Pharmacological inhibition of BAG3-HSP70 with the proposed cancer therapeutic JG-98 is toxic for cardiomyocytes

Abstract

The co-chaperone Bcl2-associated athanogene-3 (BAG3) maintains cellular protein quality control through the regulation of heat shock protein 70 (HSP70). Cancer cells manipulate BAG3-HSP70-regulated pathways for tumor initiation and proliferation, which has led to the development of promising small molecule therapies, such as JG-98, which inhibit the BAG3-HSP70 interaction and mitigate tumor growth. However, it is not known how these broad therapies impact cardiomyocytes, where the BAG3-HSP70 complex is a key regulator of protein turnover and contractility. Here, we show that JG-98 exposure is toxic in neonatal rat ventricular myocytes (NRVMs). Using immunofluorescence microscopy to assess cell death, we found that apoptosis increased in NRVMs treated with JG-98 doses as low as 10 nM. JG-98 treatment also reduced autophagy flux and altered expression of BAG3 and several binding partners involved in BAG3-dependent autophagy, including SYNPO2 and HSPB8. We next assessed protein half-life with disruption of the BAG3-HSP70 complex by treating with JG-98 in the presence of cycloheximide and found BAG3, HSPB5, and HSPB8 half-lives were reduced, indicating that complex formation with HSP70 is important for their stability. Next, we assessed sarcomere structure using super-resolution microscopy and found that disrupting the interaction with HSP70 leads to sarcomere structural disintegration. To determine whether the effects of JG-98 could be mitigated by pharmacological autophagy induction, we cotreated NRVMs with rapamycin, which partially reduced the extent of apoptosis and sarcomere disarray. Finally, we investigated whether the effects of JG-98 extended to skeletal myocytes using C2C12 myotubes and found again increased apoptosis and reduced autophagic flux. Together, our data suggest that nonspecific targeting of the BAG3-HSP70 complex to treat cancer may be detrimental for cardiac and skeletal myocytes.

Keywords: BAG3; HSP70; JG-98; cancer therapy; cardio-oncology; cardiomyocyte; cytotoxicity.

Figure 1.. Disrupting the BAG3-HSP70 complex with JG-98 causes apoptosis in cardiomyocytes.

A, Representative immunofluorescence images for NRVMs treated with DMSO and 1 nM to 10 μM JG-98; green – extracellular membrane phosphatidylserine (apoptotic), blue – healthy; scale bars = 100 μm. B, Quantification of apoptosis in the immunofluorescence images; n = 18 images per treatment from 3 separate experiments; data were analyzed by one-way ANOVA with Tukey’s post-hoc test. C, Representative western blot for HSP70 and BAG3 in HSP70 immunoprecipitation experiments from cells treated with DMSO or 1 μM JG-98. D, Quantitative densitometry analysis of the BAG3 signal intensity normalized to HSP70; n = 9 per group from 3 separate experiments; data were analyzed by two-tailed t-test. All data are presented as the mean ± SEM.

Figure 2.. JG-98 treatment reduces autophagy flux.

A, Representative western blot for LC3 in NRVMs treated with DMSO or JG-98 in the presence or absence of the autophagy inhibitor NH₄Cl. B, Quantification of LC3-II expression normalized to total protein; n = 9 per group from 3 separate experiments; data were analyzed with two-way ANOVA (interaction, p = 0.023) with Tukey’s post-hoc test. Data are presented as the mean ± SEM.

Figure 3.. Inhibition of BAG3-HSP70 alters expression of proteins involved in BAG3-dependent autophagy pathway.

A-B, Representative western blots for BAG3, HSPB8, SYNPO2, HSP70, HSPB5 in NRVMs treated with DMSO or JG-98. C-G, Quantification of BAG3 (C), HSP70 (D), HSPB8 (E), HSPB5 (F), and SYNPO2 (G) protein expression normalized to total protein loading control. H-L, qPCR analysis of bag3 (H), hspa1a (I), hspb8 (J), hspb5 (K), and synpo2 (L) mRNA expression normalized to β-actin. All treatments were with 1 μM JG-98 for 18 hours. For all, n = 9 per group from 3 separate experiments; data were analyzed by two-tailed t-test and are presented as the mean ± SEM.

Figure 4.. The BAG3-HSP70 interaction is required for maintaining BAG3, HSPB8, and HSPB5 protein stability.

