Single-cell sequencing reveals immune landscape of tumor-infiltrating lymphocytes (TILs) during non-small cell lung cancer (NSCLC) progression

Single-cell sequencing reveals immune landscape of tumor-infiltrating lymphocytes (TILs) during non-small cell lung cancer (NSCLC) progression | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Single-cell sequencing reveals immune landscape of tumor-infiltrating lymphocytes (TILs) during non-small cell lung cancer (NSCLC) progression yue li, Jinguo Liu, Hua Zhang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3879125/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 05 May, 2025 Read the published version in Genes & Immunity → Version 1 posted 9 You are reading this latest preprint version Abstract During the process of NSCLC using TILs therapy, the heterogeneity of immune cell was revealed by using combined single-cell RNA (scRNA)/ T cell receptor (scTCR) sequencing -seq data from lung adenocarcinoma (LUAD) patients. Naïve CD4 + T was increased in tumor tissue compared with circulating blood samples, activated signaling pathways were recognized, and GZMA was identified as a potential novel diagnostic biomarker. The scTCR-seq repertoire was also investigated. At transition state, macrophages ( FTL ) and dendritic ( AIF1 ) cells transferred the most CD3 TCR clones to T ( IL7R ) cells, and cytotoxicity ( NKG7 ) transported to terminal exhausted ( CCL5 ) CD8 + T cells. At transition and expansion state, T helper ( CXCL13 ) transported the most CD3 TCR clones to regulatory T ( FOXP3 ) cells. The expression profiling of cytokines, checkpoint receptors and their ligands during tumor progression were also investigated. T helper ( FTL, TNFRSF4 and TIGIT ) and regulatory T ( CTLA4, TIGIT and FTL ) show up at the initial stage of normal and metastatic samples, while cytotoxicity ( FGFBP2 , NKG7, PRF1 and CCL5 ) CD8 + T cells still appears at the final stage of normal and metastatic samples. Taken together, our study provides the single cell level of TILs in NSCLC and offers treatment strategies to overcome drug resistance. Biological sciences/Immunology/Tumour immunology Biological sciences/Genetics/Sequencing/Next-generation sequencing Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction Lung cancer is still the most predominant cancer in the world and the leading cause of cancer-related death. NSCLC including LUAD is the most common lung cancer type, accounting for nearly 85% of all patients 1 . Adoptive cell therapy (ACT) with TILs is a highly personalized immunotherapy for cancer 2 . TILs is a highly personalized immunotherapy for cancer 2 . TILs usually exist in the tumor stroma, mainly composed of CD3 + , CD4 + , CD8 + T lymphocytes 3 . In some cancers, a high density of TILs in tumor tissue is closely related to a good prognosis 4 . TILs include many phenotypic and functional heterogeneity subgroups 5 . Therefore, it is essential to characterize the tumor microenvironment of NSCLC. A recent study has revealed that a dysfunctional CD8 + TILs subset displays markers of end-stage differentiation 6 . In addition, this CD8 + TILs subgroup is accumulated in patients with advanced NSCLC, and its high abundance is related to the poor clinical response of immunotherapy 6 . Biomarkers of CD8 + T cell exhaustion were found to be related to better therapeutic effects 7 . By using scRNA-seq data, some potential prognostic biomarkers were identified, their expression levels were related to the survival time of patients, and they were lower expressed in exhausted CD4/8 + T than in CD8 + T cells. Combining scRNA-seq with scTCR-seq can effectively identify the common cloned cells between TILs and patients' peripheral blood T cells, to characterize their relationship 8 . It allows simultaneous analysis of paired TCR sequences and transcriptome to track different types of cell clones 9 . The existence of TILs in tumor can be proved by sharing the cloning and expression of TCR 8 . Here, we focused on different kinds of cells, CD4 + and CD8 + T cells in the stages of state transition, cross-tissue migration and clonal expansion, as well as their biomarkers. Since TILs contain a variety of cells, developmental processes of these cells from LUAD patients spanning different progression stages are still unclear. We would like to understand the differentiation trajectory of these cells during the process of cancer development, such as non-tumour-involved lung, primary lung adenocarcinomas and lung adenocarcinoma metastases 10 . scRNA-seq is a powerful method to reconstruct the trajectory of cell differentiation and dynamic changes of gene expression 10 . Here, we characterized the dynamics and diversity of single cell profiling across different stages of NSCLC, which contribute to elucidating host adaptive immunity and discovering novel therapeutic targets. Materials and Methods Data availability statement A comprehensive review of the literature for next generation sequencing (NGS) studies involving TILs in NSCLC yielded three recent publications: 1. Gueguen et al., referred to as “scRNA-seq and scTCR-seq data of resident and circulating precursors to tumor-infiltrating CD8 + T cell populations in lung cancer”, 2. Laughney et al., based on the “scRNA-seq transcriptional landscape of primary tumors and metastases human LUAD”, 3. Ganesan et al., referred to as “Analysis of purified populations of CD8 T cells (isolated from primary lung tumors and matched adjacent lung tissue of lung cancer patients) at the transcriptomic level by RNA sequencing” 11 – 13 . Each raw dataset was downloaded from the Gene Expression Omnibus (GEO) at the National Center for Biotechnology Information (NCBI). Raw dataset was downloaded from the Gene Expression Omnibus. Accession numbers are: https://www.ncbi.nlm.nih.gov/gds/?term=gse162498 , https://www.ncbi.nlm.nih.gov/gds/?term=gse162499 , https://www.ncbi.nlm.nih.gov/gds/?term=gse90728 , https://www.ncbi.nlm.nih.gov/gds/?term=gse123902 . Online public database The box-and-whisker plots of prognostic biomarkers in tumor and normal samples were provided by Gene Expression Profiling Interactive Analysis (RRID:SCR_018294) ( http://gepia.cancer-pku.cn/detail.php ) 13 . GEPIA used data from The Cancer Genome Atlas (TCGA) and Genotype-Tissue Expression (GTEx) 13 . |Log 2 FoldChange| > 1 and P-value < 0.01 were considered statistical significance. Genes based on individual LUAD stages were obtained from http://ualcan.path.uab.edu/analysis.html . Kaplan-Meier survival curves were retrieved from http://kmplot.com/analysis/index.php?p=service&cancer=pancancer_rnaseq . CellPhoneDB (RRID:SCR_017054) 14 is a public database of receptor–ligand interactions. Here, https://www.cellphonedb.org/ was utilized to explore the crosstalk of cell subtypes in NSCLC. The correlation intensions between specified cell types were shown as the total mean and the number of interactions. Correlations of GZMA expression with tumor purity and with the infiltration level of immune cell are estimated by TIMER2.0 ( http://timer.cistrome.org/ ) in LUAD 15 . Characterization of prognostic biomarkers in single cells Seurat (version 4.3.0) R package (version 4.1.1) was used for analysis of scRNA-seq https://www.ncbi.nlm.nih.gov/gds/?term=gse162498 and https://www.ncbi.nlm.nih.gov/gds/?term=gse123902 16 . Data was normalized with NormalizeData function. Feature counts of each cell were divided by the total counts for that cell multiplied by a scaler factor (1e4), then natural-log transformed. The normalized data were then integrated for UMAP clustering. After quality control and filtering, we performed an integrated analysis and identified the subset of clusters. FindMarkers (min.pct = 0.25) was used to find cell markers of each cluster. Clusters were annotated based on canonical cell markers ( http://xteam.xbio.top/CellMarker/ ) 17 . Uniform manifold approximation and projection (UMAP) was generated by using DimPlot . DotPlot was used visualize how feature expression changes across different identity clusters. Constructing the developmental trajectory of scRNA-seq CytoTRACE (Cellular Trajectory Reconstruction Analysis using gene Counts and Expression) was used to infer the cell differentiation trajectories for malignant cells 10 . CytoTRACE scores range from 0 to 1, while higher scores indicate higher stemness (less differentiation) and vice versa 18 . plotCytoTRACE was used to generate the predicted ordered of individual cells. Genes correlated to CytoTRACE were created by plotCytoGene. Without knowing the differentiation time or direction in advance, Monocle3 package was used to analyze the developmental trajectory 19 . learn_graph was applied to learn the trajectory graph. order_cells calculated where each cell falls in pseudotime. The function plot_genes_in_pseudotime takes a small group of genes and displays their dynamics as a function of pseudotime. scTCR-seq data analysis scRepertoire was used to merge and visualize scTCR-seq data with Seurat object of scRNA-seq data 20 . The percentage of unique clones in the bar graph across different samples was generated by quantContig function. UMAPs of clonotype in https://www.ncbi.nlm.nih.gov/gds/?term=gse162498 clusters were performed by DimPlot . Basic analysis visualizations the relative usage of genes of the TCR is used by vizGenes . Compare the clonotype diversity (Shannon and Inverse (Inv) Simpson indexes) among different cell types were handled by clonalDiversity . Counts of different clonotype cells in clusters were calculated by clonalOverlap . For looking at clonotypes by cellular origins and cluster identification, StartracDiversity is used. Network interaction of clonotypes shared between clusters along the single-cell dimensional reduction is used by clonalNetwork . Clonotypes across multiple categories were created by alluvialClonotypes . immunarch was utilized to compare the degree of clonal expansion in the repertoire of each sample 19 . Clonotype tracking in different samples was utilized by trackClonotypes. Motif of clonotype is constructed by getKmers. Software used GraphPad Prism software (GraphPad Software, Inc.) was employed to draw the mRNA expression levels of prognostic biomarkers in different patient groups and cell types of https://www.ncbi.nlm.nih.gov/gds/?term=gse123902 . Genes in https://www.ncbi.nlm.nih.gov/gds/?term=gse90728 correlated with GZMA were calculated by SigmaPlot (RRID:SCR_003210). All deep-sequencing data were analyzed in Bioconductor (RRID:SCR_006442) version 3.14 (BiocManager 1.30.19), R 4.1 1 ( R Core Team, 2021; http://www.R-project.org/ ) under Ubuntu environment (20.04). Values of *p < 0.05, **p < 0.01, ***p < 0.001 , and ****p < 0.0001 were considered significant. All codes are available at https://github.com/yueli8/TILs_NSCLC . Results High-resolution landscape of NSCLC by scRNA-seq The brief workflow is shown in Fig. 1A. Four NSCLC tumor tissue and matched circulating blood samples were used in our study. Nine types of cells were identified by UMAP clustering analysis (Fig. 1B). Most of the cells in TILs are T cells. Ten types of T cells were identified. The most abundant T cells are natural killer T and effector memory CD4 + T cells. Graph-based clusters were manually annotated based on known marker genes for the main expected cell types (Fig. 1C). The proportion of each T cell lineage varies greatly in different tumors and blood samples. Compared with blood samples, naïve CD4 + and effector memory CD8 + T cells are significantly increased in tumor, while resident memory CD8 + and regulatory T cells are decreased (Fig. 1D). Natural killer T cells contains the largest number of differentially expressed genes (DEGs). Volcano plots and Kyoto encyclopedia of genes and genomes (KEGG) pathways of the DEGs in natural killer T and naïve CD4 + T cells were shown in Fig. 1E and F. Some pathways have significant systemic anti-tumor effects in immunotherapy, such as: MARK, p53, NF-kappa B and TNF signaling. The cell differentiation pathways of Th1, Th2 and Th17 were also found, which drove CD4 + T naive cells into diverse T helper subsets. Identification of potential prognostic biomarkers of TILs in NSCLC Next, by using Kaplan-Meier plots, DEGs in T cells were used to further study their overall survival (OS) and recurrence-free survival (RFS). Best cutoff analysis showed that the low expressed of CD8A, CD8B, CD38, CD69, ENTPD1, GZMA, GZMH, MYO1F, SYNE1, TSC22D3 and XCL2 were associated with the poor OS and RFS of LUAD patients (Fig. 2A). Some of them have higher Moran’s I values, and they may largely represent markers of different T cells (Fig. 2B). We consistently observed substantially higher Moran’s I values of them and revealed that they may largely represented markers of different T cells. CD69 and TSC22D3 have peaks in exhausted CD8 + T cells, GZMA, GZMH and SYNE1 in resident memory CD8 + T cells, while CD8A and CD8B increased in effector memory CD8 + T cells. CD69 , GZMA and TSC22D3 were highly expressed in T helper cells, but low expressed in naïve CD4 + T cells. In hierarchically-clustered heatmaps, compared with other types of T cells, CD8A, GZMA, CD8B and GZMH are highly expressed in effector memory CD8 + and naïve CD4 + T cells (Fig. 2C). Furthermore, box-and-whisker plots of these genes were shown. CD69, MYO1F, SYNE1 , and TSC22D3 were significantly higher ( P-value < 0.01 ) expressed in tumor tissue than in normal samples (Fig. 2D). The expression levels of these genes spanning different stages are retrieved (Fig. 2E). They were lower in the late/advanced stage of LUAD patients, but higher in early/primary stage (except CD38 ). All these indicate that the above genes could be used as potential prognostic biomarkers of TILs in NSCLC. Identification of lower expressed genes in exhausted CD4/8 + T than in CD8 + T cells Next, we used We used the GSE123902 dataset which contains transcriptionally profiling of single cells from patients spanning different stages of LUAD progression to discover genes which are lower expressed in exhausted CD4/8 + T than in naive CD8 + T cells. First, ten types of cells were identified by using UMAP (Fig. 