Lactobacillus plantarum X1 with α-glucosidase inhibitory activity ameliorates type 2 diabetes in mice

Abstract

The cell-free supernatant of Lactobacillus plantarum X1 showed potential antidiabetic ability by inhibition of α-glucosidase activity in vitro. Oral administration of L. plantarum X1 was found to ameliorate hyperglycemia, glucose tolerance, insulin resistance, hormone levels, and lipid metabolism. And then L. plantarum X1 also partially increased the antioxidant capacity and ameliorated cytokine secretion and pancreatic damage in type 2 diabetic mice. In addition, L. plantarum X1 recovered the levels of acetic acid and elevated the levels of butyric acid in the feces of mice with diabetes markedly. In addition, type 2 diabetic mice with treatment of L. plantarum X1 showed an increase in the abundance of Bacteroidetes and a decrease in the proportion of Firmicutes, as well as an enrichment of butyrate-producing bacteria. These results demonstrated that L. plantarum X1 had the potential ability to ameliorate type 2 diabetes and the hypoglycemia ability were tightly associated with compositional changes in short-chain fatty acids and gut microbiota.

---

Introduction

Type 2 diabetes mellitus (T2DM) is a metabolic disease related to chronic hyperglycemia that causes many complications, including dysfunction of the eyes, feet, kidneys, nerves, heart, and blood vessels.¹ T2DM is characterized by insulin resistance and is accompanied by impaired insulin secretion with complicated mechanisms.² In the past few decades, rapid increases in the rate of T2DM and its high fatality rate have become a serious problem worldwide. Five classes of oral antidiabetic agents—α-glucosidase inhibitors, sulfonylurea, meglitinides, biguanides, and thiazolidinediones—are currently used for treatment of T2DM.³ However, these drugs usually have considerable side effects, such as drug resistance, abdominal bloating, flatulence, diarrhea, and secondary failure.^4–6 α-glucosidase inhibitors, a class of oral antidiabetes agents, reducing postprandial hyperglycemia by slowing the digestion of carbohydrates in the intestines are widely used in treatment of diabetes.⁷ In vitro, evaluation of α-glucosidase inhibitor activity of natural α-glucosidase inhibitors, including fruits,⁸ vegetables,⁸ plants,⁹ drinks,¹⁰ and probiotics,¹¹ offers a fast and simple strategy of screening potential hypoglycemic products.

Hypoglycemic effect of probiotics have gained wide acceptance in vivo. It has been shown that some probiotics could ameliorate and/or delay the onset of T2DM and its complications. For example, L. plantarum NCU116 and its fermented carrot juice had the potential ability to ameliorate T2DM induced by a high-fat diet and low-dose streptozotocin (STZ) in rats.¹² L. rhamnosus CCFM0528 effectively improved glucose intolerance, hyperglycemia, cytokine secretion, and pancreatic damage in mice with T2DM.¹³ Besides, many other studies have shown the favorable effects of probiotics such as L. reuteri GMNL-263, L. casei Zhang, and L. rhamnosus GG in the treatment of insulin resistance,¹⁴ glucose tolerance,¹⁵ and glucose regulation.¹⁶

Short-chain fatty acids (SCFAs) are produced from probiotic fermentation and improved inflammation response¹² and regulated an appropriate colonic pH and protected against pathological changes in the colonic mucosa.¹⁷ T2DM can be regulated by SCFAs and microbial ecology which effected energy homeostasis and glucose metabolism of host.¹⁸ Probiotics had beneficial effects on T2DM by increasing SCFAs products.¹² A metagenome-wide study presented that type 2 diabetes were assessed by gut microbiota and were characterized by compositional changes in intestinal microbiota such as decreasing of the abundance of common bacteria producing butyrate and increasing various opportunistic pathogens.¹⁹

In this study, two strains of L. plantarum with the highest and lowest α-glucosidase inhibitory activities in vitro were selected to evaluate their effects on symptoms of T2DM in a high-fat diet and STZ induced mouse model. The potential hypoglycemic pathways of L. plantarum X1 on T2DM was estimated through regulating of the concentration of SCFAs and the composition of gut microbiota.

Materials and methods

Materials and reagents

An α-glucosidase activity kit, α-glucosidase and STZ were purchased from Sigma Chemical Co. (St. Louis, MO). A blood glucose meter and blood glucose test strips were obtained from HMD Biomedical Inc. (Taiwan, China). Insulin was purchased from Novo Nordisk (Denmark). A total cholesterol (TC) kit, a triglyceride (TG) kit, a high-density lipoprotein cholesterol (HDL-C) kit, a low-density lipoprotein cholesterol (LDL-C) kit, a superoxide dismutase (SOD) kit, a glutathione (GSH) kit, and a malondialdehyde (MDA) kit were purchased from Nanjing Jiancheng Biology Engineering Institute (Nanjing, China). A glycosylated hemoglobin (HbA1c) kit, an insulin kit, a leptin kit, a tumor necrosis factor (TNF)-α kit, an interleukin (IL)-6 kit, and an IL-10 kit were purchased from Abcam (Cambridge, U.K.). Fast DNA Spin Kit and Gene Clean Turbo kit were purchased from Beijing Lianlixin Biotech Co., Ltd (MP Biomedicals, Beijing, China). Quant-iT PicoGreen dsDNA Assay Kit were purchased from Life Technologies (Carlsbad, California, USA). TruSeq DNA LT Sample Preparation Kit and MiSeq Reagent Kit were purchased from Illumina (Santiago, California, USA).

Bacterial culture

The information of the strains in this study was listed in Table 1. These strains were maintained as frozen stocks (−80 °C) in MRS broth supplemented with 30% (v/v) glycerol. These strains were consecutively reactivated at least three times with 1% (v/v) inoculum in MRS broth at 37 °C for 20 h prior to use. All of the strains were incubated in MRS broth at 37 °C for 18 h. Cell-free supernatant of the fermentation sample was obtained by centrifugation at 4600 g for 5 min and was then filtered with a 0.22 μm filter to remove bacterial cells. The precipitations of fermentation samples, that was L. plantarum, were harvested by centrifugation at 4600 g for 5 min and washed two times with phosphate-buffered saline (PBS) solution. The numbers of viable bacteria of harvested precipitations were adjusted to 8 × 10¹⁰ cfu mL⁻¹ by colony count and stored with 30% sucrose as a protectant at −80 °C. Before oral administration, the gavage samples were freshly prepared using a 10-fold dilution of the stored strains.

