Exopolysaccharides from a Scandinavian fermented milk viili increase butyric acid and Muribaculum members in the mouse gut

Highlights • Bioactivity of viili exopolysaccharide (VEPS) were determined in young male mice.• Mice drank daily tap water supplemented with VEPS, stressless administration.• A low dose VEPS modified mildly but significantly the gut microbiota.

Various structures of EPS produced by lactic acid bacteria have been determined (Laws, Gu, & Marshall, 2001;Zannini, Waters, Coffey, & Arendt, 2016). Most of these structures are determined for EPS produced by Lactobacillus and Streptococcus. Some EPSs contain only one type of monosaccharide (homopolysaccharide) and others consist of more than 2 types of monosaccharide (heteropolysaccharide). EPS is a large polymer (molecular mass of 10 4 -10 7 Da) consisting of several hundred oligosaccharide units, and the repeat unit is composed of 3-7 monosaccharides linked by αor β-glycoside bonds. The repeat unit has a principal chain (back bone) and one or more than one branch, which is chemically linked to the back bone. Because determination of the chemical structure of heteropolysaccharide is not an easy task, only a few structures of VEPS are known.
Recently, bioactivities of EPSs produced by lactic acid bacteria have been intensively investigated to find health benefits of EPSs (Korcz, Kerenyi, & Varga, 2018;Saadat, Khosroushahi, & Gargari, 2019). For example, in vitro experiments using cells, EPS (50-200 μg/mL in cell culture medium) showed various beneficial bioactivities such as immunoregulation (Kitazawa, Yamaguchi, & Itoh, 1992), anti-tumor activity (Ishiguro et al., 2017), antioxidant activity (Ling et al., 2012), and α-amylase inhibitory activity (Ayyash et al., 2020). In vivo experiments, EPS showed immunoregulation (Vinderola, Perdigón, Duarte, Farnworth, & Matar, 2006) and regulation of intestinal barrier function (Zhou et al., 2018). In these mouse experiments, a large dose of EPS (about 100 mg EPS/kg body weight/day) was administered to muse by oral gavage. Fermented milk products such as yogurts, kefir, and viili are believed to be safe and beneficial to human health. Their beneficial effects are partly due to live lactic acid bacteria (probiotics) and partly due to metabolites produced by the bacteria (Kareb & Aїder, 2019). EPS is one of the metabolites and the heteropolysaccharide content is 50-200 mg/L (Badel, Bernardi, & Michaud, 2011). The intake of EPS in the reported mouse experiments corresponds to 6 g/day in the case of adult human. To find beneficial effects of EPS on human health from the standpoint of functional foods and investigate their mechanism, in vivo experiments using practical dose of EPS are inevitable.
The composition the gut microbiota and host diet is linked closely. Morrison et al. demonstrated that in both female and male mice the reduction of soluble fiber in diet changes substantially the gut microbiome composition, resulting in a loss of taxa within the phylum Bacteroidetes and expansion of Clostrdia and Proteobacteria (Morrison, Jašarević, Howard, & Bale, 2020). Dietary fibers are fermented by the microbiota in the cecum and colon to produce acetate, propionate and butyrate (Koh, De Vadder, Kovatcheva-Datchary, & Bäckhed, 2016;Morrison & Preston, 2016;McNabney, & Henagan, 2017). These shortchain fatty acids (SCFAs) are absorbed by colonocytes and used as energy source for these cells, and SCFAs transported into circulation act as signaling molecules (Dalile, Van Oudenhove, Vervliet, & Verbeke, 2019). Recently, effects of EPS on the gut microbiota in mouse obesity model (Lim et al., 2017) and liver injury model  have been reported. In the former model, the relative abundance of Bacteroides family and Akkermansia family was significantly increased in EPS group mice. In the latter model, pre-administration of EPS for 14 days prevented the gut flora dysbiosis after intraperitoneal injection of lipopolysaccharide and D-galactosamine. It should be noted that in these two experiments 5-140 mg of EPS (corresponding to 12-336 g/day in the case of adult human) was given to each mouse by oral gavage.
The purpose of the present study is to examine whether drinking tap water supplemented with EPS ad libitum affects the gut microbiota and SCFA production in normal mouse fed ordinary diet. First, we isolated EPS from viili (VEPS). The monosaccharide composition and the scanning electron microscopy (SEM) image of VEPS indicated that the present VEPS has a structure different from those reported for Lactococcus lactis. Second, the maximum dose of VEPS was determined by comparing the water intake of VEPS group mice with that of control group mice given tap water with no supplement. Finally, tap water containing 8 μg/ mL VEPS was administered to C57BL6 mice for 28 days. Mean daily intake of VEPS was 49 ± 1 μg/day. The low dose VEPS increased significantly the relative abundance of Muribaculum in the feces. After 28 days, the relative abundance of butyrate in the feces of VEPS group was higher than that of control group.

