YY1 nitration participates in DbCM cardiomyocyte lipotoxicity by inhibiting ANXA3-induced microlipophagy

Abstract

Aims

Diabetic cardiomyopathy (DbCM), a severe complication of type 2 diabetes mellitus (T2DM), is pathologically characterized by myocardial lipid deposition leading to cardiomyocyte lipotoxic injury. Annexin A3 (ANXA3), a lipid storage inhibitory protein highly expressed in cardiomyocytes, shows altered expression in T2DM tissues, yet its regulatory role in myocardial lipid deposition remains unclear. Enhanced nitrosative stress in T2DM cardiomyocytes induces abnormal protein nitration, and while anti-nitration therapy demonstrates cardioprotective effects, the nitration-mediated regulation of ANXA3 expression requires elucidation. This study investigates ANXA3's regulatory function in T2DM-induced lipid deposition and explores nitration-mediated mechanisms underlying ANXA3 dysregulated expression.

Methods

The T2DM mouse model was established using db/db mice to investigate the pathological mechanisms of DbCM. ANXA3 was overexpressed via cardiac-specific adeno-associated virus delivery to assess its role in lipid deposition. Peroxynitrite scavengers were administered to evaluate lipid droplet accumulation reversal. Site-directed mutagenesis plasmids identified tyrosine residues undergoing nitration.

Results

1) Lipid deposition in the myocardial tissue of DbCM mice is related to downregulation of ANXA3 expression. ANXA3 regulates lipid droplet degradation in DbCM myocardial tissue though regulation of the microlipophagy-Rab7a pathway; 2) The transcription of ANXA3 gene was regulated positively by YY1; 3) In the pathological environment of T2DM, the ability of YY1 to transcribe the ANXA3 gene was decreased, which was related to its nitration at the Y¹⁸⁵ site of YY1 protein.

Conclusion

T2DM promoted cardiomyocyte YY1 nitration at Y¹⁸⁵ and Y³⁸³ residues. Y¹⁸⁵ site nitration inhibited YY1 nuclear translocation, thereby attenuating ANXA3 transcription. Microlipophagy emerged as the principal lipid droplet degradation mechanism in diabetic cardiomyocytes. ANXA3 downregulation suppresses Rab7a expression, impairing microlipophagy and causing lipid deposition that exacerbates DbCM progression. Therapeutic strategies targeting YY1 nitration inhibition or enhancing ANXA3-mediated microlipophagy demonstrate therapeutic potential for mitigating.

Graphical abstract

Abbreviations

Alanine

(A)

Adeno-associated virus 9

(AAV9)

Annexin A3

(ANXA3)

Chromatin immunoprecipitation

(ChIP)

Diabetic cardiomyopathy

(DbCM)

Fe (III) tetrakis (1-methyl-4-pyridyl) porphyrin pentachlorideporphyrin pentachloride

(FeTMPyP)

Heart failure

(HF)

High glucose and palmitic acid

(HGPA)

Lipid droplets

(LDs)

Immunoprecipitation

(IP)

Immunofluorescence

(IF)

Inducible nitric oxide synthase

(iNOS)

Nitrosative modification

(nitration)

Negative control

(NC)

Peroxynitrite

(ONOO-)

Perilipin 2

(PLIN2)

Post-translational modification

(PTMs)

Reverse transcription and quantitative polymerase chain reaction

(RT-qPCR)

Transmission electron microscopy

(TEM)

Type II diabetes mellitus

(T2DM)

Uric acid

(UA)

Yin Yang 1

(YY1)

Tyrosine

(Y)

3-nitrotyrosine

(3-NT)

1. Introduction

Diabetes mellitus is a chronic metabolic disease characterized by hyperglycemia resulting from genetic and environmental factors, and it poses a serious threat to human health. Type 2 diabetes mellitus (T2DM) accounts for more than 90% of all DM patients [1]. T2DM is associated with numerous complications, among which diabetic cardiomyopathy (DbCM) stands as one of the most severe [2]. The progression from T2DM to DbCM represents a prolonged and insidious process, initially manifesting as cardiac diastolic dysfunction and ultimately advancing to systolic dysfunction or heart failure with preserved ejection fraction (HFpEF) [3]. The pathogenesis of DbCM is complex, and no consensus has yet been reached regarding optimal management strategies for its prevention or treatment. Conventional antidiabetic agents such as sodium-glucose cotransporter 2 (SGLT2) inhibitors and glucagon-like peptide-1 (GLP-1) receptor agonists may confer myocardial benefits in diabetic patients [4,5]. Nevertheless, lipid deposition in myocardial tissue and cardiomyocyte lipotoxicity in DbCM patients induced by T2DM remain unresolved issues.

Annexin A3 (ANXA3) is a member of the annexin protein family, widely expressed in various cells, including cardiomyocytes. It usually shows abnormally high expression in multiple tumor tissues, which is of great significance for tumor diagnosis and prognosis evaluation [[6], [7], [8]]. In recent years, with advancements in proteomics and transcriptomics research, increasing attention has been directed towards the significant alterations in ANXA3 expression levels observed in various myocardial diseases, including acute myocardial infarction, dilated cardiomyopathy, and heart failure [[9], [10], [11]]. ANXA3 is a negative regulator of lipid storage [12,13]. In peripheral blood monocytes of patients with cardiovascular diseases, ANXA3 expression is markedly decreased and negatively correlated with body mass index (BMI), suggesting its potential critical role in obesity-related cardiovascular disorders [14]. T2DM represents the most significant metabolic complication of obesity [15]. Before developing diabetic cardiomyopathy, ANXA3 levels in blood samples of T2DM patients are elevated compared to healthy controls [16]. However, transcriptomic analyses of myocardial tissue from DbCM mouse models reveal downregulated ANXA3 expression [17], suggesting its potential involvement in the pathogenesis of T2DM-induced DbCM, although the precise mechanistic role remains to be elucidated.

T2DM leads to a significant generation of reactive oxygen species and reactive nitrogen species, increasing the levels of oxidative stress and nitrosative stress, which may be upstream events of T2DM-induced cell damage [18,19]. Protein post-translational modifications (PTMs) are critical processes in regulating protein localization and degradation, as well as enriching the functional diversity of proteins [20,21]. High levels of oxidative and nitrosative stress promote the nitrosative modification (nitration) of protein tyrosine residues, thereby altering protein function [22]. Preliminary studies have suggested that there is a significant alteration in the expression level of ANXA3 in the myocardium of mice. However, whether nitration plays a role in the regulation of ANXA3 expression remains to be explored.

Therefore, this study aims to clarify the role of ANXA3 in cardiomyocyte lipid deposition and the progression of DbCM and to investigate the potential mechanism underlying altered ANXA3 expression levels from the perspective of protein nitration.

2. Result

2.1. The T2DM-induced DbCM mouse model was successfully established

The db/db mouse is a model with a phenotype resembling human T2DM, which is caused by a leptin receptor gene mutation [23]. Because homozygous db/db mice have defects in reproductive ability and need to be bred by heterozygous mating, it is difficult to identify genes. Therefore, the misty (m) gene (mutually exclusive with the db locus), is introduced to maintain the diabetic characteristics of mice. Because homozygous m/m mice have obvious changes in metabolism compared with normal mice, heterozygous db/m mice of the same age are usually serving as healthy controls for db/db mice [24]. Many studies have shown that the db/db mouse is not only a T2DM model but also a good DbCM model with a typical pattern of heart function changes from a compensatory stage to a decompensatory stage [17]. We selected db/db mice at 36 weeks of age, which were identified as hyperglycemic with decompensated heart function. The appearances of db/db mice showed obesity characteristics (Fig. S1A–B), and serum biochemical tests showed that compared with db/m mice, db/db mice also had significantly higher levels of serum glucose, triglycerides, and cholesterol (Fig. S1C–E), which were in line with the characteristics of T2DM. Echocardiography was employed to assess the cardiac function. The results revealed no significant difference in ejection fraction (EF) between T2DM model mice (db/db group) and healthy controls (db/m group) (Fig. 1A–B). However, left ventricular anterior wall thickness (LVAW) and the E/E’ (Peak early diastolic mitral inflow velocity/Peak early diastolic mitral annular velocity) ratio were significantly elevated in T2DM model mice (Fig. 1A–C and D), indicating preserved systolic function but impaired diastolic function. The Tei index, a reliable parameter for evaluating global left ventricular function, showed a marked increase in T2DM model mice compared with healthy controls (Fig. 1A and E). These findings collectively demonstrate deteriorated cardiac function in the T2DM model mice. Compared to the healthy controls, the myocardial fibrosis-related gene collagen type I alpha 1 chain (Col1a1), hypertrophy-related genes myosin heavy chain gene 7 (Myh7) and natriuretic peptide A (Nppa), and inflammation-related genes tumor necrosis factor (Tnf) were up-regulated in the myocardial tissues of T2DM model mice (Fig. S1F–I). Oil Red O staining and BODIPY staining and triglycerides (TG) concentration results showed lipid deposition in the myocardial tissue of T2DM mice (Fig. 1F–G and Fig. S1J). Transmission electron microscopy (TEM) analysis revealed that the ultrastructure of the myocardial tissue in T2DM mice was disrupted, characterized by severe mitochondrial damage and a large number of lipid droplets (LDs) (Fig. 1H). The above results all indicated that T2DM mice had confirmed progression to the DbCM stage.

