Single-cell RNA sequencing reveals T and B cell-related immune features in foot and mouth disease virus-infected mice

ABSTRACT

T and B lymphocytes protect against the foot and mouth disease virus (FMDV). They are essential for long-term immunity and viral clearance, but their immunological features in FMDV infection remain unknown. Here, we performed single-cell RNA sequencing (scRNA-seq) and conducted integrative analyses to determine cell composition, gene expression, cell-cell communication state, and regulatory network of T and B cells in the spleen of FMDV- and mock-infected mice. Our results revealed variational compositions of immune cells on the subsets level. Additionally, we determined the reshaping of several essential pathways in T and B cells, such as the inhibition of antigen-presentation capacity. Finally, the virus-induced shaping of immune patterns, including cell-cell communication and regulatory networks, was also characterized during FMDV infection. Overall, our research provides valuable insights and resources for understanding the adaptive immune response in FMD.

GRAPHICAL ABSTRACT

Introduction

Foot and mouth disease (FMD) is one of the most challenging epizootics of global significance. It is also a World Organization for Animal Health (WOAH, founded as OIE) listed disease that must be reported. The disease affects cloven-hoofed ruminants such as swine, goat, and cattle and is characterized by blister-like sores in the mouth and between the hooves. Despite the low mortality in adult animals, it is highly fatal in neonates because of myocarditis or, when the dam is infected by FMD, devoid of lactation.

As a member of the Picornaviridae family, the FMDV comprises 7 serotypes (A, O, C, SAT1, SAT2, SAT3, and Asia1) with a single-stranded, positive-sense RNA genome. The genome of a mature virion particle contains a long open reading frame (ORF) encoding a single poly-protein precursor, which is cleaved into four structural proteins (VP1, VP2, VP3, and VP4) and eight non-structural proteins (L, 2A, 2B, 3A, 3B, 2C, 3C, 3D) [1,2].

Up to 2024, the incidence of FMD is still high, especially in Asia, most of Africa, and the Middle East (https://www.woah.org/en/disease/foot-and-mouth-disease/#ui-id-2). A total of 223 outbreaks have been officially reported worldwide over the past 10 years, with A, O, C, SAT1, SAT2, SAT3, and Asia1 accounting for 11.66%, 47.98%, 0.45%, 5.38%, 19.28%, 2.69%, and 1.35% of all events, along with 11.21% attributed to unknown serotypes (https://wahis.woah.org/#/event-management). Moreover, due to the ongoing globalization process, the disease continues to impose significant risks on livestock farming and international trade in animals and animal products.

Generally, regular vaccination is a major part of FMD prevention, and it is an efficient strategy for preventing FMD from being disseminated. However, due to the high mutational rate of the virus, each serotype requires a specific vaccine to confer protection to the vaccinated animals [3–6], which undoubtedly is a significant obstacle to FMD control. Current FMD vaccines are mainly divided into three categories, including inactivated virus vaccines, live attenuated vaccines, and subunit vaccines. As inactivated virus vaccines can prevent clinical infection and the inactivated virus is deprived of the ability to multiply in vaccinated animals, this vaccine dominates the current market. However, they cannot produce durable protection, and booster vaccinations are required every half year to guarantee the immune response [7,8]. Moreover, the short shelf life is another crucial shortcoming of inactivated virus vaccines [9]. In contrast, although several live attenuated vaccines have been proven to induce a strong neutralizing antibodies (Abs) response, they are unacceptable in many areas for the danger of reversion to virulence and accidental release [10]. Furthermore, a promising approach, subunit vaccines possess several advantages over inactivated and live attenuated vaccines. For instance, the production of subunit vaccines is infectious virus-free, which ensures safety. In addition, the chemically synthesized proteins and peptides are much more stable than inactivated and live attenuated vaccines, making them easier to preserve and transport. Nevertheless, several vaccines have been proven to be poorly protective in cattle [11,12], and the immunogenicity deficiency of cellular immune response has blocked its further application [13]. Thus, using proteins or peptides for FMD vaccine development still needs further research and evaluation. As a novel and promising approach for FMD control, adenoviral vector vaccines use non-replicating or replication-deficient adenoviruses – typically human or bovine adenovirus serotypes – as vectors to deliver FMDV antigens, such as the P1-2A and 3C protease genes, directly into host cells. This strategy enables both humoral and cellular immune responses without the risks associated with live virus handling [14–16]. Despite their potential, adenoviral vector vaccines face challenges related to vector immunity, production scalability, and regulatory approval. Nevertheless, ongoing research continues to optimize antigen design, expression, and delivery systems to overcome these barriers. Importantly, several FMDV vaccines – including inactivated, subunit, and vector-based types – have demonstrated the ability to elicit neutralizing antibodies that protect against both homologous and heterologous viral challenges. This underscores the central role of adaptive immunity – particularly neutralizing antibodies – in protection against FMDV infection.

Unfortunately, FMD control is further complicated under clinical conditions due to a subclinical persistent infection [17,18]. The infected animals do not develop symptoms but remain persistent virus carriers in clinical, aggravating the FMD control. Thus, there is a pressing need to understand how T and B cells impede the pathogenesis of FMDV in a latent infection state. The choice of the animal model is also vital to achieve the goal. Compared to large animal models, the mouse model is more stable and exhibits more excellent reproducibility. Several studies are using the mouse model in FMD research [19–23]. However, these studies either used mice with acquired immunodeficiency or employed virus strains adapted to mice through serial passaging. In the acquired deficiency model, it is hard to study the immune response of T and B cells, and during the passaging process, gene mutations would occur, and these mutations have unknown effects on virus pathogenicity. To avoid this, Ifnar^−/− C57BL/6 mouse model with adaptive immunocompetence and the virus strain originally from pigs were used in our study. This mouse model is suitable for FMDV infection and highly replicates the characteristics of FMDV latent infection.

Single-cell RNA sequencing (scRNA-seq), a powerful technology that enables profiling the transcriptome of individual cells, offers us a tool for analyzing the heterogeneity of cells with higher resolution than bulk RNA-seq [24,25]. Up to now, this technology has been broadly applied in the field of viral infections [26–29] but not yet in FMDV research. Since more transcriptional profiling of responding immune cells is needed during FMD, large-scale scRNA-seq may provide unbiased and comprehensive visualizations that help us to better understand the adaptive immune response under FMDV infection.

To this end, splenocytes isolated from FMDV-infected and control mice are analyzed by scRNA-seq to investigate the transcriptional profiling of T and B cells in this study. In general, the potential regulation that viral infection imposes on cell composition was examined at both the global and subset levels during FMD. Furthermore, the analyses of single-cell gene expression revealed the changes of T and B cells, indicating the underlying regulation of the adaptive immune system. In addition, the investigation of cell-cell communication and regulatory networks has also provided novel information on the cellular interactions and transcriptional regulation of T and B cells during FMD. In summary, our study first provided comprehensive cell landscapes in single-cell resolution during FMD and could be a valuable reference for exploring the pathogenesis of FMDV. This knowledge is also beneficial for potentially guiding the vaccine design.

