Universal and taxon-specific trends in protein sequences as a function of age

Abstract

Extant protein-coding sequences span a huge range of ages, from those that emerged only recently in particular lineages, to those present in the last universal common ancestor. Because evolution has had less time to act on young sequences, there might be "phylostratigraphy" trends in any properties that evolve slowly with age. Indeed, a long-term reduction in hydrophobicity and in hydrophobic clustering has been found in previous, taxonomically restricted studies. Here we perform integrated phylostratigraphy across 435 fully sequenced and dated eukaryotic species, using sensitive HMM methods to detect homology of protein domains (which may vary in age within the same gene), and applying a variety of quality filters. We find that the reduction in hydrophobic clustering is universal across diverse lineages, showing limited sign of saturation. But the tendency for young domains to have higher protein structural disorder, driven primarily by more hydrophilic amino acids, is found only among young animal domains, and not young plant domains, nor ancient domains predating the existence of the last eukaryotic common ancestor. Among ancient domains, trends in amino acid composition reflect the order of recruitment into the genetic code, suggesting that events during the earliest stages of life on earth continue to have an impact on the composition of ancient sequences.

Random peptides rich in small and disorder-promoting amino acids are less likely to be harmful

“…Given the hypothesis that young genes should first do no harm, we next tested whether the amino acids that are relatively benign in our laboratory data are similar to the amino acids enriched in purportedly young protein-coding sequences in natural populations. We took the latter from James et al 2020 trends reflect a de novo gene birth process that enriches for the most abundant amino acids in their respective lineages, such as cysteine, rather than for amino acids that promote ISD (James, et al 2020). We find that ISD still predicts harmlessness in E. coli, even though we do not find evidence it shaped de novo gene birth in its distant ancestors.…”

“…That is, if phylostratigraphy trends were due to an artifact such as homology detection bias, such an artifact would be highly unlikely to bias our random peptide analysis in the same direction. This conclusion is also supported by the fact that these slopes differ among taxa and are not correlated with amino acid changeability (James, et al 2020).…”

“…Second, previous work found that young proteins have their hydrophobic amino acids more clustered together along the primary sequence than expected by chance, while old proteins have them more interspersed (Foy, et al 2019;James, et al 2020). This hydrophobic dispersion metric (Irback, et al 1996;Irback and Sandelin 2000), which is normalized to exclude any influence from amino acid composition, did not significantly predict the fitness of our random peptides on its own (P = 0.2, likelihood ratio test in mixed model), nor did including it significantly improve our model with amino acid composition (P = 0.2, Supplemental Table S1).…”

“…Younger protein-coding sequences should be particularly constrained to first do no harm, as they have had little time to evolve more sophisticated harm-avoidance strategies (Foy, et al 2019). In support of the idea that high ISD is an accessible way to avoid harm, young animal and fungal domains (James, et al 2020) and genes (James, et al 2020), and novel overprinted viral genes (Willis and Masel 2018) have higher predicted disorder than their older counterparts. While some studies have found that putative de novo protein candidates in Saccharomyces yeasts have lower rather than higher ISD (Carvunis, et al 2012;Basile, et al 2017;Vakirlis, et al 2018), this could be an artifact of the inclusion of non-genes.…”

“…Monod and Jacob (1961) famously claimed that "anything found to be true of E. coli must also be true of Elephants." We test whether amino acids that are good for random peptides in E. coli are enriched in young animal and plant proteins from a recent phylostratigraphy dataset (James, et al 2020) of over 400 different eukaryotic species. We also test whether published, sequence-based estimators of disorder, aggregation propensity, water solubility, hydrophobic residue clustering etc.…”