Introduction

The influenza A virus (IAV) stimulates the innate immune system, activating adaptive immune cells’ proliferation, recruitment, and effector functions, which determine the host’s ability to clear the virus and the extent of tissue damage caused by the infection1. Among these immune cells, IAV-specific T lymphocytes play a crucial role. IAV-specific CD8 + T lymphocytes are the main force responsible for producing effector functions to clear the infection and enhance memory protection. They kill infected cells by binding to death receptors or using pore-forming toxins and granzyme2,3. IAV-specific CD4 + T cells provide support for defending against influenza virus infection, helping CD8 + T cells and B cells respond, and maintaining the anti-viral function of innate immune cells by producing pro-inflammatory cytokines such as TNF-α and IFN-γ2,3. At the same time, there are IAV-specific regulatory T cells (Treg), whose main function is focused on tissue repair. Non-specific innate immune responses and cytotoxic CD8 + T cell-mediated anti-infection killing cause tissue damage after IAV infection. IAV-specific Treg will help suppress this excessive tissue damage and restore tissue homeostasis4,5.

When the effect function of CD8(+) T cells is insufficient, it will exacerbate the infection, increasing morbidity and mortality6. Studies have shown that pulmonary plasmacytoid dendritic cells (pDCs) and CD8α(+) dendritic cells promote the survival and accumulation of T cells in the lung, and the absence of pDCs/CD8α(+) DCs leads to extensive apoptosis of CD8(+) T cells in the lung7,8. However, the regulation of DCs on T cells is not always positive. Other studies have found that during high-dose IAV infection, pDCs induce cytotoxic Fas(+) CD8(+) T cells to undergo apoptosis by driving the FasL-Fas(DC-T) receptor binding, thereby increasing the mortality rate of IAV infection9,10,11. In summary, pDCs have played an unknown harmful role by limiting CD8 T cell responses, and the interaction between pDCs and T cells is not entirely positive, with the boundaries and mechanisms still needing to be confirmed by more research data.

Bulk RNA-seq is a cost-effective method that provides comprehensive gene expression data. It emphasises gene interpretation and differential gene expression analysis (DGE), which is always its standard research tool, suitable for gene expression difference studies and pathway analysis of entire tissues or cell populations12,13. However, its limitation lies in the lack of identifying gene differences between cells, which may cause important information generated by cellular differences to be obscured by the “average” treatment and lost precision14,15. Single-cell RNA-seq can compensate for this deficiency. Single-cell RNA-seq provides an unbiased method for identifying cellular heterogeneity and functional diversity and implements high-throughput multi-dimensional analysis of individual cells16,17. This high-throughput sequencing technology that focuses on cell interpretation can analyse the cell map features of specifically selected cell subpopulations and associate them with the differential gene sets of bulk RNA-seq of interest. By performing a shared analysis, the ultimate core genes of the associated cell types can be locked.

In this study, we adopted a combined transcriptome and single-cell transcriptome analysis approach to link genes and cells, focusing on the changes in the subpopulations of T cells in IAV infection as a starting point. The aim was to explore more key pathways related to the life physiological processes of T cells in severe IAV infection and uncover the target genes associated with these pathways, hoping to open up new perspectives for the IAV research field and its treatment. [Fig. 1]

Fig. 1
figure 1

Research Overview. Employing a multi - omics approach integrating transcriptomics and single - cell transcriptomics, and conducting differential analysis, weighted gene co - expression network analysis (WGCNA), and Friends analysis, we further ascertained that TXN is a crucial target gene that is specifically highly expressed in proliferative T cells. Moreover, we discovered that glutathione metabolism is strongly correlated with the severity of influenza A virus (IAV) infection. During IAV infection, the development of T cells is modulated and altered. Simultaneously, our findings highlight the significant role of T cell death jointly associated with TXN and glutathione metabolism in the progression of IAV.