Material and methods

Ethics statement

All animals were handled in strict accordance with good animal practice according to the Animal Ethics Procedures and Guidelines of the People’s Republic of China. The study was approved by the Animal Ethics Committee of Lanzhou Veterinary Research Institute, Chinese Academy of Agricultural Sciences (Approved No. LVRIAEC-2023–017).

Biosafety statement

All experiments involving FMDV were conducted in the Biosafety level 3 (BSL-3) facilities at the Lanzhou Veterinary Research Institute of the Chinese Academy of Agricultural Sciences, which are approved by the Ministry of Agriculture and Rural Affairs and the China National Accreditation Service for Conformity Assessment.

Animal, cell culture, and virus

5 to 6 weeks female ifnar^−/− C57BL/6 mice were purchased from Cyagen Biosciences Inc. BHK-21 cells were cultured with DMEM (Gibco, USA) containing 10% FBS (Gibco, USA), 100 IU/mL penicillin and 100 μg/mL streptomycin in an incubator with 5% CO₂ at 37°C.The FMDV/O/GD/CHA/2010/S/BF8 strain virus was stored in the Lanzhou Veterinary Research Institute, Chinese Academy of Agricultural Sciences (LVRI, CAAS).

Animal infection

The live FMDV infection operations in this study were conducted in the biosafety level 3 (BSL-3) facilities of LVRI, CAAS. Each mouse was subcutaneously inoculated with 4.7 log₁₀ PFUs/mL of FMDV/O/GD/CHA/2010/S/BF8 strain in a volume of 100 μL or equal amounts of RPIM1640 (Gibco, USA). After viral inoculation, all mice were monitored twice daily for clinical signs including lethargy, hunching, reduced mobility, piloerection, body weight loss, and other symptoms associated with viral infection. Mice were housed under controlled environmental conditions (12-hour light/dark cycle, 20–25°C, and 50–60% humidity) with free access to food and water. Spleen samples from mock- and FMDV-infected mice were collected and subjected to scRNA-seq at the indicated time.

TCID₅₀ assay

BHK-21 cells (2 × 10⁵/well) were seeded in a 96-well plate with 100 μL DMEM medium containing 10% FBS for each well and incubated at 37°C overnight. After incubation, the medium was removed, and serially, 10-fold diluted (with DMEM containing 2% FBS) virus was added to BHK-21 cells and incubated at 37°C for 7 days. During this period, the infected cells were monitored daily, and the percentage of cell death was measured from a cell viability assay using microscopy. On 7 days post-infection (dpi), the dilution at which 50% of the cell cultures are infected (the end point) is used to calculate (Reed-Muench methods) a TCID₅₀ result mathematically and is expressed as 50% infectious dose (TCID₅₀) per milliliter (TCID₅₀/ml) [30].

Rt-qPCR

Total RNA was isolated from heart, peripheral blood, liver, lymph nodes, spleen, lung, and kidney tissues using an RNA isolating kit (TaKaRa, 9766, Dalian, China) according to the manufacturer’s protocol. To eliminate potential genomic DNA contamination, the extracted RNA was treated accordingly before being reverse transcribed into complementary DNA (cDNA) using the PrimeScript™ RT reagent kit (TaKaRa, RR036A, Dalian, China). Quantitative PCR analysis was performed on the LightCycler 480 II System (Roche, Basel, Switzerland) using the Probe qPCR Mix with UNG (TaKaRa, RR391A, Dalian, China). The primers and probe used in this study were as follows: 3D-F (ACTGGGTTTTACAAACCTGTGA), 3D-R (GCGAGTCCTGCCACGGA), and the TaqMan probe (FAM-TCCTTTGCACGCCGTGGGAC-TAMRA).

In vitro IFNγ ELISPOT assay

CD8⁺ and CD4⁺ T cells were isolated from splenocytes using a magnetic bead-based positive selection kit (Miltenyi Biotec, Germany). A total of 2 × 10⁵ purified CD8⁺ or CD4⁺ T cells were cocultured with 10⁵ DC2.4 cells, serving as antigen-presenting cells (APCs), in a 96-well Mouse IFN-γ precoated ELISPOT kit (Dakewe, 2,210,003, China). Cells were stimulated in triplicate with the specified peptide (10 µg/mL) following the manufacturer’s instructions. Positive (PMA-Ionomycin) and negative (DMSO) controls were included in all experiments to ensure assay validity.

Single-cell isolation and Single-cell RNA-seq

The spleens were rinsed twice with precooled RPMI-1640 medium supplemented with 0.04% BSA, followed by mechanical disruption using scissors. The minced tissue was then enzymatically digested with 3 mg/mL Collagenase I (Gibco, USA) at 37°C for 1 hour. After digestion, the resulting cell suspensions were filtered twice through a 40 μm mesh and collected via centrifugation at 300 g for 5 minutes. Red blood cells were subsequently lysed using a MACS lysis buffer (Germany) at 4°C for 10 minutes. The remaining cells were resuspended in 100 μL of RPMI-1640, and both cell viability and concentration were assessed using the LUNA-FL™ Counter (Logos Biosystems). Single-cell libraries were constructed using the Chromium Next GEM Single Cell 3’ GEM, Library & Gel Bead Kit v3.1 (10× Genomics) and the corresponding chip kit, following the manufacturer’s guidelines (OE Biotech Co., Ltd. Shanghai, China) [31]. Sequencing was performed on an Illumina Nova 6000 platform with PE150 read mode.

Cell quality control

Single-cell RNA sequencing data were processed using the Cell Ranger software pipeline (version 5.0.0, 10×Genomics). This pipeline was employed to demultiplex cellular barcodes, align sequencing reads to the reference genome and transcriptome via the STAR aligner, and perform read down-sampling as needed to generate normalized aggregate data across samples. The resulting gene expression matrix, containing gene counts per cell, was further analyzed using the Seurat R package (version 4.0.0) [32]. To eliminate low-quality cells and mitigate potential multiplet contamination, a stringent filtering strategy was applied. Cells were excluded based on the following criteria: (1) fewer than 200 detected genes, (2) fewer than 1,000 unique molecular identifiers (UMIs), (3) log10GenesPerUMI values below 0.7, (4) mitochondrial RNA content exceeding 20% of total UMIs, and (5) hemoglobin gene expression accounting for more than 5% of total UMIs. To further refine the dataset, the DoubletFinder package (version 2.0.2) [33] was utilized to identify and remove potential doublets.

