Fig 1.
High-throughput drug screen reveals HDAC and proteasome inhibitors as potent drug classes against synovial sarcoma.
(A) Compounds resulting in measured relative cell viability of less than 50% are annotated as hits (blue). Y-axis denotes drug target classes arranged in alphabetical order. (B) The top 40 drug screen hits out of the 900 compound screen represented by drug target class, demonstrate HDAC inhibitors as the most abundant hits in the screen. (C) Compounds that brought about greater than 90% decreased relative cell viability were scored as 1 (+++), 75.1–90% as 0.5 (++), 50–75% as 0.25 (+) and less than 50% as 0 (-). Total score across the six cell lines was calculated out of 6. (D) IC50 measurements were calculated for drug screen hits quisinostat (HDAC inhibitor), BGT-226 (PI3K/mTOR inhibitor), bortezomib (proteasome inhibitor) as compared with the current standard for synovial sarcoma treatment doxorubicin (cytotoxic DNA/RNA intercalating agent and topoisomerase inhibitor), in a panel of six human SS18-SSX positive cell lines and two control cell lines (HEK293T, MCF7). Error bars signify standard error of mean from conditions performed in triplicate.
Fig 2.
Quisinostat-mediated HDAC inhibition results in a dissociation of the driving complex in synovial sarcoma.
(A, B) Proximity ligation assay of SS18-SSX/TLE1 nuclear signal demonstrates a significant decrease in detectable protein co-localization following HDAC inhibition in SYO-1 synovial sarcoma cells. (C) Quisinostat treatment at 0.025 μM reactivates targets of SS18-SSX-mediated gene repression, EGR1 and CDKN2A, in six human synovial sarcoma cell lines. (D) Expression of EGR1, p16INKa and p14ARF (CDKN2A) protein levels increase with increasing concentrations of quisinostat, concomitant with a decrease in SS18-SSX protein levels. GAPDH was used as a loading control. Scale bars in panel A represent 20 μm. Statistical significance compared to vehicle treatment controls was determined by Student t test: * denotes p < 0.05. Error bars represent standard error of mean from conditions performed in triplicate.
Fig 3.
HDAC inhibition by quisinostat synergizes with proteasome inhibition to decrease synovial sarcoma cell viability.
(A) In all synovial sarcoma cell lines, but not HEK293T controls, the addition of 0.005 μM of bortezomib results in a downshift of approximately a full log of quisinostat, decreasing the amount of drug required to achieve the same effect as the HDAC inhibitor alone. (B) Isobologram analysis demonstrates synergy of these drug classes in synovial sarcoma cell lines (but not HEK293T controls), as increasing concentration combinations fall below the additive isoboles. (C) Combination index (CI) values calculated for the combination of bortezomib and quisinostat in synovial sarcoma are significantly less than 1, indicating synergy of the compounds is occurring in all six synovial sarcoma cell lines (but not HEK293T controls). Isobolograms and CI values were calculated using the Chou-Talalay-designed program CompuSyn. Statistical significance compared to vehicle treatment controls was determined by Student t test: * denotes p < 0.05; ** denotes p < 0.01; *** denotes p < 0.001. Error bars represent standard error of mean from conditions performed in triplicate.
Fig 4.
The synergistic effect of HDAC and proteasome inhibition is consistent within each drug class.
Additional compounds of each drug class were tested in combinational synergy studies. Quisinostat, panobinostat (pan-HDAC inhibitors), and bortezomib, carfilzomib and ixazomib (proteasome inhibitors) were studied in all combinations in the SYO-1 (A) and MoJo (B) SS18-SSX containing cell lines. CI values are less than 1 in these synovial sarcoma cell lines. CI values were calculated using the Chou-Talalay-designed program CompuSyn. Error bars represent standard error of mean from conditions performed in triplicate.
Fig 5.
HDAC inhibition prevents aggresome formation in response to proteasome inhibitors, and combination treatment leads to endoplasmic reticulum stress.
(A) Knockdown of HDAC6 results in decreased levels of LC3B in the SYO-1 synovial sarcoma cell line. (B) Proteasome inhibition as well as treatment with class I HDAC inhibitor romidepsin increases in LC3B levels, whereas quisinostat treatment decreases protein levels following treatment with proteasome inhibitors bortezomib or carfilzomib. (C) Aggresome formation is induced by bortezomib at 4 hours post treatment by PROTEOSTAT® staining analyzed by flow cytometry, an effect that is abrogated by quisinostat. The aggresome propensity factor (APF) is significantly decreased with the addition of quisinostat in the context of proteasome inhibition. (D) Endoplasmic reticulum stress markers are expressed following combination treatment and (E) a significant increase in ROS activity is measured by mean fluorescence intensity (MFI) of DCFDA, as compared with vehicle treated cells. ROS activation is abrogated by N-acetylcysteine (N-AC) treatment at 10 mM. (F) Cell death is rescued with N-AC by ~50%. Statistical significance compared to vehicle treatment controls was determined by Student t test: * denotes p < 0.05. Error bars represent standard error of mean from conditions performed in triplicate. Vinculin or α-tubulin was used as a loading control for protein analysis.
Fig 6.
HDAC and proteasome inhibition leads to apoptosis via pro-apoptosis protein activation, ROS production and caspase activation.
(A) Pro-apoptotic proteins BIM and BIK are upregulated by both quisinostat and bortezomib, and the drug combination elicits phosphorylation of anti-apoptotic protein BCL-2 in SYO-1 cells. (B) Cleavage of caspase 3 occurs following treatment with the drug combination in synovial sarcoma cell lines, demonstrated by staining with IncuCyte™ Kinetic Caspase-3/7 Apoptosis Assay Reagent, (C) inducing significant apoptosis as confirmed by Annexin-V/PI staining in the SYO-1 cell line (Q3: live, Q2: necrotic/late apoptotic, Q4: early apoptotic). (D) The low-dose quisinostat/bortezomib drug combination brings about a significant decrease in the viability of primary synovial sarcoma cells (83-SS) as compared to matched normal muscle cells derived from the same patient (83-muscle). Two-way ANOVA indicated a significant interaction between cell type and response to the drug combination (p < 0.05). (E) Tumor growth in a murine model of synovial sarcoma was significantly reduced by day 21 with the quisinostat/bortezomib combination treatment, as compared to the vehicle only control. (F) Taken together, the combination of HDAC and proteasome inhibitors results in dissociation of the SS18-SSX driving complex as well as aggresome inhibition, ER stress and ROS production, leading to apoptosis induction in synovial sarcoma. Statistical significance compared to vehicle treatment controls was determined by Student t test or two-way ANOVA where indicated: * denotes p < 0.05. Error bars represent standard error of mean from conditions performed in triplicate. Vinculin was used as a loading control for protein analysis.