Microenvironmental pH and host-pathogen co-evolution potentially influence the structural architecture of junctional...

Understanding how proteins dynamically adapt to diverse and changing physiological microenvironments is a fundamental challenge in modern biological sciences. Junctional adhesion molecules (JAMs) are a family of conserved proteins critically involved in immune regulation and cell adhesion. In this study, we investigate the evolutionary and structural dynamics of three paralogs across 274 mammalian taxa, which share similar tertiary structures but differ in isoelectric points (pI). By integrating phylogenetic modeling, partial correlation, network topology, and evolutionary molecular dynamics in physiological pH (6.5-10.5) gradient, we explored potential explanations driving this diversification. Our analysis identified JAM-B as a likely central node in the conservation network, with Lys and Cys residues as central evolutionary residues. Evolutionary mapping revealed recent episodic selection bursts across 17% to 26% of mammalian lineages, could indicate that specific functional interfaces are undergoing rapid, lineage-specific innovation. Notably, we identified episodic hotspots in JAM-A at the distal D1 viral entry interface, consistent with an ongoing host-pathogen arms race, and parallel adaptive clusters at the C-terminal motifs across all paralogs. AlphaMissense profiling revealed that acidic-> basic mutations exhibit significantly lower pathogenicity scores. In preliminary early-onset dynamics simulations, root-mean-square-deviation profiles could suggest a pI-stability relationship, JAM-A and JAM-C displayed biphasic pH-dependent deviations (at pH 8.0 and pH 8.5). Dynamics-aware evolutionary profiling identified key dynamic-conserved residues: JAM-A at Gln66, JAM-B at Gln36 and Val57, and JAM-C at several basic residues. Together, these results suggest that isoelectric divergence correlates with residue evolution and microenvironment-specific structural dynamics. Ultimately, our integrated computational framework provides genomic insights into paralog diversification, offering a testable architectural blueprint for targeted mutagenesis or therapeutic modulation of pH-sensitive adhesion processes.