Role of duodenal mucosal resurfacing in controlling diabetes in rats

Abstract

BACKGROUND

The duodenum plays a significant role in metabolic regulation, and thickened mucous membranes are associated with insulin resistance. Duodenal mucosal resurfacing (DMR), a new-style endoscopic procedure using hydrothermal energy to ablate this thickened layer, shows promise for enhancing glucose and lipid metabolism in type 2 diabetes (T2D) patients. However, the mechanisms driving these improvements remain largely unexplored.

AIM

To investigate the mechanisms by which DMR improves metabolic disorders using a rat model.

METHODS

Rats with T2D underwent a revised DMR procedure via a gastric incision using a specialized catheter to abrade the duodenal mucosa. The duodenum was evaluated using histology, immunofluorescence, and western blotting. Serum assays measured glucose, lipid profiles, lipopolysaccharide, and intestinal hormones, while the gut microbiota and metabolomics profiles were analyzed through 16S rRNA gene sequencing and ultra performance liquid chromatography-mass spectrum/mass spectrum, severally.

RESULTS

DMR significantly improved glucose and lipid metabolic disorders in T2D rats. It increased the serum levels of cholecystokinin, gastric inhibitory peptide, and glucagon-like peptide 1, and reduced the length and depth of duodenal villi and crypts. DMR also enhanced the intestinal barrier integrity and reduced lipopolysaccharide translocation. Additionally, DMR modified the gut microbiome and metabolome, particularly affecting the Blautia genus. Correlation analysis revealed significant links between the gut microbiota, metabolites, and T2D phenotypes.

CONCLUSION

This study illustrates that DMR addresses metabolic dysfunctions in T2D through multifaceted mechanisms, highlighting the potential role of the Blautia genus on T2D pathogenesis and DMR’s therapeutic impact.

Key Words: Type 2 diabetes; Duodenal mucosal resurfacing; Gut microbiota; Insulin resistance; Blautia; Gastric inhibitory peptide; Glucagon-like peptide 1

Core Tip: Duodenum is a particular metabolic signaling center, and the thickened mucous membranes cause duodenal dysfunction and promote insulin resistance. This study explored the mechanisms by which duodenal mucosal resurfacing (DMR) affects type 2 diabetes (T2D) using a rat model. It highlights the potential role of the Blautia genus in the pathogenesis of T2D and the therapeutic effect of DMR. The results provide a theoretical basis for performing DMR in humans with T2D and identify several areas requiring further research.

INTRODUCTION

Influenced by genetics, environmental factors, stress, diet, lifestyle, and various other elements, the incidence of type 2 diabetes (T2D) is on the rise, with a projected global prevalence expected to reach 12.2% by 2045[1,2]. In individuals with T2D, damage to tissues and organs increases the incidence of complications, leading to increased morbidity and mortality and affecting both longevity and quality of life. Despite the availability of lifestyle interventions and advanced medical treatment for T2D, over half of affected individuals do not attain the treatment goal of achieving a glycemic hemoglobin level ≤ 53 mmol/mol[3]. Furthermore, current T2D treatment requires daily or weekly active interventions owing to the deterioration of β-cell function and the recurrence of hyperglycemia on treatment discontinuation[4]. Therefore, disease-modifying therapeutic interventions that can revolutionize the management of T2D are urgently needed.

The pathogenesis of T2D is intricate and multifaceted, with insulin resistance serving as the central mechanism[5]. Furthermore, the contribution of the stomach and small intestine to T2D pathogenesis is being increasingly recognized, contributing to the emergence of metabolic surgery as a viable option[6,7]. Recent research has underscored the duodenum as a pivotal metabolic signaling center, where duodenal dysfunction emerges as a driving factor for insulin resistance in T2D patients, particularly linked to abnormal hyperplasia of the duodenal mucosa[8-10]. To address this mucosal abnormality, researchers at Fractyl Health developed a duodenal mucosal resurfacing (DMR) procedure, a minimally invasive endoscopic hydrothermal ablation technique that can be performed in an outpatient setting[11,12]. This procedure involves the introduction of a DMR catheter through the oral cavity into the horizontal portion of the duodenum, followed by the injection of physiological saline into the submucosa to separate the mucosal and muscular layers. Subsequently, the catheter balloon is inflated adjacent to the mucosal layer, and hot water at approximately 80 °C is injected, causing coagulation necrosis and shedding (ablation) of the abnormal mucosal cells. Within days to months after ablation, the duodenal mucosa regenerates, yielding significant improvements in the glycemic and hepatic parameters in patients with T2D. A growing body of clinical research has consistently demonstrated the effectiveness of DMR for treating T2D[13,14]. Nonetheless, the precise role of duodenal mucosa in T2D pathogenesis and the mechanisms underpinning the metabolic improvements achieved by the DMR procedure remain enigmatic.

