TRAP1 inhibits MARCH5-mediated MIC60 degradation to alleviate mitochondrial dysfunction and apoptosis of cardiomyocytes under diabetic conditions

Abstract

Mitochondrial dysfunction and cell death play important roles in diabetic cardiomyopathy, but the underlying mechanisms remain unclear. Here, we report that mitochondrial dysfunction and cell apoptosis are prominent features of primary cardiomyocytes after exposure to high glucose/palmitate conditions. The protein level of MIC60, a core component of mitochondrial cristae, is decreased via ubiquitination and degradation under these conditions. Exogenous expression of MIC60 alleviates cristae disruption, mitochondrial dysfunction and apoptosis. Moreover, we identified MARCH5 as an E3 ubiquitin ligase that specifically targets MIC60 in this process. Indeed, MARCH5 mediates K48-linked ubiquitination of MIC60 at Lys285 to promote its degradation. Mutation of the ubiquitination site in MIC60 or the MIC60-interacting motifs in MARCH5 abrogates MARCH5-mediated MIC60 ubiquitination and degradation. Silencing MARCH5 significantly alleviates high glucose/palmitate-induced mitochondrial dysfunction and apoptosis in primary cardiomyocytes. In addition to E3 ubiquitin ligases, molecular chaperones also play important roles in protein stability. We previously reported that the mitochondrial chaperone TRAP1 inhibits the ubiquitination of MIC60, but the detailed mechanism is unknown. Here, we find that TRAP1 performs this function by competing with MARCH5 for binding to MIC60. Our findings provide new insights into the mechanism underlying mitochondrial dysfunction in cardiomyocytes in diabetic cardiomyopathy.

MARCH5 promotes ubiquitination of MIC60 to induce MIC60 degradation, mitochondrial dysfunction and apoptosis in cardiomyocytes under diabetic conditions. TRAP1 inhibits MARCH5-mediated ubiquitination by competitively interacting with MIC60.

Introduction

Diabetes affects nearly 10% of the global population [1]. A major complication of diabetes is diabetic cardiomyopathy (DCM), which is highly associated with the occurrence of heart failure in diabetic patients [2]. Elucidating the mechanisms underlying DCM is a high priority for identifying therapeutic targets. Several processes have been hypothesized to be involved in DCM and include inflammation [3], impaired autophagy [4], types of cell death [5], oxidative stress [6] and glucolipid toxicity [7]. One salient feature of DCM is systolic and diastolic dysfunction, which is highly dependent on the mitochondrial ATP supply [8]. Diabetic conditions lead to mitochondrial dysfunction in cardiomyocytes, possibly inducing various types of cell death [9]. Mitochondrial dysfunction and cell death play important roles in DCM [10, 11], but the related molecular mechanisms remain unclear.

Mitochondrial cristae are deep invaginations in the inner mitochondrial membrane and are the main sites of mitochondrial ATP production [12]. Mitochondrial cristae disruption leads to mitochondrial dysfunction and cell death in ischaemia-induced cardiac injury [13]. Therefore, the role of mitochondrial cristae in DCM is worth exploring.

The mitochondrial contact site and cristae organizing system (MICOS) complex plays a crucial role in the formation of mitochondrial cristae [14]. To date, five subunits of the MICOS have been identified: MIC60, MIC27, MIC19, MIC13 and MIC10 [15]. Previous studies reported that MIC60 expression is decreased in cardiac tissues of diabetic mice and that transgenic overexpression of MIC60 attenuates cardiac and mitochondrial dysfunction [16]. However, how and why MIC60 expression is decreased remain unclear.

MARCH5 has historically been considered a transmembrane protein that is localized in mitochondria and contains a cytoplasmic RING finger domain [17]. MARCH5 regulates various physiological and pathophysiological processes in mitochondria, including mitochondrial dynamics [18], mitophagy [19] and mitochondrial import [20]. Furthermore, MARCH5 regulates numerous processes related to cell fate and function, including cell apoptosis [21], embryonic stem cell stemness [22] and cell senescence [23], under various conditions. Whether MARCH5 contributes to mitochondrial dysfunction under diabetic conditions remains unclear.

TRAP1 is a member of the HSP90 family and is localized mainly in mitochondria, where it plays a key role in mitochondrial homeostasis [24]. TRAP1, a molecular chaperone, maintains mitochondrial function and exerts its effects through protein binding and posttranslational modifications [25–27]. Our previous study showed that TRAP1 protected against mitochondrial dysfunction in renal tubular epithelial cells under conditions of diabetic nephropathy, suggesting that TRAP1 might play broader roles in organ protection in the context of diabetes [28]. In a recent study, proteins that potentially interact with TRAP1 were comprehensively identified, providing new perspectives for understanding how TRAP1 functions [29].

In our previous study, we found that TRAP1 attenuated the ubiquitin-dependent degradation of MIC60 to alleviate mitochondrial injuries [30]. Here, we report that MARCH5 promotes K48-linked ubiquitination of MIC60 at lysine 285 (Lys285, K285) to induce MIC60 degradation, a key process in cristae disruption, mitochondrial dysfunction and apoptosis in cardiomyocytes under diabetic conditions. TRAP1 inhibits MARCH5-mediated ubiquitination by competitively interacting with MIC60.

Results

Mitochondrial cristae disruption and apoptosis are prominent characteristics of diabetic conditions in vitro

Previous studies have reported that mitochondrial dysfunction participates in DCM. To fully evaluate mitochondrial function under diabetic conditions, we isolated primary neonatal mouse cardiomyocytes (NMCMs) (Fig. S1A) and incubated them in high glucose/palmitate (HG/P) medium to mimic diabetic conditions. The viability of NMCMs rapidly decreased with increasing glucose and palmitate concentrations (Fig. S1B). We finally chose 250 µM palmitate and 30 mM glucose as the appropriate concentrations for supplementation of the culture medium in the HG/P group based on the results of the cell viability assay and previous reports. Cells in the HG/P group showed severe disruption of the mitochondrial structure (Fig. 1A), including a decreased number of cristae and mitochondrial matrix vacuolization. Along with mitochondrial structural disruption, cells in the HG/P group exhibited markers of mitochondrial dysfunction, including decreased mitochondrial ATP levels (Fig. 1B), increased mitochondrial ROS levels (Fig. 1C), and decreased mitochondrial membrane potential (MMP) (Fig. 1C). We also observed an increase in cellular ROS levels (Fig. 1D).

