Introduction

The causative agent of Haemorrhagic fever with renal syndrome (HFRS) is Hantaan virus, which belongs to the family Hantaviridae and is primarily transmitted by rodents (Zheng et al., 2019). China is the most heavily affected country by HFRS, accounting for >90% of global cases, and a total of 32,462 cases of HFRS were reported in China during 2019–2022, with a fatality rate ranging from 0.1 to 15% (R. Li et al., 2023) (https://www.chinacdc.cn).

Immunoglobulin G (IgG) is one of the best-understood glycoproteins and each constant heavy 2 domain (CH2) of the fragment crystallisable (Fc) region carries a covalently attached biantennary N-glycan at the highly conserved asparagine at residue 297 (Trastoy et al., 2018). Functionally, minor alterations in the glycan composition can drastically affect the structural conformation of the Fc region with enormous IgG effector functions (Abu-Raya, Michalski, Sadarangani, & Lavoie, 2020). IgG modulates both pro- and anti-inflammatory activities depending on the moieties of the glycans. For example, sialylation of IgG-Fc reverses a pro-inflammatory process into an anti-inflammatory effect (Pereira et al., 2020). It has been evidenced that progression of dengue virus and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection is regulated by specific IgG glycosylation (Hou et al., 2021; Wang et al., 2017). The human immune response plays a fundamental role in anti-viral infections. Massive and rapidly expanding plasmablast (PB) cells are required for the recovery of HFRS patients (García et al., 2017; Y. Li et al., 2020), however, the increased amount of circulating CD27^-IgD^- B cells may partially associate with kidney dysfunction (Kerkman et al., 2021). However, there is limited report regarding the regulation of IgG glycosylation in HFRS.

In the present study, we combined scRNA-seq and flow cytometry to reveal the phenotypes of B cell responses during Hantaan virus (HTNV) infection, and explored the transcriptomic features of reactive B cell subsets. A total of 166 HFRS patients, including 65 paired HFRS samples, were involved to profile the IgG-Fc glycosylation and investigate the potential regulatory role of IgG-Fc glycosylation in HFRS pathogenesis.

Results

B cell compositional characteristics in HFRS patients

To characterize the humoral immune profiles in HFRS patients, we enrolled suspected 166 patients who were admitted to Baoji Central Hospital in Shaanxi Province, China, between October 2019 and January 2022, of whom 65 met the inclusion criteria and were included in the study (Figure 1). We identified a total of eight cell subpopulations, including subsets of CD4 T cells, CD8 T cells, CD14 monocytes, NK cells, B cells, platelet, endothelial progenitor cells, and red blood cells (Figure 2—figure supplement 1A and B). The proportions of these cell types were comparable between acute HFRS patients and healthy controls, with no significant batch effects observed (Figure 2—figure supplement 1C). CD4 T cells (31.4%) were the predominant subset prior to HTNV infection, whereas CD8 T cells (44%) became the dominant population post-infection. Notably, the proportions of NK cells (15.9% to 5.6%, p<0.001) and platelets (3.1% to 0.6%, p<0.01) significantly decreased, while the level of CD14 monocytes (15.9% to 32.3%, p<0.05) nearly doubled. However, the overall proportions of B cells did not exhibit significant changes (Figure 2—figure supplement 1D and E).

Figure 1.

To precisely elucidate the role of B cell subsets in anti-HTNV infection, we classified B cells into eight subsets, including antibody-secreting memory B cells (ASM), double-negative B cells (DN), intermediate memory B cells (IM), marginal zone-like B cells (MZB), naïve B cells (naive B), PB, quiescent resting memory B cells (RM), and exhausted tissue-like memory B cells (TLM) (Figure 2A and B, and Figure 2—figure supplement 1F). We discovered that significant expansion of ASM, PB, and RM cell populations in acute HFRS patients compared to healthy controls, accompanied by contraction of DN, IM, and naïve B cell compartments post-infection (Figure 2C). In healthy controls, naïve B cells (42.87%), DN B cells (26.65%), and TLM B cells (16.13%) constituted the majority of the B cell population. However, ASM (30.96%), naïve B cells (23.85%), and PB cells (11.12%) became the dominant subsets after HTNV infection (Figure 2D and Supplementary file 1).

Figure 2.

To validate the accuracy of our single-cell subpopulation analysis, we employed flow cytometry to quantify the B cell subsets in acute and convalescent HFRS patients (Figure 2E). The results demonstrated that ASM (39.48% vs. 24.57%, p<0.001) and PB (23.97% vs. 2.61%, p <0.001) subsets were significantly amplified in the acute phase compared to the convalescent phase (Figure 2F). Conversely, the number of naive B cells (32.78% vs. 48.92%, p=0.003) and CSM cells (18.65% vs. 27.15%, p=0.008) rapidly recovered in the convalescent stage (Figure 2F), consistent with the observations at the transcriptome level. No significant differences were observed for other B cell subtypes between the acute and convalescent stages (Figure 2—figure supplement 2). These results clearly exhibited that several B cell subpopulations were effectively activated following HTNV infection.

