A dominant-negative F-box deleted mutant of E3 ubiquitin ligase, β-TrCP1/FWD1, markedly reduces myeloma cell growth and survival in mice

Abstract

Treatment of multiple myeloma with bortezomib can result in severe adverse effects, necessitating the development of targeted inhibitors of the proteasome. We show that stable expression of a dominant-negative F-box deleted (∆F) mutant of the E3 ubiquitin ligase, SCFβ-TrCP/FWD1, in murine 5TGM1 myeloma cells dramatically attenuated their skeletal engraftment and survival when inoculated into immunocompetent C57BL/KaLwRij mice. Similar results were obtained in immunodeficient bg-nu-xid mice, suggesting that the observed effects were independent of host recipient immune status. Bone marrow stroma offered no protection for 5TGM1-∆F cells in cocultures treated with tumor necrosis factor (TNF), indicating a cell-autonomous anti-myeloma effect. Levels of p100, IκBα, Mcl-1, ATF4, total and cleaved caspase-3, and phospho-β-catenin were elevated in 5TGM1-∆F cells whereas cIAP was down-regulated. TNF also activated caspase-3 and downregulated Bcl-2, correlating with the enhanced susceptibility of 5TGM1-∆F cells to apoptosis. Treatment of 5TGM1 tumor-bearing mice with a β-TrCP1/FWD1 inhibitor, pyrrolidine dithiocarbamate (PDTC), significantly reduced tumor burden in bone. PDTC also increased levels of cleaved Mcl-1 and caspase-3 in U266 human myeloma cells, correlating with our murine data and validating the development of specific β-TrCP inhibitors as an alternative therapy to nonspecific proteasome inhibitors for myeloma patients.

Keywords: 5TGM1; FWD1/β-TrCP1; myeloma; plasmacytoma; proteasome.

Figure 1. Dominant-negative expression of β-TrCP1/FWD1 disrupts NF-κB signaling in myeloma cells

A. Representation of β-TrCP/FWD1ΔF (adapted from [12]). B. Immunoblotting for p50/105 (left panel) and p52/p100 (right panel) in lysates obtained from untreated 5TGM1-ΔF and 5TGM1-EV cells shows increased accumulation of p100 and decreased p52 levels in 5TGM1-ΔF myeloma cells. C. Cells were treated with 20 ng/ml TNF-α and lysates prepared. Immunoblotting shows increased basal levels of total IκBα in 5TGM1-ΔF cells. D. Untreated 5TGM1-ΔF and 5TGM1-EV cells were probed for ATF4, and for total and phosphorylated forms of β-catenin. Blots were normalized to actin.

Figure 2. Tumor burden is significantly reduced in disseminated myeloma mouse model bearing 5TGM1-ΔF myeloma cells

Control (Con) represents normal non-tumor-bearing mice injected with saline (n=4); EV = 5TGM1-EV-injected mice (n = 10); ΔF = 5TGM1-ΔF-injected mice (n = 10). A. 5TGM1 tumor burden assessed by serum IgG2bκ titer. B. Spleen wet weight at time of sacrifice. All mice had unequivocal evidence of myeloma tumor cells in spleen on hematoxylin and eosin (H&E)-stained sections. C. Representative photo-micrographs of serial sections of proximal tibial metaphyses from control mice (a,d) and mice intravenously inoculated with 5TGM1-EV (b,e) or 5TGM1-ΔF (c,f) myeloma cells stained either with H&E (a-c) or for tartrate-resistant acid phosphatase activity (TRAP; pinkish-red stain) to identify multi-nucleated osteoclasts (d-f). H&E-stained sections clearly show significantly increased tumor area in the bone marrow of mice inoculated with 5TGM1-EV mice B. as compared to control A. or 5TGM1-ΔF C. mice. GP= Growth Plate; B=trabecular bone; T=tumor. Arrowheads point to osteoclasts D. Tumor area per bone marrow area assessed by bone histomorphometry in the above H&E-stained sections of long bones (Counts of mice with no clearly discernible myeloma tumor in at least one leg bone: EV: 2/10; ΔF: 8/10. E. Osteoclast density represented as counts of tartrate-resistant acid phosphatase (TRAP⁺) multinucleated osteoclasts (OC; shown above) per mm bone tumor interface. In all cases, data represent mean ± SEM. NS, not significantly different; *, P < 0.05.

