Abstract
Chronic antigen exposure drives CD8+ T cell exhaustion; however, strategies to maintain a long-lived, functional CD8+ T cell population under chronic stimulation remain unclear. In this study, we demonstrate that overexpression of nitric oxide synthase–interacting protein (NOSIP) enhances the persistence of antigen-specific CD8+ T cells under chronic antigen stimulation. Notably, NOSIP overexpression preserved a less differentiated CX3CR1neg subset and inhibited its progression toward an apoptosis-prone CX3CR1hi state, which was associated with reducing cell death and promoting long-term persistence. In a tumor model, NOSIP-overexpressing CD8+ T cells exhibited improved tumor control, indicating that NOSIP-mediated persistence confers superior antitumor capacity. Furthermore, NOSIP overexpression increased the responsiveness of CD8+ T cells to programmed death-ligand 1 blockade, suggesting that NOSIP may represent a promising therapeutic target.
1 Introduction
During chronic viral infections and tumor progression, persistent antigen stimulation induces CD8+ T cell exhaustion, which is characterized by impaired effector functions, upregulation of immune checkpoint molecules, and increased apoptosis, ultimately resulting in inadequate long-term host protection (, ). Exhausted CD8+ T cells are heterogeneous and comprise two major subsets: progenitor-exhausted and terminally exhausted cells (, ). While progenitor-exhausted cells retain proliferative capacity and can respond to immunotherapies such as immune checkpoint blockades (ICBs), the dysfunctional state of terminally exhausted cells is epigenetically imprinted and irreversible. Consequently, ICB therapies, including programmed cell death protein 1 (PD-1)/programmed death-ligand 1 (PD-L1) and cytotoxic T-lymphocyte–associated protein 4 inhibitors, primarily act on the progenitor pool, leading to transient restoration of function (–). Although combination strategies—such as ICB with co-stimulatory agonists (e.g., 4-1BB, OX40) (), cytokine support (e.g., IL-7) (), and metabolic or epigenetic modulators (, )—have been explored to improve ICB efficacy, maintaining durable CD8+ T cell responses under chronic antigen exposure remains challenging. This persistent limitation underscores the need to develop novel strategies that sustain long-lived, functional CD8+ T cells.
Previously, we longitudinally tracked peptide-specific CD8+ T cells from patients with hepatocellular carcinoma (HCC) who survived for more than 10 years following peptide vaccination (). To investigate their long-term immune responses, we analyzed tetramer+ peripheral blood mononuclear cells (PBMC) collected five years after vaccination using single-cell RNA sequencing (scRNA-seq). Most peptide-specific CD8+ T cells were enriched in a dominant cluster with a memory-like phenotype and notably exhibited elevated expression of nitric oxide synthase–interacting protein (NOSIP), suggesting a potential association between NOSIP and durable T cell persistence.
NOSIP, a 34-kDa protein, was initially identified as a binding partner of endothelial nitric oxide synthase (eNOS) (). It modulates the enzymatic activity of both eNOS and neuronal nitric oxide synthase (nNOS) by altering their subcellular localization, thereby attenuating nitric oxide (NO) production (–). NOSIP has been implicated in diverse processes, including vascular remodeling, mucosal secretion, neutrophil differentiation, gastrointestinal motility and astrogliogenesis (–). In addition to regulating NOS activity, NOSIP contains a U-box domain and functions as an E3 ubiquitin ligase that modulates protein phosphatase 2A (PP2A) signaling during brain and craniofacial development (). It also participates in retinoic acid signaling by regulating Rbp1 expression, thereby influencing early neurogenesis (). Despite these diverse biological roles, the function of NOSIP in lymphocytes—particularly in CD8+ T cells—remains poorly understood. Based on our clinical observation of elevated NOSIP expression in a memory-like CD8+ T cell cluster (), we hypothesized that NOSIP may intrinsically modulate T cell persistence and function.
Here, we examined whether NOSIP overexpression could modulate CD8+ T cell differentiation and enhance their persistence in a chronic viral infection model. We found that NOSIP overexpression promotes the maintenance of a less differentiated CX3CR1neg CD8+ T cell population and reduces apoptosis under chronic antigen stimulation. These findings identify a cell-intrinsic role for NOSIP in sustaining long-term CD8+ T cell responses, with potential implications for future combinatorial immunotherapy strategies.
2 Materials and methods
2.1 ScRNA-seq data processing
Datasets () under the bioproject number PRJDB11756 in the DNA data bank of Japan (DDBJ) used: HLA-A24:hTERT461 or HLA-A24:AFP357 tetramer+ PBMC from immunized HCC patients with tumor-associated antigen vaccines.
The sequenced reads were aligned to the GRCh38 reference transcriptome, and a filtered gene-barcode matrix was generated using CellRanger version 10.0.0. Seurat version 5.4.0 R toolkit () was used for quality control and downstream analysis of scRNA-seq data taking the gene-barcode matrix as input. Specifically, count table data were initially subjected to quality control measures, including the removal of low-quality cells and genes based on read counts (between 800 and 4000) and mitochondrial gene content (<10%). Normalization techniques were then applied to account for variations in sequencing depth and library size, followed by gene selection to filter out low-expressed genes. Subsequently, dimensionality reduction techniques such as principal component analysis (PCA) were employed to identify significant sources of variation, and graph-based clustering algorithm (Louvain algorithm) was utilized to group the single cells into 4 clusters using the top 17 principal components. Doublet removal was performed using DoubletFinder (). Cell types (T cell, Monocyte, NK cell and B cell) were assigned to clusters through marker gene identification and annotation. CD8+ T cells within the T cell cluster were identified based on the following criteria: RNA count for CD8B >0 and CD4 ≤0. After the initial analysis, CD8+ T cells were re-clustered to discover two different subgroups (C0 and C1). The graph-based clustering algorithm (Louvain algorithm) was utilized to group the CD8+ T cells into subclusters with a resolution at 0.1. Differential expression analyses comparing gene expression in subgroups of CD8+ T cells were performed with the FindMarkers function of the Seurat package using the “poisson” method. Genes with P value <0.01 and log2 fold change >1 or <−1 were considered to be differentially expressed.