Dynamic characteristics of the IgG glycome in HFRS patients

HFRS patients were further stratified into three age groups (≤44, 45-59, and ≥60 years). IgG glycosylation was analyzed by ultra-performance liquid chromatography. As shown in Supplementary file 2, no significant differences in IgG glycosylation were observed between age groups, except for galactosylation. Similarly, no gender-based differences in IgG-Fc glycosylation were noted (Supplementary file 3). However, the levels of IgG with bisecting GlcNAc (15.72% vs. 14.51%, p<0.001), galactosylated IgG (74.26% vs. 71.15%, p<0.001), and sialylated IgG (22.44% vs. 21.62%, p<0.001) were significantly higher in the acute phase compared to the convalescent phase. Conversely, fucosylated IgG levels (94.29% vs. 94.89%, p<0.001) were lower in the acute phase (Supplementary file 4), indicating that IgG N-glycosylation plays a critical role in the anti-infection process.

Among the 130 paired blood samples from 65 patients, 38 (29.2%) were negative for HTNV nucleocapsid protein-specific IgG antibodies, while 92 (70.8%) were seropositive. For the 24 patients with both seronegative and seropositive samples, levels of bisecting GlcNAc (16.16% vs. 14.54%, p<0.001), galactosylation (74.12% vs. 70.94%, p=0.006), and sialylation (22.92% vs. 21.80%, p=0.008) were higher in the seronegative period, while that of fucosylation was lower (94.36% vs. 94.74%, p= 0.024) in this period (Figure 3A). Notably, bisecting GlcNAc and sialylated IgG levels decreased as HTNV-NP-specific antibody titers increased (Figure 3B), highlighting a potential link between antibody levels and IgG-Fc glycosylation.

Figure 3.

The IgG glycome in HFRS patients with seroconversion

A 4-fold or greater increase in HTNV-NP-specific antibody titers usually indicates a protective humoral immune response during the acute phase. We observed a significant increase in the Fc fucosylation (94.18% vs. 94.48%, p=0.044) and a decrease in bisecting GlcNAc (16.41% vs. 14.71%, p=0.001) post the 4-fold antibody titer increase compared to the baseline. However, no significant differences were observed in galactosylation and sialylation (Figure 3C), suggesting that fucosylation and bisecting GlcNAc may play key roles in anti-HTNV infection and recovery.

The potential source of the IgG glycome during HTNV infection

Positive correlations were between the ASM subsets and galactosylation/sialylation in the antibody Fc region, and between the PB subsets and sialylation (Figure 4A-C). Therefore, we speculated that galactosylated antibodies may primarily originate from the ASM subsets, while sialylated antibodies may derive from the ASM and PB subsets.

Figure 4.

To investigate the process of glycosylated antibody production secreted by the ASM and PB subsets, we identified DEGs among the eight B cell subpopulations. Notably, the most DEGs were observed in the ASM (470), naive (312), and TLM (333) B subpopulations, respectively (Figure 4—figure supplement 1A and Supplementary file 5). Specifically, among all the upregulated genes, the expressions of glycosylation-related, mitochondrial respiratory-related, and peptide chain elongation-related genes among the ASM, PB, and RM populations were drastically elevated after HTNV infection, including RPN2, MT-ND6, and RPS26 (Figure 4—figure supplement 1B and Supplementary file 6). On the contrary, the upregulated genes in the naïve, DN, and IM populations were related to anti-inflammatory and immunosuppressive effects, inhibition of the antioxidative function, response to environmental stress, and B-cell growth and development, such as TSC22D3, TXNIP, DUSP1, and KLF6, whereas the downregulated genes played an important role in the inflammatory response, cell motility, and antigen presentation, including LTB, ACTG1, and HLA-DRB5 (Figure 4—figure supplement 1C and Supplementary file 7).

Unexpectedly, we observed significant changes in the expression of glycosylation-related genes (GRGs) in the ASM, PB, and RM subclusters during HTNV infection compared to healthy controls. PRN1 and PRN2 were prominently upregulated, which encoded subunits of the oligosaccharyltransferase complex (Figure 4D). This implied that the process of glycosylation during HTNV infection was rapidly activated among the ASM and PB subclusters. In the ASM subcluster, the fold change of GRGs from all four glycosylation modification-related genes, aside from PRN1 and PRN2, was similar, suggesting that ASM may undergo complex and diverse glycosylation modification processes.

