, et al
Functional B7.2 and B7-H2 molecules on myeloma cells are associated with a growth advantage
.
Clin Cancer Res
2009
;
15
:
770
7
.
13.
Atanackovic
D
,
Panse
J
,
Hildebrandt
Y
,
Jadczak
A
,
Kobold
S
,
Cao
Y
, et al
Surface molecule CD229 as a novel target for the diagnosis and treatment of multiple myeloma
.
Haematologica
2011
;
96
:
1512
20
.
14.
Pojero
F
,
Flores-Montero
J
,
Sanoja
L
,
Perez
JJ
,
Puig
N
,
Paiva
B
, et al
Utility of CD54, CD229, and CD319 for the identification of plasma cells in patients with clonal plasma cell diseases
.
Cytometry B Clin Cytom
2016
;
90
:
91
100
.
15.
Ma
CS
,
Deenick
EK
. 
The role of SAP and SLAM family molecules in the humoral immune response
.
Ann N Y Acad Sci
2011
;
1217
:
32
44
.
16.
Detre
C
,
Keszei
M
,
Romero
X
,
Tsokos
GC
,
Terhorst
C
. 
SLAM family receptors and the SLAM-associated protein (SAP) modulate T cell functions
.
Semin Immunopathol
2010
;
32
:
157
71
.
17.
Dong
Z
,
Veillette
A
. 
How do SAP family deficiencies compromise immunity?
Trends Immunol
2010
;
31
:
295
302
.
18.
Graham
DB
,
Bell
MP
,
McCausland
MM
,
Huntoon
CJ
,
van Deursen
J
,
Faubion
WA
, et al
Ly9 (CD229)-deficient mice exhibit T cell defects yet do not share several phenotypic characteristics associated with SLAM- and SAP-deficient mice
.
J Immunol
2006
;
176
:
291
300
.
19.
Yousef
S
,
Kovacsovics-Bankowski
M
,
Salama
ME
,
Bhardwaj
N
,
Steinbach
M
,
Langemo
A
, et al
CD229 is expressed on the surface of plasma cells carrying an aberrant phenotype and chemotherapy-resistant precursor cells in multiple myeloma
.
Hum Vaccin Immunother
2015
;
11
:
1606
11
.
20.
Kyle
RA
,
Rajkumar
SV
. 
Criteria for diagnosis, staging, risk stratification and response assessment of multiple myeloma
.
Leukemia
2009
;
23
:
3
9
.
21.
Durie
BG
,
Salmon
SE
. 
A clinical staging system for multiple myeloma. Correlation of measured myeloma cell mass with presenting clinical features, response to treatment, and survival
.
Cancer
1975
;
36
:
842
54
.
22.
Greipp
PR
,
San Miguel
J
,
Durie
BG
,
Crowley
JJ
,
Barlogie
B
,
Blade
J
, et al
International staging system for multiple myeloma
.
J Clin Oncol
2005
;
23
:
3412
20
.
23.
Palumbo
A
,
Avet-Loiseau
H
,
Oliva
S
,
Lokhorst
HM
,
Goldschmidt
H
,
Rosinol
L
, et al
Revised international staging system for multiple myeloma: a report from International Myeloma Working Group
.
J Clin Oncol
2015
;
33
:
2863
9
.
24.
Subramanian
A
,
Tamayo
P
,
Mootha
VK
,
Mukherjee
S
,
Ebert
BL
,
Gillette
MA
, et al
Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles
.
Proc Natl Acad Sci U S A
2005
;
102
:
15545
50
.
25.
Mootha
VK
,
Lindgren
CM
,
Eriksson
KF
,
Subramanian
A
,
Sihag
S
,
Lehar
J
, et al
PGC-1alpha-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes
.
Nat Genet
2003
;
34
:
267
73
.
26.
Schwartzberg
PL
,
Mueller
KL
,
Qi
H
,
Cannons
JL
. 
SLAM receptors and SAP influence lymphocyte interactions, development and function
.
Nat Rev Immunol
2009
;
9
:
39
46
.
27.
Steinbrunn
T
,
Stuhmer
T
,
Sayehli
C
,
Chatterjee
M
,
Einsele
H
,
Bargou
RC
. 
Combined targeting of MEK/MAPK and PI3K/Akt signalling in multiple myeloma
.
Br J Haematol
2012
;
159
:
430
40
.
28.
van de Donk
NW
,
Lokhorst
HM
,
Bloem
AC
. 
Growth factors and antiapoptotic signaling pathways in multiple myeloma
.
Leukemia
2005
;
19
:
2177
85
.
29.
Yang
ZZ
,
Grote
DM
,
Ziesmer
SC
,
Manske
MK
,
Witzig
TE
,
Novak
AJ
, et al
Soluble IL-2Ralpha facilitates IL-2-mediated immune responses and predicts reduced survival in follicular B-cell non-Hodgkin lymphoma
.
Blood
2011
;
118
:
2809
20
.
30.
Rossille
D
,
Gressier
M
,
Damotte
D
,
Maucort-Boulch
D
,
Pangault
C
,
Semana
G
, et al
High level of soluble programmed cell death ligand 1 in blood impacts overall survival in aggressive diffuse large B-cell lymphoma: results from a French multicenter clinical trial
.
Leukemia
2014
;
28
:
2367
75
.
31.
Hock
BD
,
McKenzie
JL
,
Patton
WN
,
Haring
LF
,
Yang
Y
,
Shen
Y
, et al
The clinical significance of soluble CD86 levels in patients with acute myeloid leukemia and myelodysplastic syndrome
.