3A). By using hierarchical clustering of heatmaps (Fig. 3B). CD8A, CD8B, CD69, GZMA, GZMH and XCL2 were higher expressed in naive CD8 + and CD8 + T cells than other types of cells. Next, afore-mentioned genes across different stages of LUAD progression were studied. All of them are statistically lower (P-value < 0.05) expressed in metastatic than in normal or primary tumor patients (Fig. 3C). In addition, CD8A, CD8B, CD69, GZMA, GZMH and XCL2 were significantly lower ( P-value < 0.0001 ) expressed in exhausted CD4/8 + than in naive CD8 + or CD8 + T cells (Fig. 3D). Among all these genes, GZMA has the highest expression (about 1.2) in CD8 + T cells, therefore GZMA could be a potential biomarker for the prognosis of patients with LUAD. GZMA was also correlated with some genes by using bulk RNA-seq GSE90728 (Fig. 3E). It was positively correlated with cyclins, such as: CCL5, CD3D, CCL3, CD2, CCR5, CD3G, CD27 , checkpoint inhibitor: LAG3 and PDCD1 , chemokine: CXCR6 , natural killer cell: ID2 , cytokine: IFNG, GZMK, LF2 , cell phenotype: CCR5 and CD27 , and negatively correlated with CD68 and IL7R . These genes may have anti-tumor immune functions related to GZMA . Therefore, GZMA can be used as a novel prognostic marker of TILs in LUAD. Finally, the correlation between immune cell infiltration and GZMA expression was evaluated by TIMER2.0 15 . As shown in the scatter plots (Fig. 3F), GZMA negatively correlates with tumor purity (correlation = -0.427, P = 2.67e-23), suggesting that the main source of GZMA expression detected are stromal and immune cells. Also, GZMA was positively correlated with immune infiltration of CD8 + , CD4 + , regulatory, gamma delta T, myeloid dendritic cell, macrophage, and B cell. These findings further revealed a strong relationship between GZMA and immunosuppression in the inflammatory tumor microenvironment. scTCR-seq repertoire of TILs in NSCLC To investigate the diversity and dynamics of T cell repertoire of TILs in NSCLC, a combination of scRNA-seq and scTCR- seq data from GSE162500 was used 11 . Firstly, the number of unique clonotypes pre-patient was studied (Fig. 4A). Except for patient 58, the percentage of unique clonotypes in circulating blood was higher than that in tumor tissue. These findings indicated that circulating blood has more unique clonotypes than tumor tissue. In fact, by tracking the single clones of all patients, we revealed the top 25 most amplified cell clones, and these clones had significant dynamic changes in patients (Fig. 4B). Except for patient 58, all these clones had more significant amplification in tumor tissue than in circulating blood. Corresponding to Fig. 1b, the top eight most amplified cell clones were concentrated in terminally exhausted CD8 + T cells, with the first one referring to the amino acid sequence: CAASRNAGNMLTF_CASSISGTGEIGEAFF (Fig. 4C). There are more terminally exhausted CD8 + and regulatory T cells clones in tumor tissue than in circulating blood samples, and most of them have been expanded in large and medium scales. Top ten uses of TCR gene are shown in Fig. 4D. A putative binding motif for CDR3 is predominantly composed of 5-residue peptides. The first two peptides are SSSSG and GGGGF, with polar residues S and G and hydrophobic residues F and A. Interestingly, highest diversity indices were observed in dendritic cells and macrophages, while natural killer and T were the lowest (Fig. 4E). To calculate the degree of TCR repertoire overlap among various cell types, we used the overlap coefficient method. There is substantial degree of 36.4% TCR clones overlap between dendritic cell and T cells, between macrophages and T cell populations (Fig. 4F). The existence of clonal cells spanning several different tissue sites suggests the migration, transformation and expansion of the specified cell types. As expected, among all the cells, macrophages ( FTL ) and natural killer ( GNLY ) showed the highest clonal transition, while T ( IL7R ) cells showed the highest clonal expansion index (Fig. 4G and H). Moreover, macrophages ( FTL ) and dendritic ( AIF1 ) cells transfer the most CD3 TCR clones to T ( IL7R ) cells. These results implicate a potential role of different immune cells in shaping the NSCLC during T cell infiltration scTCR-seq repertoire of CD4 + and CD8 + T cells in NSCLC Next, we investigated the diversity and dynamics of CD4 + and CD8 + T cells repertoire in NSCLC. Figure 5A shows the UMAP, developmental trajectory and clonalNetwork of CD4 + T cells. T helper appeared at first, followed by effector memory or regulatory T cells. It is worth noting that naive cells will still appear, and mainly accumulate at the final stage of progression. Effector memory cells receive most clones from regulatory and naïve cells. From alluvialClonotypes in Fig. 5B, effector memory shows the high proportion. There are 13.5% clone overlapping between regulatory and T helper cells (Fig. 5C). Interestingly, highest diversity indices were observed in naïve and effector memory cells, while T helper cells was the lowest (Fig. 5D). Besides, T helper showed the highest clonal transition and expansion index (Fig. 5E). In CD8 + T cells, cytotoxic appeared at first, followed by exhausted or memory cells. Terminally exhausted cells receive most clones from pre-exhausted and resident memory cells (Fig. 5F). The proportion of terminally exhausted cells is the highest (Fig. 5G). There is 13.5% clone overlapping between cytotoxic and terminally exhausted cells (Fig. 5H). Higher diversity indices were observed in effector memory and pre-exhausted cells, while terminally exhausted was the lowest (Fig. 5I). Also, cytotoxic and resident memory showed the highest clonal transition index, while terminally exhausted cells showed the highest clonal expansion (Fig. 5J). Collectively, in CD4 + T cells, T helper ( CXCL13 ) and regulatory T ( FOXP3 ) cells have higher ability of transition and expansion. They also transfer the largest number of CD3 TCR clones to each other. In CD8 + T cells, cytotoxic ( NKG7 ) and resident memory ( GZMK ) have higher ability of transition, and terminally exhausted ( CCL5 ) have higher expansion ability. Cytotoxic ( NKG7 ) transports the most CD3 TCR clones to terminally exhausted ( CCL5 ) cells (Fig. 5K). Developmental trajectories of NSCLC during tumor progression To further understand the different cell states in NSCLC spanning different stages, the developmental trajectories of tumor cells was studied. Figure 6A shows the human tissue samples in GSE123902: adjacent non-tumor involved lung (n = 4, ‘normal’), primary LUAD (n = 8, ‘primary tumor’), and metastatic tumor of LUAD (n = 5, ‘metastasis’). UMAP of different types of cells is shown in Fig. 6B. Naïve CD4 + T cells only appeared in normal samples, while regulatory T cells in metastasis. Pre-exhausted CD8 + T cells showed up in primary tumor, while cytotoxic CD8 + T cells in metastasis. The predicted ordering of different cell types was displayed in box plots (Fig. 6C). In normal samples, macrophages first appeared, followed by CD4 + , CD8 + T cells. In primary tumor, progenitor and epithelial developed at first, then is T helper, effector CD8 + , CD4 + , pre-exhausted CD8 + T cells. Furthermore, cytotoxic CD8 + T, effector memory and recently activation CD4 + T still exist at the late stage of metastasis. Additionally, the top 10 (less differentiated; red) and bottom 10 (most differentiated; blue) genes in this dataset based on their correlation with CytoTRACE can be predicted (Fig. 6D). Some of them are significantly related to the specific states in different types of cells in pseudotime trajectory (Fig. 6E). In normal samples, FGFBP2 , NKG7 and PRF1 are associated with cytotoxic CD8 + T cells; while FTH1 , FTL and HLA-DRA are correlated with macrophages. In primary tumor samples, BTG1, CCL5, CXCR4, FTL, IL7R, and TMSB10 are highly expressed during the process of development. In metastasis samples, FTH1 and FTL were related to either CD4 + or CD8 + T cells The expression profiling of cytokines, checkpoint receptors and their ligands during tumor progression Cytokine therapy helps the immune system stop the growth of cancer cells or kill them 21 . To address the expression profiling of cytokines receptors and their ligands in different samples, 26 kinds of cytokines were selected (Fig. 7A). Most of them are highly expressed in primary tumors, while CCL5 and CCL4 are highly expressed in cytotoxic CD8 + T cells of metastasis samples. Cytokine receptors are highly expressed in macrophages of normal and primary tumor samples, while CXCR4 is highly expressed in CD4 + T cells in metastasis samples. It is generally believed that cancer cells are the only source of checkpoint ligands and are responsible for inhibiting T-cell immune responses 22 . Checkpoint receptors are highly expressed in chronic activation CD4 + T cells of normal samples, T helper of primary tumor samples, and regulatory T of metastasis samples (Fig. 7B). Next, the transcript levels of some ligands and receptor in different samples were also studied (Fig. 7C). Notably, in primary tumor samples, FTL was highly expressed in T helper cells, and IFNG was in pre-exhausted CD8 + T cells. In normal and metastatic samples, CCL5 was highly expressed in cytotoxic CD8 + T cells. TNFRSF4 , TIGIT and CTLA4 were highly expressed in regulatory T cells of metastasis samples. By integrating the data of single cell transcriptome of different samples, the development trajectory of CD8 + T cells is shown below (Fig. 7D). In normal samples, it is from effector memory, then is progenitor and cytotoxicity ( FGFBP2 , NKG7 , PRF1 and CCL5 ). In primary tumor, it is from effector ( CCL5 ) to pre-exhausted ( CCL5 and IFNG ). Cytotoxicity ( CCL5 ) still appeared in the metastatic samples. In CD4 + T cells, the development trajectory is as follows. In normal samples, it starts from chronic activation, T helper, and then naïve. In primary tumor, it is from T helper ( TNFRSF4 , TIGIT and FTL ) to effector. In metastatic samples, it begins with regulatory T ( CTLA4 , TIGIT and FTL ) to effector memory, and then recently activation. Discussion Here, we investigated the diversity of TILs during NSCLC progression. In TILs of NSCLC, many pathways are concentrated on cancer-related. The MAPK pathway leads to uncontrolled growths. IL17 is a key cytokine which induces inflammation and activates several downstream signal pathways 23 . Th1 cells mediate a cellular immune response, Th2 cells strengthen humoral response, and Th17 cells produce proinflammatory cytokine 24 . Here, the specific expression of GZMA was identified as potential new marker of TILs. GZMA is an enzyme existing in cytotoxic T lymphocyte granules, and it is considered as a major proapoptotic protease 25 . During pneumococcal pneumonia, it defenses through a mechanism that does not depend on natural killer cells. Its related pathways including IL-9, IL-15 signaling, as well as their main biological effects on different types of immune cell. Macrophage ( FTL ) and natural killer ( GNLY ) cells exhibited high clonal transition, macrophages ( FTL ) and dendritic ( AIF1 ) transfer the most CD3 TCR clones to T ( IL7R ) cells. In many cancers, FTL was significantly positively correlated with tumor infiltration by tumor-associated macrophages 26 . Granulysin ( GNLY ) exists in human cytotoxic T lymphocytes and natural killer cells 27 . AIF1 guides hematopoietic and immune responses within dendritic cells and macrophages 28 . IL7R plays a key role in the development of lymphocytes. In CD4 + T cells, T helper ( CXCL13 ) and regulatory T ( FOXP3 ) cells transfer the largest number of CD3 TCR clones to each other and have higher ability of transition and expansion. FOXP3 seems to play a role in the development and function of regulatory T cells 29 – 31 . In CD8 + T cells, cytotoxic ( NKG7 ) transfers the most CD3 TCR clones to terminally exhausted ( CCL5 ) CD8 + T cells. NKG7 was identified as a necessary component for the cytotoxic function of CD8 + T cells and it is a T-cell–intrinsic therapeutic target for enhancing cancer immunotherapy 32 . CCL5 was highly expressed in LAG3 + exhausted T cells in hepatocellular carcinoma 33 . The development trajectory of CD8 + T cells in normal samples is cytotoxicity ( FGFBP2 , NKG7 , PRF1 and CCL5 ) at the final stage. FGFBP2 protein is secreted by cytotoxic T-lymphocytes and combined with fibroblast growth factor 34 . PRF1 is thought to act by creating holes in the plasma membrane which triggers an influx of calcium and initiates membrane repair mechanisms. In primary tumor samples, pre-exhausted ( IFNG and CCL5 ) appeared at last. It may serve as better targets for immunotherapies compared to the exhausted cluster 35 . The development trajectory of CD4 + T cells in primary tumor samples is T helper ( FTL, TNFRSF4 and TIGIT ) appeared at first. TNFRSF4 has a critical role in the maintenance of an immune response due to its ability to enhance survival. TIGIT antagonist also activated T follicular helper-like cells and dendritic cells, and reduced markers of immunosuppression in regulatory T cells 36 . The regulatory T cells (Tregs or Treg cells), maintain tolerance to self-antigens, and prevent autoimmune disease 37 . Tregs in the tumor microenvironment are abundantly effector Tregs which over-express immunosuppressive molecules, such as CTLA-4 38 . Anti-CTLA-4 antibodies lead to the depletion of regulatory T cells and thus increasing the effective anti-tumor effect of CD8 + T cells 38 . In conclusion, our work may serve as potential predictive markers for chemotherapy and long-term survival prognosis, as well as new therapeutic targets or strategies to overcome drug resistance and immunosuppression. Declarations Acknowledgments We deeply thank Jie Gao for scientific discussions. Declaration of Interest Statement The authors declare that they have no competing interests. References Chen, G., Hu, J., Huang, Z., Yang, L. & Chen, M. MicroRNA-1976 functions as a tumor suppressor and serves as a prognostic indicator in non-small cell lung cancer by directly targeting PLCE1. Biochem Biophys Res Commun 473, 