Synbiotic intervention reverses alcohol drinking-induced cognitive deficits in adolescent male mice by modulating the microbiota-gut-brain axis

Abstract

Adolescence is characterized by an increased vulnerability to substance abuse, including alcohol consumption. We investigated the effects of a synbiotic intervention on disruptions of the microbiota-gut-brain axis induced by a drinking in the dark model of intermittent alcohol exposure in adolescent mice. We found that alcohol drinking induced specific shifts in gut microbiota, namely it increased Erysipelotrichaceae and reduced fecal butyric and isovaleric acids. In adulthood, other types of gut bacteria were affected such as Rhodospirillales uncultured family and Entrorhabdus uncultured bacterium. Social and nonsocial cognitive impairments were also observed, and disruptions in prefrontal cortex β-hydroxybutyrate and glutamate metabolic profile in the hippocampus were apparent. Importantly, the synbiotic restored gut microbiota alterations and exerted beneficial effects on alcohol-induced behavioral impairments and brain metabolite changes. In correlational studies, we identified two potential functional networks, one relating gut microbiota (Actinobacteria and Lactobacillaceae)-isovaleric acid with prefrontal glutamate metabolism and sociability, and the other relating SCFAs (propionic, butyric, valeric and isovaleric acids) with β-hydroxybutyrate in the hippocampus and reference memory. These results provide correlative data showing that synbiotic supplementation may restore delayed behavioral alterations induced by voluntary sub-binge alcohol drinking during adolescence through microbiota-gut-brain interactions, and might represent a potential therapeutic tool against long-term alcohol induced behavioral and molecular disturbances.

Keywords: Adolescence; behavior; brain glutamate; gut microbiota; intermittent alcohol drinking; short-chain fatty acids; β-hydroxybutyrate.

Graphical abstract

Figure 1.

Alcohol-related alterations in gut microbiota and fecal short-chain fatty acids. (A) Schematic diagram of the experimental procedures followed. One group of adolescent mice (PD30) (N = 20) consumed alcohol (EtOH, grey bars) for 4 weeks following a “drinking in the dark” (DID) protocol until adulthood (PD57), and another consumed water (H2O, white bars) for 4 weeks (N = 20). Fecal samples were collected at PD57 from all mice. No significant EtOH-related changes were observed for alpha diversity in observed taxonomic units (OTU) (B) or evenness (C). Principal component analysis (PCoA) for microbial community beta diversity in fecal samples in mice exposed to H2O (purple round symbols) or ETOH (green round symbols) during adolescence: ecological distances for (D) Bray Curtis, (E) Jaccard, (F) unweighted, and (G) weighted Unifrac indexes. Although there were no significant differences between groups for the phylum Firmicutes (H), the relative abundance of a family belonging to this phylum, namely Erysipelotrichaceae, was significantly increased in mice drinking EtOH compared to those exposed to H2O (I). The fecal concentrations of the short-chain fatty acids butyric (K) and isovaleric (M) acids were significantly decreased in mice after 4 weeks of voluntary sub-binge alcohol drinking compared to mice exposed only to H2O. No significant differences between groups were observed in propionic (J) and valeric (L) acid concentrations. Values are mean ± SEM, statistical analysis by Welch’s t-test, *p < 0.05, **p < 0.01.

Figure 2.

