The molecular and urban exposome signature of telomere length in early life: data-driven insights from birth cohort studies

Abstract

Aging is a multi-factor physiological process characterized by molecular and cellular damage, functional decline and increased vulnerability to diseases and death. As one of the hallmarks of aging that are considered to contribute to the aging process, telomere attrition, the progressive and cumulative shortening of the repetitive nucleotide sequences at the end of chromosomes, occurs in age-related diseases as well as in normal aging throughout life. Aging has been extensively explored in adulthood. Nevertheless, differences in aging and developmental trajectories in early life may already set the vulnerability differences to diseases that could persist and accumulate over time. Complex interactions of cellular and biochemical processes are involved in biological aging and are reflected in the molecular profiles of various omics layers. The advent of high-throughput omics approaches enabled a global assessment of this interaction and identification of relevant biological pathways. Environmental factors contribute to the acceleration of aging processes and progression of a wide range of age-associated diseases. Early-life exposome, encompassing all environmental factors from the prenatal period to childhood, plays a critical role in life-course aging. Yet, evidence on the environmental determinants and molecular mechanisms underlying early-life aging is scarce. In this dissertation, we aim to evaluate the omics signatures and the health effect of exposome related to telomere length in early life, based on birth cohort data collected from mother-newborn pairs at birth and during childhood. Telomere length measured at both time points was used as an indicator of biological age of the population of the corresponding age, and telomere attrition from birth to childhood was used as a marker of biological aging. This doctoral dissertation consists of our works which fall into four chapters as summarized in Table I. In the first part, we modeled the variation in newborn telomere length and telomere attrition rate from birth to around 4 years with epigenome-wide DNA methylation profiled in cord blood at birth, considering both the first-order additive effect and two-way interactions of the methylation level at CpG loci. From 850K CpG loci, distinct epigenetic signatures were found for baseline biological age and aging rates: 47 CpGs and 7 between-CpG interactions explained 76% of the variance in newborn telomere length, while 72% of the total variance in telomere attrition rate was explained by 31 CpGs and five interactions. The second part of this dissertation focused on the potential interplay between genetic variants and epigenome-wide DNA methylation in setting the newborn telomere length. Starting from candidate single nucleotide polymorphisms (SNPs), we identified 57 cis methylation quantitative trait loci, of which 22 confirmed previous findings. Indirect effects were found in five SNPs via a mediating CpG, and the effect of rs911874 (SOD2) was modified by nearby DNA methylation. These findings suggested that DNA methylation in cis might have a mediation or modification effect on the genetic difference in newborn telomere length. The investigation of omics signatures of childhood biological age was then extended to multi-omics, including genome-wide genetic variants, epigenome-wide DNA methylation, gene expression and mRNA transcriptome, plasma protein profile, and serum and urine metabolites. Through omics-wide association analysis, two CpGs (cg23686403 and cg16238918 at PARD6G gene) were identified to strongly associate with childhood telomere length, while supervised multi-omics integration revealed robust associations between the multi-omics signature of telomere length and child body mass index, with metabolites and proteins emerging as the primary contributing molecular features. Finally, we explored the health effect in biological age reflected in childhood telomere length of the longitudinal urban environment, including the built environment, air pollution, noise, meteorology, natural space, and road traffic measures that are quantified both prenatally and postnatally. We identified, via both an exposome-wide association study and a mixture effect analysis, that higher early-life exposure to natural spaces and land use diversity, as well as lower exposure to the urban built environment and traffic-related factors, were associated with longer childhood telomere length. The findings from this dissertation prove the effectiveness of our data analysis workflow, suggest potential novel targets of further investigation and an intricate interplay between telomere length, metabolism and immune responses, and provide evidence that the reductions in adverse urban exposures may promote molecular longevity from early life onward.