Abstract

   Background: Fermented milk is beneficial for metabolic disorders, while
   the underlying mechanisms of action remain unclear. This study explored
   the benefits and underlying mechanisms of Bifidobacterium longum 070103
   fermented milk (BLFM) in thirteen-week high-fat and high-sugar (HFHS)
   fed mice using omics techniques. Methods and results: BLFM with
   activated glucokinase (GK) was screened by a double-enzyme coupling
   method. After supplementing BLFM with 10 mL/kg BW per day, fasting
   blood glucose, total cholesterol (TC), low-density lipoprotein
   cholesterol (LDL-C), and leptin were significantly reduced compared
   with the HFHS group. Among them, the final body weight (BW), epididymal
   fat, perirenal fat, and brown fat in BLFM group had better change
   trends than Lacticaseibacillus rhamnosus GG fermented milk (LGGFM)
   group. The amplicon and metabolomic data analysis identified
   Bifibacterium as a key gut microbiota at regulating glycolipid
   metabolism. BLFM reverses HFHS-induced reduction in bifidobacteria
   abundance. Further studies showed that BLFM significantly reduces the
   content of 3-indoxyl sulofphate associated with intestinal barrier
   damage. In addition, mice treated with BLFM improved BW, glucose
   tolerance, insulin resistance, and hepatic steatosis. Conclusion: BLFM
   consumption attenuates obesity and related symptoms in HFHS-fed mice
   probably via the modulation of gut microbes and metabolites.

   Keywords: Bifidobacterium longum fermented milk, glucokinase,
   glycolipid metabolism disorder, gut microbiota, fecal metabolites

1. Introduction

   Changes in lifestyle and diet diversification have led to increasing
   occurrences of abnormal glucose and lipid metabolism and metabolic
   diseases [[52]1]. Gut microbiota is closely related to the host’s diet
   and metabolic parameters [[53]2]. Excessive glycolipid intake is
   related to various tissue and organ dysfunctions, increased intestine
   permeability, and gut microbiota imbalance [[54]3,[55]4]. In
   particular, a high-fat diet is significantly related to the ratio of
   Firmicutes to Bacteroides (F/B) [[56]5]. Mice that were fed Myrciaria
   dubia prevented a decrease in gut microbiota species richness caused by
   high-fat (HF) diets and high-sugar (HS) diets, reduced the ratio of
   F/B, and significantly increased Akkermansia and Bifidobacterium
   [[57]6] which are negatively correlated with metabolic diseases such as
   obesity [[58]7]. Rats fed a HF diet and exhibiting metabolic syndrome
   that were treated with Bifidobacterium adolescentis had reduced
   visceral fat accumulation [[59]8]. Gut microbiota controls metabolic
   disorders caused by a HF diet by regulating oxidative stress,
   inflammation, and metabolites [[60]9,[61]10,[62]11]. In vitro selection
   of probiotic strains to improve glucose and lipid metabolism disorders
   is mostly based on their redox ability and anti-inflammatory and
   glycolipid-related drug regulatory targets [[63]12,[64]13,[65]14].
   Glucokinase is the first rate-limiting enzyme of glycolysis regulating
   blood glucose levels in both directions between liver glucose
   metabolism and insulin secretion [[66]15]. Glucokinase activator
   SHP289-04 significantly improves hyperglycemia and hyperlipidemia in
   KKA^y mice [[67]16] and glucokinase mediates the expression of insulin
   receptor substrate 2 to alleviate islet β-cell apoptosis caused by
   endoplasmic reticulum stress [[68]17]. There are few reports of
   probiotics as new drug targets. Fermented milk is a suitable carrier
   for probiotics which contains strains and biologically active compounds
   produced during fermentation [[69]18]. The beneficial effects of L.
   rhamnosus GG have been extensively studied and have been shown to have
   good effects, especially in glycolipid metabolism [[70]19,[71]20].
   There are many reports on the improvement of glucose and lipid
   metabolism disorders [[72]21,[73]22]. However, there is a need for
   systematic research on the improvement of glucose and lipid metabolism
   parameters and the gut microbiota and its metabolites. This study
   analyzed the relationship between probiotic fermented milk that
   activates glucokinase and disorders of glucose and lipid metabolism,
   intestinal barrier, gut microbiota, and its metabolites, and proposed a
   plan for improving glucose and lipid metabolism disorders.

2. Materials and Methods

2.1. Preparation of Probiotic Fermented Milk

   A total of 87 lactic acid bacteria (LAB) strains were isolated from
   food and the feces of healthy people. Human fecal strains are derived
   from the previous isolation of the research group which received
   ethical approval [[74]23]. The DNA of the strains were identified using
   universal primers 27F: 5′-AGAGTTTGATCCTGGCCTCA-3′; 1492R:
   5′-GGTTACCTTGTTACGACTT-3′ and Sanger sequencing (Beijing Liuhe Huada
   Gene Technology Co., Ltd., Beijing, China) and stored in the bacterial
   strain bank of the Food Safety and Health Team of the Institute of
   Microbiology, Guangdong Academy of Sciences, China. Bacteria was
   cultured anaerobically in sterile tryptone-plant peptone-yeast extract
   broth (TPY) at 37 °C. Skim milk (SM) solution (12% w/v) was prepared by
   resuspending the powder (Solarbio, Beijing, China) in deionized water,
   heat-treated in a sterilization pot (Hirayama HVE-50, Tokyo, Japan) at
   105 °C for 15 min, and cooled to room temperature. LAB strains (5%
   inoculum) were added to SM and cultured anaerobically at 37 °C until
   the curd was observed for subsequent experiments.

