All sow milk samples were serially diluted in sterile phosphate-buffered saline (PBS) and inoculated anaerobically in De Man, Rogosa, and Sharpe (MRS) agar (Hopebio, Qingdao, China) at 37°C for 48–72 h (Xu and Kim, 2014). White colonies were picked and cultured in MRS broth at 37°C for 24 h (Ji et al., 2015). The crystal violet-stained colonies were observed under a microscope using the ×100 oil immersion objective lens. Gram-positive bacteria retained the color of crystalline violet staining, while Gram-negative bacteria lost the initial staining color of neutral red. Subsequently, strains were frozen at −80°C and stored in MRS broth with 50% glycerol until use.

The isolated strains were cultured in MRS broth at 37°C for 18 h under anaerobic conditions and identified via the analysis of their 16S rRNA gene sequences. Bacterial DNA was extracted using QIAamp DNA Stool Kit (Qiagen, Gaithersburg, MD, United States) according to the manufacturer’s protocols. The partial 16S rRNA genes was amplified by polymerase chain reaction (PCR) using universal primers 27F (5′-AGAGTTTGATCMTGGCTCAG-3′) and 1492R (5′-GGTTACCTTGTTACGACTT-3′). Reaction was carried out in a reaction volume of 50 μl, containing 25 μl of PrimeSTAR Max Premix 2× (TaKaRa, Shiga, Japan), 16.5 μl of ddH₂O, 0.5 μmol/l of each primer, and 50–100 ng of DNA template (Rahmdel et al., 2019). PCR was performed under the following conditions: initial denaturation at 96°C for 2 min, followed by 35 cycles of denaturation at 96°C for 10 s, annealing at 56°C for 15 s, and polymerization at 72°C for 1 min. Subsequently, the final polymerization was performed at 72°C for 10 min. PCR products obtained were sequenced by Sanger company (Sangon Biotechnology Co., Ltd., Shanghai, China).

The obtained 16S rRNA sequences were manually corrected and then subjected to Basic Local Alignment Search Tool Nucleotide (BLASTN) analysis to check the similarity with those already deposited sequences in the National Center for Biotechnology Information (NCBI). The retrieved 16S rRNA sequences were aligned using the ClustalW program. The phylogenetic tree of the aligned sequences was performed using the neighbor-joining (NJ) method in MEGA 7 software (Xu et al., 2019).

The isolated LAB strains were cultured for 24 h in modified MRS (MRS-MOPS) broth, in which inorganic phosphate (KH₂PO₄) was replaced by 0.65 g/l sodium phytate and 0.1 M3-[N-morpholino]propane sulfonic acid (Sangon Biotechnology Co., Ltd., Shanghai, China) (Raghavendra and Halami, 2009). The growth of bacteria in liquid culture media is commonly evaluated by measuring the optical density at 600 nm (OD600) by UV spectrophotometry. Colony-forming units (CFU) were determined by plate counting after gradient dilution (Simova et al., 2008). The bacterial suspension (8 μl of 10^7_10⁸ CFU/ml) was spotted on a modified MRS (mMRS) agar surface and incubated at 37°C for 24–72 h. Inorganic phosphate (KH₂PO₄) in MRS agar medium was replaced by 20 g/l MOPS, 2 g/l calcium chloride, and 2.5 g/l sodium phytate (Fischer et al., 2014). Anhydrous calcium chloride and sodium phytate were filtered through a 0.2-micron filter and added to the autoclaved medium at 55°C. Since the medium was opaque after phytic acid precipitation, the formation of a transparent zone was an indicator of phytic acid hydrolysis.

To avoid false-positive results due to hyaline circles caused by acid solubilization, the colonies on the plates were washed from the agar surface with double-distilled water after incubation and then soaked in 2% (w/v) aqueous cobalt chloride solution for 5 min. Furthermore, an equal volume of 6.25% (w/v) ammonium molybdate solution and 0.42% (w/v) ammonium metavanadate solution was used to replace cobalt chloride hexahydrate for 5 min incubation at room temperature and then removed to check the hydrolysis plate (Bhagat et al., 2020).

To further identify the phytase activity of the LAB, preliminarily screened strains were cultured in different centrifuge tubes containing 1 ml MRS broth at 37°C and 900 rpm for 24 h. Then, 50 μl of culture was transferred to 10 ml MRS broth and incubated at 37°C and 200 rpm for 72 h. The culture broth was frozen overnight at −80°C, then concentrated three times in an Ultra-Low Temperature Freezer and centrifuged at 4°C and 8,000 rpm for 10 min. The supernatant was taken and tested for phytase activity analysis with a phytase assay kit (Suzhou Comin Biotechnology Co., Ltd., Suzhou, China). Commercial phytase was directly added to MRS broth as a control. Strains with the highest phytase activity were selected for further study.

