Role of Bruton's tyrosine kinase in myeloma cell migration and induction of bone disease

Abstract

Myeloma cells typically grow in bone, recruit osteoclast precursors and induce their differentiation and activity in areas adjacent to tumor foci. Bruton's tyrosine kinase (BTK), of the TEC family, is expressed in hematopoietic cells and is particularly involved in B-lymphocyte function and osteoclastogenesis. We demonstrated BTK expression in clinical myeloma plasma cells, interleukin (IL)-6- or stroma-dependent cell lines and osteoclasts. SDF-1 induced BTK activation in myeloma cells and BTK inhibition by small hairpin RNA or the small molecule inhibitor, LFM-A13, reduced their migration toward stromal cell-derived factor-1 (SDF-1). Pretreatment with LFM-A13 also reduced in vivo homing of myeloma cells to bone using bioluminescence imaging in the SCID-rab model. Enforced expression of BTK in myeloma cell line enhanced cell migration toward SDF-1 but had no effect on short-term growth. BTK expression was correlated with cell-surface CXCR4 expression in myeloma cells (n = 33, r = 0.81, P < 0.0001), and BTK gene and protein expression was more profound in cell-surface CXCR4-expressing myeloma cells. BTK was not upregulated by IL-6 while its inhibition had no effect on IL-6 signaling in myeloma cells. Human osteoclast precursors also expressed BTK and cell-surface CXCR4 and migrated toward SDF-1. LFM-A13 suppressed migration and differentiation of osteoclast precursors as well as bone-resorbing activity of mature osteoclasts. In primary myeloma-bearing SCID-rab mice, LFM-A13 inhibited osteoclast activity, prevented myeloma-induced bone resorption and moderately suppressed myeloma growth. These data demonstrate BTK and cell-surface CXCR4 association in myeloma cells and that BTK plays a role in myeloma cell homing to bone and myeloma-induced bone disease. Am. J. Hematol. 88:463-471, 2013. © 2013 Wiley Periodicals, Inc.

Figure 1

Primary myeloma cells and IL-6 or stroma-dependent myeloma cell lines expressed BTK. A: GEP analysis demonstrated expression of BTK in bone marrow B-lymphocytes (BC, n = 6), plasma cells from healthy donors (NPC, n = 25), patients with MM (MM, n = 559), myeloma cell lines (MMCL, n = 42), cultured osteoclasts (OC, n = 8), and mesenchymal stem cells (MSCs, n = 15). Note that BTK expression is absent in MSCs, high in B-lymphocytes and variably expressed in primary myeloma plasma cells, while expressed at very low levels in most MM lines. B: Validation of BTK expression by qRT-PCR is shown for MM cells freshly obtained from nine patients, independent MM lines (OPM2, JJN3, ARP1, CAG, H929), the stroma–dependent BN line, IL-6–dependent INA6 and ANBL6 lines and osteoclast precursors (pOC). C: Western blot analyses: upper panel demonstrated variable expression of BTK in cell lines H929 and INA6, two freshly obtained primary MM plasma cell samples and osteoclast precursors (pOC). The Namwala (a human Burkitt’s Lymphoma line) cell lysate was used as positive control. Lower panel showed BTK expression in four freshly obtained myeloma plasma cell samples. The MEG-01 (Chronic Myelogenous Leukemia whole cell lysate) was used as positive control.

Figure 2

Inhibition of BTK by shRNA or LFM-A13 impeded migration of primary myeloma cells toward SDF-1. A. Knockdown of BTK in INA6 cells using lentiviral containing shRNA. Upper panel; relative expression of BTK as assessed by qRT-PCR in the INA6 cell line infected with lentiviral-containing non-target scramble (SCR) or BTK shRNA. Lower panel: Western Blot analysis demonstrating marked reduction in BTK protein in BTK knockdown INA6 cells. B: Introduction of BTK shRNA inhibited migration of INA6 MM cells toward SDF-1 (30 nM). C: Migration assay of the INA6 MM cell line in the presence of SDF-1 and its inhibition by pharmacological BTK inhibitor LFM-A13 (25 μM). D: Average effect of LFM-A13 (25 μM) on SDF-1-induced migration of primary myeloma cells from 8 patients. Cell viability in all experiments was >90%. Note significant inhibition of SDF-1-induced migration by LFM-A13. Individual migration assays for each of the primary myeloma cell samples are shown in Supporting Information Fig. 1A.

Figure 3

BTK overexpression had stimulatory effect on in vitro migration of MM cell line. CAG MM line was stably infected with BTK or control eGFP constructs as described in “Materials and Methods”. A: Expression of BTK in untreated CAG cells and in eGFP- and BTK-overexpressing CAG cells. Upper panel: relative expression of BTK as assessed by qRT-PCR. Lower panel: Western blot validation of BTK overexpression. β-actin is used as a loading control. B: Enforced expression of BTK in CAG cells increased spontaneous migration and migration toward SDF-1 compared to untreated cells or cells enforced to express eGFP.