2.2. Heterologous Expression and Purification of Glucokinase

   The amino acid sequence of human liver glucokinase (GenBank:
   [75]AAB97682.1) was entrusted to Jinweizhi Company (Suzhou Jinweizhi
   Biotechnology Co., Ltd., Suzhou, China) for gene synthesis, cloned into
   pET-28A vector, sequenced (Suzhou Jinweizhi Biotechnology Co., Ltd.),
   and transformed into BL21 Escherichia coli. His-tagged recombinant GK
   protein was overexpressed by induction with 0.5 mM isopropyl
   β-D-1-thiogalactopyranoside (IPTG) (Sanggong Bioengineering Co., Ltd.,
   Shanghai, China) at 18 °C for 48 h. Cells were centrifuged at 9000× g
   for 20 min at 4 °C, resuspended in lysis buffer (1 M NaCl, 50 mM
   phosphate buffer saline, 70 mM imidazole, 1 mg/mL lysozyme, pH = 7.6),
   and stored at −80 °C. Expression was performed according to a previous
   method with modifications [[76]24]. Cells were thawed, sonicated on ice
   using an Ultrasonic cell crusher, and centrifuged at 15,000× g for 30
   min at 4 °C. The supernatant was filtered with a 0.8 μm filter and
   placed on a 5 mL HisTrap^TM FF protein purification column, and eluted
   with elution buffer (150 mM NaCl, 50 mM phosphate buffer saline, 200 mM
   imidazole, pH = 7.2). The eluted proteins were concentrated by
   ultrafiltration MWCO 30 KDa (Millipore, Burlington, MA, USA). Protein
   purity was detected by SDS-PAGE.

2.3. In Vitro Glucokinase Enzymatic Assays

   The curd sample was taken, centrifuged at 4500× g for 5 min at 4 °C,
   and the supernatant obtained. The pH was adjusted to 7.2 with 1 moL/L
   NaOH solution and centrifuged at 12,000× g for 10 min at 4 °C. Then,
   the supernatant was filtered through a 0.22-μm membrane. The filtrate
   was stored at −80 °C. Glucokinase activity was measured using the
   6-phosphate glucose dehydrogenase coupling method by monitoring the
   amount of reduced β-nicotinamide adenine dinucleotide (NADH) at A340 nm
   [[77]25]. By adding fermented milk to the enzyme reaction system to
   detect the activity of purified recombinant human liver glucokinase,
   the fermented milk with activation effect was screened for subsequent
   experiments.

2.4. Animal Experiments

   Forty SPF male C57/BL6J mice obtained from Guangdong Medical
   Experimental Animal Center, China were raised in a barrier environment
   (23.3 °C, 50–60% relative humidity, 12 h/12 h light/dark cycle) with
   free access to food and water. Mice were adapted for 1 week, then
   divided into four groups: the control group was fed a basal diet and
   normal saline levels, the HFHS group was fed a high-fat and high-sugar
   diet and normal saline; the LGGFG group was fed a HFHS diet and 10
   mL/kg BW Lacticaseibacillus rhamnosus GG fermented milk per day, while
   the BLFM group was fed with a HFHS diet containing 10 mL/kg BW B.
   longum 070103 fermented milk per day. The basic feed had 3530 kcal/kg
   and the energy supply ratio was 20.6% protein, 12% fat, and 67.4%
   carbohydrate. The HFHS group had 4250 kcal/kg and the energy supply
   ratio was 16.46% protein, 45.65% fat, and 37.85% carbohydrate. All
   groups were continuously fed for 13 weeks, and mice BWs were recorded
   weekly. Mice feces was collected three days before the end of the
   experiment. At the end of the 13th week, the mice were fasted
   overnight, the weight recorded, anesthetized with carbon dioxide, the
   eyeballs removed for blood collection, and the mice were immediately
   killed by cervical dislocation. The liver, kidney, spleen, perirenal
   fat, epididymal fat, brown fat, ileum, and colon were dissected,
   harvested, washed with cold normal saline, and dried on filter paper.
   The liver, kidney, spleen, perirenal fat, epididymal fat, and brown fat
   were weighed to calculate the organ index (visceral weight/fasted
   weight) and fat index (different fat weight/prohibited weight after
   eating). The tissues of the same sized liver, epididymal fat, brown
   fat, and colon of each mouse were removed and fixed in 10% formalin for
   subsequent hematoxylin and eosin (H & E) staining. The remaining tissue
   was quickly placed on dry ice and stored at −80 °C. All experimental
   protocols were approved by the Experimental Animal Ethics Committee of
   the Institute of Microbiology, Guangdong Academy of Sciences
   (Guangzhou, Guangdong, China; approval number GT-IACUC202104301).

2.5. Oral Glucose Tolerance Test (OGTT)

   This was performed at the end of the 12th week of the experiment. Mice
   were fasted overnight and weighed before the test, then gavaged with
   20% w/v glucose solution at 2 g/kg BW. Blood glucose values were
   measured at 0, 30, 60, 90, and 120 min after gavage with glucose
   solution using a blood glucose meter (Yuwell, Nanjing, China), and the
   area under the curve (AUC) was calculated.

2.6. Determination of Serum Metabolites and Intestinal Barrier Index

   A blood glucose meter (Yuwell, Shanghai, China) was used on the last
   day of the experiment to detect fasting blood glucose levels in mice.
   The collected whole blood was coagulated at room temperature for two
   hours, centrifuged at 3500× g for 15 min and 4 °C, and the supernatant
   (serum) stored at −80 °C. Serum levels of total cholesterol (TC),
   thyroglobulin (TG), high density lipoprotein-C (HDL-C), low density
   lipoprotein-C (LDL-C), aspartate transferase (AST), and alanine
   transferase (ALT) were measured using a Mindray BS-480 automatic
   biochemical analyzer (Mindray Co., Ltd., Shenzhen, China), while
   insulin, leptin, and adiponectin levels were determined using a
   commercial enzyme-linked immunosorbent assay (ELISA) kit (Dongge Boye
   Biotechnology Co., Ltd., Beijing, China). The insulin resistance index
   (HOMA-IR) was expressed as fasting blood glucose (mmol/L) × fasting
   insulin (mIU/mL)/22.5.

