RSV G selection analyses support constraint of the CX3C/cystine-noose core and diversification in mucin-like regions

Zahid Ayomide Nassoro-Ally 

The RSV attachment (G) glycoprotein is a highly variable surface antigen and an important target of humoral immunity, yet it contains a short central conserved region (CCR; RSV-A2 residues 157-198) with a CX3C chemokine-mimic motif and disulfide-bonded cystine noose. A pre-specified hypothesis was tested: FEL-positive diversifying sites are enriched in the CCR and/or the CX3C motif. Using a uniform align to maximum-likelihood phylogeny to site-wise selection pipeline based on FEL, validated against influenza H3N2 haemagglutinin (positive selection significantly enriched in antigenic sites A-E; p = 0.0081), the FEL enrichment hypothesis was not supported: no mappable RSV G FEL-positive sites fell in the CCR or CX3C motif in RSV-A, RSV-B, or a descriptive pooled summary. MEME, used as an episodic-selection companion analysis, identified one CCR site in RSV-A and one in RSV-B, but zero CX3C sites in either subtype; most RSV MEME-positive sites were in the mucin-like C-terminal domain or ON1/unmappable sequence. These site-level calls are screening-level: after Benjamini-Hochberg false-discovery-rate (FDR) correction no RSV G site remained significant at q <= 0.05, whereas the influenza antigenic-site control signal did survive FDR, and a GARD screen found no recombination in either RSV G alignment. Because the FEL site-count test has low power with only 3-6 significant sites, a threshold-free companion analysis, which does not depend on any per-site cutoff, compared the per-codon dN-dS distribution across regions. Region-level distributions support stronger CCR purifying constraint in RSV-A (Kruskal-Wallis p = 0.0275; CCR vs. mucin Mann-Whitney p = 0.0066, rank-biserial = -0.26, small-to-moderate shift) and a concordant but non-significant trend in RSV-B (p = 0.074). Mapping per-codon dN-dS onto the ordered CX3C/cystine-noose core in PDB 5WN9 is consistent with purifying selection over the functional motif. These evolutionary data support, but do not by themselves establish, the rationale for monitoring and targeting the CX3C/cystine-noose core while recognizing sparse episodic-selection signals in the broader CCR.