Metabolically activated and highly polyfunctional intratumoral VISTA+ regulatory B cells are associated with tumor recurrence in early stage NSCLC.

Metabolically activated and highly polyfunctional intratumoral VISTA+ regulatory B cells are associated with tumor recurrence in early stage NSCLC. | 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 Research Article Metabolically activated and highly polyfunctional intratumoral VISTA+ regulatory B cells are associated with tumor recurrence in early stage NSCLC. Domenico Lo Tartaro, Beatrice Aramini, Valentina Masciale, Nikolaos Paschalidis, and 22 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3891288/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 14 Jan, 2025 Read the published version in Molecular Cancer → Version 1 posted 13 You are reading this latest preprint version Abstract B cells have emerged as central players in the tumor microenvironment (TME) of non-small cell lung cancer (NSCLC). However, although there is clear evidence for their involvement in cancer immunity, scanty data exist on the characterization of B cell phenotypes, bioenergetic profiles and possible interactions with T cells in the context of NSCLC. In this study, using polychromatic flow cytometry, mass cytometry, and spatial transcriptomics we explored the intricate landscape of B cell phenotypes, bioenergetics, and their interaction with T cells in NSCLC. Our analysis revealed that TME contains diverse B cell clusters, including VISTA + Bregs, with distinct metabolic and functional profiles. Target liquid chromatography-tandem mass spectrometry confirmed the expression of VISTA on B cells. Pseudotime analysis unveiled a B cell differentiation process leading to a branch formed by plasmablasts/plasma cells, or to another made by VISTA + Bregs. Spatial analysis showed colocalization of B cells with CD4 + /CD8 + T lymphocytes in TME. The computational analysis of intercellular communications that links ligands to target genes, performed by NicheNet, predicted B-T interactions via VISTA-PSGL1 axis. Notably, tumor infiltrating CD8 + T cells expressing PSGL1 exhibited enhanced metabolism and cytotoxicity. In NSCLC patients, prediction analysis performed by PENCIL revealed the presence of an association between PSGL1 + CD8 + T cells and VISTA + Bregs with lung recurrence. Our findings suggest a potential interaction between Bregs and T cells through the VISTA-PSGL1 axis, able of influencing NSCLC recurrence. Bregs NSCLC recurrence prediction PSGL-1 VISTA Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction Non-small cell lung cancer (NSCLC) is the second most common cancer worldwide [ 1 ]. Specifically, the American Cancer Society estimates 238,340 individuals (120,790 women and 117,550 men) with a lung cancer diagnosis in 2023, and 127,070 of those individuals will die for this malignant disease. Surgery is considered the gold standard for early and locally advanced stages. However, resection is often not resolutive, even for local tumors, due to the high percentage of recurrence [ 2 ] that varies between 30% and 55% for NSCLC patients, with a 5-year overall survival rate between 22% and 18.1% in all the stages [ 3 ]. In recent years, immunotherapy has emerged as a significant advancement in the treatment of tumors, and, due to its durable effects and low rate of adverse reactions, has introduced a new paradigm for the treatment of several tumors [ 4 , 5 ]. Nonetheless, the majority of initially responsive patients eventually experience immune drug resistance. It has been shown that one of the main reason of low efficacy is connected with the quantity and features of tumor infiltrating lymphocytes (TIL) [ 6 ]. The tumor microenvironment (TME) of NSCLC is a complex milieu of immune cells, stromal elements, and signaling molecules that play a pivotal role in cancer progression and influence therapy response [ 7 , 8 ]. In this environment, the infiltration and organization of immune cells in tertiary lymphoid structures (TLS) is crucial for the prolonged survival of NSCLC patients [ 9 , 10 ]. Tumor-associated TLS are indeed a privileged site for T cell differentiation and activation. Furthermore, TLS are associated with T helper type-1 (Th1) and cytotoxic immune signatures in lung, breast, and gastric cancers, indicating that they imprint the local immune microenvironment [ 11 – 13 ]. Among the diverse array of immune cells populating the TME, B cells have emerged as central players, orchestrating immune responses that profoundly influence tumor fate [ 14 , 15 ]. The repertoire of tumor-infiltrating B cell phenotypes includes naive, activated, and memory B cells, germinal center B cells, and plasma cells [ 16 ]. While generally aligning with canonical B cell subsets identified in healthy donors, tumor-infiltrating B cells demonstrate greater diversity within certain subsets, particularly within the memory compartment [ 17 ], making B cells crucial immune response modulation to tumors and lymphoid malignancies. In particular, regulatory B cells (Bregs) constitute a newly designated subset of B cells able to regulate immune responses in inflammation and more recently, cancer [ 18 – 20 ]. However, the prognostic value of B cells and related subsets together with their mechanisms of action for immune modulation in the TME are still a matter of debate. In recent years, along with PD-1 and CTLA-4, other immune modulators able to rewire the immune response within the TME have been taken into account. Indeed, V-domain Ig suppressor of T-cell activation (VISTA), an inhibitory immune checkpoint molecule, and its ligand P-selectin glycoprotein ligand-1 (PSGL-1, also known as CD162) multifunctional glycoprotein, have garnered attention for their roles in modulating immune cell function, angiogenesis, and tumor progression [ 21 – 31 ]. Here, we present a comprehensive analysis delineating the phenotypic, functional, and metabolic attributes of interacting B and T cells in patients with resectable NSCLC. Our findings highlight that a high frequency of tumor-infiltrating VISTA + Bregs together with high percentage of PSGL-1 expressing T cells are more represented in patients with cancer recurrence, and can likely predict this event. Results Bregs express VISTA and are enriched in the tumor microenvironment. In the TME, different subpopulations of B cells reside that can either promote or hamper tumor growth [ 18 , 32 , 33 ]), hence affecting patient outcomes [ 34 ]. For this reason, to directly assess the landscape of tumor-infiltrating CD19 + B cells, we used a 19-parameter flow cytometry panel. We investigated 472,627 cells isolated from resected tumor ( n = 15) paired with 538,862 cells isolated from peripheral blood ( n = 15), considered as reference. The percentage of B cells was higher in the TME when compared to blood (median: blood = 4.6, tumor = 15.8; Supplementary Fig. 1). To better characterize the landscape of CD19 + B cells, an unsupervised algorithm such as FlowSOM was used, and 20 clusters were identified (Fig. 1 A-B). Hierarchical ordering of FlowSOM clusters (rows) and markers (columns) allowed grouping of subpopulations with similar immune characteristics. Three of these clusters (C3, C4, C10) were more abundant in the peripheral blood and displayed characteristics of naïve cells (i.e., CD27 − IgD + IgM + ). One of them (C10) was identified as proliferating naïve cells expressing Ki-67. The percentages of C2, C1, C5, C12, C15 representing respectively: transitional B cells (Tr, CD27 − IgD + IgM + CD24 ++ ; in C2), class-unswitched memory B cells (MBC usw, i.e., CD27 + IgD + IgM + ; in C1 and C5) and plasmablasts (PB, CD27 ++ CD38 ++ Ki-67 + , in C12 and C15) were higher in blood when compared to TME, except for C1, which exhibited similar proportions in both blood and TME. Those in C6, C7, C13 were identified as atypical B cells (atBC, i.e., CD21 − CD27 − ) originally found in tonsils and in peripheral blood in conditions of chronic antigen stimulation, or autoimmune disease [ 35 ]. C6 was more represented in the peripheral blood, while the percentage of C7 and C13 was higher in the TME. Clusters of class-switched memory B cells (MBC sw, i.e., CD27 + IgD − IgM − ), including C11, C14, C9, and C8, were mostly represented in TME, suggesting that within mature TLS these cells might be not only activated but also in germinal center have experienced the antibody class switching and somatic hypermutation [ 36 ]. C19 was identified as proliferating VISTA + B cells (CD38 + CD138 dim Ki67 + IL-10 dim ) and its percentage was higher in blood if compared to TME. Finally, two populations of B cells expressing VISTA and IL-10 were identified (C18 and C20) and classified as B regulatory cells (hereafter referred to as ‘VISTA + Bregs’). For the first time, we reported here that a cluster of Bregs expresses VISTA. We confirmed this data by using two different approaches: first, we performed manual gating of flow cytometry data to ascertain VISTA expression on Bregs, and we showed that not only Bregs express VISTA, but also they were much more represented in the TME (Fig. 1 C-D). Then, by targeted Liquid Chromatography with Tandem Mass Spectrometry (LC-MS/MS) analysis on sorted B cells, we confirmed that VISTA is expressed by peripheral blood B cells, despite at a lower level compared to what was observed within monocytes (Fig. 1 E and Supplementary Fig. 2). Finally, by trajectory inference, also known as pseudotime analysis, we investigated the progression along the differentiation trajectory (Fig. 1 F). As expected, the differentiation process started from naïve (C3) and transitional B cells (C4); then, the trajectory revealed two distinct cellular branches: one that gave origin to plasmablasts and plasma cells (C12, C17) and the other to VISTA + Bregs cell clusters (C18 and C20), suggesting that the differentiation program that they experience is quite different from that of all other B cells. VISTA + Bregs display a metabolically activated profile and are highly polyfunctional. Given that cell function is strictly connected to metabolic profile, we employed single-cell metabolic regulome profiling (scMEP) using mass cytometry by time of flight (CyTOF) to investigate the metabolic profile of B cells within TME [ 37 ] [ 38 ]. This analysis was performed on ex vivo isolated TILs. We performed this unsupervised analysis by focusing on the 19 metabolic markers expressed by B cells (GLUT1, GAPDH, LDHA, HK2, PFKB4, G6PD, CytC, CS, IDH1, ATPA5, CD98, GLUD1/2, CD36, CPT1A, VDAC1, pACC, pPGC1α, HIF-1α, pNRF2), and we found a total of 9 scMEP states (Fig. 2 A-C). First of all, we noticed that the majority of tumor infiltrating B cells (81% of total B cells) were grouped into two states of metabolically quiescent cells (scMEP1 and 8). However, a total of 14% of cells were grouped into five states (scMEP3, 4, 5, 7 and 9) characterized by highly activated metabolism. Indeed, these states showed high level of expression of proteins belonging to glycolysis pathway (such as PFKB4 and HK2), tricarboxylic acid cycle (TCA; such as CytC and CS), amino acid pathway (such as CD98 and GLUD1/2), fatty acid oxidation (FAO; such as pACC and CPT1A), mitochondrial dynamics (such as VDAC1), pentose phosphate pathway (ppp; such as G6PD) and transcription of master regulators such as hypoxia-inducible factor-1α (HIF-1α). In particular, cells in scMEP5 state showed a high level of CD36 expression, indicating the import of fatty acids and activation of the fatty acid oxidation. scMEP2 and 6, which account for 4.8% of cells, showed mild activated metabolism, characterized by low-middle level of expression of protein related to TCA (such as CytC, ATPA5 and CS), amino acid pathway (CD98) and FAO (CD36 and pACC) accompanied by increased levels of GLUT1 (scMEP 2). In order to match the metabolic profiles to different cell phenotypes, we took into account the phenotypic markers and we found that almost all metabolically active B cells (scMEP 9, 5, 3, 7, 4, 6) expressed high levels of CD39, PSGL-1 and VISTA (Fig. 2 D-E). In particular, scMEP5 and 3 state were composed by regulatory B cells expressing high level of IL-10 and VISTA, and medium level of Ki-67 (Fig. 2 D-E). scMEP 4, 7 and 9 were mainly formed by activated B cells expressing CXCR6, a marker of lung residency, and CD38, whereas scMEP 2 and 6 were composed by cells negative for CXCR6, CD38 and VISTA (Fig. 2 D-E). Finally, all metabolically quiescent B cells (scMEP 1 and 8) were characterized by the lack of expression of all the aforementioned markers except CD73, a costimulatory molecule of B cells [ 39 ]. These results are in line with the clustering performed by using both lineage markers as well as metabolic ones (Supplementary Fig. 3). B cells are antibody-producing cells driving humoral immune responses to foreign antigens. However, B cells secrete different cytokines, which influence both pro- and anti-inflammatory immune responses. Indeed, the heterogeneity in cytokine-driven responses by B cells can range from the production of pro-inflammatory molecules such as IL-6 to the release of the immunosuppressive IL-10 [ 40 ]. For this reason, we investigated the functional profile of B cells expressing VISTA (see Methods). First, we showed that VISTA was upregulated on B cell after stimulation but not in hypoxic condition (Supplementary Fig. 4A-C). The percentage of cells expressing GM-CSF, TNF, IL-6, IL-10, IFN-γ, and TGF-β was higher in VISTA + when compared to VISTA − B lymphocytes (Fig. 2 F and Supplementary Fig. 4D-E), suggesting that these cells are much more functional and able to produce anti-inflammatory and pro-inflammatory molecules than their counterpart. The analysis of their capability to simultaneously produce different molecules, i.e. their polyfunctionality, revealed that a high percentage of VISTA + B cells were capable of produce simultaneously three (GM-CSF, TGF-β, and TNF) or four (GM-CSF, IL-10, TGF-β, and TNF) cytokines (Fig. 2 G). The broader range of cytokine production by VISTA + B cells could evidence a novel, potential role in modulating tumor-specific T cell response. B and T lymphocytes localize in the TLSs and potentially interact through VISTA-PSGL-1 axis. VISTA is the ligand of PSGL-1 in acidic pH such as that found in the TME [ 41 ]and it is known to be expressed also on monocytes and T cells [ 42 ]. For this reason, to investigate the potential role of VISTA in mediating interactions between B cells and other immune cells within the tumor microenvironment (TME), we analyzed a publicly available dataset of imaging-based spatial transcriptomics in NSCLC [ 43 ]. Here, we were able to identify, beside tumor cells (expressing KRT17) , macrophages ( C1QA) , T cells (CD3E ), B cells (MS4A1 ), and smooth muscle cells (TAGLN ) (Supplementary Fig. 5A). Notably, overall B and T cells tend to cluster together, forming the ectopic structures commonly referred to as TLS (Fig. 3 A-B, Supplementary Fig. 5B). Analyzing the Euclidean distance between cells, we observed that B cells and CD4 + T cells, with CD8 + T cells to a lesser extent, were in close proximity compared to the other cell types (Fig. 3 C). Then, to evaluate possible cell-to-cell contact between B cells other subpopulation within the TME, we calculated the number of cell-to-cell contacts. We observed CD4 + T cells exhibited a higher frequency of interactions with B cells and, to a lesser extent, with CD8 + T cells and macrophages. This suggests that B cells and CD4 + T cells tend to preferentially colocalize and likely interact within the TME (Fig. 3 C and Supplementary Fig. 5C). Thus, we further investigated what ligand–target links regulate potential crosstalk between T cells and B cells in the TME using NicheNet. In this analysis, the B cell cluster derived from imaging-based spatial transcriptomics (except plasmacells) was designated as the sender population, while the conventional CD4 + or CD8 + T cell clusters were identified as separate receiver populations. For each T cell subset, we focused our analysis on the top 45 ligand–receptor pairs. Among the well-known B and T ligand–receptor axis such as: CD40 via CD40LG , CD86 via CD28 , TGFB1 via TGFBR2 , NicheNet predicted a B-T interaction for CD4 + or CD8 + T cells via VSIR - SELPLG (Fig. 3 D, Supplementary Fig. 6) where VSIR codes for V-domain Ig suppressor of T cell activation (VISTA), while SELPLG codes for P-selectin glycoprotein ligand-1 (PSGL-1, also known as CD162). Moreover, we investigated if this predicted interaction could alter the gene-expression profile related to tumor-reactivity. We observed, both for CD4 + and CD8 + T cells, elevated potential regulatory scores among the 45 top-ranked ligands and following target genes: FASLG, IFNG, DDIT4, TNFRSF18 (GITR) and CTLA4 (Fig. 3 D, Supplementary Fig. 6). To further validate these findings, we analyzed three different scRNA-seq datasets of patients with early-stage NSCLC (Supplementary Fig. 7A-B). For each T cell subset, we focused our analysis on the top 20 ligand-receptor pairs. Remarkably, we found consistent results across both imaging-based spatial transcriptomics and scRNA-seq datasets, both in terms of ligand-receptor interactions and potential gene regulatory scores (Supplementary Fig. 7C-D). Given that B and T cells specifically co-localize and potentially interact within the TLS regions, we assessed the expression of VSIR and SELPLG within five different TLS (boxed in the Fig. 3 A). Our observations showed that VSIR expression was slightly higher on B cells, whereas SELPLG was more pronounced on both CD4 + and CD8 + T cells. Moreover, the percentage of B cells co-expressing both VSIR and IL10 was higher than that of CD4 + and CD8 + T cells (Fig. 3 E). These data offer evidence that B cells expressing VISTA and/or IL-10 are present within TLSs and may potentially interact with T cells through PSGL-1. Tumor infiltrating CD4 + T RM cells expressing PSGL-1, but not PD-1 display an activated metabolic profile. Given that VISTA likely interacts with PSGL-1 expressed by T cells, we investigated by flow cytometry the phenotype of CD4 + T cells expressing PSGL-1. We showed that the frequency of CD4 + T cells among total CD45 + CD3 + live T cells was lower in the TME if compared to blood (median blood = 67.6, tumor = 48.0; Supplementary Fig. 8). We concatenated 5.946.764 of CD4 + T cells (PBMC = 5.108.956 and TILs = 837.808) from 15 samples and analyzed them with FlowSOM. We identified 30 different clusters, whose frequencies are subsequently determined in each sample type. In particular, the dimensional reduction performed by Uniform Manifold Approximation and Projection (UMAP) showed that these clusters were distributed in six different regions (Fig. 4 A). Ten of these clusters belonging to ‘not exhausted’ cells (C1, C2, C3, C5, C6, C8, C13, C16, C17 and C21) were more abundant in the blood and displayed a phenotypic identity coincident with naïve (CD45RA + CCR7 + ; C3), central memory (CD45RA + CCR7 − ; C1, C2), tissue-resident memory-like T cells (T RM -like, i.e. , CD69 − CD103 + ; C13), effector memory (CD45RA − CCR7 − ; C5, C6, C8, C16) or cytotoxic (CD45RA dim CCR7 − GZMB + ; C17, C21) T cells. These clusters were characterized by lower levels of CTLA4, PD-1, CD96, and CD39, accompanied by higher expression of PSGL-1 (Fig. 4 A-B). Additional clusters of cells that were not in an exhausted state, such as C10 and C11, were characterized by the expression of CD69, and were notably more prevalent within the TME (Fig. 4 B). The percentages of proliferating CD39 + CD4 + T cells expressing PSGL-1, PD-1 and CTLA-4 (C26), were higher in the TME (Fig. 4 B). The remaining two clusters of likely proliferating CD4 + T cells (C4 and C7) were similar in blood and TME (Fig. 4 B). The complete absence of ‘not exhausted’ clusters within the tumor was counterbalanced by the presence of virtually unique subsets characterized by heterogeneous expression of PSGL-1 along with high levels of PD-1 and CTLA-4. These subsets included tissue-resident memory T cells (T RM , marked by CD69 + CD103 + ; C30), conventional CD39 − CD4 + T cells (Tconv CD39 − , CD39 − CD69 + CD103 − ; C12, C20, C24, C25, C15, C18, C9, C19, C14), and Tconv CD39 + CD4 + T cells (Tconv CD39 + , CD39 + CD69 + CD103 − ; C22, C23, C29, C27, C28). To identify the major phenotypic markers distinguishing tumor-infiltrating CD4 + T cells expressing PSGL-1 from those not expressing it, we manually evaluated the expression of thirteen immune markers, including T-bet, CD96, CD226, PD-1, CTLA-4, CD161, CD49a, CXCR6, Ki-67, GZMB, CCR7, CD45RA and CD39. We observed that CD4 + T cells expressing PSGL-1 displayed higher levels of GZMB, CD226, CXCR6, T-bet, CD103, CD39 and CD96, while CCR7 and CD69 were expressed predominantly by PSGL-1 negative cells (Fig. 4 C). The metabolic investigation of tumor infiltrating CD4 + T cells by the unsupervised analysis revealed 15 distinct scMEP states. In particular, 83.5% of CD4 + T cells were grouped into metabolically quiescent states (scMEP 3, 2, 15, 10 and 13), characterized by a low level of all metabolic pathway regulators such as GLUT1, HK2, LDHA, CytC, IDH1, CD36 and HIF-1α (Fig. 4 D-F). The remaining 16.5% of cells clustered into hyperactivated (scMEP 9, 5, 12, 4 and 8) to low-intermediate states (scMEP 1, 11, 14, 6 and 7) ( Fig. 4 E ) . Hyperactivated scMEP 9, 5 and 12 states were characterized by high level of almost all metabolic markers, except for GAPDH, meaning that glycolysis, pentose, TCA, amino acid metabolism as well as FAO were activated ( Fig. 4 E ) . ScMEP 4 and 8 displayed intermediate levels of TCA (such as IDH1 and ATP5A) and high level of FAO (such as CD36) paired with the complete absence of glycolytic regulator (such as GLUT1 and LDHA) (Fig. 4 E-F). ScMEP 1, 11, and 14 displayed intermediate metabolisms without the expression of CD36, except for scMEP 1, which represented 0.27% of CD4 + T cells showing upregulation of GLUT1 and HK2 ( Fig. 4 E-F ) . ScMEP 6 and 7 collectively were 11.9% of cells, both exhibiting minimal expression of metabolic regulators, except for CD98 and pPGC1α, likely suggesting that these cells primarily depend on basal levels of TCA and the FAO pathway (Fig. 4 E ) . As far as immunologic characteristics were paired to metabolic states, scMEP clusters clearly identified distinct immunological phenotypes. All metabolic quiescent cells (scMEP 3, 2, 15, 10 and 13) displayed features of resting cells (CD38 − , Ki-67 − , CD39 − and PD-1 low ) with Tconv phenotype (CD103 − ) (Fig. 4 G). On the other hand, all metabolically active scMEPs, except for scMEP 14, 6 and 7, belong to T RM phenotype (CD103 + ) and expressed PSGL-1 but not PD-1(Fig. 4 G-H). In particular, the metabolically hyperactivated state consisted of double positive CD4 + CD8 + T cells (DP; scMEP 9 and 5) or Tregs expressing IL-10 (scMEP 12) (Fig. 4 G-H). These findings align with our observation that nearly all DP cells within the tumor microenvironment were in an activated state (CD69 + CD103 − ) and undergoing proliferation (Ki-67 + ), although they exhibited lower cytotoxicity compared to their counterparts in the bloodstream (Supplementary Fig. 9). Cells exhibiting low to intermediate metabolic activity were composed by activated double-negative T cells (DN) and Treg cells expressing high levels of PD-1, CD39, and CXCR6 — a marker associated with the survival and localized expansion of effector T cells within TME [ 44 ]. These results are in line with the clustering performed by using both lineage markers as well as metabolic ones (Supplementary Fig. 10). PSGL-1-positive tumor infiltrating CD8 + T cells shows metabolic activation and enhanced cytotoxic profile. As we did for CD4 + T cells expressing PSGL-1, we assessed the heterogeneity of tumor infiltrating CD8 + T cells. The percentage of CD8 + T cells was higher in TME if compared to blood (median Blood = 23.0, median Tumor = 31.4, Supplementary Fig. 11). To further explore all the CD8 + T cells subset, we took advantage of unsupervised data analysis using FlowSOM. We investigated a total of 2.414.533 cells of which 613.023 cells from resected tumor ( n = 15, TILs) paired with 1.801.510 cells isolated from peripheral blood ( n = 15, PBMC). We identified 30 different CD8 + clusters, whose frequencies were subsequently determined in each sample type. In particular, the dimensional reduction performed using UMAP algorithm showed that these clusters were distributed in three different areas, called as ‘not exhausted’, ‘Tconv and ‘T RM CD39 + and CD39 − ’ (Fig. 5 A). Among the nine ‘not exhausted’ clusters, seven clusters (C1, C2, C4, C5, C6, C7, C18) were mostly represented within the blood and displayed a phenotypic identity coincident with naïve (CD45RA + CCR7 + ; C18), central memory (CD45RA − CCR7 + ; C5), effector memory (CD45RA − CCR7 − ; C1, C4, C6, C7) or effector memory re-expressing CD45RA (CD45RA + CCR7 − ; C2) T cells (Fig. 5 B). Similar to not exhausted CD4 + T cells, not exhausted CD8 + T cells exhibited reduced levels of CTLA4, PD-1, CD96, and CD39, along with heightened expression of PSGL-1. Other clusters belonging to Tconv (C8, C10, C11, C14, C15, C16, C17, C19, C21, C23, C27, C28), T RM CD39 + (C9, C12, C20, C22, C30) or T RM CD39 − (C25, C26, C29) cells, were enriched in TME and displayed varying levels of PSGL-1 (Fig. 5 B). Similar percentages of C3, C13 and C24 were observed in TME and blood (Fig. 5 B). As we did for tumor infiltrating CD4 + T cells, to identify the major phenotypic marker distinguishing tumor-infiltrating CD8 + T cells expressing PSGL-1 from those not expressing it, we manually evaluated the expression of thirteen markers, including T-bet, CD96, CD226, PD-1, CTLA-4, CD161, CD49a, CXCR6, Ki-67, GZMB, CCR7, CD45RA and CD39. We reported that CD8 + T cells expressing PSGL-1 displayed higher levels of GZMB, CD226, CXCR6, T-bet and CD96 (Fig. 5 C). These data, together with those of tumor infiltrating CD4 + T cells, establish GZMB, CD226, CXCR6, T-bet and CD96 as core surface markers that distinguish human PSGL-1 + and PSGL-1 − subsets within TME. Investigating the metabolic profiles of CD8 + T cells infiltrating tumors, we identified a total of 13 distinct scMEP states. Among these metabolic states, CD8 + T cells exhibited subsets characterized by varying degrees of metabolic protein expression. scMEP 1, 2, 3, and 5 displayed low level of expression metabolic proteins, suggesting reduced metabolic activity while scMEP 13, 12, and 9 displayed elevated expression of a wide array of metabolic regulators, including HK2, GAPDH, LDHA, G6PD, CytC, IDH1, CD98, CD36, and pPGC1α, indicating increased metabolic activity (Fig. 5 D-F). scMEP 8 and 10 were characterized by high expression of CD98, scMEP 11 by high level of GLUT1, scMEP 4 was distinguished by high level of expression of G6PD, scMEP 7 by GAPDH and scMEP 6 by high expression of GAPDH/PFKB4 (Fig. 5 D-F). These specific patterns of expression could suggest distinct metabolic adaptations influenced by the TME. As far as phenotypic landscape has been considered, metabolically quiescent states (such as scMEP 1, 2, 3, and 5) were Tconv resting cells (CD103 − , CD69 − , ICOS − , CD38 − , Ki-67 − , and PD-1 low ) (Fig. 5 G-H). scMEP 13, 12, and 9, which demonstrated heightened overall expression of metabolic proteins, displayed signs of activation, including CD69 + , ICOS + , CD38 + , Ki-67 + expression, and a T RM phenotype, underscoring the augmented metabolic demands of actively cycling cells. Furthermore, these latter scMEP states exhibited elevated levels of PSGL-1, CD39, CD103 and PD-1 (as shown in Fig. 5 G-H). Additionally, we observed the presence of cells with distinctive metabolic characteristics (scMEP 8, 10, 6, 7, 11, and 4). All these cells exhibited an intermediate PD-1 level alongside reduced mitochondrial capacity, decreased PSGL-1 expression, and diminished pPGC1α, which are hallmark features of T cell exhaustion (Fig. 5 G-H). These results are in line with the clustering performed by using both lineage markers as well as metabolic ones (Supplementary Fig. 12). PSGL-1 + T cells display heightened antitumor capacity. Due to its proposed role in regulating T cell trafficking within lymphoid organs and immune responses, we investigated the effector capabilities of tumor-infiltrating CD4 + and CD8 + T cells expressing PSGL-1. Our findings demonstrated that both conventional CD4 + (CD103 − FoxP3 − ; Tconv) and tissue-resident memory (T RM ; CD103 + FoxP3 − ) T cells expressing PSGL-1 + exhibited higher cytotoxicity compared to their PSGL-1 − counterparts (Fig. 6 A-C). Concerning cytokine production, we noted that PSGL-1 + T RM cells were more frequently producing TNF, while IL-2 and IFN-γ were equally produced (Fig. 6 D). There were no observed differences in IL-17 production. The polyfunctionality of both PSGL-1 + and PSGL-1 − cells was comparable, although T RM cells exhibited greater polyfunctionality compared to Tconv cells, as illustrated in Fig. 6 E. We next focused on CD8 + T cells, observing also in these case that T cells expressing PSGL-1 + exhibited higher cytotoxicity compared to their PSGL-1 − counterparts (Fig. 6 F-H). Furthermore, PSGL-1 + Tconv seemed be more cytotoxic compared to PSGL-1 + T RM (Fig. 6 G-H). In terms of cytokine production PSGL-1 + cells looked be able to produce less IL-2 among the two subsets, while IFN-γ were higher in PSGL-1 + cells. TNF production remained consistent between PSGL-1 − and PSGL-1 + cells, while no differences were observed for IL-17 production (Fig. 6 I). Analyzing the polyfunctionality, PSGL-1 + appeared be more polyfunctional compared to their PSGL-1 − counterparts, because enriched of cells able to produce simultaneously IL-2 and IFN-γ (T RM ) or IFN-γ and TNF (Tconv) (Fig. 6 J). Supervised learning framework identified clusters related to cancer relapse after surgery in early-stage NSCLC patients. To define which cell populations positively or negatively drive NSCLC recurrence, we performed a prediction analysis using PENCIL, a novel supervised learning framework [ 45 ]. Five out of fifteen patients examined with flow cytometry experienced tumor recurrence within 12 months post-surgery. By