Considering the complexity of T2D pathogenesis, DMR might exert a regulatory effect on metabolic processes via multiple pathways. In this study, we applied the DMR procedure to diabetic rats and conducted an in-depth investigation of the potential underlying mechanisms. In our preliminary experiments, we found that the clinical endoscopic and hydrothermal ablation method used in humans is not suitable for use in diabetic rats owing to the narrow lumen and thin walls of the esophagus and intestine in rats. Based on a previous study[11], we designed a DMR catheter that uses mechanical wear to induce necrosis and detachment of the duodenal mucosa in diabetic rats to simulate hydrothermal ablation used in clinical practice. Our research revealed that mechanical wear, as well as hydrothermal ablation can significantly improve metabolic function in diabetic rats and we conducted a preliminary investigation into the underlying mechanisms.

MATERIALS AND METHODS

Animals and diet protocols

Six weeks old male Sprague-Dawley rats (200 ± 20 g) were purchased from Zhejiang Charies River Animal Technology Co., Ltd. [Jiaxing, China, permission No. SCXK (Zhejiang) 2019-0001]. The rats were housed in groups of three to four per cage under the specific pathogen free conditions (temperature: 22.5 ± 2.0 °C; relative humidity: 60% ± 10%; 12-hour light-dark cycle: 07:00-19:00). The rats were acclimated for 1 week before the experiment and then randomly assigned to normal control (NC) and T2D groups. The rats in the NC group (n = 12) consumed a standard rodent diet. The rats in the T2D group (n = 40) were fed a high-fat diet (HFD) (containing 45% fat, 35% carbohydrate, 20% protein; product code: XTHF45; Jiangsu Xietong Pharmaceutical Bio-engineering Co., Ltd., Nanjing, Jiangsu Province, China) for 4 weeks to induce insulin resistance. The diabetic state is then induced by intraperitoneal injection of streptozotocin (STZ) (35 mg/kg, Sigma, United States). Rats were considered to be diabetic if their fasting blood glucose (FBG) was ≥ 11.1 mmol/L at 3, 7, and 14 days after injection of STZ[15,16]. The diabetic rats were randomly arranged to a sham (n = 16) or DMR (n = 17) group. The entire experimental period was approximately 14 weeks.

All appropriate measures were taken to minimize pain or discomfort to the animals. The Ethics Committee of the Jiangsu Center for Safety Evaluation of Drugs approved the study (No. IACUC-20220613-01, approval date: June 13, 2022) and all animal experiments were conducted in accordance with the ARRIVE guidelines.

Surgical procedures

Endoscopic and hydrothermal ablation methods are not suitable for use in rats owing to the narrow lumen and thin walls of the esophagus and duodenum; therefore, we chose gastric incision and physical mechanical wear of the duodenal mucosa. Based on previous studies[11], we designed a DMR catheter with an adjustable polyethylene terephthalate mesh balloon at the distal end, which had a certain friction force (Figure 1A). During the procedure, the balloon size was adjusted according to the size and degree of wear of the duodenal lumen. The abrasion area of the duodenum was approximately 80% without muscle layer damage, as shown in Figure 1B.

Figure 1 Design and procedure of the duodenal mucosal resurfacing experiment. A: Duodenal mucosal resurfacing (DMR) catheter under normal and balloon expansion conditions; B: Hematoxylin and eosin staining of normal duodenum and duodenum after mucosal abrasion (scale bar = 1 mm); C: Diagram of the DMR procedure; D: Animal experiment flow chart of the DMR procedure. DMR: Duodenal mucosal resurfacing; NC: Normal control; STZ: Streptozotocin; T2D: Type 2 diabetes; PBS: Phosphate buffer solution.

Isoflurane (RWD Life Science Co., Ltd., Shenzhen, China) was administered to the rats after a 12-hour fast. The DMR procedure was initiated using a 4-cm mid line incision. Then, a longitudinal incision, approximately 0.5 cm long, was made in the greater curvature of the stomach and the DMR catheter was inserted into the duodenum. Once positioned 2 cm from the pylorus, the balloon was dilated, and mucosal wear was performed through antegrade sliding, with a wear length of 6 cm-8 cm (Figure 1C). The wounds were sutured and disinfected after the surgery. Postsurgical complications and survival rates of the laboratory animals were recorded. The rats in the sham group underwent laparotomy and gastrostomy without downward movement and mucosal abrasion. The surgical time, other procedures, and postoperative care were otherwise the same in the DMR and sham groups. After the surgery, glucose or insulin solutions were administered as needed, according to the blood glucose level of the rats. After waking and turning over, the rats drank water freely. After 24 hours, a standard rodent diet was provided until the end of the study. Within 3 days after surgery, procaine penicillin (Shandong Youxin Biotechnology Co., Ltd., Binzhou, China) was injected intramuscularly at a dose of 80000 U/kg. A survival time of 72 hours was considered as the criterion for successful surgery. The 72-hour survival rates after surgery were more than 70% for the rats in the T2D-sham and T2D-DMR groups. The advantages of this surgical method are that it is intuitive and does not require laparoscopy or any other equipment. The disadvantage is that it causes significant trauma to rats, has a high mortality rate, and causes surgical complications, such as abdominal adhesions.