Fig. 1.

The T2DM-induced DbCM mouse model was successfully established.

(A) Representative images of M-mode echocardiography, Doppler echocardiography, and tissue Doppler imaging in mouse hearts, employed to assess cardiac function in mice.

(B) Statistical chart of cardiac EF (%) in mice, n = 4.

(C) Statistical chart of cardiac LVAW (mm) in mice, n = 4.

(D) Statistical chart of cardiac E/E′ ratio in mice, n = 4.

(E) Statistical chart of cardiac Tei index (%) in mice, n = 4.

(F) Representative images of Oil Red O staining were used to detect the degree of myocardial lipid deposition. The lipid droplets were manifested as red dots inside cardiomyocytes. Bars: 50 μm or 10 μm 

(G) Representative images of BODIPY staining were used to detect the degree of myocardial lipid deposition. The lipid droplets were manifested as green fluorescent dots. Bars: 10 μm 

(H) Representative images of TEM were used to detect the ultrastructural damage of cardiomyocytes. The red arrows represent lipid droplets (LDs), and the yellow arrows represent damaged mitochondria. Bars: 1 μm.

The data were presented as the mean ± SD. ∗P < 0.05, ∗∗∗P < 0.001 versus the db/m group.

2.2. The expression levels of ANXA3 were decreased in DbCM cardiomyocytes both in vivo and in vitro

To verify the effect of ANXA3 in DbCM, Western blot and immunofluorescence analyses were performed to assess ANXA3 expression level in myocardial tissue. The results demonstrated a decrease in ANXA3 expression in the myocardial tissue of DbCM mice (Fig. 2A–D). To further validate whether ANXA3 expression is reduced in DbCM cardiomyocytes, AC16 cardiomyocytes were stimulated with high glucose and palmitic acid (HGPA) to simulate hyperglycemic and hyperlipidemic conditions in vitro and establish a T2DM cellular model. AC16 cardiomyocytes are routinely cultured in low-glucose DMEM medium (5.5 mmol/L glucose). Based on the literature review, 25 mmol/L glucose is a commonly used concentration for simulating hyperglycemic conditions in vitro cellular experiments [[25], [26], [27]]. Palmitic acid (PA), the major component of saturated fatty acids and exhibiting the strongest lipotoxicity in cardiomyocytes, is typically selected to simulate hyperlipidemic conditions in vitro. Due to significant variations in PA dosage reported across studies and differing cellular sensitivities to PA, the following concentration gradients were established to screen appropriate PA concentrations for subsequent experiments: 100 μM, 200 μM, 500 μM, and 1000 μM, with stimulation durations of 24 h and 48 h. The CCK-8 assay results demonstrated that compared with the vehicle group, cell viability was significantly decreased when AC16 cardiomyocytes were stimulated with 25 mM GLU +200 μM PA for 48 h (Fig. S2A). Simultaneously, Oil Red O staining results revealed abundant lipid droplet accumulation in AC16 cardiomyocytes after 48 h of stimulation with this concentration of HGPA (Fig. S2B). Therefore, 25 mM GLU +200 μM PA combined stimulation of AC16 cardiomyocytes for 48 h was selected to establish the DbCM cell model in vitro. Our results showed that compared with the vehicle group, ANXA3 expression levels in the DbCM cell model were decreased at both mRNA and protein levels (Fig. 2E–G). These results indicate that ANXA3 expression in DbCM cardiomyocytes is decreased both in vivo and in vitro.

Fig. 2.

The expression levels of ANXA3 were decreased in DbCM cardiomyocytes both in vivo and in vitro.

(A-B) Relative ANXA3 protein expression levels in heart tissues of db/m and db/db mice analyzed by Western blot, n = 4.

(C-D) Representative IF images and quantitative analysis of ANXA3 mean fluorescence intensity in heart tissues of db/m and db/db mice (n = 4). Bars: 25 μm 

(E) Relative ANXA3 mRNA expression levels in AC16 cells analyzed by RT-qPCR (n = 6).

(F-G) Relative ANXA3 protein expression levels in AC16 cells analyzed by Western blot (n = 6).

The data were presented as the mean ± SD. ∗P < 0.05 versus the db/m group; ∗∗P < 0.01 versus the Vehicle group; ∗∗∗P < 0.001 versus the Vehicle group.

2.3. Cardiomyocyte-specific overexpression of ANXA3 delayed the progression of T2DM mice to the DbCM stage

After confirming the downregulation of ANXA3 expression in myocardial tissue of DbCM model mice, we further investigated whether cardiomyocyte-specific overexpression of ANXA3 could inhibit DbCM progression before phenotypic manifestation. Literature reports indicate that 6-month-old db/db mice are in the diabetic cardiac compensatory phase [17]. Accordingly, using adeno-associated virus serotype 9 (AAV9) as the vector, we constructed ANXA3-overexpressing virus (AAV9-cTnT-GFP-ANXA3) containing the cardiomyocyte-specific cardiac troponin T (cTnT) promoter and the control virus (AAV9-cTnT-GFP-empty). These viral constructs were administered via tail vein injection to two groups of db/db mice at 6 months of age, and the mice were then housed until 9 months of age. RT-qPCR and Western blotting results demonstrated that compared with db/db mice injected with AAV9-cTnT-GFP-empty (db/db + OE-GFP group), ANXA3 expression in the myocardial tissue of db/db mice injected with AAV9-cTnT-GFP-ANXA3 (db/db + OE-ANXA3 group) was significantly upregulated at both transcriptional and protein levels (Fig. S2C–E). However, there was no significant difference of ANXA3 protein expression in kidney and liver tissues (Fig. S2F–H), confirming successful establishment of the cardiomyocyte specific ANXA3 overexpression db/db mouse model. Next, we investigated whether cardiomyocyte-specific overexpression of ANXA3 improves the overall metabolic profile in T2DM mice. Initially, both groups of db/db mice exhibited an obese phenotype (Fig. S2I). Compared with the negative control (NC) group (db/db + OE-GFP group), the body weight of db/db mice with cardiomyocyte-specific ANXA3 overexpression (db/db + OE-ANXA3 group) showed no reduction and even a slight increase (Fig. S2J). Furthermore, there were no significant differences in serum GLU, TG, and CHOL levels between the two groups (Fig. S2K–M). Both groups of db/db mice exhibited the pathological characteristics of T2DM, indicating that cardiomyocyte-specific ANXA3 overexpression did not alter the overall metabolic profile of db/db mice, which remained in a pathological state of T2DM.

To further investigate whether cardiomyocyte-specific ANXA3 overexpression delays the progression of T2DM mice to the DbCM stage, echocardiography was performed to assess cardiac systolic and diastolic function. Compared with the negative control T2DM mice (db/db + OE-GFP group), the cardiac ejection fraction (EF) of T2DM mice with cardiomyocyte-specific ANXA3 overexpression (db/db + OE-ANXA3 group) showed no significant changes (Fig. 3A and B). However, LVAW (Fig. 3A and C), E/E′ ratio (Fig. 3A and D) and Tei index (Fig. 3A and E) were significantly reduced. These findings suggest that cardiomyocyte-specific ANXA3 overexpression did not alter cardiac systolic function but effectively improved diastolic function in T2DM mice and delayed their progression to the DbCM stage.

Fig. 3.

Cardiomyocyte-specific overexpression of ANXA3 delayed the progression of T2DM mice to the DbCM stage.

(A) Representative images of M-mode echocardiography, Doppler echocardiography, and tissue Doppler imaging in mouse hearts, employed to assess cardiac function in mice.

(B) Statistical chart of cardiac EF (%) in mice, n = 4.

(C) Statistical chart of cardiac LVAW (mm) in mice, n = 4.

(D) Statistical chart of cardiac E/E′ ratio in mice, n = 4.

(E) Statistical chart of cardiac Tei index (%) in mice, n = 4.

The data were presented as the mean ± SD. ∗P < 0.05 versus the db/db + OE-GFP group.

