A Shift in Aggregation Avoidance Strategy Marks a Long-Term Direction to Protein Evolution

Abstract

To detect a direction to evolution, without the pitfalls of reconstructing ancestral states, we need to compare "more evolved" to "less evolved" entities. But because all extant species have the same common ancestor, none are chronologically more evolved than any other. However, different gene families were born at different times, allowing us to compare young protein-coding genes to those that are older and hence have been evolving for longer. To be retained during evolution, a protein must not only have a function, but must also avoid toxic dysfunction such as protein aggregation. There is conflict between the two requirements: hydrophobic amino acids form the cores of protein folds, but also promote aggregation. Young genes avoid strongly hydrophobic amino acids, which is presumably the simplest solution to the aggregation problem. Here we show that young genes' few hydrophobic residues are clustered near one another along the primary sequence, presumably to assist folding. The higher aggregation risk created by the higher hydrophobicity of older genes is counteracted by more subtle effects in the ordering of the amino acids, including a reduction in the clustering of hydrophobic residues until they eventually become more interspersed than if distributed randomly. This interspersion has previously been reported to be a general property of proteins, but here we find that it is restricted to old genes. Quantitatively, the index of dispersion delineates a gradual trend, i.e., a decrease in the clustering of hydrophobic amino acids over billions of years.

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

“…We confirm that across our entire dataset, ISD is higher in young pfams (figure 1, linear model: R 2 = 0.13, p = 3×10 -245 ) and genes (Supplementary figure 1A, linear model for all genes: R 2 = 0.057, p =1×10 -229 ). Our improved methodology increased the steepness of the relationship between gene ISD and gene age in mouse genes above that previously estimated by Foy et al (2019) (Supplementary Fig. 1B), from -0.028 to -0.050; note that mouse genes have a steeper slope than our results across all taxa.…”

“…In Supplementary fig. 8B, we show that our improved methods for assigning gene age result in a steeper slope for mouse genes than that reported by Foy et al (2019): -0.056 instead of -0.045.…”

“…S1) ISD depends on age for whole genes, as previously reported by Foy et al (2019). Each data point consists of the average across all instances of a homologous gene family (see Methods) across all species (A) or just in mouse (B), dated according the oldest pfam the gene contains.…”

“…If the length L of a protein was not a multiple of 6, we took the average of all possible p = (L modulo 6) frames, truncating the ends appropriately. The six most hydrophobic amino acids Leu, Ile, Val, Phe, Met and Trp were scored as +1, and the remaining standard amino acids to -1, as in Irback et al (1996) and Foy et al (2019). Non-standard amino acid residue abbreviations were also scored as +1 or -1, interpreted as follows: B as corresponding to D or N and hence -1; J as corresponding to either I or L and hence +1; Z as corresponding to either E or Q and hence -1; U as selenocysteine and O as pyrrolysine, both scored as -1.…”

“…It has been claimed that younger genes encode shorter and faster evolving proteins (Lipman et al 2002;Toll-Riera et al 2012) that are more disordered, containing fewer hydrophobic amino acids (Wilson et al 2017), which are more clustered together along the primary sequence ). Foy et al (2019) suggest that the two latter trends, in hydrophobicity and in its clustering, together indicate that evolution is slow to find the sophisticated folding strategies employed by older proteins, which minimize their propensity to form aggregates within cells despite relatively high overall hydrophobicity.…”