targeting patients' clinical recurrence, PENCIL identified cells associated with tumor recurrence with an accuracy of 82% for CD4 + T cells, 79% for CD8 + T cells, and 80% for B cells (Supplementary Fig. 13A-C). Our analysis indicated that patients who experienced lung recurrence after surgery exhibited high percentages of circulating cytotoxic CD4 + cells expressing PSGL-1 (C17), CD4 + Tconv expressing PSGL-1 (C15), CD8 + T RM expressing PSGL-1 (C26), and VISTA + Bregs (C20) (Fig. 7 A-F). In contrast, patients who remained recurrence-free for 12 months post-surgery displayed high percentage of PSGL-1 − CD4 + T RM (C30) and PSGL-1 − CD8 + T RM (C12, C22, C25) (Fig. 7 C-D). Additionally, memory-unswitched B cells (C1) were elevated in both blood and tumor samples of patients without recurrence, whereas memory-switched B cells (C14) showed an increase exclusively in the tumor site of patients without recurrence (Fig. 7 E-F). To validate PENCIL prediction, we employed the multivariate Cox proportional-hazards model. In this analysis, high percentage of VISTA + Bregs (C20, [HR = 1.41; 95% CI = 0.98, 2.03]) within the TME, along with high percentage of CD8 + T RM expressing PSGL-1 (C26, [HR = 1.16; 95% CI = 0.84, 1.45]), circulating cytotoxic CD4 + cells expressing PSGL-1 (C17, [HR = 1.09; 95% CI = 0.96, 1.23]), and reduced levels of PSGL-1 − CD4 + T RM (C30, [HR = 0.64; 95% CI = 0.40, 1.01]), were predictive of an elevated risk of tumor recurrence within 12 months post-surgery (Fig. 7 G). Discussion In this study, we unraveled B and T cell landscape of the NSCLC tumor microenvironment, providing insights into their interactions, metabolic activities, and their influence on cancer recurrence post-surgery. Indeed, we unveiled for the first time not only the presence of B cells expressing VISTA within the TME of NSCLC, but also that VISTA + B cells belong to the category of Bregs. Furthermore, trajectory inference and pseudotime score revealed that VISTA + Bregs represent a unique cell fate state, branching out separately from plasma cells. Furthermore, we found that, from a bioenergetics point of view, VISTA + Bregs were characterized by high metabolic activities, including glycolysis, TCA, FAO, amino acid metabolism, and PPP, suggesting that this unique metabolic signature points toward an enhanced energy production and utilization within these cells regulated by transcription factor HIF-1α. Indeed, metabolic programs play a crucial role in the entire process of B cell development, as metabolic nutrients are essential not only for the acquisition of immune cell function, but also for the transition from effector immune cells to regulatory immune cells [ 46 , 47 ]. From a functional point of view, cytokines such as IL-10, IL-6, GM-CSF, TNF and TGF-β may drive T cell immune suppression and/or tumor progression [ 48 – 52 ], and here we show that VISTA + B cells produce various cytokines, including IL-10, TGF-β, IL-6, TNF, and GM-CSF, pointing out their immunosuppressive capacity and pro-tumor activity [ 53 ]. We unveiled that B and T cells are prevalently distributed in the TLS, but are rarely found in direct contact with tumor cells. Moreover, B cells exhibited a higher number of cell-to-cell interactions with CD4 + T cells and to a lesser extent with CD8 + T cells and macrophages, suggesting a potential role of B cells in influencing preferentially the CD4 + T cells anti-tumor immune response. Spatial colocalization of B and T cells in tumor TLS revealed a novel interaction axis between B and T cells within the TME involving VISTA and PSGL-1, alongside established interactions like CD40-CD40L and ICOS-ICOSL. This finding is consistent with recent research demonstrating the selective engagement of VISTA by PSGL-1, particularly under the acidic pH conditions commonly encountered in the TME [ 41 ]. These results suggest that, in conjunction with Tregs, VISTA + Bregs may play a role in suppressing antitumor T cell responses. This suppression may occur not only through release of immunosuppressive cytokines, but also via cell contact-dependent immune suppression mediated by the VISTA-PSGL-1 axis. Beyond adhesion, PSGL-1 has been shown to be implicated in T cell signaling pathways, influencing cytokine production and immune suppression within the TME [ 30 ]. Furthermore, recent data support that the therapeutic blockade of PSGL-1 may promotes T cell responses and melanoma tumor control [ 24 ]. Here we observed that 30% of tumor-infiltrating T cells, spanning from Tconv to T RM cells, were positive for PSGL-1 expression. Notably, these PSGL-1-positive cells exhibited a distinct phenotype compared to PSGL-1-negative counterparts. Specifically, both CD4 and CD8 T cell compartments expressing PSGL-1 shared the expression of a core set of markers, including GZMB, CD226, CXCR6, T-bet, and CD96. The expression of these markers has been associated with increased T cell activation, cytotoxicity, and lung-tissue residency [ 54 – 58 ]. Our data also revealed the heightened antitumor potential of PSGL-1 + CD4 and CD8 T cells, both in Tconv and T RM cells. These PSGL-1-positive cells exhibited enhanced production of cytotoxic molecules such as GNLY and GZMB. Additionally, they displayed distinct cytokine production patterns, with PSGL-1-positive CD4 + T RM cells secreting higher levels of IL-2 and TNF, while PSGL-1-positive CD8 + T RM cells and Tconv cells produced increased IFN-γ but lower IL-2. These results were in line with their metabolic profile. Indeed, PSGL-1 + tumor-infiltrating CD4 + and CD8 + cells exhibited hyperactivated metabolic profiles and were enriched in T RM cells. These cells showcased elevated metabolic pathways, including glycolysis, TCA, FAO, amino acid metabolism, and PPP. This heightened metabolic state suggests an increased energy demand, driven by their enhanced effector capabilities in the tumor microenvironment, as we observed. While adjuvant chemotherapy has proven beneficial for improving overall survival in NSCLC patients with stage IIA to IIIA disease, those with stage I tumors smaller than 4 cm do not typically receive adjuvant chemotherapy [ 59 ]. Even after complete lung tumor resection, a considerable proportion of these patients face the risk of relapse [ 60 ]. Our computational analyses, employing PENCIL supervised learning and Cox multivariate analysis, predicted specific immune cell clusters associated with cancer recurrence post-surgery. VISTA + tumor infiltrating Bregs (C20), circulating cytotoxic CD4 + cells expressing PSGL-1 (C17), and CD8 + T RM cells expressing PSGL-1 (C26) emerged as key clusters linked to disease recurrence. The presence of these cells is strongly associated with an increased risk of tumor relapse, highlighting their potential as prognostic markers. Collectively, these data suggest that while PSGL-1-expressing T cells exhibit highly effective antitumor capabilities, their proximity and cell-to-cell interaction with VISTA + Bregs, particularly when present at high frequencies, may lead to their suppression within the TME. Despite that, the mechanism of B and T cell interaction may be more heterogeneous. Indeed, it was recently reported that up to 60% of human Bregs express PSGL-1. Notably, PSGL-1 downregulation was observed in patients with systemic sclerosis and was correlated with a reduced extent of IL-10 production by Bregs [ 61 ]. This implies that the binding of PSGL-1 on VISTA + B cells by T cells expressing VISTA could act as a mechanism to reduce IL-10 production, serving as a brake on Bregs-mediated immune suppression. In summary, high density of VISTA + Bregs in the TLS could be associated with the tumor recurrence in NSCLC patients. Therefore, markers such as VISTA or PSGL-1 may represent a new potential therapeutic target for those patients who develop tumor recurrence or who do not respond well to anti-PD-1 or anti-CTLA-4 therapy. Methods Study design As part of the project, we recruited a cohort comprising 37 patients diagnosed with NSCLC at stages IA, IB, IIA, IIB, and IIIA. These patients were admitted to either Azienda Ospedaliero Universitaria di Modena and Reggio Emilia (University Hospital) in Modena or G.B. Morgagni—L. Pierantoni Hospital in Forlì. Detailed clinicopathologic characteristics of the patients can be found in Table 1 . The average age of the entire cohort was 71.4 ± 7.2 years. Among these patients, five had a prior history of cancer, as outlined in Table 1 . Additionally, seven patients experienced a relapse post-surgery. Importantly, none of the patients had undergone chemotherapy, radiation therapy, or any other anti-tumor treatments before tumor resection. The study protocol was approved by the local Ethical Committes (“Area Vasta Emilia Nord”, protocol number 0018007/21; “Comitato Etico della Romagna”, protocol number 7043/2021). All participants involved in this study provided written informed consent for both sample collection and data analyses. Detailed clinical information and the analysis methods employed for each sample are reported in Supplementary Table 1 . Blood collection and isolation of mononuclear cells Up to 30 mL of blood were collected from each patient in vacuettes containing ethylenediamine-tetraacetic acid (EDTA). Blood was immediately processed. Isolation of peripheral blood mononuclear cells (PBMC) was performed using ficoll-hypaque (Sentinel Diagnostics, Lymphoprep) according to standard procedures [62]. PBMC were stored in liquid nitrogen in fetal bovine serum (FBS) supplemented with 10% dimethyl sulfoxide (DMSO). Plasma was stored at −80 °C until use. Isolation of leucocytes from solid human tissues After informed consent, tumor tissue from patients undergoing surgery was collected. Primary surgical samples were transported from the operating room to the laboratory at controlled temperature of 4° C. Biopsy was washed twice with 30 ml of 1 × Dulbecco′s Phosphate Buffered Saline (D-PBS; Gibco) in a 50 ml Falcon tube. Then the tumor tissue was collected with sterile clamps in a Petri dish for cell culture. During the processing of the biopsy, residual parenchymal tissue was trimmed away using scalpels and only tumor tissue was kept and gently minced in small pieces. Small fragments obtained from the tissue were collected in gentleMACS Tube (Miltenyi Biotec) containing 2 ml of DMEM F12 medium (Gibco) supplemented with 10% FBS, 1% penicillin-streptomycin solution, 2% glutamine. Additional 2.5 ml of medium were used to wash the Petri dish and collect residual fragments, then added to the gentleMACS Tube. After mechanical disaggregation, tissue sample (in a range of 0.01 g to 1 g) was enzymatically digested to single cells by using Tumor Dissociation Kit (Miltenyi Biotec) as follows: sample was incubated for 1 h with the enzyme mix containing 200 μl of enzyme H, 100 μl of enzyme R, 25 μl of enzyme A, in 4.5 ml of complete medium at 37°C into the Gentle MACS Octo dissociator (Miltenyi Biotec). After the incubation, 5.5 ml of medium were added to the 4.5 ml suspension of the processed biopsy, then filtered using a 70 μm cell strainer and transferred into a new 50 ml Falcon tube. The used filter was washed with 2.5 ml of complete medium to collect residual cell suspension. Collected sample was centrifugated at 300g for 5 minutes, then supernatant was discharged, and pellet was resuspended with 5 ml of complete medium. The obtained suspension was gently laid on 10 ml of ficoll-hypaque (Sentinel Diagnostics, Lymphoprep) in a 15 ml Falcon tube and centrifuged at 400g for 20 minutes at 21 °C brake off and minimal acceleration. After the centrifugation, the ring containing TILs was carefully collected and washed twice with 25 ml PBS at 930g for 5 minutes at room temperature. Supernatant was removed and pellet was resuspended with 5 ml of complete medium and cells were counted using TC20 Automated Cell Counter (Bio-Rad). If not used, TILs were immediately frozen. Multicolor flow cytometry detection of surface and intracellular antigens of T lymphocytes For ex vivo immunophenotyping experiments, frozen blood and tumor samples were thawed in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS), 1% L-glutamine, 1% sodium pyruvate, 1% nonessential amino acids, antibiotics, 0.1 M 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), and 55 mM β-mercaptoethanol, referred to as R10, and also supplemented with 20 μg/ml DNase I from bovine pancreas (Sigma-Aldrich) [63]. Cells were washed with PBS and immediately stained using viability marker PromoFluor IR-840 (Promokine PromoCell, Heidelberg, Germany) for 20 min at room temperature (RT) in PBS. Then, the cells were washed with FACS buffer (PBS added with 2% FBS) and stained for chemokine receptors (CXCR6 and CCR7) for 20 min at 37°C. After washing with FACS buffer, cells were stained with combination of surface mAbs (anti-CD45, CD3, CD4, CD8, CD45RA, CD39, CD49a, CD69, CD96, CD103, CD161, CD162, CD226, PD-1 and CTLA4) for 20 min at RT (in BV buffer; 50% FACS and 50% Brilliant Stain Buffer from BD). Intracellular detection of Ki-67, T-bet and granzyme B (GZMB) was performed following fixation and permeabilization of cells with the FoxP3/transcription factor staining buffer set (eBioscience) according to manufacturer’s instructions incubating mAbs for 30 min at 4°C. The data were collected using a Cytoflex LX flow cytometer (Beckman Coulter, Brea, CA, USA), which featured five lasers (UV, 350 nm; violet, 405 nm; blue, 488 nm; yellow/green, 561 nm; red, 640 nm), all tuned at 50 mW except for UV, which was tuned at 20 mW, and had the capability to detect 21 parameters. Flow cytometry data were compensated in FlowJo using single stained controls (BD Compensation beads incubated with fluorescently conjugated antibodies). Additionally, all monoclonal antibodies were previously titrated to determine the optimal concentration. The comprehensive list and antibody titers incorporated in the flow cytometry panel can be found in Supplementary Table 2 . Multicolor flow cytometry detection of surface and intracellular antigens of B lymphocytes For the ex vivo immunophenotyping experiments, frozen blood and tumor samples were thawed in R10 supplemented with 20 μg/ml DNase I from bovine pancreas (Sigma-Aldrich). Cells were first washed with PBS and stained with the viability marker PromoFluor IR-840 (Promokine PromoCell, Heidelberg, Germany) for 20 minutes at RT. Subsequently, the cells were washed with FACS buffer and incubated with the chemokine receptor CXCR5 for 20 minutes at 37°C. After another wash with FACS buffer, cells were stained with DuraClone IM B (Beckman Coulter, Brea, CA, USA) containing the following lyophilized directly conjugated monoclonal antibodies: anti-IgD-FITC, CD21-PE, CD19-ECD, CD27-PC7, CD24-APC, CD38-AF750, anti-IgM-PB, and CD45-KrO. Additional drop-in antibodies including VISTA, CD49b, CD73, and CD138 were added in BV buffer. Intracellular detection of Ki-67 and interleukin (IL)-10 was performed after fixing and permeabilizing the cells using the FoxP3/transcription factor staining buffer set (eBioscience) according to the manufacturer’s instructions, and the monoclonal antibodies were incubated for 30 minutes at 4°C. To distinguishing true VISTA and IL-10 positive signals from background noise we used fluorescence minus one (FMO) (Supplementary Fig. 14). The comprehensive list and antibody titers incorporated in the flow cytometry panel can be found in Supplementary Table 2 . High dimensional data analysis T cells analysis Compensated Flow Cytometry Standard (FCS) 3.0 files were imported into FlowJo software version v10.7.1 and pre-analyzed by standard gating to remove doublets, aggregates, dead cells and poor flow events. For each sample, we therefore selected data from all living CD4 + or CD8 + T cells and imported them in R using flowCore package v2.4.0 [64] for a total of 5,946,764 CD4 + T cells (of which 837,808 were TIL) and 2,414,533 CD8 + T cells (of which 613,023 were TIL). The further analysis was performed using CATALYST v1.18.1 [65]. All data obtained by flow cytometry were transformed in R using hyperbolic arcsine “ arcsinh (x/cofactor)” applying manually defined cofactors where x is the fluorescence measured intensity value and cofactor as defined by Melsen and colleagues [66]. Clustering and dimensional reduction were performed using FlowSOM (version 2.4.0) and UMAP (version 0.2.8.0) algorithms, respectively. Clustering gave origin to 30 clusters of CD4 + T cells and 30 of CD8 + T lymphocytes. Details regarding the quality control (QC) of the clustering process for CD4 + and CD8 + T cells were reported in the Supplementary Fig. 15-16 and Supplementary Fig. 17-18 , respectively. B cells analysis Compensated FCS 3.0 files were imported into FlowJo software version v10.7.1 and pre-analyzed by standard gating to remove doublets, aggregates, dead cells, and identify CD19 + B cells. For each sample, we exported all living CD19 + B cells and imported them in R using flowCore package v2.4.0 for a total of or 1,011,489 CD19 + B cells (of which 472,627 were TIL). The unsupervised analysis was performed using CATALYST v1.18.1. All data were transformed in R using hyperbolic arcsine “ arcsinh (x/cofactor)” applying manually defined cofactors (where x is the fluorescence measured intensity value). Clustering and dimensional reduction were performed using FlowSOM and UMAP algorithms, respectively. Clustering gave origin to 20 clusters of CD19 + B lymphocytes. Details regarding the quality control (QC) of the clustering process for B cells can be found in Supplementary Fig. 19-20. Pseudotime analysis We applied Slingshot to CD19 + B cells data [67]. Slingshot is a tool used for inferring trajectories and pseudotime from single-cell data. Initially, we performed dimensional reduction using DiffusionMap. Subsequently, smooth branching curves were fitted to these lineages to obtain refined representations of each lineage, translating the overall knowledge of lineage structure into pseudotime at the single-cell level. The clusters obtained through FlowSOM (Total B cells = 20 clusters) were embedded within the DiffusionMap analysis. Single-cell metabolic regulome profiling (scMEP) Heavy metal conjugation of antibodies Antibodies were conjugated to heavy metal ions with commercially available MaxPar (Standard BioTools) reagents following vendor conjugation protocol for MCP9 (for cadmium) or X8 (for lanthanide). In short, antibodies were reduced with 4 mM TCEP (Thermo Fisher) for 30 min at 37 °C and washed two times. For conjugations using MaxPar reagents, metal chelation was performed by adding metal solutions (final, 0.05 M) to chelating polymers and incubating for 60 min (for cadmium) or 40 min (for lanthanide) at 37°C. Metal-loaded polymers were washed three times using a 3-kDa MWCO microfilter (Millipore) by centrifuging for 25 min at 12,000g at RT. Partially reduced antibodies and metal-loaded polymers were incubated together for 90 min at 37 °C. Cadmium conjugated antibodies were washed four times for 10 min at 5,000g at RT, while lanthanide conjugated antibodies were washed four times for 10 min at 10,000g at RT. All conjugated antibodies were collected by centrifugation (2 min, 1,000g, RT) into an inverted column in a fresh 1.6-ml collection tube and transferred into Protein LoBind tube (Eppendorf). Protein content was assessed by NanoDrop (Thermo Fisher) measurement. HRP protector buffer for cadmium conjugated antibodies (Boca Scientific) or antibody stabilizer PBS for lanthanide conjugated antibodies (Candor Bioscience) were added to reach the final antibodies concentration of 0.5mg/ml. All antibodies were stored at 4 °C. Live/dead discrimination and cell staining Cryopreserved single-cell suspensions from tumor biopsy specimens were thawed in R10 supplemented with 20 μg/ml DNase I from bovine pancreas (Sigma-Aldrich) and washed twice (5 min, 600g, RT). Before staining, tumor infiltrating leukocytes were enriched by using CD45 (TIL) MicroBeads (Miltenyi Biotec). For live/dead cell discrimination resuspended cells were stained with 1 µM of Cell-ID Cisplatin-195Pt (Standard BioTools) for 5 min at 37 °C and washed with Maxpar Cell Staining buffer (Standard BioTools). Dyed cells were stained with combination of surface mAbs (see Supplementary Table 2 ) for 30 min at RT in Maxpar Cell Staining buffer. Then, cells were washed with 2ml of Maxpar Cell Staining buffer and centrifuge for 5 min at 300g at RT. Intracellular markers detection was performed following fixation of cells with the FoxP3/transcription factor staining buffer set (eBioscience) according to manufacturer’s instructions incubating mAbs for 30 min at 4°C (see Supplementary Table 2 ). After washing with 2ml of Maxpar Cell Staining buffer, cells were placed on ice for 10 min to chill the sample and proceed for phosphoprotein staining. Each sample were resuspended using 1ml of 100% methanol and incubated for 15 min on ice. Then samples were washed twice with Maxpar Cell Staining buffer and incubated with anti-phosphoprotein antibodies mix for 30 in at RT (see Supplementary Table 2 ). Then all samples were washed twice with 2ml of Maxpar Cell Staining buffer and fixed adding 1ml of 1.6% Formaldehyde incubating for 10 min at RT. After incubation samples were centrifuged for 5 min at 800g at RT. Then all samples were resuspended in 1ml of Cell-ID intercalator-Ir (Standard BioTools) 125nM and stained overnight at 4 °C. Cells were then resuspended at 1 × 10 6 cells per ml in ddH 2 O supplemented with 1× EQ Four Element Calibration Beads (Standard BioTools) and acquired on a CyTOF2 mass cytometer (Standard BioTools). Mass cytometry data pre-processing and analysis Raw mass cytometry data were first bead normalized to remove acquisition-related influences on marker expression using CATALYST v1.18.1 pre-processing workflow. Normalized CyTOF files were exported as .fsc. Subsequently, all normalized .fcs files were uploaded into FlowJo software v10.7.1 and checked to exclude flow instabilities (using PeacoQC package), aggregates, doublets, dead cells, and non-biological events (Supplementary Fig. 21A). For each sample, we therefore selected data from living CD45 + B and T cells and imported them in R using flowCore package v2.4.0. All data obtained by CyTOF were transformed using hyperbolic arcsinh (cofactor 5). The first round of clustering and dimensional reduction was performed respectively using FlowSOM and UMAP algorithms utilizing both phenotypic and metabolic markers to clearly identify the main populations (Supplementary Fig. 21B-F). Subsequently, we selected B cells, CD4 + T cells, and CD8 + T cells, clustering them separately using exclusively metabolic markers or both as reported in Supplementary Fig. 3, 10, 12. TCF-1 and CXCL13 were excluded from the analysis due to weak signal caused by staining and conjugation issues. Calculation of metabolic scores To calculate single-cell metabolic scores, expression values (bead normalized, arcsinh transformed) from all metabolic enzymes within a given pathway (glycolysis, mitochondrial dynamics, TCA, pentose phosphate pathway, transcription, aminoacidic metabolism and fatty acid metabolism) were summed and divided by the number of channels within the pathway. The heatmap was drowned using ComplexHeatmap package v2.14.0 [68]. To calculate the ‘similarity score’ to compare the metabolism on a set of clusters we used MEM_RMSD()function from cytoMEM package v1.2.0 [69]. Briefly, MEM_RMSD calculates a normalized average root-mean-square deviation (RMSD) score pairwise between populations given their metabolic scores as input. This is meant to serve as a metric of similarity between populations. The RMSD values are then converted to percentages with the maximum RMSD in the matrix set as 100 percent, so that the final RMSD score is the percent of the maximum RMSD [Percent_max_RMSD = 100-RMSD/max_RMSD*100]. Image-based spatial transcriptomic data Spatial molecular imaging data from FFPE non-small-cell lung cancer tissue samples were retrieved from http://nanostring.com/CosMx-dataset . Transcriptomic data have been normalized using the R package Seurat applying SCTransform functions [70]. All cells were annotated using precomputed Azimuth annotations from Seurat package. ImageDimPlot function was used to plot B and T cell molecules displaying spatial correlation with each other. BuildNicheAssay function from Seurat was used to construct a new assay called ‘niche’ containing the cell type composition spatially neighboring each cell. NicheNet analysis Tumor infiltrating CD45 + T and B cells were retrieved from GSE131907, GSE148071, GSE154826 containing data of stage I, II or IIIA patients. All three datasets have been normalized and integrated using the R package Seurat applying SCTransform and IntegrateData functions [70, 71]). Principal component was computed, and 20 PCs were selected to ran UMAP and graph-based clustering with a resolution between 0.1 and 1. Clustertree was used to choose the best resolution [72]. The integrated Seurat object was annotated, and specifically, T cells and B cells were chosen. From these, CD4 + , CD8 + T cells, and CD19 + B cells (excluding plasma cells, Pro-B cells and proliferating cells; Supplementary Fig. 7A-B) were employed for running NicheNet, following the instructions outlined in the provided vignette (https://github.com/saeyslab/nichenetr; doi:10.1038/s41592-019-0667-5). During multiple separate NicheNet runs, CD4 + and CD8 + T cell subsets were set as ‘receiver’ and B cell clusters as ‘sender’ populations. For the receiver and sender cell population were retrieved expressed genes, with the key parameters being set as follows: genes expressed in at least 10% of the cells of the respective group. On these sets of genes was run NicheNet to infer the ligand-receptor network followed by finding of ‘target genes’ of top-ranked ligands (Supplementary Fig. 7C-D). Spatial transcriptomic data from NSCLC tumor-infiltrating B and T cells were obtained from the image-based dataset mentioned earlier. CD4 + , CD8 + T cells, and CD19 + B cells were selected for NicheNet analysis, following the previously described procedure. Our secondary objective was to explore whether B cells influence the expression of genes associated with tumor reactivity in CD4 + and CD8 + T cells. To achieve this, we utilized the 'tumor reactive gene signature' as defined by Olivera et al. and Guo et al. [6, 73]. The predicted targets score was based on a Pearson correlation coefficient as described in the NicheNet vignette. NicheNet ligand-receptor matrix was integrated with new information on VISTA obtained from recent publications [41, 74] Intracellular cytokine staining (ICS) of T and B cells Thawed PBMCs and TILs were rested for 1h. Then T cells were stimulated with 50 ng/ml of phorbol 12-myristate 13-acetate (PMA) and 1 μg/ml of ionomycin for 4h at 37°C in a 5% CO 2 atmosphere in R10 culture medium. For each stimulated sample, an unstimulated one was prepared as a negative control. All samples were incubated with protein transport inhibitor containing Brefeldin A (Biolegend, San Diego, CA, USA) and monensin (Biolegend, San Diego, CA, USA). After stimulation, cells were washed with PBS and stained with viability marker PromoFluor IR-840 (Promokine, PromoCell, Heidelberg, Germany) for 20 min at RT. Next, cells were washed with FACS buffer and stained with surface mAbs ( Supplementary Table 2 ). After incubation all samples were fixed and permeabilized with the FoxP3/transcription factor staining buffer set (eBioscience) according to manufacturer’s instructions incubating. Then, cells were stained with previously titrated mAbs recognizing intracellular cytokines and transcription factors for 30 min at 4°C (see Supplementary Table 2 ). B cells were primed with 10 μg/ml of oligodeoxyribonucleotides containing CpG motifs (CpG ODNs; Miltenyi Biotec) for 72h at 37°C in a 5% CO 2 atmosphere in R10 culture medium without DNase I. Then B cells were stimulated with 50 ng/ml of PMA and 1 μg/ml of ionomycin for 4 h at 37°C in a 5% CO 2 atmosphere in complete culture medium as mentioned before. For each stimulated sample, an unstimulated one was prepared as a negative control. All samples were incubated with protein transport inhibitor containing Brefeldin A (Biolegend, San Diego, CA, USA) and monensin (Biolegend, San Diego, CA, USA). After stimulation, cells were washed with PBS and stained with viability marker PromoFluor IR-840 (Promokine, PromoCell, Heidelberg, Germany) for 20 min at RT. Next, cells were washed with FACS buffer and stained with surface mAbs (see Supplementary Table 2 ). After incubation all samples were washed with FACS buffer and fixed and permeabilized with the FoxP3/transcription factor staining buffer set (eBioscience) according to manufacturer’s instructions. Then, cells were stained with previously titrated mAbs recognizing intracellular cytokines and transcription factors for 30 min at 4°C ( Supplementary Table 2 ). Gating strategy used to identify and analyze the intracellular cytokine production of CD4 + , CD8 + T lymphocytes and CD19 + B lymphocytes are reported in Supplementary Fig. 22, 23 and 24. All mAbs were previously titrated to define the optimal concentration. All samples were acquired on a Cytoflex LX flow cytometer (Beckman Coulter, Brea, CA, USA). PENCIL prediction analysis PENCIL method relies on the concept known as LWR, a machine learning strategy that introduces rejection labels in the prediction results [45]. The process of PENCIL workflow is depicted in Supplementary Fig. 25. Input data for PENCIL included hyperbolic arcsine transformed single-cell fluorescence intensity matrices and relevant cell metadata, such as cluster ID and UMAP coordinates. These data were obtained from CATALYST and imported into Seurat using the CreateSeuratObject function. We followed the standard workflows for CD4 + , CD8 + , and B cell prediction as recommended on the PENCIL GitHub page. PENCIL analyses were conducted using Python 3.11.4 with GPU acceleration (NVIDIA Quadro RTX 5000). Tuning parameters, including shuffle rate, lambda L1, and lambda L2, along with corresponding precision, recall, and f1-score, are detailed in Supplementary Fig. 13. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis Blood monocytes and B cells were isolated using CD14 or CD19 MicroBeads (Miltenyi Biotec) through magnetic enrichment techniques. Subsequently, protein extraction was carried out following the previously described method [75]. Fifty and ten µg of proteins of monocytes and B lymphocytes, respectively, were processed using FASP protocol [76]. Protein digestion was performed by trypsin in an enzyme-to-protein ration of 1:50 (w/w). The tryptic peptides were suspended in water:acetonitrile:formic acid (95:3:2) for mass spectrometry analysis. The tryptic digests were injected into a Nano UHPLC Ultimate 3000 (Thermo Fisher Scientific) coupled to Exploris™ 480 Hybrid Quadrupole-Orbitrap™ Mass Spectrometer (Thermo Fisher Scientific). Separation was achieved using a C18 EASY-Spray HPLC column (75 μm × 500 mm, 2 μm, ES903 Thermo Fisher Scientific) and elution was performed using a binary system of solvents. Solvent A was 0,1% formic acid and solvent B was 97% acetonitrile. A linear binary gradient was applied to eluate the peptides: 0–23% solvent B in solvent A for 140 min and further 35 min of 23–33% solvent B in solvent A at 300 nl/min. Data-dependent acquisition (DDA) For monocytes sample the acquisition was operated in full MS/data-dependent acquisition. The Orbitrap mass analyzer was used at a resolution of 120,000 to acquire full MS with an m/z range of 400 to 1500. MS/MS fragments were measured at an Orbitrap resolution of 15,000. Twenty of the most intense ions were isolated for MS/MS analysis. The raw data were processed using Proteome Discoverer (version 3.0 SP1, Thermo Fisher Scientific), searching with a database of human proteome (www.uniprot.org, accessed November 2022, 20327 sequences). The selected parameters for protein identification were: (i) at least 1 unique peptide; (ii) static modification carbamidomethyl on cysteines (+57.021 Da), dynamic modifications oxidation on methionine (+15.995 Da) and deamidation on asparagine and glutamine (+0.984 Da); (iii) precursor mass tolerance of 10 ppm, fragment mass tolerance of 0.02 Da; (iv) the maximum of missed trypsin cleavage sites of 1; (v) the minimum peptide length of 7. Database search analysis was performed using Sequest HT and the artificial intelligence node Chimerys (Thermo Fisher Scientific). The resulting files were imported in Skyline (v.22.2.0.527) as already described [77], to generate spectral library. Targeted MS Assay For B cell sample a targeted proteomic analysis for VISTA protein was performed by parallel reaction monitoring (PRM), as previously described [78]. The peptides of VISTA identified in monocytes sample passing the false discovery rate were exported to a text file and processed by PRM. The mass inclusion list involved mass, charge, and polarity. Lists of all peptides targeted in the PRM analyses are reported in Supplementary Fig. 2. Sample was processed in triplicate for PRM. The MS data were processed using Skyline (v.22.2.0.527) using generated spectral library as reference. Hypoxic assay PBMCs were isolated from fresh blood of eight donors ( Supplementary Table 1 ). In each sample, 30 million of PBMC were washed with PBS and stained with viability marker LIVE/DEAD™ Fixable Aqua (ThermoFisher Scientific, USA) for 20 min at RT. Next, cells were washed with FACS buffer and stained with anti-CD19-APC for 20 min at RT. After incubation all samples were washed with FACS buffer and sorted on Bigfoot Spectral Cell Sorter (ThermoFisher Scientific, USA). Purity and the gating strategy were reported in Supplementary Fig. 26. Sorted cells were plated at the concentration of 10 5 cells/100uL of R10 (without DNAse I to prevent CpG degradation) a 96-well U-bottom multiwell plate. For the stimulated condition, B cells were primed with 10 μg/ml of oligodeoxyribonucleotides containing CpG motifs (CpG ODNs; Miltenyi Biotec) for 48h at 37°C in a 5% CO 2 atmosphere (Normoxia) or at 37°C in a 1% O 2 and 5% CO 2 atmosphere (Hypoxia) within a hypoxia incubator chamber (Stemcell Technologies Inc., Canada). The cells subjected to hypoxia were quickly moved from the hypoxic chamber onto ice to prevent the swift degradation of hypoxia-responsive molecules. Subsequently, all cells were washed with cold PBS kept at 4°C and stained with viability marker LIVE/DEAD™ Fixable Aqua (ThermoFisher Scientific, USA) for 20 min at 4°C. Next, cells were rinsed with chilled FACS buffer at 4°C and then stained with surface monoclonal antibodies for 20 minutes at 4°C. Intracellular detection of HIF-1α was performed following FoxP3/transcription factor staining buffer set (eBioscience) according to manufacturer’s instructions incubating mAbs for 30 min at 4°C. All antibodies employed are listed in Supplementary Table 2 . Data acquisition was performed on a Cytoflex LX flow cytometer (Beckman Coulter, Brea, CA, USA), with ex vivo samples serving as controls for both normoxic and hypoxic conditions (Supplementary Fig. 26). Statistical analysis Statistical analyses were performed using R v4.1.0, GraphPad Prism version 8 (GraphPad Software Inc., La Jolla, USA) or SPICE software [79], unless specified otherwise. Significance of differences for the frequency of single FlowSOM clusters was determined using generalized linear mixed model (GLMM) implemented within diffcyt package v1.14.0 [80] applying FDR cutoff = 0.05. To compare distributions of manually gated subsets significance was determined by a nonparametric paired Wilcoxon rank test, unless otherwise specified in the figure legends. We used χ2 permutation test for pie chart comparison (SPICE). Kaplan-Meier method was used to analyze survival. The multivariate Cox regression analysis was conducted in R using the survival package v.3.4.0 to explore the correlation between patients' survival time and predictor variables identified by PENCIL. Declarations Ethical Approval The study protocol was approved by the local Ethical Committes (“Area Vasta Emilia Nord”, protocol number 0018007/21; “Comitato Etico della Romagna”, protocol number 7043/2021). All participants involved in this study provided written informed consent for both sample collection and data analyses. Detailed clinical information and the analysis methods employed for each sample are reported in Supplementary Table 1 . Competing Interests All authors declare no competing interests. Contributions DLT, VM, NP, FDL, AN, RB, ES, ALC, AVS, AD, MG, GM, FR, FT, FB performed experiments; DLT performed bioinformatic analysis; BA, FS, PLF, MD,FB enrolled the patients; DLT, BA, SDB, LG, DQ, AlC, ANeri, AC discussed the data; SDB and AC supervised experiments; DLT, SDB, LG and AC wrote the manuscript. Funding This work was supported by grants from: Fondazione AIRC per la ricerca sul cancro to AC, project “Role of exhausted CD8 TILs in the recurrence of resectable non-small cell lung cancer” grant number 25073. Author Contribution DLT, VM, NP, FDL, AN, RB, ES, ALC, AVS, AD, MG, GM, FR, FT, FB performed experiments; DLT performed bioinformatic analysis; BA, FS, PLF, MD,FB enrolled the patients; DLT, BA, SDB, LG, DQ, AlC, ANeri, AC discussed the data; SDB and AC supervised experiments; DLT, SDB, LG and AC wrote the manuscript. Acknowledgements SDB and LGi are Marylou Ingram Scholar of the International Society for Advancement of Cytometry (ISAC) for the period 2015–2020 and 2020–2025, respectively. Drs. Paola Paglia (ThermoFisher Scientific, Monza, Italy), Leonardo Beretta, Anis Larbi (Beckman Coulter, Milan, Italy), Dr. Paolo Santino, Ernesto Lopez, Gloria Martrus (Standard Biotools, San Francisco, CA, US) are acknowledged for their support in providing reagents and materials, for precious help and technical suggestions. We acknowledge Croce Blu (Modena hub) and its volunteers for the efficient service in transporting biological samples from Morgagni - Pierantoni Hospital to University of Modena. Finally, we gratefully acknowledge the patients who donated blood to participate in this study. Availability of data and materials The flow cytometry and mass cytometry data generated in this study are available under reasonable request. The raw data generated in this study are provided in the Source Data file. Further inquiries can be directed to the corresponding author. References Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, Bray F. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin. 2021;71:209–49. Hartwig MG, D'Amico TA. Thoracoscopic lobectomy: the gold standard for early-stage lung cancer? Ann Thorac Surg. 2010;89:2098–101. Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, 2022. CA Cancer J Clin. 2022;72:7–33. Riley RS, June CH, Langer R, Mitchell MJ. Delivery technologies for cancer immunotherapy. Nat Rev Drug Discov. 2019;18:175–96. Roy R, Singh SK, Misra S. Advancements in Cancer Immunotherapies. Vaccines (Basel) 2022, 11. Guo X, Zhang Y, Zheng L, Zheng C, Song J, Zhang Q, Kang B, Liu Z, Jin L, Xing R, et al. Global characterization of T cells in non-small-cell lung cancer by single-cell sequencing. Nat Med. 2018;24:978–85. Giraldo NA, Sanchez-Salas R, Peske JD, Vano Y, Becht E, Petitprez F, Validire P, Ingels A, Cathelineau X, Fridman WH, Sautes-Fridman C. The clinical role of the TME in solid cancer. Br J Cancer. 2019;120:45–53. Aramini B, Masciale V, Samarelli AV, Dubini A, Gaudio M, Stella F, Morandi U, Dominici M, De Biasi S, Gibellini L, Cossarizza A. Phenotypic, functional, and metabolic heterogeneity of immune cells infiltrating non-small cell lung cancer. Front Immunol. 2022;13:959114. Sautes-Fridman C, Petitprez F, Calderaro J, Fridman WH. Tertiary lymphoid structures in the era of cancer immunotherapy. Nat Rev Cancer. 2019;19:307–25. Fukuhara M, Muto S, Inomata S, Yamaguchi H, Mine H, Takagi H, Ozaki Y, Watanabe M, Inoue T, Yamaura T, et al. The clinical significance of tertiary lymphoid structure and its relationship with peripheral blood characteristics in patients with surgically resected non-small cell lung cancer: a single-center, retrospective study. Cancer Immunol Immunother. 2022;71:1129–37. Gu-Trantien C, Loi S, Garaud S, Equeter C, Libin M, de Wind A, Ravoet M, Le Buanec H, Sibille C, Manfouo-Foutsop G, et al. CD4(+) follicular helper T cell infiltration predicts breast cancer survival. J Clin Invest. 2013;123:2873–92. Goc J, Germain C, Vo-Bourgais TK, Lupo A, Klein C, Knockaert S, de Chaisemartin L, Ouakrim H, Becht E, Alifano M, et al. Dendritic cells in tumor-associated tertiary lymphoid structures signal a Th1 cytotoxic immune contexture and license the positive prognostic value of infiltrating CD8 + T cells. Cancer Res. 2014;74:705–15. Hennequin A, Derangere V, Boidot R, Apetoh L, Vincent J, Orry D, Fraisse J, Causeret S, Martin F, Arnould L, et al. Tumor infiltration by Tbet + effector T cells and CD20 + B cells is associated with survival in gastric cancer patients. Oncoimmunology. 2016;5:e1054598. Helmink BA, Reddy SM, Gao J, Zhang S, Basar R, Thakur R, Yizhak K, Sade-Feldman M, Blando J, Han G, et al. B cells and tertiary lymphoid structures promote immunotherapy response. Nature. 2020;577:549–55. Sorin M, Rezanejad M, Karimi E, Fiset B, Desharnais L, Perus LJM, Milette S, Yu MW, Maritan SM, Dore S, et al. Single-cell spatial landscapes of the lung tumour immune microenvironment. Nature. 2023;614:548–54. Fridman WH, Petitprez F, Meylan M, Chen TW, Sun CM, Roumenina LT, Sautes-Fridman C. B cells and cancer: To B or not to B? J Exp Med 2021, 218. Hu Q, Hong Y, Qi P, Lu G, Mai X, Xu S, He X, Guo Y, Gao L, Jing Z, et al. Atlas of breast cancer infiltrated B-lymphocytes revealed by paired single-cell RNA-sequencing and antigen receptor profiling. Nat Commun. 2021;12:2186. Shalapour S, Font-Burgada J, Di Caro G, Zhong Z, Sanchez-Lopez E, Dhar D, Willimsky G, Ammirante M, Strasner A, Hansel DE, et al. Immunosuppressive plasma cells impede T-cell-dependent immunogenic chemotherapy. Nature. 2015;521:94–8. Cai C, Zhang J, Li M, Wu ZJ, Song KH, Zhan TW, Wang LH, Sun YH. Interleukin 10-expressing B cells inhibit tumor-infiltrating T cell function and correlate with T cell Tim-3 expression in renal cell carcinoma. Tumour Biol. 2016;37:8209–18. Murakami Y, Saito H, Shimizu S, Kono Y, Shishido Y, Miyatani K, Matsunaga T, Fukumoto Y, Ashida K, Sakabe T, et al. Increased regulatory B cells are involved in immune evasion in patients with gastric cancer. Sci Rep. 2019;9:13083. Deng J, Li J, Sarde A, Lines JL, Lee YC, Qian DC, Pechenick DA, Manivanh R, Le Mercier I, Lowrey CH, et al. Hypoxia-Induced VISTA Promotes the Suppressive Function of Myeloid-Derived Suppressor Cells in the Tumor Microenvironment. Cancer Immunol Res. 2019;7:1079–90. DeRogatis JM, Viramontes KM, Neubert EN, Henriquez ML, Guerrero-Juarez CF, Tinoco R. Targeting the PSGL-1 Immune Checkpoint Promotes Immunity to PD-1-Resistant Melanoma. Cancer Immunol Res. 2022;10:612–25. ElTanbouly MA, Zhao Y, Nowak E, Li J, Schaafsma E, Le Mercier I, Ceeraz S, Lines JL, Peng C, Carriere C et al. VISTA is a checkpoint regulator for naive T cell quiescence and peripheral tolerance. Science 2020, 367. Hope JL, Otero DC, Bae EA, Stairiker CJ, Palete AB, Faso HA, Lin M, Henriquez ML, Roy S, Seo H, et al. PSGL-1 attenuates early TCR signaling to suppress CD8(+) T cell progenitor differentiation and elicit terminal CD8(+) T cell exhaustion. Cell Rep. 2023;42:112436. Issam Salah NEI, Marnissi F, Lakhdar A, Karkouri M, ElBelhadji M, Badou A. The immune checkpoint VISTA is associated with prognosis in patients with malignant uveal melanoma. Front Immunol. 2023;14:1225140. Kappelmayer J, Nagy B Jr.. The Interaction of Selectins and PSGL-1 as a Key Component in Thrombus Formation and Cancer Progression. Biomed Res Int. 2017;2017:6138145. Kuklinski LF, Yan S, Li Z, Fisher JL, Cheng C, Noelle RJ, Angeles CV, Turk MJ, Ernstoff MS. VISTA expression on tumor-infiltrating inflammatory cells in primary cutaneous melanoma correlates with poor disease-specific survival. Cancer Immunol Immunother. 2018;67:1113–21. Liao H, Zhu H, Liu S, Wang H. Expression of V-domain immunoglobulin suppressor of T cell activation is associated with the advanced stage and presence of lymph node metastasis in ovarian cancer. Oncol Lett. 2018;16:3465–72. Lin Y, Huang S, Qi Y, Xie L, Jiang J, Li H, Chen Z. PSGL-1 is a novel tumor microenvironment prognostic biomarker with cervical high-grade squamous lesions and more. Front Oncol. 2023;13:1052201. Tinoco R, Carrette F, Barraza ML, Otero DC, Magana J, Bosenberg MW, Swain SL, Bradley LM. PSGL-1 Is an Immune Checkpoint Regulator that Promotes T Cell Exhaustion. Immunity. 2016;44:1190–203. Wu L, Deng WW, Huang CF, Bu LL, Yu GT, Mao L, Zhang WF, Liu B, Sun ZJ. Expression of VISTA correlated with immunosuppression and synergized with CD8 to predict survival in human oral squamous cell carcinoma. Cancer Immunol Immunother. 2017;66:627–36. Germain C, Gnjatic S, Tamzalit F, Knockaert S, Remark R, Goc J, Lepelley A, Becht E, Katsahian S, Bizouard G, et al. Presence of B cells in tertiary lymphoid structures is associated with a protective immunity in patients with lung cancer. Am J Respir Crit Care Med. 2014;189:832–44. Wouters MCA, Nelson BH. Prognostic Significance of Tumor-Infiltrating B Cells and Plasma Cells in Human Cancer. Clin Cancer Res. 2018;24:6125–35. Mauri C, Menon M. Human regulatory B cells in health and disease: therapeutic potential. J Clin Invest. 2017;127:772–9. Isnardi I, Ng YS, Menard L, Meyers G, Saadoun D, Srdanovic I, Samuels J, Berman J, Buckner JH, Cunningham-Rundles C, Meffre E. Complement receptor 2/CD21- human naive B cells contain mostly autoreactive unresponsive clones. Blood. 2010;115:5026–36. Fridman WH, Meylan M, Petitprez F, Sun CM, Italiano A, Sautes-Fridman C. B cells and tertiary lymphoid structures as determinants of tumour immune contexture and clinical outcome. Nat Rev Clin Oncol. 2022;19:441–57. Hartmann FJ, Mrdjen D, McCaffrey E, Glass DR, Greenwald NF, Bharadwaj A, Khair Z, Verberk SGS, Baranski A, Baskar R, et al. Single-cell metabolic profiling of human cytotoxic T cells. Nat Biotechnol. 2021;39:186–97. De Biasi SLT, Neroni D et al. A.;. : Metabolic pathways engaged by antigen-specific T and B cells after SARS-CoV-2 vaccination in multiple sclerosis patients on different immunomodulatory drugs reveal immunosenescence and predict vaccine efficacy. PREPRINT (Version 1) available at Research Square 2023. Miller RA, Luke JJ, Hu S, Mahabhashyam S, Jones WB, Marron T, Merchan JR, Hughes BGM, Willingham SB. Anti-CD73 antibody activates human B cells, enhances humoral responses and induces redistribution of B cells in patients with cancer. J Immunother Cancer 2022, 10. de Gruijter NM, Jebson B, Rosser EC. Cytokine production by human B cells: role in health and autoimmune disease. Clin Exp Immunol. 2022;210:253–62. Johnston RJ, Su LJ, Pinckney J, Critton D, Boyer E, Krishnakumar A, Corbett M, Rankin AL, Dibella R, Campbell L, et al. VISTA is an acidic pH-selective ligand for PSGL-1. Nature. 2019;574:565–70. Rogers BM, Smith L, Dezso Z, Shi X, DiGiammarino E, Nguyen D, Sethuraman S, Zheng P, Choi D, Zhang D et al. VISTA is an activating receptor in human monocytes. J Exp Med 2021, 218. Shanshan H, Ruchir B, Carl B, Emily AB, Derek LB, Kan C, Patrick D, Dwayne D, Ryan GG, Gary G et al. High-plex Multiomic Analysis in FFPE at Subcellular Level by Spatial Molecular Imaging. bioRxiv 2022:2021.2011.2003.467020.. Di Pilato M, Kfuri-Rubens R, Pruessmann JN, Ozga AJ, Messemaker M, Cadilha BL, Sivakumar R, Cianciaruso C, Warner RD, Marangoni F, et al. CXCR6 positions cytotoxic T cells to receive critical survival signals in the tumor microenvironment. Cell. 2021;184:4512–4530e4522. Ren T, Chen C, Danilov AV, Liu S, Guan X, Du S, Wu X, Sherman MH, Spellman PT, Coussens LM, et al. Supervised learning of high-confidence phenotypic subpopulations from single-cell data. Nat Mach Intell. 2023;5:528–41. Pearce EL, Poffenberger MC, Chang CH, Jones RG. Fueling immunity: insights into metabolism and lymphocyte function. Science. 2013;342:1242454. Norata GD, Caligiuri G, Chavakis T, Matarese G, Netea MG, Nicoletti A, O'Neill LA, Marelli-Berg FM. The Cellular and Molecular Basis of Translational Immunometabolism. Immunity. 2015;43:421–34. Itakura E, Huang RR, Wen DR, Paul E, Wunsch PH, Cochran AJ. IL-10 expression by primary tumor cells correlates with melanoma progression from radial to vertical growth phase and development of metastatic competence. Mod Pathol. 2011;24:801–9. Huseni MA, Wang L, Klementowicz JE, Yuen K, Breart B, Orr C, Liu LF, Li Y, Gupta V, Li C, et al. CD8(+) T cell-intrinsic IL-6 signaling promotes resistance to anti-PD-L1 immunotherapy. Cell Rep Med. 2023;4:100878. Hong IS. Stimulatory versus suppressive effects of GM-CSF on tumor progression in multiple cancer types. Exp Mol Med. 2016;48:e242. Yan F, Du R, Wei F, Zhao H, Yu J, Wang C, Zhan Z, Ding T, Ren X, Chen X, Li H. Expression of TNFR2 by regulatory T cells in peripheral blood is correlated with clinical pathology of lung cancer patients. Cancer Immunol Immunother. 2015;64:1475–85. Nakamura K, Kitani A, Strober W. Cell contact-dependent immunosuppression by CD4(+)CD25(+) regulatory T cells is mediated by cell surface-bound transforming growth factor beta. J Exp Med. 2001;194:629–44. Yuan L, Tatineni J, Mahoney KM, Freeman GJ. VISTA: A Mediator of Quiescence and a Promising Target in Cancer Immunotherapy. Trends Immunol. 2021;42:209–27. Kallies A, Good-Jacobson KL. Transcription Factor T-bet Orchestrates Lineage Development and Function in the Immune System. Trends Immunol. 2017;38:287–97. Wein AN, McMaster SR, Takamura S, Dunbar PR, Cartwright EK, Hayward SL, McManus DT, Shimaoka T, Ueha S, Tsukui T, et al. CXCR6 regulates localization of tissue-resident memory CD8 T cells to the airways. J Exp Med. 2019;216:2748–62. Hurkmans DP, Basak EA, Schepers N, Oomen-De Hoop E, Van der Leest CH, El Bouazzaoui S, Bins S, Koolen SLW, Sleijfer S, Van der Veldt AAM et al. Granzyme B is correlated with clinical outcome after PD-1 blockade in patients with stage IV non-small-cell lung cancer. J Immunother Cancer 2020, 8. Braun M, Aguilera AR, Sundarrajan A, Corvino D, Stannard K, Krumeich S, Das I, Lima LG, Meza Guzman LG, Li K, et al. CD155 on Tumor Cells Drives Resistance to Immunotherapy by Inducing the Degradation of the Activating Receptor CD226 in CD8(+) T Cells. Immunity. 2020;53:805–823e815. Chiang EY, de Almeida PE, de Almeida Nagata DE, Bowles KH, Du X, Chitre AS, Banta KL, Kwon Y, McKenzie B, Mittman S, et al. CD96 functions as a co-stimulatory receptor to enhance CD8(+) T cell activation and effector responses. Eur J Immunol. 2020;50:891–902. Ettinger DS, Wood DE, Aisner DL, Akerley W, Bauman JR, Bharat A, Bruno DS, Chang JY, Chirieac LR, D'Amico TA, et al. Non-Small Cell Lung Cancer, Version 3.2022, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw. 2022;20:497–530. Martini N, Bains MS, Burt ME, Zakowski MF, McCormack P, Rusch VW, Ginsberg RJ. Incidence of local recurrence and second primary tumors in resected stage I lung cancer. J Thorac Cardiovasc Surg. 1995;109:120–9. Silvan J, Gonzalez-Tajuelo R, Vicente-Rabaneda E, Perez-Frias A, Espartero-Santos M, Munoz-Callejas A, Garcia-Lorenzo E, Gamallo C, Castaneda S, Urzainqui A. Deregulated PSGL-1 Expression in B Cells and Dendritic Cells May Be Implicated in Human Systemic Sclerosis Development. J Invest Dermatol. 2018;138:2123–32. Cossarizza A, Chang HD, Radbruch A, Abrignani S, Addo R, Akdis M, Andra I, Andreata F, Annunziato F, Arranz E, et al. Guidelines for the use of flow cytometry and cell sorting in immunological studies (third edition). Eur J Immunol. 2021;51:2708–3145. De Biasi S, Gibellini L, Lo Tartaro D, Puccio S, Rabacchi C, Mazza EMC, Brummelman J, Williams B, Kaihara K, Forcato M, et al. Circulating mucosal-associated invariant T cells identify patients responding to anti-PD-1 therapy. Nat Commun. 2021;12:1669. Hahne F, LeMeur N, Brinkman RR, Ellis B, Haaland P, Sarkar D, Spidlen J, Strain E, Gentleman R. flowCore: a Bioconductor package for high throughput flow cytometry. BMC Bioinformatics. 2009;10:106. Nowicka M, Krieg C, Crowell HL, Weber LM, Hartmann FJ, Guglietta S, Becher B, Levesque MP, Robinson MD. CyTOF workflow: differential discovery in high-throughput high-dimensional cytometry datasets. F1000Res 2017, 6:748. Melsen JE, van Ostaijen-Ten Dam MM, Lankester AC, Schilham MW, van den Akker EB. A Comprehensive Workflow for Applying Single-Cell Clustering and Pseudotime Analysis to Flow Cytometry Data. J Immunol. 2020;205:864–71. Street K, Risso D, Fletcher RB, Das D, Ngai J, Yosef N, Purdom E, Dudoit S. Slingshot: cell lineage and pseudotime inference for single-cell transcriptomics. BMC Genomics. 2018;19:477. Gu Z, Eils R, Schlesner M. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics. 2016;32:2847–9. Diggins KE, Greenplate AR, Leelatian N, Wogsland CE, Irish JM. Characterizing cell subsets using marker enrichment modeling. Nat Methods. 2017;14:275–8. Hafemeister C, Satija R. Normalization and variance stabilization of single-cell RNA-seq data using regularized negative binomial regression. Genome Biol. 2019;20:296. Butler A, Hoffman P, Smibert P, Papalexi E, Satija R. Integrating single-cell transcriptomic data across different conditions, technologies, and species. Nat Biotechnol. 2018;36:411–20. Zappia L, Oshlack A. Clustering trees: a visualization for evaluating clusterings at multiple resolutions. Gigascience 2018, 7. Oliveira G, Stromhaug K, Klaeger S, Kula T, Frederick DT, Le PM, Forman J, Huang T, Li S, Zhang W, et al. Phenotype, specificity and avidity of antitumour CD8(+) T cells in melanoma. Nature. 2021;596:119–25. Wang J, Wu G, Manick B, Hernandez V, Renelt M, Erickson C, Guan J, Singh R, Rollins S, Solorz A, et al. VSIG-3 as a ligand of VISTA inhibits human T-cell function. Immunology. 2019;156:74–85. Zanini G, Selleri V, Nasi M, De Gaetano A, Martinelli I, Gianferrari G, Lofaro FD, Boraldi F, Mandrioli J, Pinti M. Mitochondrial and Endoplasmic Reticulum Alterations in a Case of Amyotrophic Lateral Sclerosis Caused by TDP-43 A382T Mutation. Int J Mol Sci 2022, 23. Wisniewski JR. Filter-Aided Sample Preparation for Proteome Analysis. Methods Mol Biol. 2018;1841:3–10. Lofaro FD, Cisterna B, Lacavalla MA, Boschi F, Malatesta M, Quaglino D, Zancanaro C, Boraldi F. Age-Related Changes in the Matrisome of the Mouse Skeletal Muscle. Int J Mol Sci 2021, 22. Zhang Q, Wu L, Bai B, Li D, Xiao P, Li Q, Zhang Z, Wang H, Li L, Jiang Q. Quantitative Proteomics Reveals Association of Neuron Projection Development Genes ARF4, KIF5B, and RAB8A With Hirschsprung Disease. Mol Cell Proteom. 2021;20:100007. Roederer M, Nozzi JL, Nason MC. SPICE: exploration and analysis of post-cytometric complex multivariate datasets. Cytometry A. 2011;79:167–74. Weber LM, Nowicka M, Soneson C, Robinson MD. diffcyt: Differential discovery in high-dimensional cytometry via high-resolution clustering. Commun Biol. 2019;2:183. Tables Table 1 NSCLC patients’ clinical characteristics. SD, Standard Deviation; no, number; COPD, Chronic obstructive pulmonary disease; ADK, adenocarcinoma; CT, Computed Tomography, PET, Positron Emission Tomography. Parameters Patients (n = 37) Clinical features of patients Age at surgery, mean (± SD) 71.4 (± 7.2) Gender, no. (%) Female 19.0 (51.4) Male 18.0 (48.6) Smoking status - no. (%) Current smoker 26.0 (70.3) Former smoker 3.0 (8.1) Cigarette packs for year, mean ( ± SD) 36.3 (± 33.5) Never smoked 8.0 (21.6) Common comorbidities, no. (%) Diabetes mellitus, type II 3.0 (8.1) Hypertension 8.0 (21.6) Dyslipidaemia 3.0 (8.1) COPD 7.0 (18.9) Concomitant comorbidities* 15.0 (40.5) None 1.0 Previous tumours, no. (%) * 10.0 (27.0) Recurrence, no. (%) 7.0 (18.9) Survival follow-up, no. live (%) 35.0 (94.6) NSCLC characterization Tumour type, no. (%) Adenocarcinoma 25.0 (67.6) ADK lepidic − 0.0 (0.0) ADK acinar − 12.0 (48.0) ADK solid − 3.0 (12.0) ADK papillary − 1.0 (4.0) ADK micropapillary − 0.0 (0.0) Other* − 9.0 (36.0) Squamous-cell carcinoma 8.0 (21.6) Keratinizing − 6.0 (75.0) Nonkeratinizing − 2.0 (25.0) Other* 4.0 (10.8) pStage TNM 8th edition, no. (%) IA1 1.0 (2.7) IA2 6.0 (16.2) IA3 7.0 (18.9) IB 5.0 (13.5) IIA 5.0 (13.5) IIB 9.0 (24.3) IIIA 4.0 (10.8) Chest CT, no. (range mm) ≤ 27 mm 19.0 (15.0–27.0) > 27 mm 18.0 (30.0–97.0) FDG total-body PET, no. (range) ≤ 7.1 SUV Max Low 18.0 (1.2–7.0) > 7.1 SUV Max High 19.0 (7.2–31.0) Adjuvant radiotherapy, no. (%) 1.0 (2.7) Adjuvant chemotherapy, no. (%) 6.0 (16.2) Vascular infiltration, no. (%) 3.0 (8.1) Visceral pleural infiltration, no. (%) 11.0 (29.7) * See Supplementary Table 1 for more detailed information. Additional Declarations No competing interests reported. Supplementary Files SupplementaryTable1clinicaldata.xlsx SupplementaryTable2massandflowcytometrymAbs.docx SupplementaryfiguresTILv3DLT.pptx Cite Share Download PDF Status: Published Journal Publication published 14 Jan, 2025 Read the published version in Molecular Cancer → Version 1 posted Editorial decision: Revision requested 03 Jun, 2024 Reviews received at journal 02 Jun, 2024 Reviews received at journal 28 May, 2024 Reviewers agreed at journal 23 May, 2024 Reviewers agreed at journal 22 May, 2024 Reviewers agreed at journal 09 Apr, 2024 Reviews received at journal 07 Mar, 2024 Reviewers agreed at journal 27 Feb, 2024 Reviewers agreed at journal 20 Feb, 2024 Reviewers invited by journal 30 Jan, 2024 Editor assigned by journal 25 Jan, 2024 Submission checks completed at journal 25 Jan, 2024 First submitted to journal 23 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-3891288","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":269945639,"identity":"cd2da6c1-9600-462c-9581-05e7a3e7e3c1","order_by":0,"name":"Domenico Lo Tartaro","email":"","orcid":"","institution":"University of Modena and Reggio Emilia","correspondingAuthor":false,"prefix":"","firstName":"Domenico","middleName":"Lo","lastName":"Tartaro","suffix":""},{"id":269945640,"identity":"0aeb1583-56ab-4255-a516-edb7d11686ef","order_by":1,"name":"Beatrice Aramini","email":"","orcid":"","institution":"Ospedale G.B. Morgagni - L.Pierantoni","correspondingAuthor":false,"prefix":"","firstName":"Beatrice","middleName":"","lastName":"Aramini","suffix":""},{"id":269945641,"identity":"92637d11-c9a5-49e9-940d-a6bc40d0ae19","order_by":2,"name":"Valentina Masciale","email":"","orcid":"","institution":"University of Modena and Reggio Emilia","correspondingAuthor":false,"prefix":"","firstName":"Valentina","middleName":"","lastName":"Masciale","suffix":""},{"id":269945642,"identity":"87ce2f4c-7ed5-4be8-a798-a8afc90b84d3","order_by":3,"name":"Nikolaos Paschalidis","email":"","orcid":"","institution":"Biomedical Research Foundation of the Academy of Athens","correspondingAuthor":false,"prefix":"","firstName":"Nikolaos","middleName":"","lastName":"Paschalidis","suffix":""},{"id":269945643,"identity":"62e7efda-7ad2-48b9-83f5-5644759399d8","order_by":4,"name":"Francesco Demetrio Lofaro","email":"","orcid":"","institution":"University of Modena and Reggio Emilia","correspondingAuthor":false,"prefix":"","firstName":"Francesco","middleName":"Demetrio","lastName":"Lofaro","suffix":""},{"id":269945644,"identity":"a1dce651-dad8-40f1-8730-2fee6a4fbb53","order_by":5,"name":"Anita Neroni","email":"","orcid":"","institution":"University of Modena and Reggio Emilia","correspondingAuthor":false,"prefix":"","firstName":"Anita","middleName":"","lastName":"Neroni","suffix":""},{"id":269945645,"identity":"3f5d12a1-9a3e-4511-aafe-dd0259a8544a","order_by":6,"name":"Rebecca Borella","email":"","orcid":"","institution":"University of Modena and Reggio Emilia","correspondingAuthor":false,"prefix":"","firstName":"Rebecca","middleName":"","lastName":"Borella","suffix":""},{"id":269945646,"identity":"8bf9c745-831c-4630-afae-ff77f95aaa3b","order_by":7,"name":"Elena Santacroce","email":"","orcid":"","institution":"University of Modena and Reggio Emilia","correspondingAuthor":false,"prefix":"","firstName":"Elena","middleName":"","lastName":"Santacroce","suffix":""},{"id":269945647,"identity":"6d892813-ff3d-4fda-bcc6-67a25ee5ec77","order_by":8,"name":"Alin Liviu Ciobanu","email":"","orcid":"","institution":"University of Modena and Reggio Emilia","correspondingAuthor":false,"prefix":"","firstName":"Alin","middleName":"Liviu","lastName":"Ciobanu","suffix":""},{"id":269945648,"identity":"15651364-5a3b-40de-922b-65177c6616f7","order_by":9,"name":"Anna