2.4. ANXA3 regulated LDs accumulation in cardiomyocytes both in vivo and in vitro

Myocardial lipid deposition is a typical pathological feature of DbCM, and ANXA3 can negatively regulate lipid deposition [12,13]. Therefore, we investigated the role of ANXA3 in regulating myocardial lipid deposition in T2DM mice. Lipid accumulation in the heart leads to a significant increase in cardiac weight. The results showed that compared with the negative control group (db/db + OE-GFP group), T2DM mice with cardiomyocyte-specific overexpression of ANXA3 (db/db + OE-ANXA3 group) exhibited a significantly lower heart weight-to-body weight ratio (Fig. 4A), markedly reduced lipid deposition in myocardial tissue (Fig. 4B–C), and substantially fewer lipid droplets in cardiomyocytes with intact and visible mitochondrial cristae structures (Fig. 4D). These findings demonstrated that specific overexpression of ANXA3 in cardiomyocytes effectively reduced myocardial lipid deposition in T2DM mice and protects the ultrastructure of cardiomyocytes. Meanwhile, at the cellular level, the Oil Red O staining presented that ANXA3 overexpression reduced LDs accumulation, which was induced by HGPA (Fig. 4E), while decreased expression of ANXA3 increased HGPA-induced LDs accumulation (Fig. 4F). The above results all indicated that ANXA3 regulated LDs accumulation in cardiomyocytes both in vivo and in vitro.

Fig. 4.

ANXA3 regulated lipid deposition in cardiomyocytes both in vivo and in vitro.

(A) Statistical chart of the heart weight/body weight ratio of mice, n = 4.

(B) Representative images of Oil Red O staining were used to detect the degree of myocardial lipid deposition. The lipid droplets were manifested as red dots inside cardiomyocytes. Bars: 50 μm or 10 μm 

(C) Statistical chart of TG concentration (μmol/L) in cardiac tissue of db/db + OE-GFP versus db/db + OE-ANXA3 mice (n = 4).

(D) Representative images of TEM were used to detect the ultrastructural damage of cardiomyocytes. The yellow arrows represent LDs, and the red arrows represent damaged mitochondria. Bars: 1 μm.

(E-F) Representative images of Oil Red O staining were used to detect the degree of lipid droplet accumulation in cardiomyocytes. The lipid droplets were manifested as red dots inside cardiomyocytes. Bars: 100 μm or 20 μm.

The data were presented as the mean ± SD. ∗P < 0.05 versus the db/db + OE-GFP group; ∗∗P < 0.01 versus the db/db + OE-GFP group.

2.5. ANXA3 did not regulate myocardial free fatty acid uptake or lipolysis but positively modulated macroautophagy in T2DM mice

The primary causes of lipid deposition in myocardial tissue are increased uptake of free fatty acids (FFAs) or reduced lipid catabolism. To further investigate the specific role of ANXA3 in regulating lipid deposition in T2DM myocardial tissue, RT-qPCR was employed to detect the expression levels of FFA uptake-related genes in myocardial tissue. The results demonstrated that compared with the healthy control group (db/m group), the expression levels of key FFAs uptake-related genes cd 36 molecule (Cd36) and fatty acid binding protein 3 (Fabp3) were significantly upregulated in the myocardial tissue of DbCM mice (db/db group) (Fig. S3A–B). Correspondingly, the protein levels of CD36 and FABP3 were also elevated. (Fig. S3C–F). However, compared with the negative control group (db/db + OE-GFP group), cardiomyocyte-specific overexpression of ANXA3 in T2DM mice did not lead to significant alterations in the mRNA or protein expression levels of CD36 and Fabp3 in myocardial tissue (Fig. S3G–L). These findings indicated enhanced FFA uptake capacity in the myocardial tissue of DbCM mice, but ANXA3 did not participate in T2DM myocardial lipid deposition by regulating key genes in the FFA uptake pathway. Lipid catabolism primarily involves two pathways: lipolysis and lipophagy. The results showed no significant differences in the expression levels of key lipolysis-related genes patatin-like phospholipase domain containing 2 (Pnpla2) and lipase E (Lipe) in the myocardial tissue between healthy control mice and DbCM mice (Fig. S3M–N), indicating that the lipolysis pathway was not the primary mechanism for lipid catabolism in the myocardial tissue of DbCM mice.

We considered whether lipophagy played an important role. Lipophagy, a selective form of autophagy, selectively clears LDs through the lysosomal pathway by relying on the autophagy machinery. Current research on lipophagy remains limited, and there are no widely accepted molecular markers for detecting lipophagy. Therefore, the status of lipophagy was comprehensively evaluated by analyzing the expression levels of autophagy-related molecular markers combined with TEM observations. Firstly, we found that compared with db/m mice, the expression of autophagy substrate protein SQSTM1/p62 increased and the ratio of LC3 II/I was decreased in the myocardial tissue of db/db mice, which suggested that the autophagy level of myocardium in db/db mice was significantly reduced (Fig. 5A–C). However, the specific overexpression of ANXA3 in cardiomyocytes of db/db mice effectively reversed this phenomenon (Fig. 5D–F). Meanwhile, at the cellular level, HGPA induced autophagy deficiency in AC16 cardiomyocytes (Fig. 5G–I and Fig. S3O), which could be alleviated by ANXA3 overexpression. (Fig. 5J–L). Furthermore, when siRNA was used to downregulate the ANXA3 expression, followed by an increase in p62 expression, and a decrease in the LC3 II/I ratio, indicating reduced autophagy levels. Conversely, overexpression of ANXA3 led to a reduction in p62 expression and increased the LC3 II/I ratio, meaning increased autophagy levels (Fig. S3P–S). The above results suggested that the reduced expression of ANXA3, which was induced by T2DM, may participate in DbCM by inhibiting the macrolipophagy of cardiomyocytes.

Fig. 5.

ANXA3 did not regulate myocardial free fatty acid uptake or lipolysis but positively modulated macroautophagy in T2DM mice.

(A-C) Relative p62 protein expression level and LC3 II/I ratio in heart tissues of db/m and db/db mice analyzed by Western blot, n = 4.

(D-F) Relative p62 protein expression level and LC3 II/I ratio in heart tissues of mice analyzed by Western blot, n = 4.

(G-I) Relative p62 protein expression level and LC3 II/I ratio in AC16 cardiomyocytes analyzed by Western blot, n = 6.

(J-L) Relative p62 protein expression level and LC3 II/I ratio in AC16 cardiomyocytes analyzed by Western blot, n = 6-8.

The data were presented as the mean ± SD. ∗P < 0.05 versus the db/m group or the db/db + OE-GFP group; ∗∗P < 0.01 versus the db/m group or the vehicle group or the OE-NC + vehicle group; ∗∗∗P < 0.001 versus the vehicle group or the OE-NC + vehicle group; ^#P < 0.05 versus the OE-NC + HGPA group; ^###P < 0.001 versus the OE-NC + HGPA group.

2.6. ANXA3 participated in T2DM-induced DbCM by regulating microlipophagy

To examine whether ANXA3 mediates T2DM-induced lipid accumulation in cardiomyocytes by regulating macroautophagy, we analyzed mouse myocardial tissue by transmission electron microscopy (TEM). In cardiomyocytes of db/m mice, LDs were rarely observed, whereas autophagosomes were readily detected. In contrast, TEM analysis of DbCM mice revealed abundant LDs; however, no autophagosomes were found engulfing LDs. While, we observed LD–lysosome contacts suggestive of microlipophagy (Fig. 6A). At the same time, in T2DM mice cardiomyocytes with overexpressing ANXA3, the number of autophagosomes was increased, meanwhile without autophagosomes encapsulating LDs. Instead, more LDs came into contact with lysosomes (Fig. 6B). BODIPY staining results showed that in the myocardial tissue of db/db + OE-NC group mice there were abundant LDs, however the number of lysosomes was relatively low, with only a small portion of LDs in contact with lysosomes. In contrast, in the myocardial tissue of mice with cardiomyocyte-specific overexpression of ANXA3, the number of lysosomes was significantly increased, LDs were markedly reduced, and numerous dispersed LDs exhibited colocalization with lysosomes. (Fig. S3T), suggesting that ANXA3 might be involved in regulating the homeostasis of cardiomyocytes in the pathological environment of T2DM through two independent pathways: macroautophagy and microlipophagy. Microlipophagy is the process in which LDs directly interact with lysosomes without autophagosome formation, transferring LDs components to lysosomal phagocytosis [28].

Fig. 6.

ANXA3 participated in T2DM-induced DbCM by regulating microlipophagy.

(A-B) Representative images of TEM were used to observe the autophagy and microlipophagy in the hearts of mice. The red arrow represents autophagosomes, and the yellow arrow represents lipid droplets in contact with lysosomes. Bars: 1 μm 

(C) Relative Rab7a mRNA expression levels in heart tissues of mice analyzed by RT-qPCR, n = 4.

(D-E) Relative Rab7a protein expression levels in the heart of mice analyzed by Western blot, n = 4.

(F) Relative RAB7A mRNA expression levels in AC16 cells analyzed by RT-qPCR, n = 3-6.

