Eighty 4-week-old male Wistar rats (80–110g; Vital River Laboratory Animal Technology Co., Ltd., Beijing, China) were housed in the animal laboratory of Shandong Provincial Qianfoshan Hospital of Shandong University in specific pathogen-free housing conditions at 20–26°C and 50–60% humidity. The animal study was reviewed and approved by approved by the Institutional Animal Care and Use Committee of Shandong Provincial Qianfoshan Hospital of Shandong University. All animals were adaptively fed a standard diet for 1week (SD; 15% of the calories were fat, Laboratory Animal Center of Shandong University, Jinan, China) and then randomly divided into the following four groups (Figure 1):

(1) Control group (CON group; n=15): fed a SD during the entire study. (2) Diabetes mellitus group (DM group; n=18): received a single intraperitoneal STZ (Sigma-Aldrich, St. Louis, MO, United States) injection at 33mg/kg body weight after a 4-week HFD (40% of the calories were fat; Jiangsu Xietong Pharmaceutical Bio-engineering Co., Ltd., Nanjing, China) followed by a 12-h fast. STZ was dissolved in 0.1mol/L sodium citrate buffer (pH=4.5; Beijing Solarbio Science & Technology Co., Ltd., Beijing, China) to 0.1ml/mg. Three days after the STZ injection, the random blood glucose was measured three consecutive times (≥ 16.7mmol/L was considered a successful model of diabetic rats; Zhang et al., 2020). (3) Sham operation group (SHAM group, n=20): treated like the DM group, but a sham operation was performed 16weeks after the STZ injection. The rats continued to be fed an HFD for 8weeks after the operation and were then euthanized. (4) Sleeve gastrectomy group (SG group, n=27): treated like the DM group, but the SG was performed 16weeks after the STZ injection. The rats continued to be fed an HFD for 8weeks after the operation and were then euthanized.

After the experiments, all rats in the CON group survived, five rats in the DM group died of hyperglycemia, and two and six rats in the SHAM group died of hyperglycemia and infection, respectively. In the SG group, two rats died of hypoglycemia, four rats died of intestinal obstruction, three rats died of infection, and three rats died of gastric leakage. Namely, 15, 13, 12, and 15 rats survived in the CON, DM, SHAM, and SG groups, respectively. In addition, there were 2, 0, 2, and 2 rats in CON group, DM group, SHAM group, and SG group, respectively, which were defined as outliers by Grubbs’ test. Thus, 13, 13, 10, and 13 rats in the CON, DM, SHAM, and SG groups could be included in this study, respectively. To make the sample size of each group equal, 5, 3, 2, and 5 rats from the CON, DM, SHAM, and SG groups including outliers were randomly euthanatized by injecting an overdose of 10% chloral hydrate (6ml/kg). Consequently, each group had 10 rats.

Sleeve gastrectomy: SG was performed as previously reported (Bruinsma et al., 2015). All rats fasted for 12h before operation and were then anesthetized with 2% isoflurane gas. First, a 4cm incision in the middle of the upper abdomen was made. Then, the stomach was be dissociated until the gastric blood vessels and tissue structure were clearly distinguished. The blood vessels of the gastric fundus and the greater gastric curvature were ligated by 7–0 lines (Ningbo Cheng-He Microsurgical Instruments Factory, Ningbo, China) and then interrupted after successful ligation. Next, a 0.5cm incision parallel to the greater gastric curvature was made at the gastric fundus to remove the stomach contents. The gastric fundus and a large portion of the gastric body on the greater gastric curvature were removed, and subsequently, the remaining stomach was sutured with a 5–0 line (Ningbo Cheng-He Microsurgical Instruments Factory). The anatomical position of the abdominal cavity was restored after confirming no active bleeding. Finally, the abdominal cavity was closed layer by layer with 3–0 lines (Ningbo Cheng-He Microsurgical Instruments Factory).

Sham operation: This operation was the same as the above SG procedure before blocking the gastric blood vessels. There were no other interventions, except for exposing abdominal organs, such as the stomach, small intestine, and liver. Additionally, to maintain the same degree of stress as the SG group, we prolonged the operative and anesthesia time in line with the SG group.

The body weight of all the rats was measured before the operation and 2, 4, 6, and 8weeks after the operation. The food intake was measured at the same time points as the body weight. After euthanasia, the heart weight (HW) and tibia length (TL) of the rats were measured and the HW/TL ratio was calculated and analyzed.