Fig. 1. Mitochondrial cristae disruption and apoptosis of NMCMs are prominent characteristics under diabetic conditions in vitro.

A Transmission electron microscopy was used to assess the ultrastructure of mitochondria in NMCMs in different groups after 24 h of incubation. The average number of cristae in mitochondria in each group was quantified (n = 50). Scale bar, 500 nm. B ATP content in NMCMs in different groups after 24 h of incubation (n = 12). C Mitochondrial oxidative stress was analyzed using MitoSOX, and MMP was measured using TMRM in different groups after 24 h of incubation (n = 12). Scale bar, 100 µm. D Cellular ROS levels were measured using CELLROS reagents in different groups after 24 h of incubation. E Western blotting was used to measure the levels of proteins associated with apoptosis in NMCMs from different groups (n = 3). F Western blotting was used to measure the levels of proteins associated with ferroptosis in NMCMs from different groups (n = 3). G Western blotting was used to measure the levels of proteins associated with necroptosis in NMCMs from different groups (n = 3). H TUNEL assay of NMCMs from different groups (n = 3); the white arrows indicate TUNEL-positive NMCMs. I The viability of NMCMs in different groups was evaluated by a CCK8 assay (n = 12). The data are presented as the means ± SDs. Group comparisons were performed by one-way analysis of variance followed by Tukey’s post hoc test. **P < 0.01 vs. the NG + Veh group. ^##P < 0.01 vs. the Manni+Veh group. NG+Veh, normal glucose + vehicle. Manni+Veh, mannitose + vehicle. HG + PA, high glucose + palmitate.

Mitochondrial dysfunction might lead to various types of cell death, which are important damage mechanisms involved in DCM. Hence, we detected various types of cell death, including apoptosis, ferroptosis and necroptosis, which are associated with mitochondrial dysfunction. The levels of markers of apoptosis but not markers of ferroptosis (Fig. 1F) or necroptosis (Fig. 1G), were highly increased (Fig. 1E) in NMCMs in the HG/P group. In addition to the expression of apoptosis markers, the expression of Nrf2 was significantly increased, which might constitute a response to increased oxidative stress (Fig. 1F). The TUNEL assay also showed the presence of typical characteristics of apoptosis in NMCMs in the HG/P group (Fig. 1H). More importantly, 30 min of pretreatment with Z-VAD-FMK (an apoptosis inhibitor) but not with necrostatin-1 (a necroptosis inhibitor) or ferrostatin-1 (a ferroptosis inhibitor), significantly increased the viability of NMCMs in the HG/P group, further indicating that apoptosis was the main type of cell death occurring under this condition. Based on the above findings, HG/P treatment induced mitochondrial dysfunction and apoptosis in NMCMs.

Ubiquitination and degradation of MIC60 leads to cristae disruption, mitochondrial dysfunction and cell apoptosis under diabetic conditions in vitro

Since mitochondrial cristae disruption was a prominent characteristic of cells cultured in HG/P medium, it was crucial to determine the mechanism underlying this process. Mitochondrial cristae are maintained by the MICOS complex, which is located at cristae junctions, and other macromolecules, such Sam50 and OPA1. We first analyzed the main components of the MICOS complex, including MIC60, MIC19, MIC13 and MIC10. The protein level of MIC60 was significantly decreased in the HG/P group (Fig. 2A). Although the mRNA expression of MIC60 was also decreased, the decrease was not significant (Fig. S2A). To clarify that the decrease in the MIC60 protein level did not result from a general decrease in the number of mitochondria, we measured total mitochondrial protein levels using coomassie blue staining, as well as immunoblot analysis of the mitochondrial markers TOM20, COXIV and VDAC. None of the above proteins exhibited significantly decreased expression after HG/P treatment for 24 h (Fig. S2B). This observation prompted us to speculate that HG/P conditions enhanced the degradation of the MIC60 protein, ultimately decreasing the MIC60 protein level, but did not inhibit the transcription of MIC60 mRNA. The results of the protein half-life assay showed that the degradation of MIC60 was significantly increased in the HG/P group (Fig. S2C). MG132 but not chloroquine inhibited the degradation of MIC60 (Fig. 2B), suggesting that MIC60 was degraded through the ubiquitin–proteasome system (UPS). We analyzed the ubiquitination of MIC60, and the results confirmed that the ubiquitination of MIC60 was increased under HG/P culture conditions (Fig. 2C). Although we determined that the ubiquitination and degradation of MIC60 were increased, whether the degradation of MIC60 led to mitochondrial cristae disruption and mitochondrial dysfunction remained unknown. Thus, we constructed an MIC60 expression vector and transfected it into NMCMs. After confirming the transfection efficiency (Fig. S2D), we observed that exogenous expression of MIC60 significantly alleviated mitochondrial cristae disruption (Fig. 2D), increased MMP (Fig. 2E) and the ATP level (Fig. S2E), and decreased mitochondrial ROS levels (Fig. 2E). More importantly, exogenous expression of MIC60 increased the viability (Fig. 2F) and inhibited the apoptosis (Fig. 2G, H) of NMCMs. Collectively, these results indicated that ubiquitination and degradation of MIC60 were the key processes that led to mitochondrial cristae disruption, mitochondrial dysfunction and apoptosis in NMCMs under diabetic conditions in vitro.

Fig. 2. Ubiquitination and degradation of MIC60 leads to cristae disruption and mitochondrial dysfunction under diabetic conditions in vitro.