Full size image

Methods and materials

Data collection

The data used in our study were obtained from the public datasets available on the NCBI Gene Expression Omnibus (GEO) (https://www.ncbi.nlm.nih.gov/geo/), including microarray data of IAV-infected patients and healthy controls18. The microarray dataset GSE101702 contains peripheral blood samples from 107 patients, including 52 healthy controls, 44 patients diagnosed with severe influenza, and 63 patients with non-severe influenza. The scRNA-seq dataset CNP0001102 from CNGB Nucleotide Sequence Archive (CNSA) (https://db.cngb.org/cnsa) contains peripheral blood samples from 3 healthy controls and 2 severe influenza patients (Two patients with severe influenza were sampled respectively at two time points)19.

Microarray data processing

Based on the GSE101720 data, the differential analysis R package “Limma” was used to screen differentially expressed genes (DEGs). All DEGs were set a screening threshold of p-value < 0.05 and log2FC > 0.5. Gene Ontology Enrichment Analysis (GO) was used to annotate the functions of DEGs as biological processes (BP), molecular functions (MF), and cellular components (CC) categories, which is a commonly used method for large-scale gene function enrichment analysis. Kyoto Encyclopedia of Genes and Genomes (KEGG)20 is a database that provides comprehensive information on gene, protein, metabolite, and signalling pathway biomolecules, and the KEGG database was used for Pathway Enrichment Analysis. The GO function and KEGG pathway enrichment analysis were performed on the DEGs related to IAV infection using the R package “clusterprofiler”, and the results were visualised using the R package ggplot2. The immune infiltration analysis was performed using the Cibersort algorithm to calculate the degree of infiltration of 22 types of immune cells between the healthy group and the influenza group21.

Single-cell RNA-seq data processing and analysis

Obtain the gene-cell matrix of dataset CNP0001102 and use the R package “Seurat” to perform Unsupervised clustering Cell clustering. Select cells that express more than 300 genes but fewer than 8000 genes have a red blood cell percentage of less than 5% and a mitochondrial percentage of less than 10%. Use the “FindIntergrationAnchors” and “IntegrateData” functions to integrate the 7 batches of samples, with the dims parameter set to 20. Perform principal component analysis (PCA) and use the first 10 PCs to construct a SNN network. Use the “Louvain” algorithm to identify cell clusters with a resolution of 0.5. Finally, UMAP will be used to visualise the clustering results in a two-dimensional space. In this study, we choose the following immune cell marker genes as the label for each cell cluster and use the UMAP feature map and bubble map to show the expression of the marker genes22: Naive T (LEF1, CCR7), Effector memory CD8 + T (CD8A), Activated CD4 + T (IL7R), Circulating NKs (FCGR3A, TYROBP), Naive B cell (IGHD, TCL1A), Regulatory T cell (IL2RA), Memory B cell (AIM2, TNFRSF13B), NKs (GZMB, XCL1), Plasma cell (JCHAIN, MZB1), Monocyte (LYZ, S100A9), Proliferating CD4 + T (MKI67, TOP2A), Megakaryocyte (PPBP, PF4), DCs (PLD4, SOX4).

Pseudotime analysis

For the single-cell high-throughput dataset CNP0001102, we used the R package “monocle2” to perform cell trajectory analysis23, which inferred the cell development process or the evolution of cell subtypes based on the changes in gene expression levels of cell subpopulations over time. We built a pseudo-temporal object using the imported Seurat data, performed dimensionality reduction on cells using the DDRTree method, and then used the dimensionality reduction function to determine the type of cell differentiation state. Finally, we visualised the cell differentiation trajectory using the “plot cell trajectory” function.