Behavioral effects of voluntary sub-binge alcohol drinking during adolescence and modulation by SYN treatment. (A) Schematic diagram of the experimental procedures followed in the study. After consuming either alcohol (EtOH) or water (H2O) from PD30 to PD57, each group was separated into two groups and administered either synbiotic (SYN) treatment or vehicle (VEH: water) for 3 weeks. Behavioral tests were performed at PD80 and continued for 1–2 weeks. Brain and fecal samples were collected at the end of the experiment. Mice were tested in different behavioral paradigms. (B) a significant main effect of EtOH on immobility time in the tail suspension test was observed. No significant changes were observed in % of time in the center of an open field (C) or in the number of buried marbles (D). A significant interaction between factors was observed in the sociability index (E), social novelty index (F), and affective state discrimination index (G), with post-hoc test showing significant differences between mice exposed to EtOH-VEH and H2O-VEH and between mice exposed to EtOH-SYN and EtOH-VEH. A significant main effect of alcohol on reference memory (H) and a significant main effect of alcohol and SYN on novel object recognition (I) were observed. Values are mean ± SEM, statistical analysis by two-way ANOVA followed by a post-hoc (main effect of EtOH: ^$p < 0.05, ^$$$p < 0.001); (main effect of SYN: ^##p < 0.01); (*p < 0.05, **p < 0.01, ***p < 0.01).

Figure 3.

Changes in gut microbiota alpha and beta diversity and fecal SCFA concentrations induced by alcohol exposure during adolescence and modulation by SYN treatment. Adolescent mice were exposed to water (H2O) or alcohol (ETOH) during adolescence, and then received 3 weeks of treatment with vehicle (VEH, gray bars) or synbiotic (SYN, magenta bars). (A) a significant main effect of EtOH was observed in the diversity index of observed taxonomic units (OTU). No significant differences between groups were observed in evenness (B). Principal component analysis (PCoA) for microbial beta diversity in fecal samples for mice exposed to EtOH-VEH (filled purple circles), EtOH-SYN (filled green circles), H2O-VEH (empty purple circles), H2O-SYN (empty green circles): ecological distances for (C) Bray Curtis, (D) Jaccard, (E) unweighted, and (F) weighted Unifrac indexes. No significant differences in fecal concentrations of propionic (G), butyric (H), valeric (I), and isovaleric (J) acids were observed between groups. Values are mean ± SEM, statistical analysis by two-way ANOVA followed by a post-hoc test (main effect of EtOH, ^$p < 0.05).

Figure 4.

Changes in gut microbiota abundance induced by alcohol exposure during adolescence and modulation by SYN treatment. Adolescent mice were exposed to water (H2O) or alcohol (ETOH) during adolescence, and then received 3 weeks treatment with vehicle (VEH, gray bars) or synbiotic (SYN, magenta bars). A main effect of SYN treatment on the relative abundance of the Firmicutes (A) was observed. A significant main effect of EtOH on the relative abundance of the Actinobacteria (B) and Deferribacterota (D) was observed. For the members of Firmicutes phylum, an interaction was observed between factors in the Anaerovoraceae family (E). A significant main effect of synbiotic treatment was observed in Clostridia UCG-014 (F), Lactobacillaceae (G) and RF39 (H) families. A significant main effect of EtOH on the relative abundance of the Erysipelotrichaceae family (I) was observed. For the Proteobacteria phylum, in the uncultured family of rhodospirillales (J) a significant interaction between factors was observed. For the Deferribacterota phylum, a significant main effect of EtOH on the relative abundance of the family Deferribacteraceae (K) was observed. For the Actinobacteria phylum, in the Enterorhabdus uncultured bacterium family (L), a significant interaction between factors was observed. Values are mean ± SEM, statistical analysis by two-way ANOVA followed by a post-hoc. (main effect of EtOH: ^$p < 0.05, ^$$p < 0.01); (main effect of SYN: ^#p < 0.05); (*p < 0.05, **p < 0.01).

Figure 5.

Effects of alcohol exposure during adolescence on brain metabolite concentrations and modulation by SYN treatment. Adolescent mice were exposed to water (H2O) or alcohol (ETOH) during adolescence, and then received 3 weeks of treatment with vehicle (VEH, gray bars) or synbiotic (SYN, magenta bars). (A) a significant interaction between factors was observed for prefrontal β-hydroxybutyrate concentrations, where EtOH increased its levels and SYN countered this effect. Moreover, a significant main effect of SYN was observed in the prefrontal concentrations of GABA (B). On the other hand, hippocampal glutamate (J) was increased by EtOH in animals receiving VEH but not SYN, whereas the Glu/Gln ratio (K) was decreased by SYN. Values are mean ± SEM, statistical analysis by two-way ANOVA followed by a post-hoc. (main effect of SYN: ^#p < 0.05); (*p < 0.05).

