Screening of Potential Probiotic Lactobacillaceae and Their Improvement of Type 2 Diabetes Mellitus by Promoting PI3K/AKT Signaling Pathway in db/db Mice

The use of fecal samples is approved by a healthy fecal provider who has not taken any medication in the past year and has undergone a professional physical examination at the Chinese People's Liberation Army (PLA) General Hospital (Beijing, China). Fecal samples (10 g) from healthy adults were resuspended in 100 ml de Man, Rogosa and Sharp medium (MRS; Beijing Land Bridge Technology Co., Ltd., China) and stationary cultivated for 24 h at 37°C. Then, cultures were diluted and spread onto the MRS agar plates. After cultivation for 24 h at 37°C, colonies were picked up and added to 10 ml MRS medium. After cultivation for 12 h at 37°C, the culture broth was added to 10 ml MRS medium with 5% (v/v) inoculation volume. Finally, a diluted suspension of the culture broth was plated onto MRS agar plates, and colonies were counted.

Total DNA was extracted from bacterial cultures in MRS broth. 200 μl of cultures was collected in 1.5 ml tube and centrifuged at 9,000 × g for 2 min. After the cells were washed, excess liquid was removed, and 20 μl bacterial lysate (NaOH/SDS) was added, then incubated at room temperature for 20 min and diluted 20 times. Finally, 2 μl of supernatant was centrifuged for 2 min. The amplification procedure was pre-denaturation (94°C, 5 min), 40 cycles of denaturation (94°C, 1 min), annealing (55°C, 60 s), extension (72°C, 60 s), and a final extension at 72°C for 10 min. DNA templates were amplified by PCR using the universal primers amplifying a region of the 16S rRNA gene: 27F (5′-AGAGTTTGATCMTGGCTCAG-3′) and 1492R (5′-TACGGYTACCTTGTTACGACTT-3′) (Kunduhoglu and Hacioglu 2021). After purification of the PCR product from the agarose gel, the samples were sent for sequencing. The resulting sequences were compared with the online tool BLAST.

Based on Barapatre et al. (2015) with some variations, the α-amylase inhibitory activity was measured. 10 μl of 0.02 M sodium phosphate buffer (pH 6.9) containing α-amylase (0.5 mg/ml) and 50 μl bacterial suspension (10⁹ CFU/ml) were mixed and incubated at 37°C for 10 min. Then, 50 μl of 1% (w/v) starch dissolved in 0.02 M sodium phosphate buffer (pH 6.9) was added and further incubated at 37°C for 10 min. The reaction was terminated by adding 20 μl of DNS reagent. Then, the reaction mixture was centrifuged at 8,000 × g at 4°C for 5 min. The α-amylase activity was detected at 540 nm, and inhibitory activity was calculated by using the formula below: (1) Inhibition (%)=[1−((γ1−δ1)/(α1−β1))]×100% Inhibition\;\left( \% \right) = \left[ {1 - \left( {\left( {{\gamma _1} - {\delta _1}} \right)/\left( {{\alpha _1} - {\beta _1}} \right)} \right)} \right] \times 100\% where α₁ – the absorbance with α-amylase and without the sample, β₁ – the absorbance without α-amylase or the sample, γ₁ – the absorbance with α-amylase and the sample, and δ₁ – the absorbance with the sample but without α-amylase.

The α-glucosidase inhibitory activity was determined by the method of Nguyen et al. (2018). The mixture included 50 μl of 0.1 M phosphate buffer (pH 6.9), 20 μl of bacterial suspension (10⁹ CFU/ml), and 20 μl of p-nitro-phenyl-α-d-glucopyranoside (PNPG, 5 mM). After the mixture was incubated for 10 min at 37°C, 10 μl of 0.5 U/ml α-glucosidase solution was added. The mixture was incubated for 10 min at 37°C; then the reaction was stopped by adding 100 μl of sodium carbonate (0.02 M). The reaction mixture was centrifuged at 8,000 × g for 5 min (4°C). Then, the supernatant was diluted with 1 ml distilled water, and the absorbance was measured at 405 nm. The inhibitory activity was calculated by using the formula below: (2) Inhibition (%)=[1−((γ2−δ2)/(α2−β2))]×100% Inhibition\;\left( \% \right) = \left[ {1 - \left( {\left( {{\gamma _2} - {\delta _2}} \right)/\left( {{\alpha _2} - {\beta _2}} \right)} \right)} \right] \times 100\% where α₂ – the absorbance with α-glucosidase and without the sample, β₂ – the absorbance without α-glucosidase or the sample, γ₂ – the absorbance with α-glucosidase and the sample, and δ₂ – the absorbance with the sample but without α-glucosidase.