The fasting blood glucose (FBG) was measured by a glucometer (Roche One Touch Ultra, Lifescan, Johnson & Johnson, Milpitas, CA, United States) after a 12-h fast at the same time points as the body weight was measured. Blood was collected from the tail vein of anesthetized rats and immediately centrifuged for 8min at 3000rpm, and then, the upper serum was collected and stored in a frozen tube at −80°C. The serum insulin was measured by an ELISA kit (EZRMI-13K, Millipore, Darmstadt, Germany). Finally, the homeostasis model assessment of insulin resistance (HOMA-IR) was calculated to estimate the degree of insulin resistance. The calculation formula was: HOMA-IR=fasting serum insulin (mIU/L)×FBG (mmol/L) /22.5.

The oral glucose tolerance test (OGTT) was performed after a 12-h fast at the same time points as body weight was measured. Rats were administered with 1g/kg glucose by oral gavage, and then, their blood glucose was measured after 0, 10, 30, 60, and 120min. The area under the curve (AUC) of the OGTT (AUC_OGTT) was calculated by the trapezoidal method.

Two days after the OGTT, the insulin tolerance test (ITT) was performed after a 12-h fast. Rats were administered with 0.5IU/kg insulin (Gansulin 40R, Tonghua Dongbao Pharmaceutical Co., Ltd., Tonghua, China) by intraperitoneal injection, and then, their blood glucose was measured after 0, 10, 30, 60, and 120min. The AUC of the ITT (AUC_ITT) was calculated by the trapezoidal method.

Echocardiography was performed after the 12-h fast and before the euthanasia of rats with a Vevo 2,100 imaging system (VisualSonics, Toronto, Canada). Rats were anesthetized with 2% isoflurane. Doppler mode and M-mode imaging were performed to evaluate the cardiac structure. Furthermore, left ventricular end-diastolic diameter, left ventricular end-systolic diameter, ejection fraction, and fractional shortening were analyzed to evaluate the function of the left ventricle.

Positron-emission tomography (PET) scan was performed following a 12-h fast and before the euthanasia of rats, which were anesthetized with 2% isoflurane. After intravenous injection of 800μCi (29.6MBq) 18F-FDG, the entire rat body was continuously scanned with a PET scanner (Metis 1800, Madic Technology Co., Ltd., Linyi, China) for 60min. The PET images recorded show three dimensions: coronal, sagittal, and transverse. Additionally, the average standard uptake value (SUV_Mean) was analyzed by PMOD 4.1 software (PMOD Technologies LLC., Zurich, Switzerland) to evaluate the myocardial glucose uptake of the rats.

The heart tissue was fixed with 4% paraformaldehyde, embedded in paraffin, and sectioned into 5μm sections. The paraffin sections were stained with hematoxylin and eosin (H&E; G1003, Servicebio, Wuhan, China) to evaluate the heart structure. Furthermore, to determine collagen deposition, Sirius Red (G1018, Servicebio, Wuhan, China) and Masson (G1006, Servicebio, Wuhan, China) staining were performed on the 5μm paraffin heart tissue sections. The collagen volume fraction (CVF) was calculated by ImageJ software (National Institutes of Health, Bethesda, MD, United States) and used as a quantitative measure of collagen deposition. To determine the area of cardiac myocytes (CMs), paraffin sections of transversely cut muscle fibers were stained with wheat germ agglutinin (WGA; L4895, Sigma-Aldrich, St. Louis, MO, United States). ImageJ software was used to calculate the CM area. Oil Red O staining assessed myocardial lipid accumulation. The fresh tissue frozen sections were reheated and dried and then fixed it in the fixative solution for15 min. Oil Red solution (G1016, Servicebio, Wuhan, China) was used to stain the frozen sections. And then immersed them into isopropanol for differentiation. After stained by hematoxylin (G1004, Servicebio, Wuhan, China), followed by three washes. Finally, the sections with glycerin gelatin. All the sections were made into digital sliders by a Pannoramic Digital Slide Scanners (Pannoramic DESK, P-MIDI, P250, and P1000, 3DHISTECH Ltd., Budapest, Hungary).

Total RNA was extracted from frozen heart tissue by TRIzol reagent (G3013, Servicebio, Wuhan, China). Single-strand complementary DNA (cDNA) was synthesized by ReverTra Ace qPCR RT Kit (FSQ-101, TOYOBO Co., Ltd., Osaka, Japan). For quantitative real-time PCR, the SYBR Green Realtime PCR Master Mix (QPK-201, TOYOBO Co., Ltd.) was used following the manufacturer’s instructions. The total PCR reaction volume was 20μl, containing 0.8μl of each primer. The target-specific primers were designed by Sangon Biotech (Shanghai, China), and the specific sequences are listed in Table 1.

