This paper provides the first comprehensive analysis of the fidelity of transcription in eukaryotic cells.

Transcription errors occur in all living cells; however, it is unknown how these errors affect cellular health. To answer this question, we monitor yeast cells that are genetically engineered to display error-prone transcription. We discover that these cells suffer from a profound loss in proteostasis, which sensitizes them to the expression of genes that are associated with protein-folding diseases in humans; thus, transcription errors represent a new molecular mechanism by which cells can acquire disease phenotypes. We further find that the error rate of transcription increases as cells age, suggesting that transcription errors affect proteostasis particularly in aging cells. Accordingly, transcription errors accelerate the aggregation of a peptide that is implicated in Alzheimer's disease, and shorten the lifespan of cells. These experiments reveal a previously unappreciated role for transcriptional fidelity in cellular health and aging.

Genetic instability of the mitochondrial genome (mtDNA) plays an important role in human aging and disease. Thus far, it has proven difficult to develop successful treatment strategies for diseases that are caused by mtDNA instability. To address this issue, we developed a model of mtDNA disease in the nematode C. elegans, an animal model that can rapidly be screened for genes and biological pathways that reduce mitochondrial pathology. These worms recapitulate all the major hallmarks of mtDNA disease in humans, including increased mtDNA instability, loss of respiration, reduced neuromuscular function, and a shortened lifespan. We found that these phenotypes could be rescued by intervening in numerous biological pathways, including IGF-1/insulin signaling, mitophagy, and the mitochondrial unfolded protein response, suggesting that it may be possible to ameliorate mtDNA disease through multiple molecular mechanisms.

Significance: Accurate transcription is required for the faithful expression of genetic information. If the fidelity of transcription is compromised, protein homeostasis, cell fate, and various metabolic processes are disrupted. Here, we demonstrate that mutagens can lower the fidelity of transcription in yeast, worms, flies, and mice by promoting the misincorporation of nucleotides by RNA polymerases. These observations establish a mechanism by which the environment, our lifestyle choices, and work-related exposures may promote the age of onset, severity, and progression of various diseases.

Accurate transcription is required for the faithful expression of genetic information. However, relatively little is known about the molecular mechanisms that control the fidelity of transcription, or the conservation of these mechanisms across the tree of life. To address these issues, we measured the error rate of transcription in five organisms of increasing complexity and found that the error rate of RNA polymerase II ranges from 2.9 × 10⁻⁶ ± 1.9 × 10⁻⁷/bp in yeast to 4.0 × 10⁻⁶ ± 5.2 × 10⁻⁷/bp in worms, 5.69 × 10⁻⁶ ± 8.2 × 10⁻⁷/bp in flies, 4.9 × 10⁻⁶ ± 3.6 × 10⁻⁷/bp in mouse cells and 4.7 × 10⁻⁶ ± 9.9 × 10⁻⁸/bp in human cells. These error rates were modified by various factors including aging, mutagen treatment and gene modifications. For example, the deletion or modification of several related genes increased the error rate substantially in both yeast and human cells. This research highlights the evolutionary conservation of factors that control the fidelity of transcription. Additionally, these experiments provide a reasonable estimate of the error rate of transcription in human cells and identify disease alleles in a subunit of RNA polymerase II that display error-prone transcription. Finally, we provide evidence suggesting that the error rate and spectrum of transcription co-evolved with our genetic code.

Although vaccines and treatments have strengthened our ability to combat the COVID-19 pandemic, new variants of SARS-CoV-2 continue to emerge in human populations. Because the evolution of SARS-CoV-2 is driven by mutation, a better understanding of its mutation rate and spectrum could improve our ability to forecast the trajectory of the pandemic. Here, we use circular RNA consensus sequencing (CirSeq) to determine the mutation rate of six SARS-CoV-2 variants and perform a short-term evolution experiment to determine the impact of these mutations on viral fitness. Our analyses indicate that the SARS-CoV-2 genome mutates at a rate of ∼1.5 × 10⁻⁶/base per viral passage and that the spectrum is dominated by C → U transitions. Moreover, we find that the mutation rate is significantly reduced in regions that form base-pairing interactions and that mutations that affect these secondary structures are especially harmful to viral fitness. In this work, we show that the biased mutation spectrum of SARS-CoV-2 is likely a result of frequent cytidine deamination and that the secondary structure of the virus plays an important role in this process, providing new insight into the parameters that guide viral evolution and highlighting fundamental weaknesses of the virus that may be exploited for therapeutic purposes.