Fitness-Driven Scaling Laws between mRNA and Protein Levels

The relationship between protein and messenger RNA (mRNA) levels is a key aspect of gene expression. Although a strong mRNA-protein correlation is widely observed, the evolutionary force driving the correlation remains poorly understood, undermining the reliability of using mRNA abundance as a proxy for protein levels. Here, by balancing the cost and benefit of gene expression, we uncover an optimal scaling between mRNA (m) and protein (p) levels that maximizes cell fitness: m∼p^{γ}. Notably, we predict a lower bound of 0.5 for the exponent, and the toxicity effects of protein overexpression further increase it. For the model organism Saccharomyces cerevisiae, the simulated scaling across the genome, incorporating biologically relevant parameters, aligns well with experimental data. Genome-wide data across multiple organisms all exhibit a scaling exponent above the predicted lower bound. Our Letter rationalizes the primary role of transcriptional regulation and also the significance of translational regulation in determining protein levels from an evolutionary perspective.