Strain survivability was assessed using an in vitro digestion model. Artificial saliva, gastric juice, and intestinal juice were manufactured as described by Zhu et al. (2015). Following washing, bacteria (1 ml, 1.0 × 10⁹ CFU/ml) were simulated digestion at 37°C with peristaltic contraction, including a) suspension in 5 ml artificial saliva for 5 min; b) resuspension in 10 ml artificial gastric juice, incubation for 2 h; c) resuspension in 10 ml artificial intestinal juice, incubation for 2 h. Then, a diluted suspension of simulated digested bacteria was plated onto MRS agar plates and cultivated for 24 h at 37°C. The colonies were counted: (3) Survival rate (%)=M1M0×100% Survival\;rate\;\left( \% \right) = {{{{\rm{M}}_1}} \over {{{\rm{M}}_0}}} \times 100\% where M₁ – the number of viable bacteria after the process, M₀ – the number of viable bacteria before the process.

Bacterial adhesion to Caco-2 cell cultures was investigated according to Fonseca et al. (2021). After cultivation in MRS broth for 24 h at 37°C and washing twice with phosphate-buffered solution, the bacteria were resuspended in Dulbecco's Modified Eagle Medium (DMEM; Sigma-Aldrich Co., Ltd., USA) at a density of 10⁹ CFU/ml. Caco-2 cells were then cultivated for 1 h at 37°C in an atmosphere of 5% CO₂ with a 1 ml culture suspension of bacteria. After being washed three-time with PBS, treated with 1 ml Triton-X solution, and cultivated for 5 min at 37°C, the culture medium was serially diluted and plated onto MRS agar to count the bacterial colonies: (4) Adhesion ability (%)=N1N0×100% Adhesion\;ability\;\left( \% \right) = {{{{\rm{N}}_1}} \over {{{\rm{N}}_0}}} \times 100\% where N₁ – the number of viable bacteria after the process, N₀ – the number of viable bacteria before the process.

After incubating at 37°C for 24 h, four Lactobacillaceae strains were centrifuged for 5 min at 4°C and twice washed with sterile water to extract the bacterial cells. Then, the bacteria were freeze-dried to produce powders. Four Lactobacillaceae powders were suspended in sterile water, respectively. A diluted powder suspension was plated onto MRS agar plates to encounter the number of bacterial colonies.

An L₉ (3⁴) orthogonal table was designed. The inhibitory activities to α-amylase and α-glucosidase were used as the inspection index. L. fermentum 11 (A), L. fermentum 305 (B), L. plantarum 22 (C), and L. plantarum 25 (D) were used as the experimental factors. Each factor was designed with three levels, and the factors and levels of orthogonal tests for the inhibitory activities to α-amylase and α-glucosidase were shown in Table I.

All animals were maintained in a specific pathogen-free (SPF) environment. Under the 12-hour light cycle, the animals were permitted to access food and water at any time. In the following experiment, mice were divided into three groups after an adjustment period of 7 days: the control group (N group, 12 wt/wt mice), the diabetic model group (M group, 12 db/db mice), and the probiotic-treated group (ML group, 12 db/db mice with the Lactobacillaceae mixture intervention). The freeze-dried bacterial powders of four strains (L. plantarum 22, L. plantarum 25, L. fermentum 11, and L. fermentum 305) were mixed (the number of viable bacteria was 1.0 × 10⁹ CFU/ml for each strain). A dose of 10 ml/kg body weight was administered intragastrically to the ML group. The body weights (BW) and the FBG level of mice were recorded every week. A standard glucometer was used to determine the FBG level in the blood from the tail vein (F. Hoffmann-La Roche Co., Ltd., Switzerland).

The liver, kidney, and pancreas tissues were fixed in 4% (w/v) paraformaldehyde for 24 h, embedded in paraffin, and sectioned (5 μm). Sections were observed using an optical microscope (Motic China Group Co., Ltd., China). The sections were stained with hematoxylin and eosin (H&E), according to Zhao et al. (2021).