Table 1 Strains information

Strains	Sources	Storage
L. plantarum X1	Pickles	Collection of food Microorganisms of Jiangnan University (Wuxi, China)
L. plantarum CCFM12	Pickles
L. plantarum CCFM30	Raw milk
L. plantarum CCFM236	Yoghurt
L. plantarum CCFM311	Yoghurt
L. plantarum RS4	Stinky tofu

---

α-Glucosidase inhibitory activity (AGA) assay

The α-glucosidase inhibitory activity of strain was determined with an α-glucosidase activity kit according to the operating instructions.

The reaction mixture (220 μL) contained 200 μL of master reaction mix (200 μL of assay buffer and 8 μL p-nitrophenyl α-D-glucopyranoside substrate) and 20 μL of the sample. The sample contained three styles: (A) 10 μL of PBS and 10 μL of α-glucosidase (0.17 U mL⁻¹); (B) 10 μL of sample and 10 μL of α-glucosidase; and (C) 10 μL of sample and 10 μL of PBS. The reaction was mixed and the absorbance measured at 405 nm using microplate reader (A₄₀₅)_initial. The mixture was processed at 37 °C for 20 min, and the amount of p-nitrophenol released was quantified at 405 nm (A₄₀₅)_final using a microplate reader. One set of the reaction mixture prepared with 220 μL of water instead of the reaction was used as blank. Another set of reaction mixture prepared with 20 μL of standard and 200 μL of water was used as calibrator.

where (A₄₀₅)_calibrator indicates the value for the calibrator at 20 minutes and (A₄₀₅)_water indicates the value for water at 20 minutes.
where B₁ indicates the α-glucosidase activity of A; B₂ indicates the α-glucosidase activity of B; B₃ indicates the α-glucosidase activity of C.

Exopolysaccharide of L. Plantarum

Isolation and quantification of exopolysaccharide (EPS) were performed according to previous study.²⁰ The MRS cultures of L. Plantarum were boiled (100 °C, 10 min), and added 17% (v/v) of 85% trichloroacetic acid solution after cooling to remove protein by centrifugation. Supernatant was precipitated with ethanol (100% v/v) to obtain EPS (14 000 g, 20 min). Total EPS was determined using phenol-sulfuric acid method with glucose as a standard.

Animals

Animal experiment design was shown in Fig. 1. The study was approved by the Ethics Committee of Jiangnan University, China (JN no. 20121203-0120). The use and care of animals followed National Institutes of Health guidelines (NIH Publication 85-23, 1996). Surgical procedures were performed under sodium pentobarbital anesthesia, and all efforts were made to minimize suffering. Forty Male C57BL/6J mice (3 weeks old, inbred-specific, pathogen-free) were purchased from Shanghai Laboratory Animal Center (Shanghai, China). All mice were housed in a controlled environment (temperature, 22 °C ± 2 °C; humidity, 55% ± 5%; lights, 12 h light/dark cycle) and divided into five groups according to body weight as follows: (1) the control (C) group: mice without T2DM and oral administration of 0.25 mL 3% sucrose; (2) the model (M) group: mice with T2DM and oral administration of 0.25 mL 3% sucrose; (3) the pioglitazone group (P): mice with T2DM and oral administration of 0.25 mL pioglitazone at a dose of 10 mg kg⁻¹ body weight; (4) the L. plantarum X1(X1) group: mice with T2DM and oral administration of 0.25 mL 2 × 10⁹ L. plantarum X1; and (5) the L. plantarum CCFM30 (CCFM30) group: mice with T2DM and oral administration of 0.25 mL 2 × 10⁹ L. plantarum CCFM30. Here, in order to play a beneficial role, it is necessary to keep more than 10⁶ cfu viable bacteria in the intestinal, and which required the viable bacteria achieving 10⁹ cfu when oral administration. In addition, the amount of 10⁹ cfu was used most frequently in many studies for animal treatment by probiotics.^5,12,13 All mice administered corresponding experimental samples at 10:00 AM once a day from week 2 to week 12.

Fig. 1 Animal experimental design. C: the control group. M: the model group. P: the pioglitazone group. X1: the L. plantarum X1 group. CCFM30: the L. plantarum CCFM30.

As shown in Fig. 1, all mice were freely kept with diet and water after adapting to their new environment for 1 week. At 2 and 3 weeks, eight mice were fed a normal diet and the other mice were fed a high-fat diet (Table 2). At 4 weeks, all mice fasted for 12 h, and then the mice that had been fed a high-fat diet received STZ freshly dissolved in 50 mmol L⁻¹ citrate buffer at a dose of 100 mg kg⁻¹ body weight, and the other mice that had been fed a normal diet underwent intraperitoneal injection of an equivalent volume of 50 mmol L⁻¹ citrate buffer (pH 4.5). From 4 weeks, all mice were fed a normal diet. High-fat-diet plus STZ injection was used to induce experimental T2DM.²¹ T2DM was confirmed by a fasting blood glucose (FBG) level higher than 7 mmol L⁻¹ and a postprandial 2 h blood glucose (PBG) level higher than 11.1 mmol L⁻¹ (ref. 22) at 1 week after STZ injection.

Table 2 Composition of normal and high-fat diets

Normal chow diet		High-fat diet
Nutrient and energy composition		Nutrient and energy composition
Caloric (kcal per 100 g)	352.00	Caloric (kcal per 100 g)	524.00
Carbohydrates (g per 100 g)	52.00	Proteins (g per 100 g)	26.00
Energy (%)	60.50	Energy (%)	20.00
Proteins (g per 100 g)	22.10	Carbohydrates (g per 100 g)	26.00
Energy (%)	25.70	Energy (%)	20.00
Lipids (g per 100 g)	5.28	Lipids (g per 100 g)	35.00
Energy (%)	13.80	Energy (%)	60.00
Ingredients(g per 100 g)		Ingredients(g per 100 g)
Water	9.20	Casein	258.45
Crude protein	22.10	Cystine	3.88
Crude fat	5.10	Maltodextrin	161.53
Crude ash	5.20	Sucrose	88.91
Crude fiber	4.12	Cellulose	64.61
Nitrogen-free extract	50.00	Soybean oil	32.31
Calcium	1.24	Lard	316.6
Phosphorus	0.92	Mineral mixture M1002	12.92
Lysine	1.34	Calcium hydrophosphate	16.8
Methionine and cystine	0.78	Calcium carbonate	7.11
		Potassium citrate	21.32
		Vitamin mixture V1001	12.92
		Choline bitartrate	2.58
		Edible blue dye	0.065

---

Measurements of daily basis indicators

Body weight, food intake, water intake, FBG, and PBG were measured on the last day of every week. FBG and PBG were measured from tail blood samples using a glucometer.