(G-H) Relative Rab7a protein expression levels in AC16 cells analyzed by Western blot, n = 3-5.

The data were presented as the mean ± SD. ∗P < 0.05 versus the db/db + OE-GFP group or the si-NC group; ^#P < 0.05 versus the OE-NC group.

Rab7a (Ras-related protein Rab-7a), a lysosomal Rab protein belonging to the small GTPase family, plays significant roles in autophagy and lipophagy by regulating late endosome-to-lysosome trafficking and the maturation of autophagosomes [[29], [30], [31]]. Studies have reported that RAB7A modulates organelle-lysosome docking sites to promote microautophagy [32]. Additionally, RAB7A positively regulates the quantity and functionality of lysosomes [[33], [34], [35]], thereby indirectly influencing lipophagy. In T2DM-induced DbCM, RAB7A exhibits high colocalization with lipid droplets, facilitating their transport into lysosomes for degradation through interaction with the downstream Rab-interacting lysosomal protein. The downregulation of RAB7A expression has been identified as a primary cause of myocardial lipid accumulation in DbCM mice [29]. The ANXA family demonstrates close associations with RAB7A protein [36], but it is unknown whether ANXA3 can regulate the expression of Rab7. RT-qPCR and Western blot were analyzed for the expression of Rab7 at both mRNA and protein levels. We found that compared to the NC group mice (db/db + OE-GFP), the expression of Rab7 in the myocardial tissue of T2DM mice with overexpressing ANXA3 (db/db + OE-ANXA3) was elevated at both mRNA and protein levels (Fig. 6C–E). Furthermore, when siRNA was used to downregulate the ANXA3 expression, a decrease in Rab7 expression was observed at both the mRNA and protein levels. Conversely, overexpression of ANXA3 led to an increase in Rab7 expression at both mRNA and protein levels (Fig. 6F–H). The above results indicated that ANXA3 participated in regulating microlipophagy by positively regulating the expression of Rab7.

2.7. Microlipophagy dominated LDs degradation in T2DM-induced DbCM

Next, we further investigated whether microlipophagy serves as the primary pathway regulating lipid droplet degradation in T2DM cardiomyocytes. Immunofluorescence (IF) results showed co-localization of LDs and lysosomal marker LAMP1 (lysosomal associated membrane protein 1) in the myocardial tissue of db/db mice (Fig. 7A). We further validated at the cellular level and found that the accumulation of LDs was relatively less and LDs were highly co-localized with lysosomes in the early stages of treating AC16 cells with HGPA for 24 h. However, the accumulation of LDs gradually increased and the co-localization with lysosomes decreased as the stimulation time prolonged. This suggested that the contact of LDs and lysosomes in cardiomyocytes may be compensatorily increased in the early stage of T2DM, aiming to promote LDs degradation. The co-localization of lysosomes and LDs decreased as the disease progressed, hindering LDs degradation and causing a large accumulation of LDs (Fig. 7B–E). However, the co-localization degree between LDs and LC3 was relatively low regardless of whether they were treated with or without HGPA (co-localization score <0.5) (Fig. S4A–D).

Fig. 7.

Microlipophagy dominates LDs degradation in T2DM-induced DbCM.

(A) Representative images of BODIPY staining and IF were used to observe the co-localization of LDs and lysosomes in the hearts of mice. Bars: 10 μm or 2 μm.

(B-E) Representative images of confocal microscope and statistical charts were used to observe the co-localization of LDs and lysosomes in AC16 cells, n = 4. Bars: 8 μm or 2 μm 

(F) Representative images of Oil Red O staining were used to detect the degree of LDs accumulation in the AC16 cells. The red arrow represents lipid droplets. Bars: 50 μm.

(G-J) Relative PLIN2 protein expression levels in AC16 cells analyzed by Western blot, n = 3-4.

The data were presented as the mean ± SD. ∗P < 0.05 versus the Vehicle group; ∗∗P < 0.01 versus the Vehicle group; ^#P < 0.05 versus the HGPA group.

We considered that the absence of autophagosomes encapsulating LDs in the myocardial tissue of T2DM mice may be due to a significant decrease in autophagy levels, which cannot rule out the role of macrolipophagy in T2DM-induced lipid deposition. Thus, to further determine whether the recruitment of LDs to lysosomes depended on the formation of autophagosomes, we treated HGPA-AC16 cells with 3-MA (5 mM) to inhibit the formation of autophagosomes. The Oil Red O staining results showed that inhibiting autophagosome formation didn't exacerbate the accumulation of LDs. However, when HGPA-AC16 cardiomyocytes were treated with chloroquine (50 μM), and rapamycin (50 nM), respectively to inhibit and activate lysosomal activity, the accumulation of LDs in AC16 cells was correspondingly increased or decreased (Fig. 7F). Consistently, Western blot analysis for the expression of PLIN2 demonstrated a similar result (Fig. 7G–J). These results indicated that microautophagy, rather than macroautophagy, serves as the primary pathway regulating lipid droplet degradation in HGPA-cardiomyocytes.

2.8. YY1 positively regulated the expression of ANXA3 at the transcriptional level

Due to the downregulation of ANXA3 expression in DbCM cardiomyocytes, we further investigated the mechanism involved. Our results showed that after using MG-132 (a proteasome inhibitor blocking ubiquitination-dependent degradation), the expression of ANXA3 remained downregulated in HGPA-AC16 cells. However, when we treated HGPA-AC16 cardiomyocytes with actinomycin D (a transcriptional inhibitor), HGPA failed to induce downregulation of ANXA3 expression (Fig. S5A–C). Correspondingly, we had already confirmed that HGPA induced a decrease in ANXA3 expression at the mRNA level (Fig. 2E). Collectively, these findings indicate that HGPA primarily suppresses ANXA3 expression at the transcriptional level. Next, the transcription factors of the ANXA3 gene were used to predicted using JASPAR, PROMO and hTarget online databases, YY1 and SPI1 were obtained by taking the intersection of the predicted results (Fig. S5D). SPI1 primarily functions in hematopoiesis and immune cell regulation, with minimal expression in cardiomyocytes [37,38]. At the same time, YY1 is highly expressed in cardiomyocytes [39,40]. Therefore, we focused on YY1-mediated regulation of ANXA3 in cardiomyocytes. Chromatin immunoprecipitation (ChIP) and dual-luciferase reporter gene assay demonstrated that YY1 could bind to the promoter region of ANXA3 (Fig. 8A) and positively regulate the transcriptional activity of the ANXA3 promoter (Fig. 8B). Furthermore, YY1 was overexpressed with plasmid transfection and knocked down with si-YY1 in AC16 cells. Correspondingly, the expressions of ANXA3 at both mRNA and protein levels were increased in the YY1 overexpression group and were decreased in the YY1 knockdown group (Fig. 8C–H). These results suggested that YY1 positively regulated the expression of ANXA3 at the transcriptional level.

Fig. 8.

YY1 positively regulated the expression of ANXA3 at the transcriptional level.

(A) The binding ability of YY1 to the ANXA3 promoter in AC16 cells was detected by ChIP assay, n = 6.

(B) The ability of YY1 to active the ANXA3 promoter in HEK293 cells was detected by Dual-luciferase reporter assay, n = 6.

(C-D) Relative YY1 and ANXA3 mRNA expression levels in AC16 cells analyzed by RT-qPCR, n = 4.

(E-H) Relative YY1 and ANXA3 protein expression levels in AC16 cells analyzed by Western blot, n = 5.

The data were presented as the mean ± SD. ∗∗P < 0.01 versus the IgG group or the OE-NC group; ∗∗∗P < 0.001 versus the si-NC group (YY1) or the OE-NC group (YY1); ^#P < 0.05 versus the OE-NC group (ANXA3); ^##P < 0.01 versus the OE-NC group (ANXA3) or the si-NC group (ANXA3).