Food intake, body weight, and FBG levels were checked weekly during study. All rats were fasted overnight and were euthanized at postoperative week 3 or 6 (six rats per group) (Figure 1D). Blood samples were taken through the abdominal aorta, and the serum was stored at -80 °C after centrifugation (4000 rpm/minute, 15 minutes). Duodenum, liver and colonic contents were collected for further analysis.

Oral glucose tolerance tests

The rats were administered a glucose solution (1 g/kg) by oral gavage after overnight fasting. Blood glucose levels were measured using a glucose analyzer [OGM-161, ACON Biotech (Hangzhou) Co., Ltd., Hangzhou, Zhejiang Province, China] at 0, 15, 30, 60, and 120 minutes after glucose administration, and the area under curve of oral glucose tolerance test (OGTT) (AUC_OGTT) was computed according to the data.

Enzyme-linked immunosorbent assays and biochemical assays

The insulin (Catalog CSB-E05070r), cholecystokinin (CCK) (Catalog CSB-E08114r), gastric inhibitory peptide (GIP) (Catalog CSB-E17969r), lipopolysaccharide (LPS) (Catalog CSB-E14247r), and glucagon-like peptide 1 (GLP-1) (Catalog CSB-E08117r) enzyme-linked immunosorbent assay kits were bought from Wuhan Huamei Biotech Co., Ltd. (Wuhan, Hubei Province, China). Following the instructions, these kits were used to measure the serum levels of CCK, GIP, LPS, GLP-1, and insulin at room temperature. Serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), total cholesterol (TC), triglyceride (TG), high-density lipoprotein cholesterol (HDL-C), and low-density lipoprotein cholesterol (LDL-C) were measured using a Dimension EXL200 fully automatic biochemical meter (Siemens AG, Berlin, Germany).

Histology

The liver and duodenal segments preserved in formalin were buried with paraffin, and sectioned with 5-μm-thick. Histological changes in the duodenum were assessed using hematoxylin and eosin (HE) staining. Images were collected using the Olympus CX21 optical binocular microscope (Olympus, Tokyo, Japan) equipped with a 5 × or 40 × objective lens. The length of the duodenal villi and depth of the crypts were measured using Aperio ImageScope software. For each rat (n = 5), ten intact villi and crypts were selected for measurement. The length of villi and depth of crypts were measured using a previously reported method and the mean values were calculated for each rat[17]. The villi were measured from the apex to the villus-crypt junction, while the crypts were measured from the base to the crypt-villus junction.

Western blot

For protein extraction, the duodenal tissue was treated with RIPA buffer containing protease and phosphatase inhibitors. The protein concentration was checked by employing Bradford protein assay kit (Thermo Fisher, Catalog 23227). Then, the protein extract solution was added to the loading buffer (Biorbyt, Catalog orb90545) and boiled for denaturation. The proteins were electrophoretically separated on a BeyoGel plus PAGE prefabricated gel (Beyotime, 4%-15%, Catalog P0519S) and electro-imprinted on a polyvinylidene difluoride membrane (PALL, Catalog 65421). The 5% milk solution was used to block the membrane. The membrane was incubated overnight at 4 °C in a solution of occludin (1:1000, Abcam, Catalog ab216327), zonula occludens-1 (ZO-1) (1:5000, Santa Cruz, Catalog sc-33725) and β-actin (1:5000, Abcam, Catalog ab8226) antibody. The membranes were then incubated with horseradish peroxidase-conjugated secondary anti-rabbit antibodies (1:5000, Biosharp, Catalog BL003A). The blots were visualized using enhanced chemiluminescence reagent (Biosharp, Catalog BL523A). The Image J computer software analysis system was employed to normalize the β-actin signal and conduct semi-quantitative analysis.

Immunofluorescence and immunohistochemistry

For immunofluorescence staining, serial duodenal sections were incubated with antibodies against occludin (1:150; Abcam, Catalog ab216327) and ZO-1 (1:50; Santa Cruz Biotechnology, Catalog sc-33725), followed by the appropriate secondary antibody. The nuclei were counterdyed with 4’,6-diamidino-2-phenylindole (Sigma-Aldrich, Catalog MBD0020) at the end of the immunofluorescence procedure. Immunofluorescence images were obtained and analyzed using Image J software. Likewise, for immunohistochemically stained duodenal slices were incubated with antibodies against CCK (1:200, Absin, Catalog abs138035), GIP (1:100, Abcam, Catalog ab209792), and GLP-1 (1:2000, Starter, Catalog S0B0387).

Microbiota and metabolomics analysis

Colonic content samples were collected from the NC, T2D-sham, and T2D-DMR rats at postoperative week 6 were for 16S rRNA gene sequencing. The extraction, library preparation, sequencing, and analysis were performed by Metabo-Profile Biotechnology Co., Ltd. (Shanghai, China). Targeted metabolomic analyses of the colonic contents were performed using the Q300 metabolite analysis kit (Human Metabolomics Institute, Inc., Shenzhen, Guangdong Province, China) based on a previously published method, with modifications[18]. The specific experimental details were as described previously[19].