An oral glucose tolerance test (OGTT) was performed after 15 h of fasting on the last day of week 4 (1 week after STZ) and week 11 (1 week before the end of the experiment). All mice received oral administration of glucose at a dose of 2 g kg⁻¹ body weight, and blood glucose levels were measured at 0, 30, 60, and 120 min after oral administration of glucose.

Biochemical analysis

At the end of the experiment, serum was obtained by centrifuging whole blood at 3000 g for 10 min. Serum levels of TC, TG, HDL-C, and LDL-C, HbA1C, insulin, leptin, TNF-α, IL-6, and IL-10 were determined by enzyme-linked immunosorbent method according to the kits instructions. A homeostasis model to estimate insulin resistance (HOMA-IR) was calculated with the formula HOMA-IR = fasting insulin level × FBG/22.5.²³ The SOD activities, GSH levels, and MDA levels were measured with commercial kits purchased from Nanjing Jiancheng Biology Engineering Institute according to the manufacturer's instructions.

Histopathological examination

Pancreata were fixed in 10% formalin solution and embedded in paraffin, and paraffin samples were then cut and stained with hematoxylin and eosin (H&E) for histopathological examination. Histopathological analyses were recorded with an automatic scanner.

Determination of short-chain fatty acids (SCFAs) in cell-free supernatant and feces

SCFAs was measured as described in.²⁴ Approximately 50 μL mg⁻¹ of cell-free supernatant/fecal sample was soaked in saturated sodium chloride and then added to acidified water spiked with isotope-labeled internal SCFAs standards. Then the samples were mixed thoroughly and centrifuged. The supernatant was removed for quadruple extraction with diethyl ether and then analyzed by GC-MS.

Fecal microbiota analysis by metagenome

Gut microbiota analysis was performed according to the method described in previous studies.²⁵ Briefly, microbial genome DNA was extracted from approximately 100 mg of fecal sample using Fast DNA Spin Kit for Soil, and then the V4 region of 16S rRNA from the extracted DNA was PCR amplified using primers (forward primer, 5′-AYTGGGYDTAAAGNG-3′; reverse primer, 5′-TACNVGGGTATCTAATCC-3′) as described previously.²⁶ The PCR products were excised from a 1.5% agarose gel and purified by Gene Clean Turbo kit. Then the purified sample was quantified by Quant-iT PicoGreen dsDNA Assay Kit and concentrations were measured using a fluorescence spectrophotometer. Fifty samples of the same concentration were mixed in a library. The Library was prepared using TruSeq DNA LT Sample Preparation Kit and then sequenced for 500 + 7 cycles on an Illumina Miseq using the MiSeq Reagent Kit.

Statistical analysis

All statistical analyses were performed with GraphPad Prism 5 and Origin 8.1. Data in bar plots were mean ± SEM. The area under the curve for glucose (AUC_glucose) was determined using the trapezoidal rule. Significant differences among experimental groups (P values of less than 0.05) were assessed with Tukey's test.

Results

α-Glucosidase inhibitory activity and SCFAs and EPS level of cell-free supernatant

α-Glucosidase could hydrolyze glycosidic bonds, release glucose, and increase postprandial hyperglycemia. As shown in Fig. 2, the cell-free supernatant of L. plantarum X1, L. plantarum CCFM12, L. plantarum CCFM30, L. plantarum CCFEM236, L. plantarum CCFM311, and L. plantarum RS4 showed different inhibitory activities, with values of 41.81%, 27.14%, 24.96%, 31.18%, 27.83%, and 35.68%, respectively. In comparison to the L. plantarum CCFM30, L. plantarum X1 showed more concentrations of EPS (P < 0.05, Fig. 2B). However, the concentration of acetic acid has no significant difference between the two strains (Fig. 2B). Propionic acid and butyric acid in the cell free supernatant were not be detected. On this basis, the antidiabetic effects of L. plantarum X1 and L. plantarum CCFM30 with the highest and lowest α-glucosidase inhibitory activities were further investigated on high-fat and STZ induced type 2 diabetic mice.

Fig. 2 α-Glucosidase inhibitory activity and acetic acid concentration and EPS level of cell-free supernatant from six strains. ^*Significantly different from X1 (P < 0.05).

Body weight, food intake and water intake

Body weight, food intake, and water intake were used as direct indexes of T2DM. As shown in Fig. 3A, body weight of the diabetic mice (M, P, X1, and CCFM30 groups) was lower than that of the normal mice (the C group), whereas there was no significant difference among the five groups from weeks 4 to 12. From weeks 5 to 12, the food and water intake in the model group was significantly greater than that in the control group, whereas the food and water intake in the P and X1 group was notably lower than in the M group (Fig. 3B and C). Food intake and water intake in the CCFM30 group showed no marked difference from the M group from 5 to 12 weeks. On the whole, oral administration of L. plantarum X1 could alleviate the symptoms to a certain extent, while L. plantarum CCFM30 has no obvious effect on these direct indexes.

Fig. 3 Effect of X1 and CCFM30 on diabetes-related direct indexes. Time course of changes in body weight (A), food intake (B), and water intake (C) in five groups. (○) C group: mice without T2DM plus 3% sucrose. (□) M group: mice with T2DM plus 3% sucrose. (△) P group: mice with T2DM plus 10 mg kg⁻¹ body weight pioglitazone. (●) X1 group: mice with T2DM plus 2 × 10⁹ L. plantarum X1. (▲) CCFM30 group: mice with T2DM plus 2 × 10⁹ L. plantarum CCFM30. Data are expressed as mean ± SEM (n = 8). ^*Significantly different from C group (P < 0.05). ^#Significantly different from M group (P < 0.05).

Blood glucose

In theory, the type 2 diabetic mice had higher levels of FBG (≥7 mmol L⁻¹) and PBG (≥11.1 mmol L⁻¹). As expected, the levels of FBG and PBG in the model group were markedly higher than that in the control group after week 4 (P < 0.05) (Fig. 4). Interestingly, the animals that received pioglitazone and L. plantarum X1 showed a significant decrease in the FBG and PBG levels compared to the M group, especially in the PBG level (P < 0.05). Whereas the PBG levels in the X1 groups were also markedly higher than those in the C group from week 4. The FBG and PBG levels in the CCFM30 group were no significant difference from the model group. The results indicate that oral administration of L. plantarum X1 partially attenuates blood glucose elevation.