Weighted gene co-expression network analysis

WGCNA identifies potential biomarker genes or therapeutic targets for IAV infection based on the internal connectivity of gene sets and the correlation between gene sets and phenotypes24,25. To determine the correlation between the age or degree of infection and the cellular immune characteristics of IAV infection, we used a system biology approach based on WGCNA to identify co-expression modules between differential genes in the GSE101702 dataset. We calculated the expression variance of each gene in the GSE101702 dataset and selected the top 75% of genes with median absolute deviation (MAD) ≥ 0.01. The R package WGCNA determined the optimal soft threshold value β for constructing the network by calculating each power’s scale-free topology fitting index to ensure that the scale-free topology fitting index was close to 0.9. The adjacency relationships were then transformed into topological overlaps to measure the network connectivity of a gene. We used hierarchical clustering to identify co-expression gene modules and integrated modules with a connection degree (feature factor) greater than 0.75 to identify the most relevant modules for IAV infection characteristics. We plotted heatmaps of the clustering modules according to the degree of infection and age of infection and provided the relationship truncation points and P values for each module.

Hub gene

Using a Venn diagram, we extracted the common genes between the DEGs in the microarray dataset GSE101702 and the modules in WGCNA that were significantly correlated with the severity of influenza (r > 0.7). We quantitatively represented the correlation between these genes and the characteristics of IAV infection using the module membership (MM) and gene importance (GS) indicators for each gene. MM > 0.8 and GS > 0.2 were used as the standard for selecting Hub genes26,27. We calculated Hub genes using four analysis algorithms: bottleneck, edge penetration component (EPC), maximum group centrality (MCC), and degree. Subsequently, we retained the top 20 genes computed by each algorithm and merged them, which constituted the final differentially expressed genes (DEGs) associated with the severity of IAV infection.

Friends analysis

Based on 33 differentially expressed genes (DEGs) that are related to the severity of IAV infection, the functional similarity between these hub genes was calculated using the R package “GOSemSim”28. The genes with the most important functions in the network were identified by constructing a gene interaction network, calculating the importance index of each node, and searching for the genes with the strongest functional similarity to other differentially expressed genes.

Quantification and statistical analysis

For differential expression analysis, the Student’s T-test was used. Meanwhile, one-way ANOVA was used to distinguish statistically significant differences between groups when comparing differences between multiple groups, such as gene expression in different IAV infection degrees. The Wilcoxon rank sum test was used to detect differences in immune infiltrating cells. P-values (or adjusted P-values) less than 0.05 were considered statistically significant. We used R software and related toolkits for all statistical analyses.

Result

IAV infection suppresses T cell activation

By analysing the microarray dataset GSE101702 of 107 IAV-infected samples (influenza group) and 52 healthy samples (healthy group), we identified 1899 genes that showed significant expression differences between the infected and healthy samples. Among them were 910 upregulated genes and 989 downregulated genes [Fig. 2A]. We performed GO enrichment analysis on the upregulated and downregulated genes to investigate the physiological activities or mechanisms involved in gene expression changes in the infected group. The enrichment network diagram showed that the upregulated genes ultimately pointed to physiological activities, including defence response to the viruses, defence response to symbiont innate immune response, and regulation of defence response [Fig. 2B]. This means that during IAV infection, the body constantly improves its anti-viral infection ability to resist virus invasion. The downregulated genes ultimately pointed to physiological processes, including antigen receptor-mediated signalling pathway and T cell activation [Fig. 2C]. This result indicates that T cell activation is inhibited as IAV infection progresses, which may further affect its effector function.

Fig. 2
figure 2

Differential analysis and immune infiltration analysis of GSE101702. A Up-regulated and down-regulated differential genes. B Up-regulated gene GO biological processes enrichment bubble plot. C Up-regulated gene GO biological processes enrichment network diagram. D Heatmap of correlation between immune cell and immune infiltration-related differential gene sets. E Boxplot of immune cell infiltration proportions for 22 immune cell types.