   Briefly, 50 mg of colon sample was taken and placed in a cryogenic
   grinder together with 0.45 mL physiological saline according to a
   weight-to-volume ratio of 1:9 (g/mL), grinded for 45 s at 60 Hz, 2
   times). The supernatant was collected by centrifugation at 13,000× g
   for 15 min at 4 °C. A commercial ELISA kit was used to determine ZO-1
   and claudin-1 amounts in colon tissue supernatant (Dongge Boye
   Biotechnology, Beijing, China).

2.7. Histopathological Examination

   Dissected mouse liver, epididymal fat, brown fat, and colon tissue were
   fixed with 10% formalin solution, embedded in paraffin, sliced into
   3-mm thick sections using a semi-automatic paraffin microtome (Thermo
   Fisher Scientific, Waltham, MA, USA), and stained with H & E.
   Pathological changes were observed with a digital biological microscope
   (Nanjing Jiangnan Novel Optics Co., Ltd., Nanjing, China) using
   ScopeImage 9.0 software (Ningbo Yongxin Optics Co., Ltd., Ningbo,
   China).

2.8. Metabolome Analysis of Fecal Metabolites

   Fecal metabolites were identified according to a previous method
   [[78]26] with slight modifications. Fifty mg of mouse feces was placed
   in a centrifuge tube. Then, 200 μL of extract (2:2:1
   methanol:acetonitrile:water) was added, vortexed for 120 s, sonicated
   at 4 °C for 20 min, incubated at −20 °C for 1 h, and centrifuged at
   13,800× g for 15 min at 4 °C. The supernatant was evaporated in a
   vacuum drying oven at room temperature until dryness. The samples were
   reconstituted with 200 μL of 50% acetonitrile solution, then 20 μL of
   each sample were mixed as QC samples, and 3 μL of the spiked samples
   were injected into an Accucore C18 column (100 × 2.1 mm, 2.6 μm; Thermo
   Fisher Scientific, Waltham, MA, USA) using a Vanquish chromatography
   system and a Q Exactive™ Plus Hybrid Quadrupole-Orbitrap™ Mass
   Spectrometer (Thermo Fisher Scientific, Waltham, MA, USA). The mixture
   was separated using the following elution gradient: 0 min, 2% B; 7 min,
   50% B; 19 min, 80% B; 21 min, 98% B; 23 min, 98% B; 23.5 min, 0.2% B
   where A is either 0.1% formic acid (positive ion mode) or 5 mM acetic
   acid (negative ion mode), and B is acetonitrile. The samples were
   scanned in positive and negative ion mode at a flow rate and column
   temperature of 0.3 mL/min and 35 °C, respectively. The following
   acquisition parameters were used: resolution, 70,000; automatic gain
   control (AGC target), 1×10^6; scanning range, 70–1000 m/z; spray
   voltage, 3.5 kV. The complete MS/dd-MS2 (topN) from the full mass
   spectrum was obtained from the five strongest m/z values of the two
   ionization modes to obtain MS2 spectra of each compound using the
   following settings: resolution, 17,500; AGC target, 1 × 10^5; maximum
   IT, 50 ms; isolation window, 1.5 m/z, normalized collision energy: 20,
   40, and 60 eV. The data were analyzed using Compound Discoverer 3.1
   software (Thermo Fisher Scientific, Waltham, MA, USA) for baseline
   correction, peak differentiation, and alignment [[79]27] and the result
   (CSV file) was exported to Microsoft Excel. MetaboAnalyst 5.0
   ([80]https://www.metaboanalyst.ca/MetaboAnalyst/home.xhtml. accessed on
   05 September 2022) was used for principal component analysis (PCA),
   partial least squares discriminant analysis (PLS-DA), and orthogonal
   partial least squares discriminant analysis (OPLS-DA). Student’s t-test
   was used to calculate the statistical significance (p < 0.05) and fold
   change (FC) between metabolites. Finally, the differential metabolites
   were screened according to their projection (VIP) score. A VIP > 1.5, p
   < 0.01, and FC > 2 was considered statistically significant.
   Biochemical databases (HMDB, METLIN, and PUBMED) were used to identify
   potential markers. The metabolites with significant differences from
   the control were further analyzed for pathway enrichment analysis and
   KEGG analysis to determine the influence of consuming probiotic
   fermented milk on the fecal metabolites of mice on a HFHS diet using
   the model organism mice as a reference.

2.9. Gut Microbiota Analysis

   Bacterial genomic DNA was extracted from the colon contents using the
   E.Z.N.A.^® stool DNA kit (Omega Bio-tek, Norcross, GA, USA) and
   detected by 1% agarose gel electrophoresis. The V3-V4 variable region
   of the 16S rRNA gene was amplified by polymerase chain reaction (PCR)
   using primers 338F (5′-ACTCCTACGGGAGGCAGCAG-3′) and 806R
   (5′-GGACTACHVGGGTWTCTAAT-3′) by an ABI GeneAmp^® 9700 PCR thermocycler
   (ABI, Los Angeles, CA, USA). Cycling conditions of initial denaturation
   were 95 °C for 3 min, followed by 23 cycles of denaturing at 95 °C for
   30 s, annealing at 55 °C for 30 s, extension at 72 °C for 45 s, a
   single extension at 72 °C for 10 min, and end at 4 °C. PCR products
   were quantified with a NanoDrop 2000 spectrophotometer (Thermo Fisher
   Scientific) and separated using 2% agarose gel electrophoresis. Gel
   products were purified using an AxyPrep DNA gel extraction kit (Axygen
   Biosciences, Union City, CA, USA), quantified with a Quantus
   fluorometer (Promega, Madison, WI, USA), and detected by 2% agarose gel
   electrophoresis. A NEXTflex rapid DNA-Seq kit (Bioo Scientific, Austin,
   TX, USA) was used for library construction. Sequencing was performed
   using the Illumina MiSeq PE300 platform (Shanghai Meiji Biomedical
   Technology Co., Ltd., Shanghai, China).