Fig. 4 Effect of X1 and CCFM30 on FBG and PBG. Time course of changes in FBG (A) and PBG (B). FBG and PBG were measured on last day of every week in all groups. (○) C group: mice without T2DM plus 3% sucrose. (□) M group: mice with T2DM plus 3% sucrose. (△) P group; mice with T2DM plus 10 mg kg⁻¹ body weight pioglitazone. (●) X1 group: mice with T2DM plus 2 × 10⁹ L. plantarum X1. (▲) CCFM30 group: mice with T2DM plus 2 × 10⁹ L. plantarum CCFM30. Data are expressed as the mean ± SEM (n = 8). ^*Significantly different from C group (P < 0.05). ^#Significantly different from M group (P < 0.05).

OGTT

The results of the OGTT at week 4 and 11 were depicted in Fig. 5. At week 4, the blood glucose levels and AUC_glucose of the diabetic mice (M, P, X1, and CCFM30 groups) were significantly higher than those in the control group (P < 0.05; Fig. 5A and B). The diabetic mice received pioglitazone and L. plantarum X1 showed an effectively lower in the AUC_glucose value (16.65%, and 13.91%, respectively) than that in the M group at week 4 (P < 0.05; Fig. 5A). At week 11, a notable decrease of the PBG levels at 30, 60, and 120 min was found in the P and X1 groups in comparison to that in the M group (P < 0.05 Fig. 5C). Interestingly, compared to the M group, the AUC_glucose of the P and X1 groups was significantly reduced by 26.82% and 23.40%, respectively (P < 0.05). The AUC_glucose value in the CCFM30 group was no marked difference from the M group at week 4 and week 11. These results showed that the glucose tolerance of mice with T2DM was impaired at week 4, whereas the impairment was alleviated after oral administration of L. plantarum X1.

Fig. 5 OGTT and AUC_glucose were performed at week 4 and week 11 in five groups. (A) OGTT at week 4. (B) AUC_glucose at week 4. (C) OGTT at week 11. (D) AUC_glucose at week 11. (○) C group: mice without T2DM plus sucrose. (□) M group: mice with T2DM plus sucrose. (△) P group; mice with T2DM plus pioglitazone. (●) X1 group: mice with T2DM plus L. plantarum X1. (▲) CCFM30 group: mice with T2DM plus L. plantarum CCFM30. Data are expressed as mean ± SEM (n = 8). ^*Significantly different from C group (P < 0.05). ^#Significantly different from M group (P < 0.05).

Liver and lipid metabolic parameters

At the end of the experiment, the SOD, MDA, GSH, and GSH-Px levels were remarkably different between the control and model groups (P < 0.05; Table 3). The GSH levels in the P, X1, and CCFM30 groups were no significant difference from those in the control and model groups. The X1 and CCFM30 group showed the same recovery effects in the SOD and GSH-Px activity and MDA level compared to the M group (P < 0.05). As shown in Table 3, the serum TG, TC, HDL-C, and LDL-C levels in the M group were significantly different with those in the C group. Both L. plantarum X1 and L. plantarum CCFM30 showed significant effect on recovery of TC and HDL-C levels. L. plantarum CCFM30 exhibited better effect in decreasing TG level in T2DM whereas L. plantarum X1 showed more efficient in recovery LDL-C level. These results indicated that treatment with L. plantarum X1 and L. plantarum CCFM30 improved antioxidants and affected lipid metabolic parameters to different extents.

Table 3 Effect of X1 and CCFM30 treatments on levels of oxidative stress and serum lipids in diabetic mice^a

Parameters	C	M	P	X1	CCFM30
a C group: mice without T2DM plus 3% sucrose; M group: mice with T2DM plus 3% sucrose; P group: mice with T2DM plus 10 mg kg^−1 body weight pioglitazone; X1 group: mice with T2DM plus 2 × 10^9 L. plantarum X1; CCFM30 group: mice with T2DM plus 2 × 10^9 L. plantarum CCFM30. Data are expressed as mean ± SEM (n = 5).b Significantly different (P < 0.05) from C group.c Significantly different (P < 0.05) from M group.
SOD (U mgprot^−1)	424.52 ± 5.30^c	384.91 ± 6.33^b	433.55 ± 9.92^c	424.10 ± 5.77^c	431.21 ± 10.57^c
MDA (nmol mgprot^−1)	1.11 ± 0.11^c	1.99 ± 0.14^b	0.99 ± 0.13^c	0.90 ± 0.11^c	1.09 ± 0.22^c
GSH (μmol gprot^−1)	1.25 ± 0.05^c	0.85 ± 0.07^b	1.06 ± 0.05	1.04 ± 0.05	0.92 ± 0.05
GSH-Px (μmol gprot^−1)	153.34 ± 4.45^c	78.02 ± 5.00^b	132.58 ± 4.01^c	117.22 ± 4.66^c	95.69 ± 2.58^b
TG (mmol L^−1)	0.62 ± 0.10^c	1.18 ± 0.07^b	0.96 ± 0.01	0.88 ± 0.04	0.70 ± 0.10^c
TC (mmol L^−1)	1.76 ± 0.05^c	2.69 ± 0.07^b	2.10 ± 0.07^c	2.11 ± 0.11^c	2.11 ± 0.16^c
HDL-C (mmol L^−1)	34.12 ± 0.95^c	20.67 ± 1.22^b	25.19 ± 0.37^b^,^c	25.75 ± 0.37^b^,^c	26.87 ± 1.11^b^,^c
LDL-C (mmol L^−1)	0.40 ± 0.03^c	0.52 ± 0.01^b	0.41 ± 0.04^c	0.41 ± 0.01^c	0.48 ± 0.01

---

HbA1C, HOMA-IR and hormones

The levels of HOMA-IR, HbA1C, hormones (glycogen, insulin, and leptin) in the model group were significantly higher than those in the control group (P < 0.05, Fig. 6). HOMA-IR and fasting serum insulin level are usually used to estimate insulin resistance. The fasting serum insulin level in the X1 group was significantly different from the M group (P < 0.05), whereas the value in the CCFM30 group showed no obvious abnormality compared with that of the M group. Accordingly, treatment with L. plantarum X1 more effectively decreased HOMA-IR than that treated with L. plantarum CCFM30 (P < 0.05). L. plantarum CCFM30 could decrease the level of leptin effectively but no significant differences were observed in the levels of insulin and HbA1C compared to the M groups. However, supplementation of L. plantarum X1 to the diabetic mice caused a significant decrease in the HbA1C and hormone levels compared to the M group which also appeared in the P group (P < 0.05).