Data processing for scRNA-seq

Following quality control filtering, splenic single cells were retained for downstream analyses. Normalization of gene expression counts was performed using the NormalizeData function in the Seurat R package (version 4.0.0) [32]. Specifically, the “LogNormalize” method was applied, where gene expression values for each cell were scaled by total transcript counts, multiplied by a factor of 10,000, and log-transformed. Highly variable genes across cells were identified based on previously established methods [34] using the FindVariableGenes function (mean.function = Fast Exp Mean, dispersion.function = Fast Log VMR) in Seurat. Dimensionality reduction was conducted via principal component analysis (PCA) with the RunPCA function. A graph-based clustering approach was then employed to group cells with similar gene expression patterns using the FindClusters function. To visualize cellular distributions, Uniform Manifold Approximation and Projection (UMAP) was applied via the RunUMAP function in Seurat. Marker genes specific to each cluster were determined using the FindAllMarkers function (test.use = presto), identifying genes significantly upregulated in each cluster relative to all other cells. Cell type annotation was performed using SingleR (version 1.4.1) [35], an automated computational method for single-cell RNA sequencing (scRNA-seq) classification, leveraging the reference transcriptomic dataset “scmca” [36] to infer cell identities. Differentially expressed genes (DEGs) were identified using the FindMarkers function (test.use = presto) in Seurat, applying a significance threshold of p < 0.05 and |log2 fold change| > 0.58. Functional enrichment analysis of DEGs, including Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, was conducted in R using a hypergeometric distribution-based approach.

Gene set enrichment analysis (GSEA)

Gene Set Enrichment Analysis (GSEA) [37] was used to complete GO term enrichment analysis with the Molecular Signatures Database (MSigDB) C5 GO gene sets (Version 7.2).

Pseudotime analysis

Developmental pseudotime analysis was conducted using the Monocle2 package [38]. Initially, raw gene expression counts from the Seurat object were converted into a CellDataSet object using the importCDS function. To identify genes informative for ordering cells along the pseudotime trajectory, the differentialGeneTest function was applied, selecting genes with a q-value < 0.01. Dimensionality reduction for clustering was carried out using the reduceDimension function, followed by trajectory inference via the orderCells function with default settings. Changes in gene expression over pseudotime were visualized using the plot_genes_in_pseudotime function, allowing for the tracking of dynamic transcriptional alterations across the inferred trajectory.

Cell-cell communication analysis

CellPhoneDB (v2.0) [39] was employed to identify biologically meaningful ligand-receptor interactions from single-cell transcriptomic (scRNA-seq) data. A ligand or receptor was considered “expressed” in a given cell type if at least 10% of the cells within that cluster exhibited non-zero read counts for the corresponding gene. To assess statistical significance, cluster labels were randomly shuffled across all cells, and the interaction detection process was repeated to generate a null distribution for each ligand-receptor pair across different cell type comparisons. This permutation procedure was conducted 1,000 times, and P-values were computed based on the standard distribution curve derived from the permuted interaction scores. To construct cell-cell communication networks, interactions were established between two cell types if a ligand was expressed in one and its corresponding receptor in the other. Visualization of these interaction networks was carried out using the R packages Igraph and Circlize.

Scenic analysis

SCENIC analysis was performed using the motif database for RcisTarget and GRNboost (SCENIC version 1.1.2.2 [40], corresponding to RcisTarget 1.2.1 and AUCell 1.4.1) with default settings. Specifically, transcription factor (TF) binding motifs that were significantly enriched in gene sets were identified using the RcisTarget package. The AUCell package was then used to quantify the activity of regulon groups within individual cells. To assess the specificity of each predicted regulon across cell types, we computed the regulon specificity score (RSS) using Jensen-Shannon divergence (JSD), a metric for comparing probability distributions. This involved calculating JSD values by comparing binary regulon activity profiles with cell type assignments [41]. Additionally, the connection specificity index (CSI) for all regulons was determined using the scFunctions package (https://github.com/FloWuenne/scFunctions/).

Flow cytometry (FACS)

For flow cytometry analysis, the splenocytes were washed with 500 μL flow cytometry staining buffer (Biosharp, BL1136A) twice and subjected to anti-CD16/CD32 (BioLegend, 101,302) for 20 min on the ice for Fc receptor blocking. After that, the splenocytes were stained with an antibody mixture containing PE594-conjugated CD45 (BioLegend, 103,146), AF700-conjugated CD3 (BioLegend, 100,216), Pacific Blue-conjugated CD4 (BioLegend, 100,531), PE-conjugated CD8a (BioLegend, 100,708), PEcy7-conjugated CD19 (BioLegend, 115,519) with a 1:100 dilution ratio in a total volume of 100 μL for 30 min on the ice. Zombie Aqua^TM Fixable Viability Kit (BioLegend, 423,102) was used to determine the cell viability according to the manufacturer’s protocol. Subsequently, the stained splenocytes were washed twice and resuspended in 200 μL flow cytometry staining buffer. Finally, at least 50,000 cells were collected and analyzed using CytoFlex LX (Beckman, Germany) and FlowJo software (BD Biosciences).

Statistical analysis

All data were analyzed with GraphPad Prism software, version 8.0. Results are displayed as the mean ± standard errors. A p-value < 0.05 was considered significant.

Study period

This study was conducted in three phases. Mice infection and sample collection were performed from March 2 to 10, 2023. Single-cell RNA sequencing (scRNA-seq) was carried out from 7 March 2023 to 10 April 2023. Data processing and statistical analyses were subsequently conducted between 1 May 2023 and 30 April 2024.

Results

Characterization of FMDV infection in Ifnar^−/− mice

Firstly, the FMDV infection features in Ifnar^−/− C57BL/6 mice were characterized, and FMDV has been shown to efficiently replicate in the mouse model, with peak viral loads detected in the spleens on the 5-day post-infection (dpi) (Figure 1A). Moreover, FMDV infection significantly stimulates CD4⁺ and CD8⁺ T cells in the spleens of Ifnar⁻/⁻ C57BL/6 mice, leading to a robust production of IFNγ upon re-stimulation with the FMDV VP2 antigenic epitope (Figure 1B, C). These findings suggest that Ifnar⁻/⁻ C57BL/6 mice serve as a suitable animal model for studying FMDV infection and host immune responses.

Figure 1.

FMDV replicates efficiently in the mouse spleen and stimulates T cell activation. (a) at 0-, 3-, 5-, and 7-days post-infection, viral rna levels were determined using RT-qPCR (n = 5). (b, C) IFNγ-ELISPOT assays were conducted to confirm T cell activation with (b) CD8⁺ or (c) CD4⁺ T cells isolated from infected mice. Two independent experiments were performed (n = 5 mice per experiment) in triplicate. Data are presented as the mean number of spot-forming cells (SFC) per 10⁶ indicated T cells.

