NEW YORK – Researchers from the UK and Switzerland are teasing out interconnections between gene expression, metabolites, and other genomic features in blood samples from hundreds of individuals over time to better understand the dynamic interactions behind aging and age-related disease.
"Multiomics data also has enormous utility in identifying the functional mechanisms underlying disease states, and linking genetic variants to their downstream effects on physiology," explained King's College London's Kerrin Small, a leader in genomics in the twin research and genetic epidemiology department, who shared the findings at the American Society of Human Genetics annual meeting on Thursday.
As part of the multiomic multiple tissue human expression resource, or MultiMuTHER, project, the researchers used RNA sequencing and metabolomics to track blood gene expression and metabolite profiles, respectively, in samples collected over time from 335 female TwinsUK participants. The participants came from both identical and non-identical twin pairs and ranged in age from roughly 30 to 85 years old at the time of their first sampling visit, Small said, noting that most of the individuals were in their 50s or 60s when the study began.
Over nine years, the team collected three or more samples from each participant, generating RNA-seq profiles for 16,292 genes that were analyzed in whole blood alongside Metabolon-based profiles for nearly 1,200 metabolites in matched blood serum samples.
From these longitudinal samples, the investigators found that the collection of expressed genes tended to remain steady within each individual. And while expression levels were sometimes dialed up or down with age across the participant population, the expression of specific genes sometimes bucked that trend within a subset of individuals, shifting in the opposite direction or remaining steady over time.
"We hope to use the other variables in the dataset to determine whether these individual trajectories are environmentally, clinically, or genetically driven," Small explained, noting that the analyses done so far have taken potentially confounding factors into account, such as participants' age at study onset, seasonality, and the cell type composition of blood samples.
Along with similar analyses on transcript splicing and metabolite profiles in the participants over time, the team went on to unearth more than 105,600 gene expression-metabolite associations, which involved more than 80 percent of the genes and 95 percent of the metabolites analyzed.
"Genes showing longitudinal change over time were found to have a higher number of gene-metabolite associations than those exhibiting stable expression," Small noted, "whereas metabolites exhibiting longitudinal variation did not show a difference in the number of associated genes."
Following up on such associations, the team took a closer look at everything from the nature of the most association-rich metabolites or environmental metabolites impacting gene expression to the stability of gene-metabolite associations over time and related genotypes.
As such analyses continue to progress, the researchers are also planning to layer on clinical phenotype data to try to tease out the potential consequences of the stable and variable associations they are uncovering.
"[W]e have performed one of the largest multiomic longitudinal studies of concurrently measured gene expression and metabolite levels in whole blood, identifying over 100,000 gene-metabolite associations," Small and her co-authors concluded in an abstract for the presentation, arguing that the study "provides novel insight into the interplay between gene expression and metabolites, and may inform systems-wide approaches to projection of temporal progression of age-related diseases."