1144–1151, doi: 10.1016/j.bbrc.2016.04.030 (2016). Granhoj, J. S. et al. Tumor-infiltrating lymphocytes for adoptive cell therapy: recent advances, challenges, and future directions. Expert Opin Biol Ther 22, 627–641, doi: 10.1080/14712598.2022.2064711 (2022). Kasper, H. U., Drebber, U., Stippel, D. L., Dienes, H. P. & Gillessen, A. Liver tumor infiltrating lymphocytes: comparison of hepatocellular and cholangiolar carcinoma. World J Gastroenterol 15, 5053–5057, doi: 10.3748/wjg.15.5053 (2009). Chen, J. et al. CD8 + tumor-infiltrating lymphocytes as a novel prognostic biomarker in lung sarcomatoid carcinoma, a rare subtype of lung cancer. Cancer Manag Res 10, 3505–3511, doi: 10.2147/CMAR.S169074 (2018). Lee, S. W. et al. CD8(+) TILs in NSCLC differentiate into TEMRA via a bifurcated trajectory: deciphering immunogenicity of tumor antigens. J Immunother Cancer 9, doi: 10.1136/jitc-2021-002709 (2021). Sanmamed, M. F. et al. A Burned-Out CD8(+) T-cell Subset Expands in the Tumor Microenvironment and Curbs Cancer Immunotherapy. Cancer Discov 11, 1700–1715, doi: 10.1158/2159-8290.CD-20-0962 (2021). Hsu, C. L. et al. Exploring Markers of Exhausted CD8 T Cells to Predict Response to Immune Checkpoint Inhibitor Therapy for Hepatocellular Carcinoma. Liver Cancer 10, 346–359, doi: 10.1159/000515305 (2021). Gohil, S. H., Iorgulescu, J. B., Braun, D. A., Keskin, D. B. & Livak, K. J. Applying high-dimensional single-cell technologies to the analysis of cancer immunotherapy. Nat Rev Clin Oncol 18, 244–256, doi: 10.1038/s41571-020-00449-x (2021). Yang, H. Q., Wang, Y. S., Zhai, K. & Tong, Z. H. Single-Cell TCR Sequencing Reveals the Dynamics of T Cell Repertoire Profiling During Pneumocystis Infection. Front Microbiol 12, 637500, doi: 10.3389/fmicb.2021.637500 (2021). Gulati, G. S. et al. Single-cell transcriptional diversity is a hallmark of developmental potential. Science 367, 405–411, doi: 10.1126/science.aax0249 (2020). Gueguen, P. et al. Contribution of resident and circulating precursors to tumor-infiltrating CD8(+) T cell populations in lung cancer. Sci Immunol 6, doi: 10.1126/sciimmunol.abd5778 (2021). Laughney, A. M. et al. Regenerative lineages and immune-mediated pruning in lung cancer metastasis. Nat Med 26, 259–269, doi: 10.1038/s41591-019-0750-6 (2020). Ganesan, A. P. et al. Tissue-resident memory features are linked to the magnitude of cytotoxic T cell responses in human lung cancer. Nat Immunol 18, 940–950, doi: 10.1038/ni.3775 (2017). Efremova, M., Vento-Tormo, M., Teichmann, S. A. & Vento-Tormo, R. CellPhoneDB: inferring cell-cell communication from combined expression of multi-subunit ligand-receptor complexes. Nat Protoc 15, 1484–1506, doi: 10.1038/s41596-020-0292-x (2020). Li, T. et al. TIMER2.0 for analysis of tumor-infiltrating immune cells. Nucleic Acids Res 48, W509-W514, doi: 10.1093/nar/gkaa407 (2020). Stuart, T. et al. Comprehensive Integration of Single-Cell Data. Cell 177, 1888–1902 e1821, doi: 10.1016/j.cell.2019.05.031 (2019). Zhang, X. et al. CellMarker: a manually curated resource of cell markers in human and mouse. Nucleic Acids Res 47, D721-D728, doi: 10.1093/nar/gky900 (2019). Zhang, Z. et al. Integrated analysis of single-cell and bulk RNA sequencing data reveals a pan-cancer stemness signature predicting immunotherapy response. Genome Med 14, 45, doi: 10.1186/s13073-022-01050-w (2022). Cao, J. et al. The single-cell transcriptional landscape of mammalian organogenesis. Nature 566, 496–502, doi: 10.1038/s41586-019-0969-x (2019). Borcherding, N., Bormann, N. L. & Kraus, G. scRepertoire: An R-based toolkit for single-cell immune receptor analysis. F1000Res 9, 47, doi: 10.12688/f1000research.22139.2 (2020). Essogmo, F. E. et al. Cytokine Profile in Lung Cancer Patients: Anti-Tumor and Oncogenic Cytokines. Cancers (Basel) 15, doi: 10.3390/cancers15225383 (2023). Yeo, A. T. et al. Single-cell RNA sequencing reveals evolution of immune landscape during glioblastoma progression. Nat Immunol 23, 971–984, doi: 10.1038/s41590-022-01215-0 (2022). Saito, S., Nakashima, A., Shima, T. & Ito, M. Th1/Th2/Th17 and regulatory T-cell paradigm in pregnancy. Am J Reprod Immunol 63, 601–610, doi: 10.1111/j.1600-0897.2010.00852.x (2010). Kaiko, G. E., Horvat, J. C., Beagley, K. W. & Hansbro, P. M. Immunological decision-making: how does the immune system decide to mount a helper T-cell response? Immunology 123, 326–338, doi: 10.1111/j.1365-2567.2007.02719.x (2008). Metkar, S. S. et al. Human and mouse granzyme A induce a proinflammatory cytokine response. Immunity 29, 720–733, doi: 10.1016/j.immuni.2008.08.014 (2008). Hu, Z. W. et al. Comprehensive analysis of ferritin subunits expression and positive correlations with tumor-associated macrophages and T regulatory cells infiltration in most solid tumors. Aging (Albany NY) 13, 11491–11506, doi: 10.18632/aging.202841 (2021). Tewary, P. et al. Granulysin activates antigen-presenting cells through TLR4 and acts as an immune alarmin. Blood 116, 3465–3474, doi: 10.1182/blood-2010-03-273953 (2010). Elizondo, D. M., Brandy, N. Z., da Silva, R. L., de Moura, T. R. & Lipscomb, M. W. Allograft inflammatory factor-1 in myeloid cells drives autoimmunity in type 1 diabetes. JCI Insight 5, doi: 10.1172/jci.insight.136092 (2020). Hori, S., Nomura, T. & Sakaguchi, S. Control of regulatory T cell development by the transcription factor Foxp3. Science 299, 1057–1061, doi: 10.1126/science.1079490 (2003). Fontenot, J. D. et al. Regulatory T cell lineage specification by the forkhead transcription factor foxp3. Immunity 22, 329–341, doi: 10.1016/j.immuni.2005.01.016 (2005). Fontenot, J. D., Gavin, M. A. & Rudensky, A. Y. Foxp3 programs the development and function of CD4 + CD25 + regulatory T cells. Nat Immunol 4, 330–336, doi: 10.1038/ni904 (2003). Wen, T. et al. NKG7 Is a T-cell-Intrinsic Therapeutic Target for Improving Antitumor Cytotoxicity and Cancer Immunotherapy. Cancer Immunol Res 10, 162–181, doi: 10.1158/2326-6066.CIR-21-0539 (2022). Zheng, C. et al. Landscape of Infiltrating T Cells in Liver Cancer Revealed by Single-Cell Sequencing. Cell 169, 1342–1356 e1316, doi: 10.1016/j.cell.2017.05.035 (2017). Newman, J. H. et al. IgG4-related disease: Association with a rare gene variant expressed in cytotoxic T cells. Mol Genet Genomic Med 7, e686, doi: 10.1002/mgg3.686 (2019). Zheng, Y. et al. Immune suppressive landscape in the human esophageal squamous cell carcinoma microenvironment. Nat Commun 11, 6268, doi: 10.1038/s41467-020-20019-0 (2020). Sathe, A. et al. GITR and TIGIT immunotherapy provokes divergent multi-cellular responses in the tumor microenvironment of gastrointestinal cancers. bioRxiv, doi: 10.1101/2023.03.13.532299 (2023). Kumar, P., Saini, S., Khan, S., Surendra Lele, S. & Prabhakar, B. S. Restoring self-tolerance in autoimmune diseases by enhancing regulatory T-cells. Cell Immunol 339, 41–49, doi: 10.1016/j.cellimm.2018.09.008 (2019). Sobhani, N. et al. CTLA-4 in Regulatory T Cells for Cancer Immunotherapy. Cancers (Basel) 13, doi: 10.3390/cancers13061440 (2021). Additional Declarations There is NO conflict of interest to disclose. DECLARATION OF INTEREST STATEMENT The authors declare that they have no competing interests. Cite Share Download PDF Status: Published Journal Publication published 05 May, 2025 Read the published version in Genes & Immunity → Version 1 posted Editorial decision: revise 12 Sep, 2024 Review # 1 received at journal 13 Jun, 2024 Reviewer # 2 agreed at journal 27 May, 2024 Reviewer # 1 agreed at journal 27 May, 2024 Reviewers invited by journal 24 Mar, 2024 Submission checks completed at journal 24 Jan, 2024 First submitted to journal 23 Jan, 2024 Unknown event 22 Jan, 2024 Editor assigned by journal 19 Jan, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-3879125","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":283135607,"identity":"42e27637-9e86-4087-8cb6-4a18d8d589fe","order_by":0,"name":"yue li","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAsklEQVRIiWNgGAWjYHACxgOJ/2x4+NkbSNBzIIEtTUay5wApWhjYDtsY3HAgUrnBjeQDBx7wnOdhuMHA+OFjDlFa0hIOJEjc5mGc3cAsOXMbUVpyDA4kGNzmYZY5wMbMS5yW/A8HEhLO8bBJJBCtJQcYYgcO8PAQrUXyzDODA4kNyTwSPAebifML3/Hkhw9/NtjZ2x9vPvjhIzFaFA7AmYwNRKgHAnki1Y2CUTAKRsFIBgDXJjsS6hcI2gAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0009-0009-4953-4939","institution":"Shanghai Pharmaceuticals Holding Co Ltd","correspondingAuthor":true,"prefix":"","firstName":"yue","middleName":"","lastName":"li","suffix":""},{"id":283135608,"identity":"dd9f9329-7c9a-47ef-9680-34bf8c80daf2","order_by":1,"name":"Jinguo Liu","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Jinguo","middleName":"","lastName":"Liu","suffix":""},{"id":283135609,"identity":"4fac3b79-d303-47cb-85c1-efb0005164bd","order_by":2,"name":"Hua Zhang","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Hua","middleName":"","lastName":"Zhang","suffix":""}],"badges":[],"createdAt":"2024-01-19 15:25:31","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-3879125/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3879125/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41435-025-00330-w","type":"published","date":"2025-05-05T04:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":53583920,"identity":"e45f8d86-891a-48b5-b132-9f40b9a0c903","added_by":"auto","created_at":"2024-03-27 17:52:23","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":5807579,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eA\u003c/strong\u003e Combining specific sequencing data reveals evolution of immune landscape during NSCLC progression. By using GSE162498, prognostic biomarkers were found, then merged with GSE162499, T cell receptor (TCR) repertoire profiling was revealed. GSE123902 was used to investigate the developmental trajectories during NSCLC progression. \u003cstrong\u003eB\u003c/strong\u003escRNA-seq data of GSE162498 was analyzed. UMAP plots of cells from tumor tissue and circulating blood of patients with NSCLC. Each dot corresponds to a cell and is colored according to the different samples, cells and T cells. \u003cstrong\u003eC\u003c/strong\u003eLog-normalized expression levels of canonical marker genes for the above types of cell and T cells. Circle size represents the percentage of cells that express the gene, and colors represent the average expression value within a cluster. \u003cstrong\u003eD\u003c/strong\u003e Bar chart shows the proportion of each cell type in eight patients’ samples, where rows represent cell types and columns represent samples. Box-and-whisker plot was generated to show the percentage of different types of cells in blood and tumor samples. \u003cstrong\u003eE\u003c/strong\u003e Volcano plots of DEGs in natural killer T and naïve CD4+ T cells compared with tumor and blood. \u003cstrong\u003eF\u003c/strong\u003e Significantly enriched Kyoto encyclopedia of genes and genomes (KEGG) pathways of the DEGs in natural killer T and naïve CD4+ T cells compared with tumor and blood. X-axis represents the number of DEGs involved in the significant KEGG pathway, while y-axis represents the significant KEGG pathway.\u003c/p\u003e","description":"","filename":"fig10.png","url":"https://assets-eu.researchsquare.com/files/rs-3879125/v1/9759050b981b23cc48c15c88.png"},{"id":53582851,"identity":"0e279e2d-f0fe-404b-972f-039c3e709a73","added_by":"auto","created_at":"2024-03-27 17:44:23","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":6213697,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eA \u003c/strong\u003eSurvival analysis of selected genes in public database. Kaplan-Meier survival curve was used to analyze the OS and RFS rate of \u003cem\u003eCD8A, CD8B, CD38, CD69, ENTPD1, GZMA, GZMH, MYO1F, SYNE1, TSC22D3\u003c/em\u003e and \u003cem\u003eXCL2\u003c/em\u003e, then a log rank test was performed. Patients whose expression is higher than the best cut-off value are represented by red line, while those whose expression is lower are represented by black line.\u003cstrong\u003e B\u003c/strong\u003e The kinetic curve showed that the relative expression of \u003cem\u003eCD69, CD8A, CD8B, GZMA, GZMH, SYNE1 \u003c/em\u003eand \u003cem\u003eTSC22D3 \u003c/em\u003echanged significantly as a function of pseudotime, indicating their dynamics in different types of CD8\u003csup\u003e+\u003c/sup\u003e and CD4\u003csup\u003e+\u003c/sup\u003e T cells. These lines are approximately expressed along the trajectory using polynomial regressions. \u003cstrong\u003eC \u003c/strong\u003eHeatmap of seven potential prognostic biomarkers in different cell populations. Each vertical bar represents different types of cells. Heatmap is colored in red-to-green gradients. \u003cstrong\u003eD\u003c/strong\u003e Box-and-whisker plots showing the mRNA expression level of genes in Figure 2A between LUAD and normal lung tissues based on the GEPIA platform. \u003cem\u003e|Log2FoldChange| \u0026gt; 1\u003c/em\u003e and \u003cem\u003eP-value \u0026lt; 0.01\u003c/em\u003e were considered statistically significant. Red star indicates that there is a statistically significant difference between tumor and normal tissues. Tumor and normal samples are shown in red and gray, respectively. Y-axes represent the expression in terms of \u003cem\u003elog2 (TPM + 1)\u003c/em\u003e.\u003cstrong\u003e E\u003c/strong\u003e Box-and-whisker plots showing the mRNA expression level of genes in Figure 2A are based on individual LUAD stages. Y-axes represent the expression in terms of\u003cem\u003e log2 (TPM + 1)\u003c/em\u003e.\u003c/p\u003e","description":"","filename":"fig20.png","url":"https://assets-eu.researchsquare.com/files/rs-3879125/v1/947f8b99683cdb5c04faa258.png"},{"id":53582853,"identity":"30278abe-6e25-4268-a891-954fc39f0f74","added_by":"auto","created_at":"2024-03-27 17:44:23","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":6444055,"visible":true,"origin":"","legend":"\u003cp\u003escRNA-seq data of GSE123902 was analyzed. \u003cstrong\u003eA\u003c/strong\u003e The UMAP plot of cells from patients (n = 17) spanning the critical stages of LUAD progression shows different clusters mainly determined by cell type. Each dot corresponds to one cell and is colored according to the cell type. \u003cstrong\u003eB\u003c/strong\u003e Heatmap of seven potential prognostic biomarkers in predicting of LUAD in different cell populations. Each vertical bar represents different types of cells. Heatmap is colored in red-to-green gradients.