Profiling the splenocyte populations by scRNA-seq in FMDV-infected mice

The above findings suggest that Ifnar⁻/⁻ mice provide a valuable model for evaluating the adaptive immune response to FMDV. Based on the model, three biological replicates were selected to ensure statistical reliability while balancing ethical and logistical considerations [42–44]. All the spleens of FMDV-infected mice were sampled at 5 dpi and 10x Genomics-based scRNA-seq was conducted to comprehensively characterize the transcriptional profile of miscellaneous cell types in the spleens (Figure 2A). A total number of 8550, 8035, 8794, 10,371, 10,200, and 9867 high-quality cells were captured from 3 infected and 3 mock mice for analysis, respectively (Figure S1A). Through t-distributed stochastic neighbor embedding (t-SNE) dimensionality clustering analysis, eight major cell clusters were captured and displayed in two dimensions according to the expression of marker genes (Figure 2B-E). These cell clusters were comprised of T cells (CD3e⁺CD3d⁺CD3g⁺) [45], natural killer cells (NKs) (Nkg7⁺Ifng⁺Klrc1⁺Klrd1⁺) [46], innate lymphoid cells (ILCs) (Il7r⁺Rora⁺) [47], B cells (CD19⁺CD79a⁺CD79b⁺) [48], dendritic cells (DCs) (Flt3⁺CD209⁺Itgax⁺Tap2⁺) [49], macrophages (Macs) (Cs1r⁺Itgam⁺) [49], neutrophils (Slpi⁺Retnlg⁺Csf3r⁺) [50] and fibroblasts (Col1a1⁺Col1a2⁺) [51] (Figure 2B, C, 3A-H). The cell proportions in FMDV-infected and mock samples are shown in Figure 2E. Notably, the two groups of cells were similarly distributed across cell-type clusters, and there were no significant differences in the proportions between FMDV-infected and control mice (Figure 2D, E). To further confirm these results and investigate the potential differentiations in adaptive immune responses during FMDV infection, the flow cytometry was conducted to evaluate the compositions of T and B cell populations at different time points. Bearing a resemblance to the scRNA-seq results, the relative proportions of CD4⁺T (CD45⁺CD3⁺CD4⁺CD8a⁻), CD8⁺T (CD45⁺CD3⁺CD8a⁺CD4⁻), and B cells (CD45⁺CD19⁺) remained stable at 7, 14, 21 and 28dpi (Figure 3I). In general, these results indicated that FMDV infection did not reshape the major cell-type populations of the spleen in the mice.

Figure 2.

Single-cell transcriptome profiling of mouse spleen and verification of global changes of cell types upon FMDV infection. (a) Overview of the study design. Single cells were isolated from the spleens of mock- and FMDV-infected mice (n = 3) and simultaneously subject to scRNA-seq (10×Genomics). (b) Dot plots of the mean expression of marker genes used for cluster annotations. (c) t-SNE projection of the major eight cell subsets colored according to cell types. (d) t-SNE projection of single cells from mock- or FMDV-infected mouse spleen. (e) Bar plots of the proportions of different cell types from mock- or FMDV-infected mouse spleen.

Figure 3.

The t-SNE plot colored for gene expression in different cell types and verification of global changes of T and B cell populations upon FMDV infection at various time points. (A-H) t-SNE plot showing gene expression across multiple cell types. (a) T cells, (b) NKs, (c) B cells, (d) DCs, (e) ILCs, (f) Fibroblasts, (g) Macrophages, (h) Neutrophils. (i) Flow cytometry analysis showing the frequency of CD4⁺T cells, CD8⁺ T cells, and B cells from mock- or FMDV-infected mouse spleen. A two-tailed Student’s t-test was used to compare mock and each FMDV-infected group. All error bars correspond to sem, ns: nonsignificant.

Functional plasticity of T and B cells during FMDV infection in vivo

Since FMDV infection barely altered the overall cell composition of the spleen, including T and B cells, we performed a focused analysis of gene expression changes within these two major lymphocyte populations using single-cell RNA sequencing. A total of 3,764 DEGs in T cells and 451 DEGs in B cells were identified, with the top significantly altered genes displayed in volcano plots (Figure 4A) and heatmaps (Figure 4B). These DEGs were categorized into up- and down-regulated groups based on their expression trends. Interestingly, we observed 19 overlapping down-regulated genes and 6 overlapping up-regulated genes between T and B cells (Figure 4C), suggesting a coordinated transcriptional reprogramming in adaptive immunity during FMDV infection. To reveal the biologically meaningful patterns, the GO and GSEA analysis was conducted.

Figure 4.

(Continued).

Figure 4.

Characterization of gene expression profiles of T and B cells in FMDV-infected mouse spleen. (a) Volcano plots of differentially expressed genes in T (left) and B cells (right). (b) Heatmaps of the expression of the top 20 DEGs in T (left) and B cell (right) cluster. (c) Venn diagram of down-, up-, and overlapping regulated genes in T and B cells. (d) Go analysis of down-, up- and overlapping regulated genes in T and B cells.

In T cells, the down-regulated genes were significantly enriched in pathways involved in T cell activation, antigen processing and presentation (including MHC class I and II signaling) and T cell mediated cytotoxicity (Figure 4D, 5A). These findings indicate a functional exhaustion or suppression of T cells, which may impair effective adaptive responses to FMDV. However, the up-regulated genes were primarily associated with ribosomal biogenesis, translational initiation, and RNA metabolic processes (Figure 4D, 5B), suggesting an increase in basal cellular activity, possibly as a compensatory mechanism in response to viral challenge or stress. In B cells, the down-regulated genes were also enriched in antigen presentation pathways (Figure 4D, 5C). Additionally, protein-folding processes, such as chaperone activity, were down-regulated, suggesting impaired maturation or antigen processing capacity (Figure 4D). In contrast, up-regulated DEGs were associated with phagocytosis, complement activation, cell division and post-translational modification (Figure 4D, 5D), which may reflect B cell functional diversification or regulatory phenotypes during infection.

Figure 5.

Function changes of T and B cells in FMDV-infected mouse spleen based on GSEA. (A-D) changes in pathway activities scored in T and B cells by GSEA, with enriched go terms. (a) Downregulated in T cells. (b) Upregulated in T cells. (c) Downregulated in B cells. (d) Upregulated in B cells.

Collectively, these results suggest that while FMDV infection does not significantly alter spleen cell composition, it induces a profound functional reprogramming in T and B cells. The suppression of antigen presentation and immune activation, combined with increased apoptotic and translational processes, points to virus-induced immune evasion mechanisms, potentially facilitating FMDV persistence or systemic spread. The shared downregulation of key immune genes across T and B cells underscores a concerted impairment of adaptive immunity, which may represent a key hallmark of FMDV pathogenesis.

