Abstract

   In this study, we integrated transcriptomic and metabolomic analyses to
   achieve a comprehensive understanding of the underlying mechanisms of
   diabetic cardiomyopathy (DCM) in a diabetic rat model. Functional and
   molecular characterizations revealed significant cardiac injury,
   dysfunction, and ventricular remodeling in DCM. A thorough analysis of
   global changes in genes and metabolites showed that amino acid
   metabolism, especially the breakdown of branched-chain amino acids
   (BCAAs) such as valine, leucine, and isoleucine, is highly
   dysregulated. Furthermore, the study identified the transcription
   factor Gata3 as a predicted negative regulator of the gene encoding the
   key enzyme for BCAA degradation. These findings suggest that the
   disruption of BCAA degradation is a critical characteristic of diabetic
   myocardial damage and indicate a potential role for Gata3 in the
   dysregulation of BCAA metabolism in the context of DCM.

   Keywords: diabetic cardiomyopathy, multi-omics, amino acid metabolism,
   branched-chain amino acids

1. Introduction

   Heart failure (HF) is a major cause of death and illness worldwide
   [[33]1]. According to the World Health Organization, HF affects about
   1–2% of adults in developed countries, with a higher prevalence among
   the elderly (10% of men and 8% of women over 60) [[34]2,[35]3].
   Diabetic cardiomyopathy (DCM), first identified in the 1970s, is a
   specific type of cardiomyopathy seen in diabetic patients without
   hypertension, coronary atherosclerosis, or valvular disease [[36]4].
   The progression of DCM involves several mechanisms that contribute to
   heart dysfunction. Metabolic issues, such as poor glucose utilization
   and abnormal fatty acid metabolism, play a key role in DCM development
   [[37]5]. These metabolic problems lead to an energy imbalance in heart
   cells, causing poor contractile function and structural changes.
   Oxidative stress, defined as an imbalance between reactive oxygen
   species (ROS) production and antioxidant defenses, is another important
   factor in DCM development [[38]6]. Increased ROS production in diabetes
   can damage cell components like lipids, proteins, and DNA, leading to
   cell dysfunction and death [[39]7]. Myocardial fibrosis, characterized
   by excessive collagen and other extracellular matrix deposits, is a
   hallmark of DCM. Fibrosis disrupts the normal structure of the heart
   muscle, affecting its relaxation and contraction, and leads to
   diastolic dysfunction [[40]8]. Cardiac hypertrophy, or the increase in
   size and mass of heart muscle cells, is also common in DCM. This
   hypertrophic response is driven by factors such as insulin resistance,
   inflammation, and neurohormonal activation [[41]9].

   Several medications have been used to target the pathophysiological
   mechanisms in DCM, mainly to control high blood sugar and restore heart
   function. These include antidiabetic drugs, antioxidants, and agents
   that affect myocardial fibrosis and hypertrophy [[42]10]. Despite these
   treatments, there is a persistent increase in morbidity and mortality
   rates related to DCM, highlighting the need for further research. A
   thorough understanding of the fundamental mechanisms of DCM is crucial
   for developing more effective treatment strategies.

   Advancements in molecular biology technology have led to the
   development of high-throughput techniques such as transcriptomics and
   metabolomics. These methods provide valuable strategies for
   understanding the complex pathogenesis of diseases [[43]11,[44]12].

   In this study, we developed a DCM animal model using streptozotocin
   (STZ) and conducted multi-omics profiling of myocardial tissue. We
   performed transcriptome analysis to examine changes in RNA transcripts,
   which helped us identify the genes and pathways involved in DCM
   progression. Additionally, we conducted metabolome analysis to explore
   the global metabolic profile and identify metabolic abnormalities
   associated with DCM. By integrating transcriptomics and metabolomics
   data, we aimed to gain a comprehensive understanding of the underlying
   mechanisms of DCM, paving the way for further research and the
   development of therapeutic interventions.

2. Materials and methods

2.1. Model preparation

   The animal protocols used in this study were reviewed and approved by
   the Animal Care and Use Committee of Sichuan Greentech Bioscience Co.,
   Ltd., China, ensuring compliance with ethical guidelines for animal
   research. Male Sprague-Dawley (SD) rats, aged 6–8 weeks, were obtained
   from Sipeifu (Beijing, China) Experimental Animal Technology Co., Ltd.
   The rats were housed in a controlled environment with a consistent
   12-hour light–dark cycle, a temperature range of 20–26°C, and humidity
   levels between 40 and 70%. They had unrestricted access to water and
   standard rodent chow. After a one-week acclimatization period, the rats
   were randomly divided into two groups: a control group and a model
   group. Diabetes was induced in the model group with a single
   intraperitoneal injection of STZ at a dose of 65 mg/kg, diluted in
   citrate buffer, after an overnight fast. The control group received an
   equivalent volume of citrate buffer. Fasting venous blood samples were
   collected from the tail of each rat on days 3 and 7 after STZ
   injection. Rats in the model group with blood glucose levels exceeding
   11.1 mmol/L were considered diabetic and included in the study. This
   ensured that the model group accurately represented the diabetic
   condition for further analysis of DCM development compared to the
   control group.