Paraffin-embedded heart tissue was sectioned into 5μm sections. The paraffin sections were deparaffinized and rehydrated in xylol and alcohol. The antigen was retrieved in a microwave oven with citrate antigen retrieval solution (ZLI 9064, ZSGB-BIO, Beijing, China) after three washes with phosphate-buffered saline (PBS, G0002-2L, Servicebio). Sections were incubated overnight at 4°C with primary antibodies (Collagen I, 1:500, ab270993, Abcam, Cambridge, MA, United States; Collagen III, 1:200, ab7778, Abcam; DUSP6, 1:100, ET1602-18, Huabio, China; p-ERK1/2, 1:200, 4,370T, Cell Signaling Technology, Danvers, MA, United States), followed by three washes with PBS. Sections were then incubated with a universal two-step detection kit (PV-9000, ZSGB-BIO) following the manufacturer’s instructions. After three washes with PBS, the sections were stained with diamino-benzidine (DAB, ZLI-9018, ZSGB-BIO) and hematoxylin and then dehydrated, and transparency were performed. Finally, the sections were made into digital slides with Pannoramic Digital Slide Scanners (Pannoramic DESK, P-MIDI, P250, and P1000, 3DHISTECH Ltd.).

Total protein was extracted from frozen heart tissue using a Minute™ muscle tissue total protein extraction kit (SA-06-MS, Invent Biotechnologies, Inc., United States) following the manufacturer’s instructions. Protein samples were quantified using a BCA Protein Assay Kit (E-BC-K318-M, Elabscience, Wuhan, China). Protein samples were resolved on 10% SDS-PAGE gels (PG212, EpiZyme, Shanghai, China) and transferred onto polyvinylidene fluoride membranes (Millipore, Burlington, MA, United States). The membranes were blocked with 5% fat-free milk and incubated overnight at 4°C with primary antibodies (ANP, 1:1000, ab209232, Abcam, United States; β-MHC, 1:1000, 22280-1-AP, Proteintech, China; p38, 1:1000, ET1702-65, Huabio, China; JNK, 1:1000, RT1550, Huabio, China; ERK1/2, 1:1000, 67170-1-Ig, Proteintech, China; p-p38, 1:1000, ER1903-01, Huabio, China; p-JNK, 1:1000, 4668S, Cell Signaling Technology, USA; p-ERK1/2, 1:1000, 4370T, Cell Signaling Technology, United States; DUSP 6, 1:1000, ET1602-18, Huabio, China; β-ACTIN, 1:1000, ab8226, Abcam, United States). Then, the membranes were washed and incubated with secondary antibodies (Goat anti-Mouse IgG, 1:10000, ab216776, Abcam; Goat Anti-Rabbit IgG, 1:10000, ab6721, Abcam). The protein bands were visualized by ECL (Millipore) and quantified using ImageJ software (National Institutes of Health).

Data were analyzed using Graph Pad Prism 8.0 (GraphPad software, San Diego, CA, United States) and intergroup comparisons were performed with one-way ANOVA followed by Tukey’s multiple comparisons test. Statistical outliers were determined using the Grubbs’ test. p<0.05 was considered statistically significant. Data are presented as the mean±SEM.

The metabolic abnormalities in diabetic obese rats were significantly improved in the SG group (Figure 2). The results showed that the body weight and food intake of the DM group were significantly higher than those of the CON group, whereas SG significantly reduced these parameters (Figures 2A,B). Furthermore, compared with the SHAM group, the SG group had a lower level of FBG from 2weeks after the operation (Figure 2C). In addition to the change in FBG and the serum insulin level, we found that the insulin resistance of the SG group was significantly improved compared with that of the SHAM group as assessed by HOMA-IR (Figures 2C–E). Consistent with the above results, analysis of AUC_OGTT and AUC_ITT demonstrated a significant improvement in insulin resistance in the SG group (Figures 2F,G). All these data above collectively showed that SG can significantly improve the metabolic parameters of diabetic obese rats.

Cardiac hypertrophy was evidenced by increased HW/TL in the DM group compared with the CON group. However, the SG group had a lower HW/TL compared with the SHAM group (Figure 3A). Analyzing the mRNA level of collagen genes demonstrated that the expression of Col1a1 and Col3a1 in the DM group was higher than that in the CON group (Figure 3B). Notably, SG significantly reduced the expression of these genes compared with that in the SHAM group, yet there was no significant difference in the expression of Col4a5 and Col9a1 among these four groups (Figure 3B). Furthermore, immunohistochemistry showed that the expression intensity and distribution of collagen I and collagen III in the DM group were significantly higher than those in the CON group (Figure 3C). Furthermore, the SG group showed lower expression intensity and distribution of collagen I and collagen III compared with that in the SHAM group, but it did not decrease to the level of the CON group (Figure 3C). With respect to natriuretic peptide A (ANP) and cardiac muscle myosin heavy chain β isoform (β-MHC), the DM group had a higher expression of both, compared with the CON group (Figures 3D,E). Additionally, SG significantly decreased this expression, suggesting that the cardiac hypertrophy in the SG group was improved (Figures 3D,E). Similarly, the changes in the mRNA levels of Anp and β-mhc confirmed the above results (Figure 3F).

