LGS, LGSP, and LGSF attenuated ALD abnormality in mice

The food intake of mice is shown in Fig. 1B, and the food intake in each group was comparable. At the end of experiment, the weight of mice in ALD group was slightly higher than that in the control group and other groups (Fig. 1C). Morphological observation showed that the liver in ALD mice was apparently in a whiter color compared to control mice, indicating the fat accumulation (Fig. 1D). After treatment by MDTX, LGS, LGSP and LGSF, the liver of mice became ruddy (Fig. 1D). Furthermore, as shown in Fig. 1E, compared with the control group, the liver index was significantly increased in the ALD group (p < 0.001), which was alleviated by treatment of LGS, LGSP, and LGSF, with the most pronounced remission effect by LGS.

Histological analysis of liver tissue showed that excessive alcohol intake could lead to the appearance of a large number of lipid vacuoles in the liver and cause the destruction of hepatocyte structure, while the MTDX, LGS, LGSF and LGSP treatment apparently mitigated these histological changes (Fig. 1F, G). In addition, the protective effect of MTDX, LGS, LGSP, and LGSF on alcohol-induced hepatic steatosis was verified by examining liver fat accumulation using oil red O staining. As shown in Fig. 1H and I, the ALD group exhibited severe fat accumulation (red), which was attenuated in the MTDX, LGS, LGSP, and LGSF groups.

Overall, LGS and its LGSP and LGSF fractions had a protective effect in ALD mice.

LGS, LGSP, and LGSF protected against ALD in mice through alleviating hepatic injury and inflammation, improving dyslipidemia and promoting ethanol metabolism

Serum AST and ALT levels are sensitive indicators of liver function [31]. Alcohol intake has a significant effect on liver-specific enzymes, including AST and ALT. As shown in Fig. 2A and B, compared to control group, there was a significant elevation of AST and ALT in the ALD group, indicating that the abnormal liver function. Both MTDX and LGS treatment similarly and significantly reduced the alcohol-induced AST and ALT elevation. LGSP had a significant remission effect on ALT level, while LGSF showed alleviating effect on both AST and ALT levels.

Fig. 2

LGS and its active fractions (LGSP and LGSF) alleviated hyperlipidemia and inflammation, and improved ethanol metabolism in ALD mice. A Serum level of AST; B serum level of ALT; C serum level of TG; D serum level of TC; E serum level of HDL; F serum level of LDL; G serum level of IL-6; H serum level of TNF-α; I serum level of IL-1β; J enzymatic activity of ALDH; K enzymatic activity of ADH; L serum level of GLP-1. # P < 0.05, ## P < 0.01, ### P < 0.001 vs. ALD group; * P < 0.05, ** P < 0.01 and *** P < 0.001 vs. control group; one-way ANOVA with a post hoc Tukey test

Alcohol affects fat metabolism in the liver, causing abnormalities in processes such as fat synthesis, transport, and oxidation, which ultimately leads to fat accumulation in the liver [32]. The levels of TC, TG, LDL and HDL serve as sensitive markers for assessing liver lipid metabolism disorders. The results (Fig. 2C–F) demonstrated that MTDX and LGS significantly mitigated the ethanol-induced increase of TC, TG and LDL, and decrease of HDL. Both LGFP and LGSF significantly reduced serum TC and TG levels in ALD mice, while they had no effect on HDL. LGSP reduced LDL while LGSF did not.

Furthermore, LGS treatment significantly reduced the ethanol-induced increase of proinflammatory factors IL-6, TNF-α and IL-1β (Fig. 2G–I). MTDX only had the remission effect on IL-6 and IL-1β. LGSP decreased the serum levels of IL-6 and TNF-α, while LGSF only had an effect on IL-6 but not TNF-α. Both LGSP and LGSF had no effect on IL-1β.

Alcohol-induced liver damage might be linked to the activity of the main catalytic enzymes ADH and ALDH in the liver [33]. In this study, the enzymatic activities of ALDH and ADH were significantly lower in ALD group, and the LGS and LGSP treatment partially recovered the enzyme activities (Fig. 2J, K). Both MTDX and LGSF had no mitigating effect on activities of ALDH or ADH. In addition, although LGS, LGSP and LGSF groups showed a trend of increase on ethanol-induced serum GLP-1 level, no statistical difference was found (Fig. 2L).