   Ethical approval: The research related to animal use has been complied
   with all the relevant national regulations and institutional policies
   for the care and use of animals and has been approved by the Animal
   Care and Use Committee of Sichuan Greentech Bioscience Co., Ltd.,
   China.

2.2. Echocardiographic examinations

   During anesthesia, cardiac contraction and relaxation were
   noninvasively assessed using transthoracic echocardiography, with
   isoflurane administered via a mask at the minimum effective
   concentration. Parasternal long- and short-axis images were obtained 5
   months after STZ or placebo injection. Standard echocardiographic
   techniques were used to acquire two-dimensional, M-mode, and
   color-guided pulsed-wave Doppler images. The following parameters were
   carefully measured: left ventricular internal dimension at end diastole
   (LVIDd), left ventricular internal dimension at end systole (LVIDs),
   interventricular septal thickness at end diastole (IVSd),
   interventricular septal thickness at end systole (IVSs), left
   ventricular posterior wall thickness at end diastole (LVPWd), left
   ventricular posterior wall thickness at end systole (LVPWs), left
   ventricle end-diastolic volume (LVEDV), left ventricle end-systolic
   volume (LVESV), left ventricular fractional shortening (FS), ejection
   fraction (EF), wave E deceleration time (DT), isovolumic diastole time
   (IVRT), E/A ratio, and E/e′ ratio. Each parameter was measured three
   times by a single blinded observer to ensure accuracy and
   reproducibility.

2.3. Biochemical analysis

   At the end of the study, serum samples were collected from all
   experimental rats. The concentrations of cholesterol (CHOL),
   high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein
   cholesterol (LDL-C), triglycerides (TRIG), lactate dehydrogenase (LDH),
   and creatine kinase-MB (CK-MB) were then measured using an automated
   biochemistry analyzer.

2.4. Hemodynamics

   The rats were anesthetized with an intraperitoneal injection of 3%
   pentobarbital sodium at a dose of 60 mg/kg. To maintain a consistent
   body temperature of 37°C, the animals were placed on regulated heating
   pads, and their core temperature was continuously monitored with a
   rectal probe. A polyethylene catheter was inserted into the left
   ventricle through the right jugular artery. After a stabilization
   period of 5 min, physiological signals were recorded using a
   multi-channel physiological signal system equipped with a pressure
   transducer. The recorded parameters included heart rate (HR), left
   ventricular systolic pressure (LVSP), left ventricular diastolic
   pressure (LVDP), the maximum rate of pressure rise during systole
   (+dp/dt [max]), and the maximum rate of pressure decline during
   diastole (−dp/dt [min]).

2.5. Determination of heart weight

   Five months after diabetes induction, the animals were humanely
   euthanized, and their hearts were carefully removed. The hearts were
   dried using filter paper to remove any residual moisture before being
   weighed. The heart weight of each rat was then recorded.

2.6. Histopathology

   Heart tissues from six rats in each group were preserved in a 10%
   formalin solution. These specimens were then embedded in paraffin and
   sectioned into 5 μm thick slices. The slides were stained with
   hematoxylin-eosin (H&E) and Masson’s trichrome stains. Wheat germ
   agglutinin (WGA) staining was also used to assess cardiocyte size.
   Following the manufacturer’s guidelines, the WGA-stained sections were
   examined by a blinded expert using a microscope. Cardiocyte dimensions
   were measured using ImageJ software, with five sections per rat
   analyzed for quantification.