Fig. 6 Effects of X1 and CCFM30 on HOMA-IR, HbA1C, insulin, glycogen, and leptin levels. (A) HOMA-IR, (B) HbA1C, (C) insulin, (D) glycogen, (E) leptin. C group, mice without T2DM plus sucrose. M group, mice with T2DM plus sucrose. P group, mice with T2DM plus pioglitazone. X1 group, mice with T2DM plus L. plantarum X1. CCFM30 group, mice with T2DM plus L. plantarum CCFM30. Data are expressed as mean ± SEM (n = 5). ^*Significantly different (P < 0.05) from C group. ^#Significantly different (P < 0.05) from M group.

Cytokines

TNF-α and IL-6 are considered to be pro-inflammatory cytokines, whereas IL-10 is an anti-inflammatory cytokine. Fig. 7 presented the levels of TNF-α, IL-10, and IL-6 in mice at the end of the treatment. The TNF-α, IL-6 and IL-10 values in the model group (267.17 ng L⁻¹, 144.83 ng L⁻¹ and 522.33 pg mL⁻¹) were significantly different with those in the control group (245.17 ng L⁻¹, 114.67 ng L⁻¹ and 547.00 pg mL⁻¹). Pioglitazone exhibited a positive effect on cytokines, which significantly decreased the levels of TNF-α and IL-6 (13% and 19%) and increased IL-10 level compared with the M group (P < 0.05). Similarly, oral administration of L. plantarum X1 produced a consistent trend with the pioglitazone, but the CCFM30 group showed no remarkable difference in the levels of IL-10 and IL-6 compared to that in the M group. However, the CCFM30 group showed a significant decrease in the TNF-α level compared with the M group. These results indicated that oral administration of L. plantarum X1 could more effectively improve anti-inflammatory and reduce pro-inflammatory cytokines than treatment with L. plantarum CCFM30.

Fig. 7 Effects of X1 and CCFM30 on TNF-α, IL-10, and IL-6 levels. (A) TNF-α, (B) IL-10, (C) IL-6. Data are expressed as mean ± SEM (n = 5). ^*Significantly different (P < 0.05) from C group. ^#Significantly different (P < 0.05) from M group.

Histopathology

Impaired pancreatic islets was an important feature of type 2 diabetes mellitus, which was damaged by STZ in mimetic T2DM model. H&E staining of the pancreata displayed morphological characteristics of pancreatic islets. The normal islets in the C group were round, with a regular structure and clear boundaries (Fig. 8A). In comparison to the normal islets, the impaired islets in the M group had an irregular shape and hyperchromasia intercellular substance (Fig. 8B). H&E staining of the pancreata in the P, X1, and CCFM30 groups was restored in certain extent and the islet structure was close to that of the C groups (Fig. 8C–E). However, as showed in figures, the protective effect of the islets and the reduced rate of beta cell destruction by oral administration of L. plantarum X1 were better than that of L. plantarum CCFM30.

Fig. 8 H&E staining of pancreas. (A) C group; (B) M group; (C) P group; (D) X1 group; (E) CCFM30 group (magnification, 400×).

SCFAs composition in feces

SCFAs in feces consisted mainly of acetic acid, propionic acid, and butyric acid. Fig. 9 shows the concentrations of acetic acid, propionic acid, butyric acid, and total SCFAs at the end of the experiment. No differences were seen in the concentrations of propionic acid between the five groups. Compared to the C group, the M groups contained markedly lower concentrations of acetic acid (14.98 μmol g⁻¹) and total SCFAs (19.38 μmol g⁻¹) in feces (P < 0.05) and L. plantarum CCFM30 showed no effect on recovery of the level of acetic acid (14.95 μmol g⁻¹) and total SCFAs (20.47 μmol g⁻¹). However, the concentrations of acetic acid and total SCFAs in the X1 groups were significantly restored as pioglitazone did. In addition, although butyric acid level showed no obvious difference between C and M group, oral administration of L. plantarum X1 and L. plantarum CCFM30 both significantly increased the concentration of butyric acid in mice feces.

Fig. 9 Effects of X1 and CCFM30 on acetic acid, propionic acid, butyric acid, and SCFAs in feces. (A) Effects of X1 and CCFM30 on acetic acid, propionic acid, and butyric acid in feces. (B) Effects of X1 and CCFM30 on total SCFAs in feces. Data are expressed as mean ± SEM (n = 5). ^*Significantly different (P < 0.05) from C group. ^#Significantly different (P < 0.05) from M group.

Fecal microbiota

The composition of the microbiota is defined by high-quality, classifiable 16S ribosomal RNA (rRNA) gene sequencing. The relative populations of bacterial phyla in the C group were Firmicutes, 21.72%; Bacteroidetes, 63.19%; Proteobacteria, 0.43%; other, 14.53%; and Deferribacteres, 0.10% (Fig. 10A). The abundance of Bacteroidetes and Firmicutes in the M group was obviously different from the C group (Fig. 10C and D). The fecal microbiota was profoundly affected after treatment with pioglitazone and L. plantarum X1, which increased the abundance of Bacteroidete and decreased the proportion of Firmicutes (P < 0.05). However, only significant changes on the abundance of Firmicutes were observed after L. plantarum CCFM30 treatment compared to the M group. At the genus level, the abundance of Bacteroides was enriched in the X1 and P group compared to that in the M group as expected (P < 0.05, Fig. 10B and E). However, no significant increases in the proportion of Bacteroide were also observed in the CCFM30 group.

Fig. 10 Effects of X1 and CCFM30 on fecal microbiota. (A) Relative abundance of major phyla for five groups. (B) Relative abundance of genus for five groups. (C) Relative abundance of Bacteroidetes (phyla) for five groups. (D) Relative abundance of Firmicutes (phyla) for five groups. (E) Relative abundance of Bacteroides (genus) for five groups. Relative abundance of main phyla and genus is expressed as mean ± SEM. Values with different superscript letters significantly differ at P < 0.05.

