Induction of Synthetic Lethality by Activation of Mitochondrial ClpP and Inhibition of HDAC1/2 in Glioblastoma

Abstract

Purpose: Novel therapeutic targets are critical to unravel for the most common primary brain tumor in adults, glioblastoma (GBM). We have identified a novel synthetic lethal interaction between ClpP activation and HDAC1/2 inhibition that converges on GBM energy metabolism.

Experimental design: Transcriptome, metabolite, and U-13C-glucose tracing analyses were utilized in patient-derived xenograft (PDX) models of GBM. Orthotopic GBM models were used for in vivo studies.

Results: We showed that activation of the mitochondrial ClpP protease by mutant ClpP (Y118A) or through utilization of second-generation imipridone compounds (ONC206 and ONC212) in combination with genetic interference of HDAC1 and HDAC2 as well as with global (panobinostat) or selective (romidepsin) HDAC inhibitors caused synergistic reduction of viability in GBM model systems, which was mediated by interference with tricarboxylic acid cycle activity and GBM cell respiration. This effect was partially mediated by activation of apoptosis along with activation of caspases regulated chiefly by Bcl-xL and Mcl-1. Knockdown of the ClpP protease or ectopic expression of a ClpP D190A mutant substantially rescued from the inhibition of oxidative energy metabolism as well as from the reduction of cellular viability by ClpP activators and the combination treatment, respectively. Finally, utilizing GBM PDX models, we demonstrated that the combination treatment of HDAC inhibitors and imipridones prolonged host survival more potently than single treatments or vehicle in vivo.

Conclusions: Collectively, these observations suggest that the efficacy of HDAC inhibitors might be significantly enhanced through ClpP activators in model systems of human GBM.

Trial registration: ClinicalTrials.gov NCT04541082.

©2022 The Authors; Published by the American Association for Cancer Research.

Figures

Figure 1.

Activation of the ClpP protease along with inhibition of HDAC1/2 causes synthetic lethality in GBM model systems. A and B, GBM14 cells were treated with 2.5 nmol/L romidepsin for 24 hours and were submitted for microarray analysis followed by GSEA. Shown in A is a volcano plot (FDR-q vs. NES). Highlighted is the NES of Reactome_Pyruvate_Metabolism and Go_Lipid_Catabolic_Process gene sets derived from GSEA. Shown in B is the gene set enrichment analysis. NES, normalized enrichment score, FDR-q-value (n = 2). C and D, GBM14 cells were treated with 10 µmol/L ONC206 and 2.5 nmol/L romidepsin for 24 hours and were submitted for microarray analysis followed by GSEA. Shown in C is a volcano plot (FDR-q vs. NES). Highlighted is the NES of Go_Mitochondrial_Matrix and Go_Electron_ Transport_Chain gene sets derived from GSEA (interaction analysis of the combination treatment). Shown in D is the gene set enrichment analysis (n = 2). E, NCH644 and GBM14 cells were treated with ONC201, ONC206, or ONC212 (1 µmol/L and 10 µmol/L) in the presence or absence of panobinostat (50 nmol/L and 200 nmol/L) for 72 hours and cellular viability analysis was performed (n = 5). F, NCH644 and GBM14 cells were treated with ONC201, ONC206, or ONC212 (1 µmol/L and 10 µmol/L) in the presence or absence of romidepsin (1 nmol/L and 2.5 nmol/L) for 72 hours and cellular viability analysis was performed (n = 5). *, The statistical analysis was performed between imipridones and combination treatment. #, The statistical analysis was performed between Pb/Ro and combination treatment. **, P < 0.01; ***/****, P < 0.001; ##, P < 0.01; ####, P < 0.001. G, Cellular viability of NCH644 cells transduced with shRNA against HDAC1 or HDAC2 in the presence or absence of ONC201, ONC206, or ONC212 (n = 4). *, The statistical analysis was performed between shNT and shHDAC1/2. +, The statistical analysis was performed between imipridones and DMSO treatment. H, Protein capillary electrophoresis of NCH644 cells transduced with shRNA against HDAC1 or HDAC2. Vinculin is used as a loading control. I, Stable ClpP-wild-type or ClpP-Y118A NCH644 cells were treated with panobinostat/romidepsin for 72 hours and cellular viability analysis was performed (n = 4). Statistical significance was assessed ANOVA with Dunnett multiple comparison test in E–G and by two-tailed Student t test in I.

Figure 2.