\u003cstrong\u003e C\u003c/strong\u003e Box-and-whisker plots were generated to show the expression levels of \u003cem\u003eCD8A, CD8B, CD69, GZMA, GZMH, XCL2 \u003c/em\u003eand \u003cem\u003eCD38 \u003c/em\u003ein patients with different samples of LUAD progression. \u003cstrong\u003eD\u003c/strong\u003e Box-and-whisker plots were generated to show the expression levels of \u003cem\u003eCD8A, CD8B, CD69, GZMA, GZMH\u003c/em\u003e and \u003cem\u003eXCL2 \u003c/em\u003ein different subtypes of CD8\u003csup\u003e+\u003c/sup\u003e T cells. \u003cstrong\u003eE\u003c/strong\u003e Bulk RNA-seq GSE 90728 dataset was used to study the genes correlated to \u003cem\u003eGZMA\u003c/em\u003e. Cyclins:\u003cem\u003e CD68, CCL5, CD3D, CCL3, CD2, CCR5, CD3G, CD27\u003c/em\u003e; checkpoint inhibitor: \u003cem\u003eLAG3, PDCD1\u003c/em\u003e; chemokine: \u003cem\u003eCXCR6\u003c/em\u003e; natural killer cell:\u003cem\u003e ID2\u003c/em\u003e; cytokine: \u003cem\u003eIFNG, GZMK, IL7R, ILF2\u003c/em\u003e and cell phenotype: \u003cem\u003eCCR5, CD27\u003c/em\u003e. \u003cstrong\u003eF \u003c/strong\u003eCorrelation of \u003cem\u003eGZMA\u003c/em\u003e expression with tumor purity and the infiltration level of CD8\u003csup\u003e+\u003c/sup\u003e, CD4\u003csup\u003e+\u003c/sup\u003e, regulatory T, gamma delta T, myeloid dendritic cell, macrophage, and B cell were estimated by TIMER2.0 in LUAD.\u003c/p\u003e","description":"","filename":"fig30.png","url":"https://assets-eu.researchsquare.com/files/rs-3879125/v1/8bf8e931f74b07264479300b.png"},{"id":53582858,"identity":"f6c49632-99a8-470e-a17e-452808f28510","added_by":"auto","created_at":"2024-03-27 17:44:23","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":5476172,"visible":true,"origin":"","legend":"\u003cp\u003escTCR-seq data of GSE162499 was analyzed. \u003cstrong\u003eA\u003c/strong\u003e Relative percentage of unique clonotype that are adjusted according to the size of clonotype library. P57_B: circulating blood sample from patient 57; P57_T: tumor tissue from NSCLC patient 57; P58_B: circulating blood sample from patient 58; P58_T: tumor tissue from NSCLC patient 58; P60_B: circulating blood sample from patient 60; P60_T: tumor tissue from NSCLC patient 60; P61_B: circulating blood sample from patient 61; P61_T: tumor tissue from NSCLC patient 61. \u003cstrong\u003eB\u003c/strong\u003eAlluvial plot displaying the dynamic change of the 25 largest expanded TCR sequences from each tumor patient in the peripheral blood samples. The Y-axis represents the relative proportion corresponding to different patient groups. Each color represents a unique clonotype of a patient. \u003cstrong\u003eC \u003c/strong\u003eCorresponding to Figure1B of GSE162498 clustering, UMAP of first 8 most amplified cell clones, circulating blood and tumor tissue samples, and clonotype frequencies distributions. Clonotype distributions: cloning frequencies ≤ 100 and \u0026gt; 20, ≤ 20 and \u0026gt; 5, ≤ 5 and \u0026gt; 1 ≤ 1 and \u0026gt; 0 were defined as large, medium, small expanded and single, respectively. Left: amplified cell clones; middle: circulating blood and tumor tissue samples; right: clonotype frequencies. \u003cstrong\u003eD\u003c/strong\u003e Left: top ten relative uses of TCR genes. Right: a conserved motif constructed according to the sequence patterns of immune repertoire kmer statistics. The length of amino acids is inferred from the ruler at the bottom. Different colored letters represent different kinds of amino acid residues, and the size of letters represents the frequency of amino acids. \u003cstrong\u003eE\u003c/strong\u003eCompare the diversity (Shannon and Inverse (Inv) Simpson indexes) of clonotype in different type of cells. \u003cstrong\u003eF\u003c/strong\u003eJaccard index chart is used to compare the similarity between different type of cells. Elements are colored by the Jaccard index of clonotype overlap between the two types of cells and marked by the percentage of overlapping clones. \u003cstrong\u003eG \u003c/strong\u003eClonal expansion, migration and transition potential of all cells were quantified by STARTRAC indices. Indices were quantified for each n = 4 patient with two matched samples. Center line indicates the median value, lower and upper hinges represent the 25th and 75th percentages, respectively. \u003cstrong\u003eH\u003c/strong\u003e Sketch map showing the dynamic changes of different types of cells between blood and tumor sites.\u003c/p\u003e","description":"","filename":"fig40.png","url":"https://assets-eu.researchsquare.com/files/rs-3879125/v1/1e4be9035e6c89e1f7c5e365.png"},{"id":53582854,"identity":"ccda5703-f17a-4c6a-90ff-90041175a0c6","added_by":"auto","created_at":"2024-03-27 17:44:23","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":5198346,"visible":true,"origin":"","legend":"\u003cp\u003eCharacterization and dynamics of CD4\u003csup\u003e+\u003c/sup\u003e and CD8\u003csup\u003e+\u003c/sup\u003e T cells in NSCLC.\u003cstrong\u003e A\u003c/strong\u003e UMAP visualization showing 4 clusters of CD4\u003csup\u003e+\u003c/sup\u003e T cells and colored by clusters (left), pseudo-time trajectory of the cell (middle), and clonalNetwork (right). \u003cstrong\u003eB\u003c/strong\u003e The interaction of clonotypes across multiple categories: patients, clusters and samples. The ribbon width corresponds to the number of CD4\u003csup\u003e+\u003c/sup\u003e cells in the specified patients or clusters (left side of the ribbon) that were annotated with the cluster identity of the corresponding scRNA-seq clusters or samples (right side of the ribbon). \u003cstrong\u003eC \u003c/strong\u003eJaccard index chart is used to compare the similarity between different CD4\u003csup\u003e+\u003c/sup\u003e cells. Elements are colored by the Jaccard index of clonotype overlap between the two types of cells and marked by the percentage of overlapping clones. \u003cstrong\u003eD \u003c/strong\u003eCompare the diversity (Shannon and Inverse (Inv) Simpson indexes) of clonotype in different CD4\u003csup\u003e+\u003c/sup\u003e cells. \u003cstrong\u003eE \u003c/strong\u003eClonal expansion, migration and transition potential of different CD4\u003csup\u003e+\u003c/sup\u003e cells quantified by STARTRAC indices. \u003cstrong\u003eF \u003c/strong\u003eUMAP visualization showing 5 clusters of CD8\u003csup\u003e+\u003c/sup\u003e T cells and colored by clusters (left), pseudo-time trajectory of the cell (middle), and clonalNetwork (right). \u003cstrong\u003eG\u003c/strong\u003e The interaction of clonotypes across multiple categories: patients, clusters and samples. The ribbon width corresponds to the number of CD8\u003csup\u003e+\u003c/sup\u003e cells in the specified patients or clusters (left side of the ribbon) that were annotated with the cluster identity of the corresponding scRNA-seq clusters or samples (right side of the ribbon). \u003cstrong\u003eH\u003c/strong\u003e Jaccard index chart is used to compare the similarity between different CD8\u003csup\u003e+\u003c/sup\u003e cells. Elements are colored by the Jaccard index of clonotype overlap between the two types of cells and marked by the percentage of overlapping clones.\u003cstrong\u003e I \u003c/strong\u003eCompare the diversity (Shannon and Inverse (Inv) Simpson indexes) of clonotype in different CD8\u003csup\u003e+\u003c/sup\u003e cells. \u003cstrong\u003eJ\u003c/strong\u003e Clonal expansion, migration and transition potential of different CD8\u003csup\u003e+\u003c/sup\u003e cells quantified by STARTRAC indices. \u003cstrong\u003eK\u003c/strong\u003e Sketch map showing the dynamic changes of different CD4\u003csup\u003e+\u003c/sup\u003e and CD8\u003csup\u003e+\u003c/sup\u003e T cells between blood and tumor sites.\u003c/p\u003e","description":"","filename":"fig50.png","url":"https://assets-eu.researchsquare.com/files/rs-3879125/v1/4838b7efa8896b994806a49e.png"},{"id":53582855,"identity":"05619581-c0a3-433b-b661-ddc601837be6","added_by":"auto","created_at":"2024-03-27 17:44:23","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":6704036,"visible":true,"origin":"","legend":"\u003cp\u003eConstructing scRNA-seq GSE123902 developmental trajectories in different samples.\u003cstrong\u003e \u003c/strong\u003eNormal (left), primary tumor (middle), and metastasis (right). \u003cstrong\u003eA \u003c/strong\u003eUMAP of different samples.\u003cstrong\u003e B \u003c/strong\u003eUMAP of different types of cells. \u003cstrong\u003eC\u003c/strong\u003eBox-and-whisker plots showing the predicted ordering by CytoTRACE for individual cells within different cells. \u003cstrong\u003eD\u003c/strong\u003eIn the whole cell differentiation, the first 20 genes are related to CytoTRACE. Top (poorly differentiated, red) and bottom (well differentiated, blue). \u003cstrong\u003eE\u003c/strong\u003e The kinetics plot showed that the relative expression of \u003cem\u003eFGFBP2, NKG7, PRF1, FTH1, FTL, HLA-DRA, BTG1, CCL5, CXCR4, IL7R\u003c/em\u003e and \u003cem\u003eTMSB10\u003c/em\u003echanged significantly with pseudotime, which showed their kinetics in different cell types. These lines are approximated along the trajectory using polynomial regressions.\u003c/p\u003e","description":"","filename":"fig60.png","url":"https://assets-eu.researchsquare.com/files/rs-3879125/v1/e6c69c8e7a772bab05cd1419.png"},{"id":53582856,"identity":"35b80575-917f-43e8-9236-4c313c55f75f","added_by":"auto","created_at":"2024-03-27 17:44:23","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":11846028,"visible":true,"origin":"","legend":"\u003cp\u003eThe expression profiling of cytokines, checkpoint receptors and their ligands in normal, primary tumor and metastasis samples. Dot plot depicting the selected \u003cstrong\u003eA\u003c/strong\u003e cytokines \u003cstrong\u003eB\u003c/strong\u003e checkpoint receptors and their ligands interactions enriched in different samples. Color intensity corresponds to the mean of average expression. Dot size indicates the expressed percentage. The red box highlights highly expressed genes. The lines between the two panels represent receptor ligand pairs \u003cstrong\u003eC\u003c/strong\u003e Transcripts of \u003cem\u003eSCARA5, FTL, IFNGR2, IFNG, ACKR1, CCL5, TNFSF4, TNFRSF4, NECTIN2, TIGIT, CD86 \u003c/em\u003eand \u003cem\u003eCTLA4 \u003c/em\u003eare expressed in different types of cells across different samples. \u003cstrong\u003eD \u003c/strong\u003eSketch map showing the dynamic changes of different CD4\u003csup\u003e+\u003c/sup\u003e, CD8\u003csup\u003e+\u003c/sup\u003e T cells in different samples.\u003c/p\u003e","description":"","filename":"fig70.png","url":"https://assets-eu.researchsquare.com/files/rs-3879125/v1/a8c49daa90fa462ea1bb15f2.png"},{"id":82029589,"identity":"9f2fe4e7-777c-4727-8f98-98cf54333d95","added_by":"auto","created_at":"2025-05-06 07:09:08","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":91020276,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3879125/v1/e749938e-627b-4d1d-926d-631910f84b5c.pdf"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e conflict of interest to disclose.\nDECLARATION OF INTEREST STATEMENT\r\nThe authors declare that they have no competing interests.","formattedTitle":"Single-cell sequencing reveals immune landscape of tumor-infiltrating lymphocytes (TILs) during non-small cell lung cancer (NSCLC) progression","fulltext":[{"header":"Introduction","content":"\u003cp\u003eLung cancer is still the most predominant cancer in the world and the leading cause of cancer-related death. NSCLC including LUAD is the most common lung cancer type, accounting for nearly 85% of all patients \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. Adoptive cell therapy (ACT) with TILs is a highly personalized immunotherapy for cancer \u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. TILs is a highly personalized immunotherapy for cancer \u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. TILs usually exist in the tumor stroma, mainly composed of CD3\u003csup\u003e+\u003c/sup\u003e, CD4\u003csup\u003e+\u003c/sup\u003e, CD8\u003csup\u003e+\u003c/sup\u003e T lymphocytes \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. In some cancers, a high density of TILs in tumor tissue is closely related to a good prognosis \u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. TILs include many phenotypic and functional heterogeneity subgroups \u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Therefore, it is essential to characterize the tumor microenvironment of NSCLC.\u003c/p\u003e \u003cp\u003eA recent study has revealed that a dysfunctional CD8\u003csup\u003e+\u003c/sup\u003e TILs subset displays markers of end-stage differentiation \u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. In addition, this CD8\u003csup\u003e+\u003c/sup\u003e TILs subgroup is accumulated in patients with advanced NSCLC, and its high abundance is related to the poor clinical response of immunotherapy \u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Biomarkers of CD8\u003csup\u003e+\u003c/sup\u003e T cell exhaustion were found to be related to better therapeutic effects \u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. By using scRNA-seq data, some potential prognostic biomarkers were identified, their expression levels were related to the survival time of patients, and they were lower expressed in exhausted CD4/8\u003csup\u003e+\u003c/sup\u003e T than in CD8\u003csup\u003e+\u003c/sup\u003e T cells.\u003c/p\u003e \u003cp\u003eCombining scRNA-seq with scTCR-seq can effectively identify the common cloned cells between TILs and patients' peripheral blood T cells, to characterize their relationship \u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. It allows simultaneous analysis of paired TCR sequences and transcriptome to track different types of cell clones \u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. The existence of TILs in tumor can be proved by sharing the cloning and expression of TCR \u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. Here, we focused on different kinds of cells, CD4\u003csup\u003e+\u003c/sup\u003e and CD8\u003csup\u003e+\u003c/sup\u003e T cells in the stages of state transition, cross-tissue migration and clonal expansion, as well as their biomarkers.