2.7. RNA sequencing (RNA-seq) data analysis

   RNA was extracted from heart tissues using TRIzol (Thermo Fisher
   Scientific, USA), and sequencing libraries were prepared using the
   NEBNext^® Ultra™ RNA Library Prep Kit for Illumina^® (NEB, USA). The
   libraries were purified using AMPure XP beads and sequenced on an
   Illumina Novaseq platform. Differentially expressed genes (DEGs) were
   identified using the DESeq2 algorithm in the DESeq2 R package, with
   genes having a |log2FC| > 1 and a P-value of <0.05 considered
   differentially expressed. A volcano plot was generated using R, with
   colors indicating the filtering criteria. Additionally, Gene Ontology
   (GO) and KEGG pathway enrichment analyses, along with cluster analysis
   of DEGs, were performed using the Wekemo Bioincloud platform
   ([45]https://www.bioincloud.tech). The iRegulon plugin in Cytoscape
   software (v3.10.0) was used to predict transcription factors (TFs) for
   key genes in critical pathways, and TFs with significant expression
   patterns were selected.

2.8. Metabonomic analysis

   The non-targeted metabolomics of the heart in both control and model
   groups were examined using liquid chromatography-mass spectrometry.
   Differentially altered metabolites (DAMs) identified in both positive
   and negative ion modes were analyzed using Student’s t-tests. A partial
   least squares discriminant analysis (PLS-DA) was then conducted, with
   significance thresholds set at |Log2FC| > 1, PLS-DA VIP > 1, and
   P-value <0.05. Significant DAMs were visualized using volcano plots and
   heat maps. To further understand the biological pathways associated
   with these DAMs, a KEGG pathway analysis was performed using the Wekemo
   Bioincloud platform ([46]https://www.bioincloud.tech).

2.9. Integrated analysis of transcriptome and metabolome

   Genes and metabolites with significant differences were analyzed using
   the Pearson correlation coefficient to assess the relationship between
   DEGs and DAMs. To explore the underlying mechanisms, all DEGs and DAMs
   were integrated into MetaboAnalyst (version 5.0, www.metaboanalyst.ca)
   for joint-pathway analysis, identifying the targeted pathways. Key
   biochemical and signal transduction pathways were mapped, integrated,
   and visualized using Pathview Web and MetScape.

2.10. Identification of TFs that regulated the critical pathway

   The Cytoscape plugin iRegulon was used to analyze TFs regulating hub
   genes. iRegulon identifies regulons using motifs and tracks discovery
   within an existing network or a set of regulated genes. The cutoff
   criteria were: enrichment score threshold = 3.0, ROC threshold for AUC
   calculation = 0.03, rank threshold = 5,000, minimum identity between
   orthologous genes = 0.0, and FDR = 0.001.

2.11. Statistical analysis

   Data were analyzed using Stata v15.0 software, with results presented
   as mean ± standard deviation (x̅ ± SD). Comparisons between the two
   groups were conducted using the Student’s t-test, with statistical
   significance defined as a P-value of less than 0.05.

3. Results

3.1. The DCM rats present with systolic and diastolic dysfunction, as well as
ventricular remodeling

   Transthoracic echocardiography was performed 5 months after the onset
   of diabetes to evaluate its impact on cardiac structure and function.
   The results, shown in Table S1 and [47]Figure 1, indicate significant
   changes in the diabetic rats compared to the control group.

Figure 1.

   [48]Figure 1
   [49]Open in a new tab

   DCM resulted in left ventricular diastolic and systolic dysfunction by
   echocardiography analysis. (a–c) Measurement in M-mode of left
   ventricular EF, left ventricular posterior wall (LVPW), and heart rate
   (HR). (d–f) Measurement in Doppler echocardiography of E/A ratio,
   Isovolumic relaxation time (IVRT), and DT (n = 6–8).

   Diabetic rats exhibited significantly reduced posterior wall
   thicknesses (LVPWs and LVPWd) compared to normal rats. Additionally,
   echocardiography revealed a decline in the systolic function of the LV
   in the model group, demonstrated by notably lower EF and FS values.

   Moreover, there was a significant reduction in E, A, and e′ velocities
   in diabetic rats compared to normal rats, indicating impaired diastolic
   function. The E/A and E/e′ ratios were significantly increased in
   diabetic rats, further highlighting diastolic dysfunction.
   Additionally, diabetic rats showed prolonged mitral valve (MV) DT and
   isovolumic relaxation time (IVRT), suggesting impaired left ventricular
   relaxation.

   Interestingly, the HR in diabetic rats was significantly lower than in
   normal rats. However, no significant differences were found between the
   groups in interventricular septum thicknesses (IVSs and IVSd), left
   ventricular diameters (LVIDs and LVIDd), and left ventricular volumes
   (LVESV and LVEDV).

   These findings underscore the significant impact of diabetes on cardiac
   structure and function, evidenced by reduced posterior wall thickness,
   impaired systolic function, and compromised diastolic function. The
   prolonged MV DT and IVRT suggest a deficiency in left ventricular
   relaxation. The lower HR observed in diabetic rats may be a
   compensatory response to these cardiac dysfunctions.

