Hydrogen sulphide reduced the accumulation of lipid droplets in cardiac tissues of db/db mice via Hrd1 S‐sulfhydration

Abstract Accumulation of lipid droplets (LDs) induces cardiac dysfunctions in type 2 diabetes patients. Recent studies have shown that hydrogen sulphide (H2S) ameliorates cardiac functions in db/db mice, but its regulation on the formation of LDs in cardiac tissues is unclear. Db/db mice were injected with NaHS (40 μmol·kg‐1) for twelve weeks. H9c2 cells were treated with high glucose (40 mmol/L), oleate (200 µmol/L), palmitate (200 µmol/L) and NaHS (100 µmol/L) for 48 hours. Plasmids for the overexpression of wild‐type Hrd1 and Hrd1 mutated at Cys115 were constructed. The interaction between Hrd1 and DGAT1 and DGAT2, the ubiquitylation level of DGAT1 and 2, the S‐sulfhydration of Hrd1 were measured. Exogenous H2S ameliorated the cardiac functions, decreased ER stress and reduced the number of LDs in db/db mice. Exogenous H2S could elevate the ubiquitination level of DGAT 1 and 2 and increased the expression of Hrd1 in cardiac tissues of db/db mice. The S‐sulfhydration of Hrd1 by NaHS enhanced the interaction between Hrd1 and DGAT1 and 2 to inhibit the formation of LD. Our findings suggested that H2S modified Hrd1 S‐sulfhydration at Cys115 to reduce the accumulation of LDs in cardiac tissues of db/db mice.

lipid droplets (LDs). 7 A great amount of LDs in the heart is one of the hallmarks of DCM. 8 Nascent LD formation is believed to originate from the accumulation of triglyceride (TAG) between the two leaflets of the endoplasmic reticulum (ER) and is released into cytosol. 9 The ER plays a crucial role in the synthesis and storage of protein and lipid. 10 ER homeostasis can be disturbed by dysfunction of LD biogenesis, which causes ER stress to trigger unfolded protein response (UPR). 11 The UPR is an adaptive mechanism that leads to reduce protein translation, increase transcription of gene related to ER stress and enhance ER-associated protein degradation (ERAD). 12 Hrd1 is an ER-transmembrane E3 ubiquitin ligase, and mounting evidence reveals that this enzyme plays a crucial role in the ERAD of misfolded proteins. 13 Some studies have confirmed that Hrd1 contributed to the adaptive ER stress response, which preserves cardiac function in mouse model of pathological cardiac hypertrophy. 14 Due to LDs originating from ER, dysfunctional LDs can induce ER stress. 15 Recent studies on the formation and biological function of LDs have mainly focussed on adipose and liver, 16 but few studies have investigated the mechanisms in Hrd1 modulating the formation of LDs in the cardiovascular system.
Hydrogen sulphide (H 2 S) is considered the third gasotransmitter, after nitric oxide and carbon monoxide. Endogenous H 2 S is produced via catalysed reactions by three enzymes in mammalian cells: cystathionineβ-synthase (CBS), cystathionineγ-synthase (CSE) and 3-mercaptopyruvate sulphurtransferase . 17 H 2 S plays a key regulatory role in cardiovascular homeostasis that scavenges reactive oxygen species (ROS), reduces apoptotic signalling, modulates mitochondrial respiration and decreases inflammation. 18 Furthermore, H 2 S exerts metabolic regulation effects on hearts. 19,20 Increasing evidence suggests that the circulating levels of H 2 S are decreased in animal models of diabetes and in T2DM patients. 21,22 NaHS and Na 2 S are two well-used H 2 S donor and convenient to handle; however, H 2 S is released immediately from NaHS and Na 2 S in biological media.
Previous studies have been reported that NaHS have an immediate but long-lasting effect via protein sulfhydration or altering gene expression. 23,24 In the present study, we used a well-established type 2 diabetes model (db/db mice) to investigate the mechanism through which exogenous H 2 S regulates the formation of LDs.

| Experimental animals
Homozygous male and female ten-week-old db/db mice on a

| Echocardiographic analysis of cardiac function
Mice were sedated with avertin at a dose of 240 mg·kg -1 and placed in the supine position. Two-dimensionally guided M-mode recordings were obtained from the short-axis view at the level of the papillary muscles by using either an Acuson Sequoia system and an Acuson 15-MHz linear-array transducer or a GE Vivid 7 system with a GE S10-MHz phased-array transducer (General Electric). Left ventricular parameters were measured including ejection fraction (EF, %) and fractional shortening (FS, %).