Discussion

Inhibiting α-glucosidase activity indicated that bacteria could reduce glucose product in the intestinal tract and further prevented and treated type 2 diabetic mellitus. Cell-free supernatants were the production from probiotic fermentation, and previous reports demonstrated that cell-free supernatants^11,27 and cell-free extract¹¹ of some lactic acid bacteria possessed α-glucosidase inhibitory activity. Furthermore, the α-glucosidase inhibitory activities of cell-free supernatants were notable higher than that of cell-free extract,^11,27 so we speculated that α-glucosidase inhibitory activities of bacteria mainly focused on cell-free supernatants. Thus, inhibition of α-glucosidase in the cell-free supernatants could be as an important approach to evaluate the potential antidiabetic strains.^11,28 In this study, cell-free supernatants of six strains of L. plantarum showed α-glucosidase inhibitory activity in various degrees. In these effective strains, cell-free supernatants of L. plantarum X1 and L. plantarum CCFM30 showed the highest and lowest α-glucosidase inhibitory activity respectively (Fig. 2A). Some studies have suggested that antidiabetic activity might result from the products from probiotic fermentation such as the EPS²⁹ and SCFAs.¹² In this study, it must be noted that the level of EPS showed positive correlation with the α-glucosidase inhibitory activities of bacteria although the acetic acid concentration of these two strains in cell free supernatant has no obvious difference (Fig. 2B). Considering the probiotics administrated to mice could live a period time and produced metabolites contained acetic acid and exopolysaccharides which have hypoglycemic effects in the intestinal tract, L. plantarum X1 and L. plantarum CCFM30 were selected for further studies to explore their potential hypoglycemic activity in mouse model with T2DM.

Probiotics have been confirmed as a potent hypoglycemic material by many studies in animal models.^{12–16,30–32} So, in this study, we investigated whether the two bacteria with different α-glucosidase inhibitory activities showed distinct effects on restoration of diabetes. In this study, L. plantarum X1 more effectively attenuated the diabetic symptom on blood glucose level, glucose tolerance and insulin resistance status than L. plantarum CCFM30.

According with others reports,^5,12,33,34 in this study, the mice in the T2DM model displayed hyperglycemia, glucose intolerance, insulin resistance, and impaired β-cells of islet, such as fasting insulin level, and HOMA-IR, and most importantly impaired the function of β-cells according to histologic changes in the pancreas. Both strains used in this study showed certain effects on restore of symptom. Oral supplementation with the L. plantarum X1 markedly decreased food and water intake, FBG, PBG, OGTT, and AUC_glucose (Fig. 3–5). The evidence clearly demonstrated that oral administration of L. plantarum X1 significantly recovered almost all the phenotypes mentioned above, but supplementation with L. plantarum CCFM30 was insufficient to create the same mitigation effects. In addition, HbA1C is used to reflect the average plasma glucose concentration from 4–12 weeks. Studies have shown that treatment with L. rhamnosus CCFM0528 significantly decreased fasting blood sugar and HbA1C concentrations in mice with T2DM,¹³ which was consistent with our results on HbA1C level. These results agree with our conjecture that the cell-free supernatant of L. plantarum X1 with better α-glucosidase inhibitory activity can better delay the digestion and absorption of carbohydrates than that of L. plantarum CCFM30, thereby maintaining a blood sugar balance by reducing the postprandial blood glucose in vivo (Fig. 4B).

Hormones and cytokines play important roles in T2DM. Hormones such as insulin, glycogen, and leptin are often associated with T2DM. In this study, a high-fat diet led to leptin resistance, which further affected insulin resistance. HOMA-IR has been widely used to evaluate insulin resistance, and a high serum level of insulin is affected by insulin resistance.¹² In comparison to L. plantarum CCFM30, the better effects of L. plantarum X1 on decreasing insulin resistance resulted in a decrease of fasting serum insulin and glycogen and leptin levels, which has been reported in diabetic rats treatment of L. plantarum NCU116 (ref. 12) (Fig. 6). Except for improving insulin resistance, L. plantarum X1 also protected pancreas damage which induced by STZ (Fig. 8). Consistent with our study, L. casei CCFM0412 (ref. 5) and L. plantarum NCU116 (ref. 12) could efficiently ameliorated the impaired islets in type 2 diabetic mice. In addition, oral administration of L. plantarum X1 significantly decreased the levels of TNF-α and IL-6 and increased IL-10 in T2DM mice similar to the effect of pioglitazone (Fig. 7). TNF-α is an inflammatory cytokine that has been identified as an important factor participant in insulin resistance; IL-6 is an important cytokine involved in the inflammatory response; and IL-10 is an anti-inflammatory cytokine that inhibits antigen presentation and inflammatory cytokine production.^5,35 Agree with our discovery of pro-inflammatory and anti-inflammatory cytokines improvement by supplement of L. plantarum X1, the anti-inflammatory effect of L. casei CCFM0412 has been shown to effectively treat low-grade inflammation in T2DM.⁵ Previous reports demonstrated that supplement of L. casei CCFM0412,⁵ L. rhamnosus CCFM0528,¹³ L. plantarum CECT 7315/7316 (ref. 35) regulated systemic inflammatory response.

SCFAs have been linked to metabolic syndrome, such as T2DM and cardiovascular disease.¹² SCFAs in feces are mainly composed of acetic acid, propionic acid, and butyric acid. Propionic and butyric acids have been reported to reduce inflammation.¹² Thus, the decrease in the TNF-α and IL-6 levels and the increase in the IL-10 level might result from the production regulation of SCFAs from oral supplementation with L. plantarum X1. The SCFAs levels decreased and especially that the acetic acid level decreased in the M group and that supplementation with L. plantarum X1 could increase acetic acid and SCFAs levels compared to L. plantarum CCFM30 (Fig. 9). Interestingly, L. plantarum X1 and L. plantarum CCFM30 both significantly increased the butyric acid concentration in comparison to that in the C and M groups (Fig. 9). A previous study showed that the butyric acid concentration of fermented carrot juice supplemented with L. plantarum NCU116 was significantly increased and L. plantarum NCU116 have potential for the treatment of diabetes through change colonic pH.¹² These results are in accordance with the statement that butyric acid as a product of lactic acid bacteria fermentation which can act as a favorable factor in the growth of lactobacilli and bifidobacteria by lowering intestinal pH.³⁶

SCFAs and gut microbiota are closely linked. The abundance of Firmicutes was markedly increased by supplementation with L. plantarum X1, which was consistent to previous study that db/db diabetic mice were associated with an increase in the phylum Firmicutes, Proteobacteria and Fibrobacteres compared to lean mice.³⁷ Oral administration of L. plantarum X1 significantly increased Bacteroides, which was a kind of SCFA-producing bacteria (Fig. 10). Previous findings suggested that berberine and metformin enriched SCFA-producing bacteria, including Allobaculum, Bacteroides, Blautia, Butyricicoccus, and Phascolarctobacterium.³⁸ The main metabolite of butyric acid bacteria is butyric acid, which was conducive to the growth of beneficial bacteria and inhibited the growth of harmful bacteria.¹² Therefore, gut microbiota and SCFAs could regulate the intestinal microecology and further improve metabolic diseases such as T2DM and obesity. Thus, compositional disorder in intestinal microbiota of T2DM mice could be partial repaired by administration of L. plantarum X1.

