Molecular Noise and Loss of Gene-Expression Fidelity in Ageing

Cells of the same type do not express every gene at exactly the same level. Some variation is an ordinary consequence of probabilistic molecular events, changing cell states, and responses to the local environment. In ageing research, the term transcriptional noise is usually used more narrowly for increased cell-to-cell variability that remains after relevant biological differences have been considered. [1] [2] [6]

Gene-expression fidelity is broader. It includes whether transcription begins at the intended site, whether RNA polymerase copies a gene accurately and at an appropriate rate, and whether a coherent, correctly structured RNA product is produced. Age-associated changes have been observed at each of these levels, but they should not be collapsed into a single measurement or mechanism. [7] [8] [9] [10]

Who This Is Useful For

This page is useful for readers interpreting claims that ageing makes gene expression “noisier” or causes cells to lose their identity. It distinguishes several phenomena that can produce that description and explains why single-cell measurements, bulk-tissue averages, and direct assays of transcriptional errors answer different questions. [6] [12]

Several Meanings of Reduced Fidelity

Evidence from Individual Cells

An early single-cell study found greater expression variability for a panel of genes in cardiomyocytes from old mice than in those from young mice. The increase paralleled somatic genome rearrangements, but the study did not establish that DNA changes alone caused the expression differences. [1]

Later work extended the observation to other settings. Following immune stimulation, naive CD4 T cells from old mice showed a more variable transcriptional response than cells from young mice. Ageing mouse lung and muscle-stem-cell studies also reported greater variability within multiple cell types; the muscle study linked transcriptional heterogeneity with heterogeneous promoter methylation. [2] [4] [5]

These results are not uniform. A reanalysis applying several noise metrics to seven ageing single-cell datasets found substantial differences among tissues, datasets, and methods, with no robust universal increase. This matters because changes in cell-type composition, rare subpopulations, RNA abundance, sampling depth, and computational definitions can all be mistaken for altered noise. [6] [12]

Chromatin and Unintended Transcription Start Sites

Chromatin helps restrict access to DNA and distinguish genuine promoters from promoter-like sequences inside genes. In aged mammalian stem cells, increased transcription from internal sites was associated with lower H3K36me3 and a more promoter-like local chromatin state. Studies of cellular senescence and aged mouse liver similarly detected intragenic initiation alongside altered histone acetylation and DNA methylation. [9] [10]

These findings support a mechanism in which altered chromatin permits transcripts to begin at sites that are normally suppressed. The resulting RNAs may be shortened or otherwise abnormal, but their abundance, stability, translation, and functional importance can differ by locus and cell type. [9] [10]

RNA Polymerase Can Become Too Fast or Become Stalled

A cross-species analysis estimated that average RNA-polymerase-II elongation speed increased with age in nematodes, fruit flies, mice, rats, and human samples. Faster elongation was accompanied by changes in splicing and more RNA-to-genome mismatches. Because most changes were measured at the level of averages, this result describes altered transcriptional kinetics rather than cell-to-cell noise. [7]

A separate study of aged mouse liver found a different but compatible problem: RNA polymerase stalled at transcription-blocking DNA lesions. The effect accumulated across gene bodies, reduced complete RNA output, and disproportionately affected long genes. Ageing can therefore disturb transcription through more than one kinetic failure, and faster average movement does not imply that every polymerase moves smoothly through every gene. [8]

Potential Cellular Consequences

Variable expression can make a population respond less synchronously even when its average expression appears little changed. This was visible in old mouse T cells, where stimulation produced more heterogeneous activation-associated expression. Such variability could matter most when a tissue depends on coordinated responses, although the functional effect must be tested in each setting. [2]

Direct transcriptional mistakes may also propagate beyond RNA. In engineered yeast with error-prone transcription, errors impaired protein homeostasis and shortened cellular replicative lifespan. This demonstrates a possible causal route from transcription errors to cellular dysfunction, but it does not by itself quantify how much naturally ageing mammalian tissues are affected through that route. [11]

Evidence Quality and Interpretation

Confidence is strongest that ageing can alter multiple aspects of transcription in particular biological contexts. Independent studies have identified altered variability, cryptic initiation, elongation-speed changes, polymerase stalling, and sequence errors using different experimental systems. [2] [7] [8] [9] [10] [11]