3.2. The diabetic rats exhibit impaired systolic function, relaxation, and
increased ventricular pressure in hemodynamics

   In terms of hemodynamics, DCM rats showed significantly lower LVSP
   (106.9 ± 7.54 mmHg vs 156.33 ± 22.8 mmHg, P < 0.01), maximum rate of
   pressure rise (+dp/dt [max]) (2740.34 ± 241.02 mmHg/s vs 4764.78 ±
   650.36 mmHg/s, P < 0.01), and maximum rate of pressure decline (−dp/dt
   [min]) (2199.68 ± 162.01 mmHg/s vs 4324.01 ± 550.69 mmHg/s, P < 0.01)
   compared to the control group. Additionally, a significantly higher
   LVDP was observed in the model group (Figure S1). These results
   indicate that diabetes and hyperglycemia impair systolic function,
   relaxation, and increase ventricular pressure, which may contribute to
   the development and progression of DCM.

3.3. The DCM rats demonstrate elevated serum cardiac injury markers and
dyslipidemia

   The DCM rats also showed elevated serum cardiac injury markers and
   dyslipidemia. Serum levels of LDH and CK-MB were significantly higher
   in DCM rats compared to control rats (Figure S2a and b), suggesting
   myocardial damage due to diabetes. Furthermore, plasma lipid levels
   were altered; total CHOL, triglycerides, and LDL-C were significantly
   higher in DCM rats than in normal rats (Figure S2c–e), while there was
   no significant difference in HDL-C between the two groups (Figure S2f).

3.4. The DCM rats demonstrate structural deficits in heart

   The DCM rats also showed structural deficits in the heart. After
   hemodynamic measurements, the hearts were removed and weighed. As shown
   in Figure S3b, the heart weight of DCM rats was significantly reduced
   five months after STZ-induced diabetes. H&E staining of heart tissues
   in the control group showed regular myocardial fibers and normal
   morphology of cardiac myocytes ([50]Figure 2a). However, in the
   diabetic model group, there was slight myocardial cell death and a mild
   increase in inflammatory cells in the myocardial interstitium,
   characterized by scattered distribution of eosinophilic enhanced and
   condensed nuclei in myocardial cells and increased inflammatory cells
   within the capillaries. Masson’s trichrome staining showed no obvious
   increased collagen deposition in heart tissues. Measurement of cell
   size in transverse WGA-stained sections revealed decreased
   cardiomyocyte size in diabetic rats. Figure S3a shows representative
   images of myocytes stained with WGA in normal and DCM rats, and Figure
   S3c shows the statistical results. The analysis revealed that DCM
   caused smaller myocytes compared to normal rats (472.06 ± 38.39 μm² in
   DCM vs 587 ± 25.64 μm² in normal, P < 0.001).

Figure 2.

   [51]Figure 2
   [52]Open in a new tab

   The results of RNA-seq sequence analyses of heart. (a) The volcano plot
   showing the distribution of log2 FC and P values for the differential
   expressed genes of the hearts in response to hyperglycemia. Red dots,
   significantly up-regulated genes; green dots, significantly
   down-regulated genes; grey triangles, no significant difference. (b)
   The heat map of top 20 DEGs. The expression level of genes in different
   samples is represented by different colors. The scale at the right
   denoted normalized expression levels (red, high expression; blue, low
   expression). (c) Dot plot of the top 20 GO terms in GO analysis of
   DEGs. (d) The 20 top KEGG pathways in the KEGG analysis of DEGs. (e)
   The numbers of up-regulated and down-regulated genes in the top 20 GO
   terms. (f) The numbers of up-regulated and down-regulated genes in the
   top 20 KEGG pathways.

3.5. Transcriptome analysis reveals significant differences in gene
expression between DCM and normal rats

   Applying the cutoffs of P [adj] < 0.05 and |log2| FC > 1, we identified
   3,880 DEGs between the normal and model groups. Among these, 1,593
   genes were significantly up-regulated, while 2,287 genes were
   significantly down-regulated ([53]Figure 2a and Table S3). Notably,
   among the 3,880 DEGs, 175 TFs were identified, with 94 TFs
   significantly up-regulated and 81 significantly down-regulated (Table
   S3). To illustrate the expression patterns of these DEGs, a heat map
   and hierarchical clustering were constructed, with the top 20 DEGs
   shown in [54]Figure 2b. The heat map depicted gene expression levels
   across different samples using colors ranging from light blue (low
   expression) to dark red (high expression), highlighting distinct gene
   expression patterns between the control and model groups.