To enhance our understanding of the transcriptomic characteristics of GRG-high-expressing PB and ASM subclusters, we conducted a comprehensive pathway enrichment analysis and identified significant enrichment of pathways related to glycosylation modifications, including the endoplasmic reticulum protein-containing complex, rough endoplasmic reticulum, and protein N-linked glycosylation via asparagine (Figure 4E). Additionally, gene sets associated with glycosylation modifications showed positive regulation of these pathways (p=0.002, Figure 4F). These results indicated that glycosylation modifications in the ASM, PB, and RM subpopulations were dynamically regulated during the acute phase of HFRS, likely to satisfy the amplified anti-HTNV antibody requirement and accelerate viral clearance.

Discussion

IgG glycosylation is a key regulator of inflammatory responses for infectious diseases, acting as a molecular switch between anti-inflammatory and pro-inflammatory effects upon antigenic challenge. One of the most significant findings of this study is the elevation in IgG fucosylation and reduction in bisecting GlcNAc levels at the time of a 4-fold increase in the HTNV-NP-specific antibody titers. Afucosylation is known to enhance antibody-dependent cellular cytotoxicity (ADCC) by increasing the affinity of IgG-Fc for the FcγRIIIa receptor on natural killer cells, leading to heightened production of inflammatory cytokines such as IL-1β, IL-6, IFN-γ, and TNF-α (Larsen et al., 2021; Lu, Suscovich, Fortune, & Alter, 2018; Wieczorek, Braicu, Oliveira-Ferrer, Sehouli, & Blanchard, 2020). Increase in bisecting GlcNAc linked to Fc induces enhanced affinity for the FcγIIIa receptor and enhanced ADCC function of the antibodies (Giron et al., 2020; Hou et al., 2021; Irvine & Alter, 2020). However, rather than directly influencing the Fc receptor binding, this modification enhances antibody functionality through afucosylation, which subsequently increases the binding affinity to the FcgRIIIa receptor (Ackerman et al., 2018; Jennewein & Alter, 2017). Our findings provide compelling evidence that the dynamic alterations in glycosylation patterns, characterized by increased fucosylation and decreased bisecting GlcNAc levels during viral infection, are critically involved in disease pathogenesis and clinical recovery, suggesting their dual utility as both diagnostic biomarkers and promising therapeutic targets.

We also observed elevated galactosylation levels in the acute phase compared to the convalescent phase, a phenomenon also reported in SARS-CoV-2 infections but not in chronic diseases and cancers (Hou et al., 2021; D. Liu et al., 2018). Increased galactosylation may expand the distance between CH2 domains, exposing key amino acids for FcγRIIIa binding and enhancing ADCC functions (Houde, Peng, Berkowitz, & Engen, 2010). Simultaneously, the increase in the ASM subset, which positively correlated with galactosylation, may reflect polyclonal activation of memory B cells driven by pattern recognition receptors and cytokines, leading to the production of galactosylated antibodies. Similarly, sialylation levels were higher in the acute patients, potentially linked to the expansion of the ASM and PB subsets. In this study, the PB subset was significantly elevated in the acute phase and positively correlated with sialylation, suggesting that rapidly proliferating plasmablasts may contribute to the secretion of sialylated antibodies. However, the mechanisms behind this require further investigation.

Our findings demonstrate that the IgG-Fc glycosylation plays a pivotal role in HTNV infection and rapid recovery. Fucosylated oligosaccharides and the absence of bisecting GlcNAc are associated with favorable clinical outcomes. Importantly, specific B cell subsets, particularly ASMs and PBs, upregulated the secretion of galactosylated and sialylated antibodies, respectively. These findings would deepen our understanding of antibody glycosylation in anti-virus infection and provide a foundation for future studies aimed at optimizing glycoengineered therapeutic antibodies.

Materials and Methods

Study participants

Clinical specimens were collected from HFRS patients who were hospitalized in Baoji Central Hospital between October 2019 and January 2022. The HFRS patients were categorized into four clinical subtypes and clinical courses based on the previous reports (Ma et al., 2015; Tang et al., 2019; Zhang et al., 2022). This study was approved by the ethics committee of the Shandong First Medical University & Shandong Academy of Medical Sciences (R201937). Written informed consent was obtained from each participant or their guardians.

Single-Cell RNA Sequencing and public database

Flow cytometry antibody staining was conducted as previously described (Chakraborty et al., 2022). Dead cell and red blood cell were removed using the BD FACSAria™ III cell sorter, and PBMCs were collected for single-cell sequencing. The library construction was conducted as previously described (Jin et al., 2024).

We obtained single-cell transcriptome sequencing data from the Gene Expression Omnibus (GEO) database (GSE161354), containing six HFRS patients and two healthy volunteers, as well as healthy control samples from the Genome Sequence Archive (GSA) database (HRA000203), matched with patient demographic information.