Valeria Samarelli","email":"","orcid":"","institution":"University of Modena and Reggio Emilia","correspondingAuthor":false,"prefix":"","firstName":"Anna","middleName":"Valeria","lastName":"Samarelli","suffix":""},{"id":269945649,"identity":"c70f7dde-f453-4f0e-bf52-b49d92ed30b7","order_by":10,"name":"Federica Boraldi","email":"","orcid":"","institution":"University of Modena and Reggio Emilia","correspondingAuthor":false,"prefix":"","firstName":"Federica","middleName":"","lastName":"Boraldi","suffix":""},{"id":269945650,"identity":"699cea22-b8bb-41c2-81cb-80e0226dc74d","order_by":11,"name":"Daniela Quaglino","email":"","orcid":"","institution":"University of Modena and Reggio Emilia","correspondingAuthor":false,"prefix":"","firstName":"Daniela","middleName":"","lastName":"Quaglino","suffix":""},{"id":269945651,"identity":"30ebe359-9a66-48ed-94aa-1a1dd1a299eb","order_by":12,"name":"Alessandra Dubini","email":"","orcid":"","institution":"Ospedale G.B. Morgagni - L.Pierantoni","correspondingAuthor":false,"prefix":"","firstName":"Alessandra","middleName":"","lastName":"Dubini","suffix":""},{"id":269945652,"identity":"279a2291-bbe3-45aa-bf30-468be74edf60","order_by":13,"name":"Michele Gaudio","email":"","orcid":"","institution":"Ospedale G.B. Morgagni - L.Pierantoni","correspondingAuthor":false,"prefix":"","firstName":"Michele","middleName":"","lastName":"Gaudio","suffix":""},{"id":269945653,"identity":"7133d1d4-899c-40a9-91c9-72f870dcb807","order_by":14,"name":"Gloria Manzotti","email":"","orcid":"","institution":"Santa Maria Nuova Hospital","correspondingAuthor":false,"prefix":"","firstName":"Gloria","middleName":"","lastName":"Manzotti","suffix":""},{"id":269945654,"identity":"f02286bd-5b2c-41e5-8fe5-4ca189eedf75","order_by":15,"name":"Francesca Reggiani","email":"","orcid":"","institution":"Santa Maria Nuova Hospital","correspondingAuthor":false,"prefix":"","firstName":"Francesca","middleName":"","lastName":"Reggiani","suffix":""},{"id":269945655,"identity":"6f92345f-7372-4ae5-b628-59fa2e293a38","order_by":16,"name":"Federica Torricelli","email":"","orcid":"","institution":"Santa Maria Nuova Hospital","correspondingAuthor":false,"prefix":"","firstName":"Federica","middleName":"","lastName":"Torricelli","suffix":""},{"id":269945656,"identity":"c163abaa-ebca-4796-ad3c-1ce716bc9f0d","order_by":17,"name":"Alessia Ciarrocchi","email":"","orcid":"","institution":"Santa Maria Nuova Hospital","correspondingAuthor":false,"prefix":"","firstName":"Alessia","middleName":"","lastName":"Ciarrocchi","suffix":""},{"id":269945657,"identity":"19f68441-63a5-4864-adbd-549686184d23","order_by":18,"name":"Antonino Neri","email":"","orcid":"","institution":"Santa Maria Nuova Hospital","correspondingAuthor":false,"prefix":"","firstName":"Antonino","middleName":"","lastName":"Neri","suffix":""},{"id":269945658,"identity":"348d4bb0-e041-42c3-b29c-461fc0e58431","order_by":19,"name":"Federica Bertolini","email":"","orcid":"","institution":"Policlinico di Modena","correspondingAuthor":false,"prefix":"","firstName":"Federica","middleName":"","lastName":"Bertolini","suffix":""},{"id":269945659,"identity":"a390bf9b-6916-40fa-9587-0e03e501952a","order_by":20,"name":"Massimo Dominici","email":"","orcid":"","institution":"University of Modena and Reggio Emilia","correspondingAuthor":false,"prefix":"","firstName":"Massimo","middleName":"","lastName":"Dominici","suffix":""},{"id":269945660,"identity":"9ca268ba-5e11-431d-ae3e-141a17708f5b","order_by":21,"name":"Pier Luigi Filosso","email":"","orcid":"","institution":"University of Modena and Reggio Emilia","correspondingAuthor":false,"prefix":"","firstName":"Pier","middleName":"Luigi","lastName":"Filosso","suffix":""},{"id":269945661,"identity":"3269ba03-d2c5-4396-b94e-31bd3d6eabdf","order_by":22,"name":"Franco Stella","email":"","orcid":"","institution":"Ospedale G.B. Morgagni - L.Pierantoni","correspondingAuthor":false,"prefix":"","firstName":"Franco","middleName":"","lastName":"Stella","suffix":""},{"id":269945662,"identity":"9634902a-eee5-4c7e-ae6f-6a6c25f05d66","order_by":23,"name":"Lara Gibellini","email":"","orcid":"","institution":"University of Modena and Reggio Emilia","correspondingAuthor":false,"prefix":"","firstName":"Lara","middleName":"","lastName":"Gibellini","suffix":""},{"id":269945663,"identity":"6c3f246c-44c6-44d0-9c93-86a175888c05","order_by":24,"name":"Sara De Biasi","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA7UlEQVRIiWNgGAWjYNACNoYENiB14AMDAw9ExIBILQdnILQQ0gPUAqKYeRBCuLXIt599+IGhzCaPj/34w8O2bXdkdNt7Hz5gKPiDU4vBmXRjCYZzacVsPDkGh3PbnvGYnTlubIDPYQYMaQwSjG2HE9sYchiAWg7zmN1IY5PAp0W+/xnzD8a2/4lt/M8fHLaEaGH/gdf7IDMZ2w4ktkkkGBxmhNqCN8QMbjxjs0g4l1zMJvHG4GDPOaCWM8eYgdqN8TgsjfnGhzK7PPn+9McffpQdtjc73sb44cMfOdwOA4EEIkRGwSgYBaNgFJACALNmUOFogGw4AAAAAElFTkSuQmCC","orcid":"","institution":"University of Modena and Reggio Emilia","correspondingAuthor":true,"prefix":"","firstName":"Sara","middleName":"","lastName":"De Biasi","suffix":""},{"id":269945664,"identity":"f7e53804-5413-4d85-921a-8fa326db992b","order_by":25,"name":"Andrea Cossarizza","email":"","orcid":"","institution":"University of Modena and Reggio Emilia","correspondingAuthor":false,"prefix":"","firstName":"Andrea","middleName":"","lastName":"Cossarizza","suffix":""}],"badges":[],"createdAt":"2024-01-23 14:05:14","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-3891288/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3891288/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12943-024-02209-2","type":"published","date":"2025-01-14T15:57:31+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":50383398,"identity":"5293f967-ac1b-43a6-8fe6-1616e2e592f1","added_by":"auto","created_at":"2024-01-30 17:25:36","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":228251,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eNSCLC tumors are enriched in VISTA-expressing Bregs.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA.\u0026nbsp;\u0026nbsp;\u0026nbsp; The UMAP plot illustrates the 2D spatial distribution of 1,011,489 CD19\u003csup\u003e+\u003c/sup\u003e B cells (Blood = 538,862 and Tumor = 472,627) across 15 samples, integrated with FlowSOM-based clusters.\u003c/p\u003e\n\u003cp\u003eB.\u0026nbsp;\u0026nbsp;\u0026nbsp; Heatmap of the median marker intensities of the 14 lineage markers across the 20 cell populations obtained with FlowSOM algorithm (meta-20). The colors of cluster_id column correspond to the colors used to label the UMAP plot clusters. The color in the heatmap is referred to the median of the \u003cem\u003earcsinh\u003c/em\u003e marker expression (0–1 scaled) calculated over cells from all the samples. Blue represents lower expression, while red represents higher expression. Light gray balloon along the rows (clusters) indicates the relative sizes of the clusters. Tr, Transitional B cells; PCs, plasma cells; PB, plasmablasts; Bregs, B regulatory cells; MBC sw, switched memory B cell; MBC usw, unswitched memory B cell.\u003c/p\u003e\n\u003cp\u003eC.\u0026nbsp;\u0026nbsp;\u0026nbsp; Representative dot plots showing manual gating analysis of Bregs cells (C18 and C20) from a blood and tumor sample. Numbers in the dot plots indicate the percentage of cells identified by the gates.\u003c/p\u003e\n\u003cp\u003eD.\u0026nbsp;\u0026nbsp;\u0026nbsp; Summary of the C18 and C20 cell frequencies in blood (n = 15) and tumor (n = 15) samples, as depicted in Figure C. Wilcoxon signed-rank test was used to calculate statistical significance, p-value was indicated by * p\u0026lt; 0.05, ** p\u0026lt; 0.01, and *** p\u0026lt; 0.001.\u003c/p\u003e\n\u003cp\u003eE.\u0026nbsp;\u0026nbsp;\u0026nbsp; (Top) Precursor\u0026nbsp;peaks (MS\u003csup\u003e1\u003c/sup\u003e) of VISTA peptides identified in B cells by\u0026nbsp;nanoUHPLC-ESI-HybridQuadrupole-Orbitrap\u0026nbsp;Mass\u0026nbsp;Spectrometer from B cells replicate 2. (bottom) Table reports all identified VISTA peptides in B cells replicate 2.\u003c/p\u003e\n\u003cp\u003eF.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; DiffusionMap and pseudotime score of CD19\u003csup\u003e+\u003c/sup\u003e B cells.\u003c/p\u003e","description":"","filename":"Figure1DLT.png","url":"https://assets-eu.researchsquare.com/files/rs-3891288/v1/f7483e0f69b6f5a94c8be39a.png"},{"id":50385404,"identity":"49d499a5-2f12-42ad-ba4f-7cec93420b63","added_by":"auto","created_at":"2024-01-30 17:33:36","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":194443,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eVISTA-expressing B cells infiltrating NSCLC tumors are metabolically active and functionally suppressive.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA.\u0026nbsp;\u0026nbsp;\u0026nbsp; UMAP plot, calculated using only metabolic features, illustrates the 2D spatial distribution of tumor infiltrating 148,915 CD19\u003csup\u003e+\u003c/sup\u003e B cells from 12 samples, colored by their FlowSOM-based scMEP states.\u003c/p\u003e\n\u003cp\u003eB.\u0026nbsp;\u0026nbsp;\u0026nbsp; Heatmap displaying the enrichment (in red) or depletion (in blue) of metabolic proteins expression across CD19\u003csup\u003e+\u003c/sup\u003e B cell scMEP states, calculated utilizing exclusively metabolic features.\u003c/p\u003e\n\u003cp\u003eC.\u0026nbsp;\u0026nbsp;\u0026nbsp; UMAP visualization of CD19\u003csup\u003e+ \u003c/sup\u003eB cells colored by normalized expression of the indicated metabolic proteins.\u003c/p\u003e\n\u003cp\u003eD.\u0026nbsp;\u0026nbsp;\u0026nbsp; Heatmap showing the expression of immunological markers (excluded from FlowSOM clustering) across scMEP states.\u003c/p\u003e\n\u003cp\u003eE.\u0026nbsp;\u0026nbsp;\u0026nbsp; UMAP visualization of CD19\u003csup\u003e+\u003c/sup\u003e B cells colored by normalized expression of the indicated phenotypic proteins.\u003c/p\u003e\n\u003cp\u003eF.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; Comparison of the total production of IFN-γ, TNF, IL-6, IL-10, TGF-β, and GM-CSF between tumor infiltrating VISTA\u003csup\u003e+\u003c/sup\u003e and VISTA\u003csup\u003e−\u003c/sup\u003e B cells after in vitro stimulation. Data represent individual values from ten NSCLC patients. Mean (center bar) ± SEM (upper and lower bars). Wilcoxon signed-rank test was used to calculate statistical significance, p-value was indicated by * p\u0026lt; 0.05, ** p\u0026lt; 0.01, and *** p\u0026lt; 0.001.\u003c/p\u003e\n\u003cp\u003eG.\u0026nbsp;\u0026nbsp;\u0026nbsp; Frequency of VISTA\u003csup\u003e+\u003c/sup\u003e (blue) and VISTA\u003csup\u003e−\u003c/sup\u003e (red) tumor infiltrating B cells producing different combinations of IFN-γ, TNF, IL-6, IL-10, TGF-β, and GM-CSF after PMA/ionomycin stimulation (n = 10). Data are given as mean ± SEM. Wilcoxon signed-rank test was used to calculate statistical significance, p-value was indicated by * p\u0026lt; 0.05, ** p\u0026lt; 0.01, and *** p\u0026lt; 0.001.\u003c/p\u003e","description":"","filename":"Figure2DLT.png","url":"https://assets-eu.researchsquare.com/files/rs-3891288/v1/803618bb8800948b119e54ed.png"},{"id":50383402,"identity":"1e821ce1-f26f-479d-8412-d5d60c8d16e9","added_by":"auto","created_at":"2024-01-30 17:25:36","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":313578,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eVISTA-PSGL-1 axis mediate B and T cell-to-cell interaction in the NSCLC TLSs.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA.\u0026nbsp;\u0026nbsp;\u0026nbsp; Spatial B and T lymphocytes location within NSCLC tissue (x, y coordinates). Blue dots represent B lymphocytes while red or green represent CD4\u003csup\u003e+\u003c/sup\u003e and CD8\u003csup\u003e+\u003c/sup\u003e T lymphocytes, respectively. Scale bar 1 mm. White boxed regions identify five different tertiary lymphoid structures (TLS).\u003c/p\u003e\n\u003cp\u003eB.\u0026nbsp;\u0026nbsp;\u0026nbsp; Magnification of boxed region (#1 in A).\u003c/p\u003e\n\u003cp\u003eC.\u0026nbsp;\u0026nbsp;\u0026nbsp; (left) Plot showing Euclidian distance and (right) the number of cell-to-cell neighborhood interaction. Wilcoxon signed-rank followed by Benjamini-Hochberg correction was used to calculate statistical significance. Exact p-value were reported in the figure.\u003c/p\u003e\n\u003cp\u003eD.\u0026nbsp;\u0026nbsp;\u0026nbsp; NicheNet’s ligand activity prediction performed using imaging-based spatial RNA-seq data (see Methods). Purple square reports the predicted ligand-receptor activity between B (sender) and CD4\u003csup\u003e+\u003c/sup\u003e T cell (receiver). The predicted B cell ligands affinity was ranked by Pearson correlation coefficient (orange square). Violet square represents the potential regulatory impact of the B-T interaction on the tumor-specific T cells gene signature.\u003c/p\u003e\n\u003cp\u003eE.\u0026nbsp;\u0026nbsp;\u0026nbsp; (top) A magnification of the boxed region (#1 in A) illustrating the expression of \u003cem\u003eVSIR\u003c/em\u003e, \u003cem\u003eSELPLG\u003c/em\u003e, and \u003cem\u003eIL10\u003c/em\u003e on B cells, CD4\u003csup\u003e+\u003c/sup\u003e and CD8\u003csup\u003e+\u003c/sup\u003e T cells. (bottom) Dot plots showing the percentage of B cells, CD4\u003csup\u003e+\u003c/sup\u003e, and CD8\u003csup\u003e+\u003c/sup\u003e T cells expressing \u003cem\u003eVSIR\u003c/em\u003e, \u003cem\u003eSELPLG\u003c/em\u003e, or \u003cem\u003eVSIR\u003c/em\u003e + \u003cem\u003eIL10 \u003c/em\u003ewithin the five TLS regions, as indicated in figure panel \u003cstrong\u003eA\u003c/strong\u003e. Kruskal-Wallis test followed by Benjamini-Hochberg correction was used to calculate statistical significance. Exact p-value were reported in the figure.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e","description":"","filename":"Figure3DLT.png","url":"https://assets-eu.researchsquare.com/files/rs-3891288/v1/b6b13e73ce419bc046104ad5.png"},{"id":50383400,"identity":"1eb37a52-f574-430c-8bd7-127233d81286","added_by":"auto","created_at":"2024-01-30 17:25:36","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":335197,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eNSCLC infiltrating PSGL-1\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e+\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003eCD4\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e+\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e T cells display T\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003eRM\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e phenotype with improved metabolic and cytotoxic traits.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA.\u0026nbsp;\u0026nbsp;\u0026nbsp; UMAP plot shows the 2D spatial distribution of 5.946.764 of CD4\u003csup\u003e+\u003c/sup\u003e T cells (PBMC = 5.108.956 and TILs = 837.808) from 15 samples embedded with FlowSOM clusters.\u003c/p\u003e\n\u003cp\u003eB.\u0026nbsp;\u0026nbsp;\u0026nbsp; Heatmap of the median marker intensities of the 16 lineage markers across the 30 cell populations obtained with FlowSOM algorithm (meta30). The colors of cluster_id column correspond to the colors used to label the UMAP plot clusters. The color in the heatmap is referred to the median of the \u003cem\u003earcsinh\u003c/em\u003e marker expression (0–1 scaled) calculated over cells from all the samples. Blue represents lower expression, while red represents higher expression. Light gray balloon along the rows (clusters) indicates the relative sizes of the clusters. T\u003csub\u003eRM\u003c/sub\u003e, Tissue resident memory cells; Tconv, conventional T cells.\u003c/p\u003e\n\u003cp\u003eC.\u0026nbsp;\u0026nbsp;\u0026nbsp; (left) Representative histograms depicting the distribution of the differentially expressed markers on tumor-infiltrating PSGL-1\u003csup\u003e+\u003c/sup\u003e (red) and PSGL-1\u003csup\u003e−\u003c/sup\u003e (black) CD4\u003csup\u003e+\u003c/sup\u003e T cells. (right) Dot plots showing the MFI of the differentially expressed markers on tumor-infiltrating PSGL-1\u003csup\u003e+\u003c/sup\u003e and PSGL-1\u003csup\u003e−\u003c/sup\u003e CD4\u003csup\u003e+\u003c/sup\u003e T cells. Wilcoxon signed-rank test was used to calculate statistical significance, p-value was indicated by * p\u0026lt; 0.05, ** p\u0026lt; 0.01, and *** p\u0026lt; 0.001. The markers that remained unchanged were not shown.\u003c/p\u003e\n\u003cp\u003eD.\u0026nbsp;\u0026nbsp;\u0026nbsp; The UMAP plot, calculated using only metabolic features, illustrates the 2D spatial distribution of tumor infiltrating 391,088 CD4\u003csup\u003e+\u003c/sup\u003e T cells from 12 samples, colored by their FlowSOM-based scMEP states.\u003c/p\u003e\n\u003cp\u003eE.\u0026nbsp;\u0026nbsp;\u0026nbsp; Heatmap displaying the enrichment (in red) or depletion (in blue) of metabolic proteins expression across CD4\u003csup\u003e+\u003c/sup\u003e T cell scMEP states, calculated utilizing exclusively metabolic features.\u003c/p\u003e\n\u003cp\u003eF.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; UMAP visualization of CD4\u003csup\u003e+ \u003c/sup\u003eT cells colored by normalized expression of the indicated metabolic proteins.\u003c/p\u003e\n\u003cp\u003eG.\u0026nbsp;\u0026nbsp;\u0026nbsp; Heatmap showing the expression of immunological markers (excluded from FlowSOM clustering) across scMEP states.\u003c/p\u003e\n\u003cp\u003eH.\u0026nbsp;\u0026nbsp;\u0026nbsp; UMAP visualization of CD4\u003csup\u003e+\u003c/sup\u003e T cells colored by normalized expression of the indicated phenotypic proteins.\u003c/p\u003e","description":"","filename":"Figure4DLT.png","url":"https://assets-eu.researchsquare.com/files/rs-3891288/v1/995da2fefac13f181dcfb5fb.png"},{"id":50383399,"identity":"fa2c8d71-cae9-4994-84f8-d6552313012a","added_by":"auto","created_at":"2024-01-30 17:25:36","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":349145,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eNSCLC infiltrating PSGL-1\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e+\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003eCD8\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e+\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e T cells are highly cytotoxic and metabolically active.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA.\u0026nbsp;\u0026nbsp;\u0026nbsp; UMAP plot shows the 2D spatial distribution of 2.414.533 of CD8\u003csup\u003e+\u003c/sup\u003e T cells (PBMC = 1.801.510 and TILs = 613.023) from 15 samples embedded with FlowSOM clusters.\u003c/p\u003e\n\u003cp\u003eB.\u0026nbsp;\u0026nbsp;\u0026nbsp; Heatmap of the median marker intensities of the 16 lineage markers across the 30 cell populations obtained with FlowSOM algorithm (meta30). The colors of cluster_id column correspond to the colors used to label the UMAP plot clusters. The color in the heatmap is referred to the median of the \u003cem\u003earcsinh\u003c/em\u003e marker expression (0–1 scaled) calculated over cells from all the samples. Blue represents lower expression, while red represents higher expression. Light gray balloon along the rows (clusters) indicates the relative sizes of the clusters. T\u003csub\u003eRM\u003c/sub\u003e, Tissue resident memory cells.\u003c/p\u003e\n\u003cp\u003eC.\u0026nbsp;\u0026nbsp;\u0026nbsp; (left) Representative histograms depicting the distribution of the differentially expressed markers on tumor-infiltrating PSGL-1\u003csup\u003e+\u003c/sup\u003e (red) and PSGL-1\u003csup\u003e−\u003c/sup\u003e (black) CD8\u003csup\u003e+\u003c/sup\u003e T cells. (right) Dot plots showing the MFI of differentially expressed on tumor-infiltrating PSGL-1\u003csup\u003e+\u003c/sup\u003e and PSGL-1\u003csup\u003e−\u003c/sup\u003e CD8\u003csup\u003e+\u003c/sup\u003e T cells. Wilcoxon signed-rank test was used to calculate statistical significance, p-value was indicated by * p\u0026lt; 0.05, ** p\u0026lt; 0.01, and *** p\u0026lt; 0.001. The markers that remained unchanged were not shown.\u003c/p\u003e\n\u003cp\u003eD.\u0026nbsp;\u0026nbsp;\u0026nbsp; The UMAP plot, calculated using only metabolic features, illustrates the 2D spatial distribution of tumor infiltrating 273,043 CD8\u003csup\u003e+\u003c/sup\u003e T cells from 12 samples, colored by their FlowSOM-based scMEP states.\u003c/p\u003e\n\u003cp\u003eE.\u0026nbsp;\u0026nbsp;\u0026nbsp; Heatmap displaying the enrichment (in red) or depletion (in blue) of metabolic proteins expression across CD8\u003csup\u003e+\u003c/sup\u003e T cell scMEP states, calculated utilizing exclusively metabolic features.\u003c/p\u003e\n\u003cp\u003eF.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; UMAP visualization of CD8\u003csup\u003e+ \u003c/sup\u003eT cells colored by normalized expression of the indicated metabolic proteins.\u003c/p\u003e\n\u003cp\u003eG.\u0026nbsp;\u0026nbsp;\u0026nbsp; Heatmap showing immunological markers (excluded from FlowSOM clustering) across scMEP states.\u003c/p\u003e\n\u003cp\u003eH.\u0026nbsp;\u0026nbsp;\u0026nbsp; UMAP visualization of CD8\u003csup\u003e+\u003c/sup\u003e T cells colored by normalized expression of the indicated phenotypic proteins.\u003c/p\u003e","description":"","filename":"Figure5DLT.png","url":"https://assets-eu.researchsquare.com/files/rs-3891288/v1/4bb6e691f5319c9b9727f939.png"},{"id":50383401,"identity":"7fa45868-dcb7-46de-bef6-6987a870e629","added_by":"auto","created_at":"2024-01-30 17:25:36","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":189115,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePSGL-1-expressing CD4\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e+\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e and CD8\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e+\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e T cells are efficient in rapidly producing antitumor molecules.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA.\u0026nbsp;\u0026nbsp;\u0026nbsp; Representative dot plot illustrating the selection of CD4\u003csup\u003e+ \u003c/sup\u003eTconv and T\u003csub\u003eRM\u003c/sub\u003e based on CD103 and Foxp3 expression. Tregs were excluded from the analysis.\u003c/p\u003e\n\u003cp\u003eB.\u0026nbsp;\u0026nbsp;\u0026nbsp; Comparison of the overall production of GNLY and GZMB in PSGL-1\u003csup\u003e+\u003c/sup\u003e and PSGL-1\u003csup\u003e−\u003c/sup\u003e tumor-infiltrating CD4\u003csup\u003e+\u003c/sup\u003e Tconv and T\u003csub\u003eRM\u003c/sub\u003e lymphocytes following in vitro stimulation with PMA/ionomycin (n = 8).\u003c/p\u003e\n\u003cp\u003eC.\u0026nbsp;\u0026nbsp;\u0026nbsp; Representative dot plots showing GNLY and GZMB expression in PSGL-1\u003csup\u003e+\u003c/sup\u003e and PSGL-1\u003csup\u003e−\u003c/sup\u003e tumor-infiltrating CD4\u003csup\u003e+\u003c/sup\u003e Tconv and T\u003csub\u003eRM\u003c/sub\u003e lymphocytes.\u003c/p\u003e\n\u003cp\u003eD.\u0026nbsp;\u0026nbsp;\u0026nbsp; Comparison of the overall production of IL-2, IL-17, IFN-γ and TNF in PSGL-1\u003csup\u003e+\u003c/sup\u003e and PSGL-1\u003csup\u003e−\u003c/sup\u003e tumor-infiltrating CD4\u003csup\u003e+\u003c/sup\u003e Tconv and T\u003csub\u003eRM\u003c/sub\u003e lymphocytes following in vitro stimulation with PMA/ionomycin (n = 8).\u003c/p\u003e\n\u003cp\u003eE.\u0026nbsp;\u0026nbsp;\u0026nbsp; Frequency of PSGL-1\u003csup\u003e+\u003c/sup\u003e and PSGL-1\u003csup\u003e−\u003c/sup\u003e tumor-infiltrating CD4\u003csup\u003e+\u003c/sup\u003e Tconv and T\u003csub\u003eRM\u003c/sub\u003e lymphocytes producing different combinations of IL-2, IL-17, IFN-γ and TNF after PMA/ionomycin stimulation (n = 10). Data are given as mean ± SEM. Wilcoxon signed-rank test was used to calculate statistical significance, p-value was indicated by * p\u0026lt; 0.05, ** p\u0026lt; 0.01, and *** p\u0026lt; 0.001.\u003c/p\u003e\n\u003cp\u003eF.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; Representative dot plot illustrating the selection of CD8\u003csup\u003e+ \u003c/sup\u003eTconv and T\u003csub\u003eRM\u003c/sub\u003e based on CD103 and Foxp3 expression.\u003c/p\u003e\n\u003cp\u003eG.\u0026nbsp;\u0026nbsp;\u0026nbsp; Comparison of the overall production of GNLY and GZMB in PSGL-1\u003csup\u003e+\u003c/sup\u003e and PSGL-1\u003csup\u003e−\u003c/sup\u003e tumor-infiltrating CD8\u003csup\u003e+\u003c/sup\u003e Tconv and T\u003csub\u003eRM\u003c/sub\u003e lymphocytes following in vitro stimulation with PMA/ionomycin (n = 8).\u003c/p\u003e\n\u003cp\u003eH.\u0026nbsp;\u0026nbsp;\u0026nbsp; Representative dot plots showing GNLY and GZMB expression in PSGL-1\u003csup\u003e+\u003c/sup\u003e and PSGL-1\u003csup\u003e−\u003c/sup\u003e tumor-infiltrating CD8\u003csup\u003e+\u003c/sup\u003e Tconv and T\u003csub\u003eRM\u003c/sub\u003e lymphocytes.\u003c/p\u003e\n\u003cp\u003eI.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; Comparison of the overall production of IL-2, IL-17, IFN-γ and TNF in PSGL-1\u003csup\u003e+\u003c/sup\u003e and PSGL-1\u003csup\u003e−\u003c/sup\u003e tumor-infiltrating CD8\u003csup\u003e+\u003c/sup\u003e Tconv and T\u003csub\u003eRM\u003c/sub\u003e lymphocytes following in vitro stimulation with PMA/ionomycin (n = 8).\u003c/p\u003e\n\u003cp\u003eJ.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; Frequency of PSGL-1\u003csup\u003e+\u003c/sup\u003e and PSGL-1\u003csup\u003e−\u003c/sup\u003e tumor-infiltrating CD8\u003csup\u003e+\u003c/sup\u003e Tconv and T\u003csub\u003eRM\u003c/sub\u003e lymphocytes producing different combinations of IL-2, IL-17, IFN-γ and TNF after PMA/ionomycin stimulation (n = 10). Data are given as mean ± SEM. Wilcoxon signed-rank test was used to calculate statistical significance, P-value was indicated by * p\u0026lt; 0.05, ** p\u0026lt; 0.01, and *** p\u0026lt; 0.001.\u003c/p\u003e","description":"","filename":"Figure6DLT.png","url":"https://assets-eu.researchsquare.com/files/rs-3891288/v1/03e2a2bb8c2be453866216cc.png"},{"id":50385403,"identity":"98719305-c744-49df-999d-1a4a06da9327","added_by":"auto","created_at":"2024-01-30 17:33:36","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":179030,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEnhanced levels of tumor infiltrating VISTA\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e+\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e Bregs and PSGL-1\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e+\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e T cells are positively associated with NSCLC recurrence.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA.\u0026nbsp;\u0026nbsp;\u0026nbsp; UMAP showing CD4\u003csup\u003e+\u003c/sup\u003e T cell clusters depicted in \u003cstrong\u003eFig. 4\u003c/strong\u003e associated or not with clinical tumor relapse; cytotoxic cells (C17), CD4\u003csup\u003e+\u003c/sup\u003e Tconv expressing PSGL-1 (C15), and T\u003csub\u003eRM\u003c/sub\u003e (C30).\u003c/p\u003e\n\u003cp\u003eB.\u0026nbsp;\u0026nbsp;\u0026nbsp; (left) UMAP showing PENCIL-predicted blood and tumor CD4\u003csup\u003e+\u003c/sup\u003e T cells associated with tumor relapse in red (R), those associated with no tumor relapse in blue (NR), and unassigned cells in grey (Rejected); (right) Bar plot illustrating the percentage of indicated CD4\u003csup\u003e+\u003c/sup\u003e T cell clusters, among patients experiencing or not NSCLC relapse.\u003c/p\u003e\n\u003cp\u003eC.\u0026nbsp;\u0026nbsp;\u0026nbsp; UMAP showing CD8\u003csup\u003e+\u003c/sup\u003e T cells depicted in \u003cstrong\u003eFig. 5\u003c/strong\u003e associated or not with clinical tumor relapse; PSGL-1\u003csup\u003e− \u003c/sup\u003eT\u003csub\u003eRM\u003c/sub\u003e (C12, C22, C25) and PSGL-1\u003csup\u003e+ \u003c/sup\u003eT\u003csub\u003eRM\u003c/sub\u003e (C26).\u003c/p\u003e\n\u003cp\u003eD.\u0026nbsp;\u0026nbsp;\u0026nbsp; (left) UMAP showing PENCIL-predicted blood and tumor CD8\u003csup\u003e+\u003c/sup\u003e T cells associated with tumor relapse in red (R), those associated with no tumor relapse in blue (NR), and unassigned cells in grey (Rejected); (right) Bar plot illustrating the percentage of indicated CD8\u003csup\u003e+\u003c/sup\u003e T cell clusters, among patients experiencing or not NSCLC relapse.\u003c/p\u003e\n\u003cp\u003eE.\u0026nbsp;\u0026nbsp;\u0026nbsp; UMAP showing B cells depicted in \u003cstrong\u003eFig. 1\u003c/strong\u003e associated or not with clinical tumor relapse; memory unswitched (C1), memory-switched B cells (C14) and VISTA\u003csup\u003e+ \u003c/sup\u003eBregs (C20).\u003c/p\u003e\n\u003cp\u003eF.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; (left) UMAP showing PENCIL-predicted blood and tumor B cells associated with tumor relapse in red (R), those associated with no tumor relapse in blue (NR), and unassigned cells in grey (Rejected); (right) Bar plot illustrating the percentage of indicated B cell clusters, among patients experiencing or not NSCLC relapse.\u003c/p\u003e\n\u003cp\u003eG.\u0026nbsp;\u0026nbsp;\u0026nbsp; (left)A forest plot showing the hazard ratio (HR) and 95% confidence intervals (CI) of B and T cell clusters associated or not with tumor recurrence (from PENCIL). Diamond represents the HR, while the horizontal bars extend from the lower limit to the upper limit of the 95% CI of the estimate of the HR; (right) Table resuming HR, CI and p-value of multivariate Cox proportional-hazards analysis.\u003c/p\u003e","description":"","filename":"Figure7DLT.png","url":"https://assets-eu.researchsquare.com/files/rs-3891288/v1/11a4a11c6da3e805f696fb16.png"},{"id":74284631,"identity":"569a040e-943c-4f4c-8496-b394483d1062","added_by":"auto","created_at":"2025-01-20 16:10:00","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3945442,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3891288/v1/6efc1b61-baf0-4cd7-a0e2-027180957af9.pdf"},{"id":50383403,"identity":"f582e00e-1ad8-4b6c-8344-334380b9c09d","added_by":"auto","created_at":"2024-01-30 17:25:36","extension":"xlsx","order_by":10,"title":"","display":"","copyAsset":false,"role":"supplement","size":23789,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable1clinicaldata.