| Transmission electron microscopy assay
Ultrastructural alterations in cardiac tissues were detected by transmission electron microscopy (TEM). Cardiac tissues for TEM were cut into pieces less than 1 mm3 and fixed in 2.5% glutaraldehyde in 0.1 mol/L sodium cacodylate buffer (pH 7.4) for 4 hours. Tissues were post-fixed in osmium tetroxide and embedded in Epon 812(Electron Microscopy Sciences). Ultrathin sections were stained with uranyl acetate and lead citrate and examined under a Zeiss Axiophot microscope.

| Measurement of hydrogen sulphide level
The measurement of H 2 S level in plasma and isolated cardiac tissues followed the previously established protocol. 25 Briefly, cardiac tissues were homogenized in a 50 mmol/L ice-cold potassium phosphate buffer (pH 6.8) containing a 100 mmol/L potassium phosphate buffer, 10 mmol/L L-cysteine and 2 mmol/L pyridoxal 5′-phosphate. Plasma and cardiac sample were mixed with 10% trichloroacetic acid, respectively. The reaction was stopped by 1% zinc acetate, followed by incubation with N, N-dimethyl-p-phenylenediamine sulphate (DPD) for 15 minutes. The absorbance at 670 nm was measured with a spectrophotometer.

| Cellular experimental protocol
The cultured H9c2 were randomly divided into the following groups

| Isolation of lipid droplets in cardiac tissues
Isolation of lipid droplets in cardiac tissues was prepared by previously described methods. 26 Cardiac tissues were homogenized (100 mg cardiac tissues/1mL ice-cold Tris-EDTA buffer pH7.4 with cocktail inhibitors) on ice. Homogenates were centrifuged at 100 g at 4℃ for 10 min; then, the supernatant was centrifuged at 23,7020 g for 2 h at 4℃ (Ultracentrifuge, Beckman Optima). The LDs were con-

| Oil red O staining
The H9c2 cells were fixed with 4% paraformaldehyde for 20 minutes, and added Oil red O solution for 20 minutes, followed by a 60% isopropanol wash, staining with haematoxylin solution for 1 minutes and washed with water. Lipid droplets were visualized using the microscope (Olympus, XSZ-D2).

| LC-MS/MS analysis
Samples were lysed and trypsin digested according to our previous procedure. 27 The tryptic peptides were dissolved in 0.1% formic acid (solvent A), directly loaded onto a home-made reversed-phase analytical column (15-cm length, 75 μm i.d.). The gradient was comprised of an increase from 6% to 23% solvent B (0.1% formic acid in 98% acetonitrile) over 26 minutes, 23%-35% in 8 minutes and climbing to 80% in 3 minutes then holding at 80% for the last 3 minutes, all at a constant flow rate of 400 nL/min on an EASY-nLC 1000 UPLC system. The peptides were subjected to NSI source followed by tan- Automatic gain control (AGC) was set at 5E4. The parameter settings for mass spectrometer were referred to our published reports.

| Protein identification and quantification
The resulting MS/MS data were processed using MaxQuant search engine (v.1.5.2.8). Tandem mass spectra were searched against human UniProt database concatenated with reverse decoy database.
Trypsin/P was specified as cleavage enzyme allowing up to 4 missing cleavages. The mass tolerance for precursor ions was set as 20 ppm in First search and 5 ppm in Main search, and the mass tolerance for fragment ions was set as 0.02 Da. The false discovery rate (FDR) was adjusted to <1%, and minimum score for modified peptides was set

| Bioinformatic analysis
A volcano plot was constructed to better visualize and identify the differentially expressed proteins between groups. Hierarchical clustering analysis was carried out using Cluster 3.0 software, 29 and a heat map was produced accompanied by a dendrogram depicting the extent of similarity of protein expression among the samples.
For the convenience of gene annotation, corresponding Ensembl gene IDs of the differentially expressed proteins were used for further bioinformatics analysis. To characterize these genes, we tested them for enrichment of gene ontology (GO) biological process, cellular component and molecular function terms by using DAVIDs Functional Annotation Chart tool (Version 6.8). A P value < .05 was controlled for significant enrichment. An important portion of enriched GO terms was selected to construct a network with related proteins using Cytoscape.

| Immunoblot analysis
Western blotting was performed as described previously. Primary (Proteintech, 1:1000), anti-GAPDH (Proteintech, 1:1000). Primary antibodies were incubated overnight at 4℃. Densitometry was conducted with image processing and analysis program AlphaView.SA and the data were expressed as relative units.

