Autophagy activation can partially rescue proteasome dysfunction-mediated cardiac toxicity

Abstract

The ubiquitin-proteasome pathway and its functional interplay with other proteostatic and/or mitostatic modules are crucial for cell viability, especially in post-mitotic cells like cardiomyocytes, which are constantly exposed to proteotoxic, metabolic, and mechanical stress. Consistently, treatment of multiple myeloma patients with therapeutic proteasome inhibitors may induce cardiac failure; yet the effects promoted by heart-targeted proteasome dysfunction are not completely understood. We report here that heart-targeted proteasome knockdown in the fly experimental model results in increased proteome instability and defective mitostasis, leading to disrupted cardiac activity, systemic toxicity, and reduced longevity. These phenotypes were partially rescued by either heart targeted- or by dietary restriction-mediated activation of autophagy. Supportively, activation of autophagy by Rapamycin or Metformin administration in flies treated with proteasome inhibitors reduced proteome instability, partially restored mitochondrial function, mitigated cardiotoxicity, and improved flies' longevity. These findings suggest that autophagic inducers represent a novel promising intervention against proteasome inhibitor-induced cardiovascular complications.

Keywords: autophagy; cardiotoxicity; metformin; mitostasis; proteasome inhibitor; proteostasis.

FIGURE 1

Heart‐targeted (Gal4^ΤinCΔ4) KD of the proteasomal Prosβ5 gene results in proteome instability and reduced mitochondria number. (a) Relative Prosβ5 gene expression (vs. control) in heart tissues following Prosβ5 siRNA. (b, c) Relative (%) 26S proteasome activities (b) and ROS levels (c) in heart tissues of Prosβ5 ^RNAi (vs. control) flies. (d) Immunoblot analyses of proteome ubiquitination (Ub) and carbonylation (DNP) in flies' heart tissues after Prosβ5 KD. (e) CLSM viewing of Prosβ5 ^RNAi (vs. control) flies heart tubes stained with LysoTracker (e1), LysoTracker quantitation (e2), and immunoblotting analysis using the lysosomal marker anti‐Lamp1 (e3). (f) Relative (%) cathepsins activity in heart tissues of flies with the shown genotypes. (g) CLSM visualization of mitochondria in heart tissues of the shown fly lines after blw/ATP5A immunofluorescence staining; nuclei were counterstained with DAPI. (h) Relative expression levels (vs. control) of indicated mitochondrial genes in isolated heart tissues of the shown genotypes following Prosβ5 KD. In (a, h) gene expression was plotted vs. respective controls; RpL32/rp49 gene was used as RNA input reference. Gapdh and Actin probing in (d) and (e3), respectively, were used as protein input reference. p Values were calculated with unpaired t test. Bars, ±SD (n ≥ 3); *p < 0.05; **p < 0.01

FIGURE 2

Heart‐specific (Gal4^ΤinCΔ4) Prosβ5 KD causes cardiotoxicity, developmental disorders, and acceleration of aging‐related phenotypes. (a) Heart beats (mean from 15 female adult flies) normalized to a 30 sec period (see, Videos S1 and S2). (b) Heart rhythm calculated by beats per 5 sec; each dot represents the number of heart beats/5 sec. (c) Mitochondrial ST3/ST4 respiratory ratio in somatic tissues of the indicated genotypes. (d) Number (%) of pupae and hatched flies (at 7 and 14 days, respectively) after transferring 30 embryos per assay of the shown genotypes to culture medium. (e) Representative images (e1) and weight (%) (e2) of larvae, pupae, and female adults of control and Prosβ5 ^RNAi flies. (f) Climbing activity (%) of young transgenic flies of the indicated genotypes. (g) Longevity curves of control flies or after Prosβ5 gene KD; log‐rank, Mantel‐Cox test: Control vs. Prosβ5 ^RNAi p < 0.0001. Statistics of the longevity curves are also reported in Table S1. p values in (a–f) were calculated with unpaired t test. Bars, ±SD (n ≥ 3); *p < 0.05; **p < 0.01