T2DM could be induced by multiply factors. Among the factors, oxidative stress, which is mainly reflected in increasing production of free radicals and impairment of antioxidant defenses, was reported to play an important role in the pathogenesis and progression of T2DM and its complications.³² MDA is an end product and marker of the lipid peroxidation process.¹² SOD is an important antioxidant enzyme in organisms that can scavenge free radicals in the biological body and GSH is also reported to be one of the most important intracellular antioxidants.¹² It has been shown in previous studies that mice with untreated T2DM have a notable decrease in SOD activity and GSH concentration and an abnormally increased MDA level.⁵ In this study, both strains administrated to T2DM mice showed certain effects on antioxidant (Table 3). This may be due to the antioxidant properties of lactic acid bacteria itself, which was the subject of our previous study.¹¹ Ejtahed et al. reported that the probiotic yogurt significantly reduced MDA level and increased SOD and GSH-Px activities and further improved antioxidant status in type 2 diabetic patients.³²

Hyperglycemia and oxidative stress are related to dyslipidemia, and dyslipidemia is also a primary cause of complications of T2DM. Supplementation with L. plantarum X1 and L. plantarum CCFM30 could ameliorate lipid profiles at different extents (Table 3). These results agreed with the reports that L. plantarum NCU116 (ref. 12) and L. rhamnosus CRL981 (ref. 39) decreased the levels of serum TC, TG, and LDL-C and increased the HDL-C level. Comparing the effects on mitigation of dyslipidemia, administration of L. plantarum X1 showed more efficient than L. plantarum CCFM30 on releasing of LDL-C level which similar to the effect of pioglitazone. It is well known that high levels of serum TC and LDL-C can increase the risk of cardiovascular events. In this study, oral administration of L. plantarum X1 showed significant positive effects in decreasing the serum TC and LDL-C levels. So this strain has potential in reducing the risk of cardiovascular disease (Table 3).

In this study, L. plantarum X1 administrated to diabetic mice reduced fasting and postprandial blood glucose, improved glucose tolerance and insulin resistance, regulated inflammatory cytokine, and changed the SCFAs concentration and gut microbiota. It was reported that L. rhamnosus CCFM0528 was effective for reducing fasting and postprandial blood glucose and enhanced OGTT.¹³ Oral administration of L. reuteri GMNL-263 improved insulin resistance and abrogated the increasing level of TNF-α and IL-6 in high fructose-fed rats.¹⁴ Otherwise, previous report demonstrated that oral administration of L. plantarum NCU116 resulted in the promotion of SCFAs production in rat feces.¹² Beside bacteria, some yeast, such as Saccharomyces boulardii, also be reported that its administration could change gut microbiota in obese and type 2 diabetic db/db mice.⁴⁰ However, in this study, L. plantarum X1 only partly attenuated oxidative stress and dyslipidemia compared with L. casei CCFM0412 in previous study.⁵ To make up for this deficiency, other strains with excellent antioxidative activity and hypolipidemic function should be administrated combined with L. plantarum X1, which could prevent and recover the type 2 diabetes or other diseases more efficiently.

Conclusions

Higher α-glucosidase inhibitory activity in the cell free supernatant suggested that L. plantarum X1 had the potential ability to prevent and treat diabetes by reducing postprandial blood glucose. Evidences proved that L. plantarum X1 could more effectively alleviate the symptom of T2DM, ameliorate insulin resistance, and protect pancreatic function in an animal model than L. plantarum CCFM30. Otherwise, L. plantarum X1 also showed more excellent effect on regulation of inflammatory response and gut microbiota and SCFAs composition. In addition, as probiotics, both L. plantarum X1 could partly alleviate the relevant diabetic such as dyslipidemia and oxidative damage. Our study provided five potential pathways contributing to the beneficial effects of L. plantarum X1 on hypoglycemic in T2DM, which including dyslipidemia, oxidative stress, α-glucosidase, gut microbiota, and inflammatory (Fig. 11). These results suggested that L. plantarum X1 could be as a potential probiotics to ameliorate diabetes.

Fig. 11 Potential hypoglycemic mechanism of L. plantarum X1 on T2DM.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 31301407), the Key Projects in the National Science and Technology Pillar Program during the 12th Five-Year Plan (Nos. 2012BAD12B08), the National Basic Research Program of China (973 Program No. 2012CB720802), the Program of Introducing Talents of Discipline to Universities (B07029), the Fundamental Research Funds for the Central Universities (JUSRP51501), a project funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions and the Program of Collaborative Innovation Center of Food Safety and Quality Control in Jiangsu Province.

