ABSTRACT

The gut microbiome (GM) contains a vast diversity of microbial taxa that interact with the host, and these interactions are increasingly implicated in neurodegeneration. Notably, altered intestinal microbial compositions that correlate with gut and brain inflammatory indicators have been reported in individuals with Alzheimer’s disease (AD) and in murine models of AD. We have found that these changes in GM composition can occur even before symptoms of AD start, and that the inclusion of taxa significantly enriched during preclinical AD status improved the classification accuracy of machine learning (ML) models for preclinical AD (Ferreiro et al. 2023; accepted to Science Translational Medicine). However, despite evidence that the GM is linked to AD progression, a mechanistic understanding of how it influences neurodegeneration, especially during preclinical AD, is still missing.

We propose studying the host gut transcriptional state of AD patients in conjunction with their GM will provide a more mechanistic understanding of AD pathology. Host mRNA in stool samples, referred to as the “exfoliome,” can serve as non-invasive biomarkers for inflammatory conditions in the gut, such as enteropathy and inflammatory bowel disease (IBD). Despite such potential, the exfoliome of AD patients has not been characterized.

Here, we propose to test the hypothesis that changes to the host transcriptome that are reflective of AD can be studied through the host exfoliome, and that these discriminatory transcripts, in conjunction with GM markers, will improve classification accuracy of ML models. By interrogating ~100 banked stool samples from healthy, preclinical, and symptomatic AD patients, we propose the following three aims:

• Aim 1: Characterize the exfoliome of adults in different stages of AD
• Aim 2: Explore the correlation of the host exfoliome with classical AD biomarkers, and
• Aim 3: Test whether the inclusion of host transcripts associated with AD, as well as GM compositional and functional data can improve ML models for AD diagnosis.

This proposal seeks to improve our systems-level understanding of the gut-brain-axis in AD and to identify novel biomarkers for AD diagnosis through stool, establishing a foundation for cost-effective and noninvasive measures for AD diagnosis.

LAY SUMMARY

There is a growing body of evidence highlighting the role of the gut microbiome (GM) and inflammation in neurogenerative conditions such as Alzheimer’s disease (AD). For example, the GM composition of symptomatic AD individuals is significantly different from that of healthy individuals. However, we do not understand how this difference in GM composition affects the host.

Up until recently, the only way to study the host gut was through invasive tissue sampling. Here we propose a novel, non-invasive approach to study the effects of the GM on the host by sequencing host mRNA in stool, the exfoliome, representing the host gut response to the GM. By sequencing the mRNA found in healthy, preclinical, and symptomatic AD patient’s stools and integrating it with already existing GM composition data, we can ascertain whether the changes in the GM are reflected by the host and how these changes are related to AD progression.

Our proposed noninvasive and inexpensive method will enable us to achieve a systems-level understanding of the gut-brain axis in AD and to identify novel biomarkers for AD diagnosis from patient stool.