The colon tissues (100 mg) were homogenized in pre-cold RNA extraction solutions (1 ml), and centrifuged at 12,000 × g at 4°C for 10 min to obtain the total RNA. Reverse transcription was performed using the PrimeScript RT kit (Promega Corporation, USA), and the conditions were 42°C for 60 min then 70°C for 5 min. cDNA was amplified using SYBR Green Real-Time PCR Master Mix Kit (Servicebio Co., Ltd., China).. The amplification procedure was as follows: pre-denaturation (95°C, 10 min), denaturation (95°C, 15 s), and annealing/extension (60°C, 60 s), the dissolution curve (60–95°C, the temperature is increased by 0.3°C every 15 s). Relative fold changes were performed using the 2^−ΔΔCt method. The internal reference was GAPDH, and the primers used for qRT-PCR are shown below (Table II).

The liver tissue (100 mg) was homogenized in pre-chilled RIPA lysis buffer solution (1 ml), and centrifuged at 12,000 × g and 4°C for 10 min to extract total cell lysate. The protein concentration of cell lysate was measured using the BCA protein concentration assay kit (Servicebio Co., Ltd., China). Proteins were separated using 10% SDS-PAGE, transferred to nitrocellulose (PVDF) membranes, then blocked with 5% skim milk for 1 h at room temperature on a shaker. PVDF membranes were incubated with primary antibodies against PI3K and AKT at a 1 : 1000 dilution overnight at 4°C, followed by incubation with secondary antibodies for 30 min at room temperature. Finally, protein bands were visualized by enhanced ECL reagents, using β-actin as an internal control, and quantification of the bands was performed using the Image J program.

The cecal contents (50 mg) were vortexed in methanol/water solution (1 : 1 v/v, 1 ml) and centrifuged at 12,000 × g and 4°C for 5 min to collect the supernatant. Then the supernatant (50 μl) was derivatized in 50 μl derivatizing reagent and analyzed with UPLC-MS/MS. UPLC-MS/MS analysis was performed using an Ultra Performance Liquid Chromatography I-Class (Waters Co., Ltd., USA) coupled with a Tandem Mass Spectrometry (Waters Co., Ltd., USA). A 13-min linear gradient flow rate of 0.3 ml/min was used to inject samples into a UPLC RP Column (1.6 μm, 2.1 mm × 100 mm). Eluents A (0.1% formic acid in water) and B (methanol : isopropyl alcohol is 4 : 1) were used for the system. As follows, the solvent gradient: 5% B, 0 min; 5–15% B, 2.0 min; 15–55% B, 9.0 min; 55–100% B, 11.0 min; 100–5% B, 11.1 min; 5% B, 13 min. The Ultra Performance Liquid Chromatography operated with 150°C source temperature, 2.0 kV capillary voltage, 30.0 V cone voltage, 450°C desolvation temperature, 150 l/h cone gas flow, 1000 l/h desolvation gas flow, 0.15 ml/min collision gas flow rate. The collision gas used high-purity argon, and the rest of the gas paths used high-purity nitrogen. The data were analyzed by MassLynx software (Waters Co., Ltd., USA).

Data were analyzed with IBM^® SPSS^® 19.0 (IBM, USA) expressing as mean ± standard deviation. A one-way ANOVA was used to compare means. Statistical significance was determined by a p-value of less than 0.05.

Twenty-three isolates were obtained from human feces samples and confirmed by 16S rDNA sequencing. The sequences of isolates were uploaded to the NCBI database, and the accession number of all isolates are listed in Table III. The similarity of each isolate is over 99%. Eighteen strains were L. fermentum, and five strains were L. plantarum.

Inhibitory activities to α-amylase and α-glucosidase. The inhibitory activities to α-amylase and α-glucosidase of the 23 isolates were determined (Table IV). The α-amylase inhibitory activities of L. plantarum ranged from 17.33 ± 5.03% to 62.29 ± 0.44%; the highest value was observed for L. plantarum 22. The α-amylase inhibitory activities of L. fermentum ranged from 16.84 ± 8.34% to 58.40 ± 1.68%; the highest value was recorded for L. fermentum 11, followed by L. fermentum 305 (57.48 ± 5.04%). The α-glucosidase inhibitory activities of L. plantarum ranged from 11.45 ± 4.18% to 15.32 ± 0.89%; the highest was for L. plantarum 25, followed by L. plantarum 22 (14.89 ± 0.38%). The α-glucosidase inhibitory activities of L. fermentum ranged from 16.84 ± 8.34% to 58.40 ± 1.68%, with the highest value for L. fermentum 11, followed by L. fermentum 305 (57.48 ± 5.04%). Based on the results of the inhibitory activities of the isolates to α-glucosidase and α-amylase, four potential anti-diabetic strains (L. plantarum 22, L. plantarum 25, L. fermentum 11, and L. fermentum 305) were selected due to high inhibitory activities to α-amylase and α-glucosidase. The viable counts of the two isolates of L. plantarum and the two isolates of L. fermentum were 1.22 ± 0.47 × 10⁹ CFU/ml, 1.84 ± 0.70 × 10⁹ CFU/ml, 2.55 ± 0.30 × 10⁹ CFU/ml, and 2.04 ± 0.51 × 10⁹ CFU/ml, respectively.