References

1. A. D. Assoc, Diabetes Care, 2013, 36, S67–S74.
2. K. Murotomi, A. Umeno, M. Yasunaga, M. Shichiri, N. Ishida, T. Koike, T. Matsuo, H. Abe, Y. Yoshida and Y. Nakajima, J. Agric. Food Chem., 2015, 63, 6715–6722.
3. O. J. Phung, W. L. Baker, V. Tongbram, A. Bhardwaj and C. I. Coleman, Ann. Pharmacother., 2012, 46, 469–476.
4. A. J. Krentz and C. J. Bailey, Drugs, 2005, 65, 385–411.
5. P. Chen, Q. Zhang, H. Dang, X. Liu, F. Tian, J. Zhao, Y. Chen, H. Zhang and W. Chen, Nutrition, 2014, 30, 1061–1068.
6. A. A. Tahrani, M. K. Piya, A. Kennedy and A. H. Barnett, Pharmacol. Ther., 2010, 125, 328–361.
7. S. Bolen, L. Feldman, J. Vassy, L. Wilson, H. C. Yeh, S. Marinopoulos, C. Wiley, E. Selvin, R. Wilson, E. B. Bass and F. L. Brancati, Ann. Intern. Med., 2007, 147, 386–399.
8. J. H. Park, R. Y. Kim and E. Park, Food Sci. Biotechnol., 2012, 21, 1187–1193.
9. M. Bhat, S. S. Zinjarde, S. Y. Bhargava, A. R. Kumar and B. N. Joshi, J. Evidence-Based Complementary Altern. Med., 2011, 1–6,  DOI:10.1093/ecam/nen040.
10. M. Liu, H. Yin, G. Liu, J. J. Dong, Z. H. Qian and J. L. Miao, J. Agric. Food Chem., 2014, 62, 5548–5554.
11. P. Chen, Q. Zhang, H. Dang, X. Liu, F. Tian, J. Zhao, Y. Chen, H. Zhang and W. Chen, Food Control, 2014, 35, 65–72.
12. C. Li, Q. Ding, S. P. Nie, Y. S. Zhang, T. Xiong and M. Y. Xie, J. Agric. Food Chem., 2014, 62, 11884–11891.
13. P. Chen, Q. X. Zhang, H. Dang, X. M. Liu, F. W. Tian, J. X. Zhao, Y. Q. Chen, H. Zhang and W. Chen, J. Funct. Foods, 2014, 10, 318–326.
14. F. C. Hsieh, C. L. Lee, C. Y. Chai, W. T. Chen, Y. C. Lu and C. S. Wu, Nutr. Metab., 2013, 10, 1–14.
15. Y. Zhang, L. F. Wang, J. C. Zhang, Y. X. Li, Q. W. He, H. Li, X. Guo, J. L. Guo and H. P. Zhang, Eur. J. Nutr., 2014, 53, 221–232.
16. K. Laitinen, T. Poussa, E. Isolauri and N. A. M. I. In, Br. J. Nutr., 2009, 101, 1679–1687.
17. S. Parvez, K. A. Malik, S. A. Kang and H. Y. Kim, J. Appl. Microbiol., 2006, 100, 1171–1185.
18. D. C. Eva Lau, C. Pina-Vaz, J.-A. Barbosa and P. Freitas, Hormones, 2015, 14, 358–369.
19. J. Qin, Y. Li, Z. Cai, S. Li, J. Zhu, F. Zhang, S. Liang, W. Zhang, Y. Guan, D. Shen, Y. Peng, D. Zhang, Z. Jie, W. Wu, Y. Qin, W. Xue, J. Li, L. Han, D. Lu, P. Wu, Y. Dai, X. Sun, Z. Li, A. Tang, S. Zhong, X. Li, W. Chen, R. Xu, M. Wang, Q. Feng, M. Gong, J. Yu, Y. Zhang, M. Zhang, T. Hansen, G. Sanchez, J. Raes, G. Falony, S. Okuda, M. Almeida, E. LeChatelier, P. Renault, N. Pons, J.-M. Batto, Z. Zhang, H. Chen, R. Yang, W. Zheng, S. Li, H. Yang, J. Wang, S. D. Ehrlich, R. Nielsen, O. Pedersen, K. Kristiansen and J. Wang, Nature, 2012, 490, 55–60.
20. G. Alp and B. Aslim, Anaerobe, 2010, 16, 101–105.
21. J. Mu, A. Petrov, G. J. Eiermann, J. Woods, Y. P. Zhou, Z. Li, E. Zycband, Y. Feng, L. Zhu, R. S. Roy, A. D. Howard, C. Li, N. A. Thornberry and B. B. Zhang, Eur. J. Pharmacol., 2009, 623, 148–154.
22. A. Kautzky-Willer, J. Harreiter, D. Bancher-Todesca, A. Berger, A. Repa, M. Lechleitner and R. Weitgasser, Wien. Klin. Wochenschr., 2016, 128, S103–S112.
23. S. M. Haffner, A. S. Greenberg, W. M. Weston, H. Z. Chen, K. Williams and M. I. Freed, Circulation, 2002, 106, 679–684.
24. M. Vijay-Kumar, J. D. Aitken, F. A. Carvalho, T. C. Cullender, S. Mwangi, S. Srinivasan, S. V. Sitaraman, R. Knight, R. E. Ley and A. T. Gewirtz, Science, 2010, 328, 228–231.
25. B. Mao, D. Li, J. Zhao, X. Liu, Z. Gu, Y. Q. Chen, H. Zhang and W. Chen, J. Agric. Food Chem., 2015, 63, 856–863.
26. L. L. Zhao, G. Wang, P. Siegel, C. He, H. Z. Wang, W. J. Zhao, Z. X. Zhai, F. W. Tian, J. X. Zhao, H. Zhang, Z. K. Sun, W. Chen, Y. Zhang and H. Meng, Sci. Rep., 2013, 3, 1970–1975.
27. L. Ramchandran and N. P. Shah, LWT--Food Sci. Technol., 2010, 43, 146–152.
28. Z. Zeng, J. Y. Luo, F. L. Zuo, Y. Zhang, H. Q. Ma and S. W. Chen, J. Funct. Foods, 2016, 20, 486–495.
29. S. V. Dilna, H. Surya, R. G. Aswathy, K. K. Varsha, D. N. Sakthikumar, A. Pandey and K. M. Nampoothiri, LWT--Food Sci. Technol., 2015, 64, 1179–1186.
30. H. Yadav, S. Jain and P. R. Sinha, Nutrition, 2007, 23, 62–68.
31. T. Manaer, L. Yu, Y. Zhang, X. J. Xiao and X. H. Nabi, J. Ethnopharmacol., 2015, 169, 269–274.
32. H. S. Ejtahed, J. Mohtadi-Nia, A. Homayouni-Rad, M. Niafar, M. Asghari-Jafarabadi and V. Mofid, Nutrition, 2012, 28, 539–543.
33. K. Honda, M. Moto, N. Uchida, F. He and N. Hashizume, J. Clin. Biochem. Nutr., 2012, 51, 96–101.
34. Z. S. Xie, T. L. Truong, P. Zhang, F. G. Xu, X. J. Xu and P. Li, J. Ethnopharmacol., 2015, 172, 70–79.
35. K. W. Moore, R. D. Malefyt, R. L. Coffman and A. O'Garra, Annu. Rev. Immunol., 2001, 19, 683–765.
36. W. Z. Mehal, Nat. Rev. Gastroenterol. Hepatol., 2013, 10, 637–644.
37. L. Geurts, V. Lazarevic, M. Derrien, A. Everard, M. Van Roye, C. Knauf, P. Valet, M. Girard, G. G. Muccioli, P. Francois, W. M. de Vos, J. Schrenzel, N. M. Delzenne and P. D. Cani, Front. Microbiol., 2011, 2, 149–165.
38. X. Zhang, Y. Zhao, J. Xu, Z. Xue, M. Zhang, X. Pang, X. Zhang and L. Zhao, Sci. Rep., 2015, 5, 14405.
39. J. A. Marazza, J. G. LeBlanc, G. S. de Giori and M. S. Garro, J. Funct. Foods, 2013, 5, 1848–1853.
40. A. Everard, S. Matamoros, L. Geurts, N. M. Delzenne and P. D. Cani, mBio, 2014, 5, e01011–01014.

---