Statistical analysis

Statistical analysis was finished employing statistical product and service solutions 21.0 Version. All quantitative data were shown as mean ± SE. AUC_OGTT were computed by trapezoidal integration. Comparisons between more than two groups were performed using one-way analysis of variance (ANOVA). For comparison between two groups, we have used the Student’s t test (two-tailed). The correlations between gut microbiota, metabolites and T2D-related traits were analyzed employing Spearman’s correlation analysis. P < 0.05 was thought to be statistically meaning.

RESULTS

The DMR procedure ameliorates metabolic disorders in T2D rats

The DMR and sham surgeries were performed on the T2D rats according to the study protocol. The post-surgery body weight and food intake were comparable between the DMR and sham groups (Figure 2A and B). Both the DMR and sham groups exhibited a decline in their FBG levels on day 3 after surgery. Subsequently, the FBG levels in the sham group continued to escalate, whereas those in the DMR group decreased. From postoperative week 2, the FBG levels of T2D rats were significantly decreased by DMR procedure (all P < 0.05; Figure 2C). Furthermore, DMR procedure improved glucose tolerance at postoperative weeks 3 and 6, as evidenced by the lower values of AUC_OGTT (all P < 0.05; Figure 2D and E). Additionally, the DMR procedure led to increased serum insulin levels at postoperative week 6 (P < 0.01; Table 1). Collectively, these findings indicate that the DMR procedure significantly ameliorated metabolic disorders in blood glucose in T2D rats.

Figure 2 Effects of the duodenal mucosal resurfacing procedure on phenotypic changes in type 2 diabetes rats. A-C: Body weight, food intake and fasting blood glucose throughout the experimental period after intraperitoneal injection of streptozotocin (n = 6 or 12 mice per group); D and E: Glucose excursion curves and the area under curve of oral glucose tolerance tests performed at postoperative weeks 3 (n = 12 mice per group) and 6 (n = 6 mice per group). Data are presented as mean ± SE. ¹P < 0.01 vs normal control group. ²P < 0.001 vs normal control group. ³P < 0.01 vs type 2 diabetes-sham group. ⁴P < 0.001 vs type 2 diabetes-sham group. DMR: Duodenal mucosal resurfacing; T2D: Type 2 diabetes; NC: Normal control; STZ: Streptozotocin; OGTT: Oral glucose tolerance test; AUC: Area under curve.

Table 1 Effects of the duodenal mucosal resurfacing procedure on glucose, lipid metabolism parameters, and intestinal hormones in type 2 diabetes rats, mean ± SE (n = 5-6 rats per group).

Variable	Preoperative			Postoperative week 3			Postoperative week 6
	NC	T2D + sham	T2D + DMR	NC	T2D + sham	T2D + DMR	NC	T2D + sham	T2D + DMR
Insulin (μIU/mL)	18.86 ± 1.34	13.24 ± 1.082	13.78 ± 1.361	19.49 ± 1.15	14.62 ± 1.501	15.79 ± 1.07	19.67 ± 1.02	12.76 ± 1.312	18.07 ± 1.045
HDL-C (mmol/L)	1.39 ± 0.26	0.60 ± 0.132	0.59 ± 0.102	1.54 ± 0.25	0.66 ± 0.132	0.88 ± 0.152	1.52 ± 0.12	0.66 ± 0.162	1.03 ± 0.21
LDL-C (mmol/L)	0.96 ± 0.15	1.71 ± 0.231	1.67 ± 0.221	1.09 ± 0.16	1.68 ± 0.26	1.37 ± 0.20	1.06 ± 0.15	1.67 ± 0.231	1.10 ± 0.114
TG (mmol/L)	1.45 ± 0.19	2.15 ± 0.231	2.23 ± 0.261	1.60 ± 0.17	2.05 ± 0.21	1.87 ± 0.22	1.67 ± 0.19	2.37 ± 0.231	1.78 ± 0.144
TC (mmol/L)	1.67 ± 0.19	2.43 ± 0.211	2.45 ± 0.331	1.47 ± 0.25	2.39 ± 0.261	2.12 ± 0.25	1.76 ± 0.25	2.54 ± 0.141	1.74 ± 0.244
FFA (mmol/L)	0.51 ± 0.13	1.41 ± 0.212	1.42 ± 0.222	0.52 ± 0.06	1.30 ± 0.232	1.11 ± 0.101	0.51 ± 0.08	1.39 ± 0.282	0.98 ± 0.15
AST (U/L)	234.00 ± 28.21	314.50 ± 20.941	316.00 ± 21.491	222.17 ± 33.79	315.67 ± 28.901	304.67 ± 21.67	238.5 ± 24.66	325.17 ± 26.711	262.67 ± 19.49
ALT (U/L)	89.33 ± 14.19	169.17 ± 22.511	183.17 ± 25.082	88.67 ± 11.95	179.83 ± 20.492	156.33 ± 15.01	90.50 ± 12.86	181.33 ± 28.141	141.50 ± 21.61
GLP-1 (pg/mL)	2.37 ± 0.24	1.50 ± 0.171	1.48 ± 0.251	2.67 ± 0.15	1.62 ± 0.152	1.70 ± 0.222	2.63 ± 0.37	1.67 ± 0.151	2.30 ± 0.234
GIP (pg/mL)	103.36 ± 14.26	39.77 ± 4.553	40.90 ± 5.073	111.47 ± 17.48	40.34 ± 6.182	62.90 ± 7.891	97.89 ± 14.54	38.27 ± 7.262	72.74 ± 9.974
CCK (pg/mL)	26.61 ± 2.86	13.61 ± 2.432	15.52 ± 2.372	28.20 ± 3.21	15.02 ± 1.642	20.67 ± 1.931	28.32 ± 4.75	12.85 ± 2.561	22.92 ± 2.994

¹P < 0.05 vs normal control group.

