14 Jul 2017
Yitzhak (Tzachi) Pilpel
Weizmann Institute of Science, Israel
The rate of phenotypic errors, mainly made by RNA polymerase and the ribosome, is typically higher by orders of magnitude compared to the rate of genetic mutations. Indeed estimates are that around a fifth of proteins in cell carry a misincorporated amino acid. Here we tackle the challenge of systematically detecting and quantifying the errors in the proteome from mass spectrometry data in both E. coli and in S. cerevisiae. The error rate we observe is around 10-3 per amino acid on average but it varies considerably between amino acids, and in a manner that depends on the encoding codons. Treating bacteria with drugs that decrease proofreading within the ribosome exposed specific patterns of errors induced in each nucleotide position within codons. Substitutions tended to occur in positions that are less evolutionarily conserved and they tend to minimally affect the predicted protein stability, indicating that organisms have evolved means to minimize the level of phenotypic error when potentially detrimental. The substitution patterns we found in E. coli show significant similarity to those we detected in the yeast proteome suggesting a universal chemical basis for proteome errors. A kinetic proofreading model can explain some of the error patterns observed based on basic thermodynamics.
This is joint work with Ernest Mordret and Ariel Lindner