Based on above results, LGS had a remarkable effect on mitigating dyslipidemia, hepatic injury and inflammation, and improving ethanol metabolism. Regarding most of the detected indices, the protective effect of LGS was comparable to that of MTDX, while LGSP or LGSF displayed a relatively smaller effect.

LGS, LGSP, and LGSF mitigated intestinal damage and maintained barrier function in ALD mice

Alcohol abuse can cause damage to the intestinal mucosa, leading to epithelial cell damage and necrosis, thereby compromising mucosal barrier function [34]. Therefore, we further explored the impact of LGS, LGSP, and LGSF on intestinal repair.

As shown in Figs. 3A and S2A, H&E staining analysis of the ileum and colon showed that, compared with the control group, the ALD group showed inflammatory invasion, mucosal thickening and shedding, decreased goblet cells, shortened glandular tracts, and/or irregular arrangement of ileum villi, however, after treatment with LGS, LGSP, and LGSF, the extent of gut damage was significantly reduced, where the inflammatory cell infiltration was reduced, the intestinal epithelium was more intact, and the crypt structure was improved. According to the histopathological score, compared to the ALD group, the treatment groups’ histopathological scores were lower (Fig. S3A, B). In contrast, the positive drug MTDX had no protective effect on epithelial damage.

Fig. 3

LGS and its active fractions (LGSP and LGSF) mitigated intestinal injury, increased the expression of tight conjunction proteins, and promoted intestinal epithelial proliferation in ALD mice. A H&E staining of ileum. B TUNEL-stained ileum sections. Arrows indicate apoptotic cells. C The expression of occludin protein in ileum. D The expression of ZO-1 protein in ileum. E Immunofluorescence staining of Lyz1+ cells in ileum; red, Lyz1 positive staining; blue, DAPI staining. F Immunofluorescence staining of Muc2+ cells in ileum; red, Muc2 positive staining; blue, DAPI staining. Quantitative analysis of histopathological scores, TUNEL-positive cells, occludin or ZO-1 positively stained area, and number of Lyz1+ and Muc2+ cells are displayed in the supplementary file (Figs. S3, S7)

Additionally, we performed TUNEL apoptosis assays on the ileum (Figs. 3B, S3C) and colon (Figs. S2B, S3D), and the green stained cells represent the apoptotic cells. By counting the apoptotic cells, it was found that apoptotic cells in the ileum and colon of mice in the LGS, LGSP, and LGSF group were considerably lower than that in the ALD group.

To analyze the gut permeability, we determined the serum level of LPS. LPS, as a large molecule metabolite of bacteria that cannot readily move across gut barrier, has been used as a marker for gut permeability [26, 27]. The result showed that the serum level of LPS in LGS, LGSP, and LGSF treatment group was significantly decreased compared to the ALD group (Fig. S4). This suggests that LGS, LGSP, and LGSF presumably improved gut permeability.

The integrity of the intestinal barrier depends on tight junction proteins such as occludin, ZO-1, and other associated proteins [30]. ZO-1 and occludin play a crucial role in maintaining the integrity and permeability of the intestinal mucosa. The absence of ZO-1 and occludin increases intestinal permeability, resulting in the invasion of bacteria and viruses, which in turn affects the protective effect of the intestinal mucosa [35, 36]. Therefore, to further validate the gut barrier function, IHC was performed to further assess the levels of ZO-1 and occludin in both the ileum and colon. As shown in Figs. 3C, D, S3E–H and S5, there was an evident deletion of ZO-1 and occludin in the ileum and colon tissues of the ALD group, while in the LGS, LGSP, and LGSF treatment group, compared to the ALD group, the rates of positive expression for ZO-1 and occludin were noticeably higher. The expression of occludin in colon tissue did not show a significant difference among groups (Figs. S3F, S5A). To further elucidate the contributions of LGS, LGSP, and LGSF in preserving intestinal mucosa, we detected the expression of Paneth cell marker lyz1 and goblet cell marker muc2 in the ileum and colon by IF analysis [37]. As shown in Figs. 3E, F, S6 and S7, in ileum and colon, the number for Lyz1+ and Muc2+ cells (red) in LGS, LGSP, and LGSF groups were higher than those of the ALD group.

Overall, the above findings suggest that the use of LGS, LGSP, and LGSP inhibits intestinal damage and improves intestinal barrier function in the mouse model of ALD.