| Immunoprecipitation
The cardiac tissues and H9c2 cells were harvested and lysed as previously described. 30 Briefly, cardiac tissues or H9c2 cells lysates were diluted at 2 mg/mL. About 500 μg per sample of protein was used for immunoprecipitation. Protein A/G magnetic beads for immunoprecipitation were conjugated with antibody (10 μg antibody per 500 μg protein) and incubated with H9c2 cells or cardiac tissues lysates overnight at 4℃ with gentle rotation. Beads were collected using centrifugation at 4℃, 10 000 g for 5 minutes, and beads were washed with cell lysis buffer containing 1% PMSF, three times. The precipitates were diluted with loading buffer and boiled for 10 minutes at 100℃ and later used for Western blotting analyses to detect potential interacting proteins.

| S-sulfhydration assay
The assay was carried out as described previously. 31 Briefly, cardiac tissues and cells were homogenized in HEN buffer [250 mmol/L Hepes-NaOH (pH 7.7), 1 mmol/L EDTA, and 0.1 mmol/L neocuproine] supplemented with 100 μmol/L deferoxamine and centrifuged at 13 000 g for 30 minutes at 4℃. Lysates (240 μg) were added to blocking buffer (HEN buffer adjusted to 2.5% SDS and 20 mmol/L MMTS) at 50℃ for 20 minutes with frequent vortexing. The MMTS was then removed by acetone, and the proteins were precipitated at −20℃ for 20 minutes. After acetone removal, the proteins were resuspended in HENS buffer (HEN buffer adjusted to 1% SDS). To the suspension was added 4 mmol/L biotin-HPDP in dimethyl sulphoxide without ascorbic acid. After incubation for 3h at 25℃, biotinylated proteins were precipitated by streptavidin-agarose beads, which were then washed with HENS buffer. The biotinylated proteins were eluted by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer and subjected to Western blot analysis.

| Point mutation of Hrd1
Adenoviruses expression GFP and Hrd1-GFP were purchased from Cyagen Biosciences Inc. The full-length mouse Hrd1 with a single mutation of cysteine 115 to alanine and GFP cDNA was inserted into pM vector (Cyagen Biosciences) between the Kozak and T2A sites.
The adenovirus was added directly to H9c2 cells and after 4-6h for transfection, new fresh medium was added. The H9c2 cells were treated with different reagents after 48h, and the related proteins were detected by Western blot.

| Overexpression of Hrd1 plasmid construction
Adenoviruses expression Hrd1-FLAG were purchased from Cyagen Biosciences Inc. To construct plasmid expression Hrd1-flag proteins, the mouse Hrd1 gene was cloned into pM vector between the Kozak and T2A sites. The adenovirus was added directly to H9c2 cells and after 4-6h for transfection, new fresh medium was added. The H9c2 cells were treated with different reagents after 48 hours, and the related proteins were detected by Western blot.

| Statistical analysis
Results were analysed by using the Prism software package (GraphPad Software). Results are expressed as the mean ± standard error (SEM). More than two groups were compared using a one-way ANOVA and Bonferroni's correction. Differences between individual groups were analysed using Student's t test.

| Characteristics of db/db mice
Db/db mice, leptin receptor deficiency, are widely used as an animal model of type 2 diabetes. We first examined the parameters of the main characteristics of db/db mice. As shown in Figure S1A-E, the blood glucose level and the plasma concentrations of TAG, FFA and insulin were significantly higher in db/db mice than in the wild-type mice and the NaHS-treated wild-type mice. To detect collagen deposition in cardiac tissues, we tested the protein level of collagen I, collagen Ⅲ. Our results found that exogenous H 2 S reduced the expression of collagen I, collagen Ⅲ in db/db mice ( Figure S1F,G). Our data showed that the area of cardiomyocytes in db/db mice was obviously larger than that in wild group and db/db+NaHS groups ( Figure S1H). These results showed that db/db mice can be used as a typical type 2 diabetic animal model.

| H 2 S level and CSE expression in cardiomyocytes in hyperglycaemic and hyperlipidaemic state
An increasing number of studies have suggested that H 2 S is an important gasotransmitter generated by CSE in the cardiovascular system. 32 In the present study, we assessed the H 2 S level and CSE expression in cardiac tissues in db/db mice and in H9c2 cells. Our results showed that the H 2 S level in plasma and cardiac tissues of db/db mice was significantly lower than that of those in wild-type, wild-type and db/db mice with treatment of NaHS ( Figure 1A Figure 1C).
Additionally, the expression of CSE in cardiac tissues of db/db mice was decreased compared with those in wild-type and wild-type and db/db with treated by NaHS groups ( Figure 1D). Our previous study showed that exogenous H 2 S reduced the ubiquitylation level of CSE to inhibit the degradation of CSE. 21 Similarly, the expression levels of CSE in the groups treated with HG+Ole+Pal and PPG (an inhibitor of CSE) were also lower than those in the control and NaHS groups ( Figure 1E). These results suggested that the endogenous H 2 S level is reduced under hyperglycaemic and hyperlipidaemic state.