\u003c/p\u003e \u003cp\u003eSince TILs contain a variety of cells, developmental processes of these cells from LUAD patients spanning different progression stages are still unclear. We would like to understand the differentiation trajectory of these cells during the process of cancer development, such as non-tumour-involved lung, primary lung adenocarcinomas and lung adenocarcinoma metastases \u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. scRNA-seq is a powerful method to reconstruct the trajectory of cell differentiation and dynamic changes of gene expression \u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eHere, we characterized the dynamics and diversity of single cell profiling across different stages of NSCLC, which contribute to elucidating host adaptive immunity and discovering novel therapeutic targets.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eData availability statement\u003c/h2\u003e \u003cp\u003eA comprehensive review of the literature for next generation sequencing (NGS) studies involving TILs in NSCLC yielded three recent publications: 1. Gueguen et al., referred to as \u0026ldquo;scRNA-seq and scTCR-seq data of resident and circulating precursors to tumor-infiltrating CD8\u003csup\u003e+\u003c/sup\u003e T cell populations in lung cancer\u0026rdquo;, 2. Laughney et al., based on the \u0026ldquo;scRNA-seq transcriptional landscape of primary tumors and metastases human LUAD\u0026rdquo;, 3. Ganesan et al., referred to as \u0026ldquo;Analysis of purified populations of CD8 T cells (isolated from primary lung tumors and matched adjacent lung tissue of lung cancer patients) at the transcriptomic level by RNA sequencing\u0026rdquo; \u003csup\u003e\u003cspan additionalcitationids=\"CR12\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. Each raw dataset was downloaded from the Gene Expression Omnibus (GEO) at the National Center for Biotechnology Information (NCBI). Raw dataset was downloaded from the Gene Expression Omnibus. Accession numbers are: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/gds/?term=gse162498\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/gds/?term=gse162498\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/gds/?term=gse162499\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/gds/?term=gse162499\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/gds/?term=gse90728\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/gds/?term=gse90728\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/gds/?term=gse123902\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/gds/?term=gse123902\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eOnline public database\u003c/h2\u003e \u003cp\u003eThe box-and-whisker plots of prognostic biomarkers in tumor and normal samples were provided by Gene Expression Profiling Interactive Analysis (RRID:SCR_018294) (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://gepia.cancer-pku.cn/detail.php\u003c/span\u003e\u003cspan address=\"http://gepia.cancer-pku.cn/detail.php\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) \u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. GEPIA used data from The Cancer Genome Atlas (TCGA) and Genotype-Tissue Expression (GTEx) \u003csup\u003e13\u003c/sup\u003e. \u003cem\u003e|Log\u003c/em\u003e\u003csub\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sub\u003e\u003cem\u003eFoldChange| \u0026gt; 1\u003c/em\u003e and \u003cem\u003eP-value\u0026thinsp;\u0026lt;\u0026thinsp;0.01\u003c/em\u003e were considered statistical significance. Genes based on individual LUAD stages were obtained from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://ualcan.path.uab.edu/analysis.html\u003c/span\u003e\u003cspan address=\"http://ualcan.path.uab.edu/analysis.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Kaplan-Meier survival curves were retrieved from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://kmplot.com/analysis/index.php?p=service\u0026amp;cancer=pancancer_rnaseq\u003c/span\u003e\u003cspan address=\"http://kmplot.com/analysis/index.php?p=service\u0026amp;cancer=pancancer_rnaseq\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. CellPhoneDB (RRID:SCR_017054) \u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e is a public database of receptor\u0026ndash;ligand interactions. Here, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.cellphonedb.org/\u003c/span\u003e\u003cspan address=\"https://www.cellphonedb.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e was utilized to explore the crosstalk of cell subtypes in NSCLC. The correlation intensions between specified cell types were shown as the total mean and the number of interactions. Correlations of \u003cem\u003eGZMA\u003c/em\u003e expression with tumor purity and with the infiltration level of immune cell are estimated by TIMER2.0 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://timer.cistrome.org/\u003c/span\u003e\u003cspan address=\"http://timer.cistrome.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) in LUAD \u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eCharacterization of prognostic biomarkers in single cells\u003c/h2\u003e \u003cp\u003e \u003cem\u003eSeurat\u003c/em\u003e (version 4.3.0) \u003cem\u003eR\u003c/em\u003e package (version 4.1.1) was used for analysis of scRNA-seq \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/gds/?term=gse162498\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/gds/?term=gse162498\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/gds/?term=gse123902\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/gds/?term=gse123902\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e \u003csup\u003e16\u003c/sup\u003e. Data was normalized with \u003cem\u003eNormalizeData\u003c/em\u003e function. Feature counts of each cell were divided by the total counts for that cell multiplied by a scaler factor (1e4), then natural-log transformed. The normalized data were then integrated for UMAP clustering.\u003c/p\u003e \u003cp\u003eAfter quality control and filtering, we performed an integrated analysis and identified the subset of clusters. \u003cem\u003eFindMarkers\u003c/em\u003e (min.pct\u0026thinsp;=\u0026thinsp;0.25) was used to find cell markers of each cluster. Clusters were annotated based on canonical cell markers (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://xteam.xbio.top/CellMarker/\u003c/span\u003e\u003cspan address=\"http://xteam.xbio.top/CellMarker/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) \u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Uniform manifold approximation and projection (UMAP) was generated by using \u003cem\u003eDimPlot\u003c/em\u003e. \u003cem\u003eDotPlot\u003c/em\u003e was used visualize how feature expression changes across different identity clusters.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eConstructing the developmental trajectory of scRNA-seq\u003c/h2\u003e \u003cp\u003e \u003cem\u003eCytoTRACE\u003c/em\u003e (Cellular Trajectory Reconstruction Analysis using gene Counts and Expression) was used to infer the cell differentiation trajectories for malignant cells \u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. CytoTRACE scores range from 0 to 1, while higher scores indicate higher stemness (less differentiation) and vice versa \u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. \u003cem\u003eplotCytoTRACE\u003c/em\u003e was used to generate the predicted ordered of individual cells. Genes correlated to \u003cem\u003eCytoTRACE\u003c/em\u003e were created by \u003cem\u003eplotCytoGene.\u003c/em\u003e\u003c/p\u003e \u003cp\u003eWithout knowing the differentiation time or direction in advance, \u003cem\u003eMonocle3\u003c/em\u003e package was used to analyze the developmental trajectory \u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. \u003cem\u003elearn_graph\u003c/em\u003e was applied to learn the trajectory graph. \u003cem\u003eorder_cells\u003c/em\u003e calculated where each cell falls in pseudotime. The function \u003cem\u003eplot_genes_in_pseudotime\u003c/em\u003e takes a small group of genes and displays their dynamics as a function of pseudotime.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003escTCR-seq data analysis\u003c/h2\u003e \u003cp\u003e \u003cem\u003escRepertoire\u003c/em\u003e was used to merge and visualize scTCR-seq data with \u003cem\u003eSeurat\u003c/em\u003e object of scRNA-seq data \u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. The percentage of unique clones in the bar graph across different samples was generated by \u003cem\u003equantContig\u003c/em\u003e function. UMAPs of clonotype in \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/gds/?term=gse162498\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/gds/?term=gse162498\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e clusters were performed by \u003cem\u003eDimPlot\u003c/em\u003e. Basic analysis visualizations the relative usage of genes of the TCR is used by \u003cem\u003evizGenes\u003c/em\u003e. Compare the clonotype diversity (Shannon and Inverse (Inv) Simpson indexes) among different cell types were handled by \u003cem\u003eclonalDiversity\u003c/em\u003e. Counts of different clonotype cells in clusters were calculated by \u003cem\u003eclonalOverlap\u003c/em\u003e. For looking at clonotypes by cellular origins and cluster identification, \u003cem\u003eStartracDiversity\u003c/em\u003e is used. Network interaction of clonotypes shared between clusters along the single-cell dimensional reduction is used by \u003cem\u003eclonalNetwork\u003c/em\u003e. Clonotypes across multiple categories were created by \u003cem\u003ealluvialClonotypes\u003c/em\u003e.\u003c/p\u003e \u003cp\u003e \u003cem\u003eimmunarch\u003c/em\u003e was utilized to compare the degree of clonal expansion in the repertoire of each sample \u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. Clonotype tracking in different samples was utilized by \u003cem\u003etrackClonotypes.\u003c/em\u003e Motif of clonotype is constructed by \u003cem\u003egetKmers.\u003c/em\u003e\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eSoftware used\u003c/h2\u003e \u003cp\u003eGraphPad Prism software (GraphPad Software, Inc.) was employed to draw the mRNA expression levels of prognostic biomarkers in different patient groups and cell types of \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/gds/?term=gse123902\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/gds/?term=gse123902\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Genes in \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/gds/?term=gse90728\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/gds/?term=gse90728\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e correlated with \u003cem\u003eGZMA\u003c/em\u003e were calculated by SigmaPlot (RRID:SCR_003210). All deep-sequencing data were analyzed in Bioconductor (RRID:SCR_006442) version 3.14 (BiocManager 1.30.19), \u003cem\u003eR\u003c/em\u003e 4.1 1 (\u003cem\u003eR\u003c/em\u003e Core Team, 2021; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.R-project.org/\u003c/span\u003e\u003cspan address=\"http://www.R-project.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) under Ubuntu environment (20.04). Values of \u003cem\u003e*p\u0026thinsp;\u0026lt;\u0026thinsp;0.05, **p\u0026thinsp;\u0026lt;\u0026thinsp;0.01, ***p\u0026thinsp;\u0026lt;\u0026thinsp;0.001\u003c/em\u003e, and \u003cem\u003e****p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001\u003c/em\u003e were considered significant. All codes are available at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/yueli8/TILs_NSCLC\u003c/span\u003e\u003cspan address=\"https://github.com/yueli8/TILs_NSCLC\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eHigh-resolution landscape of NSCLC by scRNA-seq\u003c/h2\u003e \u003cp\u003eThe brief workflow is shown in Fig.\u0026nbsp;1A. Four NSCLC tumor tissue and matched circulating blood samples were used in our study. Nine types of cells were identified by UMAP clustering analysis (Fig.\u0026nbsp;1B). Most of the cells in TILs are T cells. Ten types of T cells were identified. The most abundant T cells are natural killer T and effector memory CD4\u003csup\u003e+\u003c/sup\u003e T cells. Graph-based clusters were manually annotated based on known marker genes for the main expected cell types (Fig.\u0026nbsp;1C).\u003c/p\u003e \u003cp\u003eThe proportion of each T cell lineage varies greatly in different tumors and blood samples. Compared with blood samples, na\u0026iuml;ve CD4\u003csup\u003e+\u003c/sup\u003e and effector memory CD8\u003csup\u003e+\u003c/sup\u003e T cells are significantly increased in tumor, while resident memory CD8\u003csup\u003e+\u003c/sup\u003e and regulatory T cells are decreased (Fig.\u0026nbsp;1D). Natural killer T cells contains the largest number of differentially expressed genes (DEGs). Volcano plots and Kyoto encyclopedia of genes and genomes (KEGG) pathways of the DEGs in natural killer T and na\u0026iuml;ve CD4\u003csup\u003e+\u003c/sup\u003e T cells were shown in Fig.\u0026nbsp;1E and F. Some pathways have significant systemic anti-tumor effects in immunotherapy, such as: MARK, p53, NF-kappa B and TNF signaling. The cell differentiation pathways of Th1, Th2 and Th17 were also found, which drove CD4\u003csup\u003e+\u003c/sup\u003e T naive cells into diverse T helper subsets.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eIdentification of potential prognostic biomarkers of TILs in NSCLC\u003c/h2\u003e \u003cp\u003eNext, by using Kaplan-Meier plots, DEGs in T cells were used to further study their overall survival (OS) and recurrence-free survival (RFS). Best cutoff analysis showed that the low expressed of \u003cem\u003eCD8A, CD8B, CD38, CD69, ENTPD1, GZMA, GZMH, MYO1F, SYNE1, TSC22D3\u003c/em\u003e and \u003cem\u003eXCL2\u003c/em\u003e were associated with the poor OS and RFS of LUAD patients (Fig.