Activation of the ClpP protease along with inhibition of HDAC1/2 activates a cell death with apoptotic features that is partially caspase dependent. A and B, NCH644 and GBM14 cells were treated with 10 µmol/L ONC206 and 0.2 µmol/L panobinostat and were labeled with Annexin/PI dye for flow cytometry analysis. The quantification of apoptotic cells is shown in B (n = 3). C and D, NCH644 and GBM14 cells were treated with 10 µmol/L ONC206 and 2.5 nmol/L romidepsin and were labeled with Annexin/PI dye for flow cytometry analysis. The quantification of apoptotic cells is shown in D (n = 3). E, NCH644 and GBM14 cells were treated with ONC206 and panobinostat/romidepsin in the presence or absence of 10 µmol/L z-VAD-FMK for 48 hours and were labeled with Annexin/PI dye for flow cytometry analysis. Shown is the quantification of apoptotic cells (n = 3). F and G, NCH644 and GBM14 cells were treated with 10 µmol/L ONC206 and 0.2 µmol/L panobinostat/2.5 nmol/L romidepsin for 48 hours and cell lysates were subjected for standard Western blotting with the indicated antibodies. FL, full length; CF, cleavage fragment. The quantification is shown in G (n = 3). Statistical significance was assessed ANOVA with Dunnett multiple comparison test. *, The statistic in G was compared between imipridones and combination treatment. #, The statistic in G was compared between Pb/Ro and combination treatment. *, P < 0.05; **, P < 0.01; ***, P < 0.001; #, P < 0.05; ##, P < 0.01; ###, P < 0.001.

Figure 3.

Activation of the ClpP protease along with inhibition of HDAC1/2 regulates the expression of antiapoptotic Bcl-2 family members. A and B, Standard Western blots of the antiapoptotic Bcl-2 family members (Mcl-1, Bcl-xL) in U251, GBM14, GBM12, and NCH644 cells treated with 10 µmol/L ONC206/ONC12 and 0.2 µmol/L panobinostat/2.5 nmol/L romidepsin for 24 hours. C, NCH644 and GBM14 cells were transduced with empty vector (EV) or Mcl1 adenovirus for 24 hours, treated with 10 µmol/L ONC206 and 0.2 µmol/L panobinostat/2.5 nmol/L romidepsin for 48 hours, and were labeled with Annexin/PI dye for flow cytometry analysis. Shown is the quantification of apoptotic cells (n = 3). D, NCH644 and GBM14 cells were transduced with empty vector (EV) or Bcl-xL adenovirus for 24 hours, treated with 10 µmol/L ONC206 and 0.2 µmol/L panobinostat/2.5 nmol/L romidepsin for 48 hours, and were labeled with Annexin/PI dye for flow cytometry analysis. Shown is the quantification of apoptotic cells (n = 3). Statistical significance was assessed by two-tailed Student t test. ****, P < 0.001.

Figure 4.

Activation of the ClpP protease along with inhibition of HDAC1/2 affects GBM energy metabolism. A, Stable ClpP-wild type or ClpP-Y118A U251 cells were treated with 2.5 nmol/L romidepsin for 72 hours and analyzed for OCR by a mito stress assay on a Seahorse XFe24 device. F, FCCP; OM, oligomycin; R/A, rotenone/antimycin. The graph (right) shows the OCR and coupled respiration level (n = 4–5). B, The graphic presents one turn of the TCA cycle. Glucose is metabolized to pyruvic acid (m+3; three carbons labeled). When glucose is oxidized in the TCA cycle, citric acid (m+2) is produced (two carbons are labeled). When glucose is used for anaplerosis, citric acid (m+3) is produced (three carbons are labeled). C and D, GBM14 cells were incubated in DMEM (devoid of phenol red, pyruvate and glutamine) supplemented with 25 mmol/L U-13C-glucose, 4 mmol/L glutamine, and 1.5% dialyzed FBS in the presence of 10 µmol/L ONC206, 2.5 nmol/L romidepsin or the combination for 24 hours. Shown are the total metabolites in the TCA cycle in C and the fractions of isotopologs in D (n = 3). E, Real-time PCR analysis of GBM14, U251, and NCH644 cells treated with 10 µmol/L ONC206, 2.5 nmol/L romidepsin/0.2 µmol/L panobinostat, or the combination for 24 hours. 18S is used as an internal control. Shown are means and SD (n = 4). F, CHIP-qPCR with anti-H3K27ac antibody of two different promoter regions (PGC1α and PPARD) in GBM14 treated with ONC206, 0.2 µmol/L panobinostat/2.5 nmol/L romidepsin, or the combination for 24 hours (n = 3). Statistical significance was assessed ANOVA with Dunnett multiple comparison test. **, P < 0.01; ***/****, P < 0.001. +, The statistic in D and E was compared between Pb/Ro and combination treatment.

