A team of researchers at the Cleveland Clinic has found that machine learning can help understand how gut microbial metabolites and cell receptors interact to influence the course of Alzheimer’s disease. The study has been published in Cell Reports. The researchers said that previous studies have shown that Alzheimer’s patients often experience changes in gut bacteria as the disease progresses. However, many of the mechanisms driving this ‘gut-brain axis’ are unknown.
Impact of intestinal metabolites
Intestinal metabolites are released by bacteria when they break down food. These metabolites then influence many cellular processes throughout the body. The researchers noted that this could be beneficial to health in many cases. However, in addition to Alzheimer’s disease, links between metabolites and a wide range of conditions have been documented, such as cancer and Parkinson’s disease. The significant impact that the gut microbiome has in these cases has led researchers to investigate how drugs or other treatments can prevent the harmful interactions of metabolites.
“Gut metabolites are the key to many physiological processes in our body, both positive, for our health, and negative, for disease. The problem is that we have tens of thousands of receptors and thousands of metabolites in our system, so manually figuring out which key goes into which lock is slow and expensive,” said Feixiong Cheng, PhD, director of the Cleveland Clinic Genome Center.
Machine learning
In an attempt to better understand and simplify that process, the research team used machine learning. With the aim of studying how metabolites and cell receptors interact in the context of Alzheimer’s disease. By integrating information on metabolite shapes and receptor protein structures, genetic and proteomic data, and the known impact of metabolites on patient-derived brain cells, the approach enabled researchers to analyze more than 1.09 million potential metabolite-receptor pairs.
This data was used to rank metabolites and receptors based on the likelihood that they would interact with each other. That yielded predictions about the likelihood that the couple would influence the course of Alzheimer’s disease. Next, the metabolite-receptor pairs most likely to influence Alzheimer’s disease were further examined.
Agmatine and CA3R
It is generally believed that one of the relevant metabolites, agmatine, protects brain cells against inflammation-related damage. The machine learning analysis found that agmatine is most likely to interact with a receptor known as CA3R. Further research showed that agmatine and CA3R indeed influence each other.
Treating neurons affected by Alzheimer’s with agmatine lowered CA3R levels. In addition, the agmatine-treated neurons also showed significantly lower levels of phosphorylated tau proteins, known markers of Alzheimer’s disease. The researchers conclude that these results emphasize that AI applications also have potential in other studies assessing the link between the gut microbiome and diseases. “We hope that our methods can provide the foundation to make progress in the entire field of metabolite-associated diseases and health,” said Cheng.
There is no curative treatment for Alzheimer’s yet, but progress is being made in ‘inhibiting’ the disease in its early stages, including with the help of electrical stimulation (tDCS). A lot of research is being done worldwide into diagnosing Alzheimer’s at an early stage in order to slow down the progression of the disease.