\u0026nbsp;2A).\u003c/p\u003e \u003cp\u003eSome of them have higher Moran\u0026rsquo;s I values, and they may largely represent markers of different T cells (Fig.\u0026nbsp;2B). We consistently observed substantially higher Moran\u0026rsquo;s I values of them and revealed that they may largely represented markers of different T cells. \u003cem\u003eCD69\u003c/em\u003e and \u003cem\u003eTSC22D3\u003c/em\u003e have peaks in exhausted CD8\u003csup\u003e+\u003c/sup\u003e T cells, \u003cem\u003eGZMA, GZMH\u003c/em\u003e and \u003cem\u003eSYNE1\u003c/em\u003e in resident memory CD8\u003csup\u003e+\u003c/sup\u003e T cells, while \u003cem\u003eCD8A and CD8B\u003c/em\u003e increased in effector memory CD8\u003csup\u003e+\u003c/sup\u003e T cells. \u003cem\u003eCD69\u003c/em\u003e, \u003cem\u003eGZMA\u003c/em\u003e and \u003cem\u003eTSC22D3\u003c/em\u003e were highly expressed in T helper cells, but low expressed in na\u0026iuml;ve CD4\u003csup\u003e+\u003c/sup\u003e T cells.\u003c/p\u003e \u003cp\u003eIn hierarchically-clustered heatmaps, compared with other types of T cells, \u003cem\u003eCD8A, GZMA, CD8B\u003c/em\u003e and \u003cem\u003eGZMH\u003c/em\u003e are highly expressed in effector memory CD8\u003csup\u003e+\u003c/sup\u003e and na\u0026iuml;ve CD4\u003csup\u003e+\u003c/sup\u003e T cells (Fig.\u0026nbsp;2C). Furthermore, box-and-whisker plots of these genes were shown. \u003cem\u003eCD69, MYO1F, SYNE1\u003c/em\u003e, and \u003cem\u003eTSC22D3\u003c/em\u003e were significantly higher (\u003cem\u003eP-value\u0026thinsp;\u0026lt;\u0026thinsp;0.01\u003c/em\u003e) expressed in tumor tissue than in normal samples (Fig.\u0026nbsp;2D). The expression levels of these genes spanning different stages are retrieved (Fig.\u0026nbsp;2E). They were lower in the late/advanced stage of LUAD patients, but higher in early/primary stage (except \u003cem\u003eCD38\u003c/em\u003e). All these indicate that the above genes could be used as potential prognostic biomarkers of TILs in NSCLC.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eIdentification of lower expressed genes in exhausted CD4/8\u003csup\u003e+\u003c/sup\u003e T than in CD8\u003csup\u003e+\u003c/sup\u003e T cells\u003c/h2\u003e \u003cp\u003eNext, we used We used the GSE123902 dataset which contains transcriptionally profiling of single cells from patients spanning different stages of LUAD progression to discover genes which are lower expressed in exhausted CD4/8\u003csup\u003e+\u003c/sup\u003e T than in naive CD8\u003csup\u003e+\u003c/sup\u003e T cells. First, ten types of cells were identified by using UMAP (Fig.\u0026nbsp;3A). By using hierarchical clustering of heatmaps (Fig.\u0026nbsp;3B). \u003cem\u003eCD8A, CD8B, CD69, GZMA, GZMH\u003c/em\u003e and \u003cem\u003eXCL2\u003c/em\u003e were higher expressed in naive CD8\u003csup\u003e+\u003c/sup\u003e and CD8\u003csup\u003e+\u003c/sup\u003e T cells than other types of cells.\u003c/p\u003e \u003cp\u003eNext, afore-mentioned genes across different stages of LUAD progression were studied. All of them are statistically lower \u003cem\u003e(P-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05)\u003c/em\u003e expressed in metastatic than in normal or primary tumor patients (Fig.\u0026nbsp;3C). In addition, \u003cem\u003eCD8A, CD8B, CD69, GZMA, GZMH\u003c/em\u003e and \u003cem\u003eXCL2\u003c/em\u003e were significantly lower (\u003cem\u003eP-value\u0026thinsp;\u0026lt;\u0026thinsp;0.0001\u003c/em\u003e) expressed in exhausted CD4/8\u003csup\u003e+\u003c/sup\u003e than in naive CD8\u003csup\u003e+\u003c/sup\u003e or CD8\u003csup\u003e+\u003c/sup\u003e T cells (Fig.\u0026nbsp;3D). Among all these genes, \u003cem\u003eGZMA\u003c/em\u003e has the highest expression (about 1.2) in CD8\u003csup\u003e+\u003c/sup\u003e T cells, therefore \u003cem\u003eGZMA\u003c/em\u003e could be a potential biomarker for the prognosis of patients with LUAD.\u003c/p\u003e \u003cp\u003e \u003cem\u003eGZMA\u003c/em\u003e was also correlated with some genes by using bulk RNA-seq GSE90728 (Fig.\u0026nbsp;3E). It was positively correlated with cyclins, such as: \u003cem\u003eCCL5, CD3D, CCL3, CD2, CCR5, CD3G, CD27\u003c/em\u003e, checkpoint inhibitor: \u003cem\u003eLAG3\u003c/em\u003e and \u003cem\u003ePDCD1\u003c/em\u003e, chemokine: \u003cem\u003eCXCR6\u003c/em\u003e, natural killer cell: \u003cem\u003eID2\u003c/em\u003e, cytokine: \u003cem\u003eIFNG, GZMK, LF2\u003c/em\u003e, cell phenotype: \u003cem\u003eCCR5\u003c/em\u003e and \u003cem\u003eCD27\u003c/em\u003e, and negatively correlated with \u003cem\u003eCD68\u003c/em\u003e and \u003cem\u003eIL7R\u003c/em\u003e. These genes may have anti-tumor immune functions related to \u003cem\u003eGZMA\u003c/em\u003e. Therefore, \u003cem\u003eGZMA\u003c/em\u003e can be used as a novel prognostic marker of TILs in LUAD.\u003c/p\u003e \u003cp\u003eFinally, the correlation between immune cell infiltration and \u003cem\u003eGZMA\u003c/em\u003e expression was evaluated by TIMER2.0 \u003csup\u003e15\u003c/sup\u003e. As shown in the scatter plots (Fig.\u0026nbsp;3F), \u003cem\u003eGZMA\u003c/em\u003e negatively correlates with tumor purity (correlation = -0.427, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2.67e-23), suggesting that the main source of \u003cem\u003eGZMA\u003c/em\u003e expression detected are stromal and immune cells. Also, \u003cem\u003eGZMA\u003c/em\u003e was positively correlated with immune infiltration of CD8\u003csup\u003e+\u003c/sup\u003e, CD4\u003csup\u003e+\u003c/sup\u003e, regulatory, gamma delta T, myeloid dendritic cell, macrophage, and B cell. These findings further revealed a strong relationship between \u003cem\u003eGZMA\u003c/em\u003e and immunosuppression in the inflammatory tumor microenvironment.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003escTCR-seq repertoire of TILs in NSCLC\u003c/h2\u003e \u003cp\u003eTo investigate the diversity and dynamics of T cell repertoire of TILs in NSCLC, a combination of scRNA-seq and scTCR- seq data from GSE162500 was used \u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. Firstly, the number of unique clonotypes pre-patient was studied (Fig.\u0026nbsp;4A). Except for patient 58, the percentage of unique clonotypes in circulating blood was higher than that in tumor tissue. These findings indicated that circulating blood has more unique clonotypes than tumor tissue.\u003c/p\u003e \u003cp\u003eIn fact, by tracking the single clones of all patients, we revealed the top 25 most amplified cell clones, and these clones had significant dynamic changes in patients (Fig.\u0026nbsp;4B). Except for patient 58, all these clones had more significant amplification in tumor tissue than in circulating blood.\u003c/p\u003e \u003cp\u003eCorresponding to Fig.\u0026nbsp;1b, the top eight most amplified cell clones were concentrated in terminally exhausted CD8\u003csup\u003e+\u003c/sup\u003e T cells, with the first one referring to the amino acid sequence: CAASRNAGNMLTF_CASSISGTGEIGEAFF (Fig.\u0026nbsp;4C). There are more terminally exhausted CD8\u003csup\u003e+\u003c/sup\u003e and regulatory T cells clones in tumor tissue than in circulating blood samples, and most of them have been expanded in large and medium scales.\u003c/p\u003e \u003cp\u003eTop ten uses of TCR gene are shown in Fig.\u0026nbsp;4D. A putative binding motif for CDR3 is predominantly composed of 5-residue peptides. The first two peptides are SSSSG and GGGGF, with polar residues S and G and hydrophobic residues F and A.\u003c/p\u003e \u003cp\u003eInterestingly, highest diversity indices were observed in dendritic cells and macrophages, while natural killer and T were the lowest (Fig.\u0026nbsp;4E). To calculate the degree of TCR repertoire overlap among various cell types, we used the overlap coefficient method. There is substantial degree of 36.4% TCR clones overlap between dendritic cell and T cells, between macrophages and T cell populations (Fig.\u0026nbsp;4F).\u003c/p\u003e \u003cp\u003eThe existence of clonal cells spanning several different tissue sites suggests the migration, transformation and expansion of the specified cell types. As expected, among all the cells, macrophages (\u003cem\u003eFTL\u003c/em\u003e) and natural killer (\u003cem\u003eGNLY\u003c/em\u003e) showed the highest clonal transition, while T (\u003cem\u003eIL7R\u003c/em\u003e) cells showed the highest clonal expansion index (Fig.\u0026nbsp;4G and H). Moreover, macrophages (\u003cem\u003eFTL\u003c/em\u003e) and dendritic (\u003cem\u003eAIF1\u003c/em\u003e) cells transfer the most CD3 TCR clones to T (\u003cem\u003eIL7R\u003c/em\u003e) cells. These results implicate a potential role of different immune cells in shaping the NSCLC during T cell infiltration\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003escTCR-seq repertoire of CD4\u003csup\u003e+\u003c/sup\u003e and CD8\u003csup\u003e+\u003c/sup\u003e T cells in NSCLC\u003c/h2\u003e \u003cp\u003eNext, we investigated the diversity and dynamics of CD4\u003csup\u003e+\u003c/sup\u003e and CD8\u003csup\u003e+\u003c/sup\u003e T cells repertoire in NSCLC. Figure\u0026nbsp;5A shows the UMAP, developmental trajectory and clonalNetwork of CD4\u003csup\u003e+\u003c/sup\u003e T cells. T helper appeared at first, followed by effector memory or regulatory T cells. It is worth noting that naive cells will still appear, and mainly accumulate at the final stage of progression. Effector memory cells receive most clones from regulatory and na\u0026iuml;ve cells.\u003c/p\u003e \u003cp\u003eFrom alluvialClonotypes in Fig.\u0026nbsp;5B, effector memory shows the high proportion. There are 13.5% clone overlapping between regulatory and T helper cells (Fig.\u0026nbsp;5C). Interestingly, highest diversity indices were observed in na\u0026iuml;ve and effector memory cells, while T helper cells was the lowest (Fig.\u0026nbsp;5D). Besides, T helper showed the highest clonal transition and expansion index (Fig.\u0026nbsp;5E).\u003c/p\u003e \u003cp\u003eIn CD8\u003csup\u003e+\u003c/sup\u003e T cells, cytotoxic appeared at first, followed by exhausted or memory cells. Terminally exhausted cells receive most clones from pre-exhausted and resident memory cells (Fig.\u0026nbsp;5F). The proportion of terminally exhausted cells is the highest (Fig.\u0026nbsp;5G). There is 13.5% clone overlapping between cytotoxic and terminally exhausted cells (Fig.\u0026nbsp;5H). Higher diversity indices were observed in effector memory and pre-exhausted cells, while terminally exhausted was the lowest (Fig.\u0026nbsp;5I). Also, cytotoxic and resident memory showed the highest clonal transition index, while terminally exhausted cells showed the highest clonal expansion (Fig.\u0026nbsp;5J).\u003c/p\u003e \u003cp\u003eCollectively, in CD4\u003csup\u003e+\u003c/sup\u003e T cells, T helper (\u003cem\u003eCXCL13\u003c/em\u003e) and regulatory T (\u003cem\u003eFOXP3\u003c/em\u003e) cells have higher ability of transition and expansion. They also transfer the largest number of CD3 TCR clones to each other. In CD8\u003csup\u003e+\u003c/sup\u003e T cells, cytotoxic (\u003cem\u003eNKG7\u003c/em\u003e) and resident memory (\u003cem\u003eGZMK\u003c/em\u003e) have higher ability of transition, and terminally exhausted (\u003cem\u003eCCL5\u003c/em\u003e) have higher expansion ability. Cytotoxic (\u003cem\u003eNKG7\u003c/em\u003e) transports the most CD3 TCR clones to terminally exhausted (\u003cem\u003eCCL5\u003c/em\u003e) cells (Fig.\u0026nbsp;5K).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eDevelopmental trajectories of NSCLC during tumor progression\u003c/h2\u003e \u003cp\u003eTo further understand the different cell states in NSCLC spanning different stages, the developmental trajectories of tumor cells was studied. Figure\u0026nbsp;6A shows the human tissue samples in GSE123902: adjacent non-tumor involved lung (n\u0026thinsp;=\u0026thinsp;4, \u0026lsquo;normal\u0026rsquo;), primary LUAD (n\u0026thinsp;=\u0026thinsp;8, \u0026lsquo;primary tumor\u0026rsquo;), and metastatic tumor of LUAD (n\u0026thinsp;=\u0026thinsp;5, \u0026lsquo;metastasis\u0026rsquo;).\u003c/p\u003e \u003cp\u003eUMAP of different types of cells is shown in Fig.\u0026nbsp;6B. Na\u0026iuml;ve CD4\u003csup\u003e+\u003c/sup\u003e T cells only appeared in normal samples, while regulatory T cells in metastasis. Pre-exhausted CD8\u003csup\u003e+\u003c/sup\u003e T cells showed up in primary tumor, while cytotoxic CD8\u003csup\u003e+\u003c/sup\u003e T cells in metastasis.\u003c/p\u003e \u003cp\u003eThe predicted ordering of different cell types was displayed in box plots (Fig.\u0026nbsp;6C). In normal samples, macrophages first appeared, followed by CD4\u003csup\u003e+\u003c/sup\u003e, CD8\u003csup\u003e+\u003c/sup\u003e T cells. In primary tumor, progenitor and epithelial developed at first, then is T helper, effector CD8\u003csup\u003e+\u003c/sup\u003e, CD4\u003csup\u003e+\u003c/sup\u003e, pre-exhausted CD8\u003csup\u003e+\u003c/sup\u003e T cells. Furthermore, cytotoxic CD8\u003csup\u003e+\u003c/sup\u003e T, effector memory and recently activation CD4\u003csup\u003e+\u003c/sup\u003e T still exist at the late stage of metastasis.\u003c/p\u003e \u003cp\u003eAdditionally, the top 10 (less differentiated; red) and bottom 10 (most differentiated; blue) genes in this dataset based on their correlation with CytoTRACE can be predicted (Fig.\u0026nbsp;6D). Some of them are significantly related to the specific states in different types of cells in pseudotime trajectory (Fig.\u0026nbsp;6E). In normal samples, \u003cem\u003eFGFBP2\u003c/em\u003e, \u003cem\u003eNKG7\u003c/em\u003e and \u003cem\u003ePRF1\u003c/em\u003e are associated with cytotoxic CD8\u003csup\u003e+\u003c/sup\u003e T cells; while \u003cem\u003eFTH1\u003c/em\u003e, \u003cem\u003eFTL\u003c/em\u003e and \u003cem\u003eHLA-DRA\u003c/em\u003e are correlated with macrophages. In primary tumor samples, \u003cem\u003eBTG1, CCL5, CXCR4, FTL, IL7R, and TMSB10\u003c/em\u003e are highly expressed during the process of development. In metastasis samples, \u003cem\u003eFTH1\u003c/em\u003e and \u003cem\u003eFTL\u003c/em\u003e were related to either CD4\u003csup\u003e+\u003c/sup\u003e or CD8\u003csup\u003e+\u003c/sup\u003e T cells\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eThe expression profiling of cytokines, checkpoint receptors and their ligands during tumor progression\u003c/h2\u003e \u003cp\u003eCytokine therapy helps the immune system stop the growth of cancer cells or kill them \u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. To address the expression profiling of cytokines receptors and their ligands in different samples, 26 kinds of cytokines were selected (Fig.