LGS, LGSP, and LGSF promoted intestinal epithelial repair in TNF-α-induced intestinal organoids

Intestinal organoids are three-dimensional cellular models that encompass various types of functional intestinal epithelial cells, such as epithelial cells, goblet cells, and Paneth cells [38]. In view of this, we used TNF-α (500 ng/mL) to induce 24 h to construct an intestinal organoid inflammation model [39]. In the TNF-α group, the morphology of the organoids showed a reduction in surface area, blackening and loss of part of the bud morphology; in contrast, the intestinal organoids in the control group were morphologically normal (Fig. 4A). The results demonstrated that the inflammatory injury model was well established. The number of intestinal organoids was counted by microscopic observation. There was no obvious distinction in the number of organoids among groups (Fig. 4B), while compared to TNF-α group, the calculated impairment rates showed a significant decrease in the LGS and LGSF groups, with a decrease trend in the LGSP group, but there was no statistical difference (Fig. 4C).

Fig. 4

LGS and its active fractions (LGSP and LGSF) protected against TNF-α-induced intestinal organoid injury in vitro through promoting epithelial cell differentiation and regeneration. A Intestinal organoid image under light microscopy. Scale bar = 100 μm. B Number of total organoids. C Number of injured organoids. D H&E-stained organoids. Arrow shows bud loss or reduced surface area. Scale bar = 50 μm. E EdU-stained organoids; scale bar = 20 μm. F TUNEL stained organoid sections; scale bar = 50 μm. G Lgr5 staining; red, Lgr5 positive staining; blue, DAPI staining; scale bar = 50 μm. H Lyz1 staining; red, Lyz1 positive staining; blue, DAPI staining; scale bar = 50 μm. I Muc2 staining; red, Muc2 positive staining; blue, DAPI staining; scale bar = 50 μm. Quantitative analysis of EdU+ and TUNEL+ immunofluorescence, and the number of Lgr5+, Lyz1+ and Muc2+ cells are displayed in the supplementary file (Fig. S8). # P < 0.05, ## P < 0.01, ### P < 0.001 vs. ALD group. * P < 0.05, ** P < 0.01 and *** P < 0.001 vs. control group; one-way ANOVA with a post hoc Tukey test

H&E staining (Fig. 4D) showed that the organoids in the control group are morphologically intact and the cells are tightly arranged, while the TNF-α group showed bud loss or reduced surface area, compared with the TNF-α group. After the treatment of LGS, LGSP, and LGSF, the germ emergence length and proportion of organoids were significantly higher, and the surface area was larger. In addition, EdU staining and TUNEL assay were applied to detect cell proliferation and apoptosis in each group. It was further found that after LGS, LGSP, and LGSF treatment, the positive area (red) was significantly increased (Figs. 4E, S8A), which indicated the onset of regeneration. Meanwhile, LGS, LGSP, LGSF also reduced the apoptosis of intestinal epithelial cells caused by TNF-α (Figs. 4F, S8B).

Notably, we also detected organoid epithelial cell levels in each group by IF staining of epithelial cell markers such as Lgr5, Lyz1, and Muc2. It was found that LGS, LGSP, and LGSF treatment up-regulated the TNF-α-mediated decrease of Lgr5 (Figs. 4G, S8C), Lyz1 (Figs. 4H, S8D), and Muc2 (Figs. 4I, S8E) expressions.

These findings suggest that LGS, LGSP, and LGSF can enhance the proliferation of intestinal epithelial cells, reduce apoptosis, and thus promote the recovery of intestinal function.

LGS, LGSP, and LGSF alleviated gut dysbiosis in ALD mice

Gut microbiota plays a crucial role in maintaining gut barrier function, and has been suggested tightly associated with ALD development. Thus, we further assessed the impact of LGS, LGSP, and LGSF on gut microbiota in ALD mice.

The α diversity of microorganisms was assessed using the Shannon and Chao1 indices (Fig. 5A, B). In comparison to the control group, alcoholic diet did not significantly alter the diversity of fecal microbes in mice, as the Shannon index was not changed. The values of Shannon index in the LGSP group were much higher than those of the ALD group, indicating LGSP increased microbial diversity. LGSP and LGSF also recovered ALD-mediated reduction of bacterial richness (Chao1 index).