Full size image

To investigate T cells’ migration changes and impact on the progression of IAV infection in the body, we conducted an immune infiltration analysis on the GSE101720 data. Using the Cibersort algorithm, we focused on predicting 22 types of infiltrating immune cell types. The differentially expressed genes related to immune infiltration showed that most genes positively correlated with the degree of T cell infiltration, especially CD8(+) T cells. To investigate the degree of immune cell infiltration between the healthy group (healthy group) and the IAV infection group (influenza group), we used the Cibersort algorithm to evaluate the infiltration ratio of the 22 immune cell types in each sample (a total of 107 samples) [Fig. 2D]. The boxplot results show that the infiltration of CD8(+) T cells is significantly reduced in the influenza group, and the other T cell subpopulations also show a trend of reduced infiltration [Fig. 2E]. This means that the influence of T cells on the infected area gradually weakens during IAV infection.

T-cell profile during IAV infection

Based on the single-cell sequencing dataset CNP0001102, we clustered 13 immune cell populations [Fig. 3A] from the data and annotated the cells based on differential marker genes [Fig. 3B, C]. Based on the T cell subpopulations, we found that the Effector memory CD8 + T cell population had the most significant decrease in cell number (Ctrl group 2,744, Flu group 953). In contrast, the Regulatory T cell population significantly increased significantly (Ctrl group 61, Flu group 1,607). The Naive T cell population (Ctrl group 2,824, Flu group 2,545) and the Activated CD4 + T cell population (Ctrl group 1,325, Flu group 1,295) did not show significant changes in cell number. Finally, Proliferating CD4 + T cells increased slightly [Fig. 3D]. Then, in the differential gene enrichment analysis of each T cell population in the Flu group, we found that most of the differential genes in each T cell population were downregulated [Fig. 3E] and were functionally enriched in T cell activation, activation of immune response, and positive regulation of lymphocyte activation, etc. [Fig. 3F]. The differential genes in these physiological processes mostly consisted of downregulated genes. These data confirmed the conclusion drawn from the microarray data analysis that IAV infection inhibits T cell activation and more intuitively showed that the number of CD8 + T cells (Effector memory CD8 + T) decreased sharply during IAV infection, indicating that CD8 + T cells may weaken their effect function due to a double impact of number and activation. Meanwhile, the changes in the number of regulatory T cells and CD8(+) T cells in the IAV infection group may lead to an imbalance in their ratio, which may cause a change in the immune internal environment stability during IAV infection, thereby aggravating IAV infection.

Fig. 3
figure 3

Cell clustering and differential enrichment analysis of CNP0001102. A 13 types of immune cells UMAP clustering diagram. B Expression feature map for marker genes used for cell annotation. C Expression bubble chart for marker genes. D Bar chart showing the number and percentage of immune cell populations. E Bar chart showing the number of up-regulated and down-regulated genes in differential analysis. F GO biological processes enrichment bubble chart for 5 T cell clusters.

Full size image

6IAV infection regulates T-cell development

The limited activation of T cells in IAV infection may affect their developmental process and cell trajectory differentiation, so we used the monocle algorithm to perform a retrospective analysis based on the single-cell sequencing dataset CNP0001102 [Fig. 4A]. The results show that all immune cells evolve in the order of time nodes 2-3-4-1, and we focus on extracting the trajectory of T cells for interpretation [Fig. 4B]. Compared with the control group (Ctrl group), we found that the development of almost all types of T cells in the severe IAV infection group (Flu group) is regulated, which is manifested as most cells in the Ctrl group are concentrated in the time sequence after node 1. In contrast, only a few cells in the Flu group exist after node 1, and most cells are trapped in the time sequence between nodes 3 and 1 before node 1 [Fig. 4C]. Among them, the regulation effect is most obvious for Effector memory CD8 + T, whose retrospective process was almost trapped at node 3. In addition, the trajectory differentiation of Proliferating CD4 + T in the Flu group is similar to that in the Ctrl group, implying that this subpopulation is less affected by IAV infection in terms of cell development and differentiation. In summary, the above results suggest that T cell development is significantly regulated during IAV infection, which may be related to the change of T cell activation state, thereby causing the differentiation trajectory of T cells to differ from that in the Ctrl group. Additionally, the CD8(+) T cells (Effector memory CD8 + T cells) in the Flu group mostly remained at the 3 nodes or before, indicating that their differentiation and development were significantly limited, which may seriously affect the development of their effect functions.