Isolation of EPS from viili
Viili was prepared as follows. The viili starter (1 g) was added to 1 L of cows' milk. After mixing by tumbling, the mixture was left to stand at 22.5 • C for 24 h until the milk solidified. To precipitate proteins, trichloroacetic acid (TCA) was added to the viili (final TCA concentration of 6.7%). The mixture was vigorously stirred at 4 • C for 1 h using a magnetic stirrer, and then the mixture was centrifuged at 12,000×g for 20 min at 4 • C. The supernatant was collected and mixed with 2 volumes of ethanol. The mixture was incubated at 4 • C for 12 h, and then the mixture was centrifuged at 12,000×g for 20 min at 4 • C. The precipitates were collected and resuspended in distilled water. The suspension was dialyzed against 1 L of distilled water at 4 • C for 1 day using a cellulose membrane with a molecular weight cut off of 10 kDa. Distilled water was exchanged 2 times during dialysis. The dialyzed sample (VEPS) was lyophilized using a freezing dryer (DRW040DA, ADVANTEC, Tokyo, Japan).

Animal experiments
C57BL/6J male mice at 6 weeks of age were purchased from Japan SLC and individually housed in mouse cages. For acclimatization, the mice were fed a normal diet (CE-2, CLEA Japan, Tokyo, Japan) and tap water for 2 weeks. The diet (100 g) contained 4.3 g dietary fibers. After 2 weeks of acclimatization, the mice were assigned to two groups (6 mice per group). The mice in the treatment group (EPS group) were allowed to drink ad libitum tap water containing 8 μg/mL of VEPS. The mice in the control group were allowed to drink tap water. The mice in both groups were fed the normal diet. On the 1st day and 28th day of treatment, feces were collected from individual mice. The feces were kept at − 70 • C until use. 2.9. All animal experiments were carried out in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals, and all of the protocols were approved by the Committee for Animal Research at Osaka Prefecture University (permit number 19-161).

Scanning electron microscopy (SEM) of VEPS
A small aliquot of lyophilized VEPS (1 cup of micro spatula) was treated with 4% paraformaldehyde in 5 mL of 0.1 M phosphate buffer (pH 7.4, buffer A) for 8 h at 4 • C. After incubation, the VEPS was precipitated by centrifugation. The precipitates were washed with 5 mL buffer A. The washed VEPS was collected by centrifugation and then treated with 1% tannic acid in 5 mL buffer A at 25 • C for 30 min. The VEPS was precipitated by centrifugation and the precipitates were washed with 5 mL buffer A and then washed with 5 mL distilled water at 25 • C for 30 min. The washed VEPS was dehydrated in an ethanol series: 50%, 70%, 90%, 95%, and finally tert-butanol (the alcohol solutions being exchanged every 30 min). The dehydrated VEPS was lyophilized, fixed in stubs on a double-faced metallic tape, and covered with a thin layer of gold using an ion sputter (E-1010, Hitachi, Tokyo, Japan) at 20 mA. The fixed samples were observed using a scanning electron microscope (SU1510, Hitachi, Tokyo, Japan) at an accelerating voltage of 15.0 kV.

Analysis of monosaccharide composition
VEPS (2 mg) was hydrolyzed in 0.1 mL of 2 mol/L sulfuric acid at 120 • C for 2 h. The solvent was evaporated and the hydrolysates were dissolved in 0.26 mL of 0.1 mol/L NaOH (solvent A), and then an aliquot of the solution (80 μL) was applied to a Dionex CarboPac PA1 IC column (4 × 250 mm) (Thermo Fisher Scientific, Waltham, MA). The column was developed at a flow rate of 1 mL/min using the following gradient: 0-5 min, 0% B; 5-35 min, 0-90% B (linear increase). Solvent B was solvent A containing 0.5 mol/L sodium acetate. The effluent was monitored by pulsed amperometric detection using a PAD detector (Dionex ICS-5000, Thermo Fisher Scientific). Authentic monosaccharide solutions (0-100 μM) were used to obtain calibration curves.