   GO analysis revealed that 540 terms were significantly enriched (P <
   0.05) within the DEGs, categorized into biological process, cellular
   component, and molecular function groups, comprising 380, 66, and 94
   terms, respectively (Table S4). Among the top 20 most significantly
   enriched GO terms, as shown in [55]Figure 2c, several are associated
   with mitochondrial structure or biological processes. These include the
   mitochondrial inner membrane, mitochondrial matrix, mitochondrial large
   ribosomal subunit, mitochondrial respiratory chain complex I, assembly
   of mitochondrial respiratory chain complex I, tricarboxylic acid cycle,
   ATP metabolic process, and mitochondrial electron transport. [56]Figure
   3d reveals that the top 20 most significantly enriched GO terms were
   down-regulated, indicated by the down-regulation of most genes in these
   pathways.

Figure 3.

   [57]Figure 3
   [58]Open in a new tab

   The function enrichment and hierarchical clustering of the differential
   metabolites. (a and b) Heat map of the top 20 differential metabolites
   in positive mode and negative mode. (c and d) The top 20 KEGG pathway
   enrichment and topology analysis results of positive mode and negative
   mode.

   The KEGG pathway enrichment analysis identified 82 significantly
   enriched pathways between the control and model groups (Table S5).
   There was a notable enrichment of DEGs in carbohydrate metabolism,
   energy metabolism, and amino acid metabolism. The top 20 KEGG pathways,
   mainly related to metabolic processes, included carbon metabolism, the
   citrate cycle (TCA cycle), pyruvate metabolism, glyoxylate and
   dicarboxylate metabolism, propanoate metabolism, oxidative
   phosphorylation, valine, leucine, and isoleucine degradation, as well
   as fatty acid metabolism ([59]Figure 2d). Additionally, DEGs were
   enriched in pathways related to DCM, cardiac muscle contraction, and
   dilated cardiomyopathy. [60]Figure 3d shows that the top 20 most
   significantly enriched KEGG pathways were down-regulated, indicated by
   the down-regulation of most genes in these pathways.

3.6. Metabolomic analysis reveals metabolite perturbations between in DCM
rats

   Metabolomic profiles were evaluated using PLS-DA, revealing a clear
   distinction between the control and model groups in both positive and
   negative modes (Figure S4). To identify differentially accumulated
   metabolites, the criteria of VIP > 1, P < 0.05, and fold change >2 were
   used. Based on these criteria, 53 metabolites were identified in the
   positive ion mode and 56 in the negative ion mode. In the positive ion
   mode, 30 metabolites were down-regulated in diseased tissues, while 23
   metabolites were up-regulated. In the negative ion mode, 26 metabolites
   were up-regulated, and 30 metabolites were down-regulated (Table S6).

   To visualize the overall changes in the heart metabolome, we
   constructed a clustered heat map of DAMs. This heat map provides a
   detailed view of the metabolomic alterations between the model and
   normal groups ([61]Figure 3c and d). Subsequent analysis revealed that
   these DAMs were predominantly enriched in categories such as amino
   acids, peptides and analogs, fatty acids and conjugates, carbohydrates
   and carbohydrate conjugates, and glycerophosphocholines (Table S2).

   To elucidate the pathways associated with the differentially abundant
   metabolites, we performed enrichment and topological analyses. These
   analyses revealed significant enrichment in multiple pathways. In the
   positive ion mode, notable enrichment was observed in pathways such as
   aminoacyl-tRNA biosynthesis, cofactor and cysteine metabolism,
   histidine metabolism, pantothenate and CoA biosynthesis, amino acid
   biosynthesis, and lysine degradation. In the negative ion mode,
   significant enrichment was found in galactose metabolism, arachidonic
   acid metabolism, and primary bile acid biosynthesis pathways
   ([62]Figure 3a and b, Table S7).

3.7. Integrated analysis of transcriptome and metabolome reveals disrupted
branched-chain amino acids (BCAAs) degradation in DCM rats

   To assess the correlation between the transcriptome and metabolome, the
   correlation coefficient was calculated using the Spearman method with R
   packages. The heat map analysis in [63]Figure 4a shows both positive
   and negative correlations between various genes and metabolites,
   suggesting a potential influence of the transcriptome on metabolic
   processes.

Figure 4.

