Silent mutations alter cell fitness in humans, PE screens reveal

Researchers at Peking University used the PEmax prime editing system and a library of nearly 300,000 engineered pegRNAs to introduce synonymous mutations across 3,644 protein-coding human genes in HCT116 cells. Synonymous mutations are single-nucleotide changes in coding DNA that do not alter the amino acid sequence of the resulting protein due to the redundancy of the genetic code. Traditionally viewed as biologically silent, these mutations were thought to have little or no impact on cellular function.

Their screening platform, termed PRESENT, enabled them to assess the impact of these mutations on cell proliferation. While the vast majority of synonymous mutations appeared neutral, a small subset – 409 out of 94,993 tested – exhibited measurable effects on cell fitness. These were systematically validated and found to alter biological processes such as mRNA splicing, RNA secondary structure, and translation initiation.

One prominent case, PLK1_S2 (AGT>AGC), led to altered mRNA folding and reduced ribosome binding, diminishing protein levels. Another, BUB1B_R322 (AGG>AGA), disrupted splicing by eliminating a canonical donor site, leading to aberrant transcript variants and impaired gene expression.

The team developed a machine learning model, DS Finder, trained on the screen data to predict deleterious synonymous mutations and applied it to variants in clinical databases. Their approach identified likely pathogenic mutations previously annotated as benign, such as G6PC3 c.G399A, suggesting that functional synonymous variants may be under-recognised in human disease.

These results contrast with prior yeast studies suggesting widespread functional impact from synonymous changes and instead support a model where such effects are rare but biologically meaningful in humans.

The study was led by Ying Liu and Wensheng Wei at Peking University, China. It was published in Nature Biotechnology today 24 June 2025.