| Exogenous H 2 S ameliorated DCM in db/ db mice
To further investigate the effects of H 2 S on cardiac functions, the LVEF, LVFS, LVED, LV mass and heart rate were analysed. The LVEF, LVFS and LVED values were decreased in db/db mice and improved by NaHS treatment. However, the LV mass and left ventricular enddiastolic posterior wall (LVPW(d)) in db/db mice were significantly higher than that in the NaHS-treated db/db mice (Table 1 and Figure S2A). LVEF and LVFS are dependent on heart rate; our results showed the reduction in heart rate in db/db mice compared with wild-type mice. Heart size of db/db mice was visibly and hypertrophic as demonstrated by the increased left ventricular weight/ tibia length ratio indexes compared to wild-type mice ( Table 1).
The ultrastructural morphology of cardiac tissues was observed by transmission electron microscopy (TEM), and almost no LDs were found in the cardiac tissues of the control and NaHS-treated control mice ( Figure S2B). The TEM observations also indicated that the size and number of LDs in cardiac tissues of db/db mice were increased in a time-dependent manner, whereas exogenous H 2 S significantly reduced the number and size of LDs in cardiac tissues (Figure 2A-C). As shown in Figure 2E, the staining of neutral fat with BODIPY 493/503 also confirmed that the number of droplets was abundant and TAG content in cardiac tissues of db/db mice was obviously higher than those in wild-type and db/db mice with treatment of NaHS ( Figure 2D). To further investigate whether exogenous H 2 S af-

| Exogenous H 2 S inhibited LD formation through attenuating ER stress
Some studies have demonstrated that excess saturated free fatty acids (sFAs) could trigger unfold protein response (UPR) and thereby lead to ER stress. 34 The formation of LDs was initiated by the occurrence of ER stress. 35 In this study, the expression of Bip, CHOP, p-eIF2α/ eIF2α, p-PERK/PERK, which are the hallmarks of ER stress, was significantly increased in cardiac tissues of db/db mice compared with those of wild-type and db/db mice treated with NaHS ( Figure S4A). In addition, the expression level of the abovementioned proteins in the cells treated with HG+Ole+Pal and Tg (an ER stress inducer and SR/ER Ca 2+ -ATPase inhibitor) was significantly elevated compared with those in the control cells and the cells treated with NaHS and 4-PBA (an inhibitor of ER stress) ( Figure S4B).
To further investigate whether ER stress promotes LD formation, the fluorescent probe BODIPY 558/568 (Red C12) was used to detect the accumulation of LDs in H9c2 cells. Due to the synthesis of neutral lipids in the outer and inner membranes of the ER, the location of LD formation was examined using an ER tracker. Our results showed that the number of LDs in the HG+Ole+Pal and Tg groups was markedly greater than that in the control, NaHS and 4-PBA groups ( Figure S4C and D ). These data suggested that LD formation was closely associated with the induction of ER stress in response to hyperglycaemic and hyperlipidaemic state.

| Comparative proteomic analysis of cardiac tissues from db/db and NaHS-treated db/db mice
Substantial evidence indicates that the formation of LDs is catalysed by complicated enzyme systems, which are elaborated in liver, however, it is indistinct in heart. High-through proteomics technologies combined with bioinformatics approaches have been applied for the identification of proteins associated with the cellular events that contribute to diseases. In this study, a liquid chromatography-tandem mass spectrometry (LC-MS) analysis of cardiac tissues from db/db and db/db mice treated with NaHS was performed.
A total of 2368 proteins were identified, and 1403 proteins were quantified in cardiac tissues. Using the criteria fold change >1.5 and P value < .05, 156 proteins were identified as differentially expressed (DE) between groups (Table S1). Among these, 55 proteins were upregulated and 101 proteins were down regulated in cardiac tissues in db/db mice treated by NaHS compared with db/db mice ( Figure 3A).
To obtain an overview of the function of DE proteins, a molecular function enrichment analysis was conducted ( Figure 3B), and the analysis showed enrichment of the molecular functions related to oxidoreductase activity, cytochrome c activity, sterol binding and ligase activity in the cardiac tissues of NaHS-treated db/db mice compared with db/db mice. The subcellular location of DE proteins was annotated using GO using a database. As shown in Figure 3C, the largest proportion of DE proteins was identified in cytoplasma (53), followed by mitochondria (33), nuclei (22), extracellular (21), plasma membrane (14). The major GO terms including the related biological processes were identified. DE proteins were found to be enriched in metabolic processes, myosin filament organization, mitochondria fission and respiration process ( Figure 3D).