Fig. 4
figure 4

CNP0001102 proposed temporal analysis. A The trajectory diagram of the proposed temporal analysis shows the temporal diagram (left) and the cell distribution map in the trajectory (right). B The cell distribution of T cell clusters in the proposed temporal trajectory. C The cell distribution of T cell clusters in the proposed temporal trajectory is presented separately for cell samples from the Ctrl group or the Flu group.

Full size image

Glutathione metabolism is an important pathway in IAV infection, regulating the progression of the infection

Based on the microarray dataset GSE101702, which includes patients with common IAV infections at different ages and severe infections, combined with the results of immune-related changes in the internal environment caused by IAV infection, we hypothesised whether this change is related to the severity of IAV infection or the age of the infected patient. Therefore, we used the WGCNA algorithm to construct a gene co-expression network and modules for the microarray dataset GSE101702, calculated the expression variance of each gene, and screened the top 75% of absolute median deviation (MAD) and MAD value ≥ 0.01 genes for further analysis. We chose β = 13 as the threshold for constructing a scale-free network [Fig. 5A]. We identified co-expression gene modules related to the severity of IAV infection and the age of the infected patient [Fig. 5B]. The heatmap of co-expression modules shows that the gene-to-gene correlation is stronger in the severe infection group than the non-severe infection group, and the age of the infected patient does not affect the gene-to-gene correlation [Fig. 5C]. These data suggest that as the severity of IAV infection increases, the connectivity between genes is stronger.

Fig. 5
figure 5

Use WGCNA to construct a gene co-expression network for GSE101702. A Scatter plot of power values. B Dendrogram of gene co-expression module clustering for 107 samples. C Heatmap of co-expression modules related to IAV infection severity and age. D Scatter plot of MEmaroon and MEdarkslateblue with IAV infection severity. E Bubble chart of GO biological processes and KEGG enrichment analysis of genes in MEmaroon and MEdarkslateblue.

Full size image

Next, we selected the modules (MEmaroon and MEdarkslateblue) with a correlation coefficient of > 0.7 with IAV infection severity [Fig. 5D]. We analysed their gene functions and KEGG pathways [Fig. 5E] to explore their related functions and mechanistic pathways. The gene functions in the MEmaroon module are concentrated in embryonic skeletal system development, substrate adhesion-dependent cell spreading, and the corresponding enriched pathways are Lysosome, Dilated cardiomyopathy, Hypertrophic cardiomyopathy, and Cardiac muscle contraction. The gene functions in the MEdarkslateblue module are concentrated in the leukotriene D4 biosynthetic process, glutathione catabolic process, acute inflammatory response, etc., and the corresponding enriched pathways are Glutathione metabolism and Bacterial invasion of epithelial cells. These results, particularly the enrichment of Glutathione metabolism, may help us explain the immune dysregulation observed in the analysis of the single-cell sequencing dataset CNP0001102, where there was a difference in the number of CD8(+) T(Effector memory CD8 + T) and Treg(Regulatory T) cells in the Flu group.

TXN is a target gene that is associated with T-cell proliferation after IAV infection

