The Effect of Bifidobacterium animalis subsp. lactis MN-Gup on Glucose Metabolism, Gut Microbiota, and Their Metabolites in Type 2 Diabetic Mice

Thirty 5-week-old C57BL/6J male mice, bred under Specific Pathogen-Free (SPF) conditions, were procured from Beijing Weitong Lihua Laboratory Animal Technology Co. Ltd. (Beijing, China). They were then housed in a laboratory animal facility with SPF-grade conditions at Pony Testing International Group Co., Ltd. (Beijing, China). The mice were maintained under a standard light/dark cycle (12 h light/12 h dark) with ad libitum access to both water and food. This study was conducted following approval from the institutional animal ethics committee (approval number: PONY-2023-FL-21), ensuring compliance with scientific standards and ethical guidelines. After a week-long acclimatization period, each mouse was assigned a unique identifier. These identifiers were then inputted into a randomization software program (R version 4.4.0), which allocated the animals randomly to their respective treatment groups. Group assignments were performed by a researcher (JH), who was not involved in the subsequent stages of the experiment, and during the allocation phase, the experimenters were kept blinded. Each group consisted of six mice, resulting in a total of five groups (Figure 1). Each cage, housing six male mice, served as the experimental unit, with treatments randomized and applied at the cage level to ensure that all animals within a cage received the same treatment. To mitigate potential interactions, the cages were adequately spaced to minimize visual and auditory interactions, and regular cleaning and disinfection protocols were followed to maintain the SPF environment. One group of mice was randomly selected as the normal control (NC) group and injected with a sodium citrate buffer (0.05 M, pH = 4.5) on the last day of the sixth week. The remaining four groups of mice were fed with high-fat diets for five weeks and then injected with streptozotocin (STZ, 35 mg/kg, dissolved in 0.05 M sodium citrate buffer, pH = 4.5) through intraperitoneal injection on the same day. Mice exhibiting fasting blood glucose (FBG) levels exceeding 11.1 mmol/L after 3 days of injection were classified as animals with T2DM [12].

The probiotic solid beverage products were produced by Sanhe Fucheng Biological Technology Co. (Langfang, China). The NC group continued to be fed a normal diet and physiological saline. The T2DM mice were divided into four groups, including the diabetic model control group (MC, fed a high-fat diet and physiological saline, serving as the control group), the positive control group (MF, fed a high-fat diet and 120 mg/kg/d metformin), the low-dose MN-Gup group (L-MG, fed a high-fat diet and 2 × 10⁹ CFU/kg MN-Gup), and the high-dose MN-Gup group (H-MG, fed a high-fat diet and 1 × 10¹⁰ CFU/kg MN-Gup). The body weight of mice was recorded weekly. By the end of the 13th week of intervention, feces were collected following the final gavage. The mice were euthanized the next day, blood samples were drawn, and the serum was prepared. Throughout the experiment, the behavior and health of each animal were closely monitored and recorded to ensure the reliability of the results, account for any potential interactions, detect signs of pain or distress, and establish criteria for humane endpoints to minimize animal sufferings and distress.

The FBG of mice was measured at weeks 7 and 13. Before the FBG testing, the mice fasted for 8 h. The tail was disinfected with an alcohol swab, and a small incision was made at the tip of the tail vein. The first drop of blood was discarded, and the second drop was used for glucose measurement with a glucometer (Roche Diagnostics, Mannheim, Germany). Simultaneously, blood samples were collected for subsequent indicator testing. The blood samples were centrifuged at 3500 rpm for 15 min, and insulin (Uscn life, Wuhan, China) and GLP-1 (Mesoscale, Rockville, MD, USA) were then detected using ELISA kits. Homeostasis model assessment-insulin resistance (HOMA-IR) was calculated by the following formula [13].

The glucose tolerance of all mice was measured by a glucose meter after modeling (week 7) and before death (week 13), too. Mice were given 2 g/kg·BW glucose solution (40% w/v) intragastrically after fasting for 16 h. We determined the blood glucose levels in the tail vein blood of mice at 0, 30, 60, 90, and 120 min after gastric irrigation. The results of the Oral Glucose Tolerance Test (OGTT) were quantified as the area under the curve (AUC), which was calculated using GraphPad Prism 9.0 software (San Diego, CA, USA).

The colonic and pancreatic tissues of mice were fixed in 4% paraformaldehyde for 24 h. Following fixation, the tissues underwent dehydration, paraffin embedding, and sectioning to obtain thin slices, suitable for microscopic examination. Hematoxylin and eosin (H & E) staining were applied to the sections. Finally, the stained slices were meticulously examined under an optical microscope (DMi8, Leica, Weztlar, Germany) to capture images and analyze histopathological changes.