²P < 0.01 vs normal control group.

³P < 0.001 vs normal control group.

⁴P < 0.05 vs type 2 diabetes-sham group.

⁵P < 0.01 vs type 2 diabetes-sham group.

One-way analysis of variance. DMR: Duodenal mucosal resurfacing; NC: Normal control; T2D: Type 2 diabetes; HDL-C: High-density lipoprotein cholesterol; LDL-C: Low-density lipoprotein cholesterol; TG: Triglyceride; TC: Total cholesterol; FFA: Free fatty acids; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; GIP: Gastric inhibitory peptide; GLP-1: Glucagon-like peptide 1; CCK: Cholecystokinin.

After consuming a HFD and STZ intraperitoneal injection, the serum LDL-C, TC, TG, free fatty acids, AST, and ALT levels increased significantly and HDL-C levels decreased (all P < 0.05; Table 1). These metabolic changes improved after the DMR procedure. For instance, the serum LDL-C and TC levels of T2D rats in the DMR group were signally decreased at postoperative week 6 (all P < 0.05; Table 1). Additionally, HE staining revealed that the DMR procedure reduced lipid deposition in the livers of T2D rats (Figure 3A). Collectively, these results highlight the ability of the DMR procedure to improve lipid metabolism in T2D rats.

Figure 3 Effects of duodenal mucosal resurfacing procedure on liver, enteroendocrine cells, duodenal villi, and crypts in type 2 diabetes rats. A: Hematoxylin and eosin (HE) staining of the liver in normal control (NC), type 2 diabetes (T2D) + sham and T2D + duodenal mucosal resurfacing (DMR) rats at postoperative weeks 3 and 6 (n = 5 mice per group, Scale bar = 50 μm); B: Immunohistochemical analyses of duodenal cholecystokinin, gastric inhibitory peptide, glucagon-like peptide 1 in NC, T2D + sham and T2D + DMR rats at postoperative weeks 6 (n = 4 mice per group, Scale bar = 50 μm); C: HE staining of the duodenum in NC, T2D + sham and T2D + DMR rats at postoperative weeks 3 and 6 (n = 5 mice per group, Scale bar = 500 μm); D and E: The villus length and crypt depth in NC, T2D + sham and T2D + DMR rats at postoperative weeks 3 and 6 (n = 5 mice per group). Data are presented as mean ± SE. ¹P < 0.01 vs normal control group. ²P < 0.001 vs normal control group. ³P < 0.001 vs type 2 diabetes-sham group. DMR: Duodenal mucosal resurfacing; T2D: Type 2 diabetes; NC: Normal control; CCK: Cholecystokinin; GIP: Gastric inhibitory peptide; GLP-1: Glucagon-like peptide 1.

The DMR procedure increases intestinal hormone secretion in T2D rats

Intestinal hormones play an important role in metabolic surgery[20]. The serum levels of GLP-1, GIP, and CCK decreased significantly after HFD feeding and STZ intraperitoneal injection. However, DMR procedure significantly increased the levels of these hormones (all P < 0.05; Table 1). Additionally, DMR procedure increased the number of CCK-positive cells and GLP-1-positive cells at postoperative week 6 (Figure 3B).

DMR procedure decreases villus length and crypt depth of duodenum in T2D rats

As shown in Figure 3C-E, histological findings showed increased duodenal villus length and crypt depth in diabetes rats. However, DMR procedure reversed these changes. The duodenum is an important part for the digestion and absorption of nutrients. These results suggest that DMR procedure may lessen nutrient absorption by reducing villus length and crypt depth, thereby improving glucose and lipid parameters.

DMR enhances the duodenal mucosal barrier function and decreases LPS leakage in T2D rats

The expression levels of the tight junction proteins in the duodenum were estimated using western blot and immunofluorescence. Western blot indicated that occludin and ZO-1 expression was increased by the DMR procedure (all P < 0.05; Figure 4A-C). Immunofluorescence showed similar results (all P < 0.05; Figure 4D-F). In addition, the serum LPS levels of rats in the T2D-sham group were increased, and the DMR procedure reduced serum LPS levels at postoperative week 6 (P < 0.05; Figure 4G). These results indicate that the DMR procedure significantly enhanced duodenal mucosal barrier.