Four-week-old male Kunming mice (initial body weight 18–21 g) were purchased from Hunan SJA Laboratory Animal Co., Ltd. (Changsha, Hunan, China). The mice were placed under controlled environmental conditions (temperature 22 ± 2°C, relative humidity 55 ± 2%) with light/dark cycles of 12 h. All animals were acclimatized for 3 days prior to an experiment and had free access to food and water. Standard commercially available mouse food (Jiangsu Xietong Pharmaceutical Bioengineering Co., Ltd., Jiangsu, China) was fed separately, and its specific nutritional index is shown (Supplementary Table 1). The experimental protocol was reviewed and approved by the Animal Care and Use Committee (SYXK 2014-0007) of Hunan Normal University. All experiments were performed in accordance with relevant guidelines and regulations.

Thirty-two mice were randomly divided into four groups (Supplementary Table 2), with eight mice per group, and fed different diets for 28 days: (1) control, (2) PA, (3) PA + psm16, and (4) PA + phytase. Group 1 mice were fed the basal diet. Group 2–3 mice received 1% phytic acid supplementation in the basal diet. Group 4 mice were fed a basal diet with 1% phytate and 500 U/kg commercial phytase. Group 1, 2, and 4 mice were gavaged with 200 μl PBS, and group 3 mice were gavaged with strain psm16 (1 × 10⁹ CFU in 200 μl PBS). All groups of mice were gavaged once a day for the first 18 days and then once every 3 days for a total of 21 times. Animals were fasted for 12 h and sacrificed prior to sampling.

Blood samples were placed in plastic tubes containing heparin for at least 3 h and then centrifuged at 4,000 rpm for 15 min at 4°C. The serum was recovered and stored at −80°C to be used for mineral content determination. After blood sampling, the livers, kidneys, and ceca were taken and weighed and stored at −80°C until use. The cecum wall was flushed and weighed, and the cecum contents were used in a short-chain fatty acid (SCFA) assay and pH determination. Feces were collected for an SCFA assay during the last 3 days of the experiment.

The SCFA of concentration was measured using the method of gas chromatography, as reported earlier with minor modifications (Zhou et al., 2019). Stool and cecum content samples were homogenized and diluted with distilled water. After mechanical vibration and centrifugation at 15,000 rpm for 15 min at 4°C, a mixture of supernatant fluid and 25% metaphosphoric acid solution (9:1) was allowed to stand for 3 h. Then, the mixed solution was centrifuged at 15,000 rpm for 10 min and filtered through a membrane filter (pore size 0.45 μm). SCFAs were measured with a gas chromatography (Agilent Technologies 7890B System) equipped with a DB-FFAP column (30 m × 250 μm × 0.25 μm). The carrier gas was nitrogen (flow 0.8 ml/min, split ratio 50:1, volume of sampling 1 μl). The oven, detector, and injector temperatures were 220, 280, and 250°C, respectively. The SCFA content was quantified using an external standard curve method with standard solutions of acetate, propionate, butyrate, isobutyrate, isovalerate, and valerate (Wang et al., 2017).

The pH of the cecum contents was measured by centrifugation at 8,000 rpm for 1 min at 4°C, then testing 0.8 μl of supernatant with pH paper (pH 5–9) and observing the color change.

The levels of copper, iron, zinc, manganese, magnesium, phosphorus, and calcium in mouse liver, kidney, and plasma were determined using an inductively coupled plasma emission spectrometer (ICP-OES 5110, Agilent, Santa Clara, CA, United States) (Wan et al., 2018). The samples were completely dried at 60°C and ground into powder before pretreatment, and approximately 0.2 g frozen tissue and approximately 0.2 ml plasma sample were used in each experiment. Livers and kidneys were decolorized with 1 ml hydrogen peroxide, digested with nitric acid and perchloric acid at 180°C for 100 min, then dried at 260°C, redissolved in 1% HNO₃, filtered, and subjected to an ICP analysis. The detection wavelengths were Ca (λ = 317.933 nm), Cu (λ = 324.754 nm), Fe (λ = 238.204 nm), Mg (λ = 279.553 nm), Mn (λ = 257.610 nm), P (λ = 213.618 nm), and Zn (λ = 206.200 nm). The metal and trace element content data in the liver, serum, and kidney were expressed as ppm. The content of trace elements in different dietary groups was assayed by the company SGS-CSTC Standards Technical Services Co., Ltd. (Shanghai, China) (Supplementary Table 3).

