In silico analysis of capsid and rep proteins in CRESS DNA viruses

Abstract

Abstract The circular rep-encoding single-stranded DNA viruses (CRESS DNA viruses) are among the smallest, with 2–6 kb ssDNA genomes that encode for a coat protein (C) and a replication protein (R). To comprehend the complexity and divergence of the C and R proteins, we have created predictive structural models of representative viruses infecting unique hosts from each family using the neural network-based method AlphaFold2. The Cs exhibit substantially more diversity than Rs, and while they retain the fundamental jelly-roll fold, the loops and amino-terminal ends have undergone significant conformational shifts. The Rs, on the other hand, have minimal diversification and involve alterations only to the central linker and C-terminal domains. A phylogenetic analysis of the C and R proteins based on their structures indicates evolutionary variances and components that might have aided adaption to diverse hosts and vectors. Our study also highlights the conservation of structural features involved in the interaction of R with the conserved intergenic region of the genome. The versatility of the central linker domain may be crucial for establishing interactions of R with multiple protein partners, including C.