   After demultiplexing, the resulting sequences were merged with FLASH
   (1.2.11) [[81]28] and quality filtered with fastp (0.19.6) [[82]29].
   The high-quality sequences were de-noised using the DADA2 [[83]30]
   plugin in the Qiime 2 [[84]31] (version 2020.2) pipeline with the
   recommended parameters to obtain single nucleotide resolution based on
   error profiles within samples. The sequencing depth reached 97.90%
   after noise reduction and leveling of each sequence and an ASV table
   was obtained based on the comparison of the Silva 16S rRNA gene
   database (v138). Then, species diversity analysis, principal component
   analysis, principal coordinate analysis, and non-metric analysis were
   mostly scaled through the diversity cloud analysis platform
   (cloud.majorbio.com (accessed on 6 December 2021)) of Shanghai Meiji
   Biomedical Technology Co., Ltd., Shanghai, China. The Kruskal–Wallis
   rank sum test was used to obtain species information with significant
   differences among multiple groups.

2.10. Detection Colonization Ability of B.longum 070103 in Mouse Feces by
Quantitative Real-Time PCR

   The ability of B.longum 070103 to colonize the mouse gut was assessed
   by the Light Cycler^® 96 real-time PCR system (Roche, Basel,
   Switzerland). For B.longum DNA amplification, the following primers
   were used: Forward 5′-AGCCCTGAAAGAAACAACCA-3′ and Reverse
   5′-GAACAAGGCAAGGTCCAGTC-3′. DNA amplification reaction conditions were
   denaturation at 95 °C for 120 s, followed by 45 cycles of denaturation
   at 95 °C for 20 s, and annealing at 60 °C for 60 s. The data were
   analyzed using the built-in software. All fecal DNA of BLFM was
   extracted for qPCR analysis in triplicate.

2.11. Statistical Analysis

   All data are expressed as the mean ± standard deviation with SPSS 22.0
   (IBM, Armonk, NY, USA)and Graph Pad 8.2.1 software (GraphPad Software,
   San Diego, CA, USA) used to perform one-way analysis of variance and
   determine significant differences between groups; p-values < 0.05 were
   considered to be statistically significant.

3. Results

3.1. B. longum 070103 Fermented Milk Increased Glucokinase Activity

   Recombinant glucokinase was obtained by purification using a Ni^2+
   affinity column ([85]Supplementary Figure S1) and the effect of
   fermented milk from different bacteria on glucokinase activity was
   evaluated by coupling with glucose 6-phosphate dehydrogenase. B. longum
   070103 fermented milk had a positive effect on glucokinase activation
   ([86]Table 1).

Table 1.

   Screening of glucokinase activator in vitro.
   Strain Name Strain Number Strain Source Activation Multiple
   Bifidobacterium longum 070103 human feces 1.16 ± 0.03
   Bifidobacterium bifidum 070507 human feces 1.10 ± 0.05
   Bifidobacterium longum 070106 human feces 0.81 ± 0.06
   Lactobacillus delbrueckii R37 fermented dairy products 1.09 ± 0.08
   Lactiplantibacillus plantarum R51 fermented dairy products 1.04 ± 0.08
   Lactiplantibacillus plantarum R95 fermented dairy products 1.07 ± 0.04
   Lactiplantibacillus plantarum 1–134 fermented dairy products 1.06 ±
   0.03
   Levilactobacillus brevis SR52-2 salted fish 1.05 ± 0.03
   Lactiplantibacillus plantarum R22 yogurt 0.89 ± 0.06
   Lactiplantibacillus plantarum 0101127-4 pickle 0.93 ± 0.07
   [87]Open in a new tab

   The information on the remaining 77 strains of lactic acid bacteria is
   placed in [88]Supplementary Table S1.

3.2. B. longum Fermented Milk (BLFM) Improves Obesity Caused by HFHS

   The BW of mice in the HFHS group significantly increased during the
   entire experimental period compared with the control group ([89]Figure
   1A). However, mice supplemented with LGGFM and BLFM showed a decreasing
   trend compared with HFHS, and BLFM significantly reduced the BW
   ([90]Figure 1B). In addition, the liver, kidney, spleen, perirenal fat,
   and epididymal fat coefficients of the HFHS group were significantly
   higher than control group, but there was no significant difference in
   the weight of brown fat ([91]Table 2). The liver and epididymal fat in
   the BLFM group were significantly reduced after intervention, while the
   remaining organs and tissues showed a decreasing trend compared to the
   HFHS group, but the difference was insignificant ([92]Table 2).

Figure 1.

   [93]Figure 1
   [94]Open in a new tab

   Effects of supplementing fermented milk on BW and weight gain: (A) BW
   change in mice at 13 weeks, (B) BW gain. Data are presented as means ±
   SD (n = 10) and analyzed using one-way ANOVA (* p < 0.05, **** p <
   0.0001).

Table 2.

   Indexes of organ and fat coefficient of mice in each group (%).
   Group Control HFHS LGGFM BLFM
   Liver index 3.90 ± 0.05 ^a 4.20 ± 0.13 ^b 3.78 ± 0.06 ^a 3.86 ± 0.08 ^a
   Renal index 1.26 ± 0.03 ^a 1.40 ± 0.03 ^b 1.22 ± 0.03 ^a 1.36 ± 0.02 ^b
   Spleen index 0.27 ± 0.01 ^a 0.30 ± 0.01 ^b 0.26 ± 0.01 ^a 0.28 ± 0.01
   ^ab
   Perirenal fat coefficient 0.44 ± 0.06 ^a 1.35 ± 0.18 ^bc 1.53 ± 0.21 ^c
   1.01 ± 0.09 ^b
   Epididymal fat coefficient 1.45 ± 0.16 ^a 4.31 ± 0.35 ^c 4.10 ± 0.43 ^c
   3.14 ± 0.30 ^b
   Brown fat coefficient 0.39 ± 0.03 ^a 0.43 ± 0.03 ^a 0.36 ± 0.02 ^a 0.36
   ± 0.04 ^a
   [95]Open in a new tab

   Values are means ± SD, n = 10. Means in a row without a common letter
   differ. Values with superscript letters ^a, ^b and ^c are significantly
   different across rows (p < 0.05).

