A small molecule inhibitor of ubiquitin-specific protease-7 induces apoptosis in multiple myeloma cells and overcomes bortezomib resistance

Abstract

Bortezomib therapy has proven successful for the treatment of relapsed/refractory, relapsed, and newly diagnosed multiple myeloma (MM); however, dose-limiting toxicities and the development of resistance limit its long-term utility. Here, we show that P5091 is an inhibitor of deubiquitylating enzyme USP7, which induces apoptosis in MM cells resistant to conventional and bortezomib therapies. Biochemical and genetic studies show that blockade of HDM2 and p21 abrogates P5091-induced cytotoxicity. In animal tumor model studies, P5091 is well tolerated, inhibits tumor growth, and prolongs survival. Combining P5091 with lenalidomide, HDAC inhibitor SAHA, or dexamethasone triggers synergistic anti-MM activity. Our preclinical study therefore supports clinical evaluation of USP7 inhibitor, alone or in combination, as a potential MM therapy.

Figure 1. P5091 is a selective inhibitor of USP7 activity

(A) Chemical structure of P5091. (B) Schematic representation isopeptidase reporter assay: Cleavage by the of Ub-PLA₂ isopeptidase at the carboxy-terminal glycine of the Ub releases catalytically active PLA₂, which liberate a quantifiable fluorescent product in the presence of its substrate. (C) Structure-activity relationship data for selected analogs of P5091 (compound 1 = P22074 and compound 5 = P5091). (D) P5091 demonstrates potent and specific inhibition of USP7 in vitro versus other DUBs and proteases. (E) HEK293T cells were treated with DMSO or P5091; crude cell extracts were labeled with indicated UbVME probe, followed by immunoblotting (IB) with indicated antibodies. (F) P5091 inhibits higher molecular weight poly ubiquitin chain cleavage in a dose dependent manner. NEM (N-ethyl maleimide) is a non-specific inhibitor of USP7, and served as control. (G) HCT116 (WT) and HCT116 USP7^−/− cells were treated with vehicle or indicated concentrations of P5091 for 24 hr, 48 hr, and 72 hr, and then analyzed for viability using MTT assay (mean ± SD [Standard Deviation]; n=3). Error bars indicate SD. See also Figure S1.

Figure 2. Effect of P5091 on USP7 substrate HDM2

(A) U2OS cells were transfected with HA-Ub (WT), and 48 hr post-transfection, cells were treated with DMSO or P5091 (30 μM) for 4 hr. Endogenous HDM2 ubiquitylation was analyzed by pulling down ubiquitylated proteins using TUBEs (Supplementary Methods). Cell lysates were normalized for HDM2 protein and incubated with uncojugated Agarose or TUBE1 conjugated Agarose; samples were separated by SDS-PAGE, and HDM2 Ubiquitylation was detected using anti-HDM2 antibody. Whole cell extracts (WCE) were subjected to IB using anti-HDM2 and anti-actin antibodies. (B) U2OS cells were pretreated with either vehicle or P5091 (15 μM, IC₅₀) for 2–3.5 hr, followed by addition of CHX (40 μg/ml) for the indicated times. Total proteins lysates were subjected to immunoblotting with anti-HDM2 or anti-actin antibodies. Densitometry was utilized to quantify HDM2 protein levels after normalization with actin control to obtain percent HDM2 degradation (mean ± SD; n=3). Error bars indicates SD. See also Figure S2.

Figure 3. USP7 expression and prognostic relevance in MM cells

(A) IHC analysis of BM biopsies from normal donors and MM patients to show USP7 expression. Red arrowheads indicate USP7-positive cells (Brown color). Yellow arrowheads represent normal plasma cells; and Magenta arrowheads are neutrophils. Scale bars, 50 μm (10 μm in insets). (B) Purified tumor cells from MM patients, normal peripheral blood mononuclear cells (PBMCs), and MM cell lines were analyzed for USP7 expression level by IB with anti-USP7 and anti-actin antibodies. (C and D) Kaplan-Meier plots on prognostic relevance of USP7 expression on the overall and event-free survival for MM patients. Blue line indicates patient group with higher USP7 expression and shorter survival, whereas red line represents group of patients with lower USP7 expression and longer survival. See website (http://www.ncbi.nlm.nih.gov/geo/) for gene expression data under accession number GSE39754.