2.2 Mice
Wild-type (WT) C57BL/6 mice expressing CD45.1 were purchased from the Jackson Laboratory. These mice were crossed with CD45.2+ P14 mice transgenic for a TCR specific to the H-2Db gp33–41 epitope of Lymphocytic Choriomeningitis virus (LCMV) (provided by Dr. Ruka Setoguchi from RIKEN RCAI, Japan) to generate WT P14 mice expressing both CD45.1 and CD45.2, or CD45.1 alone. WT C57BL/6 mice expressing CD45.2 were purchased from Sankyo Labo Service Corporation, Inc. Male mice aged between 6 and 10 weeks were utilized for experiments. All animal experimentation was conducted in accordance with the guidelines approved by the Committee on Animal Experimentation of Kanazawa University (AP21-017-01, AP23-017-01).
2.3 Anesthesia, blood collection, and euthanasia
Mice were anesthetized with inhaled isoflurane delivered via an induction chamber connected to a vaporizer (3.5% isoflurane). For longitudinal monitoring of LCMV-infected mice, retro-orbital blood samples were collected once per week under isoflurane anesthesia. On the terminal day, blood was collected under anesthesia, followed by euthanasia via cervical dislocation under deep anesthesia, and tissues were harvested for downstream analyses.
In the B16-gp33 tumor model, mice were anesthetized with isoflurane for tumor measurements on days 10 and 13. On day 15, tumor size was measured under anesthesia, after which mice were euthanized via cervical dislocation under deep anesthesia. Tumors were then harvested and weighed.
2.4 Virus infection
LCMV Armstrong (Arm) and Clone 13 (Cl13) strains were kindly provided by Dr. E. John Wherry, University of Pennsylvania. Each viral strain was produced and titrated as described (). Mice were infected with LCMV following established protocols. For infection, a dosage of 2×105 plaque-forming units (PFUs) of LCMV Arm was administered intraperitoneally, while 4×106 PFUs of LCMV Cl13 were administered intravenously.
2.5 Plasmid construction, retroviral transduction, and adoptive transfer of CD8+ T cells
The Gateway technology was employed in the construction of the plasmid vector, following established methodologies. The retroviral vectors MSCV (murine stem cell virus)-ccdB-VEX (violet-excited fluorescent protein) and empty vector controls (MSCV-VEX) were donated by Dr. Warren S. Pear, University of Pennsylvania. The cDNA encoding NOSIP (MR204220; OriGene) was cloned into the MSCV-ccdB-VEX plasmid to generate MSCV-NOSIP-VEX.
Retroviral (RV) particles were generated by co-transfecting HEK293T cells with MSCV-based RV expression plasmid and the pCL-Eco packaging plasmid using Lipofectamine 3000 transfection kit (Invitrogen) or jetOptimus DNA Transfection Reagent (PolyPlus). Prior to RV transduction, cell culture plates were coated with anti-mouse CD3ϵ (145-2C11; BioLegend) and anti-mouse CD28 (37.51; BioLegend) antibodies at concentrations of 5 μg/mL and 2 μg/mL, respectively, for a duration of 3 hours to facilitate T cell activation. Subsequently, bulk CD8+ T cells were isolated from the spleens of naïve WT P14 mice using the EasySepTM Mouse CD8+ T cell isolation Kit (STEMCELL Technologies), ensuring purity of the cell population. Isolated CD8+ T cells were then stimulated on the pre-coated plates with 100 U/mL recombinant human IL-2 (Peprotech) for 24 hours to induce activation. Following activation, the CD8+ T cells were transduced with RV in the presence of Polybrene (4 μg/mL; Sigma-Aldrich), employing spin infection at 2000g for 60 minutes at 30 °C.
For non-sorted studies (e.g., Supplementary Figure 2), P14 cells were cultured for 4 hours after retroviral transduction, and 1×105 of CD45.1+ P14 T cells were intravenously transferred into CD45.2+ recipient mice infected with LCMV virus 24 hours before.
For sorted studies, transduced P14 cells were cultured for 24 hours and sorted using the FACSAria Fusion (BD Biosciences) to isolate VEX+ P14 cells, indicating successful transduction. For sorting, the RV-transduced P14 cells were stained with antibodies against CD45.2, CD45.1, and CD8a, and VEX+ P14 cells were purified. Sorted cells were transferred using two designs: For single-transfer experiments, 1×105 empty or NOSIP-overexpressing VEX+ CD45.1+ P14 cells were transferred into individual CD45.2+ recipient mice. For co-transfer experiments, 1×105 P14 cells were adoptively transferred at a 1:1 ratio of empty (CD45.1+CD45.2+) and NOSIP-overexpressing (CD45.1+) cells. In some experiments, the marker combination was reversed (i.e., empty CD45.1+ and NOSIP CD45.1+CD45.2+) to counterbalance any labeling bias. Recipient mice were infected with LCMV Cl13–48 hours prior to transfer.