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-3891288/v1/48e0ea9a7b2765bc78261bf9.xlsx"},{"id":50383405,"identity":"2d94b3c3-34c3-4acd-bb06-3210e1519008","added_by":"auto","created_at":"2024-01-30 17:25:36","extension":"docx","order_by":11,"title":"","display":"","copyAsset":false,"role":"supplement","size":57090,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable2massandflowcytometrymAbs.docx","url":"https://assets-eu.researchsquare.com/files/rs-3891288/v1/2b889870a4a21310d09d3d36.docx"},{"id":50383591,"identity":"09f301e2-758b-4434-9f06-1510f941a31f","added_by":"auto","created_at":"2024-01-30 17:26:23","extension":"pptx","order_by":12,"title":"","display":"","copyAsset":false,"role":"supplement","size":707838078,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryfiguresTILv3DLT.pptx","url":"https://assets-eu.researchsquare.com/files/rs-3891288/v1/cf61875f8e5d6fb016e44b2b.pptx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Metabolically activated and highly polyfunctional intratumoral VISTA+ regulatory B cells are associated with tumor recurrence in early stage NSCLC.","fulltext":[{"header":"Introduction","content":"\u003cp\u003eNon-small cell lung cancer (NSCLC) is the second most common cancer worldwide [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Specifically, the American Cancer Society estimates 238,340 individuals (120,790 women and 117,550 men) with a lung cancer diagnosis in 2023, and 127,070 of those individuals will die for this malignant disease. Surgery is considered the gold standard for early and locally advanced stages. However, resection is often not resolutive, even for local tumors, due to the high percentage of recurrence [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e] that varies between 30% and 55% for NSCLC patients, with a 5-year overall survival rate between 22% and 18.1% in all the stages [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. In recent years, immunotherapy has emerged as a significant advancement in the treatment of tumors, and, due to its durable effects and low rate of adverse reactions, has introduced a new paradigm for the treatment of several tumors [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Nonetheless, the majority of initially responsive patients eventually experience immune drug resistance. It has been shown that one of the main reason of low efficacy is connected with the quantity and features of tumor infiltrating lymphocytes (TIL) [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe tumor microenvironment (TME) of NSCLC is a complex milieu of immune cells, stromal elements, and signaling molecules that play a pivotal role in cancer progression and influence therapy response [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. In this environment, the infiltration and organization of immune cells in tertiary lymphoid structures (TLS) is crucial for the prolonged survival of NSCLC patients [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Tumor-associated TLS are indeed a privileged site for T cell differentiation and activation. Furthermore, TLS are associated with T helper type-1 (Th1) and cytotoxic immune signatures in lung, breast, and gastric cancers, indicating that they imprint the local immune microenvironment [\u003cspan additionalcitationids=\"CR12\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Among the diverse array of immune cells populating the TME, B cells have emerged as central players, orchestrating immune responses that profoundly influence tumor fate [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. The repertoire of tumor-infiltrating B cell phenotypes includes naive, activated, and memory B cells, germinal center B cells, and plasma cells [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. While generally aligning with canonical B cell subsets identified in healthy donors, tumor-infiltrating B cells demonstrate greater diversity within certain subsets, particularly within the memory compartment [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e], making B cells crucial immune response modulation to tumors and lymphoid malignancies. In particular, regulatory B cells (Bregs) constitute a newly designated subset of B cells able to regulate immune responses in inflammation and more recently, cancer [\u003cspan additionalcitationids=\"CR19\" citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. However, the prognostic value of B cells and related subsets together with their mechanisms of action for immune modulation in the TME are still a matter of debate.\u003c/p\u003e \u003cp\u003eIn recent years, along with PD-1 and CTLA-4, other immune modulators able to rewire the immune response within the TME have been taken into account. Indeed, V-domain Ig suppressor of T-cell activation (VISTA), an inhibitory immune checkpoint molecule, and its ligand P-selectin glycoprotein ligand-1 (PSGL-1, also known as CD162) multifunctional glycoprotein, have garnered attention for their roles in modulating immune cell function, angiogenesis, and tumor progression\u003c/p\u003e \u003cp\u003e[\u003cspan additionalcitationids=\"CR22 CR23 CR24 CR25 CR26 CR27 CR28 CR29 CR30\" citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eHere, we present a comprehensive analysis delineating the phenotypic, functional, and metabolic attributes of interacting B and T cells in patients with resectable NSCLC. Our findings highlight that a high frequency of tumor-infiltrating VISTA\u003csup\u003e+\u003c/sup\u003e Bregs together with high percentage of PSGL-1 expressing T cells are more represented in patients with cancer recurrence, and can likely predict this event.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e \u003cb\u003eBregs express VISTA and are enriched in the tumor microenvironment.\u003c/b\u003e \u003c/p\u003e \u003cp\u003eIn the TME, different subpopulations of B cells reside that can either promote or hamper tumor growth [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]), hence affecting patient outcomes [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. For this reason, to directly assess the landscape of tumor-infiltrating CD19\u003csup\u003e+\u003c/sup\u003e B cells, we used a 19-parameter flow cytometry panel. We investigated 472,627 cells isolated from resected tumor (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;15) paired with 538,862 cells isolated from peripheral blood (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;15), considered as reference.\u003c/p\u003e \u003cp\u003eThe percentage of B cells was higher in the TME when compared to blood (median: blood\u0026thinsp;=\u0026thinsp;4.6, tumor\u0026thinsp;=\u0026thinsp;15.8; Supplementary Fig.\u0026nbsp;1). To better characterize the landscape of CD19\u003csup\u003e+\u003c/sup\u003e B cells, an unsupervised algorithm such as FlowSOM was used, and 20 clusters were identified (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA-B). Hierarchical ordering of FlowSOM clusters (rows) and markers (columns) allowed grouping of subpopulations with similar immune characteristics. Three of these clusters (C3, C4, C10) were more abundant in the peripheral blood and displayed characteristics of na\u0026iuml;ve cells (i.e., CD27\u003csup\u003e\u0026minus;\u003c/sup\u003eIgD\u003csup\u003e+\u003c/sup\u003eIgM\u003csup\u003e+\u003c/sup\u003e). One of them (C10) was identified as proliferating na\u0026iuml;ve cells expressing Ki-67. The percentages of C2, C1, C5, C12, C15 representing respectively: transitional B cells (Tr, CD27\u003csup\u003e\u0026minus;\u003c/sup\u003eIgD\u003csup\u003e+\u003c/sup\u003eIgM\u003csup\u003e+\u003c/sup\u003eCD24\u003csup\u003e++\u003c/sup\u003e; in C2), class-unswitched memory B cells (MBC usw, i.e., CD27\u003csup\u003e+\u003c/sup\u003eIgD\u003csup\u003e+\u003c/sup\u003eIgM\u003csup\u003e+\u003c/sup\u003e; in C1 and C5) and plasmablasts (PB, CD27\u003csup\u003e++\u003c/sup\u003eCD38\u003csup\u003e++\u003c/sup\u003eKi-67\u003csup\u003e+\u003c/sup\u003e, in C12 and C15) were higher in blood when compared to TME, except for C1, which exhibited similar proportions in both blood and TME. Those in C6, C7, C13 were identified as atypical B cells (atBC, i.e., CD21\u003csup\u003e\u0026minus;\u003c/sup\u003eCD27\u003csup\u003e\u0026minus;\u003c/sup\u003e) originally found in tonsils and in peripheral blood in conditions of chronic antigen stimulation, or autoimmune disease [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. C6 was more represented in the peripheral blood, while the percentage of C7 and C13 was higher in the TME. Clusters of class-switched memory B cells (MBC sw, i.e., CD27\u003csup\u003e+\u003c/sup\u003eIgD\u003csup\u003e\u0026minus;\u003c/sup\u003eIgM\u003csup\u003e\u0026minus;\u003c/sup\u003e), including C11, C14, C9, and C8, were mostly represented in TME, suggesting that within mature TLS these cells might be not only activated but also in germinal center have experienced the antibody class switching and somatic hypermutation [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. C19 was identified as proliferating VISTA\u003csup\u003e+\u003c/sup\u003e B cells (CD38\u003csup\u003e+\u003c/sup\u003eCD138\u003csup\u003edim\u003c/sup\u003eKi67\u003csup\u003e+\u003c/sup\u003eIL-10\u003csup\u003edim\u003c/sup\u003e) and its percentage was higher in blood if compared to TME.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eFinally, two populations of B cells expressing VISTA and IL-10 were identified (C18 and C20) and classified as B regulatory cells (hereafter referred to as \u0026lsquo;VISTA\u003csup\u003e+\u003c/sup\u003e Bregs\u0026rsquo;). For the first time, we reported here that a cluster of Bregs expresses VISTA. We confirmed this data by using two different approaches: first, we performed manual gating of flow cytometry data to ascertain VISTA expression on Bregs, and we showed that not only Bregs express VISTA, but also they were much more represented in the TME (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC-D). Then, by targeted Liquid Chromatography with Tandem Mass Spectrometry (LC-MS/MS) analysis on sorted B cells, we confirmed that VISTA is expressed by peripheral blood B cells, despite at a lower level compared to what was observed within monocytes (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eE \u003cb\u003eand\u003c/b\u003e Supplementary Fig.\u0026nbsp;2).\u003c/p\u003e \u003cp\u003eFinally, by trajectory inference, also known as pseudotime analysis, we investigated the progression along the differentiation trajectory (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eF). As expected, the differentiation process started from na\u0026iuml;ve (C3) and transitional B cells (C4); then, the trajectory revealed two distinct cellular branches: one that gave origin to plasmablasts and plasma cells (C12, C17) and the other to VISTA\u003csup\u003e+\u003c/sup\u003e Bregs cell clusters (C18 and C20), suggesting that the differentiation program that they experience is quite different from that of all other B cells.\u003c/p\u003e \u003cp\u003e \u003cb\u003eVISTA\u003c/b\u003e \u003csup\u003e \u003cb\u003e+\u003c/b\u003e \u003c/sup\u003e \u003cb\u003eBregs display a metabolically activated profile and are highly polyfunctional.\u003c/b\u003e\u003c/p\u003e \u003cp\u003eGiven that cell function is strictly connected to metabolic profile, we employed single-cell metabolic regulome profiling (scMEP) using mass cytometry by time of flight (CyTOF) to investigate the metabolic profile of B cells within TME [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e] [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e]. This analysis was performed on \u003cem\u003eex vivo\u003c/em\u003e isolated TILs. We performed this unsupervised analysis by focusing on the 19 metabolic markers expressed by B cells (GLUT1, GAPDH, LDHA, HK2, PFKB4, G6PD, CytC, CS, IDH1, ATPA5, CD98, GLUD1/2, CD36, CPT1A, VDAC1, pACC, pPGC1α, HIF-1α, pNRF2), and we found a total of 9 scMEP states (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA-C). First of all, we noticed that the majority of tumor infiltrating B cells (81% of total B cells) were grouped into two states of metabolically quiescent cells (scMEP1 and 8). However, a total of 14% of cells were grouped into five states (scMEP3, 4, 5, 7 and 9) characterized by highly activated metabolism. Indeed, these states showed high level of expression of proteins belonging to glycolysis pathway (such as PFKB4 and HK2), tricarboxylic acid cycle (TCA; such as CytC and CS), amino acid pathway (such as CD98 and GLUD1/2), fatty acid oxidation (FAO; such as pACC and CPT1A), mitochondrial dynamics (such as VDAC1), pentose phosphate pathway (ppp; such as G6PD) and transcription of master regulators such as hypoxia-inducible factor-1α (HIF-1α). In particular, cells in scMEP5 state showed a high level of CD36 expression, indicating the import of fatty acids and activation of the fatty acid oxidation. scMEP2 and 6, which account for 4.8% of cells, showed mild activated metabolism, characterized by low-middle level of expression of protein related to TCA (such as CytC, ATPA5 and CS), amino acid pathway (CD98) and FAO (CD36 and pACC) accompanied by increased levels of GLUT1 (scMEP 2).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIn order to match the metabolic profiles to different cell phenotypes, we took into account the phenotypic markers and we found that almost all metabolically active B cells (scMEP 9, 5, 3, 7, 4, 6) expressed high levels of CD39, PSGL-1 and VISTA (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eD-E). In particular, scMEP5 and 3 state were composed by regulatory B cells expressing high level of IL-10 and VISTA, and medium level of Ki-67 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eD-E). scMEP 4, 7 and 9 were mainly formed by activated B cells expressing CXCR6, a marker of lung residency, and CD38, whereas scMEP 2 and 6 were composed by cells negative for CXCR6, CD38 and VISTA (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eD-E). Finally, all metabolically quiescent B cells (scMEP 1 and 8) were characterized by the lack of expression of all the aforementioned markers except CD73, a costimulatory molecule of B cells [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. These results are in line with the clustering performed by using both lineage markers as well as metabolic ones (Supplementary Fig.\u0026nbsp;3).\u003c/p\u003e \u003cp\u003eB cells are antibody-producing cells driving humoral immune responses to foreign antigens. However, B cells secrete different cytokines, which influence both pro- and anti-inflammatory immune responses. Indeed, the heterogeneity in cytokine-driven responses by B cells can range from the production of pro-inflammatory molecules such as IL-6 to the release of the immunosuppressive IL-10 [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. For this reason, we investigated the functional profile of B cells expressing VISTA (see Methods). First, we showed that VISTA was upregulated on B cell after stimulation but not in hypoxic condition (Supplementary Fig.\u0026nbsp;4A-C). The percentage of cells expressing GM-CSF, TNF, IL-6, IL-10, IFN-γ, and TGF-β was higher in VISTA\u003csup\u003e+\u003c/sup\u003e when compared to VISTA\u003csup\u003e\u0026minus;\u003c/sup\u003e B lymphocytes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eF and Supplementary Fig.\u0026nbsp;4D-E), suggesting that these cells are much more functional and able to produce anti-inflammatory and pro-inflammatory molecules than their counterpart. The analysis of their capability to simultaneously produce different molecules, \u003cem\u003ei.e.\u003c/em\u003e their polyfunctionality, revealed that a high percentage of VISTA\u003csup\u003e+\u003c/sup\u003e B cells were capable of produce simultaneously three (GM-CSF, TGF-β, and TNF) or four (GM-CSF, IL-10, TGF-β, and TNF) cytokines (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eG). The broader range of cytokine production by VISTA\u003csup\u003e+\u003c/sup\u003e B cells could evidence a novel, potential role in modulating tumor-specific T cell response.\u003c/p\u003e \u003cp\u003e \u003cb\u003eB and T lymphocytes localize in the TLSs and potentially interact through VISTA-PSGL-1 axis.\u003c/b\u003e \u003c/p\u003e \u003cp\u003eVISTA is the ligand of PSGL-1 in acidic pH such as that found in the TME [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]and it is known to be expressed also on monocytes and T cells [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e]. For this reason, to investigate the potential role of VISTA in mediating interactions between B cells and other immune cells within the tumor microenvironment (TME), we analyzed a publicly available dataset of imaging-based spatial transcriptomics in NSCLC [\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e]. Here, we were able to identify, beside tumor cells (expressing \u003cem\u003eKRT17)\u003c/em\u003e, macrophages (\u003cem\u003eC1QA)\u003c/em\u003e, T cells \u003cem\u003e(CD3E\u003c/em\u003e), B cells \u003cem\u003e(MS4A1\u003c/em\u003e), and smooth muscle cells \u003cem\u003e(TAGLN\u003c/em\u003e) (Supplementary Fig.\u0026nbsp;5A). Notably, overall B and T cells tend to cluster together, forming the ectopic structures commonly referred to as TLS (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA-B, Supplementary Fig.\u0026nbsp;5B). Analyzing the Euclidean distance between cells, we observed that B cells and CD4\u003csup\u003e+\u003c/sup\u003e T cells, with CD8\u003csup\u003e+\u003c/sup\u003e T cells to a lesser extent, were in close proximity compared to the other cell types (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC). Then, to evaluate possible cell-to-cell contact between B cells other subpopulation within the TME, we calculated the number of cell-to-cell contacts. We observed CD4\u003csup\u003e+\u003c/sup\u003e T cells exhibited a higher frequency of interactions with B cells and, to a lesser extent, with CD8\u003csup\u003e+\u003c/sup\u003e T cells and macrophages. This suggests that B cells and CD4\u003csup\u003e+\u003c/sup\u003e T cells tend to preferentially colocalize and likely interact within the TME (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC and Supplementary Fig.\u0026nbsp;5C). Thus, we further investigated what ligand\u0026ndash;target links regulate potential crosstalk between T cells and B cells in the TME using NicheNet. In this analysis, the B cell cluster derived from imaging-based spatial transcriptomics (except plasmacells) was designated as the sender population, while the conventional CD4\u003csup\u003e+\u003c/sup\u003e or CD8\u003csup\u003e+\u003c/sup\u003e T cell clusters were identified as separate receiver populations. For each T cell subset, we focused our analysis on the top 45 ligand\u0026ndash;receptor pairs. Among the well-known B and T ligand\u0026ndash;receptor axis such as: \u003cem\u003eCD40\u003c/em\u003e via \u003cem\u003eCD40LG\u003c/em\u003e, \u003cem\u003eCD86\u003c/em\u003e via \u003cem\u003eCD28\u003c/em\u003e, \u003cem\u003eTGFB1\u003c/em\u003e via \u003cem\u003eTGFBR2\u003c/em\u003e, NicheNet predicted a B-T interaction for CD4\u003csup\u003e+\u003c/sup\u003e or CD8\u003csup\u003e+\u003c/sup\u003e T cells via \u003cem\u003eVSIR\u003c/em\u003e-\u003cem\u003eSELPLG\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD, Supplementary Fig.\u0026nbsp;6) where \u003cem\u003eVSIR\u003c/em\u003e codes for V-domain Ig suppressor of T cell activation (VISTA), while \u003cem\u003eSELPLG\u003c/em\u003e codes for P-selectin glycoprotein ligand-1 (PSGL-1, also known as CD162). Moreover, we investigated if this predicted interaction could alter the gene-expression profile related to tumor-reactivity. We observed, both for CD4\u003csup\u003e+\u003c/sup\u003e and CD8\u003csup\u003e+\u003c/sup\u003e T cells, elevated potential regulatory scores among the 45 top-ranked ligands and following target genes: \u003cem\u003eFASLG, IFNG, DDIT4, TNFRSF18\u003c/em\u003e (GITR) and \u003cem\u003eCTLA4\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD, Supplementary Fig.\u0026nbsp;6).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo further validate these findings, we analyzed three different scRNA-seq datasets of patients with early-stage NSCLC (Supplementary Fig.\u0026nbsp;7A-B). For each T cell subset, we focused our analysis on the top 20 ligand-receptor pairs. Remarkably, we found consistent results across both imaging-based spatial transcriptomics and scRNA-seq datasets, both in terms of ligand-receptor interactions and potential gene regulatory scores (Supplementary Fig.\u0026nbsp;7C-D). Given that B and T cells specifically co-localize and potentially interact within the TLS regions, we assessed the expression of \u003cem\u003eVSIR\u003c/em\u003e and \u003cem\u003eSELPLG\u003c/em\u003e within five different TLS (boxed in the Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA). Our observations showed that \u003cem\u003eVSIR\u003c/em\u003e expression was slightly higher on B cells, whereas \u003cem\u003eSELPLG\u003c/em\u003e was more pronounced on both CD4\u003csup\u003e+\u003c/sup\u003e and CD8\u003csup\u003e+\u003c/sup\u003e T cells. Moreover, the percentage of B cells co-expressing both \u003cem\u003eVSIR\u003c/em\u003e and \u003cem\u003eIL10\u003c/em\u003e was higher than that of CD4\u003csup\u003e+\u003c/sup\u003e and CD8\u003csup\u003e+\u003c/sup\u003e T cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eE). These data offer evidence that B cells expressing VISTA and/or IL-10 are present within TLSs and may potentially interact with T cells through PSGL-1.\u003c/p\u003e \u003cp\u003e \u003cb\u003eTumor infiltrating CD4\u003c/b\u003e \u003csup\u003e+\u003c/sup\u003e \u003cb\u003eT\u003c/b\u003e\u003csub\u003e\u003cb\u003eRM\u003c/b\u003e\u003c/sub\u003e \u003cb\u003ecells expressing PSGL-1, but not PD-1 display an activated metabolic profile.\u003c/b\u003e\u003c/p\u003e \u003cp\u003eGiven that VISTA likely interacts with PSGL-1 expressed by T cells, we investigated by flow cytometry the phenotype of CD4\u003csup\u003e+\u003c/sup\u003e T cells expressing PSGL-1. We showed that the frequency of CD4\u003csup\u003e+\u003c/sup\u003e T cells among total CD45\u003csup\u003e+\u003c/sup\u003eCD3\u003csup\u003e+\u003c/sup\u003e live T cells was lower in the TME if compared to blood (median blood\u0026thinsp;=\u0026thinsp;67.6, tumor\u0026thinsp;=\u0026thinsp;48.0; Supplementary Fig.\u0026nbsp;8). We concatenated 5.946.764 of CD4\u003csup\u003e+\u003c/sup\u003e T cells (PBMC\u0026thinsp;=\u0026thinsp;5.108.956 and TILs\u0026thinsp;=\u0026thinsp;837.808) from 15 samples and analyzed them with FlowSOM. We identified 30 different clusters, whose frequencies are subsequently determined in each sample type. In particular, the dimensional reduction performed by Uniform Manifold Approximation and Projection (UMAP) showed that these clusters were distributed in six different regions (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA). Ten of these clusters belonging to \u0026lsquo;not exhausted\u0026rsquo; cells (C1, C2, C3, C5, C6, C8, C13, C16, C17 and C21) were more abundant in the blood and displayed a phenotypic identity coincident with na\u0026iuml;ve (CD45RA\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e; C3), central memory (CD45RA\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e\u0026minus;\u003c/sup\u003e; C1, C2), tissue-resident memory-like T cells (T\u003csub\u003eRM\u003c/sub\u003e-like, \u003cem\u003ei.e.\u003c/em\u003e, CD69\u003csup\u003e\u0026minus;\u003c/sup\u003eCD103\u003csup\u003e+\u003c/sup\u003e; C13), effector memory (CD45RA\u003csup\u003e\u0026minus;\u003c/sup\u003eCCR7\u003csup\u003e\u0026minus;\u003c/sup\u003e; C5, C6, C8, C16) or cytotoxic (CD45RA\u003csup\u003edim\u003c/sup\u003eCCR7\u003csup\u003e\u0026minus;\u003c/sup\u003eGZMB\u003csup\u003e+\u003c/sup\u003e; C17, C21) T cells. These clusters were characterized by lower levels of CTLA4, PD-1, CD96, and CD39, accompanied by higher expression of PSGL-1 (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA-B). Additional clusters of cells that were not in an exhausted state, such as C10 and C11, were characterized by the expression of CD69, and were notably more prevalent within the TME (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB). The percentages of proliferating CD39\u003csup\u003e+\u003c/sup\u003e CD4\u003csup\u003e+\u003c/sup\u003e T cells expressing PSGL-1, PD-1 and CTLA-4 (C26), were higher in the TME (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB). The remaining two clusters of likely proliferating CD4\u003csup\u003e+\u003c/sup\u003e T cells (C4 and C7) were similar in blood and TME (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB). The complete absence of \u0026lsquo;not exhausted\u0026rsquo; clusters within the tumor was counterbalanced by the presence of virtually unique subsets characterized by heterogeneous expression of PSGL-1 along with high levels of PD-1 and CTLA-4. These subsets included tissue-resident memory T cells (T\u003csub\u003eRM\u003c/sub\u003e, marked by CD69\u003csup\u003e+\u003c/sup\u003eCD103\u003csup\u003e+\u003c/sup\u003e; C30), conventional CD39\u003csup\u003e\u0026minus;\u003c/sup\u003eCD4\u003csup\u003e+\u003c/sup\u003e T cells (Tconv CD39\u003csup\u003e\u0026minus;\u003c/sup\u003e, CD39\u003csup\u003e\u0026minus;\u003c/sup\u003eCD69\u003csup\u003e+\u003c/sup\u003eCD103\u003csup\u003e\u0026minus;\u003c/sup\u003e; C12, C20, C24, C25, C15, C18, C9, C19, C14), and Tconv CD39\u003csup\u003e+\u003c/sup\u003eCD4\u003csup\u003e+\u003c/sup\u003e T cells (Tconv CD39\u003csup\u003e+\u003c/sup\u003e, CD39\u003csup\u003e+\u003c/sup\u003eCD69\u003csup\u003e+\u003c/sup\u003eCD103\u003csup\u003e\u0026minus;\u003c/sup\u003e; C22, C23, C29, C27, C28). To identify the major phenotypic markers distinguishing tumor-infiltrating CD4\u003csup\u003e+\u003c/sup\u003e T cells expressing PSGL-1 from those not expressing it, we manually evaluated the expression of thirteen immune markers, including T-bet, CD96, CD226, PD-1, CTLA-4, CD161, CD49a, CXCR6, Ki-67, GZMB, CCR7, CD45RA and CD39. We observed that CD4\u003csup\u003e+\u003c/sup\u003e T cells expressing PSGL-1 displayed higher levels of GZMB, CD226, CXCR6, T-bet, CD103, CD39 and CD96, while CCR7 and CD69 were expressed predominantly by PSGL-1 negative cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe metabolic investigation of tumor infiltrating CD4\u003csup\u003e+\u003c/sup\u003e T cells by the unsupervised analysis revealed 15 distinct scMEP states. In particular, 83.5% of CD4\u003csup\u003e+\u003c/sup\u003e T cells were grouped into metabolically quiescent states (scMEP 3, 2, 15, 10 and 13), characterized by a low level of all metabolic pathway regulators such as GLUT1, HK2, LDHA, CytC, IDH1, CD36 and HIF-1α (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eD-F). The remaining 16.5% of cells clustered into hyperactivated (scMEP 9, 5, 12, 4 and 8) to low-intermediate states (scMEP 1, 11, 14, 6 and 7) \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE\u003cb\u003e)\u003c/b\u003e. Hyperactivated scMEP 9, 5 and 12 states were characterized by high level of almost all metabolic markers, except for GAPDH, meaning that glycolysis, pentose, TCA, amino acid metabolism as well as FAO were activated \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE\u003cb\u003e)\u003c/b\u003e. ScMEP 4 and 8 displayed intermediate levels of TCA (such as IDH1 and ATP5A) and high level of FAO (such as CD36) paired with the complete absence of glycolytic regulator (such as GLUT1 and LDHA) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE-F). ScMEP 1, 11, and 14 displayed intermediate metabolisms without the expression of CD36, except for scMEP 1, which represented 0.27% of CD4\u003csup\u003e+\u003c/sup\u003e T cells showing upregulation of GLUT1 and HK2 \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE-F\u003cb\u003e)\u003c/b\u003e. ScMEP 6 and 7 collectively were 11.9% of cells, both exhibiting minimal expression of metabolic regulators, except for CD98 and pPGC1α, likely suggesting that these cells primarily depend on basal levels of TCA and the FAO pathway (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE\u003cb\u003e)\u003c/b\u003e. As far as immunologic characteristics were paired to metabolic states, scMEP clusters clearly identified distinct immunological phenotypes. All metabolic quiescent cells (scMEP 3, 2, 15, 10 and 13) displayed features of resting cells (CD38\u003csup\u003e\u0026minus;\u003c/sup\u003e, Ki-67\u003csup\u003e\u0026minus;\u003c/sup\u003e, CD39\u003csup\u003e\u0026minus;\u003c/sup\u003e and PD-1\u003csup\u003elow\u003c/sup\u003e) with Tconv phenotype (CD103\u003csup\u003e\u0026minus;\u003c/sup\u003e) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eG). On the other hand, all metabolically active scMEPs, except for scMEP 14, 6 and 7, belong to T\u003csub\u003eRM\u003c/sub\u003e phenotype (CD103\u003csup\u003e+\u003c/sup\u003e) and expressed PSGL-1 but not PD-1(Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eG-H). In particular, the metabolically hyperactivated state consisted of double positive CD4\u003csup\u003e+\u003c/sup\u003eCD8\u003csup\u003e+\u003c/sup\u003e T cells (DP; scMEP 9 and 5) or Tregs expressing IL-10 (scMEP 12) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eG-H). These findings align with our observation that nearly all DP cells within the tumor microenvironment were in an activated state (CD69\u003csup\u003e+\u003c/sup\u003eCD103\u003csup\u003e\u0026minus;\u003c/sup\u003e) and undergoing proliferation (Ki-67\u003csup\u003e+\u003c/sup\u003e), although they exhibited lower cytotoxicity compared to their counterparts in the bloodstream (Supplementary Fig.\u0026nbsp;9). Cells exhibiting low to intermediate metabolic activity were composed by activated double-negative T cells (DN) and Treg cells expressing high levels of PD-1, CD39, and CXCR6 \u0026mdash; a marker associated with the survival and localized expansion of effector T cells within TME [\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e]. These results are in line with the clustering performed by using both lineage markers as well as metabolic ones (Supplementary Fig.\u0026nbsp;10).\u003c/p\u003e \u003cp\u003e \u003cb\u003ePSGL-1-positive tumor infiltrating CD8\u003c/b\u003e \u003csup\u003e+\u003c/sup\u003e \u003cb\u003eT cells shows metabolic activation and enhanced cytotoxic profile.