Figure 4

Inhibition of BTK impeded migration and homing of myeloma cells to bone in vivo. A: X-ray radiograph of a SCID-rab mouse demonstrating the model used for the study. B: Luciferase-expressing INA6 MM cells were pretreated with LFM-A13 (50 μM) or 0.01% DMSO overnight and then intravenously (IV) injected in SCID-rab mice (5 × 10⁶ cells/mouse, 3 mice/group). Live-animal imaging taken 2 hr after cell injection demonstrated localization of control INA6 cells (CONT) but not localization of LFM-A13–treated INA6 cells in implanted bones. C: Individual implanted bones were crunched and subjected to ex vivo bioluminescence imaging. Note reduced bioluminescence of bones from mice injected with LFM-A13–pretreated INA6 MM cells. Numbers on the side of each well represent bioluminescence intensity (p/sec/cm²/sr). D: Quantification of bioluminescence intensity showed an approximately six fold decrease (P <0.03) in homing of INA6 MM cells treated with LFM-A13 compared to the control cells. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 5

BTK expression was associated with cell-surface CXCR4, and BTK was phosphorylated (activated) by SDF-1 in myeloma cells. A: Representative histograms of myeloma primary (n = 3) and cell line (n = 3) samples demonstrating expression of cell-surface CXCR4 (% of cell-surface expression). B: Correlation coefficients indicate the strength of the linear relationship between the percent of cells expressing cell-surface CXCR4 (assessed by flow cytometry) and BTK expression (assessed by qRT-PCR) in myeloma cells (n = 33) including 28 freshly obtained primary myeloma cells (●) and the 5 cell lines CAG, U266, INA6, H929, and ANBL6 (▲). C: Immunoblot showing increase in phosphorylation of BTK (P-BTK) at Tyrosine 223 residue (activation state) over time by SDF-1 stimulation of INA6 cultured with no serum (starvation) for 3 hr. Note high baseline phosphorylation of BTK and no further phosphorylation of BTK by SDF-1 in serum condition. D: Anti-BTK immunoblot representing total phosphorylated BTK in INA6 cell lysates immunoprecipitated using 4G10 phosphotyrosine immunobeads. Increased BTK phosphorylation is observed at 5 min of SDF-1 stimulation only in serum-starved INA6 cells. E: Anti-BTK immunoblot representing total phosphorylated BTK in serum-starved H929 cell lysates immunoprecipitated using 4G10 phosphotyrosine immunobeads. Rapid phosphorylation of BTK at 2 minutes and decrease within 5 min of SDF-1 stimulation is observed. F: Anti-BTK immunoblot representing total phosphorylated BTK in serum-starved primary myeloma plasma cell lysates immunoprecipitated using 4G10 phosphotyrosine immunobeads. Increased BTK phosphorylation is observed at 5 min of SDF-1 stimulation. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 6

BTK expression was higher in cell-surface CXCR4-sorted myeloma cells. Three MM cell lines (INA6, H929 and CAG) and a patient’s myeloma cells were sorted based on cell-surface CXCR4. A: Representative fluorescence-activated cell sorting (FACS) analysis of myeloma cells sorted by CXCR4⁺ and CXCR4⁻ cell-surface expression. The top panels demonstrate flow cytometry staining with control IgG (left) and CXCR4 antibody (right). The bottom panels demonstrate CXCR4⁻ (left) and CXCR4⁺ (right) myeloma cells after sorting. B: Expression of BTK in one primary cell sample and indicated MM cell lines sorted by CXCR⁺ and CXCR4⁻ as assessed by qRT-PCR. Data are expressed as fold change between CXCR4+ and CXCR4− sorted MM cells from each sample. C: Immunohistochemistry demonstrated greater BTK staining in primary myeloma cells sorted with CXCR4⁺ than with CXCR4⁻ (left panels ×200 original magnification; right panels ×400 original magnification). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 7

BTK inhibition reduced migration and differentiation of osteoclast precursors and bone resorption activity of mature osteoclasts. A: Flow cytometry demonstrating expression of cell-surface CXCR4 in osteoclast precursors sorted from blood using CD14 immunomagnetic beads. B: Migration assays demonstrated migration of osteoclast precursors toward SDF-1 (30 nM), an effect that was suppressed by LFM-A13 (25 μM). C: LFM-A13 inhibited differentiation of osteoclast precursors in medium supplemented with RANKL and M-CSF in a dose-related manner (each P value indicates significant dose effect against untreated control). D: Representative photographs demonstrating TRAP staining in osteoclast precursor cultures treated with vehicle (CONT) or 10 μM LFM-A13. E: LFM-A13 (10 μM) suppressed pit formation by mature osteoclasts on dentine slices. All functional assays were performed in triplicate. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 8

In vivo, LFM-A13 markedly inhibited myeloma-induced bone disease and moderately reduced myeloma growth. SCID-rab mice engrafted with primary myeloma cells, which express high levels of BTK (see Fig. 1, Primary #2), were treated with LFM-A13 intraperitoneally (n = 10, 40 mg/kg, twice daily) or vehicle (n = 10) for 3 weeks. A: Bone Mineral Density (BMD) changes in implanted, myelomatous bones demonstrated reduced BMD in the control group but maintained BMD levels after treatment with LFM-A13. B: X-ray images of implanted, myelomatous bones prior to treatment (Pre-Rx) and at experiment’s end (Final) in five representative mice from each group (Control and LFM-A13). Note bone loss in mice treated with the control compared to bone preservation in mice after treatment with LFM-A13. C: Histological sections of the implanted myelomatous bones revealed a reduced number of osteoclasts (OC) when treated with LFM-A13. D: MM burden monitored by measurement of circulating human immunoglobulins demonstrated reduced MM growth after treatment with LFM-A13, at a level close to statistical significance (P < 0.07).