2.6 Adoptive transfer of CD8+ T cells to uninfected mice
Prior to RV transduction, cell culture plates were coated with anti-mouse CD3ϵ and anti-mouse CD28 antibodies at concentrations of 1 μg/mL and 0.5 μg/mL, respectively, for 3 hours to facilitate T cell activation. Subsequently, bulk CD8+ T cells were isolated from the spleens of naïve WT P14 mice using the EasySepTM Mouse CD8+ T cell isolation Kit. Isolated CD8+ T cells were then stimulated on pre-coated plates with 20 U/mL recombinant human IL-2 for 24 hours to induce activation. Following activation, the CD8+ T cells were transduced with RV in the presence of Polybrene (4 μg/mL), employing spin infection at 2000g for 60 minutes at 30 °C. P14 cells were cultured for 1 day after retroviral transduction, and 4×105 VEX+ P14 cells were adoptively transferred at a 1:1 ratio of empty (CD45.1+CD45.2+) and NOSIP-overexpressing (CD45.1+) cells. After 7 days of adoptive transfer, spleens were harvested and analyzed.
2.7 Flow cytometry
Lymphocytes from peripheral blood, spleen, and liver were analyzed. Spleens and liver tissues were collected and homogenized using a 70 µm cell strainer. Liver immune cells were further purified using Percoll gradient centrifugation. Red blood cells were lysed using the RBC lysis buffer. The antibodies used for flow cytometry are listed in Supplementary Table 1.
For cell surface staining, 2×106 cells were seeded in a 96-well plate and first resuspended in PBS with Fixable Viability Stain 440UV at 4 °C for 30 minutes to exclude dead cells from analysis. After washing, cells were then incubated with anti-mouse CD16/CD32 antibody for 15 minutes at room temperature to minimize non-specific binding. Following Fc blocking, cells were incubated with antibodies in staining buffer (PBS containing 2% fetal bovine serum and EDTA) at 4 °C for 30 minutes. The antibodies used included CD8, CD45.1, CD45.2, CD44, CD62L, CD27, KLRG1, CX3CR1, Ly108, PD-1, TIGIT, TIM3. For intracellular cytokine staining, cells were incubated with 1 μg/mL gp33 peptide (MBL) in the presence of 0.1% brefeldin A and 0.1% monensin (both from eBioscience) at 37 °C for 4 hours. Following surface marker staining, cells were fixed with 2% PFA for 20 minutes at 4 °C to preserve VEX signal, followed by cytokine staining with Cytofix/Cytoperm Fixation/Permeabilization Kit (BD Biosciences). The antibodies used included: CD45.1, CD8, CD45.2, CD62L, CD44, TNF-α, granzyme B, IFN-γ, IL-2. For intranuclear transcription factor staining, cells were fixed with 2% PFA for 20 minutes at 4 °C to preserve VEX signal, fixed and permeabilized with the FoxP3/Transcription Factor Staining Buffer Set (eBioscience). The antibodies used included: CD45.2, CD62L, CD8a, CD45.1, PD-1, CD44, IRF4, EOMES, T-bet, Ki-67, TOX, TCF1/TCF7. Flow cytometry analysis was conducted using a FACSymphony A5 instrument, and data were analyzed using FlowJo v10 software (BD Biosciences).
2.8 Western blotting
Proteins were extracted from RV-transduced P14 CD8+ T cells that had been sorted for VEX+ expression 24 hours after transduction. Following sorting, a fraction of the cells was used for adoptive transfer, and the remaining cells were lysed for protein analysis. Cells were lysed in RIPA buffer (Millipore) supplemented with Protease Inhibitor Cocktail (Roche) and PhosSTOP (Roche). Protein concentration was determined using a BCA assay, and equal amounts of protein were loaded into SDS-PAGE gels with a 5–20% gradient (Fujifilm). The proteins were then transferred to PVDF membrane Immobilon-P (Millipore). Following transfer, membranes were blocked with 5% skim milk (Fujifilm) in TBST (Tris-buffered saline with Tween-20) for 1 hour and then incubated overnight at 4 °C with a primary rabbit polyclonal antibody against NOSIP (Abcam) diluted at 1:1000. β-actin served as an internal control (Cell Signaling Technologies). The next day, membranes were washed and incubated with HRP-linked anti-rabbit IgG secondary antibody (Cell Signaling Technologies) at room temperature for 30 minutes, followed by three additional washes with TBST. Finally, the protein bands were visualized using Clarity Western ECL Substrate kit (Bio-Rad).
2.9 Apoptosis analysis
CD8+ T cells were enriched on day 0 and retrovirally transduced on day 1 as described above. On day 2, transduced P14 cells were sorted to isolate VEX+ cells for downstream assays. Sorted VEX+ P14 cells were then co-cultured at a 1:1 ratio of empty (CD45.1+) and NOSIP-overexpressing (CD45.1+CD45.2+) cells (total 1×105 cells per well) in 6-well plates under two stimulation conditions: (1) transient stimulation: cells were continuously stimulated by 100 U/mL recombinant human IL-2; (2) chronic stimulation: cells were continuous exposure to 1 ng/mL gp33 peptide and 10U/mL recombinant human IL-2. After 2 days of culture (day 4), cells were harvested and apoptosis was assessed using the MEBCYTO Apoptosis Kit (MBL). Cells were first incubated with anti-CD45.1, CD45.2, CX3CR1, and CD8 antibodies at 4 °C for 30 min in the dark, then washed twice with staining buffer. Cells were then resuspended in binding buffer and stained with annexin V-FITC (1:200) and Propidium Iodide (PI; 1:1000) at room temperature for 15 min in the dark. Stained cells were passed through a 70 µm nylon cell strainer and immediately analyzed by flow cytometry using the FACSymphony A5. Apoptotic status was defined as follows: live cells (Annexin V− PI−), early apoptotic cells (Annexin V+ PI−), late apoptotic cells (Annexin V+ PI+), and necrotic cells (Annexin V− PI+).