Heterogeneity of subtypes of T and B cells in the spleen in FMDV-infected mouse

To understand the detailed characteristics of T and B cells involved in FMDV infection, these two cell populations were selected for further subclass analysis. In our study, fifteen and seven subclusters of T and B cells were identified (Figure 6A-D) according to the classic specific gene expression signatures (Figure 6E,F).

Figure 6.

Subclusters of T and B cells in FMDV-infected mice spleen. (a, C) umap projection of T and B cell subsets in mock- and FMDV-infected mice. (a) T cells, (c) B cells. (b, d) T and B cell subsets proportions in mock- and FMDV-infected mice. (b) T cells, (d) B cells. (E-F) heatmaps of marker genes among T (e) and B cells (f) subtypes.

Unlike the stability of global T and B cell clusters, subcluster analysis revealed substantial heterogeneity and dynamic remodeling within subset populations (Figure 6B,D). Importantly, an increase in effector and proliferative T cell subtypes was observed, such as cytotoxic CD8⁺ T cells (CD8⁺Ccl4⁺Ly6c2⁺), effector CD4⁺ T cells (Ifng⁺Nkg7⁺), and proliferative T cells (CD4⁺CD8⁺Mki67⁺) (Figure 6B). These populations are typically associated with heightened immune activity, suggesting that FMDV infection triggers a robust cellular immune response. Moreover, the increase of Th1 (Il10⁺ Il2⁺) cells may reflect a protective immune effort to control viral replication while minimizing immune-mediated tissue damage. Central memory CD4⁺ T cells (CD4⁺Tgfbr3⁺Dapl1⁺) were also slightly increased. This type of cell is responsible for long-term immune surveillance and rapid recall responses upon re-exposure to pathogens, suggests that adaptive immunity is not only activated but also being programmed for durable memory after FMDV infection. In contrast, a decrease was observed in subpopulations associated with immune regulation, including Th2 (Il4⁺ P2rx7⁺), Th17 (Il22⁺ Il17a⁺ Ccr6⁺), and gdT1 (IL17a⁺ Trdc⁺ Scart2⁺) subsets (Figure 6B). This shift likely reflects a functional reprogramming of the immune landscape during FMDV infection, favoring cytotoxic effector responses and Th1-biased immunity. The downregulation of IL-17-producing subsets in particular may serve to limit excessive inflammation and immunopathology within lymphoid tissues, while enabling a more focused antiviral response. A decrease in NK (Klra8⁺ Eya1⁺ Styk1⁺) and ILC (Dscam⁺ Kit⁺ Cited4⁺) subsets was also observed following FMDV infection (Figure 6B). These decreases may suggest that as the infection progresses, the immune system reprograms toward a more adaptive, cytotoxic T cell-dominated response, likely due to antigen-driven expansion. The drop in NK and ILCs could also reflect their early engagement and subsequent resolution or suppression, aligning with the temporal dynamics of antiviral immunity. Interestingly, naïve CD8⁺ T cells (CD8⁺Ly22⁺Ly6c⁺), central memory CD8⁺T cells (CD8⁺Plaur⁺Auts2⁺), gdT17 (IL17a⁺Trdc⁺Klra7⁺), exhausted CD4⁺ T cells (Lag3⁺CD244⁺) and CD4⁺regulatory T cells (Tregs) (Foxp3⁺Dst⁺) remained unchanged (Figure 6B). This may indicate that the immune system has not yet reached a state of exhaustion or suppression at this time point, or that FMDV may selectively modulate these compartments to its advantage. Moreover, B cells were classified into plasma cells and six other subclusters (B1–B6) based on transcriptional profiles (Figure 6C, D). Among them, cluster B1 (Hs3st1⁺ Cystm1⁺ Mpp7⁺) constituted the dominant subset in the mock-infected group but dropped sharply from over 50% to less than 25% post-infection. Conversely, clusters B2 (Il16⁺ Ccr6⁺), B3 (Tifa⁺ Fam129c⁺), B4 (Rapgef4⁺ Bicra⁺), B5 (Srm⁺ Mettl1⁺) showed varying degrees of expansion and plasma cells (Cd9⁺Prg2⁺) held steady following FMDV infection (Figure 6D). While the functional characterization of these subsets remains limited, further investigation is required to determine their precise immunological roles and potential influence in FMDV infection.

Next, pseudotime analysis was used to investigate the cell states and internal relationships among T cell clusters after FMDV infection. In general, compared with mock-infected groups, more CD8⁺ but not CD4⁺ T cells are present at the end of the branches in FMDV-infected groups (Figure 7A). For CD8⁺ T cells, we found that component 1 split CD8⁺ T cells into naïve/memory cells and effectors, indicating a shift from a resting state to an activated state of CD8⁺ T cells (Figure 7B, C). Component 2 divided the cells into two branches correlated with cytotoxic or mitotic-related genes. Interestingly, proliferative T cells were significantly enriched at the end of the lower branch and strongly correlated with mitosis, like mki67. In contrast, almost no proliferative T cells were distributed on the upper branch, indicating disparate differentiation directions of terminal CD8⁺ T cells (Figure 7C). Unlike proliferative T cells, cytotoxic CD8⁺ T cells were observed to be distributed on both branches and displayed a preferable aggregation at the tail.To further explore the features of these two terminal CD8⁺ T cells. Four function modules were identified by gene expression analysis along the two differentiation branches (Figure 7D). According to GO analysis, module 1, which contains genes upregulated in all cell types, was related to multiple necessary biological processes such as cell cycle, cell division, and DNA replication (Figure 7D, E). In contrast, genes belonging to module 2, which was upregulated explicitly in the lower branch, were highly enriched in functions that involved in the host response to infection, such as NK cell-mediated cell killing, inflammatory response, adaptive immune response, and T cell proliferation, showed the participation of immune system after FMDV infection (Figure 7D, E). Moreover, module 3 was specifically upregulated in the upper branch and highly related to the transcription process and post-translational modification (Figure 7D, E). Nevertheless, module 4, which is mainly upregulated in the pre-branch, was involved in translation and ribosome-associated processes (Figure 7D, E). Together, these findings suggest that FMDV infection induces a bifurcated immune strategy, with one arm of CD8⁺ T cells entering a proliferative and inflammatory trajectory, likely contributing to viral clearance, while another arm potentially undergoes functional adaptation that may support sustained activity or regulation. These results align with studies in viral infections, where T cell responses diversify into functional subsets to balance viral control and immunopathology [52,53].

Figure 7.

(Continued).

Figure 7.