Single-Cell Data Analysis

The FastQC software was used to evaluate the data obtained to ensure the quality of the raw sequencing data. The raw data were mapped to the human reference genome (GRCh38, https://cf.10xgenomics.com/supp/cell-exp/refdata-gex-GRCh38-2020-A.tar.gz) using Cell Ranger. The Seurat R package (Version 4.4.0) was utilized for the merging and clustering of single-cell data. The Harmony R package and the anchor module of Seurat were used to remove batch effects between samples and groups for cell clustering. t-distributed Stochastic Neighbor Embedding (t-SNE) and Uniform Manifold Approximation and Projection (UMAP) were employed for dimensionality reduction and visualization of individual cells.

Cell Type Annotation

UMAP and t-SNE were used to reduce the dimension of all cells and to cluster them in a two-dimensional space based on shared features. The Single R (Version 2.2.0) package was used to independently infer the cell origin of each single cell based on a reference transcriptome dataset of pure cell types, for unbiased cell type identification from single-cell RNA sequencing data. Subsequently, specific highly expressed genes were combined with literature review for manual annotation to ultimately determine cell types. Classic biomarkers for specific cell types were used to identify cells in different clusters.

Differential Gene Identification and Functional Analysis

The FindMarkers() function in the Seurat package was used to identify differentially expressed genes (DEGs) between different cell groups, with criteria of |log2FC| > 0.25 and p-value<0.05, pct.1>0.25. The ClusterProfiler R package was utilized for GO, KEGG, and GSEA enrichment analyses, and the ggplot2 and enrichplot R packages were used for visualization.

Measurement of HTNV specific antibodies

Levels of serum IgM and IgG antibodies to HTNV nucleocapsid protein (NP) were assessed through the enzyme-linked immunosorbent assay (ELISA) kit (Wantai BioPharm, Beijing, China) in accordance with the manufacturer’s instructions.

Analysis of IgG glycans

The process of isolating, labelling, purifying, and analyzing plasma IgG was performed using previously established methods (Hou et al., 2019; P. Liu et al., 2023).

Virus-specific B-cell detection

Virus-specific B-cell against HTNV were performed as previously described (Liechti & Roederer, 2019; Song et al., 2018).

Statistical analysis

Categorical variables were compared using chi-square tests. Normality of numeric variables was assessed with the Kolmogorov-Smirnov test; non-normally distributed continuous variables were analyzed using non-parametric tests. Statistical analyses were performed with R software version 4.1.1 (R Core Team, New Zealand) and GraphPad Prism 8 (GraphPad Software, San Diego, CA), with a significance level of α = 0.05 (two-tailed).

Supplementary materials

Figure 2—figure supplement 1.

Figure 2—figure supplement 2.

Figure 4—figure supplement 1.

Additional information

Funding

Data availability

The data underlying this article will be shared upon request to the corresponding author. The following dataset was generated:

Zhang H, Wang Y (2022) Increased CD4⁺CD8⁺ double positive T cells during Hantaan virus infection ID GSE161354. In the public domain at GEO (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE161354).

Zheng Y, Liu X (2020) A human circulating immune cell landscape in aging and COVID-19 ID HRA000203. In the public domain at GSA (https://ngdc.cncb.ac.cn/gsa-human/browse/HRA000203).

Author contributions

Haifeng Hou and Weifeng Shi designed and supervised this study. Chuansong Quan, Lu Wang, Jiming Gao, Yaoni Li, Houqiang Li, Hongzhi Liu, Qian Li, and Liqiong Zhao collected the samples and performed experiments. Chuansong Quan, Lu Wang, Jiming Gao, Xiaoyu Xu, Zixuan Gao, Wenxu Ruan, and Weijia Xing analyzed the data and performed the statistical analyses. Lu Wang, Chuansong Quan, Jiming Gao, and Xiaoyu Xu wrote the manuscript. Haifeng Hou, Weifeng Shi, Weijia Xing, and Michael J. Carr reviewed and edited the manuscript. All authors had full access to all of the data in the study and accept responsibility for the decision to submit for publication.

Additional files

Supplementary file 1. The proportion of different B cell subpopulations between healthy controls and acute HFRS patients.

Supplementary file 2. Difference in the main IgG glycome features of HFRS patients between age groups.

Supplementary file 3. Difference in the main IgG glycome features of HFRS patients between sex groups.

Supplementary file 4. Differentially expressed genes of different B cell subgroups between healthy controls and acute HFRS patients.

Supplementary file 5. Differentially expressed genes of different B cell subgroups a between health and acute HFRS groups

Supplementary file 6. Differential expression genes in the ASM, PB and RM B cell subpopulations.

Supplementary file 7. Differential expression genes in the DN, IM and naive B cell subpopulations.