Summary
A Nature study tested nearly all single amino-acid mutations in RSV’s fusion protein and measured their effects on two antibodies used to prevent severe disease in infants. Resistance-associated sequence patterns were rare in current surveillance data, while the results provide a framework for tracking viral escape.
A study published in Nature has mapped how nearly all single amino-acid mutations in the respiratory syncytial virus (RSV) fusion protein affect viral cell entry and neutralization by antibodies used to protect infants. The work identified additional mutations that can reduce the activity of nirsevimab and clesrovimab, and found that RSV sequences with high predicted escape scores accounted for less than 1% of sequences in the surveillance data analysed.
The researchers combined a biophysical model of antibody binding with pseudovirus deep mutational scanning. Their results also explain why nirsevimab resistance has been observed more often in RSV subtype B than in subtype A, despite the antibody recognizing a similar region of the viral protein in both subtypes.
How the study measured antibody escape
RSV’s fusion protein, known as F, helps the virus enter cells. Nirsevimab and clesrovimab bind different regions of the prefusion form of this protein and are licensed options for preventing severe RSV disease in infants.
The researchers created two independent libraries of lentiviral pseudoviruses displaying RSV F from the laboratory-adapted subtype A Long strain. The libraries covered 99% and 99.4% of all possible amino-acid mutations in the protein’s ectodomain. The pseudoviruses could complete only a single round of cell entry and did not encode the other viral proteins required for a fully replicative RSV infection.
The team first measured how each mutation affected F-mediated entry into 293T-TIM1 cells. It then tested mutations that retained sufficient entry function for their effects on neutralization by the IgG and Fab forms of nirsevimab, clesrovimab and several other antibodies. IgG has two antigen-binding arms, whereas a Fab fragment has one.
That distinction helped explain the subtype difference. Nirsevimab binds F from subtype A with higher affinity than F from subtype B. When affinity is high enough, bivalent IgG binding can buffer the effect of a mutation that weakens one antibody-binding interaction. The same mutation can therefore have little effect on IgG neutralization of subtype A while reducing neutralization of subtype B. Fab measurements lack this two-arm buffering and were consequently more informative across the two genetic backgrounds.
Two known nirsevimab escape mutations, K68Q and (K/N)201S, reduced Fab neutralization against both subtypes but reduced IgG neutralization only against subtype B, matching the model’s prediction. The study also identified previously unreported mutations near the nirsevimab binding site, including changes at positions 67, 73, 206–207, 210–211 and 215–216.
What the mutation map found
The F protein was not equally tolerant of change across its structure. Regions such as the fusion peptide were highly constrained because many mutations impaired cell entry. The nirsevimab binding site was more tolerant of mutation than the lateral-face region targeted by clesrovimab. This creates more opportunities for mutations that preserve entry while weakening antibody binding in the nirsevimab epitope.
For clesrovimab, the strongest effects were concentrated around residues 426–470. The experiments confirmed previously reported changes and identified additional mutations at positions 426, 429, 433 and 470. Mutations that moderately reduced clesrovimab neutralization generally had a larger effect on Fab than IgG, consistent with buffering by high-affinity bivalent binding.
The researchers used the measurements to calculate escape scores for natural RSV sequences and integrated them into updated Nextstrain phylogenetic trees. Sequences with high predicted nirsevimab or clesrovimab escape scores were rare, represented by less than 1% of the sequences examined, and were scattered across the trees rather than showing sustained spread. Pseudoviruses carrying F proteins from nearly all of the high-scoring natural strains showed reduced neutralization compared with control strains.
The map also showed how suptavumab, an earlier antibody that failed a phase 3 trial against RSV subtype B, was vulnerable to mutations around positions 172 and 173. In particular, S173L reduced suptavumab neutralization while having little effect on F-mediated cell entry. The finding illustrates how antibody resistance can become consequential when a mutation preserves the viral protein’s essential function.
Implications for surveillance and antibody design
The study provides an experimental reference for interpreting newly sequenced RSV F proteins. Instead of treating every sequence difference as equally important, surveillance systems can focus on mutations experimentally associated with reduced antibody neutralization and flag combinations with high predicted escape scores.
The results also support development of antibodies with distinct or complementary escape profiles. Candidate antibodies RSM01, 1A2 and 1B6 were affected by mutation patterns that differed from one another and, in several cases, from those affecting nirsevimab or clesrovimab. Such differences could help researchers assess antibody combinations and design molecules that remain effective against a wider range of viral variants.
The measurements primarily describe single mutations in the F background of one subtype A laboratory strain. The researchers validated selected results in subtype B pseudoviruses, but the effects of mutations can depend on the surrounding sequence. The escape scores also assume that individual mutation effects combine additively; one subtype A strain with a high nirsevimab score did not show reduced neutralization, possibly because of an interaction between S211R and a nearby R213S mutation. Continued genomic and phenotypic surveillance is therefore needed to determine whether rare resistant strains spread in human populations.