A Western blotting was used to measure the protein levels of MIC60, MIC10, MIC13 and MIC10 in NMCMs from different groups (n = 3). B Western blotting was used to measure the protein level of MIC60 in different groups (n = 3). C Western blotting was used to measure MIC60 expression and ubiquitination in whole-cell lysate (WCL) and IP samples (n = 3). D Mitochondrial oxidative stress was determined using MitoSOX, and MMP was measured using TMRM in different groups (n = 12). Scale bar, 100 µm. E Transmission electron microscopy was used to observe the ultrastructure of mitochondria in NMCMs from different groups. The average number of cristae in one mitochondrion of each group was quantified (n = 5). F Western blotting was used to measure the levels of proteins associated with apoptosis in NMCMs from different groups (n = 3). G TUNEL assay of NMCMs from different groups (n = 3); the white arrows indicate TUNEL-positive NMCMs. H The viability of NMCMs in different groups was evaluated by a CCK8 assay (n = 12). Scale bar, 500 nm. The data are presented as the means ± SDs. Group comparisons were performed by one-way analysis of variance followed by Tukey’s post hoc test. **p < 0.01. ^##p < 0.01 vs. the 8 h or 12 h group. ns nonsignificant.

MARCH5 is the E3 ubiquitin ligase that specifically interacts with and ubiquitinates MIC60

Ubiquitination of MIC60 plays a key role in mitochondrial dysfunction, but the mechanism underlying this process remains unclear. To investigate whether a specific E3 ubiquitin ligase is involved in this process, we performed a series of screens. First, we predicted the E3 ubiquitin ligases that could interact with MIC60 with the UbiBrowser database (Fig. S3A). Then, the E3 ubiquitin ligases that interact with MIC60 were coimmunoprecipitated and identified using mass spectrometry (MS; Fig. S3B). Finally, we found that ubiquitination of MIC60 occurred in mitochondria (Fig. S3C). Based on the above results (Fig. S3D), we chose four E3 ubiquitin ligases that potentially target MIC60 (NEDD4, SYVN1, Parkin and MARCH5) and constructed the corresponding expression vectors. After cotransfection of the expression vectors for these E3 ubiquitin ligases with the MIC60 expression vector into 293T cells, we observed that MARCH5 significantly promoted the ubiquitination of MIC60 (Fig. 3A). The results of co-IP assays indicated both endogenous and exogenous interactions between MARCH5 and MIC60 (Fig. 3B, C). Immunofluorescence staining showed partial colocalization of MIC60 and MARCH5 (Fig. 3D). We further performed an in vitro pulldown assay using purified GST-MARCH5 expressed in Escherichia coli and purified Myc-MIC60. GST-MARCH5 but not GST directly interacted with MIC60 (Fig. 3E). Moreover, the C-terminal domain of MARCH5 was crucial for its binding to MIC60, similar to the findings for other reported substrates [31] (Fig. 3F, G). MIC60 interacted with MARCH5 through its N-terminal domain (Fig. 3H, I). Taken together, these results indicate that MARCH5 is the specific E3 ubiquitin ligase of MIC60.

Fig. 3. MARCH5 is the E3 ubiquitin ligase that specifically interacts with and ubiquitinates MIC60.

A Western blotting was used to measure MIC60 expression and ubiquitination in WCL and IP samples from 293T cells transfected with the indicated plasmids. B Western blots of WCL and co-IP samples collected with anti-MARCH5 or anti-MIC60 magnetic beads from NMCMs. C Western blots of WCL and co-IP samples collected with anti-Myc or anti-His magnetic beads from 293T cells transfected with the indicated plasmids. D Representative confocal micrographs showing the colocalization of MARCH5 and MIC60 in NMCMs. Scale bars, 50 µm. E Western blots of GST pull-down samples performed with an anti-MIC60 or anti-GST antibody. F Schematic representation of the truncated Flag-MARCH5 protein. G Western blots of WCL and IP samples from 293T cells transfected with the indicated plasmids performed with an anti-His antibody. H Schematic representation of the truncated His-MIC60 protein. I Western blots of WCL and IP samples from 293T cells transfected with the indicated plasmids performed with an anti-Myc antibody.

MARCH5 mediates K48-linked ubiquitination and degradation of MIC60 under diabetic conditions in vitro

MARCH5 directly interacts with and promotes the ubiquitination of MIC60, but whether it promotes the degradation of MIC60 to induce cristae disruption and mitochondrial dysfunction is unclear. The results of the protein half-life assay showed that exogenous expression of MARCH5 obviously promoted the degradation of MIC60 and that this effect was reversed by MG132 treatment (Fig. 4A). The results of ubiquitination assays also showed that exogenous expression of MARCH5 promoted K48-linked but not K63- or K11-linked ubiquitination of MIC60, consistent with the role of MARCH5 in promoting MIC60 degradation (Fig. 4B). To investigate the role of MARCH5 in the ubiquitination and degradation of MIC60 under diabetic conditions in vitro, we separately transfected two small interfering RNAs (siRNAs) targeting MARCH5 into 293T cells. After confirming the knockdown efficiency, we observed that MARCH5 knockdown significantly decreased the HG/P-induced ubiquitination of MIC60 (Fig. 4C). However, exogenous expression of MARCH5 CS (C65S), a form of the protein that lacks ubiquitin ligase activity, did not affect the ubiquitination or protein level of MIC60 (Fig. 4D). Next, we measured the mRNA and protein levels of MARCH5 in NMCMs after HG/P treatment (Fig. 4E, F). Immunofluorescence staining showed increased colocalization of MIC60 and MARCH5 after HG/P treatment (Fig. 4G). The results of co-IP assays also revealed an increased interaction between MARCH5 and MIC60 after HG/P treatment (Fig. 4H). The results of ubiquitination assays showed that exogenous expression of MARCH5 but not MARCH5 CS promoted the ubiquitination of MIC60 (Fig. 4I) and that silencing MARCH5 decreased the ubiquitination of MIC60 after HG/P treatment (Fig. 4J). These results indicated that the E3 catalytic activity of MARCH5 was required for the degradation of MIC60. More importantly, the mutual interaction between these two proteins was increased after HG/P treatment, ultimately promoting the ubiquitination and degradation of MIC60.

Fig. 4. MARCH5 mediates K48-linked ubiquitination and degradation of MIC60 under diabetic conditions in vitro.