2.10 B16-gp33 tumor model
The right flank of WT C57BL/6 mice (recipient mice, CD45.2+) was shaved before subcutaneous injection of 2×105 B16-gp33 cells at day 0. On day 3 after tumor cell inoculation, VEX+ P14 cells transduced with empty or NOSIP-overexpressing (2 × 105 cells per mouse, CD45.1+) were sorted and adoptively transferred intravenously. Tumor dimensions (length and width) were measured on days 10, 13, and 15 using a digital caliper, and tumor volume was calculated using the formula: tumor volume = (4/3)×π×(length/2)×(width/2)². Mice were sacrificed on day 15, and the excised tumors were weighed to assess tumor weight.
2.11 Antibody treatment in vivo
PD-L1 blocking assays were performed following LCMV Cl13 infection and adoptive co-transfer of sorted VEX+ P14 cells as described above. Mice received intraperitoneal injections of 200 µg per dose per mouse of anti–PD-L1 antibody (clone 10F.9G2™; Bio X Cell) or rat IgG2b isotype control (LTF-2; Bio X Cell). Repeated injections of anti-PD-L1 or isotype control antibodies were administered starting on day 8 post-infection, then every 3 days for a total of 5 doses. All mice were sacrificed on day 23 post-infection for flow cytometry analysis.
2.12 NO assay
CD8+ T cells were enriched on day 0 and retrovirally transduced on day 1 as described above. On day 2, transduced P14 cells were sorted to isolate VEX+ cells for downstream assays. Sorted VEX+ P14 cells were then co-cultured at a 1:1 ratio of empty (CD45.1+) and NOSIP-overexpressing (CD45.1+CD45.2+) cells (total 1×105 cells per well) in 6-well plates under chronic stimulation conditions (1 ng/mL gp33 peptide and 10 U/mL recombinant human IL-2). After 2 days of culture (day 4), cells were harvested, and intracellular NO was assessed using the Nitric Oxide Assay Kit (Flow Cytometry Orange; Abcam). Cells were first incubated with NO Orange Dye in 37 °C/5% CO2 incubator for 30 min, followed by a wash in PBS. As a positive control, cells were treated with NONOate in Assay Buffer at 37 °C. Then, cells were incubated with anti-CD45.1 and CD45.2 antibodies at 4 °C for 30 min in the dark, then washed twice with staining buffer. Stained cells were passed through a 70 µm nylon cell strainer and immediately analyzed by flow cytometry using the FACSCanto II (BD Biosciences).
2.13 Statistical analysis
All quantitative data are presented as mean ± SEM and were analyzed using GraphPad Prism software version 10.4.1. Paired t-tests and unpaired two-tailed Student’s t-tests with Welch’s correction were used for comparisons between two groups. For multiple group comparisons, either one-way repeated measures ANOVA with Holm–Sidak’s multiple comparisons test or two-way ANOVA with Šídák’s multiple comparisons test was applied, as appropriate. P-values < 0.05 were considered statistically significant. Statistical significance is indicated as follows: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Trends (0.05 < P < 0.1) are indicated where relevant.
3 Results
3.1 NOSIP is enriched in a memory-like CD8+ T cell cluster in peptide vaccine–treated patients
To further investigate the transcriptional context of NOSIP expression, we reanalyzed a previously reported scRNA-seq dataset of tetramer+ PBMC from peptide vaccine-treated HCC patients. PBMC collected at 1 and 5 years after vaccination were combined to enable an integrated analysis of the transcriptional landscape and were subjected to unsupervised clustering and UMAP visualization (Figure 1A). Unsupervised clustering of CD8+ T cells identified two major transcriptional clusters, visualized by UMAP projection (Figure 1B). Feature plot analysis revealed that NOSIP expression was preferentially enriched in cluster 1, which was characterized by high expression of the progenitor-associated transcription factor TCF7 (), whereas cluster 0 exhibited elevated expression of the effector-associated genes GZMB and CX3CR1 (, ) (Figure 1C). Violin plot analysis further confirmed that NOSIP expression was higher in the TCF7+ CD8+ T cell cluster compared with the effector-like cluster (Figure 1D). Consistent with these observations, differential gene expression analysis demonstrated enrichment of NOSIP, TCF7, and SELL () in cluster 1, whereas GZMB and CX3CR1 were enriched in cluster 0 (Supplementary Figure 1). Together, these results place NOSIP within the transcriptional landscape of CD8+ T cell differentiation and indicate that NOSIP expression is associated with a memory-like CD8+ T cell cluster observed in long-term surviving patients following peptide vaccination.
Figure 1
3.2 NOSIP overexpression enhances CD8+ T cell persistence during chronic viral infection
Because NOSIP expression was enriched in a memory-like CD8+ T cell cluster identified by scRNA-seq analysis, we next asked whether NOSIP might influence CD8+ T cell differentiation and persistence in vivo. To evaluate the impact of NOSIP overexpression on these processes, we first performed single-transfer experiments using both the acute lymphocytic choriomeningitis virus (LCMV) Armstrong strain and the chronic LCMV Clone 13 (Cl13) strain—two well-established models for studying CD8+ T cell responses in vivo () (Supplementary Figure 2A). Equal numbers (1 × 105) of CD45.1+ P14 cells (T cell receptor-transgenic cells specific for the LCMV gp33–41 epitope) transduced with retrovirus encoding empty or NOSIP were adoptively transferred into CD45.2+ recipient mice. Overexpression of NOSIP protein was confirmed by Western blot analysis (Supplementary Figure 2B). Following LCMV infection, empty and NOSIP-overexpressing P14 cells were analyzed within the VEX+ gate (Supplementary Figure 2C). In the acute infection model, the relative percentage of VEX+ P14 cells began to decline from day 15 in peripheral blood (PB), with NOSIP-overexpressing P14 cells exhibiting a slightly slower contraction than empty control (Supplementary Figures 2D, E). Notably, in the chronic infection model, NOSIP-overexpressing P14 cells tended to persist at higher frequencies in PB, particularly by days 29 and 42 (Supplementary Figures 2F, G). Consistently, NOSIP-overexpressing P14 cells remained at higher levels in the spleen on day 42 (Supplementary Figures 2H–K). These results indicate that NOSIP overexpression promotes the long-term persistence of CD8+ T cells, with a more pronounced effect under chronic antigen exposure.