Experimental data were initially compiled using the WPS Education Test Edition and then analyzed using IBM SPSS Statistics 20.0. The normality and homoscedasticity of the data were tested before performing a one-way ANOVA. Non-parametric tests were used to ensure the accuracy of the analysis for data that did not meet the assumption of normality and homogeneity of variance, and the Duncan and LSD methods were used to meet the requirements. Phytase data were processed with GraphPad Prism 8 (GraphPad Software, San Diego, CA, United States). The results were expressed as the mean ± standard error, and significance was defined as a p-value < 0.05.

After cultivation of microbes from sow milk samples, a total of 30 strains were isolated based on colony morphology and 16Sr RNA gene sequence to include as much diversity as possible at the species level. A phylogenetic tree was constructed using an NJ analysis with MEGA 7, which indicated the relative phylogeny of the isolates in comparison to reference strains. Lactobacillus (11/30), Pediococcus pentosaceus (9/30), and Lactococcus (8/30) were the most abundant species in the isolated microbial flora. Minor populations of other species such as Pediococcus acidilactici (1/30) and Leuconostoc lactis are (1/30) were also seen in the phylogenetic tree (Figure 1).

The isolates were screened for extracellular phytase production with a plate screening method. Psm6, psm15, psm16, and psm17 showed the ability to degrade phytic acid. Psm15 and psm17 are P. pentosaceus strains, psm16 is a Lactococcus lactis subsp. lactis strain, and psm6 is a L. lactis strain. Of the four isolates, strain psm16 showed the largest degradation transparent circle on the plate (Supplementary Figure 1) and had the strongest phytase activity (Figure 2). For further verification, microscopic observation indicated that isolated strain psm16 was cocci (Supplementary Figure 2). Strain psm16 was deposited in the China General Microbiological Culture Collection Center (CGMCC 22933).

The weights of liver, kidney, cecum, and cecum contents and the lengths of large intestine (LIL) and small intestine (SIL) in mice were measured and compared with the animal’s final weight (Table 1). The addition of phytic acid to the diet as well as the gavaged strain had no effect on the final weight of the mice. There were no significant differences in LIL, SIL, liver weight, kidney weight, and cecum weight due to the dietary treatments in mice. In contrast, the addition of strain psm16 significantly increased the weight of the cecum contents (p < 0.045), and the addition of commercial phytase had a profound effect on the cecum wall weight (p < 0.010). Compared with the basic group, the pH value in the cecum was relatively lower in the phytic acid group and the addition of psm16 restored the pH to the value of the control group and was acidic, but the pH value became slightly alkaline after the addition of phytase (Supplementary Figure 3).

We investigated whether the various diets influenced SCFA concentrations in mice. Compared to the control, there were changes in the cecal and fecal SCFA levels in mice fed with phytic acid (Table 2). Propionate, butyrate, isovalerate, valerate, and total SCFA were lower in the ceca of mice on phytic acid diets than on basal diet. In addition, acetic acid was significantly reduced. When mice were orally gavaged with strain psm16, the concentrations of acetate (p < 0.015), propionate (p < 0.014), butyrate (p < 0.036), and total SCFA (p < 0.009) in the ceca of mice were significantly increased. Compared with the group fed phytic acid alone, the groups fed phytic acid and treated with strain psm16 had higher fecal acetate (p < 0.034), propionate (p < 0.000), butyrate (p < 0.000), and total SCFA (p < 0.009) levels. Additionally, significant increases in the acetate and total SCFA levels were observed in the phytase group.

To verify whether strain psm16 affects the content of mineral elements in mice, we studied the calcium, magnesium, copper, zinc, iron, phosphorus, and manganese levels in the liver, kidney, and plasma. Mineral elements in the various diets were determined, and no differences were evident (Supplementary Table 3). In mice fed with phytic acid, the levels of Fe (p < 0.048) and Zn in the liver were lower than in the basal diet group, whereas Cu (p < 0.021) and Zn (p < 0.017) were significantly higher when strain psm16 was given orally, and Fe was increased in the phytase group (Table 3). In kidney tissue, phytic acid significantly decreased the levels of Fe (p < 0.002) and Mn (p < 0.010). Gavage with strain psm16 also changed the mineral level in the kidney: a significant increase of Mn and Zn (p < 0.022) was seen, whereas the levels of Cu (p < 0.007) and Mg (p < 0.001) were significantly decreased. Addition of phytase together with a phytic acid diet increased the content of Mn (p < 0.010) and decreased the contents of Cu (p < 0.007) and Mg (p < 0.001) in the kidney (Table 3). The levels of Ca (p < 0.000), Mn (p < 0.000), and Zn (p < 0.000) in plasma were significantly increased by gavage with strain psm16. Cu (p < 0.040) was significantly increased in the phytase group compared with the basal diet group (Table 3). In addition, the content of P in the kidney and the plasma was lower after adding phytic acid, but it returned to normal levels after treatment with strain psm16 or phytase, although the differences were not statistically significant.