Identification of biochemically neutral positions in liver pyruvate kinase

Abstract

In the goal of interpreting human exomes, when predictive programs assign functional importance to some positions, they implicitly assume the existence of non-important positions -those that accommodate many side chain chemistries without altering function ("neutral"). However, very few (if any) experimental studies have demonstrated the existence of neutral positions. We sought experimental evidence for neutral positions using human liver pyruvate kinase (hLPYK) as a model system. To that end, we used multiple evolutionary criteria to identify 20 possibly neutral positions. Nine positions were further tested with a total of 117 amino acid substitutions. Although "all" potential hLPYK functions can never be explored, we measured effects on 5 parameters associated with substrate/ligand affinities and allosteric coupling. At each position, the aggregate outcomes of multiple variants were used to quantify "neutrality" scores. Three of the nine positions showed perfect neutral scores in all 5 parameters; a fourth position had high neutral scores. Although our strategy for predicting positions had low predictive power for the identification of neutral positions, all positions had neutral scores that were much higher than positions in functional sites. Given this evidence for the existence of neutral positions, similar studies should be carried out for other proteins to generate a database of well characterized neutral positions that can then be available to benchmark and validate predictions about amino acid substitutions.

Functional tunability from a distance: Rheostat positions influence allosteric coupling between two distant binding sites

“…Empirically, for datasets the size of the hLPYK set, we have found that rheostat scores above 0.5 indicate strong rheostat character of the position 3 . In contrast to rheostat scores, neutral scores must be above 0.7 to be considered a high score 12 . For toggle scores, we previously defined a strong toggle threshold of "at least 8 severe variants" 14 The results from this study confirm that substitutions distant to the two allostericallycoupled binding sites can exert full tunability over allosteric coupling (e.g., full bin occupancies in Figure 4 and rheostat scores of 1 in Figure 5).…”

“…To identify the appropriate width and number of histogram bins, the RheoScale calculator assessed the full experimental dataset for (i)minimal and maximal values for the parameter being assessed, (ii) average error of experimental measurements, and (iii) average number of variants available per position. Since completing our first study of "local" hLPYK rheostat positions 3 , we have obtained and collated a much larger dataset of hLPYK parameters and used them to refine RheoScale histogram analyses 12,13 . These include measurements for additional variants and multiple, independent measurements of wildtype hLPYK (Supplemental Information).…”

Rheostat functional outcomes occur when substitutions are introduced at nonconserved positions that diverge with speciation

“…[3][4][5][6][7] As such, substitutions at these "rheostat" positions provide ready opportunities for fine-tuning function during evolution and protein engineering. The rheostat substitution behavior is strikingly different from (i) positions for which most substitutions abolish function (a "toggle" substitution outcome; e.g., 2 ) and (ii) positions that can accommodate a range of substitutions without any change in protein function (a "neutral" substitution outcome; e.g., 8 ). Both toggle and neutral positions are often associated with their own evolutionary signatures: conserved positions are expected to exhibit toggle substitution behavior, whereas highly nonconserved positions may be expected to exhibit neutral substitution behavior.…”

“…The lower structures show two views of the structure of a single monomer, with stars approximating the locations of the active (yellow), allosteric inhibitor (green), and allosteric activator (cyan) sites. Magenta spheres identify positions with strong phylogeny scores tested in this study: 107, 177, 192, 259, 107, 423, and 538 thresholds for the rheostat, toggle, and neutral scores that were used to assign an overall substitution behavior to each position 2,4,8 ; those thresholds are indicated in Figure 3 by dashed horizontal lines.…”

“…The other three phylogenetic positions-156, 177, and 259-did not exhibit strong rheostatic substitution behavior in any single parameter, but they also failed to meet the criteria previously defined for neutral positions. 8 To better evaluate the non-neutrality of these positions, we combined all available measurements for each position (all variants, each with up to 5 functional parameters) to generate a "composite neutral" score (Figure 4). Overall, the phylogenetic positions of this study had fewer neutral outcomes (lower composite neutral scores) than those in a prior study that searched for neutral positions.…”