To strengthen these observations and enable direct comparison within the same host, we next performed co-transfer experiments using the Cl13 model. CD45.1+CD45.2+ or CD45.1+ P14 cells were retrovirally transduced with an empty or NOSIP encoding vector, respectively (Figure 2A). Equal numbers (5 × 104) of VEX+ P14 cells were sorted on day 2 post-transduction and co-transferred into CD45.2+ recipient mice previously infected with LCMV Cl13 (day 0). The gating strategy for distinguishing empty and NOSIP-overexpressing P14 cells is shown in Supplementary Figure 3. Longitudinal flow cytometric analysis of PB revealed a progressive dominance of NOSIP-overexpressing P14 cells from day 10 to day 31 post-infection (Figure 2B). Although the initial input ratio was 1:1, a clear divergence appeared by day 10, with the ratio shifting to 1:9 by day 31 (Figure 2C). The numbers of VEX+ P14 cells in both groups peaked around day 10 post-infection. Between days 10 and 17, NOSIP-overexpressing P14 cells were stably maintained, whereas empty-transduced cells had already begun to decline. Empty P14 cells decreased sharply and were nearly lost by day 24, while NOSIP-overexpressing P14 cells contracted more gradually and persisted at significantly higher levels on day 31 (Figure 2D). To determine whether this effect extended beyond the circulation, we analyzed the distribution of VEX+ P14 cells in the spleen and liver on day 31. Consistent with the results observed in PB, NOSIP-overexpressing P14 cells exhibited significantly higher frequencies and absolute numbers in both tissues than empty control (Figures 2E–J). Collectively, these results demonstrate that NOSIP overexpression enhances the persistence of antigen-specific CD8+ T cells, particularly during chronic viral infection. In an additional transfer experiment using naïve, uninfected recipients, NOSIP-overexpressing P14 cells were also detected at a higher proportion on day 7 than empty control (Supplementary Figure 4), suggesting that NOSIP overexpression may confer a baseline persistence advantage after transfer. However, because the chronic infection model showed a clearer and more sustained divergence between the two populations, we focused our subsequent analyses on the chronic setting.
Figure 2
3.3 NOSIP overexpression sustains a less differentiated CX3CR1neg CD8+ T cell population
To further characterize the phenotype of NOSIP-overexpressing CD8+ T cells, we performed surface marker profiling to evaluate activation and memory (CD44, CD62L), differentiation (CD27, CX3CR1, Ly108, KLRG1), and exhaustion states (PD-1, TIGIT, TIM-3) (Supplementary Figure 5A) (). In PB, compared with empty control, NOSIP-overexpressing P14 cells consistently exhibited lower frequencies of CX3CR1+ cells and higher frequencies of PD-1+ cells (Supplementary Figure 5B). This difference was further demonstrated by a progressive rightward shift in CX3CR1 intensity among empty control from day 10 to day 31, whereas NOSIP-overexpressing P14 cells maintained lower levels (Supplementary Figure 5C). Given the established relationship between CX3CR1 expression and T cell differentiation (, ), we subdivided VEX+ P14 cells into three CX3CR1-defined subsets: negative (neg), intermediate (int), and high (hi) (Figure 3A). In PB, NOSIP-overexpressing P14 cells consistently displayed higher frequencies of CX3CR1neg cells and lower frequencies of CX3CR1hi cells than empty control at both day 10 and day 31 post-infection (Figures 3B, C). Interestingly, cell number analysis showed that CX3CR1neg cells were also increased in number, whereas CX3CR1hi cells were not reduced in number and were even significantly increased on day 31, despite their lower frequency (Supplementary Figure 5D). A similar pattern was observed in the spleen and liver on day 31 (Figures 3D, E). Correspondingly, subset distributions shifted toward a greater proportion of CX3CR1neg (Figures 3F, G). Moreover, in the single-transfer LCMV Cl13 model, longitudinal PB analysis from day 8 to day 42 revealed that NOSIP-overexpressing P14 cells maintained a larger CX3CR1neg pool, whereas this subset gradually declined in empty control (Supplementary Figures 6A, B). A comparable trend was observed in the spleen on day 42 (Supplementary Figures 6C, D). Collectively, these findings suggest that the long-term persistence advantage conferred by NOSIP overexpression is accompanied by sustained restraint on the transition toward high CX3CR1 expression.
Figure 3
Since previous reports have shown that CX3CR1neg cells exhibit higher PD-1 expression than CX3CR1hi cells (), we next compared PD-1 levels across the three subsets. As expected, PD-1 expression was highest in the CX3CR1neg subset and lowest in CX3CR1hi cells (Supplementary Figure 7A), suggesting that elevated PD-1 levels in the NOSIP group reflect an increased proportion of CX3CR1neg cells. Because previous studies have used Ly108 together with CX3CR1 to further characterize CD8+ T cell differentiation status (, ), we next incorporated this marker combination into our analysis. In splenic VEX+ P14 cells on day 31, NOSIP-overexpressing cells showed an increased frequency of Ly108+CX3CR1− cells, which are considered progenitor-exhausted cells (Supplementary Figures 7B, C). These findings suggest that NOSIP overexpression is associated with the preservation of a less-differentiated Ly108+CX3CR1− population. To further assess the differentiation status of the three CX3CR1-defined subsets (), we examined the expression of established markers—CD27, Ly108 and KLRG1 (Figures 3H, I). In NOSIP-overexpressing cells, the CX3CR1neg subset was enriched for CD27+, Ly108+, and KLRG1− cells, consistent with a less differentiated phenotype, whereas the CX3CR1hi subset was predominantly CD27−, Ly108−, and KLRG1+, indicative of terminal differentiation. Similar trends were observed in empty control (Supplementary Figures 7D, E), reinforcing the association between CX3CR1 expression and differentiation state. Collectively, these results demonstrate that NOSIP overexpression delays the transition of CX3CR1neg cells toward the CX3CR1hi terminal phenotype, thereby preserving a less differentiated pool that may contribute to its persistence advantage.