The pseudotime analysis of T cells in FMDV-infected mice spleen. (a) Deduced developmental trajectory of T cells in mock- and FMDV-infected mice. (b) Deduced developmental trajectory of CD8⁺ T cells during FMDV infection. (c) Deduced developmental trajectory of individual subsets of CD8⁺ T cells during FMDV infection. (d) Gene expression dynamics model for CD8⁺ T cells based on the pseudotime results. (e) Go analysis of different gene expression modules of CD8⁺ T cells based on the pseudotime results.

In conclusion, the study found and analyzed the heterogeneity of the subpopulations of T and B cells in FMDV-infected mice spleens. Various cell states and functions were also identified after infection, indicating the activation of T and B cells.

Construction of the cell communication atlas in FMDV-infected spleen

Intercellular communication is essential for cell functions. To further investigate whether FMDV infection altered the cell communication atlas in the spleen, CellPhoneDB was used to analyze all the 26 cell clusters described above, including the subtypes of T and B cells. Firstly, increased inferred interactions and enhanced interaction strength were observed, indicating the strengthening of cell-cell communication in the infected group (Figure 8A). Furthermore, our results showed the potential variation in both outgoing and incoming sides for all types of cells during FMDV infection (Figure 8B, C). Among the whole cell types, T cells, especially the CD8⁺ T cells, presented the most intense interactions with other cells (Figure 8B) and took on many more roles as signal receivers than senders (Figure 8C). By contrast, B cell subsets B1-B6 displayed significantly increasing interaction strength with T cells on outgoing sides (Figure 8B) and showed more activation as signal senders rather than receivers (Figure 8C).

Figure 8.

Cell – cell communications in FMDV-infected mouse spleen. (a) the bar plot shows the interactions and interaction strength of different cells in the spleens of mock- and FMDV-infected mice. (b) heatmaps of the number of interactions (left) and interaction strength (right) of different cells. The Y axis shows the ligand cell types, while the X axis represents the receptor cell types. The colored bar on the top and right represents the signal value of each cell type as a source, sender, or receiver. (c) the scatter plot shows the interaction strength of different cells as signal senders or receivers. (d) the information flow of signaling pathways according to cell – cell interaction analysis. The pathways with blue and red text are enriched in mock- and FMDV-infected mouse spleen, respectively. The pathways with black text are neither enriched in mock- nor FMDV-infected mouse spleen. (e) the circle plot shows the ifn ii signaling pathway-related interactions among different cells. The number of interactions is proportionate to the thickness of the connecting line.

Next, the potential regulations of signal pathways from the angle of cell-cell communications were explored. Notably, pro-inflammatory signaling pathways such as tumor necrosis factor (TNF), interleukin2 (IL2), and interferon-gamma (IFN II) were significantly activated in the infected group. In contrast, the anti-inflammatory signaling like IL10 was more related to the mock-infected group (Figure 8D). These results showed a possible inflammatory condition in the infected spleen. At the same time, remarkable activations in multiple critical pathways were also observed in the FMDV-infected group, such as fibronectin (FN1, highly involved in cell adhesion, maintenance of cell shape [54], osteoblast mineralization [55] and binds cell surfaces and various compounds including collagen, fibrin, heparin, DNA, and actin) [56], thy-1 membrane glycoprotein (THY1, plays a role in cell-cell or cell-ligand interactions) [57] and CD23 (has essential roles in the regulation of IgE production and the differentiation of B cells) [58] signaling pathways (Figure 8D). All these findings suggested the potential alterations of cell-cell communications resulting from FMDV infection, including a shift from stable to a potentially inflammatory condition in the spleen and the disturbance in cell adhesion, cell-cell interaction, and cell development functions.

Since IFN-γ was essential for antiviral functions, the IFN-II signaling pathway network was subsequently analyzed to understand the underlying mechanism of the abovementioned change. Compared with the control group, the B cell subcluster B3 presented a distinct activity in receiving IFN-γ related signals from other cells (Figure 8E). Moreover, central memory CD8⁺ T cells, cytotoxic CD8⁺ T cells, NKs, ILCs, th1, and effector CD4⁺ T cells showed stronger secretory activities in IFN-γ associated signals (Figure 8E). These results suggested the activation of the IFN-II signaling pathway during FMDV infection; however, the role of subset B3 in this process still needs to be revealed in more studies.

As impressive changes were observed in the antigen-presenting process of both T and B cells (Figure 4D, 5A, B), the study finally investigated the interactions between antigen-presenting cells (APCs) and T cells via classical major histocompatibility complex (MHC). To this end, bubble plots were drawn to display the potential communication probabilities (Figure 9A-F). As the results showed, all types of APCs (including DCs, macrophages, and B cells) presented more frequent interactions with CD8⁺ T cells during FMDV infection. Notably, the classical MHC-I (H2-k1 and H2d1) expressed by APCs preferred interacting with CD8b1 but not CD8a. Among all APCs, macrophages showed the most significant increase in interacting with CD8⁺ T cells before and after FMD-infection (Figure 9A-C). Bearing similarity with the results, the interactions between APCs and CD4⁺ T cell clusters via classical MHC-II (H2-A and H2-E) and CD4 molecular also increased during FMDV infection except subtype Th17 (Figure 9D-F). Furthermore, the interaction probabilities dramatically dropped between Th17 and B cells/DCs after FMDV infection (Figure 9D, E), while no signals were detected between macrophages and Th17 (Figure 9F). All these findings above indicated that FMDV infection altered the intercellular communication networks, leading to the activation or inhibition of multiple pathways.

Figure 9.

The cell-cell communications in antigen presenting process during FMDV infection. (A-F) bubble plots showing the interaction probabilities between APCs and T cells via classical MHC molecules. (a) B cells-T cells via MHC I, (b) DCs-T cells via MHC I, (c) macrophages-T cells via MHC I, (d) B cells-T cells via MHC ii, (e) DCs-T cells via MHC ii, (f) macrophages-T cells via MHC ii.