Fig. 5

LGS and its active fractions (LGSP and LGSF) altered gut microbial structure in ALD mice. A Shannon index. B Chao1 index. C PCoA analysis based on ASV level. D Percent of community abundance on phylum level. E Abundance of Actinobacteriota. F Abundance of Firmicutes. G Abundance of Bacteroidota. H Abundance of Verrucomicrobiota. # P < 0.05, ## P < 0.01, ### P < 0.001 vs. ALD group; * P < 0.05, ** P < 0.01 and *** P < 0.001 vs. control group; one-way ANOVA with a post hoc Tukey test

Beta diversity can be used to intuitively compare the structure of microbe communities and analyze the differences in microbe communities between groups. PCoA analysis based on ASV level (Fig. 5C) showed that ALD mice were far separated from the control mice, showing vast distinct microbial structures. The MTDX group was placed close to the ALD group, suggesting that MTDX treatment had minimal effect on microbial compositions. However, after treatment of LGS, LGSP and LGSF, it is observed that the microbial structure was altered in a way moving towards the control group. The LGS and ALD groups were relatively close in distance, and LGSP and LGSF groups were significantly distinguished from the ALD group, with the LGSF group was the most obvious which was close to the control group. The above results suggested that the intervention of LGS, LGSP, and LGSF affected the structure of the microbiota to various extents, and LGSF caused the greatest changes in the structure of the microbiome.

Subsequently, the microbial composition was assessed. At the phylum level (Fig. 5D), Actinobacteriota (Fig. 5E), Firmicutes (Fig. 5F), Bacteroidetes (Fig. 5G), and Verrucomicrobiota (Fig. 5H) were the dominant bacterial groups, constituting over 90% of the total sequences. The proportion of Actinobacteriota was decreased significantly from 33.68 ± 4.3% in the control group to 8.15 ± 3.44% in ALD group, which was further increased to 15.88 ± 7.77% (MTDX group), 16.78 ± 5.82% (LGS group), 14.45 ± 6.69% (LGSP group), 22.96 ± 5.9% (LGSF group), with statistical difference observed for LGSF group (Fig. 5E). The proportion of Firmicutes increased significantly from 49.04 ± 7.43% in the control group to 74.65 ± 10.91% in the ALD group, and a significant decrease was observed after treatment of MTDX and LGSP (Fig. 5F). In addition, the proportion of Bacteroidota was increased specifically in the LGSP group (Fig. 5G). Compared with the control group (2.34 ± 2.14%), the percentage of Verrucomicrobiota was increased to 17.97 ± 10.56% in the ALD group, of which showed a significant decrease in the LGS (4.2 ± 3.3%), LGSP (1.48 ± 0.84%), and LGSF (1.03 ± 0.78%) groups (Fig. 5H).

In order to find the specific major microbes between the groups, a linear discriminant analysis (LDA) effect size (LEfSe) analysis was performed (Fig. 6A). The results showed that certain bacterial genera were enriched in each group (Fig. 6A): Coriobacteriaceae-UCG-002 in the control group; Parasutterella, Romboutsia and Escherichia-shigella in the ALD group; Faecalibaculum and Lactobacillus in the LGS group; Enterorhabdus, Alloprevotella, and Bacteroides in the LGSP group; and Monoglobus in the LGSF group.

Fig. 6

LGS and its active fractions (LGSP and LGSF) mediated specific microbial changes in ALD mice. A LEfSe analysis of the microbial community in each group. The significantly altered microbial genera included the following: B Coriobacteriaceae-UCG-002; C Parasutterella; D Romboutsia; E Escherichia-shigella; F Akkermansia; G Faecalibaculum; H Lactobacillus; I Bacteroides; J Enterorhabdus; K Monoglobus; L Bacillus; M Bifidobacterium. # P < 0.05, ## P < 0.01, ### P < 0.001 vs. ALD group; * P < 0.05, ** P < 0.01 and *** P < 0.001 vs. control group; one-way ANOVA with a post hoc Tukey test