   [64]Figure 4
   [65]Open in a new tab

   Integrated analysis of transcriptome and metabolome, (a) The top 50
   heat plot of the correlations between metabolites (columns) and genes
   (rows). The red and blue colors show the positive and negative
   correlations between transcriptomics and metabolomics data (*P < 0.05,
   **P < 0.01, ***P < 0.05). (b) Numbers of DEGs and metabolites in
   comapped significant enrichment pathways by joint KEGG enrichment
   analysis. (c) Schematic diagram of the valine, leucine, and isoleucine
   degradation pathway, the red hexagonal boxes with green borders
   represent the DAMs, with larger boxes indicating up-regulated DAMs and
   smaller boxes representing down-regulated DAMs. The blue circles with
   green borders represent the DEGs, with larger circles indicating
   up-regulated DEGs and smaller circles representing down-regulated DEGs.
   (d) Valine, leucine, and isoleucine degradation pathway in DCM hearts,
   expression changes of target genes and metabolites are mapped by
   colors: green color – statistically significant down-regulated genes,
   red color – statistically significant up-regulated genes, yellow color
   – statistically significant up-regulated metabolites, and blue color –
   statistically significant down-regulated metabolites.

   To further investigate the relationship between DEGs and DAMs, we
   mapped them to the KEGG pathway database. This enabled the
   identification of pathways significantly enriched with both
   differential metabolites and DEGs. A joint KEGG enrichment analysis of
   the transcriptome and metabolome revealed six significant pathways
   ([66]Figure 4b). Among these, the degradation of valine, leucine, and
   isoleucine was the most significantly dysregulated at both the gene and
   metabolite levels. A total of 36 DEGs and 2 DAMs were involved in this
   pathway’s dysregulation. Visualization of the integrated pathway-level
   analysis from transcriptomics and metabolomics data using Pathview
   showed that the expression of most genes associated with this pathway
   is down-regulated, including BCAT2, BCKDHA, BCKDHB, DBT, and DLD genes
   ([67]Figure 4c). BCAT2 encodes branched-chain amino acid transferase,
   which is involved in the first reversible step of the degradation
   reaction, converting BCAAs to their respective branched-chain α-keto
   acids (BCKAs). Subsequently, BCKAs undergo irreversible oxidative
   decarboxylation mediated by branched-chain α-keto acid dehydrogenase
   (BCKDH) to produce coenzyme A (CoA) esters. BCKDH comprises four
   catalytic subunits: BCKDHA/E1α, BCKDHB/E1β, DBT/E2, and DLD/E3, with
   BCKDHA and BCKDHB encoding the key catalytic subunits of BCKDH. Our
   study detected an increase in NADH, indicating an inhibition of BCKD
   kinase activity. Along with the down-regulation of BCAA catabolic gene
   expression, there was a significant up-regulation of intramyocardial
   levels of the BCAA valine.

3.8. Gata3 are the key TFs that regulate valine, leucine, and isoleucine
degradation

   To further understand the regulation of valine, leucine, and isoleucine
   degradation in DCM, key DEGs including BCAT2, BCKDHA, BCKDHB, DBT, and
   DLD were analyzed using the iRegulon plugin in Cytoscape software to
   identify TFs. A total of 40 TFs were predicted to regulate BCAT2,
   BCKDHA, BCKDHB, DBT, or DLD. Among these, Zc3h7a, Prdm14, Gata3, and
   Gata2 were predicted to regulate four out of the five key genes.
   Further analysis found that Gata3, detected in our RNA sequencing data,
   was significantly up-regulated. Consequently, Gata3 was identified as a
   negative regulator in the valine, leucine, and isoleucine degradation
   pathway, as its target genes BCAT2, BCKDHB, DBT, and DLD were
   down-regulated ([68]Figure 5).

Figure 5.

   Figure 5
   [69]Open in a new tab

   The TF-target gene regulatory network. Circles represent genes, and
   hexagons represent TFs.

4. Discussion

   In our study, we found that the disruption of valine, leucine, and
   isoleucine degradation is a critical characteristic of myocardial
   damage induced by diabetes, as revealed by integrated transcriptome and
   metabolome analysis in DCM rats. This down-regulation of BCAA
   degradation in diabetic hearts is likely due to the increased
   expression of Gata3, a TF that regulates the key enzymes involved in
   valine, leucine, and isoleucine metabolism, as identified using the
   iRegulon plugin in Cytoscape software.

   BCAAs, which include leucine, isoleucine, and valine, are essential
   nutrients primarily obtained from the diet. They are crucial for
   protein synthesis and energy metabolism and serve as important
   nutritional signals with significant physiological and pathological
   effects [[70]13]. Research in animal models and clinical studies has
   shown that abnormalities in BCAA catabolism are a prominent metabolic
   feature in cardiac diseases [[71]13,[72]14,[73]15,[74]16].

