Gut microbial taxa elevated by dietary sugar disrupt memory function

The study investigates whether excessive consumption of added sugars during early life impairs adult hippocampal-dependent memory via the gut microbiome. Prior work indicates that gut microbes modulate neurocognitive development, with early life as a critical window, and that Western diet components (high saturated fat and sugar) produce lasting memory impairments, especially when exposure occurs during juvenile/adolescent development. The hippocampus is particularly vulnerable to such dietary factors. The purpose here is to test whether sugar-induced changes in specific gut bacterial populations during early life are causally related to hippocampal-dependent memory impairments observed in adulthood.

Evidence shows: (1) the gut microbiome influences neurocognitive development, with abnormalities in germ-free rodents reversible only when colonization occurs early in life; (2) diet is a major determinant of microbiota diversity, with Western diets causing dysbiosis; (3) consumption of Western diet factors, including sugar alone, impairs hippocampal function more strongly when consumed during juvenile/adolescent periods; (4) the precise biological pathways linking diet to neurocognitive dysfunction remain unclear, and the functional relationship between diet-induced microbiome changes and early-life neurocognitive outcomes is poorly understood.

Subjects: Juvenile male Sprague Dawley rats (postnatal day, PN 26–28; 50–70 g), singly housed under a 12:12 light/dark cycle with ad libitum water and standard chow. All procedures were IACUC-approved. Experiment 1 (early-life sugar exposure): Rats were assigned to either an 11% w/v sugar solution (65% fructose, 35% glucose) modeling human SSBs (SUG) or an extra bottle of water (CTL), with ad libitum access to chow and water. Food/solution intake and body weight were monitored thrice weekly. Behavioral testing: NOIC at PN 60 (hippocampal-dependent episodic contextual memory), Zero Maze at PN 67 (anxiety-like behavior), body composition at PN 70, and intraperitoneal glucose tolerance test (IP GTT) at PN 84. Blinded scoring was used. Fecal and cecal samples were collected at PN 104. In a separate cohort (n=6/group), NOR and Open Field (OF) were tested around PN 60; dorsal hippocampus punches collected at PN 65 and stored at −80 °C. Experiment 2 (taxa-specific enrichment): Rats received twice-daily gavage for 7 days of either saline or an antibiotic cocktail (Vancomycin 50 mg/kg, Neomycin 100 mg/kg, Metronidazole 100 mg/kg) plus ampicillin (1 mg/mL) in drinking water; sterile housing and handling were maintained. Thirty-six hours after the last antibiotic dose, ABX-treated rats received oral gavage with a 1:1 mix of Parabacteroides distasonis (ATCC 8503) and Parabacteroides johnsonii (DSM 18315) or saline. Bacteria were grown anaerobically and authenticated by full-length 16S rRNA sequencing; enrichment dose was 10 CFU of each strain in 500 µL pre-reduced PBS, co-administered 1:1. Behavioral schedule: NOIC (PN 50), NOR (PN 60), Zero Maze (PN 62), OF (PN 64), IP GTT (PN 73), body composition (PN 76); sacrifice at PN 83 with dorsal hippocampus punches and cecal samples collected and stored at −80 °C. Behavioral assays: NOIC (contextual object recognition with 3-day protocol and counterbalancing; discrimination index calculated as time exploring context-novel object divided by total exploration). NOR (5-min familiarization with identical objects; 3-min test with one novel object after 5-min delay; tracked with Any-Maze). Zero Maze (time in open sections and entries). OF (10-min exploration; distance traveled and center time measured). Microbiome profiling: For sugar-fed vs control rats, DNA extraction followed Earth Microbiome Project protocols using MO BIO PowerSoil; V4 region amplified with 515F/806R; libraries barcoded, pooled, purified, and sequenced on Illumina MiSeq 2×150 bp at UCSD; sequences deposited in Qiita Study 11255; demultiplexed with QIIME1 split libraries; reads trimmed to 100/150 bp; OTUs matched at 97% identity. For Parabacteroides-enriched vs control rats, DNA was extracted with Qiagen DNeasy PowerSoil; V4 amplicons (515F-806R) prepared, triplicate PCRs pooled and purified; Illumina MiSeq 2×250 bp; ASVs generated via Deblur; taxonomy/rarefaction via QIIME2-2019.10. Hippocampal transcriptomics: Total RNA from hippocampus (sugar vs control; Parabacteroides vs control) extracted with RNeasy Lipid Tissue Mini Kit; libraries prepared with NuGen Universal Plus mRNA-seq; sequenced on Illumina NextSeq 550 (SR75, ~25M reads/sample) at USC Genome Core. QC via FastQC; trimming with Trim Galore; quantification with Salmon to Rnor6 transcriptome; gene-level counts via tximport. PCA identified outliers; one control and one treatment sample (Parabacteroides experiment) removed. Differential expression and pathways: DESeq2 used for DEGs (genes with mean raw count >1 in >50% samples retained); significant DEGs defined at BH FDR<0.05; VST normalization and z-scores for heatmaps. Pathway enrichment (unadjusted P<0.01 gene sets) via Enrichr against KEGG, Gene Ontology (BP/CC/MF), and WikiPathways; pathways significant at FDR<0.05. Statistics: Data as mean±SEM. Repeated-measures ANOVA for body weight and intake; ANCOM for taxonomic differences with Sidak post hoc tests; AUC calculated for IP GTT; other comparisons via two-tailed unpaired t-tests. Normality confirmed; Grubbs’ test (alpha=0.05) for outlier exclusion.

- Early-life sugar consumption impaired adult hippocampal-dependent memory: rats given an 11% sugar solution from adolescence showed reduced discrimination of the context-novel object in NOIC at PN 60, without changes in total object exploration time. NOR performance (perirhinal cortex-dependent under these conditions) was unaffected. - Anxiety-like behavior and general activity were unchanged by early-life sugar: no differences in time in open sections or entries in the Zero Maze; no differences in distance traveled or center time in the Open Field. - Metabolic effects: early-life sugar did not alter total kcal intake, body weight, or adiposity; rats compensated for liquid sugar calories by reducing chow intake. However, sugar-exposed rats exhibited impaired glucose tolerance in IP GTT during adulthood. - Gut microbiome alterations: adolescent sugar intake increased the abundance of Parabacteroides distasonis and Parabacteroides johnsonii; these taxa were negatively correlated with hippocampal function. - Causality via taxa enrichment: antibiotic-conditioned enrichment with P. distasonis and P. johnsonii in adolescent rats impaired hippocampal-dependent memory in adulthood, mirroring the sugar-induced deficit. - Hippocampal transcriptomics: early-life sugar altered gene expression in intracellular kinase and synaptic neurotransmitter signaling pathways; Parabacteroides enrichment altered pathways related to metabolic function, neurodegenerative disease, and dopaminergic signaling.