SDS-polyacrylamide gel electrophoresis (SDS-PAGE) of VEPS
SDS-PAGE was carried out using precast gels (Any kD precast gel, Bio-Rad) according to the manufacturer's instructions. Proteins were stained with Coomassie Brilliant Blue G-250 using CBB-G250 solution (Bio-Rad, CA, USA). To visualize polysaccharides, periodic acid-Schiff (PAS) staining (Zacharius, Zell, Morrison, & Woodlock, 1969) was performed. Briefly, immediately after the electrophoresis, the gels were fixed in 12.5% TCA for 30 min, rinsed lightly with distilled water, and then immersed in 1% periodic acid (in 3% acetic acid) for 50 min. After periodic acid treatment, the gels were washed in distilled water overnight (water being exchanged 3 times). The washed gels were immersed in fuchsin-sulfite solution (Sigma-Aldrich, MO, USA) in the dark for 50 min and then washed with 0.5% metabisulfite for 10 min. This washing was carried out 3 times. The gels were then washed in distilled water overnight and stored in 5% acetic acid.

Analysis of the composition of gut microbiota by 16S rRNA sequencing
On the 1st day and 28th day, feces were collected individually from mice in the control and VEPS groups. Extraction of DNA from feces, amplification of the V3-V4 region of 16S rRNA (sense primer, 5 ′ -TCGTCGGCAGCGTCAGATGTGTATAAGAGA-CAGCCTACGGGNGGCWGCAG-3 ′ ; antisense primer, 5 ′ -GTCTCGTGGGCTCGGAGATGTGTATAAGAGA-CAGGACTACHVGGGTATCTAATCC-3 ′ ), indexing by sequencing adapters, and sequencing using MiSeq (Illumina, San Diego, CA) were performed as described previously (Harada et al., 2020). Post-processing sequencing data were analyzed using the QIIME2 pipeline (Bolyen et al., 2019). After denoising using Deblur (Hyde, Gonzalez, & Knight, 2017), the filtered sequence data were classified against Silva (v132). Principal coordinate analysis (PCoA) and hierarchical cluster analysis were performed using R software. Bacterial 16SrRNA sequencing data sets have been deposited in DDBJ Sequence Read Archive under the accession number DRA012528.

Analysis of short chain fatty acids
Feces (about 3.0 g) were homogenized in 10 volumes (v/w) of distilled water using a polytron homogenizer, and the homogenate was centrifuged at 3000×g for 10 min. An aliquot of the supernatant (1 mL) was further centrifuged at 13,000×g for 10 min. The supernatant (500 μL) was transferred to a new 1.5-mL tube and mixed with 500 μL acetonitrile. The mixture was incubated on ice for 10 min and then centrifuged at 13,000×g for 10 min. An aliquot of the supernatant (40 μL) was mixed with 135 μL of 50% aqueous acetonitrile and 25 μL of 2.0 mmol/L heptanoic acid (in 50% aqueous acetonitrile). 2-NPH derivatization of SCFAs in this sample (200 μL) was performed according to the method used by Peters et al. (2004). The sample was mixed with 20 μL of 40 mmol/L 2-NPH•HCl (in 50% aqueous ethanol) and 40 μL of 125 mmol/L EDC•HCl (in 50% aqueous ethanol containing 1.5% pyridine). The mixture was incubated at 60 • C for 15 min. The reaction was stopped by cooling the mixture on ice. To remove excess ethanol and acetonitrile, the cooled mixture was subjected to evaporation for 30 min using a centrifugal evaporator. Salts and excess reagents in the resultant sample were removed by solid phase extraction using MonoSpinC18 (GL Science, Tokyo, Japan) as follows. The spin column was washed with 300 μL methanol and then equilibrated with 300 μL 0.1% aqueous TFA (solvent C). The sample was diluted with 300 μL solvent C, and the diluted sample was applied to the spin column. The column was washed with 300 μL solvent C twice. The 2-NPH-derivatized fatty acids were then eluted from the monolith with 50 μL of 0.1% TFA in 90% aqueous acetonitrile (solvent D): solvent C = 50:50 (v/v).
The eluted sample was 5-fold diluted with solvent C and a 20-μL aliquot of the diluted sample was applied to an InertSustain C18 column (2.1 mm × 100 mm, GL Science). The column was developed at the flow rate of 150 μL/min according to the following time program: 0-5 min, 10% B; 5-10 min, 10-20% B; 10-50 min, 20-90% B; 50-59 min, 90% B; 59-60 min, 90-10% B. The absorbance at 400 nm was monitored. Chromatograms were analyzed using the chromatography software Clarity (DataApex, Prague, Czech). Calibration curves of the fatty acids were constructed by plotting the peak area ratio of each fatty acid to heptanoic acid against concentration. HPLC analysis was performed using a Prominence HPLC system (Shimadzu, Kyoto, Japan).