   Elevated plasma BCAA levels have been observed in individuals with HF,
   and epidemiologically, increased circulating BCAA levels are considered
   biomarkers for HF and can predict adverse outcomes in these patients
   [[75]17–[76]24]. Studies in both animal models and clinical settings
   have also demonstrated that a deficiency in BCAA catabolism is a
   prominent metabolic characteristic in cardiac diseases such as
   compensated and decompensated HF and dilated cardiomyopathy
   [[77]13,[78]25–[79]32].

   Upon ingestion, BCAAs contribute to protein synthesis and cellular
   signaling. The homeostasis of BCAAs is tightly regulated, and excess
   BCAAs are degraded via a catabolic pathway in the mitochondrial matrix
   [[80]33]. Initially, BCAAs are converted into branched alpha-ketoacids
   (BCKAs) by branched-chain amino acid transferases (BCATs)
   [[81]34,[82]35]. BCATs exist in two isoforms: the cytosolic form (BCATc
   or BCAT1) and the mitochondrial form (BCATm or BCAT2) [36]. The
   irreversible oxidation of BCKAs, the rate-limiting step of BCAA
   catabolism, is performed by the BCKA dehydrogenase (BCKDH) complex
   [[83]33,[84]37]. The BCKDH complex comprises three components: an E1
   decarboxylase with α/β subunits encoded by BCKDHA and BCKDHB genes; an
   E2 dihydrolipoyl transacylase encoded by the DBT gene; and an E3
   dihydrolipoyl dehydrogenase encoded by the DLD gene [[85]38]. The
   activity of BCKDH is negatively regulated by the phosphorylation of the
   E1α subunit by BCKD kinase, while dephosphorylation by protein
   phosphatase 2Cm (PP2Cm) activates BCKDH [[86]39]. Additionally, an
   increased NADH/NAD + ratio inhibits BCKDH activity [[87]33,[88]40]. In
   our study, the expression of the gene encoding BCAT2 and the subunits
   (Bckdha/E1α, Bckdhb/E1β, and DLD) of BCKDH was down-regulated,
   indicating a disruption in the transcription of key enzymes involved in
   BCAA degradation. Along with the down-regulation of BCAA catabolic gene
   expression, there was a significant accumulation of intramyocardial
   BCAAs, particularly valine.

   The impairment in cardiac BCAA catabolism is unlikely to affect energy
   supply in HF per se, as BCAA oxidation contributes less than 5% to ATP
   production, even in a healthy heart [[89]41,[90]42]. However, animal
   studies have shown that disruptions in BCAA metabolism can affect
   cardiac function, vascular remodeling, glucose and lipid metabolism,
   inflammation, and tissue fibrosis [[91]43–[92]45].

   Deficiencies in BCAA catabolism can directly impair cardiac function
   [[93]29,[94]46]. In a mouse model with inactivation of the BCKDHA gene,
   which encodes the E1α subunit of the BCKDH complex, a complete blockage
   of BCAA catabolic flux resulted in rapid and severe contractile defects
   in the heart [[95]47]. Studies on myocyte-specific PP2Cm-KO mice
   indicate that deficient BCAA catabolism in the heart can directly
   impair cardiac function and accelerate pressure overload-induced
   cardiomyopathy [[96]47]. Our study also observed systolic and diastolic
   dysfunction in DCM rats, accompanied by significant disruption in BCAA
   degradation, suggesting that impaired BCAA degradation may be a
   critical mechanism in diabetes-induced heart injury.

   Inflammation is a key factor in many metabolic diseases, including
   diabetes, obesity, and cardiovascular diseases [[97]48–[98]50]. Liu et
   al. [[99]51] reported that long-term high BCAA supplementation
   increased BCKA levels, inflammation, and tissue fibrosis (liver and
   kidney) in rodents. Additionally, a significant relationship was found
   between isoleucine and IL-6 in patients with metabolic syndrome,
   suggesting that isoleucine may be a potential predictive biomarker for
   the pro-inflammatory state of metabolic syndrome [[100]52]. In our
   study, we observed a mild increase in inflammatory cells in the
   myocardial interstitium, especially within the capillaries, compared to
   the control group.