3.4 NOSIP overexpression enhances the antitumor activity of CD8+ T cells
To further evaluate the functional capacity of NOSIP-overexpressing CD8+ T cells, we first analyzed key transcription factors associated with T cell fate. The expression levels of TCF1, TOX, and EOMES did not differ significantly between empty and NOSIP-overexpressing P14 cells (Supplementary Figures 8A, B). In contrast, Ki-67 expression showed a trend toward higher levels in NOSIP-overexpressing P14 cells, suggesting a potentially enhanced proliferative capacity within the less differentiated pool. Because NOSIP overexpression preferentially sustains the CX3CR1neg pool—which has been reported to exhibit lower cytotoxicity than CX3CR1hi cells ()—we next assessed cytokine production. Upon ex vivo stimulation of splenic P14 cells on day 31 post-infection in the LCMV Cl13 model, NOSIP-overexpressing cells produced levels of IFN-γ, TNF-α, IL-2, and granzyme B comparable to those of empty control (Supplementary Figure 8C), indicating that NOSIP overexpression did not impair cytokine production at this stage.
Nevertheless, less differentiated CD8+ T cells have been shown to mediate superior tumor control in vivo compared with more differentiated counterparts (). We next examined whether NOSIP-overexpressing cells could elicit effective antitumor responses in vivo. A B16-gp33 tumor model was employed, in which mice received adoptive transfers of either empty or NOSIP-overexpressing P14 cells (Figure 4A). Mice receiving NOSIP-overexpressing P14 cells exhibited significantly delayed tumor progression relative to empty control (Figure 4B). At the experimental endpoint, tumor weights were also significantly lower in the NOSIP-overexpressing group (Figure 4C), indicating that NOSIP-overexpressing CD8+ T cells retain the ability to mediate potent antitumor responses.
Figure 4
3.5 NOSIP overexpression enhances CD8+ T cell responsiveness to PD-L1 blockade
Given that ICB preferentially reinvigorates progenitor-like exhausted CD8+ T cells—a less differentiated subset with self-renewal capacity (, )—we hypothesized that NOSIP overexpression would augment the proliferative expansion of CD8+ T cells in response to PD-L1 blockade. To test this hypothesis, LCMV Cl13-infected mice that had received co-transfers of empty and NOSIP-overexpressing P14 cells were treated with repeated administrations of anti-PD-L1 or isotype control antibodies (Figure 5A). By day 23 post-infection, NOSIP-overexpressing cells had become dominant in PB, increasing from 50% at transfer to approximately 75% in both isotype- and anti-PD-L1 antibody-treated groups (Supplementary Figures 9A, B). PD-L1 blockade significantly increased the numbers of NOSIP-overexpressing P14 cells while exerting minimal effects on the empty-transduced group (Figure 5B), suggesting that NOSIP overexpression enhances CD8+ T cell sensitivity to PD-L1 blockade.
Figure 5
Moreover, previous studies have demonstrated that progenitor-like populations preferentially expand following PD-L1 blockade (). To determine whether a similar effect occurred in our system, we analyzed CX3CR1 expression in PB on day 23. Because of the limited number of empty P14 cells, subset analysis was restricted to the NOSIP-overexpressing P14 cells (Figure 5C). Anti-PD-L1 antibody treatment increased the CX3CR1neg fraction while significantly reducing the proportion of CX3CR1hi cells, indicating a shift toward a less differentiated phenotype (Figure 5D). Although the frequency of CX3CR1neg cells showed only a modest upward trend, cell number analysis revealed a significant expansion of the CX3CR1neg population, with CX3CR1int cells also exhibiting an increasing trend and CX3CR1hi cells maintained at comparable levels under anti-PD-L1 antibody treatment (Supplementary Figure 9C). Collectively, these results demonstrate that NOSIP overexpression enhances the proliferative response of CD8+ T cells to PD-L1 blockade and promotes the maintenance of a less differentiated CX3CR1neg pool associated with long-term persistence.
3.6 NOSIP overexpression promotes CD8+ T cell persistence by limiting differentiation into an apoptosis-prone CX3CR1hi state
Because NOSIP-overexpressing CD8+ T cells exhibited enhanced antitumor activity and preferential maintenance of less differentiated phenotypes, we next sought to elucidate the mechanism by which NOSIP promotes CD8+ T cell persistence under chronic stimulation. Previous studies have demonstrated that less differentiated CD8+ T cell subsets possess greater self-renewal capacity and are less susceptible to apoptosis, whereas terminally differentiated cells are more prone to cell death (). To investigate this, we performed in vitro co-cultures of equal numbers of sorted empty and NOSIP-overexpressing P14 cells under transient or chronic stimulation and analyzed their persistence (Figure 6A; Supplementary Figure 10A). The frequency of NOSIP-overexpressing P14 cells exhibited a modest increase compared with empty control under chronic stimulation on day 4, although the difference was not statistically significant (Figure 6B; Supplementary Figure 10B). Annexin V and PI staining revealed that NOSIP-overexpressing P14 cells consistently displayed lower frequencies of late apoptotic and necrotic cells, together with a higher proportion of live cells under chronic stimulation conditions (Figures 6C, D), with a similar but less pronounced trend observed under transient stimulation in vitro. These findings indicate that NOSIP overexpression is associated with reduced cell death, particularly under chronic stimulation.