3.3. BLFM Improves Hyperglycemia and Dyslipidemia in HFHS-Fed Mice

   The effect of BLFM supplementation on glucose and lipid metabolism
   disorders in HFHS fed mice was tested by analyzing the relevant
   markers. A glucose tolerance test was performed at the end of the 12th
   week. The fasting blood glucose, AUC (glucose tolerance), fasting
   insulin, HOMA-IR index, serum LDL-C, TG, and TC concentrations in the
   HFHS group significantly increased compared with the control group
   ([96]Figure 2A–D,F,H). The HFHS group had severe glucose intolerance
   and dyslipidemia. In contrast, HFHS mice supplemented with LGGFM and
   BLFM showed significantly improved metabolic parameters after
   intervention compared with the HFHS group. The serum LDL-C and TC
   concentrations in the BLFM group significantly decreased ([97]Figure
   2F,G), the HDL-C concentration significantly increased ([98]Figure 2E),
   and the TG content showed an insignificant downward trend ([99]Figure
   2H). In addition, the decreased AUC of the OGTT indicates a significant
   improvement in glucose tolerance. There was no significant difference
   between LGGFM and BLFM in improving glucose and lipid metabolism
   indicators, suggesting that BLFM can also improve glucose and lipid
   metabolism disorders compared with LGGFM.

Figure 2.

   [100]Figure 2
   [101]Open in a new tab

   Supplementing fermented milk improves glucose and lipid metabolism
   disorders in mice: (A) fasting blood glucose, (B) OGTT and AUC, (C)
   fasting serum insulin, (D) HOMA-IR, (E) HDL-C, (F) LDL-C, (G) total
   cholesterol, (H) triacylglycerol. Data are presented as means ± SD (n =
   10) and analyzed using one-way ANOVA (* p < 0.05, ** p < 0.01, *** p <
   0.001, **** p < 0.0001).

3.4. BLFM Improves HFHS Induced Liver Steatosis, Inhibits White Adipose
Tissue (WAT) Proliferation

   A HFHS diet caused liver steatosis of the HFHS group through obvious
   lipid accumulation according to histological analysis of H&E-stained
   liver ([102]Figure 3A), while intervention with LGGFM and BLFM
   significantly reduced the size and number of lipid droplets. ALT and
   AST are biomarkers for hepatotoxicity and early metabolic syndrome
   [[103]32,[104]33,[105]34]. Serum ALT and AST levels were significantly
   higher in the HFHS group than in the control group. Serum ALT and AST
   levels were significantly reduced in mice supplemented with LGGFM and
   BLFM, and BLFM was reduced to a greater degree than LGGFM ([106]Figure
   3D,E). This shows that BLFM supplementation reduces liver damage caused
   by HFHS. Histopathological analysis revealed that the lipid droplet
   size in the epididymal fat and brown adipose tissue of the HFHS group
   was significantly larger than that of the control group. The adipose
   tissue had smaller lipid droplets after intervention of the HFHS group
   with LGGFM and BLFM treatment ([107]Figure 3B,C). High plasma leptin
   levels are correlated with body fat mass [[108]35], with leptin
   resistance caused by decreased tissue sensitivity to prolonged high
   circulating leptin levels [[109]36] and increased susceptibility to
   obesity. Supplementation with LGGFM and BLFM significantly reduced
   leptin resistance caused by obesity in mice fed a HFHS diet
   ([110]Figure 3F).

Figure 3.

   [111]Figure 3
   [112]Open in a new tab

   Effect of supplementing fermented milk on pathological changes of
   liver, adipose tissue and colon tissue in mice: (A–C) H&E of liver
   tissue sections, epididymal white adipose tissue (epididymal WAT),
   brown fat (BAT), (D,E) glutamic pyruvic transaminase (ALT) and glutamic
   oxalacetic transaminase (AST), (F) leptin. Data are presented as means
   ± SD (n = 10) and analyzed using one-way ANOVA (** p < 0.01, *** p <
   0.001, **** p < 0.0001).

3.5. BLFM Modulates Gut Microbiota

   Since glucose and lipid metabolism disorder is closely related to gut
   microbiota, this study examined the relationship between the
   supplementation of fermented milk in the diet and the gut microbiota.
   The significantly lower Shannon index in the HFHS group compared with
   that in the control group indicates that the community diversity and
   richness of the samples were lower in the HFHS group ([113]Figure 4A).
   A significant increase in the gut microbiota diversity was observed
   after LGGFM and BLFM intervention ([114]Figure 4B–D) with Firmicutes,
   Bacteroides, and Actinomycetes being the most abundant taxa. The F/B
   ratio in the HFHS group was significantly increased compared with the
   control group. LGGFM and BLFM intervention significantly and
   insignificantly decreased the F/B, respectively ([115]Figure 4E,F). The
   main microbiota consisted of norank_f_Muribaculaceae, Allobaculum,
   Lactobacillus, Dubosiella, Faecalibaculum, and Bifidobacterium
   ([116]Figure 4G). The abundance of norank_f_Muribaculaceae,
   Bifidobacterium, Alloprevotella, norank_f_Oscillospiraceae, Alistipes,
   Bacteroides, and Parabacteroides in the HFHS group was lower than those
   in the control group, while Allobaculum and Faecalibaculum abundance
   was higher than those of the control group. Supplementation of LGGFM
   and BLFM to the HFHS group significantly increased the abundance of
   norank_f_Muribaculaceae, Bifidobacterium, and other bacteria
   ([117]Figure 4H). At the same time, the feces of mice fed with BLFM
   were evaluated to detect the colonization ability of B. longum, and the
   number was 4.36 ± 0.013 lg/g of feces ([118]Supplementary Figure S2).

