The L. intestinalis (ATCC49335) used in this study was purchased by Beijing Beina Chuanglian Biotechnology Research Institute (Beijing, China). Unless otherwise stated, bacterial strains were grown in MRS Broth (MRSB) (Qingdao Hope Bio-technology Corporation Ltd.) or on MRS Agar (MRSA) plates at 37°C.

Fifty female C57BL/6 mice (aged 6 weeks, 17 ± 0.5 g and aged 4 weeks, 14 ± 0.5 g) were randomly divided into 5 groups with arbutin solution of 0, 0.1, 0.2, 0.4, and 1 mg/ml (20) and fed maintenance diet lasted 3 weeks. We found that arbutin 0.4 mg/ml was most effective in improving intestinal index; thus, twenty mice were treated with antibiotics and antibiotics + 0.4 mg/ml arbutin for 3 weeks. Then, we transplanted fecal microbes from oral 0.4 mg/ml arbutin mice to mice pretreated with antibiotics for 1 week and then normal feeding for 2 weeks. Finally, twenty mice were pretreated with antibiotics for a week, 1.3 × 10⁹ colony-forming unit (CFU)/ml L. intestinalis was intragastric to mice for 1 week, and then normal feeding for 2 weeks. All the animals were purchased from Hunan SJA Laboratory Animal Corporation Ltd. (Changsha, China) and used in this study. All the experimental animals were allowed free access to food and drinking water, and subjected to 12-h light-dark cycles, controlled temperature (23 ± 2°C), and humidity (45–60%) during the experiment. The basic diet was described in our previous study (21).

Intestinal HE staining was performed. The jejunal and ileal segments were fixed in 4% paraformaldehyde solution. The sections were first treated with xylene and ethanol solution for 15 and 5 min, respectively, then stained with hematoxylin for 5 min, rinsed with water for 5 min, then stained with eosin solution for 1–3 min, and then washed with ethanol and sealed. Finally, the villi, and crypt morphology were observed under a microscope.

Serum samples were separated after centrifugation at 1,500 × g for 10 min at 4°C and 100 μl serum was transferred into another tube. Serum biochemical parameters were determined using an Automatic Biochemistry Analyzer (Cobas c 311, Roche).

To eradicate commensal bacteria, filter-sterilized drinking water was supplemented with ampicillin (0.5 mg/ml, Meilunbio), gentamicin (0.5 mg/ml, Meilunbio), metronidazole (0.5 mg/ml, Meilunbio), neomycin (0.5 mg/ml, Meilunbio), and vancomycin (0.25 mg/ml, Meilunbio) for 1 week. Antibiotics were purchased from Dalian Meilun Biotechnology Corporation Ltd. (Dalian, China). Before fecal microbiota transplantation, the native gut microbiota in one group of C57 female mice (n = 10 biologically independent animals per group) was deleted by administering drinking water containing a cocktail of antibiotics for 1 week. Fecal samples of ∼200 mg were then collected from arbutin (0.4 mg/ml)-fed mice and resuspended in 2.0 ml normal saline. Fecal samples were mixed and centrifuged at 1,000 × g, and the microbiota supernatants were transplanted into the microbiota-depleted mice by gavaging with 0.2 ml per mice for 1 week. After transplantation, two groups of mice were administrated with a standard diet and regular water.

Total genome DNA from ileal chyme and mucosa was extracted using cetyltrimethylammonium bromide (CTAB) method. DNA concentration and purity were monitored on 1% agarose gels. According to the concentration, DNA was diluted to 1 ng/μl using sterile water. 16S rDNA genes of distinct regions (16S V3-V4) were amplified using a specific primer (515F-806R) with the barcode. All the PCR reactions were carried out with 15 μl of the Phusion^® High-Fidelity PCR Master Mix (New England Biolabs), 2 μM of forward and reverse primers, and about 10 ng of template DNA. Sequencing libraries were generated using the TruSeq^® DNA PCR-Free Sample Preparation Kit (Illumina, United States) following the manufacturer’s recommendations and index codes were added according to our previous study (22). Microbial communities were investigated by iTag sequencing of 16S rDNA genes (23, 24).

All the statistical analyses were performed using the one-way ANOVA and t-test analysis in SPSS version 20.0 software (SPSS Incorporation, Chicago, IL, United States). The data are expressed as the means ± SEM). P < 0.05 was considered statistically significant. All the figures in this study were drawn using GraphPad Prism version 8.0.

Final body weight was not obviously changed (Figure 1A), but the relative weight and weight/length of the intestine were significantly increased by arbutin administration at 0.2 and 0.4 mg/ml (P < 0.05) (Figures 1B,D), and arbutin did not alter the intestinal length, villus width and crypt depth (Figures 1C,F,H). Thus, we continued to investigate the intestinal pathology section; the results showed that the villus length was increased by 0.4 and 1.0 mg/ml arbutin, villus area was enhanced by 0.2 and 0.4 mg/ml arbutin, and villus length/crypt depth (L/D) was higher at 0.4 mg/ml arbutin (P < 0.05) (Figures 1E,G,I,J).