Figure 6
Consistent with the findings from the LCMV infection model, NOSIP-overexpressing P14 cells maintained a higher proportion of CX3CR1neg cells and a lower frequency of CX3CR1hi cells in vitro under both stimulation conditions (Figures 6E, F). To determine whether the reduced cell death observed in NOSIP-overexpressing P14 cells was associated with the differentiation state, we next assessed annexin V/PI categories within CX3CR1 subsets under chronic stimulation (Figure 6G). The CX3CR1hi subset exhibited significantly higher frequencies of apoptotic and necrotic cells, whereas the CX3CR1neg subset was more resistant to cell death (Figure 6H). A similar pattern was observed under transient stimulation (Supplementary Figures 10C, D). Furthermore, intracellular cytokine staining demonstrated that NOSIP-overexpressing P14 cells produced levels of TNF-α, IFN-γ, and IL-2 comparable to those of empty control under both stimulation conditions (Supplementary Figure 11), consistent with the in vivo findings. Collectively, these results suggest that NOSIP overexpression is associated with enhanced CD8+ T cell persistence, accompanied by preservation of an apoptosis-resistant, less differentiated subset and reduced progression toward an apoptosis-prone terminal phenotype.
4 Discussion
Our study indicates that NOSIP overexpression prolongs CD8+ T cell persistence during chronic infection, accompanied by a restrained transition from a less differentiated CX3CR1neg subset to a terminally differentiated CX3CR1hi state. Consistent with this, NOSIP overexpression maintains a CX3CR1neg pool that is more resistant to apoptosis. Notably, combining PD-L1 blockade with NOSIP overexpression resulted in a further enrichment of this persistent subset, suggesting that NOSIP may modulate differentiation trajectories to support sustained CD8+ T cell persistence. Interestingly, a recent scRNA-seq study of tumor-draining lymph nodes from patients with colorectal cancer reported that NOSIP expression progressively decreases along metallothionein-positive CD8+ T cell differentiation trajectories, further supporting our observation that NOSIP is associated with the maintenance of less-differentiated T cell states (). To the best of our knowledge, our study provides the first functional evidence implicating NOSIP in regulating CD8+ T cell fate in a murine model of chronic infection, thereby extending its biological relevance to adaptive immunity.
Depletion of NOSIP has been reported to increase apoptosis in neural progenitor cells and neurons (). NOSIP binds to NOS isoforms and suppresses their enzymatic activity, thereby reducing NO production. It may also function as an E3 ubiquitin ligase that inhibits PP2A signaling (, ). Both NO and PP2A pathways are well-established regulators of cell fate (–). A recent study demonstrated that NOSIP overexpression in hepatocellular carcinoma models suppressed apoptosis while increasing intracellular NO levels (). Notably, NO has been shown to exert diverse effects and may not directly induce apoptosis in T cells (), leaving the link between NOSIP-mediated NO/PP2A signaling and lymphocyte survival unresolved. In our model, NOSIP overexpression preferentially maintained the CX3CR1neg subset, which exhibited significantly lower rates of apoptosis than the terminally differentiated CX3CR1hi subset. This observation is consistent with previous findings that less differentiated CD8+ T cells are intrinsically more resistant to apoptosis and possess greater self-renewal capacity (, ). However, in the in vitro model, the differences between the empty control and NOSIP-overexpressing P14 cells were less pronounced than those observed in vivo. These findings suggest that, in addition to apoptosis resistance, multiple factors may contribute to the phenotype observed in vivo. Although NOSIP overexpression did not enhance cytokine production under chronic stimulation, this pattern aligns with the functional profile of progenitor-like cells, which primarily serve as a proliferative reservoir rather than immediate cytolytic effectors (). The precise mechanism by which NOSIP restrains the progression from a less differentiated state to terminal differentiation remains to be elucidated. A recent study revealed that PP2A, as an interactor of the scaffold protein AMBRA1, participates in the differentiation program of CD8+ T cells through activation of transcription factor EB (). Moreover, NO has also been reported to influence T cell development (48). In our preliminary analysis under chronic stimulation in vitro, NO Orange Dye staining showed a slight reduction in intracellular NO level in NOSIP-overexpressing P14 cells compared with empty control. However, the overall signal was minimal relative to the positive control, suggesting that only minimal NO production occurs under this condition (Supplementary Figure 12). Thus, while this result is consistent with the possibility that NOS-related signaling may be affected by NOSIP in this context, further studies will be required to clarify the underlying mechanism.
Although NOSIP has been primarily investigated in the context of NOS and PP2A regulation, emerging proteomic evidence suggests that its interaction network may be broader. For instance, a proteomic screen identified NOSIP as a potential cargo of the nuclear transport receptor importin 13 (), implying that NOSIP may associate with molecular partners beyond the currently recognized NOS and PP2A, thereby expanding its functional repertoire. However, the NOSIP interactome in immune cells remains largely uncharacterized. Large-scale interactome and proteomic mapping platforms could be utilized to systematically identify NOSIP-binding partners and signaling networks in T cells (49, 50). Such analyses will be essential to elucidate how NOSIP modulates CD8+ T cell fate.
