Evidence for long-lasting alterations in the fecal microbiota following prenatal alcohol exposure

Abstract

Background: There is growing evidence that the gut microbiota can be shaped by early-life experiences/exposures, with long-term consequences for brain, behavior, and health. Changes in the gut microbiota have also been identified in neurodevelopmental disorders including Autism Spectrum Disorder and schizophrenia. In contrast, no studies to date have investigated whether the gut microbiota is altered in individuals with Fetal Alcohol Spectrum Disorder (FASD), the neurodevelopmental disorder that results from prenatal alcohol exposure (PAE). The current study was designed to assess the impact of PAE on the fecal microbiota.

Methods: We used a rodent model in which pregnant Sprague-Dawley rats were provided with an EtOH-containing diet or a control diet throughout gestation. Fecal samples were collected from adult male and female animals and 16s rRNA sequencing was performed.

Results: Overall, PAE rats showed greater richness of bacterial species, with community structure investigations demonstrating distinct clustering by prenatal treatment. In addition, prenatal treatment and sex-specific alterations were observed for many specific microbes. For example, in males, Bacteroides and Bifidobacterium, and in females, Faecalitalea and Proteus, differed in abundance between PAE and control rats.

Conclusions: Taken together, these results show for the first time that PAE has a long-lasting and sex-specific impact on the fecal microbiota. Further research is needed that considers fetal microbiota in the development of new interventions in FASD.

Keywords: Fetal Alcohol Spectrum Disorder; development; microbiota; sex differences.

Figure 1.

Assessments of α and β diversity between PAE and C animals. A) Boxplots displaying α diversity measures of observed richness and the Shannon index. Statistical analyses: Welch two sample t-tests. B) Bray-Curtis non-metric multidimensional scaling (NMDS) plot in 2 dimensions (stress value = 0.196230). Ellipsoids represent 95% confidence intervals and were plotted for each prenatal treatment group, with significant differences identified by PERMANOVA. C) Relative abundance plot representing the top 80% of OTUs across all samples (56 genera), with each column representing one animal pair. The top 5 most abundant genera are shown in the figure legend, for clarity. *p<0.05; #: trend p=0.08; PAE: prenatal alcohol exposure; n=8 cages of two rats/prenatal treatment (males, females).

Figure 2.

Differentially abundant taxa (males and females). Data was statistically grouped and filtered at the genus level before DESeq2 analysis. Taxa were ordered according to decreasing log₂ fold change. Only taxa showing significant differential abundances (adjusted p-value ≤ 0.05) are shown. Statistical analysis: Median of ratios method in the DESeq2 package. Points represent unique genera and are colored according to phylum. Points above the line (positive differences in differential abundance) depict taxa that are overrepresented in PAE compared to control animals, and those below the line (negative differences in differential abundance) depict taxa that are underrepresented in PAE compared to control animals. NA: indicates microbes that were unidentifiable at the genus level (identified at the family level in the figure); n=3–5 cages of two rats/prenatal treatment, per sex.

Figure 3.

Assessments of α and β diversity between PAE and C animals with stratification by sex. A) Boxplot displaying α diversity measures of observed richness and the Shannon index. Statistical analysis: Two-way ANOVA (prenatal treatment × sex) followed by post-hoc testing using Tukey’s HSD test. B) Bray-Curtis non-metric multidimensional scaling (NMDS) plots in 2 dimensions (stress value = 0.1289576 and 0.2529185 for males and females, respectively). Ellipsoids represent 95% confidence intervals and were plotted for each prenatal treatment, by sex, with significant differences identified by PERMANOVA. C) Relative abundance plot representing the top 80% of OTUs across all samples (56 genera), with each column representing one animal pair. The top 5 most abundant genera are shown in the figure legend, for clarity. *p<0.05; #: trend p=0.099 (male), p=0.135 (female); PAE: prenatal alcohol exposure; n=3–5 cages of two rats/prenatal treatment, per sex.

Figure 4.

Bray-Curtis non-metric multidimensional scaling (NMDS) plots in 2 dimensions, illustrating the differences in community composition between males and females in the (A) control group and (B) PAE group. Ellipsoids represent 95% confidence intervals and were plotted for each prenatal treatment, by sex, with significant differences identified by PERMANOVA. Distinct clustering by sex was observed in controls (A), with only a trend for sex differences detected in PAE animals (B). *p<0.05; # p=0.117; PAE: prenatal alcohol exposure; n=3–5 cages of two rats/prenatal treatment, per sex.

Figure 5.

Differentially abundant taxa in A) males and B) females. Data was statistically grouped and filtered at the family level before DESeq2 analysis. Taxa were ordered according to decreasing log₂ fold change. Only taxa with significant differential abundance (adjusted p-value ≤ 0.05) are shown. Statistical analysis: Median of ratios method in the DESeq2 package. Points represent unique genera and are colored according to phylum. Points above the line (positive differences in differential abundance) depict taxa that are overrepresented in PAE compared to control animals, and those below the line (negative differences in differential abundance) depict taxa that are underrepresented in PAE compared to control animals. NA: indicates microbes that were unidentifiable at the genus level (identified at the family level in the figure); n=3–5 cages of two rats/prenatal treatment, per sex.

Figure 6.

Venn diagrams illustrating the number of shared and unique OTUs: A) by prenatal treatment and sex and in B) females only, and C) in males only. A) PAE and control males had the highest number (35) of unique OTUs (35). PAE and control females were found to have 31, and 18 unique OTUs, respectively. 10 OTUs were exclusively shared between PAE and control females. By contrast, PAE and control males shared 7 OTUs. Overall, PAE females had the greatest OTU diversity (202 OTUs), followed by PAE males (198 OTUs), control males (171 OTUs), and control females (162 OTUs). B) In females, PAE and controls had a 52.3% similarity in overall OTU composition. By contrast, in males, C) a 43.6% similarity was identified.