\u003c/b\u003e\u003c/p\u003e \u003cp\u003eAs we did for CD4\u003csup\u003e+\u003c/sup\u003e T cells expressing PSGL-1, we assessed the heterogeneity of tumor infiltrating CD8\u003csup\u003e+\u003c/sup\u003e T cells. The percentage of CD8\u003csup\u003e+\u003c/sup\u003e T cells was higher in TME if compared to blood (median Blood\u0026thinsp;=\u0026thinsp;23.0, median Tumor\u0026thinsp;=\u0026thinsp;31.4, Supplementary Fig.\u0026nbsp;11). To further explore all the CD8\u003csup\u003e+\u003c/sup\u003e T cells subset, we took advantage of unsupervised data analysis using FlowSOM. We investigated a total of 2.414.533 cells of which 613.023 cells from resected tumor (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;15, TILs) paired with 1.801.510 cells isolated from peripheral blood (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;15, PBMC). We identified 30 different CD8\u003csup\u003e+\u003c/sup\u003e clusters, whose frequencies were subsequently determined in each sample type. In particular, the dimensional reduction performed using UMAP algorithm showed that these clusters were distributed in three different areas, called as \u0026lsquo;not exhausted\u0026rsquo;, \u0026lsquo;Tconv and \u0026lsquo;T\u003csub\u003eRM\u003c/sub\u003e CD39\u003csup\u003e+\u003c/sup\u003e and CD39\u003csup\u003e\u0026minus;\u003c/sup\u003e\u0026rsquo; (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA). Among the nine \u0026lsquo;not exhausted\u0026rsquo; clusters, seven clusters (C1, C2, C4, C5, C6, C7, C18) were mostly represented within the blood and displayed a phenotypic identity coincident with na\u0026iuml;ve (CD45RA\u003csup\u003e+\u003c/sup\u003e CCR7\u003csup\u003e+\u003c/sup\u003e; C18), central memory (CD45RA\u003csup\u003e\u0026minus;\u003c/sup\u003e CCR7\u003csup\u003e+\u003c/sup\u003e; C5), effector memory (CD45RA\u003csup\u003e\u0026minus;\u003c/sup\u003e CCR7\u003csup\u003e\u0026minus;\u003c/sup\u003e; C1, C4, C6, C7) or effector memory re-expressing CD45RA (CD45RA\u003csup\u003e+\u003c/sup\u003e CCR7\u003csup\u003e\u0026minus;\u003c/sup\u003e; C2) T cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB). Similar to not exhausted CD4\u003csup\u003e+\u003c/sup\u003e T cells, not exhausted CD8\u003csup\u003e+\u003c/sup\u003e T cells exhibited reduced levels of CTLA4, PD-1, CD96, and CD39, along with heightened expression of PSGL-1. Other clusters belonging to Tconv (C8, C10, C11, C14, C15, C16, C17, C19, C21, C23, C27, C28), T\u003csub\u003eRM\u003c/sub\u003e CD39\u003csup\u003e+\u003c/sup\u003e (C9, C12, C20, C22, C30) or T\u003csub\u003eRM\u003c/sub\u003e CD39\u003csup\u003e\u0026minus;\u003c/sup\u003e (C25, C26, C29) cells, were enriched in TME and displayed varying levels of PSGL-1 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB). Similar percentages of C3, C13 and C24 were observed in TME and blood (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB). As we did for tumor infiltrating CD4\u003csup\u003e+\u003c/sup\u003e T cells, to identify the major phenotypic marker distinguishing tumor-infiltrating CD8\u003csup\u003e+\u003c/sup\u003e T cells expressing PSGL-1 from those not expressing it, we manually evaluated the expression of thirteen markers, including T-bet, CD96, CD226, PD-1, CTLA-4, CD161, CD49a, CXCR6, Ki-67, GZMB, CCR7, CD45RA and CD39. We reported that CD8\u003csup\u003e+\u003c/sup\u003e T cells expressing PSGL-1 displayed higher levels of GZMB, CD226, CXCR6, T-bet and CD96 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC). These data, together with those of tumor infiltrating CD4\u003csup\u003e+\u003c/sup\u003e T cells, establish GZMB, CD226, CXCR6, T-bet and CD96 as core surface markers that distinguish human PSGL-1\u003csup\u003e+\u003c/sup\u003e and PSGL-1\u003csup\u003e\u0026minus;\u003c/sup\u003e subsets within TME.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eInvestigating the metabolic profiles of CD8\u003csup\u003e+\u003c/sup\u003e T cells infiltrating tumors, we identified a total of 13 distinct scMEP states. Among these metabolic states, CD8\u003csup\u003e+\u003c/sup\u003e T cells exhibited subsets characterized by varying degrees of metabolic protein expression. scMEP 1, 2, 3, and 5 displayed low level of expression metabolic proteins, suggesting reduced metabolic activity while scMEP 13, 12, and 9 displayed elevated expression of a wide array of metabolic regulators, including HK2, GAPDH, LDHA, G6PD, CytC, IDH1, CD98, CD36, and pPGC1α, indicating increased metabolic activity (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eD-F). scMEP 8 and 10 were characterized by high expression of CD98, scMEP 11 by high level of GLUT1, scMEP 4 was distinguished by high level of expression of G6PD, scMEP 7 by GAPDH and scMEP 6 by high expression of GAPDH/PFKB4 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eD-F). These specific patterns of expression could suggest distinct metabolic adaptations influenced by the TME. As far as phenotypic landscape has been considered, metabolically quiescent states (such as scMEP 1, 2, 3, and 5) were Tconv resting cells (CD103\u003csup\u003e\u0026minus;\u003c/sup\u003e, CD69\u003csup\u003e\u0026minus;\u003c/sup\u003e, ICOS\u003csup\u003e\u0026minus;\u003c/sup\u003e, CD38\u003csup\u003e\u0026minus;\u003c/sup\u003e, Ki-67\u003csup\u003e\u0026minus;\u003c/sup\u003e, and PD-1\u003csup\u003elow\u003c/sup\u003e) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eG-H). scMEP 13, 12, and 9, which demonstrated heightened overall expression of metabolic proteins, displayed signs of activation, including CD69\u003csup\u003e+\u003c/sup\u003e, ICOS\u003csup\u003e+\u003c/sup\u003e, CD38\u003csup\u003e+\u003c/sup\u003e, Ki-67\u003csup\u003e+\u003c/sup\u003e expression, and a T\u003csub\u003eRM\u003c/sub\u003e phenotype, underscoring the augmented metabolic demands of actively cycling cells. Furthermore, these latter scMEP states exhibited elevated levels of PSGL-1, CD39, CD103 and PD-1 (as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eG-H). Additionally, we observed the presence of cells with distinctive metabolic characteristics (scMEP 8, 10, 6, 7, 11, and 4). All these cells exhibited an intermediate PD-1 level alongside reduced mitochondrial capacity, decreased PSGL-1 expression, and diminished pPGC1α, which are hallmark features of T cell exhaustion (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eG-H). These results are in line with the clustering performed by using both lineage markers as well as metabolic ones (Supplementary Fig.\u0026nbsp;12).\u003c/p\u003e \u003cp\u003e \u003cb\u003ePSGL-1\u003c/b\u003e \u003csup\u003e \u003cb\u003e+\u003c/b\u003e \u003c/sup\u003e \u003cb\u003eT cells display heightened antitumor capacity.\u003c/b\u003e\u003c/p\u003e \u003cp\u003eDue to its proposed role in regulating T cell trafficking within lymphoid organs and immune responses, we investigated the effector capabilities of tumor-infiltrating CD4\u003csup\u003e+\u003c/sup\u003e and CD8\u003csup\u003e+\u003c/sup\u003e T cells expressing PSGL-1. Our findings demonstrated that both conventional CD4\u003csup\u003e+\u003c/sup\u003e (CD103\u003csup\u003e\u0026minus;\u003c/sup\u003eFoxP3\u003csup\u003e\u0026minus;\u003c/sup\u003e; Tconv) and tissue-resident memory (T\u003csub\u003eRM\u003c/sub\u003e; CD103\u003csup\u003e+\u003c/sup\u003eFoxP3\u003csup\u003e\u0026minus;\u003c/sup\u003e) T cells expressing PSGL-1\u003csup\u003e+\u003c/sup\u003e exhibited higher cytotoxicity compared to their PSGL-1\u003csup\u003e\u0026minus;\u003c/sup\u003e counterparts (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eA-C). Concerning cytokine production, we noted that PSGL-1\u003csup\u003e+\u003c/sup\u003e T\u003csub\u003eRM\u003c/sub\u003e cells were more frequently producing TNF, while IL-2 and IFN-γ were equally produced (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eD). There were no observed differences in IL-17 production. The polyfunctionality of both PSGL-1\u003csup\u003e+\u003c/sup\u003e and PSGL-1\u003csup\u003e\u0026minus;\u003c/sup\u003e cells was comparable, although T\u003csub\u003eRM\u003c/sub\u003e cells exhibited greater polyfunctionality compared to Tconv cells, as illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eE. We next focused on CD8\u003csup\u003e+\u003c/sup\u003e T cells, observing also in these case that T cells expressing PSGL-1\u003csup\u003e+\u003c/sup\u003e exhibited higher cytotoxicity compared to their PSGL-1\u003csup\u003e\u0026minus;\u003c/sup\u003e counterparts (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eF-H). Furthermore, PSGL-1\u003csup\u003e+\u003c/sup\u003e Tconv seemed be more cytotoxic compared to PSGL-1\u003csup\u003e+\u003c/sup\u003e T\u003csub\u003eRM\u003c/sub\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eG-H). In terms of cytokine production PSGL-1\u003csup\u003e+\u003c/sup\u003e cells looked be able to produce less IL-2 among the two subsets, while IFN-γ were higher in PSGL-1\u003csup\u003e+\u003c/sup\u003e cells. TNF production remained consistent between PSGL-1\u003csup\u003e\u0026minus;\u003c/sup\u003e and PSGL-1\u003csup\u003e+\u003c/sup\u003e cells, while no differences were observed for IL-17 production (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eI). Analyzing the polyfunctionality, PSGL-1\u003csup\u003e+\u003c/sup\u003e appeared be more polyfunctional compared to their PSGL-1\u003csup\u003e\u0026minus;\u003c/sup\u003e counterparts, because enriched of cells able to produce simultaneously IL-2 and IFN-γ (T\u003csub\u003eRM\u003c/sub\u003e) or IFN-γ and TNF (Tconv) (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eJ).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eSupervised learning framework identified clusters related to cancer relapse after surgery in early-stage NSCLC patients.\u003c/b\u003e \u003c/p\u003e \u003cp\u003eTo define which cell populations positively or negatively drive NSCLC recurrence, we performed a prediction analysis using PENCIL, a novel supervised learning framework [\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]. Five out of fifteen patients examined with flow cytometry experienced tumor recurrence within 12 months post-surgery. By targeting patients' clinical recurrence, PENCIL identified cells associated with tumor recurrence with an accuracy of 82% for CD4\u003csup\u003e+\u003c/sup\u003e T cells, 79% for CD8\u003csup\u003e+\u003c/sup\u003e T cells, and 80% for B cells (Supplementary Fig.\u0026nbsp;13A-C). Our analysis indicated that patients who experienced lung recurrence after surgery exhibited high percentages of circulating cytotoxic CD4\u003csup\u003e+\u003c/sup\u003e cells expressing PSGL-1 (C17), CD4\u003csup\u003e+\u003c/sup\u003e Tconv expressing PSGL-1 (C15), CD8\u003csup\u003e+\u003c/sup\u003e T\u003csub\u003eRM\u003c/sub\u003e expressing PSGL-1 (C26), and VISTA\u003csup\u003e+\u003c/sup\u003e Bregs (C20) (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eA-F). In contrast, patients who remained recurrence-free for 12 months post-surgery displayed high percentage of PSGL-1\u003csup\u003e\u0026minus;\u003c/sup\u003e CD4\u003csup\u003e+\u003c/sup\u003e T\u003csub\u003eRM\u003c/sub\u003e (C30) and PSGL-1\u003csup\u003e\u0026minus;\u003c/sup\u003e CD8\u003csup\u003e+\u003c/sup\u003e T\u003csub\u003eRM\u003c/sub\u003e (C12, C22, C25) (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eC-D). Additionally, memory-unswitched B cells (C1) were elevated in both blood and tumor samples of patients without recurrence, whereas memory-switched B cells (C14) showed an increase exclusively in the tumor site of patients without recurrence (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eE-F). To validate PENCIL prediction, we employed the multivariate Cox proportional-hazards model. In this analysis, high percentage of VISTA\u003csup\u003e+\u003c/sup\u003e Bregs (C20, [HR\u0026thinsp;=\u0026thinsp;1.41; 95% CI\u0026thinsp;=\u0026thinsp;0.98, 2.03]) within the TME, along with high percentage of CD8\u003csup\u003e+\u003c/sup\u003e T\u003csub\u003eRM\u003c/sub\u003e expressing PSGL-1 (C26, [HR\u0026thinsp;=\u0026thinsp;1.16; 95% CI\u0026thinsp;=\u0026thinsp;0.84, 1.45]), circulating cytotoxic CD4\u003csup\u003e+\u003c/sup\u003e cells expressing PSGL-1 (C17, [HR\u0026thinsp;=\u0026thinsp;1.09; 95% CI\u0026thinsp;=\u0026thinsp;0.96, 1.23]), and reduced levels of PSGL-1\u003csup\u003e\u0026minus;\u003c/sup\u003e CD4\u003csup\u003e+\u003c/sup\u003e T\u003csub\u003eRM\u003c/sub\u003e (C30, [HR\u0026thinsp;=\u0026thinsp;0.64; 95% CI\u0026thinsp;=\u0026thinsp;0.40, 1.01]), were predictive of an elevated risk of tumor recurrence within 12 months post-surgery (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eG).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, we unraveled B and T cell landscape of the NSCLC tumor microenvironment, providing insights into their interactions, metabolic activities, and their influence on cancer recurrence post-surgery. Indeed, we unveiled for the first time not only the presence of B cells expressing VISTA within the TME of NSCLC, but also that VISTA\u003csup\u003e+\u003c/sup\u003e B cells belong to the category of Bregs. Furthermore, trajectory inference and pseudotime score revealed that VISTA\u003csup\u003e+\u003c/sup\u003e Bregs represent a unique cell fate state, branching out separately from plasma cells. Furthermore, we found that, from a bioenergetics point of view, VISTA\u003csup\u003e+\u003c/sup\u003e Bregs were characterized by high metabolic activities, including glycolysis, TCA, FAO, amino acid metabolism, and PPP, suggesting that this unique metabolic signature points toward an enhanced energy production and utilization within these cells regulated by transcription factor HIF-1α. Indeed, metabolic programs play a crucial role in the entire process of B cell development, as metabolic nutrients are essential not only for the acquisition of immune cell function, but also for the transition from effector immune cells to regulatory immune cells [\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e]. From a functional point of view, cytokines such as IL-10, IL-6, GM-CSF, TNF and TGF-β may drive T cell immune suppression and/or tumor progression [\u003cspan additionalcitationids=\"CR49 CR50 CR51\" citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e], and here we show that VISTA\u003csup\u003e+\u003c/sup\u003e B cells produce various cytokines, including IL-10, TGF-β, IL-6, TNF, and GM-CSF, pointing out their immunosuppressive capacity and pro-tumor activity [\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eWe unveiled that B and T cells are prevalently distributed in the TLS, but are rarely found in direct contact with tumor cells. Moreover, B cells exhibited a higher number of cell-to-cell interactions with CD4\u003csup\u003e+\u003c/sup\u003e T cells and to a lesser extent with CD8\u003csup\u003e+\u003c/sup\u003e T cells and macrophages, suggesting a potential role of B cells in influencing preferentially the CD4\u003csup\u003e+\u003c/sup\u003e T cells anti-tumor immune response. Spatial colocalization of B and T cells in tumor TLS revealed a novel interaction axis between B and T cells within the TME involving VISTA and PSGL-1, alongside established interactions like CD40-CD40L and ICOS-ICOSL. This finding is consistent with recent research demonstrating the selective engagement of VISTA by PSGL-1, particularly under the acidic pH conditions commonly encountered in the TME [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. These results suggest that, in conjunction with Tregs, VISTA\u003csup\u003e+\u003c/sup\u003e Bregs may play a role in suppressing antitumor T cell responses. This suppression may occur not only through release of immunosuppressive cytokines, but also via cell contact-dependent immune suppression mediated by the VISTA-PSGL-1 axis.\u003c/p\u003e \u003cp\u003eBeyond adhesion, PSGL-1 has been shown to be implicated in T cell signaling pathways, influencing cytokine production and immune suppression within the TME [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. Furthermore, recent data support that the therapeutic blockade of PSGL-1 may promotes T cell responses and melanoma tumor control [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. Here we observed that 30% of tumor-infiltrating T cells, spanning from Tconv to T\u003csub\u003eRM\u003c/sub\u003e cells, were positive for PSGL-1 expression. Notably, these PSGL-1-positive cells exhibited a distinct phenotype compared to PSGL-1-negative counterparts. Specifically, both CD4 and CD8 T cell compartments expressing PSGL-1 shared the expression of a core set of markers, including GZMB, CD226, CXCR6, T-bet, and CD96. The expression of these markers has been associated with increased T cell activation, cytotoxicity, and lung-tissue residency [\u003cspan additionalcitationids=\"CR55 CR56 CR57\" citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e]. Our data also revealed the heightened antitumor potential of PSGL-1\u003csup\u003e+\u003c/sup\u003e CD4 and CD8 T cells, both in Tconv and T\u003csub\u003eRM\u003c/sub\u003e cells. These PSGL-1-positive cells exhibited enhanced production of cytotoxic molecules such as GNLY and GZMB. Additionally, they displayed distinct cytokine production patterns, with PSGL-1-positive CD4\u003csup\u003e+\u003c/sup\u003e T\u003csub\u003eRM\u003c/sub\u003e cells secreting higher levels of IL-2 and TNF, while PSGL-1-positive CD8\u003csup\u003e+\u003c/sup\u003e T\u003csub\u003eRM\u003c/sub\u003e cells and Tconv cells produced increased IFN-γ but lower IL-2. These results were in line with their metabolic profile. Indeed, PSGL-1\u003csup\u003e+\u003c/sup\u003e tumor-infiltrating CD4\u003csup\u003e+\u003c/sup\u003e and CD8\u003csup\u003e+\u003c/sup\u003e cells exhibited hyperactivated metabolic profiles and were enriched in T\u003csub\u003eRM\u003c/sub\u003e cells. These cells showcased elevated metabolic pathways, including glycolysis, TCA, FAO, amino acid metabolism, and PPP. This heightened metabolic state suggests an increased energy demand, driven by their enhanced effector capabilities in the tumor microenvironment, as we observed.\u003c/p\u003e \u003cp\u003eWhile adjuvant chemotherapy has proven beneficial for improving overall survival in NSCLC patients with stage IIA to IIIA disease, those with stage I tumors smaller than 4 cm do not typically receive adjuvant chemotherapy [\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e]. Even after complete lung tumor resection, a considerable proportion of these patients face the risk of relapse [\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e]. Our computational analyses, employing PENCIL supervised learning and Cox multivariate analysis, predicted specific immune cell clusters associated with cancer recurrence post-surgery. VISTA\u003csup\u003e+\u003c/sup\u003e tumor infiltrating Bregs (C20), circulating cytotoxic CD4\u003csup\u003e+\u003c/sup\u003e cells expressing PSGL-1 (C17), and CD8\u003csup\u003e+\u003c/sup\u003e T\u003csub\u003eRM\u003c/sub\u003e cells expressing PSGL-1 (C26) emerged as key clusters linked to disease recurrence. The presence of these cells is strongly associated with an increased risk of tumor relapse, highlighting their potential as prognostic markers.\u003c/p\u003e \u003cp\u003eCollectively, these data suggest that while PSGL-1-expressing T cells exhibit highly effective antitumor capabilities, their proximity and cell-to-cell interaction with VISTA\u003csup\u003e+\u003c/sup\u003e Bregs, particularly when present at high frequencies, may lead to their suppression within the TME. Despite that, the mechanism of B and T cell interaction may be more heterogeneous. Indeed, it was recently reported that up to 60% of human Bregs express PSGL-1. Notably, PSGL-1 downregulation was observed in patients with systemic sclerosis and was correlated with a reduced extent of IL-10 production by Bregs [\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e]. This implies that the binding of PSGL-1 on VISTA\u003csup\u003e+\u003c/sup\u003eB cells by T cells expressing VISTA could act as a mechanism to reduce IL-10 production, serving as a brake on Bregs-mediated immune suppression.\u003c/p\u003e \u003cp\u003eIn summary, high density of VISTA\u003csup\u003e+\u003c/sup\u003e Bregs in the TLS could be associated with the tumor recurrence in NSCLC patients. Therefore, markers such as VISTA or PSGL-1 may represent a new potential therapeutic target for those patients who develop tumor recurrence or who do not respond well to anti-PD-1 or anti-CTLA-4 therapy.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eStudy design\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAs part of the project, we recruited a cohort comprising 37 patients diagnosed with NSCLC at stages IA, IB, IIA, IIB, and IIIA. These patients were admitted to either Azienda Ospedaliero Universitaria di Modena and Reggio Emilia (University Hospital) in Modena or G.B. Morgagni\u0026mdash;L. Pierantoni Hospital in Forl\u0026igrave;. Detailed clinicopathologic characteristics of the patients can be found in \u003cstrong\u003eTable 1\u003c/strong\u003e. The average age of the entire cohort was 71.4\u0026thinsp;\u0026plusmn;\u0026thinsp;7.2 years. Among these patients, five had a prior history of cancer, as outlined in \u003cstrong\u003eTable 1\u003c/strong\u003e. Additionally, seven patients\u0026nbsp;experienced a relapse post-surgery. Importantly, none of the patients had undergone chemotherapy, radiation therapy, or any other anti-tumor treatments before tumor resection. The study protocol was approved by the local Ethical Committes (\u0026ldquo;Area Vasta Emilia Nord\u0026rdquo;, protocol number 0018007/21; \u0026ldquo;Comitato Etico della Romagna\u0026rdquo;, protocol number 7043/2021). All participants involved in this study provided written informed consent for both sample collection and data analyses. Detailed clinical information and the analysis methods employed for each sample are reported in \u003cstrong\u003eSupplementary Table 1\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eBlood collection and isolation of mononuclear cells\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eUp to 30 mL of blood were collected from each patient in vacuettes containing ethylenediamine-tetraacetic acid (EDTA). Blood was immediately processed. Isolation of peripheral blood mononuclear cells (PBMC) was performed using ficoll-hypaque (Sentinel Diagnostics, Lymphoprep) according to standard procedures\u0026nbsp;[62]. PBMC were stored in liquid nitrogen in fetal bovine serum (FBS) supplemented with 10% dimethyl sulfoxide (DMSO). Plasma was stored at \u0026minus;80 \u0026deg;C until use.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIsolation of leucocytes from solid human tissues\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAfter informed consent, tumor tissue from patients undergoing surgery was collected. Primary surgical samples were transported from the operating room to the laboratory at controlled temperature of 4\u0026deg; C. Biopsy was washed\u0026nbsp;twice with 30 ml of 1\u0026thinsp;\u0026times; Dulbecco\u0026prime;s Phosphate Buffered Saline (D-PBS; Gibco) in a 50 ml Falcon tube. Then the tumor tissue was collected with sterile clamps in a Petri dish for cell culture. During the processing of the biopsy, residual parenchymal tissue was trimmed away using scalpels and only tumor tissue was kept and gently minced in small pieces. Small fragments obtained from the tissue were collected in gentleMACS Tube (Miltenyi Biotec) containing 2 ml of DMEM F12 medium (Gibco) supplemented with 10% FBS, 1% penicillin-streptomycin solution, 2% glutamine. Additional 2.5 ml of medium were used to wash the Petri dish and collect residual fragments, then added to the gentleMACS Tube. After mechanical disaggregation, tissue sample (in a range of 0.01 g to 1 g) was enzymatically digested to single cells by using Tumor Dissociation Kit (Miltenyi Biotec) as follows: \u0026nbsp;sample was incubated for 1 h with the enzyme mix containing 200 \u0026mu;l of enzyme H, 100 \u0026mu;l of enzyme R, 25 \u0026mu;l of enzyme A, in 4.5 ml of complete medium at 37\u0026deg;C into the Gentle MACS Octo dissociator (Miltenyi Biotec). After the incubation, 5.5 ml of medium were added to the 4.5 ml suspension of the processed biopsy, then filtered using a 70 \u0026mu;m cell strainer and transferred into a new 50 ml Falcon tube. The used filter was washed with 2.5 ml of complete medium to collect residual cell suspension. Collected sample was centrifugated at 300g for 5 minutes, then supernatant was discharged, and pellet was resuspended with 5 ml of complete medium. The obtained suspension was gently laid on 10 ml of ficoll-hypaque (Sentinel Diagnostics, Lymphoprep) in a 15 ml Falcon tube and centrifuged at 400g for 20 minutes at 21 \u0026deg;C brake off and minimal acceleration. After the centrifugation, the ring containing TILs was carefully collected and washed\u0026nbsp;twice with 25 ml PBS at 930g for 5 minutes at room temperature. Supernatant was removed and pellet was resuspended with 5 ml of complete medium and cells were counted using TC20 Automated Cell Counter (Bio-Rad). If not used, TILs were immediately frozen.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMulticolor flow cytometry detection of surface and intracellular antigens of T lymphocytes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFor \u003cem\u003eex vivo\u003c/em\u003e immunophenotyping experiments, frozen blood and tumor samples were thawed in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS), 1% L-glutamine, 1% sodium pyruvate, 1% nonessential amino acids, antibiotics, 0.1 M 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), and 55 mM \u0026beta;-mercaptoethanol, referred to as R10, and also supplemented with 20 \u0026mu;g/ml DNase I from bovine pancreas (Sigma-Aldrich)\u0026nbsp;[63]. Cells were washed with PBS and immediately stained using viability marker PromoFluor IR-840 (Promokine PromoCell, Heidelberg, Germany) for 20 min at room temperature (RT) in PBS. Then, the cells were washed\u0026nbsp;with FACS buffer (PBS added with 2% FBS)\u0026nbsp;and stained for chemokine receptors (CXCR6 and CCR7) for 20 min at 37\u0026deg;C. After washing with FACS buffer, cells were stained with combination of surface mAbs (anti-CD45, CD3, CD4, CD8, CD45RA, CD39, CD49a, CD69, CD96, CD103, CD161, CD162, CD226, PD-1 and CTLA4) for 20 min at RT (in BV buffer; 50% FACS and 50% Brilliant Stain Buffer from BD). Intracellular detection of Ki-67, T-bet and granzyme B (GZMB) was performed following fixation and permeabilization of cells with the FoxP3/transcription factor staining buffer set (eBioscience) according to manufacturer\u0026rsquo;s instructions incubating mAbs for 30 min at 4\u0026deg;C.\u003c/p\u003e\n\u003cp\u003eThe data were collected using a Cytoflex LX flow cytometer (Beckman Coulter, Brea, CA, USA), which featured five lasers (UV, 350 nm; violet, 405 nm; blue, 488 nm; yellow/green, 561 nm; red, 640 nm), all tuned at 50 mW except for UV, which was tuned at 20 mW, and had the capability to detect 21 parameters. Flow cytometry data were compensated in FlowJo using single stained controls (BD Compensation beads incubated with fluorescently conjugated antibodies). Additionally, all monoclonal antibodies were previously titrated to determine the optimal concentration. The comprehensive list and antibody titers incorporated in the flow cytometry panel can be found in \u003cstrong\u003eSupplementary Table 2\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMulticolor flow cytometry detection of surface and intracellular antigens of B lymphocytes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFor the \u003cem\u003eex vivo\u003c/em\u003e immunophenotyping experiments, frozen blood and tumor samples were thawed in R10 supplemented with 20 \u0026mu;g/ml DNase I from bovine pancreas (Sigma-Aldrich). Cells were first washed with PBS and stained with the viability marker PromoFluor IR-840 (Promokine PromoCell, Heidelberg, Germany) for 20 minutes at RT. Subsequently, the cells were washed with FACS buffer and incubated with the chemokine receptor CXCR5 for 20 minutes at 37\u0026deg;C. After another wash with FACS buffer, cells were stained with DuraClone IM B (Beckman Coulter, Brea, CA, USA) containing the following lyophilized directly conjugated monoclonal antibodies: anti-IgD-FITC, CD21-PE, CD19-ECD, CD27-PC7, CD24-APC, CD38-AF750, anti-IgM-PB, and CD45-KrO. Additional drop-in antibodies including VISTA, CD49b, CD73, and CD138 were added in BV buffer. Intracellular detection of Ki-67 and interleukin (IL)-10 was performed after fixing and permeabilizing the cells using the FoxP3/transcription factor staining buffer set (eBioscience) according to the manufacturer\u0026rsquo;s instructions, and the monoclonal antibodies were incubated for 30 minutes at 4\u0026deg;C. To distinguishing true VISTA and IL-10 positive signals from background noise we used fluorescence minus one (FMO) (Supplementary Fig. 14). The comprehensive list and antibody titers incorporated in the flow cytometry panel can be found in \u003cstrong\u003eSupplementary Table 2\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eHigh dimensional data analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eT cells analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCompensated Flow Cytometry Standard (FCS) 3.0 files were imported into FlowJo software version\u0026nbsp;v10.7.1\u0026nbsp;and pre-analyzed by standard gating to remove doublets, aggregates, dead cells and poor flow events. For each sample, we therefore selected data from all living CD4\u003csup\u003e+\u003c/sup\u003e or CD8\u003csup\u003e+\u003c/sup\u003e T cells and imported them in R using flowCore package v2.4.0\u0026nbsp;[64]\u0026nbsp;for a total of 5,946,764 CD4\u003csup\u003e+\u003c/sup\u003e T cells (of which\u0026nbsp;837,808 were TIL) and 2,414,533\u0026nbsp;CD8\u003csup\u003e+\u003c/sup\u003e T cells (of which 613,023 were TIL). The further analysis was performed using CATALYST v1.18.1\u0026nbsp;[65]. All data obtained by flow cytometry were transformed in R using hyperbolic arcsine \u0026ldquo;\u003cem\u003earcsinh\u0026nbsp;\u003c/em\u003e(x/cofactor)\u0026rdquo; applying manually defined cofactors where x is the fluorescence measured intensity value and cofactor as defined by Melsen and colleagues\u0026nbsp;[66]. Clustering and dimensional reduction were performed using FlowSOM (version 2.4.0) and UMAP (version 0.2.8.0) algorithms, respectively. Clustering gave origin to 30 clusters of CD4\u003csup\u003e+\u003c/sup\u003e T cells and 30 of CD8\u003csup\u003e+\u003c/sup\u003e T lymphocytes. Details regarding the quality control (QC) of the clustering process for CD4\u003csup\u003e+\u003c/sup\u003e and CD8\u003csup\u003e+\u003c/sup\u003e T cells were reported in the Supplementary Fig. 15-16 and Supplementary Fig. 17-18\u003cstrong\u003e,\u003c/strong\u003e respectively.