Gene regulatory networks of T and B cells

Transcription factors (TFs) and the corresponding downstream targets participate in the regulation of cell functions to a great extent. These two components also comprise a comprehensive and complex gene regulatory network. Thus, a better understanding of each cell’s regulons (composed of TFs and their target genes) states enables us to touch the underlying mechanisms of the shifts of cell functions under the infection of FMDV. To this end, single-cell regulatory network inference and clustering (SCENIC) analysis was conducted to explore the regulon states of T and B cells. As expected, different subclusters showed distinct regulon activities quantified by regulon activity score (RAS) (Figure 10A, B). Moreover, potential changes in regulon activities were observed before and after FMDV infection (Figure 10C, D), and among all the differently expressing regulons, Fli1, Irf2, Foxo1, Jun, Fosb, Jund, Rel, Nfkb1, and Stat1 appeared simultaneously in T and B cells (Figure 10C, D). Furthermore, Rel, Nfkb1, and Irf2 presented contrary tendencies in T and B cells. The RAS of Rel and Nfkb1 increased in T cells but decreased in B cells, while that of Irf2 was the opposite. Rel is an essential component of nuclear factor-kappa B (NF-κB, which is formed by the Rel-like domain-containing proteins like RELA/p65, RELB, NFKB1/p105, NFKB1/p50, REL, and NFKB2/p52), participating in regulating many biological processes such as inflammation, immunity, differentiation, cell growth, tumorigenesis, and apoptosis. Nfkb1, which encodes NF-κB p105 subunit, is also an essential factor to NF-κB. P105 is the precursor of the active p50 subunit, while the latter is associated with RELA/p65 to form the NF-κB p65-p50 complex. Furthermore, the p65-p50 complex acts as a transcription factor binding to κB consensus sequence 5”-GGRNNYYCC-3,” which belongs to the enhancer region of genes involved in the immune response [59,60]. Thus, Rel and Nfkb1 can exert synergistic functions through regulating NF-κB related target genes. Notably, Rel and Nfkb1 simultaneously showed high RAS in exhausted CD4⁺ T cells whereas dropped in plasma cells (Figure 10A, B), indicating the regulated NF-κB related function of these cells. Irf, which promotes differentiation toward the B lineage by inhibiting the T-cell instructive Notch signaling pathway through the specific transcriptional repression of Notch downstream target genes, plays a key role in differentiating lymphoid progenitors into B and T lineages [61]. In T cells, IRF-related regulons such as Irf1, Irf2, and Irf4 presented various RAS in different subclusters. Among these, the RAS of Irf2 regulons dropped in Th17 and effector CD4⁺ T cells while dramatically increasing in B2 and B4 clusters (Figure 10A, B). Since RAS is positively correlated with the activity of regulons in cells, it suggests that irf2 is activated in B2 and B4, and Irf2 has been shown to regulate the B cell proliferation and antibody production [62], this may suggest an activation of humoral immunity against FMDV infection.

Figure 10.

Gene regulatory networks in FMDV-infected mouse spleen. (A-B) heatmaps showing the ras of different subsets of T (a) and B cells (b). (C-D) heatmaps showing the ras of t(c) and B cells (d) before and after the infection of FMDV. (E-F) the csi matrix highlights regulon-to-regulon correlation across T (e) and B cells (f) four distinct regulon modules were determined in both T and B cells. Regulons shared between T and B cells were labeled by asterisks.

Next, the regulons were classified into different modules based on the functions, and the connection specificity index (CSI) was used to determine the potential associativity between individual regulons. The analysis of T and B cells resulted in 44 and 42 regulons across four function modules, respectively (Figure 10E, F). The nine regulons (highlighted by asterisks) shared between T and B cells gathered in modules 2 and 4 in T cells and uniformly distributed in all modules in B cells. For T cells, regulons belonging to the same module presented pretty high CSI, including Rel and Nfkb1 described above (Figure 10E), indicating their co-regulatory capacity of downstream genes. Unsurprisingly, similar results were observed in B cells. Additionally, irf2 showed high CSI with regulons across different modules, like stat1 from module 1, fli1, and foxo1 from module 4 (Figure 10F), suggesting its extensive function for B cells.

Finally, the potential metergasis of T and B cells was examined from the perspective of regulatory networks by analyzing regulon activity states and cell subset compositions. To this end, Th1, Th2, B1, and B2 subclusters that showed the most significant changes during FMDV infection (Figure 6A-D) were chosen for further analysis. The downstream target genes belonging to the top enriched regulons in these subclusters were subsequently subjected to KEGG enrichment analysis (Figure 11A-D). As the results showed, six signaling pathways were mainly enriched in these subclusters, including NF-κB, TNF, apoptosis, Th1 and Th2 differentiation, Th17 differentiation, and IL-17 signaling pathway. The results suggested the tanglesome impacts of FMDV infection on inflammation, cell differentiation and cell death processes [63–65].

Figure 11.

Potential transcription regulatory heterogeneity in subsets of T and B cells during FMDV infection. (A-D) kegg analysis for downstream target genes of regulons in subset B1 (a), B2 (b), Th1 (c) and TH2 (d). The shared pathways were labeled by red boxes.

Discussion

FMD is one of the most severe domestic animal diseases throughout the world. Despite FMD vaccines effectively eliciting protective Abs that have been applied for over 60 years, the significant pathogen has not yet been eliminated [1,66,67]. This mainly results from an incomprehensive knowledge of the interaction between viruses and the hosts’ immune system. In this study, an integrative transcriptomic landscape with high-resolution spleen immune cells after FMDV infection was described for the first time by scRNA-seq technology. Our findings revealed the proportional changes in the subclusters of T and B cells, although these changes were not observed at the global level. We identified the DEGs of T and B cells and characterized the corresponding functions, investigated the cell-cell communication and transcription regulatory networks in the spleens of FMDV-infected mice. These data enabled us to explore the changes from various angles in the T and B cells of infected mice compared to healthy controls, thereby accelerating a better understanding of the host adaptive immune response against FMDV.

In our study, we observed that the overall immune cell composition in the spleen, particularly the proportions of T and B cells, remained relatively stable following FMDV infection. This finding appears to contrast with reports of lymphopenia and substantial T cell depletion in the peripheral blood of infected animals, as noted in previous studies [68,69]. We believe this discrepancy can be attributed to differences in samples, animal models, and disease kinetics. The spleen, as a secondary lymphoid organ, plays a crucial role in filtering blood and supporting immune cell proliferation. It may retain lymphocytes during infection, whereas the peripheral blood reflects dynamic immune cell trafficking and redistribution. Several studies, including those on LCMV and influenza (IAV), show that immune cells often migrate from circulation into lymphoid and peripheral tissues during viral infections [70]. Moreover, the use of mice model might present different immune features to natural hosts. Furthermore, the stage of infection and the severity of disease also influence immune cell profiles. Our sampling may have captured a phase in which splenic lymphoid architecture is maintained, while previous studies may have examined time points reflecting peak systemic lymphopenia. Nevertheless, the stable global cell composition after infection observed in our research seemed consistent with another study on the Japanese encephalitis virus (JEV) [71].