Next, we created a Venn diagram to compare the DEGs in the GSE101702 dataset with the genes in the WGCNA module significantly correlated with IAV severity (p < 0.05, r > 0.7). We selected 137 genes that were common to both sets [Fig. 6A]. Then, we identified the Hub genes among these 137 genes using four different algorithms and combined the results to obtain a final list of 33 Hub genes [Fig. 6B]. We further refined the selection of these Hub genes using the Friends analysis, ultimately retaining 10 key genes that play important roles in IAV disease progression, namely VNN1, TXN, S100A9, S100A8, S100A6, S100A12, MMP9, CSTA, CD63, and ARG1 [Fig. 6C]. Furthermore, the gene expression heatmap confirmed that these key genes were significantly upregulated in the IAV severe infection group [Fig. 6D]. Next, we attempted to identify the key genes related to T cells by clustering the single-cell sequencing dataset CNP0001102 into five T cell subpopulations and intersecting the differentially expressed genes in the GSE101702 dataset and the key modules in the WGCNA with each subpopulation. The results showed that only the Proliferating CD4 + T cell cluster shared genes with the Hub genes, namely TXN and S100A6 [Fig. 6E]. Therefore, we conducted expression analysis of these two genes based on the single-cell sequencing dataset CNP0001102. The violin plot shows that S100A6 is highly expressed in Naive T, Effector memory CD8 + T, Activated CD4 + T, Regulatory T cell, and Proliferating CD4 + T [Fig. 6F]. TXN, on the other hand, shows specific high expression in Proliferating CD4 + T and exhibits a significant downregulation in the Flu group in Proliferating CD4 + T [Fig. 6F]. In summary, the above data were screened through overlapping analysis of key genes, ultimately identifying TXN as a target gene related to T cell proliferation after IAV infection.

Fig. 6
figure 6

IAV Target Gene Mining. A A Venn diagram of the 137 genes shared by the DEGs in GSE101720 and the modules in WGCNA with a correlation of > 0.7 to IAV infection severity. B Four different algorithms were used to calculate the Hub genes and the results were combined to obtain 33 Hub genes. C The cloud-rain diagram of the Friends analysis, where the abscissa represents the FSIM value (the geometric mean of the gene based on the MF, CC, and BP ontologies of Gene Ontology) D Boxplot of gene expression levels of the 10 candidate target genes. E Construct a Venn diagram for the following three entities: the differentially expressed genes of GSE101702, the genes in the key modules (MEmaroon and MEdarkslateblue) within the Weighted Gene Co-expression Network Analysis (WGCNA) of GSE101702, and the differentially expressed genes in the Proliferating CD4 + T cluster of CNP0001102. F Gene expression violin plots of S100A6 in each T cell cluster. G Gene expression violin plots of TXN in each T cell cluster.

Full size image

Discussion

The precise dynamics of T cell subsets during IAV infection have yet to be fully understood. In this study, we used a multi-omics approach to explore the functional changes and important mechanisms involved in severe IAV infection in T cells and finally reported our findings. Our study has shown the phenomenon of massive apoptosis in Effector memory CD8 + T cells during severe IAV infection. However, the significant reduction in the proportion of Effector memory CD8 + T cells has not impacted the host’s ability to fight infection.

Another interesting finding was that the number and proportion of regulatory T cells (Regulatory T) were significantly increased in the group with severe IAV infection, consistent with the reported accumulation of Treg in the lungs of IAV-infected animals29. However, it is well known that Treg is beneficial for maintaining self-immune tolerance, which can repair tissue damage caused by excessive anti-viral response. However, the significant difference in the proportion of Effector memory CD8 + T and Regulatory T cells in the group with severe IAV infection raises questions about how the infected system balances the anti-viral state induced by Effector memory CD8 + T cells and the self-tolerance induced by Regulatory T cells. Or, how does the body maintain the effector state of active CD8(+) T cells in the case of a significant reduction in the number and proportion of Effector memory CD8 + T cells in severe IAV infection?

Our subsequent analysis provided an answer to this question. In the WGCNA analysis, we found that genes enriched in the modules closely associated with IAV infection were significantly enriched in Glutathione metabolism. Based on a literature survey, there have been reports that glutathione (GSH) has a role in maintaining Treg function30. Combining our data, we can hypothesise that the Glutathione metabolism pathway may be beneficial in assisting Treg in repairing excess tissue damage during severe IAV infection, helping to maintain the stability of the immune environment.

