Intracellular NAD⁺ depletion enhances bortezomib-induced anti-myeloma activity

Abstract

We recently demonstrated that Nicotinamide phosphoribosyltransferase (Nampt) inhibition depletes intracellular NAD⁺ content leading, to autophagic multiple myeloma (MM) cell death. Bortezomib has remarkably improved MM patient outcome, but dose-limiting toxicities and development of resistance limit its long-term utility. Here we observed higher Nampt messenger RNA levels in bortezomib-resistant patient MM cells, which correlated with decreased overall survival. We demonstrated that combining the NAD⁺ depleting agent FK866 with bortezomib induces synergistic anti-MM cell death and overcomes bortezomib resistance. This effect is associated with (1) activation of caspase-8, caspase-9, caspase-3, poly (ADP-ribose) polymerase, and downregulation of Mcl-1; (2) enhanced intracellular NAD⁺ depletion; (3) inhibition of chymotrypsin-like, caspase-like, and trypsin-like proteasome activities; (4) inhibition of nuclear factor κB signaling; and (5) inhibition of angiogenesis. Furthermore, Nampt knockdown significantly enhances the anti-MM effect of bortezomib, which can be rescued by ectopically overexpressing Nampt. In a murine xenograft MM model, low-dose combination FK866 and Bortezomib is well tolerated, significantly inhibits tumor growth, and prolongs host survival. Taken together, these findings indicate that intracellular NAD⁺ level represents a major determinant in the ability of bortezomib to induce apoptosis in MM cells and provide proof of concept for the combination with FK866 as a new strategy to enhance sensitivity or overcome resistance to bortezomib.

Figure 1

Nampt is a cytoplasmic protein with prognostic relevance in bortezomib-treated MM patients. (A-B) MM cell lines, MM patients’ CD138⁺ cells, and PBMCs from healthy donors or MM patients were used to characterize subcellular distribution patterns of Nampt expression by immunofluorescence (A) or western blot analysis (B) using anti-Nampt and specific antibodies. (C) Expression levels (log₁₀ transformed) for Nampt transcript in CD138⁺ cells from MM patients after bortezomib (n = 163) or dexamethasone (n = 70) therapy, according to gene expression profile arrays generated at Millenium Pharmaceuticals (GSE9782). Nampt expression levels (Affimetrix probeset 217739_s_at) among responders (R) or nonresponders (NR) within Bz (n = 85 for R and n = 78 for NR) or Dex (n = 30 for R and n = 40 for NR) groups were plotted on the horizontal axis against the log₁₀-transformed normalized expression units on the vertical axis. For each therapy group, P values comparing R and NR are shown. (D) Kaplan–Meier overall survival curves of MM patients treated with bortezomib (n = 163) according to Nampt mRNA expression above or below the median value of 250.18, based on gene expression omnibus dataset GSE9782. The blue line indicates a patient group with lower Nampt expression and longer survival, whereas the red line represents a group of patients with higher Nampt expression and shorter survival. Bz, bortezomib; DAPI, 4,6 diamidino-2-phenylindole; Dex, dexamethasone; HMMCs, human multiple myeloma cell lines.

Figure 2

Combination of low doses of FK866 and bortezomib induces synergistic anti-MM activity. (A) RPMI-8226/S, U266, MM1S, MM1R ANBL6/WT, and ANBL6-BR cells were treated with or without increasing doses of FK866 (1–3 nM) for 48 hours, and then vehicle or bortezomib (over a range of concentrations depending on cell line) were added for a further 48 hours. Viability was assessed using PI staining and FACS analysis. Data presented are means of triplicate ± SD (n = 3; P < .05 for all cell lines). CI values refer to the highest drug concentrations used. (B) Purified patient bortezomib-sensitive and -resistant MM (CD138⁺) cells were pretreated with FK866 for 48 hours; bortezomib was then added for an additional 48 hours, followed by cell death analysis using PI staining and FACS analysis. Data are mean ± standard deviation (SD) of triplicate samples (P < .05 for all patient samples). A CI less than 1 indicates synergism. (C) PBMCs from healthy donors were treated as in panel B with indicated concentration of FK866, bortezomib, or their combination, and then analyzed for viability as described above.

Figure 3

Mechanisms mediating the anti-MM activity of FK866 plus bortezomib. (A) MM-1S, U266, ANBL6/WT, and ANBL6/BR cells were pretreated with or without low-dose FK866 (3 nM) for 24 hours, and then bortezomib (2 nM) was added for an additional 24 hours. Cells were then harvested, and whole-cell lysates were subjected to immunoblot analysis using anti-PARP, anti-caspase 3, anti-caspase 8, anti-caspase 9, anti-Mcl-1, anti bcl-2, or anti-actin antibodies. (B) Bortezomib-sensitive (ANBL6/WT) and bortezomib-resistant (ANBL6/BR) cells were treated with FK866 (3 nM), bortezomib (2 nM), or combined therapy for 72 hours, followed by Annexin V/PI staining and flow cytometry analysis. CF, cleaved fragment; FL, full length.