   The disruption of valine, leucine, and isoleucine degradation has been
   reported to affect glucose and lipid metabolism as well as
   mitochondrial function and structure. Several animal studies have shown
   that the accumulation of BCAAs and their metabolites can directly
   inhibit pyruvate dehydrogenase (PDH) activity, leading to a significant
   reduction in glucose uptake and oxidation [[101]53–[102]55].
   Specifically, high BCAA levels disrupt mitochondrial pyruvate (the end
   product of glucose oxidation) utilization by inhibiting PDH activity
   [[103]54]. Regarding lipid metabolism, the metabolism of BCAAs and
   fatty acids appears to be functionally related, as previous studies
   have shown that KLF15 regulates both the BCAA catabolic pathway and
   PPARα activity, which are involved in fatty acid metabolism in
   cardiomyocytes [[104]13,[105]56–[106]59]. Additionally, studies in
   animal models have indicated that E1α knockdown, which suppresses fatty
   acid oxidation, is associated with changes in the expression of
   molecules that regulate fatty acid metabolism [[107]60]. Research on
   non-alcoholic fatty liver disease (NAFLD) and breast cancer has also
   shown that the loss of BCAA catabolism significantly contributes to
   mitochondrial dysfunction [[108]61–[109]64]. In our study, alongside
   the significant disruption of BCAA catabolism, GO pathway analysis
   revealed a significant enrichment of DEGs in terms related to
   mitochondrial structure or biological processes. These included the
   mitochondrial inner membrane, mitochondrial matrix, mitochondrial large
   ribosomal subunit, mitochondrial respiratory chain complex I, assembly
   of mitochondrial respiratory chain complex I, tricarboxylic acid cycle,
   ATP metabolic process, and mitochondrial electron transport. Most of
   the genes in these categories were down-regulated, indicating impaired
   mitochondrial function and structure. KEGG pathway analysis of DEGs
   revealed significant enrichment in pathways related to carbon
   metabolism, the citrate cycle (TCA cycle), pyruvate metabolism,
   glyoxylate and dicarboxylate metabolism, propanoate metabolism,
   oxidative phosphorylation, and fatty acid metabolism. Furthermore, the
   metabonomic analysis indicated that the DAMs were predominantly
   enriched in sets of metabolites such as peptides and analogs, fatty
   acids and conjugates, carbohydrates and carbohydrate conjugates, and
   glycerophosphocholines. These findings suggest that the disruption of
   branched-chain amino acid metabolism occurs concurrently with
   abnormalities in glucose and lipid metabolism and mitochondrial
   dysfunction. However, the causal relationship among these factors
   requires further investigation.

   GATA-3, part of the GATA family of TFs, plays a crucial role in
   regulating hematopoietic stem cells and their derivatives
   [[110]65,[111]66]. Studies in mGATA3KO mice have shown that the
   deficiency of GATA3-positive macrophages can improve cardiac function
   following myocardial infarction or pressure overload hypertrophy
   [[112]67–[113]69]. In our study, GATA3 was significantly up-regulated
   and identified as a TF regulating the key enzymes involved in the
   degradation of valine, leucine, and isoleucine. GATA3’s involvement in
   amino acid metabolism regulation is supported by studies in breast
   cancer, where Glutaminase 2 (GLS2), transcriptionally regulated by the
   estrogen receptor co-factor GATA3, is more common in hormone
   receptor-positive tumors [[114]70,[115]71]. Furthermore, research on
   clear cell renal cell carcinoma has demonstrated the tumor
   growth-inducing functions of the AMPK–GATA3–ECHS1 pathway. In this
   pathway, BCAA accumulation, increased by ECHS1 down-regulation, acts as
   a promoter of mTORC1 and de novo fatty acid synthesis, thereby
   enhancing cell proliferation [[116]72,[117]73]. Our study predicted
   GATA3 as a negative regulator of the genes encoding key enzymes in the
   degradation of valine, leucine, and isoleucine.

   In summary, our results indicate that the disruption of BCAA
   degradation is a critical characteristic of myocardial damage induced
   by diabetes. This disruption occurs concurrently with abnormalities in
   glucose and lipid metabolism, damage to mitochondrial structure and
   function, significant cardiac dysfunction, cardiac remodeling, and a
   mild increase in inflammatory cells in the myocardial interstitium,
   particularly within the capillaries. GATA3 was identified as a TF
   regulating key catalytic enzymes involved in branched-chain amino acid
   metabolism. While the causal relationship between the disruption of
   BCAA degradation and other factors, such as cardiac dysfunction,
   abnormalities in glucose and lipid metabolism, and increased
   inflammatory cell infiltration in DCM, requires further investigation,
   our findings provide valuable insights for future studies on DCM.

Supplementary Material

   Supplementary material
   Supplementary Table 3
   Supplementary Table 4
   Supplementary Table 5
   Supplementary Table 6
   Supplementary Table 7

Acknowledgements