2.9. T2DM attenuated the function of YY1 by promoting nitrosative stress

To further investigate the regulatory mechanism of T2DM on YY1 in the heart, Western blot analysis was performed to assess YY1 expression. The results showed that T2DM and HGPA did not alter YY1 expression levels under both in vivo and in vitro conditions (Fig. 9A–D). However, a significant decrease in YY1 nuclear translocation was observed in cardiomyocytes of DbCM mice (Fig. 9E–F). At the cellular level, HGPA-treated AC16 cells exhibited reduced YY1 nuclear translocation compared with the vehicle group (Fig. 9G–H). To explore the underlying mechanism, we focused on post-translational modifications (PTMs). Given that the T2DM is characterized by both elevated oxidative stress and increased production of reactive nitrogen species (RNS) which resulting in the elevated nitrosative stress, we hypothesized that protein nitration, a PTM driven by the potent oxidant peroxynitrite (ONOO⁻) might be involved [41,42]. 3-nitrotyrosine (3-NT) is the footprint of protein tyrosine residue nitration in vivo. In our study, IF was used to detect the 3-NT level, the results showed that the level of 3-NT in the myocardial tissue of DbCM mice was significantly higher than that in the control group mice (Fig. 9I and Fig. S6A). The generation of peroxynitrite (ONOO⁻) is the basis for inducing protein nitration. After using iron porphyrin Fe (III) tetrakis (1-methyl-4-pyridyl) porphyrin pentachloride (FeTMPyP) and uric acid (UA) as ONOO⁻ scavengers. Both IF and Western blot showed that these two ONOO⁻ scavengers could reverse the reduction of YY1 nuclear translocation induced by HGPA (Fig. 9J–K and Fig. S6B–D). Correspondingly, the decrease of ANXA3 in the AC16 cells induced by HGPA could also be reversed by ONOO⁻ scavengers (Fig. S6E–F), and even the HGPA-induced LDs accumulation (Fig. S6E, G-H). ONOO⁻-mediated signaling pathways, potentially involving protein nitration, play a critical role in impairing YY1 protein nuclear translocation and promoting LDs accumulation in diabetic cardiomyopathy.

Fig. 9.

T2DM attenuated the function of YY1 by promoting nitrosative stress.

(A-B) Relative YY1 protein expression levels in the cardiac tissue of mice analyzed by Western blot, n = 4.

(C-D) Relative YY1 protein expression levels in AC16 cells analyzed by Western blot, n = 6.

(E-F) Representative images of IF and the statistical chart were used to observe the nuclear translocation situation of YY1 in the cardiac tissue of mice. Bars: 25 μm, n = 4.

(G-H) Relative YY1 protein expression levels in the nucleus of AC16 cells analyzed by Western blot, n = 4-5.

(I) Representative images of IF were used to observe the expression level of 3-NT in the cardiac tissue of mice. Bars: 25 μm.

(J-K) Relative YY1 protein expression levels in the nucleus of AC16 cells analyzed by Western blot, n = 3.

The data were presented as the mean ± SD. ∗P < 0.05 versus the Vehicle group; ∗∗∗P < 0.001 versus the db/m group; ^#P < 0.05 versus the HGPA group; ^###P < 0.01 versus the HGPA group.

2.10. Nitrated Y¹⁸⁵ residue of YY1 was involved in inhibiting the transcription of ANXA3

The human YY1 protein contains six tyrosine residues in its sequence, all of which represent potential nitration sites. To identify nitrated tyrosine residues in YY1, recombinant plasmid vectors encoding YY1 mutants with tyrosine-to-alanine substitutions (YY1–Y→A) were constructed. Immunoprecipitation (IP) analysis identified Y¹⁸⁵ and Y³⁸³ as nitration sites of YY1 (Fig. 10A–B). Notably, the HGPA-induced reduction in YY1 nuclear translocation was rescued by the YY1^Y185A mutant, but not by YY1^Y383A (Fig. 10C–D). Similarly, YY1^Y185A remained the mRNA level of YY1 unchanged, but the reduced transcription of ANXA3 induced by HGPA was reversed (Fig. 10E–F). Evolutionary conservation analysis revealed that the Y¹⁸⁵ site of YY1 is conserved across multiple species (Fig. S7A). These results suggested that both the Y¹⁸⁵ and Y³⁸³ sites of YY1 could undergo nitration. However, only the Y¹⁸⁵ site of YY1 nitration participated in the functional regulation of YY1.

Fig. 10.

Tyr185 site of YY1 nitrated was involved in inhibiting the transcription of ANXA3.

(A-B) Relative nitration levels of YY1 in HEK293 cells analyzed by IP, n = 7.

(C-D) Relative YY1 expression levels in the nucleus of AC16 cells analyzed by Western blot, n = 5.

(E-F) Relative YY1 and ANXA3 mRNA expression levels in AC16 cells analyzed by RT-qPCR, n = 6.

The data were presented as the mean ± SD. ∗P < 0.05 versus the WT group; ∗∗P < 0.01 versus the WT group; ∗∗∗P < 0.001 versus the WT group; ^##P < 0.01 versus the WT + HGPA group.

2.11. The concentration of ANXA3 was decreased in serum of T2DM patients

Given that ANXA3 is a secreted protein that can be detected in serum, we confirmed that the concentration of ANXA3 in the serum of heart failure (HF) patients with T2DM was lower than that of HF patients without T2DM (Fig. 11A), and the serum ANXA3 concentration was negatively correlated with serum glucose concentration (Fig. 11B). Overall, these results implied that ANXA3 is associated with T2DM.

Fig. 11.

The concentration of ANXA3 decreased in the serum of T2DM patients.

(A) The ANXA3 concentration in the serum of patients analyzed by Elisa analysis, n = 29.

(B) Correlation analysis of serum glucose and ANXA3 expression levels, n = 58.

The data were presented as the mean ± SD. ∗∗∗P < 0.001 versus the HF without T2DM group.

3. Discussion

T2DM-induced dyslipidemia is an important cause of DbCM [43]. It is a long and hidden process for T2DM to develop into DbCM. The early manifestation of DbCM is myocardial diastolic dysfunction, which progresses to systolic dysfunction or HFpEF in the late stage, and develops into HF as the condition progresses eventually. The pathogenesis of DbCM is rather complex, and at present, there is no consensus on the best management strategy for preventing or treating cardiovascular complications related to DM. The drug therapy commonly used to treat DM such as the sodium-glucose cotransport inhibitors and glucagon-like peptide-1 receptor agonists may have beneficial effects on the myocardium in DM patients [44,45]. However, the issues of lipid accumulation and severe lipid toxicity caused by T2DM remain unresolved. Here, we demonstrated that microlipophagy played a crucial role in DbCM, which was beneficial for reducing LDs accumulation and protecting cardiomyocytes from lipid toxicity damage. Subsequently, T2DM mice with specific overexpression of ANXA3 in myocardial tissue and the HGPA cell model were used to demonstrate that ANXA3 participated in the regulation of microlipophagy and macroautophagy of cardiomyocytes by regulating the expression of Rab7. Following this, we discovered that the level of nitrative stress was increased under long-term T2DM, which promoted nitration of the Y¹⁸⁵ site of YY1 and reduced its transcription of ANXA3, which is the key mechanism for the downregulation of ANXA3 expression in DbCM cardiomyocytes. Finally, we found that as a secreted protein, the concentration of ANXA3 was also reduced in the serum of T2DM patients, and negatively correlated with blood glucose concentration, which may provide some new ideas for the prevention and treatment of DbCM.

ANXA3 is a secreted protein highly expressed in multiple organs throughout the body. Previous studies on the function of ANXA3 were limited to tumor cells. Still, increasing transcriptomic and proteomic analyses have revealed that the expression of ANXA3 also undergoes significant changes in other non-tumor pathological environments in recent years, suggesting that ANXA3 may also play an important role in the steady-state regulation of non-tumor cells [10,46]. Through transcriptome sequencing analysis in Wu et al.'s study, ANXA3 expression was decreased in the myocardial tissue of DbCM mice [17]. Still, its role and mechanism in cardiomyocytes are not yet clear. We validated both in vivo and in vitro that ANXA3 expression was downregulated in DbCM cardiomyocytes, and overexpressed ANXA3 could effectively delay the progression of T2DM to DbCM. Phenotypically speaking, overexpression of ANXA3 in cardiomyocytes significantly reduces LDs accumulation in DbCM cardiomyocytes both in vivo and in vitro. A study has shown that ANXA3 negatively regulates lipid storage through modulation of vesicular trafficking in ccRCC cells [12]. However, there are no reports on how ANXA3 reduces cardiomyocyte lipid deposition induced by T2DM.

The primary cause of T2DM-induced lipid deposition in myocardial tissue is increased uptake or reduced catabolism of FFAs in cardiomyocytes. In FFAs' uptake pathways, the fatty acid transporter CD36 in cardiomyocytes serves as a core regulatory factor. It binds to extracellular FFAs and actively transports them into the cell, mediating approximately 70% of long-chain fatty acid transmembrane transport [47]. Subsequently, the cardiac fatty acid-binding protein Fabp3 transports FFAs to mitochondria or peroxisomes to prevent lipid droplet accumulation and lipotoxicity. However, under lipid metabolism imbalance, elevated CD36 expression exacerbates FFAs uptake by cardiomyocytes, while increased Fabp3 expression accelerates intracellular FFAs transport and further enhances CD36-mediated FFAs uptake. When cardiomyocyte FFAs uptake capacity exceeds lipid droplet metabolic capacity, accumulated intracellular lipid droplets exert lipotoxic effects, causing myocardial injury [48]. This study found upregulated expression of FFAs uptake-related genes in myocardial tissue of DbCM mice, consistent with previous findings [49], indicating enhanced cardiac FFAs uptake in DbCM mice. However, in T2DM mice with myocardial-specific ANXA3 overexpression, FFAs uptake gene expression exhibited no significant alterations, indicating that ANXA3 does not reduce myocardial lipid deposition in T2DM mice by decreasing FFAs uptake. Based on these findings, we hypothesized whether lipid catabolism played an important role in lipid deposition induced by T2DM.