A Western blotting of WCL with anti-MIC60 and anti-His antibodies was performed to measure MIC60 expression and ubiquitination in 293T cells transfected with the indicated plasmids. B After transfection of 293T cells with the indicated plasmids for 48 h, Western blotting of WCL and IP samples with the indicated antibodies was performed. C After transfection with the indicated siRNA for 48 h, cells were treated as indicated for 24 h. Then, Western blotting of WCL and IP samples from 293T cells was performed with the indicated antibodies. D After transfection with the indicated plasmids for 48 h, Western blotting of WCL and IP samples from 293T cells was performed with the indicated antibodies. E RT‒qPCR was used to measure the mRNA level of MARCH5 in NMCMs from different groups. F Western blots of WCLs from NMCMs in different groups performed with an anti-MARCH5 antibody. G NMCMs were transfected with the indicated plasmids for 48 h and were then treated as indicated for 24 h. Representative confocal micrographs showing the colocalization of His-MARCH5 and MIC60 in NMCMs. Scale bars, 10 µm. H Western blots of WCL and IP samples from NMCMs in different groups performed with the indicated antibodies. I NMCMs were transfected with the indicated plasmids for 48 h and were then treated as indicated for 24 h. Western blots of WCL and IP samples from NMCMs in different groups performed with the indicated antibodies. J NMCMs were transfected with the indicated siRNAs for 48 h and were then treated as indicated for 24 h. Western blots of WCL and IP samples from NMCMs in different groups performed with the indicated antibodies. The data are presented as the means ± SDs. Group comparisons were performed by one-way analysis of variance followed by Tukey’s post hoc test. **p < 0.01 vs. the 0 h group. MARCH5 CS, MARCH5 C65S.

MARCH5 mediates the ubiquitination of MIC60 at Lys285

Next, we sought to investigate the site in MIC60 ubiquitinated by MARCH5. First, we co-transfected the Myc-MIC60 and His-MARCH5 expression vector or control vector. Myc-MIC60 was purified, and the ubiquitination sites were identified using MS. Ubiquitination of MIC60 Lys285 was increased after cotransfection with His-MARCH5 compared with the control vector (Fig. 5A). We analyzed the amino acid sequences of MIC60 from different species and found that Lys285 of MIC60 was conserved in different species (Fig. 5B). Then, we mutated several lysine residues to arginine (R) based on the predicted ubiquitination sites in MIC60 (Fig. 5C). The results of ubiquitination assays showed that the ubiquitination of K285R-mutant MIC60 was significantly decreased compared to that of wild-type MIC60 or other mutants (Fig. 5D). Moreover, the HG/P-induced and MARCH5-mediated degradation of K285R-mutant MIC60 was decreased (Fig. 5E, F). All these results indicated that Lys285 site was critical for MARCH5-mediated ubiquitination of MIC60.

Fig. 5. MARCH5 mediates Lys285 ubiquitination and degradation of MIC60.

A Potential ubiquitination sites in MIC60 were identified by IP followed by LC–MS/MS. B Lys285 of MIC60 was a conserved site across species. C Predicted ubiquitination sites in MIC60. D Western blots of WCL or IP samples from 293T cells transfected with the indicated plasmids. E, F 293T cells were transfected with the indicated plasmids for 48 h and treated as indicated for 24 h. Western blots of WCL samples from 293T cells performed with the indicated antibodies. CHX (20 µM) was used to inhibit protein synthesis.

MARCH5 knockdown alleviates mitochondrial cristae disruption, mitochondrial dysfunction and apoptosis under diabetic conditions in vitro

We next investigated whether MARCH5 knockdown can alleviate mitochondrial cristae disruption and mitochondrial dysfunction in cardiomyocytes under diabetic conditions in vitro. First, we confirmed the silencing efficiency of MARCH5 (Fig. 6A, B). MARCH5 knockdown increased the ATP production (Fig. 6C) and strongly alleviated mitochondrial cristae disruption (Fig. 6D). MARCH5 knockdown also decreased mitochondrial ROS production and increased the MMP (Fig. 6E). Additionally, MARCH5 knockdown inhibited the apoptosis (Figs. 6F, G) and increased the viability (Fig. 6H) of NMCMs. In conclusion, silencing MARCH5 inhibited mitochondrial cristae disruption, mitochondrial dysfunction and apoptosis in NMCMs under diabetic conditions in vitro.

Fig. 6. MARCH5 knockdown alleviates mitochondrial cristae disruption and mitochondrial dysfunction under diabetic conditions in vitro.

A Western blots of WCLs from NMCMs transfected with the indicated siRNAs for 48 h. B RT‒qPCR was used to measure the mRNA level of MARCH5 in NMCMs transfected with the indicated siRNAs for 48 h. C NMCMs were transfected with the indicated siRNAs for 48 h and were then treated as indicated for 24 h. ATP content in NMCMs in different groups after 24 h of incubation (n = 12). D NMCMs were transfected with the indicated siRNAs for 48 h and were then treated as indicated for 24 h. Transmission electron microscopy was used to evaluate the ultrastructure of mitochondria in NMCMs in different groups. The average number of cristae in one mitochondrion of each group was quantified (n = 50). Scale bar, 500 nm. E NMCMs were transfected with the indicated siRNAs for 48 h and were then treated as indicated for 24 h. Mitochondrial oxidative stress was assessed using MitoSOX, and MMP was measured using TMRM in different groups. F TUNEL assay of NMCMs from different groups (n = 3); the white arrows indicate TUNEL-positive NMCMs. G Western blotting was used to measure the levels of proteins associated with apoptosis in NMCMs from different groups (n = 3). H The viability of NMCMs in different groups was evaluated by a CCK8 assay (n = 12). The data are presented as the means ± SDs. Group comparisons were performed by one-way analysis of variance followed by Tukey’s post hoc test. **p < 0.01, *p < 0.05.