\u0026nbsp;7A). Most of them are highly expressed in primary tumors, while \u003cem\u003eCCL5\u003c/em\u003e and \u003cem\u003eCCL4\u003c/em\u003e are highly expressed in cytotoxic CD8\u003csup\u003e+\u003c/sup\u003e T cells of metastasis samples. Cytokine receptors are highly expressed in macrophages of normal and primary tumor samples, while \u003cem\u003eCXCR4\u003c/em\u003e is highly expressed in CD4\u003csup\u003e+\u003c/sup\u003e T cells in metastasis samples.\u003c/p\u003e \u003cp\u003eIt is generally believed that cancer cells are the only source of checkpoint ligands and are responsible for inhibiting T-cell immune responses \u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. Checkpoint receptors are highly expressed in chronic activation CD4\u003csup\u003e+\u003c/sup\u003e T cells of normal samples, T helper of primary tumor samples, and regulatory T of metastasis samples (Fig.\u0026nbsp;7B).\u003c/p\u003e \u003cp\u003eNext, the transcript levels of some ligands and receptor in different samples were also studied (Fig.\u0026nbsp;7C). Notably, in primary tumor samples, \u003cem\u003eFTL\u003c/em\u003e was highly expressed in T helper cells, and \u003cem\u003eIFNG\u003c/em\u003e was in pre-exhausted CD8\u003csup\u003e+\u003c/sup\u003e T cells. In normal and metastatic samples, \u003cem\u003eCCL5\u003c/em\u003e was highly expressed in cytotoxic CD8\u003csup\u003e+\u003c/sup\u003e T cells. \u003cem\u003eTNFRSF4\u003c/em\u003e, \u003cem\u003eTIGIT\u003c/em\u003e and \u003cem\u003eCTLA4\u003c/em\u003e were highly expressed in regulatory T cells of metastasis samples.\u003c/p\u003e \u003cp\u003eBy integrating the data of single cell transcriptome of different samples, the development trajectory of CD8\u003csup\u003e+\u003c/sup\u003e T cells is shown below (Fig.\u0026nbsp;7D). In normal samples, it is from effector memory, then is progenitor and cytotoxicity (\u003cem\u003eFGFBP2\u003c/em\u003e, \u003cem\u003eNKG7\u003c/em\u003e, \u003cem\u003ePRF1\u003c/em\u003e and \u003cem\u003eCCL5\u003c/em\u003e). In primary tumor, it is from effector (\u003cem\u003eCCL5\u003c/em\u003e) to pre-exhausted (\u003cem\u003eCCL5\u003c/em\u003e and \u003cem\u003eIFNG\u003c/em\u003e). Cytotoxicity (\u003cem\u003eCCL5\u003c/em\u003e) still appeared in the metastatic samples.\u003c/p\u003e \u003cp\u003eIn CD4\u003csup\u003e+\u003c/sup\u003e T cells, the development trajectory is as follows. In normal samples, it starts from chronic activation, T helper, and then na\u0026iuml;ve. In primary tumor, it is from T helper (\u003cem\u003eTNFRSF4\u003c/em\u003e, \u003cem\u003eTIGIT\u003c/em\u003e and \u003cem\u003eFTL\u003c/em\u003e) to effector. In metastatic samples, it begins with regulatory T (\u003cem\u003eCTLA4\u003c/em\u003e, \u003cem\u003eTIGIT\u003c/em\u003e and \u003cem\u003eFTL\u003c/em\u003e) to effector memory, and then recently activation.\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eHere, we investigated the diversity of TILs during NSCLC progression. In TILs of NSCLC, many pathways are concentrated on cancer-related. The MAPK pathway leads to uncontrolled growths. \u003cem\u003eIL17\u003c/em\u003e is a key cytokine which induces inflammation and activates several downstream signal pathways \u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e. Th1 cells mediate a cellular immune response, Th2 cells strengthen humoral response, and Th17 cells produce proinflammatory cytokine \u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eHere, the specific expression of \u003cem\u003eGZMA\u003c/em\u003e was identified as potential new marker of TILs. \u003cem\u003eGZMA\u003c/em\u003e is an enzyme existing in cytotoxic T lymphocyte granules, and it is considered as a major proapoptotic protease \u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. During pneumococcal pneumonia, it defenses through a mechanism that does not depend on natural killer cells. Its related pathways including IL-9, IL-15 signaling, as well as their main biological effects on different types of immune cell.\u003c/p\u003e \u003cp\u003eMacrophage (\u003cem\u003eFTL\u003c/em\u003e) and natural killer (\u003cem\u003eGNLY\u003c/em\u003e) cells exhibited high clonal transition, macrophages (\u003cem\u003eFTL\u003c/em\u003e) and dendritic (\u003cem\u003eAIF1\u003c/em\u003e) transfer the most CD3 TCR clones to T (\u003cem\u003eIL7R\u003c/em\u003e) cells. In many cancers, \u003cem\u003eFTL\u003c/em\u003e was significantly positively correlated with tumor infiltration by tumor-associated macrophages \u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. Granulysin (\u003cem\u003eGNLY\u003c/em\u003e) exists in human cytotoxic T lymphocytes and natural killer cells \u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. \u003cem\u003eAIF1\u003c/em\u003e guides hematopoietic and immune responses within dendritic cells and macrophages \u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. \u003cem\u003eIL7R\u003c/em\u003e plays a key role in the development of lymphocytes.\u003c/p\u003e \u003cp\u003eIn CD4\u003csup\u003e+\u003c/sup\u003e T cells, T helper (\u003cem\u003eCXCL13\u003c/em\u003e) and regulatory T (\u003cem\u003eFOXP3\u003c/em\u003e) cells transfer the largest number of CD3 TCR clones to each other and have higher ability of transition and expansion. \u003cem\u003eFOXP3\u003c/em\u003e seems to play a role in the development and function of regulatory T cells \u003csup\u003e\u003cspan additionalcitationids=\"CR30\" citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIn CD8\u003csup\u003e+\u003c/sup\u003e T cells, cytotoxic (\u003cem\u003eNKG7\u003c/em\u003e) transfers the most CD3 TCR clones to terminally exhausted (\u003cem\u003eCCL5\u003c/em\u003e) CD8\u003csup\u003e+\u003c/sup\u003e T cells. \u003cem\u003eNKG7\u003c/em\u003e was identified as a necessary component for the cytotoxic function of CD8\u003csup\u003e+\u003c/sup\u003e T cells and it is a T-cell\u0026ndash;intrinsic therapeutic target for enhancing cancer immunotherapy \u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. \u003cem\u003eCCL5\u003c/em\u003e was highly expressed in LAG3\u003csup\u003e+\u003c/sup\u003e exhausted T cells in hepatocellular carcinoma \u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe development trajectory of CD8\u003csup\u003e+\u003c/sup\u003e T cells in normal samples is cytotoxicity (\u003cem\u003eFGFBP2\u003c/em\u003e, \u003cem\u003eNKG7\u003c/em\u003e, \u003cem\u003ePRF1\u003c/em\u003e and \u003cem\u003eCCL5\u003c/em\u003e) at the final stage. FGFBP2 protein is secreted by cytotoxic T-lymphocytes and combined with fibroblast growth factor \u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. \u003cem\u003ePRF1\u003c/em\u003e is thought to act by creating holes in the plasma membrane which triggers an influx of calcium and initiates membrane repair mechanisms. In primary tumor samples, pre-exhausted (\u003cem\u003eIFNG\u003c/em\u003e and \u003cem\u003eCCL5\u003c/em\u003e) appeared at last. It may serve as better targets for immunotherapies compared to the exhausted cluster \u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe development trajectory of CD4\u003csup\u003e+\u003c/sup\u003e T cells in primary tumor samples is T helper (\u003cem\u003eFTL, TNFRSF4\u003c/em\u003e and \u003cem\u003eTIGIT\u003c/em\u003e) appeared at first. \u003cem\u003eTNFRSF4\u003c/em\u003e has a critical role in the maintenance of an immune response due to its ability to enhance survival. \u003cem\u003eTIGIT\u003c/em\u003e antagonist also activated T follicular helper-like cells and dendritic cells, and reduced markers of immunosuppression in regulatory T cells \u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe regulatory T cells (Tregs or Treg cells), maintain tolerance to self-antigens, and prevent autoimmune disease \u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. Tregs in the tumor microenvironment are abundantly effector Tregs which over-express immunosuppressive molecules, such as CTLA-4 \u003csup\u003e38\u003c/sup\u003e. Anti-CTLA-4 antibodies lead to the depletion of regulatory T cells and thus increasing the effective anti-tumor effect of CD8\u003csup\u003e+\u003c/sup\u003e T cells \u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIn conclusion, our work may serve as potential predictive markers for chemotherapy and long-term survival prognosis, as well as new therapeutic targets or strategies to overcome drug resistance and immunosuppression.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe deeply thank Jie Gao for scientific discussions.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDeclaration of Interest Statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eChen, G., Hu, J., Huang, Z., Yang, L. \u0026amp; Chen, M. MicroRNA-1976 functions as a tumor suppressor and serves as a prognostic indicator in non-small cell lung cancer by directly targeting PLCE1. Biochem Biophys Res Commun 473, 1144\u0026ndash;1151, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.bbrc.2016.04.030\u003c/span\u003e\u003cspan address=\"10.1016/j.bbrc.2016.04.030\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGranhoj, J. S. \u003cem\u003eet al.\u003c/em\u003e Tumor-infiltrating lymphocytes for adoptive cell therapy: recent advances, challenges, and future directions. Expert Opin Biol Ther 22, 627\u0026ndash;641, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/14712598.2022.2064711\u003c/span\u003e\u003cspan address=\"10.1080/14712598.2022.2064711\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKasper, H. U., Drebber, U., Stippel, D. L., Dienes, H. P. \u0026amp; Gillessen, A. Liver tumor infiltrating lymphocytes: comparison of hepatocellular and cholangiolar carcinoma. World J Gastroenterol 15, 5053\u0026ndash;5057, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3748/wjg.15.5053\u003c/span\u003e\u003cspan address=\"10.3748/wjg.15.5053\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, J. \u003cem\u003eet al.\u003c/em\u003e CD8\u0026thinsp;+\u0026thinsp;tumor-infiltrating lymphocytes as a novel prognostic biomarker in lung sarcomatoid carcinoma, a rare subtype of lung cancer. Cancer Manag Res 10, 3505\u0026ndash;3511, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2147/CMAR.S169074\u003c/span\u003e\u003cspan address=\"10.2147/CMAR.S169074\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee, S. W. \u003cem\u003eet al.\u003c/em\u003e CD8(+) TILs in NSCLC differentiate into TEMRA via a bifurcated trajectory: deciphering immunogenicity of tumor antigens. J Immunother Cancer 9, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1136/jitc-2021-002709\u003c/span\u003e\u003cspan address=\"10.1136/jitc-2021-002709\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSanmamed, M. F. \u003cem\u003eet al.\u003c/em\u003e A Burned-Out CD8(+) T-cell Subset Expands in the Tumor Microenvironment and Curbs Cancer Immunotherapy. Cancer Discov 11, 1700\u0026ndash;1715, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1158/2159-8290.CD-20-0962\u003c/span\u003e\u003cspan address=\"10.1158/2159-8290.CD-20-0962\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHsu, C. L. \u003cem\u003eet al.\u003c/em\u003e Exploring Markers of Exhausted CD8 T Cells to Predict Response to Immune Checkpoint Inhibitor Therapy for Hepatocellular Carcinoma. Liver Cancer 10, 346\u0026ndash;359, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1159/000515305\u003c/span\u003e\u003cspan address=\"10.1159/000515305\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGohil, S. H., Iorgulescu, J. B., Braun, D. A., Keskin, D. B. \u0026amp; Livak, K. J. Applying high-dimensional single-cell technologies to the analysis of cancer immunotherapy. Nat Rev Clin Oncol 18, 244\u0026ndash;256, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41571-020-00449-x\u003c/span\u003e\u003cspan address=\"10.1038/s41571-020-00449-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang, H. Q., Wang, Y. S., Zhai, K. \u0026amp; Tong, Z. H. Single-Cell TCR Sequencing Reveals the Dynamics of T Cell Repertoire Profiling During Pneumocystis Infection. Front Microbiol 12, 637500, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fmicb.2021.637500\u003c/span\u003e\u003cspan address=\"10.3389/fmicb.2021.637500\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGulati, G. S. \u003cem\u003eet al.\u003c/em\u003e Single-cell transcriptional diversity is a hallmark of developmental potential. Science 367, 405\u0026ndash;411, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1126/science.aax0249\u003c/span\u003e\u003cspan address=\"10.1126/science.aax0249\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGueguen, P. \u003cem\u003eet al.\u003c/em\u003e Contribution of resident and circulating precursors to tumor-infiltrating CD8(+) T cell populations in lung cancer. Sci Immunol 6, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1126/sciimmunol.abd5778\u003c/span\u003e\u003cspan address=\"10.1126/sciimmunol.abd5778\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLaughney, A. M. \u003cem\u003eet al.\u003c/em\u003e Regenerative lineages and immune-mediated pruning in lung cancer metastasis. Nat Med 26, 259\u0026ndash;269, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41591-019-0750-6\u003c/span\u003e\u003cspan address=\"10.1038/s41591-019-0750-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGanesan, A. P. \u003cem\u003eet al.