FIGURE 3

Heart‐targeted (Gal4^ΤinCΔ4) Atg8a ^OE in adult flies suppresses the toxic effects of proteasome KD and partially rescues mitochondria and cardiac functionality. (a–c) Relative (%) 26S proteasome activities (a), ROS levels (b), and cathepsins activity (c) in the heart tissues of the indicated transgenic lines. (d) CLSM visualization of shown fly lines heart tubes following LysoTracker staining (d1) and LysoTracker quantitation (d2). (e) CLSM visualization of mitochondria following blw/ATP5A immunofluorescence staining of transgenic flies' heart tissues. (f) Relative expression levels of mitochondrial genes in heart tissues of the indicated transgenic flies. (g) Mitochondrial ST3/ST4 respiratory efficiency rates from somatic tissues of flies expressing the shown transgenes, specifically in heart. (h) Heart beats (mean from 15 female adult flies) normalized to a 30 sec period (see, Videos S1 and S3). (i) Heart rhythm calculated by beats per 5 sec; each dot represents the number of heart beats/5 sec. (j) Climbing activity (%) of young transgenic flies vs. control. (k) Longevity curves of the indicated transgenic lines; log‐rank, Mantel‐Cox test: Prosβ5 ^RNAi vs. Prosβ5 ^RNAi, GFP‐Atg8a ^ΟΕ p < 0.001. Statistics of the longevity curves are also reported in Table S1. Gene expression in (f) was plotted vs. respective controls; RpL32/rp49 gene expression was used as RNA input reference. p Values were calculated with one‐way ANOVA with Kruskal–Wallis test in (a–d, f–h, j) and with unpaired t test in (i). Bars, ±SD (n ≥ 3); *p < 0.05; **p < 0.01

FIGURE 4

MET treatment promotes autophagy and reduces proteotoxic stress caused by PIs administration. (a) Immunoblot analyses of dissected flies' hearts after treatment with MET; samples were probed with antibodies against foxo, ref(2)P/p62, p‐Ampka, and Atg8a/GABARAP. (b) CLSM visualization of flies' heart tubes following LysoTracker and immunofluorescence Atg8a/GABARAP staining (b1) along with quantitation of lysosomes number (b2) and immunoblotting analysis using the lysosomal marker anti‐Lamp1 (b3). (c) Relative (%) cathepsins activity in heart tissues of control flies and flies exposed to MET and/or CFZ, BTZ. (d) Relative (%) 26S proteasome activities in heart tissues of flies after treatment with the indicated drugs. (e) Immunoblot analyses of total protein ubiquitination (Ub) (e1) or carbonylation (DNP) (e2) in heart tissues of flies treated with the shown drugs. (f) Relative (%) ROS levels in heart tissues following treatment of flies with the indicated drugs. Concentrations of used drugs were as follows: MET (1 mM), CFZ (50 μM), and BTZ (1 μM). Flies were treated with the indicated drugs for 14 days. Asterisk (*) in (a) indicates the lipidated Atg8a form. Actin probing in (a), ponceau S staining in (b3), and Gapdh probing in (e) were used as reference for protein input. p Values were calculated with one‐way ANOVA with Kruskal–Wallis test. Bars, ±SD (n ≥ 3); *p < 0.05; **p < 0.01

FIGURE 5

MET induces mitochondrial biogenesis and restores mitochondria and cardiac functionality when co‐administrated with PIs. (a) CLSM visualization of heart tissues mitochondria following blw/ATP5A immunofluorescence staining. (b) Relative expression levels of shown mitochondrial genes following treatment (or not) of flies with the indicated drugs. (c) Mitochondrial ST3/ST4 respiratory efficiency rates (vs. control) from somatic tissues of flies exposed to shown drugs for 14 days. (d) Heart beats (mean from 20 female adult flies) normalized to a 30 sec period (see, Videos S5 and S6). (e) Heart rhythm calculated by beats per 5 sec; each dot represents the number of heart beats/5 sec. (f) Locomotion (climbing) activity (%) and (g) longevity curves of flies exposed (or not) to the shown drugs; log‐rank, Mantel‐Cox test: Control vs. MET p = 0.5, control vs. CFZ p < 0.0001, CFZ vs. ΜET + CFZ p = 0.03, control vs. BTZ p < 0.0001, BTZ vs. ΜET + BTZ p < 0.0001. Statistics of the longevity curves are also reported in Table S1. Concentrations of used drugs were as follows: (MET 1 mM), CFZ (50 μM), and BTZ (1 μM). Gene expression in (b) was plotted vs. respective controls. RpL32/rp49 gene expression in b was used as input reference. p Values were calculated in (b–d, f) with one‐way ANOVA with Kruskal–Wallis test and in (e) with unpaired t test. Bars, ±SD (n ≥ 3); *p < 0.05; **p < 0.01