Landscape of somatic mutations in aging human heart muscle cells

Abstract

Using single-cell whole-genome sequencing, we identified and characterized the landscape of somatic single-nucleotide variants (sSNVs) in single cardiomyocytes from individuals across the human lifespan. Aged cardiomyocytes were found to have a higher burden of sSNVs and show mutational signatures that suggest failed repair of oxidative DNA damage.

EXPERT OPINION

This study shows that somatic mutations accumulate with age in individual human cardiomyocytes. The mutations likely result from oxidative damage, reflecting the heart’s high energy demand. This new information is important because these accumulated mutations could also cause cellular dysfunction even though most cardiomyocytes rarely divide. This study could shed light on why the aging human heart is more prone to dysfunction under many different circumstances.” Richard T. Lee, Harvard University, Cambridge, MA, USA

FROM THE EDITOR

Recent studies have provided insight into the somatic mutation landscape in various normal tissues to understand its contribution to aging. Here is a technically difficult-to-perform study that provides the first of these data for human heart.” Editorial Team, Nature Aging

The question

Age is the single most important risk factor for many heart diseases, including atherosclerosis and heart failure. However, despite being so clinically relevant, our understanding of how aging promotes disease progression is incomplete. It is thought the cumulative exposure of traditional risk factors, such as hypertension, hypercholesterolemia and heritable traits, determine the effect of aging on the heart. Yet clinical evidence indicates that some individuals at low or intermediate risk as based on traditional risk factors still experience heart disease¹, suggesting that additional, unidentified factors might be important. The unmet question in the field is: what else is driving heart disease in the older population? Accumulating evidence over the last five years suggests that non-inherited acquired mutations affecting somatic cells are a hallmark of aging in many human cell types². We hypothesized that such mutations are an important player in heart disease and potentially other age-related disorders.

The discovery

Single-nucleotide variants, the most common form of somatic mutations, have long been considered critical in cancer, although their role in non-cancer genomes has long been difficult to determine owing to technical challenges³. Evidence from the last five years suggests that somatic mutations are important in common diseases, including coronary artery disease. Moreover, generic patterns of somatic mutations (so-called mutational signatures) have been linked to specific mutagens and biological processes, such as tobacco smoke or impaired DNA repair⁴. Understanding the mutational signatures and their mechanism of formation might lead us to discover the mechanism of DNA damage and disease progression in the aging heart. With this in mind, we evaluated the landscape of sSNVs and associated mutational signatures in the aging human heart muscle cell (cardiomyocyte).

Using single-cell whole-genome sequencing combined with the LiRA algorithm, a method for phasing-based somatic mutation calling, we evaluated the genome-wide burden of sSNVs from single cardiomyocyte nuclei derived from post-mortem hearts of infant, middle-aged and aged individuals. Our analysis revealed a significantly higher burden of sSNVs in the aged heart muscle cell (Fig. 1a), corresponding to a rate of 124 sSNVs per cell per year. Moreover, we found that sSNVs were distributed broadly across the genome. Decomposition of cardiac sSNVs into COSMIC cancer signatures highlighted two specific mutagenesis processes in aging heart, which are associated with oxidative damage and defective DNA repair pathways (Fig. 1b). Finally, our analysis of mutational signatures suggested a model whereby oxidative stress leads to an increased burden of damaged DNA bases, which might overwhelm excision repair pathways⁵ in aged cardiomyocytes. Cardiomyocytes with more copies of chromosomes can better tolerate the damage than diploid cardiomyocytes, where mutations in genes would be especially damaging if both alleles are affected (Fig. 1c).

Fig. 1 |. sSNVs in the aging heart.

a, sSNV density in cardiomyocytes across the human lifespan. Each data point represents estimated sSNVs in a single cardiomyocyte. b, Schematic of proposed mechanism of sSNV occurrence in the heart. Cardiomyocytes with increased age show elevated levels of oxidative stress (ROS), which results in a mismatched pairing of nucleotides in the genome. Overwhelmed repair machinery in aged cardiomyocytes might not function effectively to repair DNA damage, leading to fixed sSNVs. BER, base excision repair; KO, knockout; MMR, mismatch repair; NER, nucleotide excision repair. c, Cardiomyocytes with higher ploidy can better tolerate the deleterious effect of accumulated somatic mutations. © 2022, Choudhury, S. et al., CCBY 4.0.

The interpretation

Our study revealed that aging results in increased generation of oxidative DNA lesions, decreased repair, or both, and ultimately leads to fixed double-stranded sSNVs in cardiomyocytes. Our analysis indicates that the number of sSNVs and the likelihood of disrupting essential gene function in human cardiomyocytes significantly increases with age, suggesting that the age-related cellular dysfunction in aged cardiomyocytes could be partially due to somatic mutations. Functional studies will be needed to determine the causal relationship between mutational burden and age-associated decrease in cardiac function.

A limitation of our study is that we restricted our analysis to sSNV burden, whereas cardiomyocytes could carry other types of somatic mutations, such as insertions, deletions, copy number variations and structural variations. Broader genomic coverage for variant calling and larger sample sizes will ultimately provide everimproved estimates of sSNV accumulation rate and mutational signatures to identify specific causative mechanisms. A further limitation is that our sample might be biased towards cells that amplified well and evenly, given that the multiple displacement amplification (MDA) method used requires extensive quality control testing to identify well-amplified samples. Nevertheless, our findings provide an early view into the somatic mutation landscape of terminally differentiated cardiomyocytes. Elucidating the cardiomyocyte mutational landscape might help us to understand the mechanism of aging in the heart, with implications for the design of new treatments to target age-related heart diseases.

BEHIND THE PAPER.

Given the intensive oxidative metabolism of cardiomyocytes and the fact that neurons from the aging brain show evidence of oxidative damage, we wanted to decipher whether DNA damage over time results in the accumulation of fixed, double-stranded somatic mutations in cardiomyocytes. To achieve this goal, we had to overcome two main roadblocks. The first was to isolate cardiomyocyte nuclei with high purity; we spent months before figuring out that ploidy status was the most efficient and confident method to identify cardiomyocyte nuclei. The second roadblock was the quantitative detection of the somatic mutations in tetraploid single cardiomyocytes after whole-genome amplification and whole-genome sequencing because of the nature of the polyploidization in the cardiomyocyte. Our bioinformatic collaborators had to be very innovative to come up with the idea of modifying LiRA to call the sSNV from the tetraploid cardiomyocyte. S.C.