Human metapneumovirus (HPMV) is a significant pathogen that causes lower respiratory tract infections. Given the weak immunogenicity thereof, and the few relevant studies, the utility of the viral membrane protein G as a vaccine remains controversial. In this study, the G extracellular domain (RMG) of HMPV was expressed either alone or fused with the cholera toxin B subunit (CTB) and “cross-reacting material 197” (CRM197) carrier proteins (giving G-CTB/G and CRM197), to enhance immunogenicity. The non-glycosylated G protein (REG) expressed in Escherichia coli served as a control. SDS-PAGE and anti-His tag Western blotting verified that each protein was successfully expressed and correctly identified. BALB/c mice were immunized with each protein and subjected to challenge with HMPV. The results showed that, although immunization with RMG alone failed to induce potent neutralizing antibodies, it modestly reduced viral loads in the lungs of mice. However, the pathological damage caused by lung inflammation was more aggravated than that of the control challenge group. The level of specific IgG antibody induced by the recombinant G-CTB was significantly higher than that elicited by RMG. Compared to the RMG group, the viral load in the lungs of the G-CTB group tended to be reduced. Also, the damage caused by lung inflammation was significantly alleviated. Our study proves that HMPV G may be a valuable antigen in terms of HMPV vaccine development and offers a promising strategy for modulating the immunogenicity and safety thereof.

It has been demonstrated that infants and young children exhibit immune tolerance as a consequence of immature immune systems, which are characterized by a natural Th2 bias. RSV infection has been reported to result in acute lower respiratory infection (ALRI), while formalin-inactivated vaccination has been observed to exacerbate Th2 responses, consequently leading to enhanced respiratory disease (ERD). Transcriptomic data from three independent cohorts of RSV-infected infants were analyzed (GSE246622 served as the discovery and train set; GSE105450 and GSE188427 were used as validation sets). Immune infiltration analysis revealed immunological characteristics, which were then used to perform unsupervised clustering using feature-related genes. WGCNA was used to identify co-expressed gene modules, while Mfuzz and TCseq were employed to analyze temporal expression patterns. Machine learning models were developed using a refined panel of candidate genes. Severe symptoms of RSV infection exhibited a strong correlation with age, with younger infants demonstrating more intense inflammatory responses from neutrophils, macrophages, mast cells and dendritic cells. A predictive model was constructed using ten co-expressed genes: The following genes were identified: MCEMP1, FCGR1B, ANXA3, FAM20A, CYSTM1, GYG1, ARG1, SLPI, BMX and SMPDL3A. It was observed that infants of a younger demographic demonstrated a heightened degree of immunosuppression and pronounced innate immune activation in patients of severe symptoms with RSV infection. However, eosinophils exhibited minimal involvement in these processes. These gene models pertaining to the neutrophil, macrophage or mast cell was found to be a relatively effective predictor in patients of severe symptoms.