A large population analysis has identified a mutational hotspot at human transcription start sites. The work, published in Nature Communications, shows that the first 100 base pairs after a gene’s start site accumulate mutations at rates 35 percent higher than expected. The findings point to a previously overlooked source of heritable changes and add needed precision to models used across genetic research.
“These sequences are extremely prone to mutations and rank among the most functionally important regions in the entire human genome, together with protein-coding sequences,” said Donate Weghorn of the Centre for Genomic Regulation in Barcelona.
The team examined transcription start sites across 150,000 genomes from the UK Biobank and 75,000 genomes from gnomAD. They paired these data with mosaic mutation information from eleven family studies. The combined approach helped detect variants that arise shortly after conception, when early embryonic cell divisions introduce mutations in some cell lineages but not others.
The study also reports an excess of mutations near start sites for genes associated with cancer, brain function, and limb development. Variants in these regions were common among very rare alleles but diminished among older, more widespread alleles. The pattern suggests that many of these changes are removed by natural selection over generations.
Mechanistic analyses point to transcription itself as a contributor. The machinery that initiates RNA synthesis often pauses, restarts, or moves bidirectionally near the start site. These events can expose single DNA strands to damage during rapid early cell divisions. Some alterations persist because repair pathways cannot always keep pace.
Accurate models of mutational burden depend on correct baseline expectations. “If a model doesn’t know this region is naturally mutation-rich, it might expect, say, 10 mutations but observe 50. If the correct baseline is 80, then 50 means fewer than expected and is a sign harmful changes are being removed by natural selection. You would completely miss the importance of that gene,” said Weghorn.
Overall, these findings indicate that transcription start sites contribute measurably to the human germline mutation load and should be handled explicitly in future variant interpretation frameworks.
Publication Details
Cortés Guzmán, M., Castellano, D., Serrano Colomé, C. et al. Transcription start sites experience a high influx of heritable variants fueled by early development. Nat Commun 16, 10120 (2025). https://doi.org/10.1038/s41467-025-66201-0
“These sequences are extremely prone to mutations and rank among the most functionally important regions in the entire human genome, together with protein-coding sequences,” said Donate Weghorn of the Centre for Genomic Regulation in Barcelona.
The team examined transcription start sites across 150,000 genomes from the UK Biobank and 75,000 genomes from gnomAD. They paired these data with mosaic mutation information from eleven family studies. The combined approach helped detect variants that arise shortly after conception, when early embryonic cell divisions introduce mutations in some cell lineages but not others.
The study also reports an excess of mutations near start sites for genes associated with cancer, brain function, and limb development. Variants in these regions were common among very rare alleles but diminished among older, more widespread alleles. The pattern suggests that many of these changes are removed by natural selection over generations.
Mechanistic analyses point to transcription itself as a contributor. The machinery that initiates RNA synthesis often pauses, restarts, or moves bidirectionally near the start site. These events can expose single DNA strands to damage during rapid early cell divisions. Some alterations persist because repair pathways cannot always keep pace.
Accurate models of mutational burden depend on correct baseline expectations. “If a model doesn’t know this region is naturally mutation-rich, it might expect, say, 10 mutations but observe 50. If the correct baseline is 80, then 50 means fewer than expected and is a sign harmful changes are being removed by natural selection. You would completely miss the importance of that gene,” said Weghorn.
Overall, these findings indicate that transcription start sites contribute measurably to the human germline mutation load and should be handled explicitly in future variant interpretation frameworks.
Publication Details
Cortés Guzmán, M., Castellano, D., Serrano Colomé, C. et al. Transcription start sites experience a high influx of heritable variants fueled by early development. Nat Commun 16, 10120 (2025). https://doi.org/10.1038/s41467-025-66201-0