Figure 4.

   [119]Figure 4
   [120]Open in a new tab

   FM modulates the gut microbiota: (A) Shannon index, (B–D) Plots of
   PCoA, NMDS, and PCA scores based on the Bray-Curties distance
   algorithm, (E) Relative abundance of gut microbiota at the phylum
   level, (F) Ratio of Firmicutes to Bacteroidetes, (G) Relative abundance
   of gut microbiota at the genus level, (H) Relative abundance of
   bacterial communities at the genus level among different groups, (I)
   H&E of colon tissue sections, (J) The content of Claudin-1 in the
   colon, (K) The content of ZO-1 in the colon. Data are presented as
   means ± SD (n = 10) and analyzed using one-way ANOVA (ns, not
   significant, * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001).

   In addition, the HFHS group had sparse intestinal villi and the glands
   were reduced compared with the control group indicating that HFHS
   caused colon damage. LGGFM and BLFM intervention reverted the
   intestinal villi to their dense form and significantly the gland was
   restored to a similar size as that of the control ([121]Figure 4I).
   ELISA showed that colonic ZO-1 and Claudin-1 levels were significantly
   lower in the HFHS group than those in the control group, indicating
   that the integrity of the intestinal mucosal barrier was destroyed with
   a HFHS diet. ZO-1 and Claudin-1 significantly increased after LGGFM and
   BLFM intervention compared with the HFHS group ([122]Figure 4J,K).
   Therefore, BLFM intervention significantly improved mucosal
   permeability and restored the integrity of the intestinal mucosal
   barrier by regulating tight junction proteins.

3.6. BLFM Improves HFHS Induced Fecal Metabolites in Mice

   Changes in the gut microbiota inevitably affect the biotransformation
   of dietary components by the gut microbiota. The contents of
   metabolites in feces in a HFHS diet following intervention with BLFM
   were analyzed using a non-targeted metabolomics approach with LC-MS.
   The analysis of PCA ([123]Figure 5A), PLS-DA ([124]Figure 5B), and
   OPLS-DA ([125]Figure 5C) showed that the distribution of the samples
   was significantly different, suggesting that the HFHS diet
   significantly changed the fecal metabolites of mice compared with the
   normal diet. A VIP value > 1.5, FC > 2, and t test p value < 0.01
   revealed 44 different metabolites between the HFHS group and BLFM group
   according to the discriminant analysis of the orthogonal partial least
   squares method ([126]Supplementary Table S2). Enrichment analysis and
   pathway enrichment analysis of the 44 different metabolites showed that
   most of the metabolites were concentrated in the metabolism of alanine,
   aspartic acid, glutamic acid, pantothenic acid, coenzyme A
   biosynthesis, citric acid cycle, β-alanine metabolism, purine
   metabolism, arginine biosynthesis, and butyrate metabolism ([127]Figure
   5D). The contribution ratios of different metabolites to metabolic
   pathways are shown in [128]Figure 5E. After intervention with BLFM, the
   abundance of 22 feces metabolites significantly increased, including
   4-phenylbutyric acid, oleanolic acid, pentadecanoic acid,
   docosatrienoic acid, 13-hydroxy-α-tocopherol, and L-(+)-aspartic acid.
   Meanwhile, a significant reduction in the abundance of 22 metabolites
   was observed, including ɣ-hydroxybutyric acid (GHB), 3-indoxyl
   sulphate, and other substances. The heat map analysis illustrates the
   altered levels of the differential metabolites in the HFHS and BLFM
   groups ([129]Figure 5F).

Figure 5.

   [130]Figure 5
   [131]Open in a new tab

   Supplementation of fermented milk changes the fecal metabolites and
   related KEGG pathways of mice on HFHS diet: (A) principal component
   analysis (PCA), (B) Partial Least Squares Discriminant Analysis
   (PLS-DA), (C) Orthogonal Partial Least Squares Discriminant Analysis
   (OPLS-DA), (D) KEGG pathway enrichment histogram, (E) Overview of
   pathway analysis, (F) heatmap of differential metabolites.

3.7. Correlation among Gut Microbiota, Fecal Metabolites, and Pathological
Indicators

   Spearman’s correlation coefficient analysis revealed significant
   correlation between 14 physiological and biochemical indicators and the
   top 50 genera of gut microbiota in the HFHS and BLFM groups. In
   particular, Bifidobacterium was negatively correlated with WG, AUC,
   insulin, HOMA-IR, HDL-C, LDL-C, TC, TG, AST, ALT, and leptin, and
   positively correlated with claudin-1 ([132]Figure 6A). The abundances
   of norank_f__norank_o__Rhodospirillales, norank_f__Peptococcaceae, and
   norank_f__UCG-010 were negatively correlated with AUC, insulin,
   HOMA-IR, LDL-C, TC, AST, ALT, and leptin, and positively correlated
   with claudin-1. The abundances of UBA1819, Lachnospiraceae_UCG-001, and
   Colidextribacter were negatively correlated with insulin, HOMA-IR,
   LDL-C, TC, AST, and ALT. Claudin-1 was positively correlated with the
   abundances of UBA1819 and Colidextribacter, while AUC was negatively
   correlated with these bacteria. Leptin was negatively correlated with
   UBA1819 and Lachnospiraceae_UCG-001.

Figure 6.

   [133]Figure 6
   [134]Open in a new tab

   Correlation analysis of gut microbiota, fecal metabolites, and glucose
   and lipid metabolism indexes: (A) Heatmap of correlations between
   metabolic parameters and the top 50 genera in gut microbiota abundance,
   (B) Heatmap of correlations between nine important differential fecal
   metabolites and the top 50 genera in gut microbiota abundance, (C)
   Heatmap of correlations between metabolic parameters and important
   differential fecal metabolites. The legend on the right is the color
   range for different R values, red for positive correlation and blue for
   negative correlation. (* p < 0.05, ** p < 0.01, *** p < 0.001).