It is well known that activated T cells release reactive oxygen species (ROS). At the same time, increasing ROS levels can widely activate the innate immune system and further enhance T cell effect functions31. However, high concentrations of ROS can inhibit glyceraldehyde-3-phosphate dehydrogenase, impeding glycolysis32, and this glucose metabolism is crucial for T cell activation and proliferation33. Increasing the synthesis of acetyl-CoA through glucose metabolism drives the production of cellular energy and promotes cell proliferation34. Currently, the GSH antioxidant system activated by ROS plays a crucial role. The GSH antioxidant system initiates the metabolic reprogramming of activated T cells, helping T cells make up for their demand for acetyl-CoA and avoid T cell death-induced immune regulation disorder35. In summary, the GSH antioxidant system links activated T cells’ metabolic reprogramming with immune homeostasis regulation.

We finally selected and retained 10 core genes from the Hub gene by analysing Friends. Among them, VNN1 (vascular non-inflammatory molecule 1) encodes an extracellular enzyme with pantothenate hydrolase activity, and related studies have reported that the metabolic pathway catalysed by the VNN1 pantothenate hydrolase can effectively inhibit the progression of respiratory IAV infection36. S100A6, S100A8, S100A9, and S100A12 encode proteins that belong to the S100 calcium-binding protein family, which all have regulatory roles in inflammatory processes and immune responses37,38,39. Furthermore, there have been reports that extracellular S100A9 modulates the inflammatory response and disease progression during IAV infection by exaggerating pro-inflammatory responses, cell death, and viral pathogenicity40. MMP9 (Matrix metalloproteinase-9) is considered a therapeutic target for IAV infection, and related studies have reported that MMP-9 gene knockout mice exhibit more effective adaptive immune responses to IAV41. CSTA (Cystatin A) is a cysteine proteinase inhibitor family member, and reports have indicated that it enhances the caspase pathway to exert antitumor effects42. CD63 encodes a protein that belongs to the transmembrane tetraspan family, and there have been reports that CD63 maintains viral synapses, which may increase the possibility of HIV-1 infection success43. The ARG1 (Arginase 1) related metabolic pathway includes the hypertrophic pathway of L-citrulline metabolism and the innate immune system. Induction of its expression has been reported as a key feature of lung CD4 + T cells during mouse influenza infection, and in the conditional knockout of CD4 + T cells Arg1, the influenza virus is more effectively cleared44.

There are also reports that the Thioredoxin (TXN) system has a positive regulatory effect on lymphocyte proliferation. TXN in T cells can provide electrons to synthesise new deoxyribonucleotides, thereby accelerating cell proliferation45. This provides good scientific evidence for our Friends analysis based on HUB genes, confirming that TXN upregulation is crucial for the rapid expansion of T cells after IAV infection. Meanwhile, there are also studies linking TXN to GSH. GSH is the main antioxidant in mammals, and the TXN system is also a key antioxidant system. The report indicates that simultaneous inhibition of the GSH and TXN pathways is a well-established route to trigger cell death46. Our study also uncovers the GSH metabolic pathway and TXN, which may provide a new direction for studying T cell death in IAV infection.

Regarding another one of our findings, which is that IAV severe infection regulates T cell development, there are also relevant studies that serve as a counterpoint. Research has shown that the brief thymus atrophy caused by IAV infection in mouse experiments has different effects on the development dynamics of conventional T cells and Tregs47. However, in past studies and our study, the specific mechanisms by which IAV infection affects T-cell development have yet to be fully explored. This study confirms that IAV infection regulates T cell development, but we are still determining whether this regulation is positive or negative for T cells. More exploration is needed to elucidate this phenomenon regarding T cell development in IAV infection.

Our study conducted a multi-omics analysis to identify TXN as a key gene related to T cells in severe IAV infection groups and found that its expression was downregulated in proliferative T cells in infected groups. Meanwhile, we identified Glutathione metabolism as playing an important role in IAV infection through WGCNA, and the combination of this key gene TXN may help advance research on non-apoptotic forms of T cell death in IAV infection.