Figure 4. P5091 inhibits USP7 activity, induces MM cell death, and overcomes bortezomib-resistance

(A) MM.1S cells were treated with DMSO or with P5091 for 3 hr; and cellular USP7 was immunoprecipitated under native conditions and analyzed for USP7 activity using the Ub-EK_L assay (p < 0.05; n=2). (B) MM.1S cells were treated with DMSO or with P5091 or bortezomib for 3 hr; and protein lysates were analyzed for proteasome activity. Raw data were normalized to percent proteasome activity in control (as 100%) versus drug-treated cells (mean ± SD; n=3). Protein lysates from control, P5091-, or bortezomib-treated MM.1S cells were subjected to IB with anti-Ub antibody. (C) MM cell lines were treated with DMSO or with P5091 for 48 hr, followed by assessment for cell viability (mean ± SD; p < 0.001 for all cell lines; n=3). (D) MM.1S cells were cotransfected with scr-siRNA plus empty-vector, USP7-siRNA plus empty-vector, or USP7-siRNA plus active USP7 (WT), and subjected to growth analysis using CellTiter-Glo assay (mean ± SD; n=3). Immunoblot shows USP7 expression in cells transfected with scr-siRNA plus empty-vector (lane 1), USP7-siRNA plus empty-vector (lane 2), or USP7-siRNA plus active USP7 (WT) (lane 3) (Inset). (E) Purified patient MM cells were treated with P5091 for 48 hr, and analyzed for viability (mean ± SD of triplicate cultures; p < 0.001 for all patient samples). (F) MM cells from 6 bortezomib-resistant patients (Pt #1-#6) were treated in a side-by-side manner with either bortezomib or P5091 for 48 hr and analyzed for viability (mean ± SD of triplicate cultures; p < 0.001 for all patient samples). (G) Bortezomib-sensitive (ANBL-6.WT) and -resistant (ANBL-6.BR) cells were treated with either bortezomib or P5091 for 48 hr, followed by assessment for cell viability. ratio (ANBL-6.BR/ANBL-6.WT) of The bar graph shows the IC₅₀ P5091 and bortezomib (mean ± SD; n = 2). (H) Normal PBMCs were treated with P5091 for 48 hr, and analyzed for viability (mean ± SD of quadruplicate cultures). (I) MM.1S cells were cultured with or without BMSCs for 72 hr, and DNA synthesis was measured by ³H-TdR uptake (mean ± SD of triplicate cultures; p < 0.001 for all samples). (J) MM.1S cells were pretreated with inhibitors of caspase-8, caspase-9, or pan-caspase for 1 hr, and then P5091 (12.5 μM) or bortezomib (10 nM) was added for another 24 hr, followed by analysis of viability. Error bars (A–J) indicate SD. See also Figure S3.

Figure 5. Mechanisms mediating anti-MM activity of P5091

(A) MM.1S and RPMI-8226 cells were treated with P5091 (6.25 μM and 12.5 μM) for 24 hr; and protein lysates were subjected to IB with anti-HDM2, anti-HDMX, anti-p53, anti-p21, or anti-actin antibodies. (B) MM.1S cells were pretreated with P5091 (6.25 μM) for 4 hr, followed by addition of CHX (50 μg) for the final 0–120 mins of the experiment. Protein lysates were subjected to IB using indicated antibodies. (C) MM.1S cells were transfected with either scr- or HDM2-siRNA. 24 hr post-transfection, cells were treated with vehicle or indicated concentrations of P5091 for an additional 24 hr, followed by analysis of cell viability (mean ± SD; n = 3). (D) p53^−/− and p53^−/−/MDM2^−/− MEFs were treated with P5091 for 48 hr, followed by analysis of cell viability (mean ± SD; n = 3). (E) MM.1S cells were transfected with scr- or p53-siRNA, followed by treatment with P5091 or Nutlin-3A for 48 hr; and then analyzed for viability (mean ± SD; n = 3). (F) HCT116 p53 (WT) and HCT116 p53^−/− cells were treated with DMSO, P5091, or Nutlin-3A for 72 hr, followed by assessment for cell viability (mean ± SD; n=3). (G) ARP-1 MM cells were treated with DMSO or P5091 for 48 hr, and analyzed for cell viability (mean ± SD; n=3). Inset: Lysates from control- and P5091-treated ARP-1 cells were analyzed for HDM2 and p21 levels using IB. (H) MM.1S cells were treated with DMSO or P5091 (6.25 μM) for 12 hr; cell lysates were immunoprecipitated with anti-p21, anti-HDM2, or corresponding isotype control (mock) antibodies, and the immune complexes were then subjected to IB with anti-HDM2 and anti-p21 antibodies. (I) MM.1S cells were transfected with scr- or p21-siRNA, followed by treatment with P5091 for 48 hr; and then analyzed for viability (mean ± SD; n = 3). Percent cell viability was normalized (as 100%) for scr- or p21-siRNA controls, respectively. (J) HCT116 (WT) and HCT116 p21^−/− cells were treated with DMSO or with P5091 for 48 hr, followed by assessment for cell viability (mean ± SD; p < 0.001; n=3). Error bars in C-G, I, and J indicate SD. See also Figure S4.