Shanghai Majorbio Bio-pharm Technology Co., Ltd. (Shanghai, China) performed the gut microbiota analysis of fresh feces. The DNA of collected colonic content samples was extracted using E.Z.N.A. DNA Kit. Primers 338F (5′-ACTCCTACGGGAGGCAGCAG-3′) and 806R (5′-GGACTACHVGGGTWTCTAAT-3′) were used for PCR amplification of the V3–V4 region of bacterial 16S ribosomal RNA gene. The PCR product was retrieved from a 2% agarose gel and purified using the AxyPrep DNA Gel Extraction Kit from Axygen Biosciences (Union City, CA, USA). Following purification, quantification was carried out utilizing the Quantus™ Fluorometer, manufactured by Promega (Madison, WI, USA). Samples were sequenced on an Illumina MiSeq PE300 platform (Illumina, San Diego, CA, USA). The optimized sequence was obtained by splicing and quality control of sequencing data. Optimizing sequence: 16S rRNA sequences were classified into operational taxonomic units (OTUs), with a threshold of 97% identity, with Uparse (version 7.0.1090).

SCFAs were detected using gas chromatography. Add 50 μL of 50% sulfuric acid to 20 mg of fecal sample, and homogenize the sample for 3 min at 4 °C. Then add 10 μL of internal standard (2.927 mmol/L 2-ethylbutyric acid) and 500 μL of ether, and homogenize for 1 min. Centrifuge at 4600× g for 5 min at 4 °C, and collect 350 μL of the supernatant. Repeat this step. Add 500 μL of ether to the pellet, homogenize, and after centrifugation, collect 400 μL of the supernatant. Finally, filter the combined supernatants through a 0.22 μm filter membrane and inject them into brown sample vials for analysis. The specific experimental conditions for gas chromatography were established based on the method developed by Zhang et al. [14].

As shown in Figure 2a, the weight gain of the mice fed high-fat diets was significantly (p < 0.05) higher than that of the mice fed normal diets at week 7. Then, STZ was injected at the end of the seventh week. The FBG of the model group was significantly increased compared to that of the NC group (p < 0.05), and the FBG was higher than 11.1 mmol/L. This suggested that a mouse model of T2DM was successfully established in the current study. Subsequently, different intervention agents were administered to the corresponding groups, leading to significant alterations in the glucose metabolism in T2DM mice.

As seen in Figure 2b, there were no significant differences observed in the FBG levels among the groups of T2DM mice before the intervention. As the intervention progressed, both the low-dose and high-dose MN-Gup intervention groups (L-MG group and H-MG group) exhibited a notable decrease in the FBG compared to the MC group (p < 0.05). Remarkably, similar trends were also observed in the area under the curve (AUC) of the OGTT (Figure 2c), incremental AUC (Figure 2d) and HOMA-IR (Figure 2e). Following the intervention, the MN-Gup intervention groups showed a significant reduction in the AUC of the OGTT (p < 0.05) and HOMA-IR (p < 0.05) compared to the MC group. There were no significant differences in the AUC of the OGTT and HOMA-IR among the MF group, L-MG group, and H-MG group after intervention. In addition, we found no significant differences in insulin among the groups (Figure S1a). However, concerning the FBG levels, the MF group demonstrated significantly lower levels compared to the L-MG group post-intervention (p < 0.05), while no significant difference was observed between the MF group and the H-MG group. In summary, these findings suggest that both low-dose and high-dose MN-Gup exhibit promising hypoglycemic and insulin resistance-alleviating effects. Furthermore, the high-dose MN-Gup group exhibited a more pronounced reduction in fasting glucose levels compared to the low-dose MN-Gup group, demonstrating a similarity to the positive control drug metformin. This observation suggests a potential dose-dependent relationship of MN-Gup.

We employed a histopathological examination of colon and pancreatic tissues (H & E staining) to analyze the degree of lesions (Figure 3). In the NC group, the colonic epithelium and mucosal architecture remained intact. Conversely, the MC group exhibited partial epithelial cell damage and sloughing. However, both the L-MG and H-MG groups demonstrated colonic epithelial and mucosal structures akin to those of the NC group. The pancreatic tissue findings mirrored those observed in the colon. In the NC group, the pancreatic islet structure of mice remained intact, with a plentiful and evenly distributed population of islet cells. Conversely, in the MC group, severe atrophy of the pancreatic islet tissue was observed, accompanied by a reduction in islet area. However, in both the L-MG and H-MG groups, there was a mitigation of pancreatic tissue pathology. There was an increase in the number of islet cells, indicative of restoration.

Fecal samples from the T2DM mice were MiSeq sequenced, then clustered for 97% similarity and annotated using the silva132/16s_bacteria database. The results of the diversity analysis are illustrated in Figure 4a–c. The Shannon and Sobs indices of the H-MG group exhibited a marked increase in comparison to those of the MC group (p < 0.05). Conversely, the Simpson index for the H-MG group demonstrated a significant decrease relative to that of the MC group (p < 0.05). A similar trend was evident in the L-MG group, where the Sobs index was notably higher than that in the MC group (p < 0.05). Although no significant difference was observed between the MC group and the L-MG group regarding the Shannon and Simpson indices, their trends aligned with those of the H-MG group.