   Correlation analysis of the abundance of fecal metabolites and gut
   microbiota showed that Bifidobacterium was positively correlated with
   13-hydroxy-alpha-tocopherol, docosatrienoic acid, oleanolic acid,
   pentadecanoic acid, L-(+)-aspartic acid, 4-phenylbutyric acid, and
   2-arachidonoylglycerol, while a negative correlation was observed with
   Bifidobacterium and 3-indoxyl sulphate or GHB. The abundance of
   norank_f__UCG-010 and norank_f__norank_o__Rhodospirillales was
   positively correlated with 13-hydroxy-alpha-tocopherol, docosatrienoic
   acid, oleanolic acid, pentadecanoic acid, and L-(+)-aspartic acid,
   while a negatively correlation was observed with 3-indoxyl sulphate. It
   is worth noting that the abundances of Rikenella and Parvibacter were
   proportional to 3-indoxyl sulphate ([135]Figure 6B).

   The nine metabolites were significantly correlated with at least one
   physiological and biochemical index with a total of 65 significantly
   correlations. Among them, 3-indoxyl sulphate was positively correlated
   with FBG, AUC, insulin, HOMA-IR, LDL-C, TC, AST, ALT, and leptin. There
   was a negative correlation between the expression of claudin-1 and
   ZO-1. Claudin-1 and ALT were negatively correlated with
   13-hydroxy-alpha-tocopherol, docosatrienoic acid, oleanolic acid,
   pentadecanoic acid, L-(+)-aspartic acid, 4-phenylbutyric acid, and
   2-arachidonoylglycerol. Meanwhile, they were positively correlated with
   3-indoxyl sulphate and GHB ([136]Figure 6C). These results indicate
   that BLFM reduces the production of 3-indoxyl sulphate by increasing
   the abundance of bifidobacteria and metabolites, such as docosatrienoic
   acid and oleanolic acid. This partially regulates the gut microbiota,
   improves the intestinal barrier, and alleviates glucose and lipid
   metabolism disorders caused by HFHS.

4. Discussion and Conclusions

   The health benefits of fermented dairy products are closely related to
   the strains and metabolites produced during fermentation, especially
   the presence of active ingredients such as probiotics, peptides,
   conjugated linoleic acid, etc. [[137]37]. Glucokinase is the
   rate-limiting enzyme in the first step of glycolysis. A small change in
   its activity alters the glucose-stimulated insulin secretion threshold
   of pancreatic β-cells and regulates the steady state of glucose
   [[138]38]. Supplementation of probiotic fermented milk has a positive
   effect on obesity and diabetes [[139]39,[140]40,[141]41,[142]42].

   In this study, a screening system was constructed by the heterologous
   expression and purification of human liver glucokinase in E. coli.
   Secondly, probiotic strains isolated from the feces of healthy
   individuals were used to ferment milk. Finally, the supernatant of
   fermented milk was used to screen the fermented milk with glucokinase
   activation. This research is the first attempt to screen fermented milk
   with probiotics that activate glucokinase. BLFM significantly improved
   glucose and lipid metabolism disorders in mice caused by high fat and
   high sugar levels by regulating gut microbiota and changing metabolite
   composition.

   B. longum 070103 fermented milk significantly reduced the weight gain
   induced by high-fat and high-sugar diets, improved the steady state of
   glucose and lipid metabolism, and reduced fatty liver degeneration and
   white fat proliferation. This is consistent with other studies on
   fermented milk with probiotics [[143]22,[144]43,[145]44,[146]45]. This
   study also proved that BLFM significantly increased the production of
   claudin-1 and ZO-1 and repaired the intestinal barrier. Impaired
   intestinal barrier integrity is considered to be an important cause of
   various systemic metabolic diseases such as insulin resistance,
   obesity, and type 2 diabetes [[147]46,[148]47,[149]48,[150]49]. The
   improvement of the intestinal barrier was consistent with the
   improvement in the colon tissue observed by H & E staining. The
   intestinal barrier is closely related to the gut microbiota and
   intestinal metabolites. Therefore, it is possible that the regulation
   of intestinal microbes leads to changes in metabolites, repairs the
   intestinal barrier, and improves the disorder of glucose and lipid
   metabolism [[151]50]. In our mouse model, the overall structure of the
   gut microbiota changes after BLFM intervention, thereby improving the
   imbalance of the microbiota induced by HFHS diet. Moreover, changes in
   the gut microbiota structure are related to physiological and
   biochemical indicators of glucose and lipid metabolism disorders.

   Differences in diet can lead to changes in the gut microbiota. Gut
   microbiota imbalance is the most important factor affecting the
   intestinal barrier and is associated with various metabolic diseases
   [[152]51]. The intake of different food components has a significant
   effect on microbe levels [[153]2]. For example, the intake of navy
   beans and broccoli reduced the F/B ratio [[154]9,[155]52], and our
   supplementation with BLFM similarly reduced the F/B ratio compared.
   Therefore, we can intervene in the treatment of certain diseases
   through food. The high-fat/carbohydrate diet significantly reduces the
   abundance and diversity of the gut microbiota with significantly
   reduced Bacteroides leading to obesity and dyslipidemia [[156]53]. In
   contrast, increasing the abundance of such bacteria alleviates disease.
   Supplementation of obese women with inulin-type fructans reduces the
   metabolic risk factors associated with higher fecal short-chain fatty
   acid concentrations, and significantly enriches B. longum, pseudo-chain
   Bifidobacterium, Bifidobacterium adolescentis, and Bacillus, with B.
   longum negatively correlated with serum lipopolysaccharide and
   endotoxin [[157]54]. Bifidobacterium is considered a target for
   regulating obesity microbiota [[158]55]. In this study, Spearman’s
   correlation analysis was used to determine the degree of correlation
   between changes in gut microbiota following supplementation with
   fermented milk, and host metabolic parameters. Supplementation with
   fermented milk significantly increased the abundance of
   norank_f_Muribaculaceae, Bifidobacterium, Alloprevotella,
   norank_f__Oscillospiraceae, Alistipes, Bacteroides, and
   Parabacteroides. These bacteria alleviate glucose and lipid metabolism
   disorders caused by obesity and a high-fat diet
   [[159]56,[160]57,[161]58,[162]59,[163]60] by the production of succinic
   acid and secondary bile acids. In contrast, Allobaculum and
   Faecalibaculum were significantly enriched in the HFHS group, which are
   associated with the development of metabolic diseases [[164]3,[165]61],
   and supplementation with fermented milk reduced the abundance of these
   bacteria. In this study, supplementation of fermented milk to the mouse
   diet significantly increased the abundance of Bifidobacteria which
   alleviating metabolic disorders. Bifidobacteria were negatively
   correlated with WG, AUC, insulin, HOMA-IR, HDL-C, LDL-C, TC, TG, AST,
   ALT, and leptin and positively correlated with claudin-1. In summary,
   supplementation with BLFM specifically regulates key gut microbiota,
   thereby improving host metabolism.