Cancer
2003
;
98
:
1681
8
.
32.
Huang
SY
,
Lin
HH
,
Lin
CW
,
Li
CC
,
Yao
M
,
Tang
JL
, et al
Soluble PD-L1: a biomarker to predict progression of autologous transplantation in patients with multiple myeloma
.
Oncotarget
2016
;
7
:
62490
502
.
33.
Chaidos
A
,
Barnes
CP
,
Cowan
G
,
May
PC
,
Melo
V
,
Hatjiharissi
E
, et al
Clinical drug resistance linked to interconvertible phenotypic and functional states of tumor-propagating cells in multiple myeloma
.
Blood
2013
;
121
:
318
28
.
34.
Yang
Y
,
Shi
J
,
Tolomelli
G
,
Xu
H
,
Xia
J
,
Wang
H
, et al
RARalpha2 expression confers myeloma stem cell features
.
Blood
2013
;
122
:
1437
47
.
35.
Matsui
W
,
Wang
Q
,
Barber
JP
,
Brennan
S
,
Smith
BD
,
Borrello
I
, et al
Clonogenic multiple myeloma progenitors, stem cell properties, and drug resistance
.
Cancer Res
2008
;
68
:
190
7
.
36.
Hosen
N
,
Matsuoka
Y
,
Kishida
S
,
Nakata
J
,
Mizutani
Y
,
Hasegawa
K
, et al
CD138-negative clonogenic cells are plasma cells but not B cells in some multiple myeloma patients
.
Leukemia
2012
;
26
:
2135
41
.
37.
Gao
M
,
Kong
Y
,
Yang
G
,
Gao
L
,
Shi
J
. 
Multiple myeloma cancer stem cells
.
Oncotarget
2016
;
7
:
35466
77
.
38.
Sayos
J
,
Martin
M
,
Chen
A
,
Simarro
M
,
Howie
D
,
Morra
M
, et al
Cell surface receptors Ly-9 and CD84 recruit the X-linked lymphoproliferative disease gene product SAP
.
Blood
2001
;
97
:
3867
74
.
39.
Li
C
,
Iosef
C
,
Jia
CY
,
Han
VK
,
Li
SS
. 
Dual functional roles for the X-linked lymphoproliferative syndrome gene product SAP/SH2D1A in signaling through the signaling lymphocyte activation molecule (SLAM) family of immune receptors
.
J Biol Chem
2003
;
278
:
3852
9
.
40.
Chan
B
,
Lanyi
A
,
Song
HK
,
Griesbach
J
,
Simarro-Grande
M
,
Poy
F
, et al
SAP couples Fyn to SLAM immune receptors
.
Nat Cell Biol
2003
;
5
:
155
60
.
41.
Cannons
JL
,
Yu
LJ
,
Hill
B
,
Mijares
LA
,
Dombroski
D
,
Nichols
KE
, et al
SAP regulates T(H)2 differentiation and PKC-theta-mediated activation of NF-kappaB1
.
Immunity
2004
;
21
:
693
706
.
42.
Morra
M
,
Lu
J
,
Poy
F
,
Martin
M
,
Sayos
J
,
Calpe
S
, et al
Structural basis for the interaction of the free SH2 domain EAT-2 with SLAM receptors in hematopoietic cells
.
EMBO J
2001
;
20
:
5840
52
.
43.
Martin
M
,
Del Valle
JM
,
Saborit
I
,
Engel
P
. 
Identification of Grb2 as a novel binding partner of the signaling lymphocytic activation molecule-associated protein binding receptor CD229
.
J Immunol
2005
;
174
:
5977
86
.
44.
Zhang
J
,
Zhang
F
,
Niu
R
. 
Functions of Shp2 in cancer
.
J Cell Mol Med
2015
;
19
:
2075
83
.
45.
Chang-Yew Leow
C
,
Gerondakis
S
,
Spencer
A
. 
MEK inhibitors as a chemotherapeutic intervention in multiple myeloma
.
Blood Cancer J
2013
;
3
:
e105
.
46.
Hallek
M
,
Bergsagel
PL
,
Anderson
KC
. 
Multiple myeloma: increasing evidence for a multistep transformation process
.
Blood
1998
;
91
:
3
21
.
47.
Boucher
MJ
,
Morisset
J
,
Vachon
PH
,
Reed
JC
,
Laine
J
,
Rivard
N
. 
MEK/ERK signaling pathway regulates the expression of Bcl-2, Bcl-X(L), and Mcl-1 and promotes survival of human pancreatic cancer cells
.
J Cell Biochem
2000
;
79
:
355
69
.
48.
Musgrove
EA
. 
Cyclins: roles in mitogenic signaling and oncogenic transformation
.
Growth Factors
2006
;
24
:
13
9
.
49.
Barille
S
,
Akhoundi
C
,
Collette
M
,
Mellerin
MP
,
Rapp
MJ
,
Harousseau
JL
, et al
Metalloproteinases in multiple myeloma: production of matrix metalloproteinase-9 (MMP-9), activation of proMMP-2, and induction of MMP-1 by myeloma cells
.
Blood
1997
;
90
:
1649
55
.
50.
Kridel
SJ
,
Chen
E
,
Kotra
LP
,
Howard
EW
,
Mobashery
S
,
Smith
JW
. 
Substrate hydrolysis by matrix metalloproteinase-9
.
J Biol Chem
2001
;
276
:
20572
8
.

Supplementary data