To further understand the role of arbutin, lipid parameters in serum were determined (Figures 2A–F). Arbutin at 0.4 mg/ml significantly enhanced the content of serum glucose (Glu) (P < 0.05) (Figure 2A). Nevertheless, arbutin at 0.2 mg/ml lowered the content of total cholesterol (TC) and high-density lipoprotein (HDL) (P < 0.05) (Figures 2C,D), and low-density lipoprotein (LDL) was lowered at 0.1 and 0.2 mg/ml (P < 0.05) (Figure 2E). These results suggested that arbutin can improve intestinal development and serum lipid parameters.

To investigate the effects of arbutin on gut microbes, we determined the microbiome by 16S rDNA sequencing at 0.4 mg/ml (Figures 3, 4). Venn diagram showed that 597 and 111 different operational taxonomic units (OTUs) were found in the control and arbutin groups and contained the same 540 OTUs (Figure 3A), rarefaction curve indicated that the sample capacity and sample depth were reasonable (Figure 4B). Arbutin significantly decreased the α-diversity index [observed species, Shannon index, phylogenetic diversity (PD), Simpson index, Chao1, and abundance-based coverage estimator (ACE)] (P < 0.05) (Figures 3B–D, 4A,C–E). Meanwhile, the β-diversity index was reduced (P < 0.05) (Figure 3E), and principal component analysis showed that there were different zones of intestinal microflora between the control group and arbutin (Figure 3F). At the phylum level, the relative abundance of Actinobacteria and Proteobacteria was clearly lowered by arbutin (P < 0.05) (Figure 3G). At the species level, 0.4 mg/ml arbutin markedly increased the abundance of Lactobacillus intestinalis (P < 0.05) (Figures 3H, 4F), while the abundance of Bifidobacterium animalis, Bacillus velezensis, Lachnospiraceae bacterium_M18-1, Eubacterium sp_14-2, Helicobacter ganmani, Lachnospiraceae bacterium_10-1, Lachnospiraceae bacterium_615, Planoglabratella opercularis, Pseudoflavonifractor sp_Marseille-P3106, Clostridium leptum, Clostridium sp_ASF356, Dubosiella newyorkensis, Burkholderiales bacterium_YL145, Desulfovibrio sp_ABHU2SB, Firmicutes_bacterium CAG_194_44_15, Clostridium sp_Culture-27, and Ruminiclostridium sp_KB18 was lowered compared to control (P < 0.05) (Figures 3H, 4F).

The intestinal microbiota has been shown to regulate intestinal development (25) and host metabolism (18). To further determine the role of intestinal microbiota, 4 weeks mice were given an antibiotics cocktail for 1 week with oral arbutin solution (0.4 mg/ml). Predictably, arbutin significantly enriched the villi width compared to the antibiotics group in the jejunum (P < 0.05) (Figures 5A,C), but villus length, villus area, crypt depth, L/D were not changed (Figure 5B,D–F), and there was a tendency to enhance the ileal villi index (Figures 5G–L). Then, we further collected feces from mice administered with arbutin 0.4 mg/ml and transplanted them to mice with an antibiotics cocktail. Fecal microflora transplantation significantly improved intestinal pathologies, such as jejunal villus length (Figures 6A,B), jejunal villus length/villus width (L/D) (Figures 6A,F), and ieal villus areas (Figure 6J). But jejunal villus width, jejunal villus area, jejunal crypt depth, ileal villus legth, ileal villus width, ileal crypt depth and ileal L/D were uninfluential (Figures 6C–E,H,I,K,L). In summary, gut microbes contributed improving intestinal development.

We have found that the abundance of L. intestinalis (Lin) was markedly enhanced by arbutin and was the most abundant bacterium in the gut (Figures 3H, 4F). L. intestinalis was often found in the gut of the host, which was treated for various diseases (26–28) and metabolic disorders (29, 30), but the effect of L. intestinalis on gut development and host lipid metabolism was unclear. Thus, we investigated the role of Lin on intestinal pathology and used Lin monocolonization (31) with an antibiotics cocktail for 1 week. Interestingly, after an antibiotics cocktail for 1 week, Lin monocolonization clearly increased the villus length, crypt depth, and villus areas (Figures 7A–C,E), and there was a tendency to elevate the number of goblet cells (Figures 7A,G). Whereas villus width and L/D were not changed by L. intestinalis (Figures 7D,F).

In order to verify the previous results, we co-cultured arbutin and L. intestinalis to investigate the growth of L. intestinalis in vitro. The results showed that arbutin significantly promotes the growth of L. intestinalis (P < 0.05) (Figure 8).