Aggregation gatekeepers modulate protein homeostasis of aggregating sequences and affect bacterial fitness

The most common mechanism by which proteins aggre-gate consists in the assembly of short hydrophobicprimary sequence segments into extended b-structuredagglomerates. A significant enrichment of charged resi-dues is observed at the flank of these aggregation-pronesequence segments, suggesting selective pressure againstaggregation. These so-called aggregation gatekeepers actby increasing the intrinsic solubility of aggregatingsequences in vitro, but it has been suggested that theycould also facilitate chaperone interactions. Here, weaddress whether aggregation gatekeepers affect bacterialfitness. In Escherichia coli MC4100 we overexpressedGFP fusions with an aggregation-prone segment of s32(further termed s32b) flanked by gatekeeper and non-gatekeeper residues and measured pairwise competitivegrowth. We found that the identity of flanking residueshad significant effect on bacterial growth. Overexpressionof s32b flanked by its natural gatekeepers displayed thegreatest competitive fitness, followed by other combina-tions of gatekeepers, while absence of gatekeepers strong-ly affects bacterial fitness. Further analysis showed thediversity of effects of gatekeepers on the proteostasis ofs32b including synthesis and degradation rates, in vivoaggregation propensity and chaperone response. Ourresults suggest that gatekeeper residues affect bacterialfitness not only by modulating the intrinsic aggregationpropensity of proteins but also by the manner in whichthey affect the processing of s32b–GFP by the proteinquality control machinery of the cell. In view of theseobservations, we hypothesize that variation at gatekeeperpositions offers a flexible selective strategy to modulatethe proteostatic regulation of proteins to the match intrin-sic aggregation propensities of proteins with required ex-pression levels.Keywords: amyloidosis/gatekeeper/protein aggregation/protein expression/proteostasisIntroductionOne of the most common mechanisms whereby proteins ag-gregate consists in the b-stand association of shortaggregation-nucleating sequence segments into insolubleaggregates. Protein misfolding and aggregation generallyleads to the suppression of native protein function, and inmany cases it also appears to elicit wild-type independentgain-of-function leading to neurodegenerative diseases suchas Alzheimer’s disease or Parkinson’s disease, systemic amy-loidoses, cancer and metabolic diseases (Chiti and Dobson,2006). The fact that most proteins tend to aggregate, evenunder native conditions, to form amorphous b-aggregates orcross-b amyloid structures shows that b-aggregation is ageneric structural propensity of protein polypeptide chainsthat is in direct competition with native protein folding(Dobson, 2001; Rousseau et al., 2006a; Fitzpatrick et al.,2011). Indeed, it has been demonstrated that the vast majorityof proteins possess at least one and often multipleaggregation-nucleating amino acid segments (Rousseau et al.,2006c; Goldschmidt et al., 2010). Moreover, aggregation pro-pensity and globular structure of proteins are intimately tiedas aggregation-prone segments and generally contribute to thehydrophobic core (Linding et al., 2004;De Baets et al., 2011).As a result, although protein folding is a thermodynamicallydetermined process (Anfinsen and Scheraga, 1975), foldingefficiency is often challenged by non-native misfolding andaggregation propensities (De Simone et al., 2011).Given the undissociatable ties of protein aggregationwith globular structure and function on the one hand, and onthe other hand the detrimental consequences of proteinaggregation on the biochemistry of the cell, organisms haveevolved control mechanisms to keep aggregation in check(Monsellier and Chiti, 2007; Monsellier et al., 2007;Reumers et al., 2009a; Reumers et al., 2009b; de Groot andVentura, 2010). One of these strategies includes a sophisti-cated protein quality control system that consists of molecu-lar chaperones and proteases (Mogk et al., 1999; Guisbertet al., 2004; Balch ., 2008). These proteins prevent un-wanted interactions between unfolded proteins, improve theefficiency of de novo protein folding of newly synthesizedproteins, remove misfolded proteins by degradation and reso-lubilize protein aggregates for subsequent refolding or deg-radation (Muchowski and Wacker, 2005; Soti et al., 2005).Apart from having developed dedicated protein qualitycontrol machinery, selective pressure has also shaped proteinsequences in a way to minimize aggregation (Rousseauet al., 2006c; Monsellier and Chiti, 2007; de Groot andVentura, 2010). Whereas aggregation nucleating sequencesegments can generally not be eliminated altogether withoutaffecting the hydrophobic core of globular domains (Lindinget al., 2004), their aggregation propensity can be minimizedby placing charged residues like arginine (R), lysine (K),