Statistical analysis
Data are expressed as means ± S.E. of at least three replicates for each sample. Statistical analyses were performed using Statcel4 software (OMS, Tokyo, Japan). The difference between two groups was evaluated using analysis of variance (one-way ANOVA) followed by unpaired Student's t-test at p < 0.05. For comparison of multiple samples, the Tukey-Kramer test was used. The SDS-PAGE of the purified VEPS revealed no band stained with Coomassie blue (the left panel of Fig. 1A), indicating that VEPS contained no detectable levels of protein. A relatively broad PAS-stained band (the right panel of Fig. 1A) was detected near the upper edge of the separation gel. When 10 μg of VEPS was applied (the third lane in Fig. 1A), a thin PAS-stained band appeared just below the broad band, indicating that VEPS contained at least two polysaccharides with different molecular sizes.

Isolation and characterization of VEPS
The SEM image VEPS showed a structure of entangled mixture of long fibers and flakes (Fig. 1B). This structure is distinct from that of EPS from Lactococcus lactis F-mou in which a flake-like structure dominates (Nehal et al., 2019), that of EPS from Lactobacillus plantarum KX041 in which flakes with rough surface dominates (Xu, Cui, et al., 2019), and that of EPS from Lactococcus garvieae C47 in which stick-like and flakelike structures are layered (Ayyash et al., 2020). The fibrous nature of VEPS may be related to the ropy nature of viili.
As shown in Fig. 1C, VEPS contained rhamnose, glucose, galactose, and glucosamine in the molar ratio of rhamnose/glucose/galactose/ glucosamine = 3.6/1.0/3.4/5.2. We compared monosaccharide composition of the EPS produced by Lactococcus lactis species in Table 1. The EPS produced by L. Iactis subsp. cremoris SBT 0495 contains rhamnose, glucose, galactose, and phosphorus in the molar ratio of 1.3:1.8:2.1:1.0. (Nakajima, Hirota, Toba, Itoh, & Adachi, 1992). The VEPS isolated in this study contains much higher amount of glucosamine compared to other EPSs. The acetylation level of glucosamine in VEPS is remained to be determined. Elucidation of VEPS structure is important to understand its biological activities.  Each mouse was given pre-weighted diet and its remnants after 7 days were weighted to determine weekly food intake. Data are presented as mean ± SE.

Effects of the EPS on mice weight and food intake
The concentration of VEPS in tap water was determined by measuring the water intake of VEPS group mice. The control group mice drank 9.0 ± 0.3 mL tap water/day. When the VEPS concentration was increased up to 8.0 μg/mL, the daily water intake decreased to 6.1 ± 0.2 mL. Based on these results, we determined the maximal acceptable VEPS concentration in tap water to be 8.0 μg/mL. Fig. 2A provides the   Fig. 3. Composition of the fecal microbiota at the genus level A. Relative abundance of operational taxonomic units (OTUs) before (Day 0) and after intake of VEPS (Day 28). VEPS group mice (mouse no. 4-6 and 10-12) drank tap water containing 8.0 μg/mL VEPS for 28 days. Control group (mouse no. 1-3 and 7-9) drank only tap water. B-E. Mean relative abundance of bacteria belonging to Lactobacillus (B), Lachnospiraceae family (C), Muribaculum (D), and Bacteroidales order (E), respectively. Data expressed as mean ± SE. Asterisk indicates statistically significant difference (p < 0.05). F. Principal coordinate analysis (PCoA) plot. G. The hierarchical cluster analysis. In F and G, each feces sample is shown as a single circle. experimental design. Daily intake of about 50 μg VEPS did not affect the body weight gain (Fig. 2B). Throughout the experiment, the VEPS group mice ate the same amount of food (3.1-3.7 g/day) as the control mice did (Fig. 2C). Because the normal diet contained 4.3 % (w/w) dietary fibers, the mice took 130-160 mg dietary fibers per day. Therefore, the intake of VEPS (50 μg/day) is much lower than that of dietary fibers.
Lim et al. fed C57BL/6J mice with high-fat diet containing 5% (w/w) exopolysaccharides (EPS) isolated from kefir grains and observed the reduction in body weight gain of EPS-treated mice (Lim et al., 2017). In their experiments, daily intake of EPS was 140 mg, about 3000-fold higher dose compared to the present study. The different results on body weight gain indicate the importance of diet used and dose of EPS to investigate effects of EPS on C57BL/6J mice. Fig. 3A shows the bacterial composition at the genus level of each fecal sample. The dominant genera were Lactobacillus, those belonging to Lachnospiraceae family, Muribaculum, and those belonging to Bacteroidales order in all the fecal samples (35-80% of total bacteria) except that collected from mouse no. 7 on day 0, which showed abnormal microbial composition compared to other samples due to unknown reason. Principal coordinates analysis showed no significant clustering of fecal samples based on day and group (Fig. 3F). Phylogenetic tree analysis revealed no clustering of samples (Fig. 3G). These results indicate that small dose of VEPS did not induce a large alteration in the composition of the gut microbiota.