Figure 6. P5091 inhibits xenografted human MM cell growth, prolongs survival, and blocks angiogenesis in CB-17 mice

(A) Mice bearing human MM.1S MM tumors were treated with vehicle or P5091 (10 mg/kg; IV) twice weekly for 3 consecutive weeks. Data are presented as mean tumor volume ± SD (10 mice/group). (B) Kaplan-Meier survival plot shows survival of mice. (C) Tumors from vehicle- or P5091-treated mice were analyzed for USP7 activity. Inset: Immunoblot shows equal USP7 protein input. HDM2 and p21 levels in mice tumors were analyzed by IB. Tumor sections from control and P5091-treated mice were immunostained with anti-HDM2 antibody and DAPI. Scale bars, 50 μm. (D) Tumor sections from control and P5091-treated mice were immunostained with BrdU and Ki67 antibodies. Nuclear staining was performed with DAPI. White rectangle in the middle control panel represents the 5 X enlarged area. White Scale bars, 50 μm. Yellow scale bars, 10 μm. (E) Apoptotic cells in tumors sectioned from control- or P5091-treated mice were identified by immunostaining for activated caspase-3 (Red cells), TUNEL (Green cells), and DAPI (blue). White rectangle represents 5 X enlarged area. Black arrowheads point at apoptotic cells. Scale bars, 50 μm (10 μm in the inset). (F) Tumor sections from control and P5091-treated mice were subjected to immunostaining with anti-PECAM1, -VEGFR2, -LYVE1, or -ENG antibodies. Nuclear staining was performed with DAPI. Scale bars, 50 μm. (G) Tumor sections from control and P5091-treated mice were subjected to dual immunostaining with PECAM1 and ACTA2. White rectangle represents 5 X enlarged area. Scale bars, 50 μm. (H) Quantification of ACTA2-positive cells shown in (G). *p < 0.05. Error bars represent SD. See also Figure S5.

Figure 7. P5091 targets USP7 in vivo, and exhibits anti-tumor activity in distinct animal models

(A) Mice bearing human p53-null ARP-1 MM tumors were treated with either P5091 (10 mg/kg; IV) or vehicle control twice weekly for 3 consecutive weeks. Average and standard deviation of tumor volume (mm³) is shown from mice versus time when tumor was measured (mean tumor volume ± SD, 6 mice/group). (B) Kaplan-Meier plot shows survival in mice. (C) Tumors from vehicle- or P5091-treated mice were analyzed for USP7 activity. Immunoblot shows that equal USP7 protein input. (D) Average and standard deviation of tumor volume (mm³) is shown from mice (n = 5/group) versus time when tumor was measured. RPMI-8226 cells (5 × 10⁶ cells/mouse) were implanted in the rear flank of female mice. On Day 28–30, mice were randomized to treatment with vehicle, MTD doses of P5091 (20 mg/kg), or with bortezomib (1 mg/kg) on a twice-weekly schedule for 3 weeks. (E) SCID-hu mice bearing INA-6 MM tumors (5 mice/group) were treated with either vehicle alone or P5091 (10 mg/kg), and mouse serum samples were analyzed for shIL-6R. (F) Mice bearing CCRF-CEM leukemic tumors were treated with either vehicle or P5091 (10 mg/kg) twice weekly for 3 consecutive weeks. Data are presented as mean tumor volume ± SD (8 mice/group). (G) Kaplan-Meier survival plot shows survival of mice. Error bars indicate SD. See also Figure S6.

Figure 8. Combination of P5091 and lenalidomide, SAHA, or Dex trigger synergistic anti-MM activity

(A) MM.1S cells were treated for 48 hr with P5091, lenalidomide, or P5091 plus lenalidomide; and then assessed for viability using MTT assays. Isobologram analysis shows the synergistic cytotoxic effect of P5091 and lenalidomide. The graph (right panel) is derived from the values given in the Table (left panel). The numbers 1–9 in graph represent combinations shown in the Table. FAcom = Fraction of viable cells. Combination index (CI) < 1 indicates synergy. (B) MM.1S cells were treated for 48 hr with P5091, SAHA, or P5091 plus SAHA; and then assessed for viability using MTT assays. Synergistic anti-MM activity was analyzed as in panel A. (C) MM.1S cells were treated for 48 hr with P5091, Dex, or P5091 plus Dex; and then assessed for viability using MTT assays. Synergistic anti-MM activity was analyzed as in panel A.