TRAP1 inhibited MARCH5-mediated K48-linked ubiquitination and degradation of MIC60 at Lys285 under diabetic conditions in vitro

In our previous study, we found that TRAP1 inhibited the ubiquitination of MIC60 under conditions of extracellular acidosis [30]. However, whether TRAP1 can inhibit the ubiquitination of MIC60 by regulating MARCH5 under diabetic conditions is unclear. Here, we observed that the expression of TRAP1 was decreased after HG/P treatment (Fig. 7A). First, we speculated that TRAP1 regulated the expression or function of MARCH5. Unexpectedly, exogenous expression of TRAP1 significantly increased the protein level of MIC60 after HG/P treatment for 24 h (Fig. 7B) but had no effect on the expression of MARCH5 (Fig. 7C). The results of co-IP assays showed that TRAP1 did not interact with MARCH5 (Fig. 7D). Interestingly, exogenous expression of TRAP1 inhibited the K48-linked ubiquitination and degradation of MIC60 after HG/P treatment (Fig. 7E, G), but this was not observed in MIC60 (K285R) (Fig. S4A, B) even though TRAP1 could also interact with it (Fig. S4C). Furthermore, co-expression of TRAP1 and MARCH5 in 293T cells significantly inhibited MARCH5-mediated ubiquitination and degradation of MIC60 (Fig. 7F, H). These results indicated that TRAP1 inhibited MARCH5-mediated K48-linked ubiquitination and degradation of MIC60 at Lys285 under diabetic conditions in vitro.

Fig. 7. TRAP1 inhibits MARCH5-mediated ubiquitination of MIC60 by competitively binding to MIC60.

A NMCMs were treated as indicated for 24 h. Western blots of WCLs from NMCMs performed with the indicated antibodies. B, C NMCMs were transfected with the indicated plasmids for 48 h and were then treated as indicated for 24 h. Western blots of WCLs from NMCMs performed with the indicated antibodies. D Western blots of WCL and IP samples from NMCMs performed with the indicated antibodies. E, F NMCMs were transfected with the indicated plasmids for 48 h and were then treated as indicated. Western blots of WCLs from NMCMs performed with the indicated antibodies. CHX (20 µM) was used to inhibit protein synthesis. G Western blots of IP samples from NMCMs performed with the indicated antibodies. H 293T cells were transfected with the indicated plasmids for 48 h. Western blots of WCL and IP samples from 293T cells performed with the indicated antibodies. I, J NMCMs were transfected with the indicated plasmids for 48 h. Western blots of WCL and IP samples from NMCMs. K Western blots of WCL and IP samples from 293T cells transfected with the indicated plasmids performed with an anti-Myc antibody.

TRAP1 competitively inhibits the binding of MARCH5 to MIC60

We showed that TRAP1 did not directly regulate MARCH5 but inhibited MARCH5-mediated ubiquitination and degradation of MIC60. Next, we investigated whether TRAP1 affects the interaction between MARCH5 and MIC60. The results of co-IP assays showed that exogenous expression of TRAP1 decreased the interaction between MARCH5 and MIC60 (Fig. 7I, J). Furthermore, TRAP1 interacted with the N-terminal domain of MIC60, indicating that MARCH5 and TRAP1 interact with the same domain of MIC60 (Fig. 7K). Therefore, we concluded that TRAP1 competitively interacted with the same domain of MIC60 to inhibit MARCH5-mediated ubiquitination of MIC60.

Exogenous expression of TRAP1 alleviated mitochondrial dysfunction and cell apoptosis under diabetic conditions in vitro

Next, we investigated whether exogenous expression of TRAP1 can alleviate mitochondrial dysfunction and cell apoptosis. Exogenous expression of TRAP1 alleviated cristae disruption (Fig. 8A), increased the ATP level (Fig. 8B) and MMP (Fig. 8C), and decreased mitochondrial ROS levels (Fig. 8C). Furthermore, exogenous expression of TRAP1 inhibited the apoptosis (Fig. 8D, E) and increased the viability (Fig. 8F) of NMCMs. Therefore, exogenous expression of TRAP1 alleviated mitochondrial dysfunction and apoptosis in NMCMs under diabetic conditions in vitro.

Fig. 8. Exogenous expression of TRAP1 alleviated mitochondrial dysfunction and cell apoptosis under diabetic conditions in vitro.

A Transmission electron microscopy was used to evaluate the ultrastructure of mitochondria in NMCMs from different groups. The average number of cristae in one mitochondrion of each group was quantified (n = 5). B NMCMs were transfected with the indicated siRNAs for 48 h and were then treated as indicated for 24 h. ATP content in NMCMs in different groups after 24 h of incubation (n = 12). C Mitochondrial oxidative stress was determined using MitoSOX, and MMP was measured using TMRM in different groups (n = 12). Scale bar, 100 µm. D Western blotting was used to measure the levels of proteins associated with apoptosis in NMCMs from different groups (n = 3). E TUNEL assay of NMCMs from different groups (n = 3); the white arrows indicate TUNEL-positive NMCMs. F The viability of NMCMs in different groups was evaluated by a CCK8 assay (n = 12). Scale bar, 500 nm. The data are presented as the means ± SDs. Group comparisons were performed by one-way analysis of variance followed by Tukey’s post hoc test. **p < 0.01. ns nonsignificant.

Discussion

Mitochondrial dysfunction and subsequent cell death play crucial roles in DCM, but the molecular mechanisms are not yet well understood. Here, we found that mitochondrial cristae disruption was a prominent characteristic of NMCMs under diabetic conditions. The morphology of mitochondrial cristae determines the assembly of the mitochondrial respiratory chain and respiratory efficiency [32]. Disruption of mitochondrial cristae has been described in various metabolic diseases and is highly associated with mitochondrial dysfunction [33]. Furthermore, mitochondrial dysfunction might lead to various types of cell death, including apoptosis, ferroptosis and necroptosis [34]. We found that apoptosis was the main type of cell death in NMCMs at 24 h of culture in HG/P medium, a finding similar to that in a recent report [5]. Furthermore, necroptosis was reported to be the main type of cell death occurring at the late stage of DCM [5]; this is an interesting observation, and further exploration is needed to determine whether and how mitochondrial dysfunction is involved in the switching of cell death types. Overall, these observations indicate that mitochondrial cristae disruption might be a key process in DCM.

Mitochondrial cristae are maintained mainly by the MICOS complex [35]. To clarify the mechanism underlying the disruption of mitochondrial cristae, we measured the expression of the main components of the MICOS complex, including MIC60, MIC25, MIC19 and MIC10. Interestingly, the differences in the mRNA levels of MICOS complex components were not statistically significant, but the protein level of MIC60 was significantly decreased.

