Amino acid homorepeats act as buffers to maintain proteostasis and constrain the compatible sequence space of proteomes

Proteins are composed of combinations of 20 amino acids, yet the principles governing their sequence compatibility within living cells remain poorly understood. Here, we evaluated the “neutrality”—an integrated measure of harmful versus beneficial properties—of amino acid homorepeats (PolyX) in Saccharomyces cerevisiae . Using the genetic tug-of-war (gTOW) method to quantify maximum expression limits, we found that neutrality is highly amino-acid-specific and intrinsically linked to physicochemical properties. Hydrophobic repeats (I, V, W, F, Y) were universally harmful and aggregation-prone, while hydrophilic repeats (E, S, N, Q) not only exhibited high neutrality but also actively suppressed the proteotoxic stress and aggregation of host proteins. These experimental neutrality values show a striking correlation with the occurrence and length of homorepeats across diverse eukaryotic proteomes. Our findings suggest that PolyX neutrality defines the “compatible sequence space” of proteins, where certain sequences like Poly ₁₀ E act as evolutionary buffers to maintain proteostasis. This study provides a quantitative framework for understanding how physical constraints on protein solubility shape the evolution of proteomes and offers a new metric for rational protein design.