Although recent works have shown the research into global populations of T and B cells, those studies did not analyze the potential internal state changes of the two dominating cell types for the adaptive immune system. Here, our work identified the DEGs of both T and B cells with single-cell resolution under FMDV infection and determined the functional shifts through enrichment analysis. According to GO analysis, the downregulated genes for T and B cells were highly connected with immune response and antigen-presenting process. In contrast, the upregulated genes were closely related to ribosome/translation pathways in T cells and multifunctional pathways in B cells. To further verify these results, GSEA was subsequently conducted to investigate the functional regulations of T and B cells. From GSEA results, the inhibition of both antigen-presenting process and cellular immunity in T and B cells were determined after FMDV infection, indicating the existence of immunosuppression of adaptive immunity [72,73]. Notably, the immune escaping of FMDV by disrupting the antigen-presenting process has already been proved in other studies [68,69,74,75]. Interestingly, two upregulated genes in both T and B cells named gm26917 and gm42418 were identified. Other research has found that the long non-coding RNAs encoded by these two genes were enriched in NLRP3 inflammasomes in macrophages [76]. This result indicated the potential regulation of inflammatory response in T and B cells by FMDV infection. Moreover, the occurrence of cmss1 in the intersection of up-regulated genes of T and B cells suggests its possible profound effects on FMDV infection, which, however, require further research to uncover. Future studies could focus on validating the expression of cmss1 at the protein level using flow cytometry or immunoblotting, performing gene knockout or knockdown experiments in vitro or in animal models to assess its immunological role and involvement in FMDV-induced immune modulation. Additionally, its potential interactions with known immune regulators could also be explored to determine whether it contributes to viral immune evasion or host defense.

Since T and B cells are classified into several subtypes, and each executes various biological functions, there is great significance in looking into the underlying changes of these cell clusters. In this respect, a significant remodeling to various extents was observed in several subtypes. Among these, the increase of Th1 and the impairment of the Th2 subtype suggested the activation of cellular immune response but not humoral immune response at 5dpi of FMDV [77,78]. Moreover, the plasma cells showed a relatively stable proportion. These findings emphasized the importance of cellular immunity in FMDV pathogenesis. The pseudotime analysis found that CD8⁺ T cells showed two cell fates along the major bifurcations. Along the upper branch, CD8⁺ T cells evolved toward antiviral functions marked with cytotoxic-related genes. Along the lower branch, the development of CD8⁺ T cells revolved around both antiviral activity and cell proliferation. Interestingly, during this process, a proliferative T cell cluster (a CD8 and CD4 double positive cluster marked by proliferative related gene mki67) that may play essential roles in immune regulation and T cell development was identified. This hypothesis was then corroborated by the gene expression analysis in a pseudotime-dependent manner. According to the results, the cell killing, inflammation, adaptive immune response, and T cell proliferation-related genes were enriched in the lower branch, dominated by the proliferative T cell cluster.

Importantly, these functional alterations in T and B cells could be contextualized within the known immune evasion mechanisms of FMDV. FMDV encodes several proteins which suppress host innate immunity by targeting interferon signaling, disrupting NF-κB activation, and cleaving host transcription/translation factors [79–81]. These viral strategies inhibit the early innate alarm signals that normally drive antigen presentation and T/B cell priming. In our study, impaired antigen-presenting capacity, altered cell-cell interactions, and reduced Th2-associated activity could be downstream consequences of such viral interference.

Studies indicate that viruses promote infection by exploiting and manipulating cell-cell communications [82–84]. However, whether similar mechanisms are involved in the functional regulation of T and B cells during FMDV infection remains unknown. In cell-cell communication analysis, the enhancement in both interaction strength and frequency among different immune cells in FMDV-infected mice spleens compared with mock infection were determined. By ligand-receptor analysis, the activations of T and B cells on the outgoing or incoming sides during FMD were identified. The enhanced communications between individual cell subsets related to the IFN-II pathway, which is essential for the immunological process were also observed. Moreover, antigen-presenting process-related interactions between APCs and T lymphocytes were investigated and CD8b1 was found to be the regnant receptor rather than CD8a. To better understand the cell regulatory states of T and B cells during FMD, SCENIC analysis was conducted based on scRNA-seq data. A total of nine regulons were identified, and significant changes were made simultaneously in both T and B cells. Intriguingly, some regulons were proved to exert synergistic functions like Rel and Nfkb1. Overall, through analyses of cell-cell communication and regulatory networks, some cell features not described before were characterized, and these might have significant impacts on driving FMD.

Nevertheless, there are still some improvable points in our study. Firstly, although the Ifnar^−/− mice have been proven susceptible to FMDV, the discrepancy may still exist between the mouse model and the natural hosts. Future studies could employ large animal models, such as swine or bovine to better simulate the immunopathogenesis of FMDV under more physiologically relevant conditions. Secondly, our study detected FMDV-infected and mock mouse spleen only at a one-time point (5dpi); thus, it could not reflect comprehensive infection dynamics during the FMDV pathogenetic process. To solve this, future experiments could include longitudinal sampling across multiple time points, which would allow for analysis of immune cell composition, activation status, and gene expression changes over the course of FMDV infection. Another limitation of this study is the absence of experimental validation of the DEGs identified through scRNA-seq. However, the study provided a comprehensive transcriptomic overview of cellular heterogeneity and signaling dynamics within the system. Future research may therefore perform targeted experiments to investigate the biological importance of candidate genes and pathways identified in this dataset. We believe that the insights gained from our data-driven analysis offer a valuable framework for understanding the cellular landscape and can serve as a reference for future studies that aim to experimentally validate specific gene targets. Despite these limitations, this research can enlighten future studies about FMD immune response and pathogenesis.

In summary, our study provided a single-cell transcriptome landscape for FMDV at single-cell resolution in mouse spleen. The heterogeneity of cell composition was determined and the characteristics of regulations in gene expression, cell communication, and regulatory networks in T and B cells were confirmed. These results will be helpful for better understanding the adaptive immune response during FMDV infection and offer a robust resource for effective vaccine design and diagnostic approaches.

Funding Statement

We acknowledge support from the National Key R&D Program of China [2021YFD1800300]; the Support Program for Longyuan Youth and Fundamental Research Funds for the Universities of Gansu Province [2024QNTD18]; the Foundation for Innovation Groups of Basic Research in Gansu Province [24JRRA783 and 23JRRA561]; the Key Project of National Natural Sciences Foundation of China [32330107]; the Key Science and Technology Foundation of Gansu Province [22ZD6NA001]; and the Natural Sciences Foundation of Gansu Province [24JRRA013].

Disclosure statement

No potential conflict of interest was reported by the author(s).

Animal experiments statement

All animal experiments in this study were conducted in compliance with the ARRIVE guidelines to ensure ethical and transparent reporting.

Data availability statement

The data that support the findings of this study are openly available at https://doi.org/10.6084/m9.figshare.29836901.v1 [85]. The raw scRNA-seq data are publicly available in NCBI GEO repository at https://www.ncbi.nlm.nih.gov/geo/and can be found under the accession number GSE279768.

Supplemental data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/21505594.2025.2580141