Figure 4 Effects of the duodenal mucosal resurfacing procedure on the duodenal mucosal barrier function in type 2 diabetes rats. A: Protein extracts of duodenum were subjected to western blot analysis with antibodies to occludin, zonula occludens-1 (ZO-1) or β-actin; B and C: The blots of occludin and ZO-1 were quantified using image J software and normalized to β-actin levels (n = 3 mice per group); D: Representative images of immunofluorescence staining for occludin (green), ZO-1 (red) and 4’,6-diamidino-2-phenylindole (blue) of duodenum (Scale bar = 50 μm); E and F: The relative fluorescence intensity of occludin and ZO-1 (n = 5 mice per group); G: Serum lipopolysaccharide levels of rats at postoperative weeks 3 and 6 (n = 6 mice per group). Data are presented as mean ± SE. ¹P < 0.05 vs normal control group. ²P < 0.01 vs normal control group. ³P < 0.001 vs normal control group. ⁴P < 0.05 vs type 2 diabetes-sham group. ⁵P < 0.01 vs type 2 diabetes-sham group. ⁶P < 0.001 vs type 2 diabetes-sham group. DMR: Duodenal mucosal resurfacing; T2D: Type 2 diabetes; NC: Normal control; LPS: Lipopolysaccharide; ZO-1: Zonula occludens-1; DAPI: 4’,6-diamidino-2-phenylindole.

DMR modifies the gut microbiome and metabolites in T2D rats

High-throughput sequencing of the 16S rRNA gene was performed to study the effects of DMR on the microbiota profiles of colonic content samples from T2D rats. Specific analysis of α-diversity showed that, there were no marked discrepancies in microbiome richness between groups, but multiformity and evenness differed markedly between groups (Figure 5A-C). Moreover, principal coordinates analysis of β-diversity showed differences in the composition of gut microbiota among the NC, T2D-sham, and T2D-DMR groups (Figure 5D). At the phylum level, Firmicutes, Bacteroidetes, Actinobacteria and Proteobacteria were predominant in all groups, with the proportion exceeding 96% (Supplementary Figure 1A). The relative abundance of Firmicutes was increased in the T2D-sham group. However, the relative abundance of Bacteroidetes and Proteobacteria was decreased in the T2D-sham group. At the genus level, the relative abundance of some genera were disparate among the three groups. For example, the relative abundance of Blautia was merely 1.21% in the NC group, which increased to 7.29% in the T2D-sham group, and recovered to 3.84% in the T2D-DMR group (Supplementary Figure 1B).

Figure 5 Effects of the duodenal mucosal resurfacing procedure on gut microbiota in type 2 diabetes rats. A-C: Analysis of the alpha-diversity values including Chao1, Simpson, and Pielou’s evenness index; D: Principal coordinates analysis of β-diversity; E and F: Graphics of linear discriminant analysis effect size confirmed the most variously abundant taxa of gut microbiota in the normal control, type 2 diabetes-sham, and type 2 diabetes-duodenal mucosal resurfacing groups (n = 5-6 mice per group). DMR: Duodenal mucosal resurfacing; NC: Normal control; T2D: Type 2 diabetes; LDA: Linear discriminant analysis.

In addition, linear discriminant analysis effect size analysis revealed that the intestinal microbiota underwent significant changes at different taxonomic levels. The DMR procedure reversed the changes in abundance of Proteobacteria in the colonic contents at the phylum level and changes in the abundance of Blautia, Jeotgaleliccus, Rothia, Turicibacter, Phascolarctobacter, Staphylococcus and Actinomyces at the genus level (Figure 5E and F). Mapping the gene families of these 16S rRNA characteristic sequences to the Kyoto encyclopedia of genes and genomes database to predict related pathways revealed that they were mainly associated with the metabolic pathways of carbohydrates and amino acids (Supplementary Figure 2). In summary, these results indicate that the DMR procedure regulates the gut microbiota in T2D rats.

Ultra-high performance liquid chromatography with quadrupole time-of-flight mass spectrometry was used to profile the metabolites across the three groups. The total amounts of amino acids, bile acids, pyridines, benzenoids, peptides and carnitines differed significantly between the T2D-DMR, T2D-sham, and NC groups (Figure 6A). Principal component analysis revealed that each group of samples was distinctly arranged into disparate blocks, indicating that T2D and the DMR procedure generated distinct metabolite profiles (Figure 6B). Candidate biomarkers were confirmed employing ANOVA and the Kruskal-Wallis test. 87 altered metabolites with variable importance in projection > 1 and P < 0.05 were chosen. Compared with NC group, a total of 14 metabolites were increased, and 73 metabolites were reduced in the T2D-sham group. Further analysis revealed that the DMR procedure reversed the changes in 25 metabolites (Supplementary Table 1). Pathway-associated metabolite sets database and the rno databank were used for pathway enrichment analysis of the differential metabolites (Figure 6C and D). The results showed that the metabolism pathway affected by the DMR procedure might be involved with β oxidation and biosynthesis of fatty acids, biosynthesis and metabolism of amino acids, the uric acid cycle, and the glucose-alanine cycle.