The survival rate of the four strains is shown in Table V. The survival rate of the strains after the simulated gastrointestinal digestion test were 12.42 ± 2.84%, 9.10 ± 1.12%, 5.86 ± 0.52%, and 8.82 ± 2.50%, respectively. The adhesion rates of L. plan tarum 22 and L. plantarum 25 were 6.94 ± 0.27% and 6.91 ± 0.11%, respectively, while the adhesion rates of L. fermentum 11 and L. fermentum 305 were 6.09 ± 0.39% and 6.37 ± 0.28%, respectively.

Table VI shows that the optimal condition of α-amylase inhibitory activities was A₁B₁C₁D₁ (the optimal concentration of each strain in Lactobacillaceae powders was 10¹⁰ CFU/ml). The maximal inhibitory activity to α-amylase was 93.18 ± 1.19%. The effect of the factors in descending order was A > D > B > C. A, B, and D were significant factors, while C was a non-significant factor. L. fermentum strains 11 and 305 and L. plantarum 25 were significant factors, while L. plantarum 22 was a non-significant factor.

As was shown in Table VII, the inhibitory activities to α-glucosidase were most affected by L. fermentum 305, followed by L. fermentum 11, L. plantarum 22, and L. plantarum 25. The optimal conditions of α-glucosidase inhibitory activities were A₁B₁C₁D₁, and the maximal inhibitory activities to α-glucosidase was 75.33 ± 2.89%. The effect of the factors in descending order was B > A > C > D. A, B, C, and D were significant factors. The optimal concentration of each strain in Lactobacillaceae powders was 10¹⁰ CFU/ml, and each strain was a significant factor.

The BW of the mice was significantly increased in the M and ML group compared with the N group in week 12, respectively (Fig. 1). Compared to the M group, the BW of the mice decreased slightly in the ML group in week 12. FBG was significantly increased in the M group compared with the N group in week 12. As compared to the M group, the FBG of the mice significantly decreased in the ML group in week 12. During the first four weeks of the treatment, the FBG of the mice in the M group increased and then remained at a higher level. Interestingly, the Lactobacillaceae mixture-treated db/db mice showed ameliorated FBG from weeks 6 to 12 (p < 0.05). In week 12, the ML group displayed an FBG of 22.4% lower than the M group.

Fig 1.

In the N group, the liver, kidney, pancreas, and colon structures were well-organized and showed no pathological signs. Compared to the N group, in the M group, the following was observed: 1) many vacuoles, irregularly arranged and disordered cells in the liver (Fig. 2a); 2) degeneration of glomerular and tubular structures and fat accumulation in the kidney (Fig. 2b); 3) islet atrophy and focal necrosis in the pancreas (Fig. 2c); and 4) a thinner intestinal wall, degeneration of structures and decreasion of cells (Fig. 2d). However, the Lactobacillaceae culture revealed favorable influences on the injuries of liver, kidney, pancreas, and colon respectively.

Fig 2.

As shown in Fig. 3, the mRNA expression levels of IL-6 and IL-1β were significantly increased. In contrast, the mRNA expression levels of IL-10 and ZO-1 were significantly decreased in the M group compared with the N group. There was a significant increase in the mRNA expression levels of IL-10 in the ML group compared to the M group, while the mRNA expression levels of IL-1β significantly decreased. Lactobacillaceae isolates slightly increased the mRNA expression levels of ZO-1 but decreased that of IL-6.

Fig 3.

The protein expression levels of PI3K and AKT in the liver are shown in Fig. 4. Compared with the N group, the protein expression levels of PI3K and AKT in the M group were significantly decreased. There was a significant increase in the protein expression levels of AKT in the ML group compared with the M group. In contrast, the protein expression levels of PI3K increased, but not significantly.

Fig 4.

The contents of SCFAs in each group are shown in Fig. 5. Acetic acid, propionic acid, and butyric acid contents in the M group were significantly lower than in the N group. There was a significant increase in acetic acid and butyric acid contents in the ML group compared to the M group. however, no significant difference was detected in the contents of propionic acid between the M and ML group.

Fig 5.