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eB cells analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCompensated FCS 3.0 files were imported into FlowJo software version v10.7.1 and pre-analyzed by standard gating to remove doublets, aggregates, dead cells, and identify CD19\u003csup\u003e+\u003c/sup\u003e B cells. For each sample, we exported all living CD19\u003csup\u003e+\u003c/sup\u003e B cells and imported them in R using flowCore package v2.4.0 for a total of or 1,011,489\u0026nbsp;CD19\u003csup\u003e+\u003c/sup\u003e B cells (of which 472,627 were TIL). The unsupervised analysis was performed using CATALYST v1.18.1. All data were transformed in R using hyperbolic arcsine \u0026ldquo;\u003cem\u003earcsinh\u0026nbsp;\u003c/em\u003e(x/cofactor)\u0026rdquo; applying manually defined cofactors (where x is the fluorescence measured intensity value). Clustering and dimensional reduction were performed using FlowSOM and UMAP algorithms, respectively. Clustering gave origin to 20 clusters of CD19\u003csup\u003e+\u003c/sup\u003e B lymphocytes. Details regarding the quality control (QC) of the clustering process for B cells can be found in Supplementary Fig. 19-20.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePseudotime analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe applied Slingshot to CD19\u003csup\u003e+\u003c/sup\u003e B cells data\u0026nbsp;[67]. Slingshot is a tool used for inferring trajectories and pseudotime from single-cell data. Initially, we performed dimensional reduction using DiffusionMap. Subsequently, smooth branching curves were fitted to these lineages to obtain refined representations of each lineage, translating the overall knowledge of lineage structure into pseudotime at the single-cell level. The clusters obtained through FlowSOM (Total B cells = 20 clusters) were embedded within the DiffusionMap analysis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSingle-cell metabolic regulome profiling (scMEP)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eHeavy metal conjugation of antibodies\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAntibodies were conjugated to heavy metal ions with commercially available MaxPar (Standard BioTools) reagents following vendor conjugation protocol for MCP9 (for cadmium) or X8 (for lanthanide). In short, antibodies were reduced with 4\u0026thinsp;mM TCEP (Thermo Fisher) for 30\u0026thinsp;min at 37\u0026thinsp;\u0026deg;C and washed two times. For conjugations using MaxPar reagents, metal chelation was performed by adding metal solutions (final, 0.05\u0026thinsp;M) to chelating polymers and incubating for 60\u0026thinsp;min (for cadmium) or 40 min (for lanthanide) at 37\u0026deg;C. Metal-loaded polymers were washed three times using a 3-kDa MWCO microfilter (Millipore) by centrifuging for 25\u0026thinsp;min at 12,000g at RT. Partially reduced antibodies and metal-loaded polymers were incubated together for 90\u0026thinsp;min at 37\u0026thinsp;\u0026deg;C. Cadmium conjugated antibodies were washed four times for 10\u0026thinsp;min at 5,000g at RT, while lanthanide conjugated antibodies were washed four times for 10\u0026thinsp;min at 10,000g at RT. \u0026nbsp;All conjugated antibodies were collected by centrifugation (2\u0026thinsp;min, 1,000g, RT) into an inverted column in a fresh 1.6-ml collection tube and transferred into Protein LoBind tube (Eppendorf). Protein content was assessed by NanoDrop (Thermo Fisher) measurement. HRP protector buffer for cadmium conjugated antibodies (Boca Scientific) or antibody stabilizer PBS for lanthanide conjugated antibodies (Candor Bioscience) were added to reach the final antibodies concentration of 0.5mg/ml. All antibodies were stored at 4\u0026thinsp;\u0026deg;C.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLive/dead discrimination and cell staining\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCryopreserved single-cell suspensions from tumor biopsy specimens were thawed in R10 supplemented with 20 \u0026mu;g/ml DNase I from bovine pancreas (Sigma-Aldrich) and washed twice (5\u0026thinsp;min, 600g, RT). Before staining, tumor infiltrating leukocytes were enriched by using CD45 (TIL) MicroBeads (Miltenyi Biotec). For live/dead cell discrimination resuspended cells were stained with 1 \u0026micro;M of Cell-ID Cisplatin-195Pt (Standard BioTools) for 5 min at 37\u0026thinsp;\u0026deg;C and washed with Maxpar Cell Staining buffer (Standard BioTools). Dyed cells were stained with combination of surface mAbs (see \u003cstrong\u003eSupplementary Table 2\u003c/strong\u003e) for 30 min at RT in Maxpar Cell Staining buffer. Then, cells were washed with 2ml of Maxpar Cell Staining buffer and centrifuge for 5\u0026thinsp;min at 300g at RT. Intracellular markers detection was performed following fixation of cells with the FoxP3/transcription factor staining buffer set (eBioscience) according to manufacturer\u0026rsquo;s instructions incubating mAbs for 30 min at 4\u0026deg;C (see \u003cstrong\u003eSupplementary Table 2\u003c/strong\u003e). After washing with 2ml of Maxpar Cell Staining buffer, cells were placed on ice for 10 min to chill the sample and proceed for phosphoprotein staining. Each sample were resuspended using 1ml of 100% methanol and incubated for 15 min on ice. Then samples were washed twice with Maxpar Cell Staining buffer and incubated with anti-phosphoprotein antibodies mix for 30 in at RT (see \u003cstrong\u003eSupplementary Table 2\u003c/strong\u003e). Then all samples were washed twice with 2ml of Maxpar Cell Staining buffer and fixed adding 1ml of 1.6% Formaldehyde incubating for 10 min at RT. After incubation samples were centrifuged for 5 min at 800g at RT. Then all samples were resuspended in 1ml of Cell-ID intercalator-Ir (Standard BioTools) 125nM and stained overnight at 4\u0026thinsp;\u0026deg;C. Cells were then resuspended at 1 \u0026times; 10\u003csup\u003e6\u003c/sup\u003e cells per ml in ddH\u003csub\u003e2\u003c/sub\u003eO supplemented with 1\u0026times; EQ Four Element Calibration Beads (Standard BioTools) and acquired on a CyTOF2 mass cytometer (Standard BioTools).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMass cytometry data pre-processing and analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eRaw mass cytometry data were first bead normalized to remove acquisition-related influences on marker expression using CATALYST v1.18.1 pre-processing workflow. Normalized CyTOF files were exported as .fsc. Subsequently, all normalized .fcs files were uploaded into FlowJo software v10.7.1 and checked to exclude flow instabilities (using PeacoQC package), aggregates, doublets, dead cells, and non-biological events (Supplementary Fig. 21A). For each sample, we therefore selected data from living CD45\u003csup\u003e+\u003c/sup\u003e B and T cells and imported them in R using flowCore package v2.4.0. All data obtained by CyTOF were transformed using hyperbolic\u0026nbsp;\u003cem\u003earcsinh\u003c/em\u003e (cofactor 5). The first round of clustering and dimensional reduction was performed respectively using FlowSOM and UMAP algorithms utilizing both phenotypic and metabolic markers to clearly identify the main populations (Supplementary Fig. 21B-F). Subsequently, we selected B cells, CD4\u003csup\u003e+\u003c/sup\u003e T cells, and CD8\u003csup\u003e+\u003c/sup\u003e T cells, clustering them separately using exclusively metabolic markers or both as reported in Supplementary Fig. 3, 10, 12. TCF-1 and CXCL13 were excluded from the analysis due to weak signal caused by staining and conjugation issues.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCalculation of metabolic scores\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo calculate single-cell metabolic scores, expression values (bead normalized,\u0026nbsp;\u003cem\u003earcsinh\u003c/em\u003e transformed) from all metabolic enzymes within a given pathway (glycolysis, mitochondrial dynamics, TCA, pentose phosphate pathway, transcription, aminoacidic metabolism and fatty acid metabolism) were summed and divided by the number of channels within the pathway. The heatmap was drowned using ComplexHeatmap package v2.14.0\u0026nbsp;[68]. To calculate the \u0026lsquo;similarity score\u0026rsquo; to compare the metabolism on a set of clusters we used\u0026nbsp;MEM_RMSD()function from cytoMEM\u0026nbsp;package v1.2.0\u0026nbsp;[69]. Briefly,\u0026nbsp;MEM_RMSD\u0026nbsp;calculates a normalized average root-mean-square deviation (RMSD) score pairwise between populations given their metabolic scores as input. This is meant to serve as a metric of similarity between populations. The RMSD values are then converted to percentages with the maximum RMSD in the matrix set as 100 percent, so that the final RMSD score is the percent of the maximum RMSD [Percent_max_RMSD = 100-RMSD/max_RMSD*100].\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eImage-based spatial transcriptomic data\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSpatial molecular imaging data from FFPE non-small-cell lung cancer tissue samples were retrieved from\u0026nbsp;\u003ca href=\"http://nanostring.com/CosMx-dataset\"\u003ehttp://nanostring.com/CosMx-dataset\u003c/a\u003e . Transcriptomic data have been normalized using the R package Seurat applying SCTransform functions\u0026nbsp;[70]. All cells were annotated using precomputed Azimuth annotations from Seurat package. ImageDimPlot function was used to plot B and T cell molecules displaying spatial correlation with each other. BuildNicheAssay function from Seurat was used to construct a new assay called \u0026lsquo;niche\u0026rsquo; containing the cell type composition spatially neighboring each cell.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNicheNet analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTumor infiltrating CD45\u003csup\u003e+\u003c/sup\u003e T and B cells were retrieved from GSE131907, GSE148071, GSE154826 containing data of stage I, II or IIIA patients. All three datasets have been normalized and integrated using the R package Seurat applying SCTransform and IntegrateData functions\u0026nbsp;[70, 71]). Principal component was computed, and 20 PCs were selected to ran UMAP and graph-based clustering with a resolution between 0.1 and 1. Clustertree was used to choose the best resolution\u0026nbsp;[72]. The integrated Seurat object was annotated, and specifically, T cells and B cells were chosen. From these, CD4\u003csup\u003e+\u003c/sup\u003e, CD8\u003csup\u003e+\u003c/sup\u003e T cells, and CD19\u003csup\u003e+\u003c/sup\u003e B cells (excluding plasma cells, Pro-B cells and proliferating cells; Supplementary Fig. 7A-B) were employed for running NicheNet, following the instructions outlined in the provided vignette (https://github.com/saeyslab/nichenetr; doi:10.1038/s41592-019-0667-5). During multiple separate NicheNet runs, CD4\u003csup\u003e+\u003c/sup\u003e and CD8\u003csup\u003e+\u003c/sup\u003e T cell subsets were set as \u0026lsquo;receiver\u0026rsquo; and B cell clusters as \u0026lsquo;sender\u0026rsquo; populations. For the receiver and sender cell population were retrieved expressed genes, with the key parameters being set as follows: genes expressed in at least 10% of the cells of the respective group. On these sets of genes was run NicheNet to infer the ligand-receptor network followed by finding of \u0026lsquo;target genes\u0026rsquo; of top-ranked ligands (Supplementary Fig. 7C-D).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSpatial transcriptomic data from NSCLC tumor-infiltrating B and T cells were obtained from the image-based dataset mentioned earlier. CD4\u003csup\u003e+\u003c/sup\u003e, CD8\u003csup\u003e+\u003c/sup\u003e T cells, and CD19\u003csup\u003e+\u003c/sup\u003e B cells were selected for NicheNet analysis, following the previously described procedure. Our secondary objective was to explore whether B cells influence the expression of genes associated with tumor reactivity in CD4\u003csup\u003e+\u003c/sup\u003e and CD8\u003csup\u003e+\u003c/sup\u003e T cells. To achieve this, we utilized the 'tumor reactive gene signature' as defined by Olivera et al. and Guo et al.\u0026nbsp;[6, 73]. The predicted targets score was based on a Pearson correlation coefficient as described in the NicheNet vignette. NicheNet ligand-receptor matrix was integrated with new information on VISTA obtained from recent publications\u0026nbsp;[41, 74]\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIntracellular cytokine staining (ICS) of T and B cells\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThawed PBMCs and TILs were rested for 1h. Then T cells were stimulated with 50 ng/ml\u0026nbsp;of phorbol 12-myristate 13-acetate (PMA) and 1 \u0026mu;g/ml of ionomycin\u0026nbsp;for 4h at 37\u0026deg;C in a 5% CO\u003csub\u003e2\u003c/sub\u003e atmosphere in R10 culture medium. For each stimulated sample, an unstimulated one was prepared as a negative control. All samples were incubated with protein transport inhibitor containing Brefeldin A (Biolegend, San Diego, CA, USA) and monensin (Biolegend, San Diego, CA, USA). After stimulation, cells were washed with PBS and stained with viability marker PromoFluor IR-840 (Promokine, PromoCell, Heidelberg, Germany) for 20 min at RT. Next, cells were washed with FACS buffer and stained with surface mAbs (\u003cstrong\u003eSupplementary Table 2\u003c/strong\u003e). After incubation all samples were fixed and permeabilized with the FoxP3/transcription factor staining buffer set (eBioscience) according to manufacturer\u0026rsquo;s instructions incubating. Then, cells were stained with previously titrated mAbs recognizing intracellular cytokines and transcription factors for 30 min at 4\u0026deg;C (see \u003cstrong\u003eSupplementary Table 2\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eB cells were primed with\u0026nbsp;10\u0026nbsp;\u0026mu;g/ml\u0026nbsp;of\u0026nbsp;oligodeoxyribonucleotides containing CpG motifs (CpG ODNs; Miltenyi Biotec) for 72h at 37\u0026deg;C in a 5% CO\u003csub\u003e2\u003c/sub\u003e atmosphere in R10 culture medium without DNase I. Then B cells were stimulated with 50 ng/ml of PMA and 1 \u0026mu;g/ml of ionomycin for 4 h at 37\u0026deg;C in a 5% CO\u003csub\u003e2\u003c/sub\u003e atmosphere in complete culture medium as mentioned before. For each stimulated sample, an unstimulated one was prepared as a negative control. All samples were incubated with protein transport inhibitor containing Brefeldin A (Biolegend, San Diego, CA, USA) and monensin (Biolegend, San Diego, CA, USA). After stimulation, cells were washed with PBS and stained with viability marker PromoFluor IR-840 (Promokine, PromoCell, Heidelberg, Germany) for 20 min at RT. Next, cells were washed with FACS buffer and stained with surface mAbs (see \u003cstrong\u003eSupplementary Table 2\u003c/strong\u003e). After incubation all samples were washed with FACS buffer and fixed and permeabilized with the FoxP3/transcription factor staining buffer set (eBioscience) according to manufacturer\u0026rsquo;s instructions. Then, cells were stained with previously titrated mAbs recognizing intracellular cytokines and transcription factors for 30 min at 4\u0026deg;C (\u003cstrong\u003eSupplementary Table 2\u003c/strong\u003e). Gating strategy used to identify and analyze the intracellular cytokine production of CD4\u003csup\u003e+\u003c/sup\u003e, CD8\u003csup\u003e+\u003c/sup\u003e T lymphocytes and CD19\u003csup\u003e+\u003c/sup\u003e B lymphocytes are reported in Supplementary Fig. 22, 23 and 24. All mAbs were previously titrated to define the optimal concentration. All samples were acquired on a Cytoflex LX flow cytometer (Beckman Coulter, Brea, CA, USA).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePENCIL prediction analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePENCIL method relies on the concept known as LWR, a machine learning strategy that introduces rejection labels in the prediction results\u0026nbsp;[45]. The process of PENCIL workflow is depicted in Supplementary Fig. 25. Input data for PENCIL included hyperbolic arcsine transformed single-cell fluorescence intensity matrices and relevant cell metadata, such as cluster ID and UMAP coordinates.\u0026nbsp;These data were obtained from CATALYST and imported into Seurat using the \u003cem\u003eCreateSeuratObject\u003c/em\u003e function. We followed the standard workflows for CD4\u003csup\u003e+\u003c/sup\u003e, CD8\u003csup\u003e+\u003c/sup\u003e, and B cell prediction as recommended on the PENCIL GitHub page. PENCIL analyses were conducted using Python 3.11.4 with GPU acceleration (NVIDIA Quadro RTX 5000). Tuning parameters, including shuffle rate, lambda L1, and lambda L2, along with corresponding precision, recall, and f1-score, are detailed in Supplementary Fig. 13.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLiquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBlood monocytes and B cells were isolated using CD14 or CD19 MicroBeads (Miltenyi Biotec) through magnetic enrichment techniques. Subsequently, protein extraction was carried out following the previously described method\u0026nbsp;[75]. Fifty and ten \u0026micro;g of proteins of monocytes and B lymphocytes, respectively, were processed using FASP protocol\u0026nbsp;[76]. Protein digestion was performed by trypsin in an enzyme-to-protein ration of 1:50 (w/w). The tryptic peptides were suspended in water:acetonitrile:formic acid (95:3:2) for mass spectrometry analysis. The tryptic digests were injected into a Nano UHPLC Ultimate 3000 (Thermo Fisher Scientific) coupled to Exploris\u0026trade; 480 Hybrid Quadrupole-Orbitrap\u0026trade; Mass Spectrometer (Thermo Fisher Scientific). Separation was achieved using a C18 EASY-Spray HPLC column (75 \u0026mu;m \u0026times; 500 mm, 2 \u0026mu;m, ES903 Thermo Fisher Scientific) and elution was performed using a binary system of solvents. Solvent A was 0,1% formic acid and solvent B was 97% acetonitrile. A linear binary gradient was applied to eluate the peptides: 0\u0026ndash;23% solvent B in solvent A for 140 min and further 35 min of 23\u0026ndash;33% solvent B in solvent A at 300 nl/min.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eData-dependent acquisition (DDA)\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eFor monocytes sample the acquisition was operated in full MS/data-dependent acquisition. The Orbitrap mass analyzer was used at a resolution of 120,000 to acquire full MS with an m/z range of 400 to 1500. MS/MS fragments were measured at an Orbitrap resolution of 15,000. Twenty of the most intense ions were isolated for MS/MS analysis. The raw data were processed using Proteome Discoverer (version 3.0 SP1, Thermo Fisher Scientific), searching with a database of human proteome (www.uniprot.org, accessed November 2022, 20327 sequences). The selected parameters for protein identification were: (i) at least 1 unique peptide; (ii) static modification carbamidomethyl on cysteines (+57.021 Da), dynamic modifications oxidation on methionine (+15.995 Da) and deamidation on asparagine and glutamine (+0.984 Da); (iii) precursor mass tolerance of 10 ppm, fragment mass tolerance of 0.02 Da; (iv) the maximum of missed trypsin cleavage sites of 1; (v) the minimum peptide length of 7. Database search analysis was performed using Sequest HT and the artificial intelligence node Chimerys (Thermo Fisher Scientific). The resulting files were imported in Skyline (v.22.2.0.527) as already described\u0026nbsp;[77], to generate spectral library.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eTargeted MS Assay\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eFor B cell sample a targeted proteomic analysis for VISTA protein was performed by parallel reaction monitoring (PRM), as previously described\u0026nbsp;[78]. The peptides of VISTA identified in monocytes sample passing the false discovery rate were exported to a text file and processed by PRM. The mass inclusion list involved mass, charge, and polarity. Lists of all peptides targeted in the PRM analyses are reported in Supplementary Fig. 2. Sample was processed in triplicate for PRM. The MS data were processed using Skyline (v.22.2.0.527) using generated spectral library as reference. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eHypoxic assay\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePBMCs were isolated from fresh blood of eight donors (\u003cstrong\u003eSupplementary Table 1\u003c/strong\u003e). In each sample, 30 million of PBMC were washed with PBS and stained with viability marker LIVE/DEAD\u0026trade; Fixable Aqua (ThermoFisher Scientific, USA) for 20 min at RT. Next, cells were washed with FACS buffer and stained with anti-CD19-APC for 20 min at RT. After incubation all samples were washed with FACS buffer and sorted on Bigfoot Spectral Cell Sorter (ThermoFisher Scientific, USA). Purity and the gating strategy were reported in Supplementary Fig. 26. Sorted cells were plated at the concentration of 10\u003csup\u003e5\u003c/sup\u003e cells/100uL of R10 (without DNAse I to prevent CpG degradation) a 96-well U-bottom multiwell plate. For the stimulated condition, B cells were primed with 10\u0026nbsp;\u0026mu;g/ml\u0026nbsp;of\u0026nbsp;oligodeoxyribonucleotides containing CpG motifs (CpG ODNs; Miltenyi Biotec) for 48h at 37\u0026deg;C in a 5% CO\u003csub\u003e2\u003c/sub\u003e atmosphere (Normoxia) or at 37\u0026deg;C in a 1% O\u003csub\u003e2\u003c/sub\u003e and 5% CO\u003csub\u003e2\u003c/sub\u003e atmosphere (Hypoxia) within a hypoxia incubator chamber (Stemcell Technologies Inc., Canada). The cells subjected to hypoxia were quickly moved from the hypoxic chamber onto ice to prevent the swift degradation of hypoxia-responsive molecules. Subsequently, all cells were washed with cold PBS kept at 4\u0026deg;C and stained with viability marker LIVE/DEAD\u0026trade; Fixable Aqua (ThermoFisher Scientific, USA) for 20 min at 4\u0026deg;C. Next, cells were rinsed with chilled FACS buffer at 4\u0026deg;C and then stained with surface monoclonal antibodies for 20 minutes at 4\u0026deg;C. Intracellular detection of HIF-1\u0026alpha; was performed following FoxP3/transcription factor staining buffer set (eBioscience) according to manufacturer\u0026rsquo;s instructions incubating mAbs for 30 min at 4\u0026deg;C. All antibodies employed are listed in \u003cstrong\u003eSupplementary Table 2\u003c/strong\u003e. Data acquisition was performed on a Cytoflex LX flow cytometer (Beckman Coulter, Brea, CA, USA), with ex vivo samples serving as controls for both normoxic and hypoxic conditions (Supplementary Fig. 26).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eStatistical analyses were performed using R v4.1.0, GraphPad Prism version 8 (GraphPad Software Inc., La Jolla, USA) or SPICE software\u0026nbsp;[79], unless specified otherwise. Significance of differences for the frequency of single FlowSOM clusters was determined using generalized linear mixed model (GLMM) implemented within diffcyt package v1.14.0\u0026nbsp;[80]\u0026nbsp;applying FDR cutoff = 0.05. To compare distributions of manually gated subsets significance was determined by a nonparametric paired Wilcoxon rank test, unless otherwise specified in the figure legends. We used \u0026chi;2 permutation test for pie chart comparison (SPICE). Kaplan-Meier method was used to analyze\u0026nbsp;survival. The multivariate Cox regression analysis was conducted in R using the survival package v.3.4.0 to explore the correlation between patients' survival time and predictor variables identified by PENCIL.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003e\u003cstrong\u003eEthical Approval\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThe study protocol was approved by the local Ethical Committes (\u0026ldquo;Area Vasta Emilia Nord\u0026rdquo;, protocol number 0018007/21; \u0026ldquo;Comitato Etico della Romagna\u0026rdquo;, protocol number 7043/2021). All participants involved in this study provided written informed consent for both sample collection and data analyses. Detailed clinical information and the analysis methods employed for each sample are reported in \u003cstrong\u003eSupplementary Table\u0026nbsp;1\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting Interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll authors declare no competing interests.\u003c/p\u003e\n\u003ch2\u003eContributions\u003c/h2\u003e\n\u003cp\u003eDLT, VM, NP, FDL, AN, RB, ES, ALC, AVS, AD, MG, GM, FR, FT, FB performed experiments; DLT performed bioinformatic analysis; BA, FS, PLF, MD,FB enrolled the patients; DLT, BA, SDB, LG, DQ, AlC, ANeri, AC discussed the data; SDB and AC supervised experiments; DLT, SDB, LG and AC wrote the manuscript.\u003c/p\u003e\n\u003ch2\u003eFunding\u003c/h2\u003e\n\u003cp\u003eThis work was supported by grants from: Fondazione AIRC per la ricerca sul cancro to AC, project \u0026ldquo;Role of exhausted CD8 TILs in the recurrence of resectable non-small cell lung cancer\u0026rdquo; grant number 25073.\u003c/p\u003e\n\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\n\u003cp\u003eDLT, VM, NP, FDL, AN, RB, ES, ALC, AVS, AD, MG, GM, FR, FT, FB performed experiments; DLT performed bioinformatic analysis; BA, FS, PLF, MD,FB enrolled the patients; DLT, BA, SDB, LG, DQ, AlC, ANeri, AC discussed the data; SDB and AC supervised experiments; DLT, SDB, LG and AC wrote the manuscript.\u003c/p\u003e\n\u003ch2\u003eAcknowledgements\u003c/h2\u003e\n\u003cp\u003eSDB and LGi are Marylou Ingram Scholar of the International Society for Advancement of Cytometry (ISAC) for the period 2015\u0026ndash;2020 and 2020\u0026ndash;2025, respectively. Drs. Paola Paglia (ThermoFisher Scientific, Monza, Italy), Leonardo Beretta, Anis Larbi (Beckman Coulter, Milan, Italy), Dr. Paolo Santino, Ernesto Lopez, Gloria Martrus (Standard Biotools, San Francisco, CA, US) are acknowledged for their support in providing reagents and materials, for precious help and technical suggestions. We acknowledge Croce Blu (Modena hub) and its volunteers for the efficient service in transporting biological samples from Morgagni - Pierantoni Hospital to University of Modena. Finally, we gratefully acknowledge the patients who donated blood to participate in this study.\u003c/p\u003e\n\u003ch2\u003eAvailability of data and materials\u003c/h2\u003e\n\u003cp\u003eThe flow cytometry and mass cytometry data generated in this study are available under reasonable request. The raw data generated in this study are provided in the Source Data file. Further inquiries can be directed to the corresponding author.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eSung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, Bray F. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin. 2021;71:209\u0026ndash;49.\u003c/li\u003e\n \u003cli\u003eHartwig MG, D\u0026apos;Amico TA. Thoracoscopic lobectomy: the gold standard for early-stage lung cancer? Ann Thorac Surg. 2010;89:2098\u0026ndash;101.\u003c/li\u003e\n \u003cli\u003eSiegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, 2022. CA Cancer J Clin. 2022;72:7\u0026ndash;33.\u003c/li\u003e\n \u003cli\u003eRiley RS, June CH, Langer R, Mitchell MJ. Delivery technologies for cancer immunotherapy. Nat Rev Drug Discov. 2019;18:175\u0026ndash;96.\u003c/li\u003e\n \u003cli\u003eRoy R, Singh SK, Misra S. Advancements in Cancer Immunotherapies. Vaccines (Basel) 2022, 11.\u003c/li\u003e\n \u003cli\u003eGuo X, Zhang Y, Zheng L, Zheng C, Song J, Zhang Q, Kang B, Liu Z, Jin L, Xing R, et al. Global characterization of T cells in non-small-cell lung cancer by single-cell sequencing. Nat Med. 2018;24:978\u0026ndash;85.\u003c/li\u003e\n \u003cli\u003eGiraldo NA, Sanchez-Salas R, Peske JD, Vano Y, Becht E, Petitprez F, Validire P, Ingels A, Cathelineau X, Fridman WH, Sautes-Fridman C. The clinical role of the TME in solid cancer. Br J Cancer. 2019;120:45\u0026ndash;53.\u003c/li\u003e\n \u003cli\u003eAramini B, Masciale V, Samarelli AV, Dubini A, Gaudio M, Stella F, Morandi U, Dominici M, De Biasi S, Gibellini L, Cossarizza A. Phenotypic, functional, and metabolic heterogeneity of immune cells infiltrating non-small cell lung cancer. Front Immunol. 2022;13:959114.\u003c/li\u003e\n \u003cli\u003eSautes-Fridman C, Petitprez F, Calderaro J, Fridman WH. Tertiary lymphoid structures in the era of cancer immunotherapy. Nat Rev Cancer. 2019;19:307\u0026ndash;25.\u003c/li\u003e\n \u003cli\u003eFukuhara M, Muto S, Inomata S, Yamaguchi H, Mine H, Takagi H, Ozaki Y, Watanabe M, Inoue T, Yamaura T, et al. The clinical significance of tertiary lymphoid structure and its relationship with peripheral blood characteristics in patients with surgically resected non-small cell lung cancer: a single-center, retrospective study. Cancer Immunol Immunother. 2022;71:1129\u0026ndash;37.\u003c/li\u003e\n \u003cli\u003eGu-Trantien C, Loi S, Garaud S, Equeter C, Libin M, de Wind A, Ravoet M, Le Buanec H, Sibille C, Manfouo-Foutsop G, et al. CD4(+) follicular helper T cell infiltration predicts breast cancer survival. J Clin Invest. 2013;123:2873\u0026ndash;92.\u003c/li\u003e\n \u003cli\u003eGoc J, Germain C, Vo-Bourgais TK, Lupo A, Klein C, Knockaert S, de Chaisemartin L, Ouakrim H, Becht E, Alifano M, et al. Dendritic cells in tumor-associated tertiary lymphoid structures signal a Th1 cytotoxic immune contexture and license the positive prognostic value of infiltrating CD8\u0026thinsp;+\u0026thinsp;T cells. Cancer Res. 2014;74:705\u0026ndash;15.