The bacterial community composition of different groups was studied by PCoA analysis based on the Bray–Curtis distance (Figure 4d). The results showed a clear separation among groups (R² = 0.3986, p = 0.001, PC1 = 24.11%, PC2 = 17.38%). The confidence ellipses of the L-MG group and H-MG group overlapped, indicating that these two bacterial communities had similarities. Comparing each group separately, it was observed that the confidence ovals of the NC group (Figure S2a, R² = 0.4348, p = 0.003, PC1 = 48.71%, PC2 = 17.82%), L-MG group (Figure S2b, R² = 0.301, p = 0.003, PC1 = 44.43%, PC2 = 24.19%), and H-MG group (Figure S2c, R² = 0.3849, p = 0.003, PC1 = 43.10%, PC2 = 19.41%) did not overlap with those of the MC group. However, the confidence ellipses of the L-MG group and the H-MG group overlapped (Figure S2d, R² = 0.1506, p = 0.061, PC1 = 28.55%, PC2 = 23.30%), indicating that these two bacterial communities had similarities.

Species annotation results revealed that the predominant phyla were Firmicutes, Proteobacteria, Actinobacteria, and Bacteroidota (Figure S3). The analysis of microbiota composition at the genus level showed that the top ten species were Escherichia-Shigella, Staphylococcus, Aerococcus, Dubosiella, Proteus, Bifidobacterium, Kurthia, norank_f__Muribaculaceae, Enterococcus, and Lactobacillus (Figure 5e). After six weeks of MN-Gup intervention, some bacterial abundances changed. Among the top 10 strains in relative abundance, seven strains showed significant differences among groups. Bifidobacterium increased in the L-MG group and H-MG group (Figure 5d). The abundance of Bifidobacterium in the H-MG group was notably higher compared to that in the MC group (p < 0.05). In terms of the mean relative abundance of the strains, the relative abundance of Bifidobacterium doubled in the L-MG group and tripled in the H-MG group compared to the MC group. However, the relative abundance of Escherichia-Shigella (Figure 5a) and Staphylococcus (Figure 5b) in the H-MG group was significantly lower than that in the MC group (p < 0.05). Furthermore, the abundance of Aerococcus (Figure 5c) and Kurthia (Figure 5e) exhibited higher levels in the MF, L-MG, and H-MG groups, while they were less abundant in the NC and MC groups. Conversely, the relative abundance of norank_f__Muribaculaceae (Figure 5f) and Lactobacillus (Figure 5g) in the NC group was notably higher than that in the other groups.

Moreover, linear discriminant analysis effect size (LEfSe) can help to find good biomarkers of microbial communities [16]. LEfSe analysis (Figure 5h) showed the MC group was significantly enriched with Escherichia-Shigella and Staphylococcus. Consistent with the results of previous strain analysis, Aerococcus and Kurthia were markedly enriched in the L-MG group, whereas Bifidobacterium was significantly enriched in the H-MG group.

Gut microbiota-derived metabolites, such as short-chain fatty acids (SCFAs), are recognized as pivotal molecular mediators, bridging the interaction between the microbiota and the host [17]. To investigate the potential mechanism of intestinal flora intervention in attenuating T2DM induced by a high-fat diet and STZ, we sought to analyze the levels of SCFAs in the fecal samples. As shown in Figure 6, in terms of total acids (Figure 6a), both the L-MG group and H-MG group exhibited significantly higher levels compared to the MC group (p < 0.05). Specifically, acetic acid (Figure 6b) was notably elevated in these two groups compared to the MC group (p < 0.05), while propionic acid (Figure 6c) and butyric acid (Figure 6d) levels in the L-MG group and H-MG group did not show significant differences compared to those in the MC group.

Given the well-established role of SCFAs in mediating gut microbiota–host interactions, and building upon the analysis of SCFAs, we now redirect our attention to GLP-1, a pivotal incretin hormone intricately involved in the maintenance of glucose homeostasis [18]. As shown in Figure 6e, GLP-1 levels were notably elevated in both the L-MG and H-MG groups compared to the MC group (p < 0.05). Furthermore, there was no significant difference in the GLP-1 levels between the L-MG and H-MG groups. However, the mean GLP-1 level in the H-MG group was higher than that in the L-MG group.

A correlation analysis can reveal the correlation among complex factors [19]. Correlation heat maps were used to analyze the relationship between the top ten abundant species of different environmental factors at the genus level (Figure 7). Firstly, we observed a significant negative correlation between the relative abundances of Escherichia-Shigella and Staphylococcus and FBG, the AUC of the OGTT, and HOMA-IR. Conversely, the relative abundances of Lactobacillus exhibited a negative correlation with the FBG and the AUC of the OGTT. Secondly, we noted a positive correlation between the relative abundances of Dubosiella and Lactobacillus and the total acid content and the acetic acid content. Additionally, the relative abundance of Bifidobacterium demonstrated a significant positive correlation with the GLP-1 content, which in turn exhibited negative correlations with the incremental AUC.

The, Origin Data.xlsx, contains detailed data sets and additional information related to the findings presented in this study. Additional inquiries should be addressed to the corresponding author. Supplementary Materials File