\u003c/em\u003e Tissue-resident memory features are linked to the magnitude of cytotoxic T cell responses in human lung cancer. Nat Immunol 18, 940\u0026ndash;950, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/ni.3775\u003c/span\u003e\u003cspan address=\"10.1038/ni.3775\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEfremova, M., Vento-Tormo, M., Teichmann, S. A. \u0026amp; Vento-Tormo, R. CellPhoneDB: inferring cell-cell communication from combined expression of multi-subunit ligand-receptor complexes. Nat Protoc 15, 1484\u0026ndash;1506, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41596-020-0292-x\u003c/span\u003e\u003cspan address=\"10.1038/s41596-020-0292-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, T. \u003cem\u003eet al.\u003c/em\u003e TIMER2.0 for analysis of tumor-infiltrating immune cells. Nucleic Acids Res 48, W509-W514, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/nar/gkaa407\u003c/span\u003e\u003cspan address=\"10.1093/nar/gkaa407\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStuart, T. \u003cem\u003eet al.\u003c/em\u003e Comprehensive Integration of Single-Cell Data. \u003cem\u003eCell\u003c/em\u003e 177, 1888\u0026ndash;1902 e1821, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.cell.2019.05.031\u003c/span\u003e\u003cspan address=\"10.1016/j.cell.2019.05.031\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, X. \u003cem\u003eet al.\u003c/em\u003e CellMarker: a manually curated resource of cell markers in human and mouse. Nucleic Acids Res 47, D721-D728, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/nar/gky900\u003c/span\u003e\u003cspan address=\"10.1093/nar/gky900\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, Z. \u003cem\u003eet al.\u003c/em\u003e Integrated analysis of single-cell and bulk RNA sequencing data reveals a pan-cancer stemness signature predicting immunotherapy response. Genome Med 14, 45, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s13073-022-01050-w\u003c/span\u003e\u003cspan address=\"10.1186/s13073-022-01050-w\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCao, J. \u003cem\u003eet al.\u003c/em\u003e The single-cell transcriptional landscape of mammalian organogenesis. Nature 566, 496\u0026ndash;502, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41586-019-0969-x\u003c/span\u003e\u003cspan address=\"10.1038/s41586-019-0969-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBorcherding, N., Bormann, N. L. \u0026amp; Kraus, G. scRepertoire: An R-based toolkit for single-cell immune receptor analysis. \u003cem\u003eF1000Res\u003c/em\u003e 9, 47, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.12688/f1000research.22139.2\u003c/span\u003e\u003cspan address=\"10.12688/f1000research.22139.2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEssogmo, F. E. \u003cem\u003eet al.\u003c/em\u003e Cytokine Profile in Lung Cancer Patients: Anti-Tumor and Oncogenic Cytokines. Cancers (Basel) 15, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/cancers15225383\u003c/span\u003e\u003cspan address=\"10.3390/cancers15225383\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYeo, A. T. \u003cem\u003eet al.\u003c/em\u003e Single-cell RNA sequencing reveals evolution of immune landscape during glioblastoma progression. Nat Immunol 23, 971\u0026ndash;984, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41590-022-01215-0\u003c/span\u003e\u003cspan address=\"10.1038/s41590-022-01215-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSaito, S., Nakashima, A., Shima, T. \u0026amp; Ito, M. Th1/Th2/Th17 and regulatory T-cell paradigm in pregnancy. Am J Reprod Immunol 63, 601\u0026ndash;610, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/j.1600-0897.2010.00852.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1600-0897.2010.00852.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKaiko, G. E., Horvat, J. C., Beagley, K. W. \u0026amp; Hansbro, P. M. Immunological decision-making: how does the immune system decide to mount a helper T-cell response? Immunology 123, 326\u0026ndash;338, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/j.1365-2567.2007.02719.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1365-2567.2007.02719.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMetkar, S. S. \u003cem\u003eet al.\u003c/em\u003e Human and mouse granzyme A induce a proinflammatory cytokine response. Immunity 29, 720\u0026ndash;733, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.immuni.2008.08.014\u003c/span\u003e\u003cspan address=\"10.1016/j.immuni.2008.08.014\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHu, Z. W. \u003cem\u003eet al.\u003c/em\u003e Comprehensive analysis of ferritin subunits expression and positive correlations with tumor-associated macrophages and T regulatory cells infiltration in most solid tumors. Aging (Albany NY) 13, 11491\u0026ndash;11506, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.18632/aging.202841\u003c/span\u003e\u003cspan address=\"10.18632/aging.202841\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTewary, P. \u003cem\u003eet al.\u003c/em\u003e Granulysin activates antigen-presenting cells through TLR4 and acts as an immune alarmin. Blood 116, 3465\u0026ndash;3474, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1182/blood-2010-03-273953\u003c/span\u003e\u003cspan address=\"10.1182/blood-2010-03-273953\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eElizondo, D. M., Brandy, N. Z., da Silva, R. L., de Moura, T. R. \u0026amp; Lipscomb, M. W. Allograft inflammatory factor-1 in myeloid cells drives autoimmunity in type 1 diabetes. JCI Insight 5, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1172/jci.insight.136092\u003c/span\u003e\u003cspan address=\"10.1172/jci.insight.136092\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHori, S., Nomura, T. \u0026amp; Sakaguchi, S. Control of regulatory T cell development by the transcription factor Foxp3. Science 299, 1057\u0026ndash;1061, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1126/science.1079490\u003c/span\u003e\u003cspan address=\"10.1126/science.1079490\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2003).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFontenot, J. D. \u003cem\u003eet al.\u003c/em\u003e Regulatory T cell lineage specification by the forkhead transcription factor foxp3. \u003cem\u003eImmunity\u003c/em\u003e 22, 329\u0026ndash;341, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.immuni.2005.01.016\u003c/span\u003e\u003cspan address=\"10.1016/j.immuni.2005.01.016\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2005).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFontenot, J. D., Gavin, M. A. \u0026amp; Rudensky, A. Y. Foxp3 programs the development and function of CD4\u0026thinsp;+\u0026thinsp;CD25\u0026thinsp;+\u0026thinsp;regulatory T cells. Nat Immunol 4, 330\u0026ndash;336, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/ni904\u003c/span\u003e\u003cspan address=\"10.1038/ni904\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2003).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWen, T. \u003cem\u003eet al.\u003c/em\u003e NKG7 Is a T-cell-Intrinsic Therapeutic Target for Improving Antitumor Cytotoxicity and Cancer Immunotherapy. Cancer Immunol Res 10, 162\u0026ndash;181, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1158/2326-6066.CIR-21-0539\u003c/span\u003e\u003cspan address=\"10.1158/2326-6066.CIR-21-0539\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZheng, C. \u003cem\u003eet al.\u003c/em\u003e Landscape of Infiltrating T Cells in Liver Cancer Revealed by Single-Cell Sequencing. \u003cem\u003eCell\u003c/em\u003e 169, 1342\u0026ndash;1356 e1316, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.cell.2017.05.035\u003c/span\u003e\u003cspan address=\"10.1016/j.cell.2017.05.035\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNewman, J. H. \u003cem\u003eet al.\u003c/em\u003e IgG4-related disease: Association with a rare gene variant expressed in cytotoxic T cells. Mol Genet Genomic Med 7, e686, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/mgg3.686\u003c/span\u003e\u003cspan address=\"10.1002/mgg3.686\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZheng, Y. \u003cem\u003eet al.\u003c/em\u003e Immune suppressive landscape in the human esophageal squamous cell carcinoma microenvironment. Nat Commun 11, 6268, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41467-020-20019-0\u003c/span\u003e\u003cspan address=\"10.1038/s41467-020-20019-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSathe, A. \u003cem\u003eet al.\u003c/em\u003e GITR and TIGIT immunotherapy provokes divergent multi-cellular responses in the tumor microenvironment of gastrointestinal cancers. bioRxiv, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1101/2023.03.13.532299\u003c/span\u003e\u003cspan address=\"10.1101/2023.03.13.532299\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKumar, P., Saini, S., Khan, S., Surendra Lele, S. \u0026amp; Prabhakar, B. S. Restoring self-tolerance in autoimmune diseases by enhancing regulatory T-cells. Cell Immunol 339, 41\u0026ndash;49, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.cellimm.2018.09.008\u003c/span\u003e\u003cspan address=\"10.1016/j.cellimm.2018.09.008\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSobhani, N. \u003cem\u003eet al.\u003c/em\u003e CTLA-4 in Regulatory T Cells for Cancer Immunotherapy. Cancers (Basel) 13, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/cancers13061440\u003c/span\u003e\u003cspan address=\"10.3390/cancers13061440\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"genes-and-immunity","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"genes","sideBox":"Learn more about [Genes \u0026 Immunity](http://www.nature.com/gene/)","snPcode":"41435","submissionUrl":"https://mts-gene.nature.com/cgi-bin/main.plex","title":"Genes \u0026 Immunity","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-3879125/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3879125/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eDuring the process of NSCLC using TILs therapy, the heterogeneity of immune cell was revealed by using combined single-cell RNA (scRNA)/ T cell receptor (scTCR) sequencing -seq data from lung adenocarcinoma (LUAD) patients. Na\u0026iuml;ve CD4\u003csup\u003e+\u003c/sup\u003e T was increased in tumor tissue compared with circulating blood samples, activated signaling pathways were recognized, and \u003cem\u003eGZMA\u003c/em\u003e was identified as a potential novel diagnostic biomarker. The scTCR-seq repertoire was also investigated. At transition state, macrophages (\u003cem\u003eFTL\u003c/em\u003e) and dendritic (\u003cem\u003eAIF1\u003c/em\u003e) cells transferred the most CD3 TCR clones to T (\u003cem\u003eIL7R\u003c/em\u003e) cells, and cytotoxicity (\u003cem\u003eNKG7\u003c/em\u003e) transported to terminal exhausted (\u003cem\u003eCCL5\u003c/em\u003e) CD8\u003csup\u003e+\u003c/sup\u003e T cells. At transition and expansion state, T helper (\u003cem\u003eCXCL13\u003c/em\u003e) transported the most CD3 TCR clones to regulatory T (\u003cem\u003eFOXP3\u003c/em\u003e) cells. The expression profiling of cytokines, checkpoint receptors and their ligands during tumor progression were also investigated. T helper (\u003cem\u003eFTL, TNFRSF4\u003c/em\u003e and \u003cem\u003eTIGIT\u003c/em\u003e) and regulatory T (\u003cem\u003eCTLA4, TIGIT and FTL\u003c/em\u003e) show up at the initial stage of normal and metastatic samples, while cytotoxicity (\u003cem\u003eFGFBP2\u003c/em\u003e, \u003cem\u003eNKG7, PRF1\u003c/em\u003e and \u003cem\u003eCCL5\u003c/em\u003e) CD8\u003csup\u003e+\u003c/sup\u003e T cells still appears at the final stage of normal and metastatic samples. Taken together, our study provides the single cell level of TILs in NSCLC and offers treatment strategies to overcome drug resistance.\u003c/p\u003e","manuscriptTitle":"Single-cell sequencing reveals immune landscape of tumor-infiltrating lymphocytes (TILs) during non-small cell lung cancer (NSCLC) progression","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-03-27 17:44:18","doi":"10.21203/rs.3.rs-3879125/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"revise","date":"2024-09-12T10:09:14+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"This content is not available.","date":"2024-06-13T12:06:20+00:00","index":1,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2024-05-27T19:51:06+00:00","index":2,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2024-05-27T07:13:52+00:00","index":1,"fulltext":"This content is not available."},{"type":"reviewersInvited","content":"","date":"2024-03-24T04:03:40+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-01-24T13:50:43+00:00","index":"","fulltext":""},{"type":"submitted","content":"Genes \u0026 Immunity","date":"2024-01-23T09:38:12+00:00","index":"","fulltext":""},{"type":"checksFailed","content":"","date":"2024-01-22T15:40:36+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-01-19T15:20:56+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"genes-and-immunity","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"genes","sideBox":"Learn more about [Genes \u0026 Immunity](http://www.nature.com/gene/)","snPcode":"41435","submissionUrl":"https://mts-gene.nature.com/cgi-bin/main.plex","title":"Genes \u0026 Immunity","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"7dd58171-e5d2-4ecb-9c06-0f9d4503d341","owner":[],"postedDate":"March 27th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":29809770,"name":"Biological sciences/Immunology/Tumour immunology"},{"id":29809771,"name":"Biological sciences/Genetics/Sequencing/Next-generation sequencing"}],"tags":[],"updatedAt":"2025-05-06T07:08:15+00:00","versionOfRecord":{"articleIdentity":"rs-3879125","link":"https://doi.org/10.1038/s41435-025-00330-w","journal":{"identity":"genes-and-immunity","isVorOnly":false,"title":"Genes \u0026 Immunity"},"publishedOn":"2025-05-05 04:00:00","publishedOnDateReadable":"May 5th, 2025"},"versionCreatedAt":"2024-03-27 17:44:18","video":"","vorDoi":"10.1038/s41435-025-00330-w","vorDoiUrl":"https://doi.org/10.1038/s41435-025-00330-w","workflowStages":[]},"version":"v1","identity":"rs-3879125","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3879125","identity":"rs-3879125","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}