   There is a high correlation between the gut microbiome and the fecal
   metabolome which is associated with host health and diet [[166]62]. A
   high-fat diet destroys the intestinal barrier through various
   mechanisms such as the destruction of the tight junctions of the
   intestinal epithelium, imbalance of gut microbiota, oxidative stress,
   and inflammation [[167]63]. The permeability of the intestinal barrier
   aggravates metabolic disorders caused by a high-fat diet
   [[168]64,[169]65]. Hyperglycemia increases intestinal permeability in
   obese and diabetic mice leading to microbial metabolites entering the
   host and increasing the risk of infection [[170]66]. In this study, the
   high-fat diet group significantly increased the content of 3-indoxyl
   sulphate and 4-hydroxybutyric acid in the feces. Moreover, 3-indoxyl
   sulphate was strongly correlated with all metabolic parameters. This
   metabolite inhibits DPR1 (dynamin-related protein 1) in intestinal
   epithelial cells by enhancing IRF1 (interferon regulatory factor 1)
   expression, inhibiting mitochondrial autophagy, and inducing intestinal
   barrier damage [[171]67]. This substance is also considered biomarkers
   for patients with type 2 diabetic atherosclerosis [[172]68]. GHB is an
   inhibitory neurotransmitter that can inhibit glucagon production.
   However, studies have shown that it can stimulate the secretion of
   human glucagon in contrast to impaired regulation of glucagon secretion
   in type 2 diabetes patients with hyperglucagonemia [[173]69]. BLFM
   supplementation significantly reduced 3-indoxyl sulphate and
   4-hydroxybutyric acid levels while increasing oleanolic acid,
   pentadecanoic acid, and L- (+)-aspartic acid. These metabolites are
   related to the alleviation of glucose and lipid metabolism disorders
   [[174]70,[175]71,[176]72,[177]73].

   In summary, BLFM counteracted glucose and lipid metabolism disorders
   caused by high fat and sugar by adjusting the gut microbiota, changing
   fecal metabolites, and repairing the intestinal barrier. BLFM regulated
   the gut microbiota, increased F/B, significantly increased the
   abundance of Bifidobacteria, and promoted the production of
   pentadecanoic acid, 4-phenylbutyric acid, and docosatrienoic acid in
   mice with metabolic disorders induced by diet. Trienoic acid, oleanolic
   acid, 13-hydroxy-α-tocopherol, and L-(+)-aspartic acid production
   significantly reduced GHB and 3-indoxyl sulphate and repaired the
   damaged intestinal barrier. These results indicate that B. longum
   070103 fermented milk can be considered a dietary supplement to
   alleviate glucose and lipid metabolism disorders.

Supplementary Materials

   The following supporting information can be downloaded at:
   [178]https://www.mdpi.com/article/10.3390/nu14194050/s1, Figure S1:
   SDS-PAGE of G6PDH and GK. Figure S2: Colonization ability of
   Bifidobacterium longum 070103. Table S1: Screening of glucokinase
   activator in vitro (77 samples). Table S2: Differential metabolites
   between BLFM and HFHS groups.
   [179]Click here for additional data file.^ (198KB, zip)

Author Contributions

   T.J., Y.L., L.L., T.L. and M.D. contributed equally to this work. T.J.,
   Y.L., L.L., T.L. and M.D. conducted the animal study, wrote the
   manuscript. L.Y., J.Y., R.Y. and H.Z., performed data analysis and
   organized the literature and figures. M.C., Y.D., J.Z., J.W., X.X. and
   Q.W. provided funding and reviewed/edited the manuscript. All authors
   have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

   All experimental involving human participants were reviewed and
   approved by Ethics Committee of The First Affiliated Hospital of
   Guangdong Pharmaceutical University (reference 2017-98). The
   experimental protocols were approved by the Experimental Animal Ethics
   Committee of the Institute of Microbiology, Guangdong Academy of
   Sciences (Guangzhou, Guangdong, China; approval number
   GT-IACUC202104301).

Informed Consent Statement

   Informed consent was obtained from all subjects involved in the study.

Data Availability Statement

   16S amplicon data from animal feces of C57/BL6J mice
   [180]https://submit.ncbi.nlm.nih.gov/subs/sra/SUB12024050/overview
   (accessed on 7 September 2022) Submission ID: SUB12024050, BioProject
   ID: PRJNA877441.

Conflicts of Interest

   The authors declare no conflict of interest.

Funding Statement

   This work was financially supported by the National Key R&D Program
   Project of China (2021YFA0910200), the National Natural Science
   Foundation of China (21977020), the Key R&D Program of Guangdong
   Province (2018B020205002), the Innovation Development Project of
   Guangdong Academy of Sciences (2020GDASYL-20200301002) and the Project
   by the Department of Science and Technology of Guangdong Province
   (2019QN01N107).

Footnotes

   Publisher’s Note: MDPI stays neutral with regard to jurisdictional
   claims in published maps and institutional affiliations.

References