Proteome-wide Convergence of Amino Acid Composition in Phage–Host Systems

抄録

Background: Understanding the selective pressures that shape amino acid usage is a fundamental challenge in evolutionary biology. We previously proposed a "mutual constraint" model, suggesting that the cytoplasmic protein pool serves as both a product of the genome and a resource for new protein synthesis, creating a feedback loop that stabilizes species-specific proteome profiles.

Methods: To test this hypothesis, we evaluated the compositional alignment between bacteriophages and their hosts using a comprehensive dataset of 117 phage species and 67 bacterial host species. We calculated the species-specific proteome-wide mean amino acid composition for each organism and quantified the differences using the Jensen-Shannon (JS) distance.

Results: Our analysis revealed that specific phage–host pairs exhibit significantly closer compositional proximity (median JS distance = 0.076) compared to random inter-species combinations (median = 0.110). This alignment was strikingly lifestyle-dependent: temperate phages showed a highly concentrated reciprocal convergence, with approximately 34% of pairs occupying the highest proximity bin in a two-dimensional distance map. In contrast, virulent phages displayed a broader and more scattered distribution.

Discussion: Furthermore, computational frame-shift experiments on Escherichia coli K-12 and lambda phage demonstrated that genomic signatures alone cannot dictate amino acid profiles, suggesting that selective pressure acts directly on the amino acid composition to minimize metabolic friction within the host’s resource environment.

Conclusion: These findings demonstrate that the host's nutritional landscape is a primary determinant of viral proteome evolution. Our study supports the existence of universal metabolic constraints governing the evolution of obligate intracellular entities and provides a new quantitative framework for investigating host–parasite co-evolution.