Lipid catabolism primarily occurs through two pathways: lipolysis and lipophagy. In the lipolytic pathway, TG is specifically hydrolyzed into 1,2-diacylglycerol and FFAs by adipose triglyceride lipase encoded by the PNPLA2 gene [50], followed by further hydrolysis of 1,2-diacylglycerol into monoacylglycerol by hormone-sensitive lipase encoded by the LIPE gene [51]. Finally, monoacylglycerol is hydrolyzed into glycerol and FFAs by monoacylglycerol lipase. This study demonstrated that the expression levels of lipolytic enzyme genes Pnpla2 and Lipe remained unchanged in myocardial tissue of DbCM mice, suggesting that lipolysis may not be the predominant pathway regulating lipid droplet degradation in this model. Lipophagy, a selective autophagy dependent on activation of autophagy-related molecules, has recently gained significant attention as a lipid degradation pathway. Due to the lack of widely accepted specific markers, the occurrence of lipophagy requires comprehensive evaluation by combining expression levels of macroautophagy protein markers and TEM observations. First, we confirmed insufficient autophagy in both the myocardial tissue of T2DM-induced DbCM mice and HGPA-induced DbCM cellular models. Notably, myocardial-specific overexpression of ANXA3 in T2DM mice significantly upregulated myocardial autophagy levels. Furthermore, in vitro experiments demonstrated that ANXA3 positively regulates cardiomyocyte autophagy. Moreover, ANXA3 overexpression in cardiomyocytes effectively prevented HGPA-induced autophagy deficiency.

Surprisingly, through TEM observation, we did not find autophagosomes that encapsulate LDs in the heart of DbCM mice, but we discovered some lysosomes that came into contact with LDs. Due to the reduced overall autophagy levels in myocardial tissue of DbCM mice, which may be a critical reason for the difficulty in directly detecting autophagosomes engulfing LDs, the potential role of macroautophagy in lipid droplet degradation within DbCM cardiomyocytes cannot be excluded. Furthermore, the regulatory effect of ANXA3 on macroautophagy pathways in T2DM mouse cardiomyocytes cannot be excluded. However, we observed increased contact between lysosomes and LDs in the myocardial tissue of T2DM mice that overexpressed ANXA3, without autophagosomes encapsulating LDs. This result suggested that although ANXA3 regulated the level of autophagy in cardiomyocytes, in T2DM cardiomyocytes, ANXA3 regulated lipid catabolism through microlipophagy independent of macrolipophagy.

As an important way to reduce lipid deposition, lipophagy has received widespread attention in recent years. It is divided into three types: macrolipophagy, microlipophagy, and chaperone-mediated lipophagy. Studies have shown that during starvation, the adipose triglyceride lipase encoded by the PNPLA2 gene mediates lipolysis of large lipid droplets to reduce their size. Subsequently, smaller lipid droplets form stable membrane contact sites (MCSs) with lysosomes, enabling direct binding and transfer into lysosomes for degradation, thereby reducing intracellular lipid droplet accumulation. This process is termed micro-lipophagy [28]. Unlike macro-lipophagy, micro-lipophagy does not rely on autophagosome formation; instead, lipid droplets are directly engulfed through lysosomal contact. It remains controversial whether lipophagy exists in DbCM [52,53]. However, some studies have observed co-localization of lysosomes and LDs in DbCM [53,54], which is consistent with our research. Compared to the more thoroughly studied macrolipophagy, research on microlipophagy is still limited. Microlipophagy promotes the absorption of LDs by directly binding LDs to lysosomes. However, there is no consensus on molecular markers for macrolipophagy [55]. We further investigated whether micro-lipophagy serves as the predominant pathway regulating LDs degradation in the myocardial tissue of DbCM mice. First, this study observed colocalization between LDs and lysosomal marker LAMP1 protein in myocardial tissue of DbCM mice through IF assays. Subsequently, in vitro experiments revealed that during the early stage of HGPA stimulation (24 h), cardiomyocytes developed a small number of LDs with high colocalization between LDs and lysosomes, indicating active lysosomal engulfment of LDs to attenuate their accumulation at this phase. However, this phenomenon diminished with prolonged HGPA stimulation (48 h), which might be a critical reason for the massive LD accumulation observed in cardiomyocytes during later stages of HGPA stimulation. Furthermore, regardless of HGPA treatment, the colocalization between LDs and autophagosome marker LC3 remained low in cardiomyocytes, suggesting minimal involvement of macro-lipophagy (dependent on autophagosome formation) in lipid droplet clearance during HGPA-induced lipid accumulation. Further validation was conducted at the molecular and morphological levels. We treated AC16 cells with autophagosome formation inhibitor 3-MA, lysosome inhibitor chloroquine and lysosome activator rapamycin, and observed that inhibiting the formation of autophagosomes did not increase LDs accumulation in HGPA-treated AC16 cardiomyocytes, but inhibiting lysosome activity did increase LDs accumulation. Conversely, increasing lysosome activity could reduce LDs accumulation in DbCM cardiomyocytes, which indicates that T2DM-induced LDs accumulation in cardiomyocytes mainly through an autophagosome-independent microlipophagy pathway.

However, compared with the extensively studied macro-lipophagy, research on micro-lipophagy remains limited, and its molecular markers have yet to be definitively identified. RAB7A, a member of the Ras small GTPase family and a lysosomal Rab protein localized to late endosomes and lysosomal membranes, interacts with Rilp protein to participate in membrane tethering processes [56]. The RAB7A-Rilp complex is recruited to late endosomes and lysosomes during various autophagy processes, regulating autophagosome-lysosome fusion, mediating early-to-late endosome maturation and late endosome-lysosome trafficking, and playing critical roles in both macroautophagy and microautophagy regulation [57]. Previous studies reported that lysosomal RAB7A promotes mitochondrial-lysosome contacts to facilitate mitophagy [58]. Moreover, RAB7A is indispensable in lipid droplet degradation [59], as it recruits Rilp to form the RAB7A-Rilp complex in myocardial tissue of T2DM mice, enhancing lipid droplet-lysosome colocalization and micro-lipophagy to eliminate lipid droplets in cardiomyocytes [54]. Additionally, RAB7A is a key mediator of lysosome biogenesis [33], and in multiple cellular models, it positively regulates lysosomal quantity and activity, indirectly modulating micro-lipophagy pathways [34,35]. ANXA family members play pivotal roles in vesicle trafficking, membrane-cytoskeleton interactions, and membrane-membrane contacts. ANXA1 was initially reported to mediate early endosome fusion in a calcium-dependent manner, and later shown to participate in forming endoplasmic reticulum-late endosome membrane contact sites [60]. ANXA2, ANXA5, and ANXA6 interact with early endosomes, autophagosomes/lysosomes, and late endosomes/lysosomes, respectively [36], underscoring their essential roles in diverse autophagy pathways. Furthermore, the ANXA family is closely associated with RAB7A. For instance, elevated ANXA6 expression negatively regulates RAB7A levels, promoting intracellular cholesterol accumulation [61]. However, whether ANXA3 modulates RAB7A expression remains unknown. This study demonstrated that ANXA3 positively regulates RAB7A expression both in vivo and in vitro, which may represent a key mechanism by which ANXA3 modulates micro-lipophagy in DbCM cardiomyocytes.

Although macro-lipophagy, dependent on macroautophagy-related molecular activation, is not the predominant pathway mediating lipid droplet degradation in T2DM myocardial tissue, and ANXA3 cannot regulate macro-lipophagy in cardiomyocytes, the regulatory role of ANXA3 in cardiomyocyte macroautophagy remains critical during the progression from T2DM to DbCM. Cardiomyocytes are terminally differentiated cells with limited regenerative capacity; thus, maintaining appropriate autophagy levels for intracellular material recycling is essential. Numerous studies have demonstrated that insufficient cardiomyocyte autophagy leads to mitochondrial dysfunction [62], elevated oxidative stress [63], and exacerbated inflammatory responses [64], while moderately enhancing autophagy effectively mitigates DbCM progression [65]. Therefore, downregulation of ANXA3 expression, which suppresses cardiomyocyte autophagy, represents a key mechanism underlying T2DM-induced myocardial injury. The combined effects of ANXA3 downregulation-induced micro-lipophagy impairment and autophagy deficiency collectively contribute to DbCM pathogenesis. As this study focused on elucidating ANXA3's regulatory mechanisms in T2DM myocardial lipid catabolism, its role in macroautophagy regulation was not thoroughly explored. Future investigations employing proteomic and transcriptomic sequencing are planned to delineate the molecular mechanisms by which ANXA3 modulates cardiomyocyte autophagy, which holds significant potential for delaying T2DM-induced DbCM progression.