Figure 3. ΔF mutant attenuates myeloma cell growth in a cell-autonomous mode

A subcutaneous plasmacytoma model, in which tumor cells were inoculated subcutaneously in flank of syngeneic naïve mice, was used to determine the role of the bone marrow microenvironment. A. GFP expression of 5TGM1-EV and 5TGM1-ΔF cells analyzed by flow cytometry immediately before inoculation in mice. A single peak for each cell type indicates relative homogeneity of GFP expression in either population. Consistent with this observation, median GFP expression of the gated M1 populations did not differ significantly from the median GFP for all cells. B. Mice were inoculated subcutaneously with 5TGM1-EV cells (EV, n = 5) or 5TGM1-ΔF cells (ΔF, n = 5). Representative mice are shown (left panel); whole-body optical images of tumor-emitting green fluorescence in anesthetized live mice inoculated with above cells taken post-tumor inoculation (right panel). Tumor growth in mice inoculated with 5TGM1-ΔF cells (bottom) was markedly inhibited compared with those inoculated with 5TGM1-EV cells (top). C. Quantitative analysis of tumor volume (days 14 and 21 post-tumor cell inoculation) and excised tumor wet weight (at the end of the experiment; day 23 post-tumor cell inoculation). Days 14, 21 and 23 refer to the number of days after tumor cell inoculation in the flank of mice. Expression of ΔF in 5TGM1 cells almost completely inhibited plasmacytoma growth in vivo. Data represent mean ± SEM; *, P < 0.05. D. Tumor tissue was harvested from 5TGM1-EV- and 5TGM1-ΔF-injected mice, disaggregated, sieved, stained with annexin V-phycoerythrin and 7-AAD, and analyzed by flow cytometry to quantify apoptotic cells (annexin V⁺, 7-AAD⁻; lower-right quadrant). Apoptotic cells in tumor tissue harvested from 5TGM1-ΔF-inoculated mice increased by 10-fold compared with 5TGM1-EV tumors.

Figure 4. Bone marrow stromal cells do not protect 5TGM1-ΔF myeloma cells from TNF-α-induced apoptosis

A. Cytotoxic and proliferative effects of TNF-α on 5TGM1-EV and 5TGM1-ΔF myeloma cells were evaluated in medium containing 2% FBS with a standard MTS assay. Data represent mean ± SEM; *, P < 0.05. B. 5TGM1-EV or 5TGM1-ΔF cells were treated with either 100 ng/ml of IL-6 or 20 ng/mL of TNF-α for 24 h, washed them in PBS, and analyzed for apoptosis by flow cytometry after staining with propidium iodide. C. 5TGM1-EV or 5TGM1-ΔF cells were grown on a layer of 14M1 BMSCs and visualized for GFP expression after treatment with 20 ng/ml of TNF-α for 72 h. Viability of 5TGM1-ΔF cells decreased irrespective of the presence of BMSCs.

Figure 5. 5TGM1-ΔF myeloma cells exhibit constitutive upregulation of proapoptotic factors

A. Untreated 5TGM1 or 5TGM1-ΔF cells were probed for Mcl-2, Bcl-2, XIAP, cIAP and caspase-3 and normalized to actin. B. Western blotting of cell lysates shows reduced steady-state level of Bcl-2 protein in 5TGM1-ΔF cells that is rapidly cleared over time after treatment with 20 ng/ml TNF-α. C. Cells were treated with 10 μg/ml of cycloheximide for the indicated time points and harvested; immunoblotting of cell lysates shows significantly elevated half-life of procaspase-3 and time-dependent increase in cleaved caspase-3 in untreated 5TGM1-ΔF cells. D. Cells were treated for 15 min with TNF-α (0, 1, 10, and 20 ng/ml; top panel) or IL-6 (0, 1, 10, and 100 ng/ml; bottom panel) and harvested after 3 h into RIPA buffer followed by immunoblotting. Cleaved products of Mcl-1 are increased in untreated 5TGM1-ΔF cells.

Figure 6. Inhibition of β-TrCP by PDTC significantly reduces overall tumor burden in myeloma-bearing mice

5TGM1-bearing mice were injected with either PDTC dissolved in saline (prepared fresh as needed and injected intraperitoneally) or PBS (n = 5 each). A. Serum IgG2bκ levels were measured 30 days after tumor cell inoculation. Data represent mean ± SEM; *, P < 0.05 B. Immunoblotting of 5TGM1 lysates shows that PDTC promotes dose-dependent cleavage of Mcl-1. Arrowhead points to Mcl-1v (32 kDa) C. U266 cells were treated with 10, 50, 100 and 250 μM PDTC for 24 h and lysates probed for Mcl-1 and caspase-3.