MIC60 is a key component of the MICOS complex, and deletion of MIC60 results in disruption of the complex [36]. Here, we found that the protein level but not the mRNA level of MIC60 was significantly decreased, suggesting that posttranslational regulation of MIC60 resulted in a decrease in its protein level. Hence, we determined the half-life of MIC60 and found that its degradation was significantly increased under diabetic conditions in vitro. Furthermore, treatment with the proteasome inhibitor MG132 but not the lysosomal inhibitor chloroquine inhibited the degradation of MIC60, indicating that the degradation of MIC60 was dependent on the UPS. Indeed, in subsequent experiments, the ubiquitination of MIC60 was found to be increased. More importantly, exogenous expression of MIC60 significantly alleviated cristae disruption, mitochondrial dysfunction and apoptosis in NMCMs. These results indicated that the ubiquitination and degradation of MIC60 were key processes in cristae disruption, mitochondrial dysfunction and cell apoptosis under diabetic conditions. These findings were consistent with a previous study conducted on diabetic MIC60-transgenic rats [16], and thus, MIC60 might be a key factor in DCM [37].

We found that the degradation of MIC60 under diabetic conditions was highly dependent on the UPS. Previous studies have reported that various E3 ubiquitin ligases participate in the progression of DCM, mainly by regulating the degradation of various transcription factors, including Nrf2 [38] and PPAR [39]. Our research is the first to show that mitochondrial inner membrane proteins are degraded via the UPS to promote DCM. The homoeostasis of mitochondrial inner membrane proteins is maintained by mitochondrial proteases [40] and mitophagy [41]. Recent studies have reported that the UPS also regulates the degradation of mitochondrial inner membrane proteins [42]. These studies suggested that the UPS might have important effects on mitochondrial protein homeostasis, and we first demonstrated the degradation of MIC60 via the UPS during the progression of diabetes-associated mitochondrial dysfunction.

In subsequent experiments, we found that MARCH5 was the E3 ubiquitin ligase that specifically mediated the ubiquitination of MIC60. MARCH5 is a RING-type E3 ubiquitin ligase in mitochondria [43]. Previous studies have shown that MARCH5 regulates mitochondrial fission [44], mitochondrial import [45] and mitophagy [31]. Here, we found its novel role in regulating mitochondrial cristae formation by promoting K48-linked ubiquitination and degradation of MIC60 under diabetic conditions. MARCH5 can promote K48- or K61-linked ubiquitination under different conditions, with K48-linked ubiquitination mainly inducing protein degradation [46] and K61-linked ubiquitination altering protein localization or function [17, 47]. This observation is consistent with our finding that MARCH5-mediated K48-linked ubiquitination of MIC60 caused its degradation. More importantly, MARCH5 knockdown significantly alleviated cristae disruption, mitochondrial dysfunction and apoptosis in NMCMs, indicating that MARCH5 might be a therapeutic target for DCM.

TRAP1 belongs to the mitochondrial HSP90 protein family [48]. In our previous study, we found that TRAP1 mitigated MIC60 ubiquitination and degradation induced by extracellular acidification, but the detailed mechanism was unclear [30]. Here, we found that exogenous expression of TRAP1 inhibited the K48-linked ubiquitination and degradation of MIC60 after HG/P treatment, but this was not observed in MIC60 (K285R), even though TRAP1 could also interact with it. In addition, exogenous expression of TRAP1 also inhibited MARCH5-mediated K48-linked ubiquitination and degradation of MIC60. These results indicated that TRAP1 relied on Lys285 of MIC60 to inhibit MARCH5-mediated K48-linked ubiquitination under diabetic conditions. More interestingly, TRAP1 did not directly interact with or regulate MARCH5 but instead competed with MARCH5 for binding to the same domain of MIC60, ultimately inhibiting MARCH5-mediated ubiquitination of MIC60. Taken together, we observed that TRAP1 inhibited MARCH5-mediated K48-linked ubiquitination of MIC60 at Lys285 under diabetic conditions. Previous studies reported that TRAP1 inhibited the ubiquitination of several proteins, such as Sorcin [49] and β-catenin [26]. These results indicated that TRAP1 might regulate the ubiquitination and degradation of various proteins to perform molecular chaperone functions. This phenomenon revealed a possible explanation for the mechanism by which TRAP1 inhibits the ubiquitination of client proteins. Moreover, exogenous expression of TRAP1 alleviated cristae disruption, mitochondrial dysfunction and apoptosis in NMCMs. Thus, TRAP1 might play important roles in various pathophysiological processes by binding to crucial client proteins.

In summary, we demonstrated that TRAP1 competitively inhibited MARCH5-mediated ubiquitination and degradation of MIC60 to alleviate cristae disruption, mitochondrial dysfunction and apoptosis in cardiomyocytes under diabetic conditions. Inhibition of MIC60 ubiquitination and degradation might constitute a specific approach for treating DCM.

Materials and methods

NMCM isolation and culture

NMCMs were isolated from 1- to 2-day-old neonatal mice. In brief, the mice were disinfected with 75% ethanol, and their hearts were excised with medical forceps and quickly washed with ice-cold D-Hank’s solution (BL559A, Biosharp, China). The isolated hearts were subjected to digestion with 0.1% trypsin for 8 h at 4 °C followed by 3 rounds of digestion with 0.08% collagenase type II (17101015, Gibco, Grand Island, NY, USA) for 5 min each (room temperature, 80 rpm). The supernatants were transferred to new tubes and neutralized with 15 ml of DMEM (C11995500BT, Gibco, Grand Island, NY, USA) supplemented with 10% foetal bovine serum (FBS; 10437028, Gibco, Grand Island, NY, USA). The collected supernatants were centrifuged (1000 rpm, 5 min), and the precipitates were resuspended and filtered through a 70 µm cell strainer (352350, Falcon, Corning, NY, USA). The filtered supernatants were centrifuged again (1000 rpm, 5 min) and resuspended in DMEM supplemented with 10% FBS. The cells were cultured in cell culture plates for 1.5 h for removal of noncardiomyocytes. Finally, the purified cardiomyocytes were transferred and cultured in DMEM supplemented with 10% FBS and BrdU (19–160, Sigma‒Aldrich, St. Louis, MO, USA) in cell culture dishes in an incubator at 37 °C in 5% CO₂.