Finally, we investigated the mechanism by which T2DM induced a decrease in ANXA3 expression. Many studies have shown that elevated levels of nitrative stress are upstream events of T2DM-induced DbCM [66,67], and anti-nitrative stress can effectively alleviate T2DM-induced myocardial injury. As a high-energy-consuming organ, the heart primarily depends on mitochondria as its main energy source. T2DM severely damages the structure and function of mitochondria, releasing large amounts of ROS, which increases the oxidative stress level of the heart and generating abundant superoxide anions [68,69]. Meanwhile, T2DM increases the expression of iNOS, which may be related to the elevated expression levels of inflammatory factors. Upregulated iNOS generates a large amount of nitric oxide, which binds with superoxide anions to form peroxynitrite (ONOO⁻) and undergoes nitration with protein tyrosine to form 3-nitrotyrosine, thereby regulating protein function [70,71]. Here, we confirmed that HGPA inhibited the transcription of ANXA3 by promoting nitration at the Y¹⁸⁵ site of YY1. This study reveals the regulatory effect of YY1 nitration on its function, enriching the pathological mechanism of DbCM from the perspective of PTMs.

Indeed, our research has some limitations. First, we did not knock out ANXA3 in the early stage of T2DM to observe whether knocking out ANXA3 during the cardiac compensatory phase would aggravate DbCM. Second, we did not mutate the Y¹⁸⁵ site of YY1 to alanine (A) in vivo to observe whether the mutation could effectively prevent DbCM.Third, we were unable to obtain myocardial tissue samples from DbCM patients, and ANXA3 levels in serum samples may not directly reflect changes in the heart. Furthermore, the transcriptional regulation of RAB7A by ANXA3 was not validated, nor could we rule out the possibility of the direct role of ANXA3 in microlipophagy. Additionally, lipolysis was assessed only at the mRNA level of Pnpla2 and Lipe, without examining the levels of corresponding proteins, their localization, or enzymatic activity. Finally, lipid droplet accumulation in certain AC16 cell experiments (e.g., Fig. 4E and F, S6H) was demonstrated solely by Oil Red O staining, a qualitative method.

4. Conclusion

Here, we identified that in DbCM cardiomyocytes, significantly elevated nitrative stress levels induce nitration modifications at the Y¹⁸⁵ and Y³⁸³ sites of transcription factor YY1. Specifically, nitration at the Y¹⁸⁵ site of YY1 inhibits nuclear translocation of the YY1 protein, thereby attenuating its transcriptional activity toward the ANXA3 gene and leading to downregulated ANXA3 protein expression. Microlipophagy serves as the primary pathway mediating lipid droplet degradation in cardiomyocytes under T2DM-induced pathological conditions. Downregulation of ANXA3 suppresses the expression of RAB7A (a core microlipophagy-related protein), resulting in impaired microlipophagy and consequent lipid accumulation in myocardial tissues. Additionally, reduced ANXA3 expression inhibits macroautophagy levels in cardiomyocytes. The combined insufficiency of macroautophagy and microlipophagy collectively contributes to the pathogenesis and progression of DbCM. This study aims to advance the understanding of T2DM-induced DbCM mechanisms and explore therapeutic strategies targeting protein nitration modifications. By enhancing cardiomyocyte microlipophagy/autophagy levels to delay DbCM progression, our findings provide novel insights and potential therapeutic targets for early intervention and pharmacological interventions of DbCM.

5. Materials and methods

5.1. Human serum sample

The serum samples of HF patients with or without T2DM (patients with malignant tumors, liver and renal failure, chronic gastritis, outflow tract obstruction, pregnancy and lactation women, and infectious diseases were excluded), and 58 were collected from Qingdao Municipal Hospital. The commercial ELISA kit (ELK Biotechnology, Wuhan, China) was used to detect ANXA3 levels in the serum of HF patients with or without T2DM. The experiments were conducted according to the manufacturer's instructions. All procedures performed in studies involving human participants followed the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. All subjects have been informed of the purpose and nature of this study and given written informed consent (No. 2024-KY-031).

5.2. Animal experiments

5.2.1. Animal model establish

8-week-old Male db/m and db/db mice were purchased from Shanghai Model Organisms Co., Ltd (Shanghai, China) and fed with a normal diet to 36 weeks of age in an SPF-level barrier facility. The db/db mice were used to establish the DbCM mouse model and the db/m mice were used as the Control group (n = 4). Echocardiography was used to detect the cardiac function before sacrifice. Then, the mice were euthanized by intraperitoneal injection of 1% pentobarbital sodium (80 mg/kg) for further analysis.

Another batch of 8-week-old male db/db mice was also purchased from Shanghai Model Organisms Co., Ltd (Shanghai, China) and fed with a normal diet to 24 weeks of age. They were randomly divided into two groups to inject adeno-associated virus 9 (AAV9): the db/db + GFP group which injected AAV9-cTnT-GFP (5E+11 vg) via the tail vein of db/db mice and another group (db/db + ANXA3) injected with AAV9-cTnT-ANXA3 (5E+11 vg) via the tail vein of db/db mice (n = 4). Then these mice were fed with the normal diet to 36 weeks of age in an SPF-level barrier facility. Before being sacrificed, echocardiography was used to detect the cardiac function of mice. Then, the mice were euthanized by intraperitoneal injection of 1% pentobarbital sodium (80 mg/kg) for further analysis. The AAV9 was brought from Obio TechnologyCorp.,Ltd (Shanghai, China).

All animal procedures followed the “Guiding Principles in the Use and Care of Animals” published by the National Institutes of Health (NIH Publication No. 85-23, Revised 1996) and approved by the Institutional Animal Care and Use Committee of Capital Medical University (Ethical number: AEEI-2024-147).

5.2.2. Animal cardiac ultrasound

The cardiac ultrasound was used to monitor the cardiac functions of each group of mice. 1% isoflurane was used to anesthetize mice by inhalation. Then Vevo 2100 system (Visual Sonics Inc., Canada) was performed to record EF, LVAW, E/E’, isovolumetric contraction time (ICT), isovolumic relaxation time (IRT) and ejection time (ET). The Tei index was calculated according to the Tei index (ICT + IRT)/ET formula, and the Tei index reflected the overall function of the left ventricle.

5.2.3. Biochemical analysis

The blood samples of mice were collected from the abdominal aorta and left to stand at room temperature for 30 min. After coagulation, they were centrifuged to obtain serum at 3000 g, 4 °C for 15 min. The serum samples were loaded into the Hitachi 7020 automatic biochemical analyzer (HITACHI, Tokyo, Japan) to detect the concentration of blood sugar, cholesterol and triglyceride.

5.2.4. Transmission electron microscope (TEM)

The tip of cardiac tissue fixed with 2.5% glutaraldehyde, was incubated with 2% OsO4 and dehydrated in ethanol. Then, the heart tissue was subjected to critical point drying, sputtered with gold, and observed using an HT7700 TEM (HITACHI, Tokyo, Japan).

5.2.5. Freezing tissue section staining

After being fixed with 4% paraformaldehyde, heart tissues were embedded with OCT and frozen section. Then the standard hematoxylin and eosin staining (Solarbio, G1120), Masson staining (Solarbio, G1346), Sirius Red staining (Solarbio, G1472) and Oil Red O (Solarbio, G1261) were performed to detect the changes in heart tissues. The experiments were conducted according to the manufacturer's instructions.

5.2.6. Immunofluorescence (IF)

After restoring the frozen sections to room temperature, blocked them with 5%bovine serum albumin (w/v)for 30 min, then incubated them with anti-ANXA3 antibody (Proteintech, 11804-1-AP; 1:200 [v/v]), anti-YY1 antibody (Proteintech, 22156-1-AP; 1:200 [v/v]), anti-LAMP1 antibody (Proteintech, 21997-1-AP; 1:200 [v/v]) and anti-3-NT antibody (Millipore, Massachusetts, 06284; 1:400 [v/v]) for 16 h respectively. Washed the sections slightly with PBS and added DAPI containing sealing agent dropwise.

5.2.7. BODIPY staining

The experiments were conducted according to the manufacturer's instructions (LABLEAD, B3162; 2 μM).

Using fluorescence confocal microscopy for imaging and Image J software for fluorescence intensity analysis.

5.3. Cell culture, plasmid, siRNA and transfection

AC16 cells and HEK293 cells were purchased from the American Type Culture Collection (Maryland, USA) and used for in vitro research. Cells were cultured in High glucose media (VivaCell Biosciences, C3103-0500) or Low glucose media (VivaCell Biosciences, C3120-0500) supplemented with 10% (v/v) fetal bovine serum, and were incubated in 5% CO₂/95% O₂ at 37 °C. Once the cells reach 90% fusion, prepare for cell passage.