Subsequently, one-way ANOVA was performed to determine the significantly altered microbes in groups. The key feature of microbial alterations induced by LGS, LGSP or LGSF could be summarized as follows: (1) recovery of ALD-mediated abnormal changes; (2) specifically enrichment of certain gut microbes. In the ALD group, the proportion of Coriobacteriaceae-UCG-002 (Fig. 6B) was decreased significantly, while that of Parasutterella (Fig. 6C), Romboutsia (Fig. 6D), and Escherichia-shigella (Fig. 6E) was significantly increased. Notably, the treatment of LGS, LGSP or LGSF could at least partially recover these ALD-mediated changes. LGSF significantly increased the Coriobacteriaceae-UCG-002 abundance in ALD mice (Fig. 6B). LGS, LGSP and LGSF all significantly decreased ALD-induced increase of Parasutterella and Romboutsia (Fig. 6C, D). LGS and LGSF could reverse the ALD-mediated enrichment of Escherichia-shigella (Fig. 6E). Furthermore, certain gut microbes were remarkably enriched after treatment of LGS, LGSP or LGSF: Akkermansia (Fig. 6F) in MTDX group, Faecalibaculum (Fig. 6G) and Lactobacillus (Fig. 6H) in the LGS group, Bacteroides (Fig. 6I), and Enterorhabdus (Fig. 6J) in the LGSP group, and Monoglobus (Fig. 6K) in the LGSF group. With a closer look, the Bacillus (Fig. 6L) were also significantly increased in LGS and LGSP groups, and Bifidobacterium (Fig. 6M) was enriched in the LGS group. The discrepancy in the effect of LGS, LGSP and LGSF may be due to their specific regulation of gut microbes or microbial interaction.

These results suggest that the microbial changes induced by ALD can be restored by LGS, LGSP, and LGSF treatment, and some of them are specific microbes in response to them.

LGS, LGSP, and LGSF promoted the proliferation of specific bacterial strains in vitro

Based on the microbial analysis, there were specifically altered gut microbes in the LGS, LGSP, and LGSF groups, including the Faecalibaculum, Lactobacillus, Bacteroides, Enterorhabdus, Monoglobus, Bacillus, and Bifidobacterium. To see whether there were specific bacterial strains, we investigated the species level alterations. As can be seen from Fig. S9A, several uncultured or unclassified strains were identified in these genera. In particular, the Bacteroides spp., B. sartorii, was found significantly elevated in the LGSP group (Fig. S9B), and the Dubosiella spp., D. newyorkensis, was significantly increased in the LGS and LGSP groups (Fig. S9C).

To further investigate the direct effect of LGS and its fractions on specific bacteria strains, in vitro culture was conducted. In addition to the B. sartorii and D. newyorkensis, the selected bacteria also included the Lactobacillus, Bifidobacterium strains (B. adolescentis, and B. bifidum), and Bacillus strain (B. coagulans), as there have been reports showing the anti-ALD effect of these strains [40, 41]. In order to avoid the interference of glucose in the experimental results, in this experiment, sugar-free culture medium was used in the incubation, ensuring that main carbon source was from the supplemented LGS, LGSP, or LGSF.

The results showed that LGS and LGSP, but not the LGSF, significantly promoted the growth of Lactobacillus (Fig. 7A). Similarly, LGS and LGSP significantly increased the proliferation of B. sartorii, and LGSF had a minimal effect (Fig. 7B). LGS, LGSP and LGSF all promoted the growth of B. coagulans (Fig. 7C), B. adolescentis (Fig. 7D), and B. bifidum (Fig. 7E). LGS and LGSP promoted the growth of D. newyorkensis, but the LGSF had no effect (Fig. 7F). Overall, the promoting effect of LGSF on most of bacteria strains, except the B. coagulans, was smaller than that of LGS and LGSP. This may be due to the fact that both LGS and LGSP contained rich carbon source for bacterial growth. Furthermore, the in vitro finding was not exactly consistent with the in vivo results. The probable reason is that flavonoids prevent harmful bacteria from growing and reduce the competitive pressure in the gut [42], thus creating favorable conditions for the growth of probiotics, while in vitro incubation is a single microbe proliferation experiment, which ignores the bacteria-to-bacteria interactions.

Fig. 7

LGS and its active fractions (LGSP and LGSF) promoted the in vitro growth of specific bacterial strains. A Growth curve of Lactobacillus. B Growth curve of Bacteroides sartorii. C Growth curve of Bacillus coagulans. D Growth curve of Bifidobacterium adolescentis. E Growth curve of Bifidobacterium bifidum. F Growth curve of Dubosiella newyorkensis. Glucose (Glu) was used as positive control throughout the experiments. *** P < 0.001 vs. medium group; one-way ANOVA with a post hoc Tukey test

Taken together, LGS, LGSP, and LGSF promoted the proliferation of specific bacterial strains in vitro.