Effects of VEPS on mice fecal microbiota
Interestingly, the relative abundance of Muribaculum increased in all VEPS-treated mice (Fig. 3A). The mean relative abundance of Muribaculum of VEPS group mice increased from 6.3% to 16.2% (Fig. 3D).
No significant change was observed for other dominant bacterial groups (Fig. 3B, 3C, and 3E).
Muribaculaceae family, previously called as family S24-7, is a dominant bacterial group in the mouse gut (Seedorf et al., 2014;Lagkouvardos et al., 2016;Lagkouvardos et al., 2019). Muribaculum is the first described genus belonging to this family. Recently, two species, M. intestinale DSM 28989 T and M. gordoncarteri sp. nov. (DSM 108194 T ), were isolated from faces of conventionally raised C57BL/6J mice (Miyake et al., 2020). These bacterial cells are Gram-stain-negative, strictly anaerobic, lack oxidase and catalase activities, and have genes coding a variety of enzymes involving the degradation of polysaccharides. In colitis mouse model induced by dextran sulfate sodium, the relative abundance of Muribaculaceae was negatively correlated with pro-inflammatory cytokines, and positively correlated with the expression levels of tight junction proteins and mucin2 (Yan et al., 2019). Therefore, bacterial species belonging to Muribaculum seem to be important for keeping normal conditions of the mouse gut.
The effects of oral gavage of a large dose of EPS on the gut microbiota have been investigated. For example, Xu et al. administered about 8 mg EPS produced by Lactobacillus buchneri TCP016 to female BALB/c mice by daily gavage for 14 days (Xu, Aruhan et al., 2019). This EPS treatment did not modulate largely the gut microbiota but significantly alleviated the gut microbiome alterations caused by intraperitoneal injection of lipopolysaccharide/D-galactosamine (acute liver injury model). In an aging mouse model, daily administration of about 1 mg EPS produced by Lactobacillus plantarum YW11 by oral gavage prevented the decrease of the gut microbiota diversity of mice injected subcutaneously D-galactose solution (Zhang et al., 2017). These reports suggest that EPS is useful for improvement of the dysbiosis found in these diseases. However, the large dose of EPS is highly problematic in application to food Fig. 4. Composition of fecal SCFA and lactic acid. The mean relative abundance of acetic acid (A), propionic acid (B), butyric acid (C), lactic acid (D), valeric acid (E), isovaleric acid (F), and isobutyric acid (G) was calculated for control group mice (Control) and VEPS group mice (VEPS) on days 0 and 28. Data are expressed as mean ± SE. Asterisks indicate statistically significant differences (p < 0.05). supplement because 1-8 mg EPS/day corresponds to 3-24 g EPS/day in the case of human adult.

Effects of VEPS on mice fecal SCFAs
We measured SCFAs and lactic acid in feces. Unexpectedly, in control group mice, the relative abundance of butyric acid significantly decreased from 15.0 ± 1.7 % to 9.5 ± 1.3 %, whereas in VEPS group the relative abundance of butyric acid did not changed (Fig. 4C). Other SCFAs and lactic acid showed no significant change after 28 days (Fig. 3A, 3B, 3D-G). Acetic acid, propionic acid, and butyric acid are produced by bacteria metabolizing fiber non-digestible in the gut, and these fecal SCFAs decrease in aged mice (18-20 months) compared to young mice (2-3 months) (Lee et al., 2020). In the present study, mice were 2 months old on day 0, and became 3 months old at the end of the experiments. Therefore, aging is probably not the reason for the decrease in the fecal butyric acid in the control mice. Low dose VEPS seems to prevent the reduction of butyric acid in the mouse gut.

Exopolysaccharides with new monosaccharide composition (VEPS)
were isolated from viili. The highest dose of VEPS (50 μg/day) compatible with drinking ad libitum was administered via tap water to young male mice fed normal diet. VEPS specifically increased the relative abundance of the genus Muribaculum in the gut of microbiome. The decrease of fecal butyric acid during feeding was found and it was prevented by VEPS. This is the first report on the effects of low dose VEPS on the gut microbiome of normal young mice.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.