Figure 5.

The combination treatment of imipridones along with HDAC inhibitors is highly dependent on mitochondrial ClpP. A, GBM14 cells were transduced with three different shRNA against CLPP, were treated with 10 µmol/L ONC206 and 0.2 µmol/L panobinostat/2.5 nmol/L romidepsin for 72 hours, and cellular viability analysis was performed (n = 4). B, Standard Western blots of GBM14 transduced with three different shRNA against ClpP. Actin is used as a loading control. C, Stable shCLPP GBM14 cells were transduced with empty vector (EV) or an adenovirus carrying CLPP cDNA for 24 hours (CLPP-OE), were treated with 10 µmol/L ONC206 and 0.2 µmol/L panobinostat/2.5 nmol/L romidepsin for 72 hours, and cellular viability analysis was performed (n = 4). Statistical significance was assessed ANOVA with Dunnett multiple comparison test in A and C. *, The statistical analysis was performed between shNT and shClpP. +, The statistical analysis was performed between empty vector (EV) and ClpP-OE. D, Standard Western blots of stable shCLPP GBM14 cells were transduced with empty vector (EV) or adenovirus (CLPP-OE) for 24 hours. E, U251, GBM14, and NCH644 cells were transduced with empty vector (EV), ClpP wild-type, or ClpP D190A lentiviral particles and were treated with 10 µmol/L ONC206, 0.2 µmol/L panobinostat/2.5 nmol/L romidepsin, or the combination for 72 hours and cellular viability analysis was performed (n = 4). F, Standard Western blot of U251, GBM14, and NCH644 cells transduced with empty vector (EV), ClpP wild-type, or ClpP D190A lentiviral particles. G, Standard Western blot of GBM14 cells transduced with ClpP wild-type or ClpP D190A lentiviral particles and were treated with 10 µmol/L ONC206, 0.2 µmol/L panobinostat/2.5 nmol/L romidepsin, or the combination for 24 hours. H and I, GBM14 cells were transfected with siRNA against CLPP; treated with 10 µmol/L ONC206, 2.5 nmol/L romidepsin, or the combination of both for 24 hours; and analyzed for OCR by a mito stress assay on a Seahorse XFe24 device. F, FCCP; OM, oligomycin; R/A, rotenone/antimycin. The graph in I shows the OCR and coupled respiration level (n = 4–5). J and K, Stably transduced cells expressing CIpP wild-type or ClpP D190A treated with 10 µmol/L ONC206, 2.5 nmol/L romidepsin, or the combination of both for 24 hours were analyzed for OCR by mito stress assay on a Seahorse XFe24 device. The graph in K show the OCR and coupled respiration level (n = 5). Statistical significance was assessed by two-tailed Student t test in E, I, and K. *, P < 0.05; ***/****, P < 0.001.

Figure 6.

Combined treatment with imipridones and HDAC inhibitors elicits enhanced antiglioma activity in PDX models. A–D, GBM12 and GBM43 cells were implanted into the subcutis of immunocompromised Nu/Nu mice and were treated with vehicle, ONC206 (50 mg/kg), romidepsin (0.5 mg/kg), or combination treatment of both three times per week after tumors were established. The tumor volumes over time are shown in A and C and the tumor volumes on the last day of the experiment are shown in B and D. E–G, Tumors from the experiment in A were fixed and stained with H&E, TUNEL, or Ki-67. H–J, Survival analysis of immunocompromised Nu/Nu mice implanted with GBM12-luc 15 × 104 cells and administered with vehicle, ONC212 (50 mg/kg), panobinostat (5 mg/kg), or combination treatment of both two times per week after 8 days post intracranial. The log-rank test was used to assess statistical significance (vehicle n = 12, ONC212 n = 8, Pb n = 10, and combination n = 14). The representative IVIS image from each group after 22 days of implantation was shown in J. K–M, Survival analysis of C57BL/6NTac mice implanted with GL261-luc 10×104 cells and administered with vehicle, ONC212 (50 mg/kg), panobinostat (5 mg/kg), or combination treatment of both two times per week for panobinostat and one time per week for ONC212 after 8 days post intracranial. The log-rank test was used to assess statistical significance (vehicle n = 11, ONC212 n = 9, Pb n = 9, and combination n = 11). The representative IVIS image from each group after 17 days of implantation is shown in M. Scale bar, 50 µm. Statistical significance was assessed ANOVA with Dunnett multiple comparison test.