Stratification of Gut Microbiota Profiling Based on Autism Neuropsychological Assessments

Abstract

Autism spectrum disorder (ASD) is a neurodevelopmental disorder. Investigations of gut microbiota (GM) play an important role in deciphering disease severity and symptoms. Overall, we stratified 70 ASD patients by neuropsychological assessment, based on Calibrated Severity Scores (CSSs) of the Autism Diagnostic Observation Schedule-Second edition (ADOS-2), Child Behavior Checklist (CBCL) and intelligent quotient/developmental quotient (IQ/DQ) parameters. Hence, metataxonomy and PICRUSt-based KEGG predictions of fecal GM were assessed for each clinical subset. Here, 60% of ASD patients showed mild to moderate autism, while the remaining 40% showed severe symptoms; 23% showed no clinical symptoms, 21% had a risk of behavior problems and 56% had clinical symptoms based on the CBCL, which assesses internalizing problems; further, 52% had no clinical symptoms, 21% showed risk, and 26% had clinical symptoms classified by CBCL externalizing problems. Considering the total CBCL index, 34% showed no clinical symptoms, 13% showed risk, and 52% had clinical symptoms. Here, 70% of ASD patients showed cognitive impairment/developmental delay (CI/DD). The GM of ASDs with severe autism was characterized by an increase in Veillonella, a decrease in Monoglobus pectinilyticus and a higher microbial dysbiosis index (MDI) when compared to mild-moderate ASDs. Patients at risk for behavior problems and showing clinical symptoms were characterized by a GM with an increase of Clostridium, Eggerthella, Blautia, Intestinibacter, Coprococcus, Ruminococcus, Onthenecus and Bariatricus, respectively. Peptidoglycan biosynthesis and biofilm formation KEGGs characterized patients with clinical symptoms, while potential microbiota-activated PPAR-γ-signaling was seen in CI/DD patients. This evidence derived from GM profiling may be used to further improve ASD understanding, leasing to a better comprehension of the neurological phenotype.

Keywords: autism; autism severity; gut microbiota; microbial dysbiosis index (MDI); neuropsychological phenotypes.

Figure 1

Stratification representation of ASD patients based on anamnestic and neuropsychological features.

Figure 2

Beta diversity analysis based on Bray–Curtis dissimilarity algorithm calculated for ASD patients stratified by the confounding factor of age. PERMANOVA p-value is reported.

Figure 3

Alpha diversity of ASD patients grouped by CBCL_EXT and CBCL_TOT. Alpha diversity representation of ASD patients grouped by CBCL_EXT and CBCL_TOT based on Shannon (A), Simpson (B) and Chao1 (C) indices. Comparisons amongst the three subgroups (no symptoms; at risk; and symptoms) were performed by Kruskal–Wallis tests, while the sub-group pairwise comparisons were performed through a post-hoc Mann–Whitney test (ns: p-value > 0.05; ** p-value ≤ 0.01; * p-value < 0.05).

Figure 4

Microbial biomarkers discriminating ASD patients grouped by CBCL_EXT. Distribution of statistically significant genera for ASD patients stratified by CBCL_EXT (Kruskal–Wallis test, p-value < 0.05). The pairwise comparisons were obtained by a post hoc Mann–Whitney test (ns: p-value > 0.05; *** p-value ≤ 0.001; ** p-value ≤ 0.01; * p-value < 0.05).

Figure 5

Prediction of functional KEGG pathways performed by PICRUSt2 to discriminate ASD patients grouped by IQ/DQ. Statistically significant functional pathways were identified for ASDs patients stratified based on the IQ/DQ feature by the Limma Voom method (A) and LEfSe with LDA > 2 (B) (p-value < 0.05).

Figure 6

MDI distribution for ASD patients grouped by CSS. Boxplot of MDI distribution (percentage) in patients stratified by CSS.