Figure 4

Combined low doses of FK866 and bortezomib overcome the survival advantage conferred by the BM microenvironment and inhibit in vitro capillary-like tube formation. (A) MM1S cells were cultured for 72 hours in BMSC-coated or -uncoated wells with control media, FK866, bortezomib, or FK866 plus bortezomib. Cell proliferation was assessed by [³H]thymidine incorporation assay. Data are mean ± SD of triplicate samples. Error bars represent SD. (B) MM1S cells were treated with FK866, bortezomib, or their combination in the presence or absence of rhIL-6 (10 ng/mL) or rhIGF-1 (100 mg/mL) for 72 hours; DNA synthesis was then determined by [³H]thymidine uptake. The results are mean ± SD of triplicate samples. (C) HUVEC were treated with FK866 (10 nM), bortezomib (2.5 nM), or their combination for 8 hours and then assessed for in vitro angiogenesis using Matrigel capillary-like tube structure formation assay. Endothelial cell tube formation was analyzed by microscopy (magnification: ×40/ DIC NA 0.75). Image is representative of 3 experiments with similar results (left). Bar graph represents quantification of tube formation in left panel in response to indicated stimuli: branch points in several random views fields/well were counted, values were averaged, and statistical significance of differences was measured using the Student t test (right). Error bars represent SD. CPM, [³H]thymidine incorporation.

Figure 5

Role of Nampt in modulating response to bortezomib. (A) MM1S cells were infected with either lentiviral construct expressing control scrambled or shRNA targeting Nampt. Total protein extracts were then subjected to immunoblot analysis with anti-Nampt or anti-GAPDH antibodies (left). The effect of Nampt knockdown on bortezomib response was assessed by measuring viability of infected cells (with shRNA scramble or targeting Nampt) after bortezomib treatment (1.25–5 nM) by using PI staining followed by FACS analysis (right). (B) Representative immunoblot images showing Nampt overexpressed in MM1S and U266 cells. Antitubulin monoclonal antibody served as loading control (top). Relative expression of Nampt protein was calculated by taking the ratio of the densitometry signal for Nampt to tubulin in each sample using Image J software (bottom). Cells overexpressing Nampt were subjected to bortezomib treatment of 24 h, and then viability was measured by MTT analysis (right). (C) U266 and MM1S cells were infected with a specific lentiviral shNampt or scramble control and then treated with 2 nM bortezomib for 24 hours. Thereafter, cells were used for cell lysates preparation, and Nampt, Mcl-1, bcl-2, and tubulin were detected by immunoblotting.

Figure 6

Combination treatment inhibits UPS and NF-κB pathway in MM cells. (A) MM cell lines were treated with FK866 (3 nM), bortezomib (2 nM), or combined therapy for 3 hours. Cells were then harvested, and intracellular NAD⁺ level was measured using an enzyme cyclic assay and normalized to total cell number. Data are mean ± SD of 2 independent experiments. (B) MM1S cells were treated with FK866 (3 nM) for 48 hours, and bortezomib (2 nM) was added for the last 6 hours. Cell extracts were then analyzed for 20S proteasome activities (CT-L, C-L, and T-L). Results are percentage inhibition of proteasome activities in drug in comparison with vehicle control–treated cells. Data are representative of 3 independent experiments. (C) MM1S and U266 cells were treated with FK866 (3 nM), bortezomib (2 nM), or combined therapy for 24 hours. Whole-cell lysates were then immunoblotted using antiubiquitin and antiactin Abs. Blots shown are representative of 3 independent experiments. (D) MM1S cells were treated with FK866 (3 nM), bortezomib (2 nM), or their combination for 24 hours. Cells were then fixed and stained with 4′,6-diamidino-2-phenylindole (blue) and antiubiquitin Ab. (E) MM1S cells were cultured with FK866 (3 nM), bortezomib (2 nM), or their combination for 6 hours, with TNF-α (10 ng/mL) added for the last 20 minutes. Cytoplasmic and nuclear extracts were subjected to western blotting using specific antibody for analysis of NF-κB canonical (anti–p-NF-κBp65 and -p-IκB) and noncanonical (-NF-κBp52, -RelB) activity. (F) MM1S cells were cultured with FK866 (3 nM), bortezomib (2 nM), or the combination for 6 hours, with TNF-α (10 ng/mL) added for the last 20 minutes. Immunocytochemical analysis was performed using anti-pospho-NF-κBp65 antibody. DAPI (4′,6-diamidino-2-phenylindole) was used to stain nuclei.

Figure 7

FK866 plus bortezomib trigger synergistic inhibition of human MM cell growth in vivo. (A) Average and SD of tumor volume (mm³) from groups of mice (n = 7 per group) versus time (days) when tumor was measured. MM1S cells (5 × 10⁶ in 100 µL of serum-free RPMI-1640 medium) were implanted in the flank of CB17 SCID mice. After tumor detection, mice were randomized to intraperitoneal treatment with vehicle, FK866, bortezomib, or their combination at the indicated doses over 3 weeks. A significant decrease in tumor growth was noted in combination-treated mice versus vehicle-treated mice (P = .0045 after the first week and P < .001 at the end of treatment). Data are mean tumor volume ± SD. Error bars represent mean ± SD. (B) Kaplan–Meier survival plot showing survival for mice treated with vehicle, FK866, bortezomib, or their combination at the indicated concentrations. FK866 plus bortezomib-treated mice show significantly increased survival (P = .007) in comparison with vehicle-treated mice. The mean overall survival was 20 days in the vehicle-treated cohort versus 37 days in the combination-treated cohort. (C) Micrographs show tumors sectioned on day 30 (endpoint) from vehicle-, FK866- (30 mg/kg), bortezomib- (0.5 mg/kg), or combination-treated mice immunostained for Ki-67, caspase 3, or CD31 expression. Photographs are representative of 2 mice receiving each treatment. (D) Cell lysates were prepared from tumor tissues harvested from treated and untreated mice and then analyzed by western blot analysis for Mcl-1 and Bcl-2 protein level.