Cells were divided into the following treatment: (i) Vehicle group:5.5 mmol glucose; (ii) HGPA group:25 mM glucose + 200 μM palmitic acid (PA); (iii) FeTMPyP (Cayman Chemical, 75854) reverse group: 25 mM glucose + 200 μM PA +20 μM FeTMPyP; (iv) UA (Cayman Chemical, 16219) reverse group: 25 mM glucose + 200 μM PA +20 μM UA. FeTMPyP and UA were added to the culture medium half an hour in advance for pretreatment.

Plasmids and siRNA were transfected with Lipo6000 (Beyotime, C0526), and all steps were performed according to the reagent instructions. Specific siRNA oligonucleotides and plasmids targeting human ANXA3 and YY1 were purchased from HanBio (Shanghai, China). RT-qPCR was used to analyze mRNA expression levels 24 h after transfection, and Western blot was used to analyze protein expression levels 48 h after transfection.

5.4. Western blot (WB)

The RIPA lysis buffer (APPLYGEN, C1053) containing 1% (LABLEAD, C0101 [v/v]) protease inhibitor and 1% (LABLEAD, C0104 [v/v]) phosphatase inhibitorwere used to obtain the total protein from heart tissue or cells. Used 20 μg proteins obtained from heart tissues or cells to load and electrophorese in SDS-PAGE and then transfer to a polyvinylidene difluoride membrane. Blocked with 5% (w/v) non-fat-dried milkat room temperature for 1 h. Then the membranes were incubated with the anti-ANXA3 antibody (Proteintech, 11804-1-AP; 1:1000 [v/v]), the anti-YY1 antibody (Proteintech, 22156-1-AP; 1:1000 [v/v]) the anti-PLIN2 antibody (Proteintech, 15294-1-AP; 1:1000 [v/v]), the anti-SQSTM1/p62 antibody (Cell Signaling Technology, 23214; 1:1000 [v/v]), the anti-LC3 antibody (Cell Signaling Technology, 12741; 1:1000 [v/v]), the anti-Rab7 antibody (Proteintech, 55469-1-AP; 1:1000 [v/v])or the rabbit anti-α-Tubulin antibody (ABclonal, AC031; 1:1000 [v/v])overnight at 4 °C. The next day, the membranes afterwashed with TBST buffer three times for 10 minuteseach, incubated with either horseradish peroxidase-conjugated goat anti-rabbit IgG or goat anti-mouse IgG (Zhongshan Golden Bridge Biotechnology, ZB2301 and ZB23051; 40000 [v/v]) for 1 h at room temperature. Repeated the above steps to wash the membranes three times with TBST. ECL Plus substrate (MedChemExpress, HY-K2005) was applied to the blot. The images were captured by the gel documentation system (Bio-Rad, USA) and the optical density of protein bands was analyzed using gel software Image Lab 3.0.

5.5. Real-time PCR analysis (RT-qPCR)

Using TRIzol (Sigma-Aldrich, t9424) to extract the total RNA from AC16 cells, and use 1 μg RNA to synthesize cDNA by HiScript II QRT SuperMix for qPCR (+gDNA wiper) (Vazyme, Nanjing, R223). All steps were performed according to the manufacturer's instructions. Using 7500 Real-Time PCR systems (Applied Biosystems)to perform RT-qPCR. According to the instructions, mixed SYBR Master Mix (Vazyme, Q211-01), DEPC water, primers and cDNA together. The mixture was followed by an initial denaturation at 95 °C for 5 min, and then 40 amplification cycles, each consisting of 95 °C for 10 s and 60 °C for 30 s. The expression of mRNA was normalized to the endogenous TUBA1. Details of all primers used in this study are provided in Supplementary Table 1.

5.6. Chromatin immunoprecipitation (ChIP)

The chromatin fragments were obtained from AC16 cells and used SimpleChIP Plus Sonication Chromatin IP Kit (Cell Signaling Technology, 56383) to conduct the ChIP experiment. All steps were performed according to the manufacturer's instructions.

5.7. Dual-luciferase reporter assay

AC16 cells were planted into the 24-well plate and prepared for transfection. The transfection mixture was divided into four groups: (i) 100 ng renilla luciferase plasmid, 400 ng ANXA3 promoter negative control-firefly luciferase plasmid, and 400 ng pcDNA3.1-negative control plasmid; (ii) 100 ng renilla luciferase plasmid, 400 ng ANXA3 promoter negative control-firefly luciferase plasmid and 400 ng wild-type (WT) YY1 plasmid; (iii) 100 ng renilla luciferase plasmid, 400 ng ANXA3 promoter-firefly luciferase plasmid and 400 ng pcDNA3.1- negative control plasmid; (iv) 100 ng renilla luciferase plasmid, 400 ng ANXA3 promoter-firefly luciferase plasmid and 400 ng WT YY1 plasmid. These plasmids were purchased from HanBio (Shanghai, China). The other steps of transfection referred to the Lipo6000's instructions (Beyotime, C0526).

5.8. Site-directed mutation and cell transfection

The mutation of the tyrosine residue to A in human-derived YY1 plasmids was generated commercially by Hanbio Company (Shanghai, China). The plasmids were transfected into HEK293s by using Lipo6000 (Beyotime, C0526). The steps of transfection were referred to the manufacturer's instructions.

5.9. Immunoprecipitation (IP)

The total protein of cells was extracted by lysis buffer (Tris-HCl 20 mM, Triton 0.1% (v/v), NaCl 100 mM, PMSF 100 μM). Then the protein lysis buffer was gently mixed with 20 μl Protein A/G PLUS-Agarose (Santa Cruz Biotechnology, sc-2003) at 4 °C for 2 h. Centrifuge at 2500g for 5 min at 4 °C, taking the supernatant, and add 2.5 μg anti-Flag antibody (Medical and Biological Laboratories co., M185-3L) with gentle mixing for 8 h at 4 °C. Then added 50 μl Protein A/G PLUS-Agarose (Santa Cruz Biotechnology, sc-2003) to co-incubated for 6 h. Centrifuge at 2500g for 5 min, taking the sediment. Washing the precipitate separately with washing buffer (Tris-HCl 20 mM, Triton 0.1% [v/v])4 times for 10 min each time, and then added with 20 μl 2 × SDS-PAGE buffer and boiled at 99 °C for 10 min followed by Western blot. The steps of the Western blot were the same as in Materials and Methods 5.4. The primary antibodies used for immunoblotting were anti-nitrotyrosine antibody (Millipore, 06284; 1:1000 [v/v]) and anti-Flag antibody (Medical and Biological Laboratories co., M185-3L; 1:10000 [v/v]).

5.10. Statistical analysis

Using SPSS v13.0 to perform statistical analysis. The results were expressed as mean ± standard deviation (SD). When the statistical data followed the normal distribution, the non-paired student t-test was used to compare the two groups, and the one-way analysis of variance (ANOVA) was used to analyze three or more groups. When the statistical data did not follow the normal distribution, the Mann-Whitney test was used to compare the two groups, and the Kruskal-Wallis test was used to analyze three or more groups. Correlations were analyzed using linear correlation analysis. When P < 0.05, the difference was considered statistically significant.

Ethics statement

All human serum samples used in this study were collected with informed consent and approved by the clinical trial ethics committee of Qingdao Municipal Hospital (No. 2024-KY-031). All programs comply with the 1964 Helsinki Declaration and its subsequent revisions or similar ethical standards. Supplementary Table 2 summarizes the general clinical characteristics of the patients. The Ethics Review Committee approved animal studies for Animal Experimentation at Capital Medical University. All animal experiments were conducted according to the policies instituted by the National Institutes of Health Guide for the Care and Use of Laboratory Animals (Ethical number: AEEI-2024-147).

Funding statement

Funding was provided by the National Natural Science Foundation of China (No. 32471233)and Beijing Natural Science Foundation (No.L248045, 7242005) to Dr. Wen Wang.

CRediT authorship contribution statement

Jiayin Chai: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Visualization, Writing – original draft. Lijie Han: Data curation, Formal analysis, Validation, Visualization. Degang Mo: Methodology, Resources. Yunfei Bai: Methodology. Yiming Yang: Data curation, Validation, Visualization. Yuqing Ding: Validation. Shuai Chen: Validation. Xiangning Kong: Supervision. Xinyu Zhu: Supervision. Lijia Xu: Validation. Hongyan Dai: Methodology, Resources. Wen Wang: Conceptualization, Funding acquisition, Project administration, Resources, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no conflict of interest.

Appendix A. Supplementary data

The following is/are the supplementary data to this article:

Data availability

Data will be made available on request.