Confidence is lower that a single quantity called “molecular noise” rises universally with age or can be compared directly across tissues. Single-cell RNA sequencing captures only part of each cell's RNA, and inferred variability is sensitive to cell classification, expression level, tissue composition, sequencing protocol, and statistical method. Cross-dataset analyses therefore provide an important check on claims made from one tissue or one metric. [6] [12]

What This Does Not Mean

• It does not mean all variation is harmful; controlled heterogeneity is part of normal cell biology, and some ageing datasets do not show increased noise. [3] [6]
• It does not mean an altered tissue average proves that individual cells have become noisier; cell proportions and cell-type-specific expression both change with age. [6] [12]
• It does not mean transcriptional variability, cryptic initiation, polymerase stalling, and copying errors are interchangeable measurements. [7] [8] [10]
• It does not establish loss of gene-expression fidelity as the single cause of ageing; current evidence places it among interacting molecular changes whose importance varies by context. [6] [12]

Practical Interpretation Examples

• If a bulk tissue shows wider expression differences with age: First ask whether its mix of cell types changed; that observation alone does not identify cell-intrinsic noise. [6] [12]
• If aged cells cluster less tightly in single-cell data: The pattern may reflect greater biological variability, but cell quality, RNA content, hidden subtypes, and the chosen distance metric also require examination. [6]
• If abnormal RNA fragments increase: Cryptic initiation, altered splicing, incomplete transcription, and RNA degradation are distinct explanations that require different assays. [7] [8] [9]

Summary

Ageing is associated with several forms of reduced gene-expression fidelity: some cell populations become more variable, normally silent transcription start sites can become active, RNA polymerase kinetics can change, and direct transcription errors can increase. These findings form a credible but heterogeneous body of evidence. The most defensible conclusion is that fidelity can decline through multiple, context-dependent mechanisms—not that every ageing tissue undergoes one uniform rise in molecular noise. [6] [7] [8] [9] [10]

References

1. Bahar, R., et al. (2006). “Increased cell-to-cell variation in gene expression in ageing mouse heart.” Nature. https://www.nature.com/articles/nature04844
2. Martinez-Jimenez, C. P., et al. (2017). “Aging increases cell-to-cell transcriptional variability upon immune stimulation.” Science. https://pmc.ncbi.nlm.nih.gov/articles/PMC5405862/
3. Sarnoski, E. A., et al. (2017). “Noise reduction as an emergent property of single-cell aging.” Nature Communications. https://www.nature.com/articles/s41467-017-00752-9
4. Angelidis, I., et al. (2019). “An atlas of the aging lung mapped by single cell transcriptomics and deep tissue proteomics.” Nature Communications. https://www.nature.com/articles/s41467-019-08831-9
5. Hernando-Herraez, I., et al. (2019). “Ageing affects DNA methylation drift and transcriptional cell-to-cell variability in mouse muscle stem cells.” Nature Communications. https://www.nature.com/articles/s41467-019-12293-4
6. Ibañez-Solé, O., et al. (2022). “Lack of evidence for increased transcriptional noise in aged tissues.” eLife. https://elifesciences.org/articles/80380
7. Debès, C., et al. (2023). “Ageing-associated changes in transcriptional elongation influence longevity.” Nature. https://www.nature.com/articles/s41586-023-05922-y
8. Gyenis, A., et al. (2023). “Genome-wide RNA polymerase stalling shapes the transcriptome during aging.” Nature Genetics. https://www.nature.com/articles/s41588-022-01279-6
9. McCauley, B. S., et al. (2021). “Altered chromatin states drive cryptic transcription in aging mammalian stem cells.” Nature Aging. https://www.nature.com/articles/s43587-021-00091-x
10. Sen, P., et al. (2023). “Spurious intragenic transcription is a feature of mammalian cellular senescence and tissue aging.” Nature Aging. https://www.nature.com/articles/s43587-023-00384-3
11. Vermulst, M., et al. (2015). “Transcription errors induce proteotoxic stress and shorten cellular lifespan.” Nature Communications. https://www.nature.com/articles/ncomms9065
12. The Tabula Muris Consortium. (2020). “A single-cell transcriptomic atlas characterizes ageing tissues in the mouse.” Nature. https://www.nature.com/articles/s41586-020-2496-1