Cell treatment and reagents

Sodium palmitate was dissolved in 20% (w/v) BSA at 70 °C to obtain the palmitate solution (10 mM). The palmitate/BSA solution was diluted with DMEM (1:40) to the indicated concentration (250 µM) before use. The final concentrations of palmitate and glucose were 250 µM and 30 mM, respectively, in the HG/P medium. The control cells were treated with 0.5% BSA and 5.5 mM glucose.

Z-VAD-FMK (20 µM, HY-16658B, MedChemExpress) was used to inhibit apoptosis. Ferrostatin-1 (20 µM, HY-100579, MedChemExpress) was used to inhibit ferroptosis. Necrostatin-1 (20 µM, HY-15760, MedChemExpress) was used to inhibit necroptosis.

Cell line cultures

HEK 293T cells (GNHu17) were purchased from the Cell Bank of the Chinese Academy of Sciences and cultured in Dulbecco’s modified Eagle’s medium (DMEM; 11965175, Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% FBS (10437028, Gibco, Grand Island, NY, USA) and 1% penicillin/streptomycin (cat. no. 516106, Sigma‒Aldrich) at 37 °C in 5% CO₂.

Transmission electron microscopy (TEM)

Mitochondrial ultrastructure was examined by transmission electron microscopy (Hitachi HT7700, Tokyo, Japan) using conventional methods. NMCMs were fixed with 2.5% glutaraldehyde in 0.1 mol/L phosphate buffer (pH 7.4) and then reacted with 1% OsO4. After dehydration, thin sections were stained with uranyl acetate and lead citrate, and digital images were acquired.

ATP content measurement

ATP levels were measured using an Enhanced ATP Assay Kit (S0027, Beyotime) according to the instructions. In brief, cells in the different groups were lysed with 100 μl of lysis buffer provided with the kit and were then centrifuged at 12000 × g at 4 °C for 20 min. Then, 10 μl of the supernatant was mixed with 100 μl of ATP detection solution. The luminescence intensity of each ATP standard and sample was measured using a luminometer (Varioskan Flash, Thermo Fisher Scientific). The ATP concentrations were calculated according to the ATP standard curve and normalized to the protein concentrations in the supernatants.

Cell viability

A Cell Counting Kit-8 (CCK8; CK04; Dojindo) was used to assess cell viability according to the manufacturer’s instructions. In brief, cells were seeded in 96-well plates at a density of 10,000 cells per well in complete culture medium. Ten microlitres of testing solution was added to each well after processing, and the plates were incubated for an additional 1.5 h. The optical density was measured at 450 nm in a microplate reader (Multiskan^TM FC; Thermo Fisher Scientific, Inc.)

TUNEL assay

A TUNEL assay kit (C1086, Beyotime) was used to evaluate apoptosis according to the manufacturer’s instructions. TUNEL-positive cells were observed and counted under a confocal microscope (Leica TCS SP8).

MMP assay

Tetramethylrhodamine, methyl ester (TMRM; I34361, Thermo Fisher Scientific) was used to measure the MMP according to the manufacturer’s instructions. In brief, NMCMs were cultured with 50 nmol/l TMRM in serum-free medium at 37 °C for 30 min, washed 3 times with PBS and then observed under a confocal microscope (Leica TCS SP8).

Mitochondrial oxidative stress assay

mtSOX reagents (MT14, Dojindo) were used to analyze mitochondrial oxidative stress according to the manufacturer’s instructions. In brief, NMCMs were cultured with 20 nmol/l mitochondrial superoxide detection reagent in serum-free medium at 37 °C for 30 min, washed 3 times with PBS and then observed under a confocal microscope (Leica TCS SP8).

RNA isolation, reverse transcription PCR (RT-PCR) and quantitative real-time qPCR (RT-qPCR)

Total cellular RNA was extracted using TRIzol reagent (15596018, Invitrogen), and total RNA concentrations were measured using a NanoDrop spectrophotometer (Thermo). Then, RT-PCR was conducted using a kit (R323-01, Vazyme, China). RT-qPCR was carried out using another kit (Q711-02, Vazyme, China).

The primer pairs used were as follows:

MARCH5 (mouse): Forward TTAGAGTTTCGTCGTCGCAGTCAAG, Reverse GTCAGCTCGCTCCATAACATCCAG.

MIC60 (mouse): Forward TTAGAGTTTCGTCGTCGCAGTCAAG, Reverse TTTCGTCAATCATCGCTACCCTTCG.

MIC19 (mouse): Forward AGTCCTCTCCATCTGGCTCTAAGTC, Reverse CGCATCACTCGGTCTTTCTCTTCC.

MIC13 (mouse): Forward AAGGGAAGTGTGGCTGGAGGAG, Reverse CACCGACAAGGCTGACATGACTG.

MIC10 (mouse): Forward TCCGAGCTGGGCA, Reverse CCATGTAGAAGATACGGAGCCTGA.

Cell transfection

Plasmid transfection was conducted as follows:

The coding sequences of the indicated proteins were inserted into the pcDNA 3.1 vector. The plasmids were mixed with Lipofectamine 3000 (L3000, Invitrogen; ratio of 1:1.5) and 100 μl of Opti-MEM™ (51985034, Invitrogen), and each mixture was added to one well of a six-well plate. After 72 h, the transfection efficiency was determined by PCR and Western blotting.

siRNA transfection was conducted as follows:

The sequences of the siRNAs targeting MARCH5 are shown below. TransMessenger Transfection Reagent (301525, QIAGEN) was used for siRNA transfection according to the manufacturer’s instructions. Then, 72 h post-infection, the transfection efficiency was determined by PCR and Western blotting.

siMARCH5#1 (mouse): sense GCUUAGACUAUGGCGCAAAUATT, antisense UAUUUGCGCCAUAGUCUAAGCTT.

siMARCH5#2 (mouse): sense GUGACAGUGAUGCAGGUUGUATT, antisense UACAACCUGCAUCACUGUCACTT.

siMARCH5#3 (mouse): sense CAGAGCAAGAAGAAGCAUAAATT, antisense UUUAUGCUUCUUCUUGCUCUGTT.

siMARCH5#1 (human): sense GCUUAGACUGUGGCGCAAAUATT, antisense UAUUUGCGCCACAGUCUAAGCTT.

siMARCH5#2 (human): sense GAAUUGCGUUUGUUGCCAUAATT, antisense UUAUGGCAACAAACGCAAUUCTT.

siMARCH5#3 (human): sense GUGACAGUGAUGCAGGUUGUATT, antisense UACAACCUGCAUCACUGUCACTT.