A-B, Representative western blots for BAG3, HSPB8, SYNPO2, HSP70, HSPB5 in NRVMs treated with DMSO or 1 μM JG-98 in the presence of cycloheximide (CHX) for 0–24 hours. C-E, Quantification of BAG3 (C), HSPB8 (D), and HSPB5 (E) protein expression normalized to total protein over the time-course of CHX treatment with DMSO or JG-98; n = 3 for each time point from 3 separate experiments; data were plotted using one-phase decay equation; data were analyzed by two-tailed t-test; *p < 0.05, **p < 0.01, ***p < 0.001. F, Presentation of protein half-life for BAG3, HSPB8, and HSPB5 obtained from the preceding graphs; each point for a given protein represents a separate experiment; data were analyzed by two-tailed t-test. All data are presented as the mean ± SEM.

Figure 5.. BAG3-HSP70 is required for maintaining sarcomere structure.

A, Representative immunofluorescence images of NRVMs treated with DMSO or 1 μM JG-98 for 18 hours and immunostained for BAG3 and α-actinin. B, Quantification of sarcomere disarray in the DMSO and JG-98 groups; n = 36 cells per group from 3 separate experiments. C, Representative immunofluorescence images of NRVMs treated with DMSO or 1 μM JG-98 for 1 hour, followed by heat shock (HS) at 42 °C for two hours, then fixed and immunostained for BAG3 and α-actinin. D, Quantification of sarcomere disarray in the DMSO/HS and JG-98/HS groups; n = 24 cells per group from 3 separate experiments. Images were acquired at 63X magnification; scale bars = 10 μm. Data were analyzed by two-tailed t-test.

Figure 6.. The negative effects of JG-98 on apoptosis and sarcomere structure are partially ameliorated with rapamycin co-treatment.

A, Representative immunofluorescence images of NRVMs treated with DMSO or 1 μM JG-98 with or without 30 nM rapamycin; green – extracellular membrane phosphatidylserine (apoptotic), blue – healthy; scale bars = 100 μm. B, Quantification of apoptosis in the four groups; n = 23 images per group from 3 separate experiments; data are presented as the mean ± SEM and were analyzed by two-way ANOVA (interaction = 0.0039) with Tukey’s post-hoc test. C, Representative immunofluorescence images of NRVMs treated with DMSO or 1 μM JG-98 with or without 30 nM rapamycin for 18 hours and immunostained for BAG3 and α-actinin; images were acquired at 63X magnification; scale bars = 10 μm. D, Quantification of sarcomere structural disarray in the four groups; n = 32 cells per group from 3 separate experiments; data were analyzed by two-way ANOVA (interaction = 0.0075) with Tukey’s post-hoc test.

Figure 7.. The effects of JG-98 on C2C12 skeletal myotubes.

A, Quantification of apoptosis identified by immunofluorescence imaging in C2C12 myotubes treated with DMSO or JG-98 for 18 hours; one-way ANOVA, Tukey post-hoc. B, Representative western blot for LC3 in C2C12s treated with DMSO or JG-98 in the presence or absence of NH₄Cl. C, Quantification of LC3-II expression normalized to total protein; n = 9 per group from 3 separate experiments; data were analyzed by two-way ANOVA (interaction = 0.0104), Tukey post-hoc. D, Representative western blots for BAG3, HSPB8, SYNPO2, HSP70, and HSPB5 in C2C12s treated with DMSO or JG-98 for 18 hours. E-G, Quantification of SYNPO2 (E), HSP70 (F), and BAG3 (G) protein expression normalized to total protein; n = 9 per group from 3 separate experiments; two-tailed t-test. H-J, qPCR analysis of synpo2 (H), hspa1a (I), and bag3 (J) expression normalized to β-actin; n = 9 per group from 3 separate experiments; two-tailed t-test. K, Representative western blots for BAG3, HSPB8, SYNPO2, HSP70, and HSPB5 in C2C12s treated with DMSO or 1 μM JG-98 in the presence of cycloheximide (CHX). L-N, Quantification of BAG3 (L), HSPB8 (M), and HSPB5 (N) expression normalized to total protein over the time-course of CHX treatment; n = 3 for each time point from 3 separate experiments; data were plotted using one-phase decay equation. Data are presented as the mean ± SEM.