Gut microbial composition mediates the relationship between intestinal inflammation and amyloid burden before cognitive decline

• Published in: Alzheimer's & dementia Vol. 19; no. S10
• Main Authors: Heston, Margo B., Hong, Qilin, González, Antonio, Betthauser, Tobey J, Chin, Nathaniel A., Hanslik, Kendra L., Zarbock, Katie R, Kerby, Robert L., Asthana, Sanjay, Johnson, Sterling C, Knight, Rob, Kaddurah‐Daouk, Rima F., Tang, ZhengZheng, Ulland, Tyler K., Rey, Federico E., Bendlin, Barbara B
• Format: Journal Article
• Language: English
• Published: 01.12.2023
• ISSN: 1552-5260; 1552-5279
• DOI: 10.1002/alz.081862

Background Gut microbial composition is associated with Alzheimer’s disease (AD) pathology, and recent findings (Heston et al, medRxiv 2022) have identified an increase in intestinal inflammation with greater age and with higher cortical amyloid burden. Although gut microbiome composition and intestinal inflammation have independently been linked to AD pathology, it remains unclear whether the microbiome and gut inflammation are interrelated in the context of preclinical AD. Here we used phylogenetically‐based mediation (PhyloMed) to test whether the gut microbiome mediates the relationship between gut inflammation and cortical amyloid burden, and to highlight bacterial clades that may be specifically implicated as mediators. Method 78 participants (Table 1) were enrolled in the Wisconsin Alzheimer’s Disease Research Center and Wisconsin Registry for Alzheimer’s Prevention, collected fecal samples, and completed C11‐PiB PET. Figure 1 illustrates the processing and analysis for each data stream. Fecal calprotectin was measured to quantify intestinal inflammation, and 16S rRNA sequencing characterized the fecal microbiome. Testing the full cohort and the cognitively unimpaired (CU) subset separately (n = 73), PhyloMed was performed at every taxonomic level, resulting in 14 models. Covariates included age, sex, body mass index, Bristol Stool Scale score, cognitive diagnosis/amyloid status, pseudocontinuous APOE AD risk score, and age difference between PET scan and fecal sample. Result For both the full and CU cohorts, models at the amplicon sequence variant (ASV) level showed that gut microbiome composition was a significant mediator (p<0.05); species‐ and genus‐level models were marginally significant (p<0.1, Benjamini‐Hochberg FDR) (Table 2). ASVs from the family Lachnospiraceae (Figure 2A) were highlighted as significantly mediating taxa (p<.05, FDR). Interestingly, these taxa were negatively associated with calprotectin, while they were positively associated with PiB DVR (Figures 2B, 2C). Conclusion Even before cognitive decline, the gut microbiome mediated the relationship between calprotectin, a measure of gut inflammation, and cortical PiB DVR, a measure of amyloid burden. Taxa within Lachnospiraceae were implicated as specific mediators, linked with lower inflammation and higher amyloid burden. Future work will use germ‐free mouse models to functionally elucidate the role of these bacteria in inflammation and AD.