| Exogenous H 2 S regulated the interaction with DGATs and Hrd1
We subsequently tested the expression of Hrd1 to validate the results from the proteomic analysis and found that the expression of Hrd1 was significantly decreased in cardiac tissues and LDs extracted from the hearts of db/db mice compared with those obtained from  To investigate the involvement of Hrd1 in the formation of LDs, we detected the expression of DGAT1 and 2, which are key enzymes that catalyse TAG synthesis. The expression of DAGT1 and 2 in cardiac tissues of db/db mice was significantly higher than that F I G U R E 3 Proteomic analysis of cardiac tissues from db/db mice and with the treatment of NaHS. A, Volcano plot showing the quantitative protein expression in cardiac tissues from db/db mice and NaHS-treated db/db mice. Proteins showing differential expression with fold change >1.5 are marked in colour. B, Heatmap of the molecular functions of the differentially expressed proteins between db/ db mice and NaHS-treated db/db mice (Q means the ratio of db/db-NaHS group to db/db, Q1 < 0.5, Q2 = 0.5-0.66, Q3 = 1.5-2, Q4 > 2; red indicates strong functional enrichment, and the blue colour shows weak functional enrichment). C, Subcellular distribution of the differentially expressed proteins. D, Gene ontology between db/db mice and NaHS-treated db/db mice in NaHS-treated db/db mice ( Figure 4D). To explore the causes for the down-regulation of the expression of DGAT1 and 2 after NaHS treatment, the ubiquitylation levels of DGAT1 and 2 were measured.
Our coimmunoprecipitation (Co-IP) results showed that the ubiquitylation levels of DGAT1 and 2 was significantly increased in cardiac tissue of db/db mice after NaHS treatment compared with that of db/db mice ( Figure 4E). Additionally, PYR41, an inhibitor of ubiquitinactivating enzyme (E1), and MG132, a 26S proteasome inhibitor, were used, and the PYR41 and MG132 treatments increased the protein levels of DGAT1 and 2 in H9c2 cells ( Figure 4F). Moreover, the Co-IP results confirmed the interaction between Hrd1 and DGAT1 and 2 ( Figure 4G).
To further assess whether Hrd1 modulated the ubiquitylation level of DGAT1 and 2, we constructed an Adeno-Hrd1 expression plasmid.
Western blot assay showed that the adeno-Hrd1 plasmid-infected H9c2 cells showed significantly increased expression of Hrd1 compared with the empty plasmid-infected H9c2 cells ( Figure 5A). Our results also revealed that overexpression of Hrd1 did not change the expression of DGAT1 and DGAT2 after HG+Ole+Pal or NaHS treatment ( Figure 5B).
Furthermore, we found that overexpression of Hrd1 also did not affect

| D ISCUSS I ON
The results from the current study show the following: (a) exogenous H 2 S decreased the number of LDs in cardiomyocytes of db/db mice; (b) exogenous H 2 S elevated the protein level of Hrd1 and the Ssulfhydration of Cys115 in Hrd1 to enhance the interaction between Hrd1 and DGAT1 and 2 and to thereby decrease LD formation.
T2DM is associated with altered lipid metabolism, which leads to increase lipolysis and elevated fatty acid concentration in circulatory system and their enhanced uptake of fatty acids by peripheral tissues, including the heart and arteries. 37  represent the means ± SE. *P < .05, **P < .01, ***P < .001, n = 5

| INNOVATION
This study revealed the following: (a) H 2 S protect against diabetic cardiomyopathy by reducing the accumulation of LDs and (b) Ssulfhydrated Hrd1 at cysteine 115 elevates the ubiquitination level of DGAT1 and 2 to inhibit the LD formation. These results uncover a new mechanism through which H 2 S regulates the formation of LDs to ameliorate diabetic cardiomyopathy (Figure 7).

ACK N OWLED G EM ENTS
We thank Jingjie PTM BioLab Co. Ltd. (Hangzhou, China) for the mass spectrometry analysis. This work was supported by the National Natural Science Foundation of China (81670344, 81970317 and 81970411).

CO N FLI C T O F I NTE R E S T S
The authors declare that they have no conflict of interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data sets used and/or analysed in the current study are available from the corresponding author upon reasonable request.