Figure 6 Metabolites and prospective metabolic pathways related to type 2 diabetes and the duodenal mucosal resurfacing procedure. A: Stacked bar chart of relative abundance statistics for the median values of various metabolites in the colonic contents of rats; B: Two-dimensional principal component analysis score map of colonic contents samples and corresponding box plot of principal component analysis scores; C: Bar chart of small molecule pathway database-associated metabolite sets used in the pathway enrichment analysis; D: Bubble chart of the rno library pathway analysis. ¹P < 0.05. ²P < 0.01. ³P < 0.001. DMR: Duodenal mucosal resurfacing; NC: Normal control; SCFA: Short-chain fatty acid; T2D: Type 2 diabetes; BAs: Bile Acids; PC1: Principal component 1; PC2: Principal component 2; TCA: Taurocholic acid.

Correlations between gut microbiota, metabolites, and traits associated with T2D

Significant correlations were observed between T2D-related traits, gut microbiota, and metabolites. Specifically, the abundance of the Blautia genus showed positive association with FBG, LPS, and TC levels, while displaying negative correlation with GIP and insulin levels. In contrast, the abundance of the Jeotgalelicus, Rothia, Phascolarctobacter, and Actinomyces genera exhibited negative correlation with FBG, LPS or TC levels, while displaying positive correlation with GIP or insulin levels (Figure 7A). Similarly, the majority of the 25 metabolites influenced by the DMR procedure were correlated with glucose and lipid parameters, as well as LPS and intestinal hormones (Figure 7B). Further analysis revealed that levels of the Blautia and Turiciactor genera were positively correlated with maltose/lactose, maltotriose, and 3-methyladipic acid levels, and negatively correlated with ketoleucane levels (Figure 7C). Conversely, the levels of Phascolarctobacterum, Rothia, Jeotgallicus genera were negatively correlated with maltose/lactose, maltotriose, and 3-methyladipic acid, and positively correlated with ketoleucane levels (Figure 7C).

Figure 7 Heat map of the Spearman rank-correlation coefficients between differential microbes, metabolites and type 2 diabetes-related traits. A: Spearman rank-correlation coefficients between characteristic microbes and type 2 diabetes (T2D)-related traits at the genus level; B: Spearman rank-correlation coefficients between metabolites and T2D-related traits; C: Spearman rank-correlation coefficients between characteristic microbes and metabolites at the genus level. DMR: Duodenal mucosal resurfacing; FBG: Fasting blood glucose; GIP: Gastric inhibitory peptide; GLP-1: Glucagon-like peptide 1; LDL-C: Low-density lipoprotein cholesterol; LPS: Lipopolysaccharide; NC: Normal control; SCFA: Short-chain fatty acid; T2D: Type 2 diabetes; TC: Total cholesterol; DHA: Docosahexaenoic acid; DPA: Docosapentaenoic acid; GHDCA: Glycohyodeoxycholate; LCA: Lithocholic acid; TCDCA: Taurochenodeoxycholate; THDCA: Taurohyodeoxycholic acid; UDCA: Ursodeoxycholic acid; HDCA: α-hyodeoxycholic acid; CA: Cholic acid; UCA: Ursocholic acid.

DISCUSSION

In this study, we successfully performed DMR surgery in T2D rats. Although our mechanical abrasion method is different from the clinical hydrothermal ablation, both methods significantly improved diabetes control, confirming the importance of the duodenal mucosa in the pathogenesis and treatment of T2D. Our findings strongly suggest that the DMR-induced improvements in glucose and lipid metabolism in T2D rats involved multiple pathways.

Numerous studies have reported that the levels of gastrointestinal hormones such as GLP-1 and CCK increase significantly after metabolic surgery[21]. The upregulation of endogenous GLP-1 signaling is considered the primary pathway for enhancing glucose metabolism after metabolic surgery[22,23]. This study also found that the amount of GLP-1-positive cells in the duodenum increased after DMR. CCK plays a role in satiety stimulation and insulin secretion[24,25], contributing to metabolic control. The DMR procedure shares similarities with duodenal-jejunal bypass sleeve (DJBS) surgery in that it reduces nutrient absorption in the duodenum. Speck et al[26] found an increase in serum GIP levels after DJBS surgery, which partially explains the elevated serum GIP levels following the DMR procedure. However, the effects of endogenous GIP after surgical metabolic interventions are not fully understood. Endogenous GLP and GIP might function synergistically to yield greater metabolic benefits. The DMR procedure had no impact on the expression level of GIP-positive cells in the duodenum. The increase in serum GIP and GLP-1 levels may be related to a rise in the amount of GIP or GLP-1 positive cells in other intestinal segments or the DMR procedure may directly stimulate the secretion of GIP and GLP-1. However, further research is needed to verify these phenomena.

Previous studies have revealed that high-fat and high-sugar diets can induce duodenal mucosal hyperplasia in rodents, triggering insulin resistance signaling and contributing to metabolic diseases[9]. This concept underpins the design of the DMR procedure[8]. Additionally, diabetes can promote mucosal hyperplasia, unrelated to food intake. This mucosal thickening is primarily attributable to the elongation of the intestinal villi and deepening of the crypts[27,28]. This increases the intestinal surface area, enhancing digestion and absorption in rats. The potential mechanism whereby the DMR procedure improves the control of T2D may involve reducing nutrient absorption by modulating villus length and crypt depth.