Figure 6.

Spearman correlations between behavioral performance and brain metabolite changes and between changes in fecal SCFA and brain metabolite concentrations in mice exposed to EtOH-VEH and EtOH-SYN. In the EtOH-VEH group, the sociability index negatively correlated with the Glu/Gln ratio in the PFC. Social novelty positively correlated with PFC BHB, and reference memory showed a negative correlation with PFC Gln (A). Reference memory positively correlated with BHB in the HPC (B). GABA in the PFC correlated positively with propionic and valeric acids (C). BHB in the HPC correlated positively with propionic, butyric, valeric, and isovaleric acids (D). In the EtOH-SYN group, novel object recognition (F) performance and butyric acid (H) correlated negatively with Glu/Gln in the HPC. In the EtOH-SYN group, no significant correlations were found between brain metabolites and behavioral tests (E) or SCFA (G). Gln: Glutamine; Glu: glutamate. The dependent variables in the behavioral tests are: sociability index, social novelty index, affective state discrimination index, immobility time in the tail suspension test, novel object discrimination index for object recognition test and ratio of time in novel arm for reference memory test. The color-coded side bar represents the R of Spearman (from −1 to 1), and significant correlations are marked with asterisks (*p < 0.05, **p < 0.01).

Figure 7.

Spearman correlates between behavioral performance and gut microbiota abundance changes and fecal SCFA, between brain metabolites and gut microbiota changes, and between gut microbiota and SCFA in mice exposed to EtOH-VEH and EtOH-SYN. In the EtOH-VEH group, the phylum Actinobacteria correlated positively with sociability index, and the Anaerovoracocaceae family correlated positively with affective state discrimination. The Rhodospirillales family correlated positively with immobility time, and Clostridia correlated positively with the novel object recognition index (A). The dependent variables in the behavioral tests are: sociability index, social novelty index, affective state discrimination index, immobility time in the tail suspension test, novel object discrimination index for object recognition test and ratio of time in novel arm for reference memory test. In addition, actinobacteria correlated positively and Lactobacillaceae correlated negatively with isovaleric acid, while Deferribacteraceae correlated negatively and RF39 uncultured bacterium correlated negatively with propionic acid (B). In the PFC, Actinobacteria correlated negatively with Glu/Gln whereas Lactobacillaceae correlated positively with glutamic acid and Glu/Gln. Moreover, Deferribacteraceae and RF39 correlated positively with glutamic acid (C). In the HPC, Deferribacteraceae correlated negatively and RF39 correlated positively with glutamine (D). In the EtOH-SYN group, the phylum Proteobacteria correlated positively with affective discrimination, social novelty and with immobility time. Also, object recognition correlated positively with Clostridia UGC and negatively with Enterorhabdus (E) and Lactobacillaceae correlated positively with isovaleric acid (F). In the PFC, BHB correlated positively with Actinobacteria and Eryspelotrichaceae correlated negatively with glutamic acid (G). In contrast, BHB correlated negatively with Enterorhabdus uncultured, Eryspelotrichaceae correlated negatively with glutamine and glutamic acid and Rhodospirillales correlated positively with Glu/Gln (H).The color-coded side bar represents the R of Spearman (from −1 to 1), and significant correlations are marked with asterisks (*p < 0.05, **p < 0.01, ***p < 0.001).

Figure 8.

Schematic representation of the gut microbiota-PFC glutamate metabolism-sociability network (A) and the SCFA-HPC BHB-spatial memory relationship (B). Red arrows indicate Spearman positive correlation whereas green arrows indicate negative correlation.