\u003c/li\u003e\n \u003cli\u003eHennequin A, Derangere V, Boidot R, Apetoh L, Vincent J, Orry D, Fraisse J, Causeret S, Martin F, Arnould L, et al. Tumor infiltration by Tbet\u0026thinsp;+\u0026thinsp;effector T cells and CD20\u0026thinsp;+\u0026thinsp;B cells is associated with survival in gastric cancer patients. Oncoimmunology. 2016;5:e1054598.\u003c/li\u003e\n \u003cli\u003eHelmink BA, Reddy SM, Gao J, Zhang S, Basar R, Thakur R, Yizhak K, Sade-Feldman M, Blando J, Han G, et al. B cells and tertiary lymphoid structures promote immunotherapy response. Nature. 2020;577:549\u0026ndash;55.\u003c/li\u003e\n \u003cli\u003eSorin M, Rezanejad M, Karimi E, Fiset B, Desharnais L, Perus LJM, Milette S, Yu MW, Maritan SM, Dore S, et al. Single-cell spatial landscapes of the lung tumour immune microenvironment. Nature. 2023;614:548\u0026ndash;54.\u003c/li\u003e\n \u003cli\u003eFridman WH, Petitprez F, Meylan M, Chen TW, Sun CM, Roumenina LT, Sautes-Fridman C. B cells and cancer: To B or not to B? J Exp Med 2021, 218.\u003c/li\u003e\n \u003cli\u003eHu Q, Hong Y, Qi P, Lu G, Mai X, Xu S, He X, Guo Y, Gao L, Jing Z, et al. Atlas of breast cancer infiltrated B-lymphocytes revealed by paired single-cell RNA-sequencing and antigen receptor profiling. Nat Commun. 2021;12:2186.\u003c/li\u003e\n \u003cli\u003eShalapour S, Font-Burgada J, Di Caro G, Zhong Z, Sanchez-Lopez E, Dhar D, Willimsky G, Ammirante M, Strasner A, Hansel DE, et al. Immunosuppressive plasma cells impede T-cell-dependent immunogenic chemotherapy. Nature. 2015;521:94\u0026ndash;8.\u003c/li\u003e\n \u003cli\u003eCai C, Zhang J, Li M, Wu ZJ, Song KH, Zhan TW, Wang LH, Sun YH. Interleukin 10-expressing B cells inhibit tumor-infiltrating T cell function and correlate with T cell Tim-3 expression in renal cell carcinoma. Tumour Biol. 2016;37:8209\u0026ndash;18.\u003c/li\u003e\n \u003cli\u003eMurakami Y, Saito H, Shimizu S, Kono Y, Shishido Y, Miyatani K, Matsunaga T, Fukumoto Y, Ashida K, Sakabe T, et al. Increased regulatory B cells are involved in immune evasion in patients with gastric cancer. Sci Rep. 2019;9:13083.\u003c/li\u003e\n \u003cli\u003eDeng J, Li J, Sarde A, Lines JL, Lee YC, Qian DC, Pechenick DA, Manivanh R, Le Mercier I, Lowrey CH, et al. Hypoxia-Induced VISTA Promotes the Suppressive Function of Myeloid-Derived Suppressor Cells in the Tumor Microenvironment. Cancer Immunol Res. 2019;7:1079\u0026ndash;90.\u003c/li\u003e\n \u003cli\u003eDeRogatis JM, Viramontes KM, Neubert EN, Henriquez ML, Guerrero-Juarez CF, Tinoco R. Targeting the PSGL-1 Immune Checkpoint Promotes Immunity to PD-1-Resistant Melanoma. Cancer Immunol Res. 2022;10:612\u0026ndash;25.\u003c/li\u003e\n \u003cli\u003eElTanbouly MA, Zhao Y, Nowak E, Li J, Schaafsma E, Le Mercier I, Ceeraz S, Lines JL, Peng C, Carriere C et al. VISTA is a checkpoint regulator for naive T cell quiescence and peripheral tolerance. Science 2020, 367.\u003c/li\u003e\n \u003cli\u003eHope JL, Otero DC, Bae EA, Stairiker CJ, Palete AB, Faso HA, Lin M, Henriquez ML, Roy S, Seo H, et al. PSGL-1 attenuates early TCR signaling to suppress CD8(+) T cell progenitor differentiation and elicit terminal CD8(+) T cell exhaustion. Cell Rep. 2023;42:112436.\u003c/li\u003e\n \u003cli\u003eIssam Salah NEI, Marnissi F, Lakhdar A, Karkouri M, ElBelhadji M, Badou A. The immune checkpoint VISTA is associated with prognosis in patients with malignant uveal melanoma. Front Immunol. 2023;14:1225140.\u003c/li\u003e\n \u003cli\u003eKappelmayer J, Nagy B Jr.. The Interaction of Selectins and PSGL-1 as a Key Component in Thrombus Formation and Cancer Progression. Biomed Res Int. 2017;2017:6138145.\u003c/li\u003e\n \u003cli\u003eKuklinski LF, Yan S, Li Z, Fisher JL, Cheng C, Noelle RJ, Angeles CV, Turk MJ, Ernstoff MS. VISTA expression on tumor-infiltrating inflammatory cells in primary cutaneous melanoma correlates with poor disease-specific survival. Cancer Immunol Immunother. 2018;67:1113\u0026ndash;21.\u003c/li\u003e\n \u003cli\u003eLiao H, Zhu H, Liu S, Wang H. Expression of V-domain immunoglobulin suppressor of T cell activation is associated with the advanced stage and presence of lymph node metastasis in ovarian cancer. Oncol Lett. 2018;16:3465\u0026ndash;72.\u003c/li\u003e\n \u003cli\u003eLin Y, Huang S, Qi Y, Xie L, Jiang J, Li H, Chen Z. PSGL-1 is a novel tumor microenvironment prognostic biomarker with cervical high-grade squamous lesions and more. Front Oncol. 2023;13:1052201.\u003c/li\u003e\n \u003cli\u003eTinoco R, Carrette F, Barraza ML, Otero DC, Magana J, Bosenberg MW, Swain SL, Bradley LM. PSGL-1 Is an Immune Checkpoint Regulator that Promotes T Cell Exhaustion. Immunity. 2016;44:1190\u0026ndash;203.\u003c/li\u003e\n \u003cli\u003eWu L, Deng WW, Huang CF, Bu LL, Yu GT, Mao L, Zhang WF, Liu B, Sun ZJ. Expression of VISTA correlated with immunosuppression and synergized with CD8 to predict survival in human oral squamous cell carcinoma. Cancer Immunol Immunother. 2017;66:627\u0026ndash;36.\u003c/li\u003e\n \u003cli\u003eGermain C, Gnjatic S, Tamzalit F, Knockaert S, Remark R, Goc J, Lepelley A, Becht E, Katsahian S, Bizouard G, et al. Presence of B cells in tertiary lymphoid structures is associated with a protective immunity in patients with lung cancer. Am J Respir Crit Care Med. 2014;189:832\u0026ndash;44.\u003c/li\u003e\n \u003cli\u003eWouters MCA, Nelson BH. Prognostic Significance of Tumor-Infiltrating B Cells and Plasma Cells in Human Cancer. Clin Cancer Res. 2018;24:6125\u0026ndash;35.\u003c/li\u003e\n \u003cli\u003eMauri C, Menon M. Human regulatory B cells in health and disease: therapeutic potential. J Clin Invest. 2017;127:772\u0026ndash;9.\u003c/li\u003e\n \u003cli\u003eIsnardi I, Ng YS, Menard L, Meyers G, Saadoun D, Srdanovic I, Samuels J, Berman J, Buckner JH, Cunningham-Rundles C, Meffre E. Complement receptor 2/CD21- human naive B cells contain mostly autoreactive unresponsive clones. Blood. 2010;115:5026\u0026ndash;36.\u003c/li\u003e\n \u003cli\u003eFridman WH, Meylan M, Petitprez F, Sun CM, Italiano A, Sautes-Fridman C. B cells and tertiary lymphoid structures as determinants of tumour immune contexture and clinical outcome. Nat Rev Clin Oncol. 2022;19:441\u0026ndash;57.\u003c/li\u003e\n \u003cli\u003eHartmann FJ, Mrdjen D, McCaffrey E, Glass DR, Greenwald NF, Bharadwaj A, Khair Z, Verberk SGS, Baranski A, Baskar R, et al. Single-cell metabolic profiling of human cytotoxic T cells. Nat Biotechnol. 2021;39:186\u0026ndash;97.\u003c/li\u003e\n \u003cli\u003eDe Biasi SLT, Neroni D et al. A.;. : Metabolic pathways engaged by antigen-specific T and B cells after SARS-CoV-2 vaccination in multiple sclerosis patients on different immunomodulatory drugs reveal immunosenescence and predict vaccine efficacy. \u003cem\u003ePREPRINT (Version 1) available at Research Square\u003c/em\u003e 2023.\u003c/li\u003e\n \u003cli\u003eMiller RA, Luke JJ, Hu S, Mahabhashyam S, Jones WB, Marron T, Merchan JR, Hughes BGM, Willingham SB. Anti-CD73 antibody activates human B cells, enhances humoral responses and induces redistribution of B cells in patients with cancer. J Immunother Cancer 2022, 10.\u003c/li\u003e\n \u003cli\u003ede Gruijter NM, Jebson B, Rosser EC. Cytokine production by human B cells: role in health and autoimmune disease. Clin Exp Immunol. 2022;210:253\u0026ndash;62.\u003c/li\u003e\n \u003cli\u003eJohnston RJ, Su LJ, Pinckney J, Critton D, Boyer E, Krishnakumar A, Corbett M, Rankin AL, Dibella R, Campbell L, et al. VISTA is an acidic pH-selective ligand for PSGL-1. Nature. 2019;574:565\u0026ndash;70.\u003c/li\u003e\n \u003cli\u003eRogers BM, Smith L, Dezso Z, Shi X, DiGiammarino E, Nguyen D, Sethuraman S, Zheng P, Choi D, Zhang D et al. VISTA is an activating receptor in human monocytes. J Exp Med 2021, 218.\u003c/li\u003e\n \u003cli\u003eShanshan H, Ruchir B, Carl B, Emily AB, Derek LB, Kan C, Patrick D, Dwayne D, Ryan GG, Gary G et al. High-plex Multiomic Analysis in FFPE at Subcellular Level by Spatial Molecular Imaging. \u003cem\u003ebioRxiv\u003c/em\u003e 2022:2021.2011.2003.467020..\u003c/li\u003e\n \u003cli\u003eDi Pilato M, Kfuri-Rubens R, Pruessmann JN, Ozga AJ, Messemaker M, Cadilha BL, Sivakumar R, Cianciaruso C, Warner RD, Marangoni F, et al. CXCR6 positions cytotoxic T cells to receive critical survival signals in the tumor microenvironment. Cell. 2021;184:4512\u0026ndash;4530e4522.\u003c/li\u003e\n \u003cli\u003eRen T, Chen C, Danilov AV, Liu S, Guan X, Du S, Wu X, Sherman MH, Spellman PT, Coussens LM, et al. Supervised learning of high-confidence phenotypic subpopulations from single-cell data. Nat Mach Intell. 2023;5:528\u0026ndash;41.\u003c/li\u003e\n \u003cli\u003ePearce EL, Poffenberger MC, Chang CH, Jones RG. Fueling immunity: insights into metabolism and lymphocyte function. Science. 2013;342:1242454.\u003c/li\u003e\n \u003cli\u003eNorata GD, Caligiuri G, Chavakis T, Matarese G, Netea MG, Nicoletti A, O\u0026apos;Neill LA, Marelli-Berg FM. The Cellular and Molecular Basis of Translational Immunometabolism. Immunity. 2015;43:421\u0026ndash;34.\u003c/li\u003e\n \u003cli\u003eItakura E, Huang RR, Wen DR, Paul E, Wunsch PH, Cochran AJ. IL-10 expression by primary tumor cells correlates with melanoma progression from radial to vertical growth phase and development of metastatic competence. Mod Pathol. 2011;24:801\u0026ndash;9.\u003c/li\u003e\n \u003cli\u003eHuseni MA, Wang L, Klementowicz JE, Yuen K, Breart B, Orr C, Liu LF, Li Y, Gupta V, Li C, et al. CD8(+) T cell-intrinsic IL-6 signaling promotes resistance to anti-PD-L1 immunotherapy. Cell Rep Med. 2023;4:100878.\u003c/li\u003e\n \u003cli\u003eHong IS. Stimulatory versus suppressive effects of GM-CSF on tumor progression in multiple cancer types. Exp Mol Med. 2016;48:e242.\u003c/li\u003e\n \u003cli\u003eYan F, Du R, Wei F, Zhao H, Yu J, Wang C, Zhan Z, Ding T, Ren X, Chen X, Li H. Expression of TNFR2 by regulatory T cells in peripheral blood is correlated with clinical pathology of lung cancer patients. Cancer Immunol Immunother. 2015;64:1475\u0026ndash;85.\u003c/li\u003e\n \u003cli\u003eNakamura K, Kitani A, Strober W. Cell contact-dependent immunosuppression by CD4(+)CD25(+) regulatory T cells is mediated by cell surface-bound transforming growth factor beta. J Exp Med. 2001;194:629\u0026ndash;44.\u003c/li\u003e\n \u003cli\u003eYuan L, Tatineni J, Mahoney KM, Freeman GJ. VISTA: A Mediator of Quiescence and a Promising Target in Cancer Immunotherapy. Trends Immunol. 2021;42:209\u0026ndash;27.\u003c/li\u003e\n \u003cli\u003eKallies A, Good-Jacobson KL. Transcription Factor T-bet Orchestrates Lineage Development and Function in the Immune System. Trends Immunol. 2017;38:287\u0026ndash;97.\u003c/li\u003e\n \u003cli\u003eWein AN, McMaster SR, Takamura S, Dunbar PR, Cartwright EK, Hayward SL, McManus DT, Shimaoka T, Ueha S, Tsukui T, et al. CXCR6 regulates localization of tissue-resident memory CD8 T cells to the airways. J Exp Med. 2019;216:2748\u0026ndash;62.\u003c/li\u003e\n \u003cli\u003eHurkmans DP, Basak EA, Schepers N, Oomen-De Hoop E, Van der Leest CH, El Bouazzaoui S, Bins S, Koolen SLW, Sleijfer S, Van der Veldt AAM et al. Granzyme B is correlated with clinical outcome after PD-1 blockade in patients with stage IV non-small-cell lung cancer. J Immunother Cancer 2020, 8.\u003c/li\u003e\n \u003cli\u003eBraun M, Aguilera AR, Sundarrajan A, Corvino D, Stannard K, Krumeich S, Das I, Lima LG, Meza Guzman LG, Li K, et al. CD155 on Tumor Cells Drives Resistance to Immunotherapy by Inducing the Degradation of the Activating Receptor CD226 in CD8(+) T Cells. Immunity. 2020;53:805\u0026ndash;823e815.\u003c/li\u003e\n \u003cli\u003eChiang EY, de Almeida PE, de Almeida Nagata DE, Bowles KH, Du X, Chitre AS, Banta KL, Kwon Y, McKenzie B, Mittman S, et al. CD96 functions as a co-stimulatory receptor to enhance CD8(+) T cell activation and effector responses. Eur J Immunol. 2020;50:891\u0026ndash;902.\u003c/li\u003e\n \u003cli\u003eEttinger DS, Wood DE, Aisner DL, Akerley W, Bauman JR, Bharat A, Bruno DS, Chang JY, Chirieac LR, D\u0026apos;Amico TA, et al. Non-Small Cell Lung Cancer, Version 3.2022, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw. 2022;20:497\u0026ndash;530.\u003c/li\u003e\n \u003cli\u003eMartini N, Bains MS, Burt ME, Zakowski MF, McCormack P, Rusch VW, Ginsberg RJ. Incidence of local recurrence and second primary tumors in resected stage I lung cancer. J Thorac Cardiovasc Surg. 1995;109:120\u0026ndash;9.\u003c/li\u003e\n \u003cli\u003eSilvan J, Gonzalez-Tajuelo R, Vicente-Rabaneda E, Perez-Frias A, Espartero-Santos M, Munoz-Callejas A, Garcia-Lorenzo E, Gamallo C, Castaneda S, Urzainqui A. Deregulated PSGL-1 Expression in B Cells and Dendritic Cells May Be Implicated in Human Systemic Sclerosis Development. J Invest Dermatol. 2018;138:2123\u0026ndash;32.\u003c/li\u003e\n \u003cli\u003eCossarizza A, Chang HD, Radbruch A, Abrignani S, Addo R, Akdis M, Andra I, Andreata F, Annunziato F, Arranz E, et al. Guidelines for the use of flow cytometry and cell sorting in immunological studies (third edition). Eur J Immunol. 2021;51:2708\u0026ndash;3145.\u003c/li\u003e\n \u003cli\u003eDe Biasi S, Gibellini L, Lo Tartaro D, Puccio S, Rabacchi C, Mazza EMC, Brummelman J, Williams B, Kaihara K, Forcato M, et al. Circulating mucosal-associated invariant T cells identify patients responding to anti-PD-1 therapy. Nat Commun. 2021;12:1669.\u003c/li\u003e\n \u003cli\u003eHahne F, LeMeur N, Brinkman RR, Ellis B, Haaland P, Sarkar D, Spidlen J, Strain E, Gentleman R. flowCore: a Bioconductor package for high throughput flow cytometry. BMC Bioinformatics. 2009;10:106.\u003c/li\u003e\n \u003cli\u003eNowicka M, Krieg C, Crowell HL, Weber LM, Hartmann FJ, Guglietta S, Becher B, Levesque MP, Robinson MD. CyTOF workflow: differential discovery in high-throughput high-dimensional cytometry datasets. \u003cem\u003eF1000Res\u003c/em\u003e 2017, 6:748.\u003c/li\u003e\n \u003cli\u003eMelsen JE, van Ostaijen-Ten Dam MM, Lankester AC, Schilham MW, van den Akker EB. A Comprehensive Workflow for Applying Single-Cell Clustering and Pseudotime Analysis to Flow Cytometry Data. J Immunol. 2020;205:864\u0026ndash;71.\u003c/li\u003e\n \u003cli\u003eStreet K, Risso D, Fletcher RB, Das D, Ngai J, Yosef N, Purdom E, Dudoit S. Slingshot: cell lineage and pseudotime inference for single-cell transcriptomics. BMC Genomics. 2018;19:477.\u003c/li\u003e\n \u003cli\u003eGu Z, Eils R, Schlesner M. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics. 2016;32:2847\u0026ndash;9.\u003c/li\u003e\n \u003cli\u003eDiggins KE, Greenplate AR, Leelatian N, Wogsland CE, Irish JM. Characterizing cell subsets using marker enrichment modeling. Nat Methods. 2017;14:275\u0026ndash;8.\u003c/li\u003e\n \u003cli\u003eHafemeister C, Satija R. Normalization and variance stabilization of single-cell RNA-seq data using regularized negative binomial regression. Genome Biol. 2019;20:296.\u003c/li\u003e\n \u003cli\u003eButler A, Hoffman P, Smibert P, Papalexi E, Satija R. Integrating single-cell transcriptomic data across different conditions, technologies, and species. Nat Biotechnol. 2018;36:411\u0026ndash;20.\u003c/li\u003e\n \u003cli\u003eZappia L, Oshlack A. Clustering trees: a visualization for evaluating clusterings at multiple resolutions. Gigascience 2018, 7.\u003c/li\u003e\n \u003cli\u003eOliveira G, Stromhaug K, Klaeger S, Kula T, Frederick DT, Le PM, Forman J, Huang T, Li S, Zhang W, et al. Phenotype, specificity and avidity of antitumour CD8(+) T cells in melanoma. Nature. 2021;596:119\u0026ndash;25.\u003c/li\u003e\n \u003cli\u003eWang J, Wu G, Manick B, Hernandez V, Renelt M, Erickson C, Guan J, Singh R, Rollins S, Solorz A, et al. VSIG-3 as a ligand of VISTA inhibits human T-cell function. Immunology. 2019;156:74\u0026ndash;85.\u003c/li\u003e\n \u003cli\u003eZanini G, Selleri V, Nasi M, De Gaetano A, Martinelli I, Gianferrari G, Lofaro FD, Boraldi F, Mandrioli J, Pinti M. Mitochondrial and Endoplasmic Reticulum Alterations in a Case of Amyotrophic Lateral Sclerosis Caused by TDP-43 A382T Mutation. Int J Mol Sci 2022, 23.\u003c/li\u003e\n \u003cli\u003eWisniewski JR. Filter-Aided Sample Preparation for Proteome Analysis. Methods Mol Biol. 2018;1841:3\u0026ndash;10.\u003c/li\u003e\n \u003cli\u003eLofaro FD, Cisterna B, Lacavalla MA, Boschi F, Malatesta M, Quaglino D, Zancanaro C, Boraldi F. Age-Related Changes in the Matrisome of the Mouse Skeletal Muscle. Int J Mol Sci 2021, 22.\u003c/li\u003e\n \u003cli\u003eZhang Q, Wu L, Bai B, Li D, Xiao P, Li Q, Zhang Z, Wang H, Li L, Jiang Q. Quantitative Proteomics Reveals Association of Neuron Projection Development Genes ARF4, KIF5B, and RAB8A With Hirschsprung Disease. Mol Cell Proteom. 2021;20:100007.\u003c/li\u003e\n \u003cli\u003eRoederer M, Nozzi JL, Nason MC. SPICE: exploration and analysis of post-cytometric complex multivariate datasets. Cytometry A. 2011;79:167\u0026ndash;74.\u003c/li\u003e\n \u003cli\u003eWeber LM, Nowicka M, Soneson C, Robinson MD. diffcyt: Differential discovery in high-dimensional cytometry via high-resolution clustering. Commun Biol. 2019;2:183.\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cdiv class=\"gridtable\"\u003e\n \u003ctable id=\"Tab1\" border=\"1\"\u003e\n \u003ccaption\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Bold\"\u003eNSCLC patients\u0026rsquo; clinical characteristics.\u003c/span\u003e SD, Standard Deviation; no, number; COPD, Chronic obstructive pulmonary disease; ADK, adenocarcinoma; CT, Computed Tomography, PET, Positron Emission Tomography.\u003c/div\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Italic\"\u003eParameters\u003c/span\u003e\u003c/div\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Italic\"\u003ePatients (n\u0026thinsp;=\u0026thinsp;37)\u003c/span\u003e\u003c/div\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Italic\"\u003eClinical features of patients\u003c/span\u003e\u003c/div\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Bold\"\u003eAge at surgery, mean (\u0026plusmn;\u0026thinsp;SD)\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e71.4 (\u0026plusmn;\u0026thinsp;7.2)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Bold\"\u003eGender, no. (%)\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eFemale\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e19.0 (51.4)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eMale\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e18.0 (48.6)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Bold\"\u003eSmoking status - no. (%)\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eCurrent smoker\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e26.0 (70.3)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eFormer smoker\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e3.0 (8.1)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eCigarette packs for year, mean (\u003cspan class=\"Bold\"\u003e\u0026plusmn;\u003c/span\u003e\u0026thinsp;SD)\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e36.3 (\u0026plusmn;\u0026thinsp;33.5)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eNever smoked\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e8.0 (21.6)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Bold\"\u003eCommon comorbidities, no. (%)\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eDiabetes mellitus, type II\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e3.0 (8.1)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eHypertension\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e8.0 (21.6)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eDyslipidaemia\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e3.0 (8.1)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eCOPD\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e7.0 (18.9)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eConcomitant comorbidities*\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e15.0 (40.5)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eNone\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e1.0\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Bold\"\u003ePrevious tumours, no. (%)\u003c/span\u003e *\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e10.0 (27.0)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Bold\"\u003eRecurrence, no. (%)\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e7.0 (18.9)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Bold\"\u003eSurvival follow-up, no. live (%)\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e35.0 (94.6)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"BoldItalic\"\u003eNSCLC characterization\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Bold\"\u003eTumour type, no. (%)\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Italic\"\u003eAdenocarcinoma\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"BoldItalic\"\u003e25.0 (67.6)\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eADK lepidic\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u0026minus; 0.0 (0.0)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eADK acinar\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u0026minus; 12.0 (48.0)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eADK solid\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u0026minus; 3.0 (12.0)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eADK papillary\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u0026minus; 1.0 (4.0)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eADK micropapillary\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u0026minus; 0.0 (0.0)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eOther*\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u0026minus; 9.0 (36.0)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Italic\"\u003eSquamous-cell carcinoma\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"BoldItalic\"\u003e8.0 (21.6)\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eKeratinizing\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u0026minus; 6.0 (75.0)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eNonkeratinizing\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u0026minus; 2.0 (25.0)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Italic\"\u003eOther*\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"BoldItalic\"\u003e4.0 (10.8)\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Bold\"\u003epStage TNM 8th edition, no. (%)\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eIA1\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e1.0 (2.7)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eIA2\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e6.0 (16.2)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eIA3\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e7.0 (18.9)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eIB\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e5.0 (13.5)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eIIA\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e5.0 (13.5)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eIIB\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e9.0 (24.3)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003eIIIA\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e4.0 (10.8)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Bold\"\u003eChest CT, no. (range mm)\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u0026le;\u0026thinsp;27 mm\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e19.0 (15.0\u0026ndash;27.0)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u0026gt;\u0026thinsp;27 mm\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e18.0 (30.0\u0026ndash;97.0)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Bold\"\u003eFDG total-body PET, no. (range)\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u0026le;\u0026thinsp;7.1 SUV Max Low\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e18.0 (1.2\u0026ndash;7.0)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u0026gt;\u0026thinsp;7.1 SUV Max High\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e19.0 (7.2\u0026ndash;31.0)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Bold\"\u003eAdjuvant radiotherapy, no. (%)\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e1.0 (2.7)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Bold\"\u003eAdjuvant chemotherapy, no. (%)\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e6.0 (16.2)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Bold\"\u003eVascular infiltration, no. (%)\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e3.0 (8.1)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e\u003cspan class=\"Bold\"\u003eVisceral pleural infiltration, no. (%)\u003c/span\u003e\u003c/div\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cdiv class=\"SimplePara\"\u003e11.0 (29.7)\u003c/div\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003ctfoot\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\"\u003e* See Supplementary Table\u0026nbsp;1 for more detailed information.\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tfoot\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\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":"molecular-cancer","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"molc","sideBox":"Learn more about [Molecular Cancer](http://gsejournal.biomedcentral.com/)","snPcode":"12943","submissionUrl":"https://submission.nature.com/new-submission/12943/3","title":"Molecular Cancer","twitterHandle":"@SN_Oncology","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Bregs, NSCLC, recurrence, prediction, PSGL-1, VISTA","lastPublishedDoi":"10.21203/rs.3.rs-3891288/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3891288/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eB cells have emerged as central players in the tumor microenvironment (TME) of non-small cell lung cancer (NSCLC). However, although there is clear evidence for their involvement in cancer immunity, scanty data exist on the characterization of B cell phenotypes, bioenergetic profiles and possible interactions with T cells in the context of NSCLC.\u003c/p\u003e \u003cp\u003eIn this study, using polychromatic flow cytometry, mass cytometry, and spatial transcriptomics we explored the intricate landscape of B cell phenotypes, bioenergetics, and their interaction with T cells in NSCLC. Our analysis revealed that TME contains diverse B cell clusters, including VISTA\u003csup\u003e+\u003c/sup\u003e Bregs, with distinct metabolic and functional profiles. Target liquid chromatography-tandem mass spectrometry confirmed the expression of VISTA on B cells. Pseudotime analysis unveiled a B cell differentiation process leading to a branch formed by plasmablasts/plasma cells, or to another made by VISTA\u003csup\u003e+\u003c/sup\u003e Bregs. Spatial analysis showed colocalization of B cells with CD4\u003csup\u003e+\u003c/sup\u003e/CD8\u003csup\u003e+\u003c/sup\u003e T lymphocytes in TME. The computational analysis of intercellular communications that links ligands to target genes, performed by NicheNet, predicted B-T interactions \u003cem\u003evia\u003c/em\u003e VISTA-PSGL1 axis. Notably, tumor infiltrating CD8\u003csup\u003e+\u003c/sup\u003e T cells expressing PSGL1 exhibited enhanced metabolism and cytotoxicity. In NSCLC patients, prediction analysis performed by PENCIL revealed the presence of an association between PSGL1\u003csup\u003e+\u003c/sup\u003eCD8\u003csup\u003e+\u003c/sup\u003e T cells and VISTA\u003csup\u003e+\u003c/sup\u003e Bregs with lung recurrence. Our findings suggest a potential interaction between Bregs and T cells through the VISTA-PSGL1 axis, able of influencing NSCLC recurrence.\u003c/p\u003e","manuscriptTitle":"Metabolically activated and highly polyfunctional intratumoral VISTA+ regulatory B cells are associated with tumor recurrence in early stage NSCLC.","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-01-30 17:25:31","doi":"10.21203/rs.3.rs-3891288/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-06-03T07:57:22+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-06-02T14:46:19+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-05-28T06:56:20+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"124745075767127228318616766910927701429","date":"2024-05-23T15:15:12+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"118681746344818967159451357080799242139","date":"2024-05-22T04:48:15+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"fe0f0eaa-5887-49ae-84e9-d39a2943a672","date":"2024-04-09T15:30:02+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-03-07T22:15:01+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"59555b7a-46c2-486a-aa2a-cb92990085d7","date":"2024-02-27T14:05:10+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"01c265dd-06c6-4cb8-84ae-cf7d1c0a3ff6","date":"2024-02-20T11:33:49+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-01-30T14:01:36+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-01-26T00:40:45+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-01-26T00:40:45+00:00","index":"","fulltext":""},{"type":"submitted","content":"Molecular Cancer","date":"2024-01-23T14:01:15+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"molecular-cancer","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"molc","sideBox":"Learn more about [Molecular Cancer](http://gsejournal.biomedcentral.com/)","snPcode":"12943","submissionUrl":"https://submission.nature.com/new-submission/12943/3","title":"Molecular Cancer","twitterHandle":"@SN_Oncology","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"a41b3879-d764-4b82-88f0-17d377e2b71b","owner":[],"postedDate":"January 30th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-01-20T16:02:14+00:00","versionOfRecord":{"articleIdentity":"rs-3891288","link":"https://doi.org/10.1186/s12943-024-02209-2","journal":{"identity":"molecular-cancer","isVorOnly":false,"title":"Molecular Cancer"},"publishedOn":"2025-01-14 15:57:31","publishedOnDateReadable":"January 14th, 2025"},"versionCreatedAt":"2024-01-30 17:25:31","video":"","vorDoi":"10.1186/s12943-024-02209-2","vorDoiUrl":"https://doi.org/10.1186/s12943-024-02209-2","workflowStages":[]},"version":"v1","identity":"rs-3891288","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3891288","identity":"rs-3891288","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}