Dysfunction in the gut barrier, gut microbiota, and metabolic products is closely associated with T2D[29-31]. Numerous studies have provided evidence that the gut microbiota produce a effect in T2D. However, specific taxonomic groups associated with T2D vary considerably between studies[29]. In our research, we primarily observed significant correlations between certain bacteria and T2D-related parameters at the genus level. Among these, Blautia were the most strongly correlated with T2D-related parameters. As a genus within the Lachnospiraceae family, several human and rodent studies have shown an association between Blautia and T2D, although these associations have varied between studies. Some studies have suggested that Blautia, as an anaerobic probiotic, is negatively associated with metabolic disorders[32], and that oral administration of Blautia wexlerae ameliorates obesity and T2D by altering the gut microbiota[33]. Conversely, some clinical studies have shown that there is no correlation between Blautia and T2D[34]. Nevertheless, most clinical and animal studies have reported an increase in the abundance of Blautia in T2D or prediabetic individuals[35-38], with decreases observed after metabolic surgery[39,40]. Anhê et al[40] found that following bariatric surgery gut microbiota, specifically higher Parabacteroides and lower Blautia abundance, improved blood glucose independent of changes in obesity, insulin, or insulin resistance. Consistent with most studies, this study demonstrated an increase in the Blautia genus in T2D rats, which significantly decreased after the DMR procedure. Further investigation revealed that the level of the Blautia genus was strongly correlated with T2D-related traits. Additionally, Leclercq et al[41] found that an increase in Blautia abundance was linked to high intestinal permeability. Blautia may increase the intestinal barrier permeability, promote LPS leakage, trigger insulin resistance, and consequently elevate blood glucose levels. Based on the correlation between Blautia and metabolites, Blautia may participate primarily in amino acid biosynthesis and metabolism. However, the precise role of Blautia after the DMR procedure warrants further investigation. In addition, we only studied the impact of DMR on the gut microbiota in the colon. Future research should include an investigation of changes in the gut microbiota in duodenum.

The pathogenesis of T2D is multifaceted, and DMR improves metabolic disorder through multiple interconnected pathways[10]. The changes induced by the DMR procedure interact with each other. This study suggests that the DMR procedure initially reduce villus length and crypt depth, thereby decreasing nutrient absorption. This may lead to improvements in the gut microbiota and gut hormone secretion, and a reduction in LPS production. These alterations in the gut microbiota, and metabolite and gut hormone levels may regulate the proliferation and apoptosis of intestinal epithelial cells, enhance barrier function, reduce LPS leakage, reduce inflammation, and ultimately reduce insulin resistance, thereby enhancing glucose and lipid metabolism (Figure 8).

Figure 8 Schematic representation of the mechanisms by which the duodenal mucosal resurfacing procedure improves the control of metabolic disorders. This diagram illustrates the hypothesized mechanisms through which duodenal mucosal resurfacing procedure improves blood glucose and liver metabolic disorders in a type 2 diabetes rat model. The black solid line arrows indicate promotion effects; the black dashed line arrows indicate potential promotion effects; the black inverted arrows indicate the content or activity decrease; the blue forward arrows indicate the content or activity increase; and the T-shaped solid line without an arrow indicates inhibition. CCK: Cholecystokinin; DMR: Duodenal mucosal resurfacing; GIP: Gastric inhibitory peptide; GLP-1: Glucagon-like peptide 1; LPS: Lipopolysaccharide.

This study has several limitations. First, we only conducted short-term postoperative observation. In the future, long-term changes in glucose and lipid metabolism and mucosal changes should be investigated. Second, this was a preliminary observational study without in-depth exploration of the mechanisms. In-depth research on the role of Blautia in the pathogenesis and treatment of T2D, as well as the interplay between the microbiota, villi, and intestinal hormones is needed. Third, the surgical methods used in the animal experiments differed from those of the DMR procedure used in clinical trials. Clinical DMR is a minimally invasive endoscopic procedure that causes minimal trauma in humans. We performed laparotomy, which causes significant trauma in T2D rats. To eliminate interference from surgery, we used a sham-surgery group as a control. In addition, hot water has been used to damage the mucosal layer in clinical trials, whereas we used mechanical force in this study. Although the methods are different, the common goal was to cause necrosis and shedding of mucosal cells without damaging the muscle layer.

CONCLUSION

Collectively, our findings strongly indicate that the duodenal mucosa is important in the pathogenesis and treatment of T2D, and that DMR may effectively alleviate insulin resistance and metabolic disorders in T2D rats through a multitude of pathways. These mechanisms may include a reduction of villus length and crypt depth, augmentation of intestinal hormone secretion, reinforcement of intestinal barrier function, reduction in LPS leakage, and modulation of gut microbiota and metabolites. This study serves as a preliminary investigation and further animal experiments are required to corroborate these findings. Nevertheless, our research provides theoretical groundwork for the potential clinical application of the DMR procedure in humans with T2D.

ACKNOWLEDGEMENTS

The authors wish to thank Dr. Wang HP for providing technical guidance with the animal surgery, and Mr. Han DZ and Nanjing Kangyou Medical Technology Co., Ltd. for designing the DMR catheters.