Immunoprecipitation assay

Protein A/G magnetic beads (HY-K0202, MedChemExpress, USA) were incubated with the indicated antibodies at 4 °C for 2 h. Cell lysates (500 μg of total protein per sample) were incubated with beads conjugated to the indicated antibodies at 4 °C for 6 h. The beads were washed with PBST (PBS containing 0.5% Triton X-100) 5 times and then boiled at 95 °C for 10 min with 1× SDS‒PAGE loading buffer. Proteins were detected by liquid chromatography–tandem mass spectrometry (LC‒MS/MS) or western blotting.

LC‒MS/MS

LC‒MS/MS analyses were conducted by Fitgene Biotechnology Co. (Guangzhou, China). In brief, immunoprecipitated (IP) samples were boiled at 95 °C for 10 min with 1× SDS‒PAGE loading buffer, proteins were separated by SDS‒PAGE, and the gels were subjected to silver staining or Coomassie blue staining. Bands of interest were excised and digested. The contained peptides were extracted and dissolved in 2% acetonitrile and 0.1% formic acid. Liquid chromatography was performed using Acclaim PepMap RSLC C18 (Thermo, 160454) and Acclaim PepMap 75 µm X 150 mm (Thermo, 160321) columns. The separated peptides were analyzed with a mass spectrometer (Thermo Scientific Q Exactive). MS data were searched in Sorcerer2-SEQUEST using the reviewed Swiss-Prot database.

Immunofluorescence staining

Cells were washed 3 times with cold PBS, fixed with 4% paraformaldehyde at 37 °C for 15 min and permeabilized with 0.1% Triton X-100 at room temperature for 15 min. Then, the cells were blocked with 5% BSA for 30 min before incubation with the indicated primary antibodies and secondary antibodies (ab150113 and ab150078, Abcam) at 4 °C for 8 h.

Cells were observed using a laser scanning confocal microscope (Leica TCS SP8). Analyses of fluorescence intensities and colocalization were performed using ImageJ (V2.1.4.8; National Institutes of Health).

Western blot and protein half-life analyses

NMCMs were lysed in Cell Lysis Buffer for Western and IP (P0013, Beyotime, China) supplemented with a protease inhibitor cocktail (78430, Thermo Fisher Scientific, Inc.) for 30 min on ice. After the lysates were centrifuged at 4 °C and 12,000 × g for 20 min, the supernatants were collected. Protein concentrations were measured with a BCA Assay Kit (23225, Thermo Fisher Scientific, Inc.).

Twenty micrograms of total protein was boiled at 95 °C for 10 min with 5× SDS‒PAGE loading buffer and then separated by SDS‒PAGE. The proteins were transferred to PVDF membranes (IPVH00010; Millipore, USA), and the membranes were blocked with 5% nonfat milk (P0216; Beyotime, China) for 1 h at room temperature. Then, the membranes were incubated first with the indicated primary antibodies at 4 °C overnight and then with the corresponding secondary antibodies for 1 h at room temperature. Bands were visualized by chemiluminescence with Immobilon Western Chemiluminescent HRP Substrate (WBKLS; Millipore) in a ChemiDoc Imaging System (Bio-Rad, CA, USA). Semiquantification of band densities was performed using ImageJ (V2.1.4.8; National Institutes of Health).

For the protein half-life analysis, cells were treated with cycloheximide (CHX; 20 µM, HY-12320, MedChemExpress), MG132 (10 µM, HY-13259, MedChemExpress) or chloroquine phosphate (CQ; 10 µM, HY-17589A, MedChemExpress) for different times.

Antibodies

The antibodies specific for the following proteins and tags were obtained from the indicated sources: Bax (2772, CST), Bcl2 (ab182858, Abcam), Caspase 9 (9504, CST), Cleaved caspase 9 (9509, CST), Caspase 3 (14220, CST), cleaved caspase 3 (94530, CST), p-MLKL (ab196436, Abcam), MLKL (ab243142, Abcam), RIP (3493, CST), RIP3 (95702, CST), Nrf2 (12721, CST), SLC7A11 (26864-1-AP, Proteintech), GPX4 (67763-1-Ig, Proteintech), ACSL4 (22401-1-AP, Proteintech), MIC60 (10179-1-AP, Proteintech), MIC60 (ab110329, Abcam), TRAP1 (NBP2-47597, Novus), TRAP1 (92345, CST), TRAP1 (sc-13557, Santa Cruz), MIC19 (ab224565, Abcam), MIC13 (PA5-69966, Thermo Fisher), ubiquitin (Ub; 3936, CST), K48-Ub (8081, CST), HA (66006-2-Ig, Proteintech), Flag (66008-4-Ig, Proteintech), His (66005-1-Ig, Proteintech), Myc (60003-2-Ig, Proteintech), GST (66001-2-Ig, Proteintech), MARCH5 (19168S, CST), GAPDH (60004-1-Ig, Proteintech), and beta-tubulin (66240-1-Ig, Proteintech). Uncropped Western blot images are shown in the supplemental material.

Statistical analysis

The data are presented as the mean ± SD from three independent experiments and were analyzed using SPSS 25.0 (IBM Corp.). One-way analysis of variance followed by Tukey’s post hoc test was carried out to analyze differences between groups. P < 0.05 was considered to indicate a statistically significant difference.

Author contributions

LZ, GL, and TZ conceived of the project. LZ, SC, GL, TZ supervised the experiments, analyzed the data, and wrote the manuscript. LZ, YL, LL performed cellular and molecular experiments. All the authors discussed the results and commented on the manuscript.

Funding

This work were supported in part by the National Natural Science Foundation of China (Nos. 82270872, 82171604, 81971759 and 82300909), Natural Science Foundation of Guangdong (No. 2022A1515012249 and 2023B1515020108) and Science and Technology Planning Project of Guangzhou (No. 202206010089).

Data availability

All the data used during the study are available from the corresponding author on request.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41418-023-01218-w.