Moreover, echocardiography, an important method for evaluating heart function, showed that SG significantly improved the heart function and left ventricular wall thickness of diabetic obese rats (Figure 4A). The SG group showed smaller left ventricular end-diastolic diameter and left ventricular end-systolic diameter, approximately 2mm and 1mm reduction, respectively, compared with the SHAM group (Figures 4B,C). Furthermore, SG significantly elevated the ratio of ejection fraction and fractional shortening, by approximately 20 and 10%, respectively (Figures 4D,E).

All these results suggest that SG improved the degree of myocardial fibrosis and diabetes-induced cardiac hypertrophy, and it had a positive effect on the heart function impaired by DM.

Multiple studies have shown that DM can result in obstruction of myocardial glucose uptake (Steneberg et al., 2018). Thus, we evaluated myocardial glucose uptake by PET scans (Figure 4F). The SUV_Mean of the CON group and SG group was significantly higher than that of the DM group and SHAM group, respectively; however, the SUV_Mean of the SG group did not reach the level of the CON group (Figure 4G). These results suggest that SG improved the diabetes-induced obstruction of myocardial glucose uptake.

Many studies have reported that diabetes causes changes in myocardial histomorphology (Sassi et al., 2017; Pabel et al., 2018). Compared with the CON group, the CMs in the DM group were disordered, the myocardial fibers were fragmented and broken, and the cytoplasmic distribution was uneven. Notably, SG significantly ameliorated these changes (Figure 5A). To evaluate CM size, WGA staining was used to analyze the CM area. The results showed that DM led to larger CMs, whereas in the SG group, the area of CMs was significantly smaller than that in the SHAM group (Figures 5B,F). Myocardial fibrosis, a significant feature of cardiac hypertrophy, was evaluated by the CVF, which was determined by Sirius Red staining and Masson staining (Figures 5C,D). The SG group showed a significantly reduction in CVF, compared with the SHAM group (Figure 5G). Oil Red O staining was performed to evaluate the lipid accumulation. As a result, compared with the SHAM group, SG group showed a significantly reduction in lipid accumulation (Figure 5E). Namely, SG reduced the size of CMs and improved their arrangement and the cytoplasmic distribution. Furthermore, it significantly improved myocardial lipid accumulation and fibrosis.

Based on previous studies showing that the activation of MAPK signaling pathways resulted in cardiac hypertrophy (Singh et al., 2017), we investigated the effect of SG on these pathways. Western blot was performed to analyze the expression of MAPKs. The myocardial expression of p-p38, p-JNK, and p-ERK1/2 was significantly higher in the DM group and SHAM group, compared with the CON group, but these three MAPKs were significantly downregulated after SG (Figures 6A,B). The results of the ratio of p-p38 to total p38, p-JNK to total JNK, and p-ERK1/2 to total ERK1/2 were consistent with the above result (Figures 6C–E). Taken together, we deduced that SG ameliorated diabetes-induced cardiac hypertrophy, at least partly, associated with inhibition of MAPK signaling pathways.

DUSPs specifically dephosphorylate and inactivate MAPKs (Liu and Molkentin, 2016). ERK1/2 plays a key role in cell growth, differentiation, and proliferation (Mutlak and Kehat, 2021). Previous study has reported that DUSP6 may inhibit ERK1/2 (Ramkissoon et al., 2019). Based on these theories, we investigated the myocardial expression of DUSP6. As we speculated, Western blot showed that the DUSP6 expression in the DM group was significantly lower than that in the CON group, whereas it was significantly upregulated after SG (Figures 6F,G). The real-time PCR results were consistent with the protein expression results (Figure 6H). These data suggest that the upregulation of DUSP6 may correlate with SG.

To determine the effect of DUSP6 on ERK1/2 dephosphorylation, immunohistochemistry and quantitative analysis were performed. Two consecutive paraffin-embedded heart tissue sections were incubated with primary antibodies for DUSP6 and p-ERK1/2. We observed a clear negative correlation between DUSP6 and p-ERK1/2 expression (Figure 6I). These results were consistent with our previous results (Figures 6A–H). Namely, the results indicated that in the heart tissue, DUSP6 promoted ERK1/2 dephosphorylation, thereby partially ameliorating diabetes-induced cardiac hypertrophy.