ICB exerts its therapeutic efficacy primarily through progenitor-like CD8+ T cells possessing self-renewal capacity (51–53). PD-L1 blockade has been reported to preferentially expand the CX3CR1neg subset, whereas the CX3CR1int and CX3CR1hi subsets exhibit minimal changes (). In our chronic infection model, PD-L1 inhibition markedly enhanced the expansion of NOSIP-overexpressing P14 cells in PB but had limited effects on empty control. Consistent with previous studies (, ), this expansion was accompanied by a phenotypic shift toward a less differentiated state, characterized by an increased frequency of CX3CR1neg and a decreased frequency of CX3CR1hi fractions. Importantly, this reduced frequency of CX3CR1hi cells should not be interpreted as depletion of the effector-like compartment. In both the PD-L1 blockade setting and the co-transfer LCMV Cl13 infection model, the numbers of NOSIP-overexpressing CX3CR1hi P14 cells were maintained at comparable levels or were even increased compared with empty control, whereas CX3CR1neg cells were expanded, and CX3CR1int cells showed a trend toward an increase. Thus, the observed phenotypic shift primarily reflects expansion of the less differentiated CX3CR1neg population rather than loss of the CX3CR1hi subset. These findings suggest that NOSIP maintains a progenitor-like CD8+ T cell compartment that remains preferentially responsive to PD-L1 blockade. Collectively, our results indicate that NOSIP not only sustains a persistent CD8+ T cell reservoir but also potentiates the therapeutic effects of immune checkpoint blockade.
While our study provides novel insights into the role of NOSIP in sustaining a less differentiated, apoptosis-resistant CD8+ T cell population, several limitations should be acknowledged. First, NOSIP has been reported to interact with eNOS and nNOS isoforms, which are primarily expressed in neuronal and vascular tissues (, ). However, its relevance to inducible nitric oxide synthase (iNOS)—the predominant isoform in immune cells (48, 54)—remains unclear. Whether NOSIP directly modulates iNOS activity in CD8+ T cells and how this interaction affects NO/PP2A signaling cascades requires further mechanistic investigation. Second, we did not directly assess the functional consequences of combining NOSIP overexpression with PD-L1 blockade, as the present study was specifically designed to delineate differentiation and persistence mechanisms. Future work using infection and tumor models will be necessary to address this limitation. Third, our study relied on NOSIP overexpression systems, which could induce nonphysiological alterations in cell differentiation. Therefore, we cannot conclude that NOSIP inherently contributes to the maintenance of less differentiated populations under physiological conditions. Definitive evidence will require NOSIP loss-of-function studies, such as knockout models. Finally, although our findings were established in a murine chronic infection model, their relevance to human immunity remains to be confirmed. Future studies employing human samples and clinically relevant tumor models will be essential to determine whether targeting NOSIP can preserve progenitor-like T cell reservoirs and enhance the durability of checkpoint-based immunotherapies. Collectively, these findings position NOSIP modulation as a promising strategy to sustain long-term T cell immunity in chronic infection and cancer.
Statements
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Ethics statement
The animal study was approved by the committee on Animal Experimentation of Kanazawa University (AP21-017-01, AP23-017-01). The study was conducted in accordance with the local legislation and institutional requirements.
Author contributions
SL: Conceptualization, Methodology, Writing – review & editing, Data curation, Formal analysis, Investigation, Visualization, Writing – original draft. TT: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing – review & editing. YS: Investigation, Writing – review & editing, Funding acquisition. SF: Investigation, Methodology, Writing – review & editing. YT: Investigation, Methodology, Writing – review & editing. MU: Investigation, Writing – review & editing. HN: Conceptualization, Methodology, Writing – review & editing. JK: Resources, Writing – review & editing. MKo: Resources, Writing – review & editing. EM: Conceptualization, Funding acquisition, Supervision, Writing – review & editing. MKu: Conceptualization, Funding acquisition, Methodology, Resources, Supervision, Writing – review & editing.
Funding
The author(s) declared that financial support was received for this work and/or its publication. This study was supported by Grant-in-Aid for Scientific Research (B: 23K24109 and 24K02435) and Grant-in-Aid for Research Activity Start-up (25K23807) from the Japan Society for the Promotion of Science, the Japan Agency for Medical Research and Development under Grant Number JP24fk0310509, JP25fk0310531, and JP26fk0310531, and the Kanazawa University SAKIGAKE project.
Acknowledgments
This work was the result of using research equipment shared in MEXT Project for promoting public utilization of advanced research infrastructure (Program for supporting construction of core facilities) Grant Number JPMXS0440300024. We thank Editage [https://www.editage.com] for editing and reviewing this manuscript for English language.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The authors declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Correction note
A correction has been made to this article. Details can be found at: 10.3389/fimmu.2026.1933173.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fimmu.2026.1755657/full#supplementary-material
References
1
McLaneLMAbdel-HakeemMSWherryEJ. Cd8 t cell exhaustion during chronic viral infection and cancer. Annu Rev Immunol. (2019) 37:457–95. doi: 10.1146/annurev-immunol-041015-055318
2
WherryEJ. T cell exhaustion. Nat Immunol. (2011) 12:492–9. doi: 10.1038/ni.2035
3
MillerBCSenDRAl AbosyRBiKVirkudYVLaFleurMWet al. Subsets of exhausted cd8+ t cells differentially mediate tumor control and respond to checkpoint blockade. Nat Immunol. (2019) 20:326–36. doi: 10.1038/s41590-019-0312-6
4
SiddiquiISchaeubleKChennupatiVFuertes MarracoSACalderon-CopeteSPais FerreiraDet al. Intratumoral tcf1+pd-1+cd8+ t cells with stem-like properties promote tumor control in response to vaccination and checkpoint blockade immunotherapy. Immunity. (2019) 50:195–211.e10. doi: 10.1016/j.immuni.2018.12.021
5
PaukenKEWherryEJ. Overcoming t cell exhaustion in infection and cancer. Trends Immunol. (2015) 36:265–76. doi: 10.1016/j.it.2015.02.008
6
SenDRKaminskiJBarnitzRAKurachiMGerdemannUYatesKBet al. The epigenetic landscape of t cell exhaustion. Science. (2016) 354:1165–9. doi: 10.1126/science.aae0491
7
SunQHongZZhangCWangLHanZMaD. Immune checkpoint therapy for solid tumours: clinical dilemmas and future trends. Signal Transduct Target Ther. (2023) 8:320. doi: 10.1038/s41392-023-01522-4
8
WeiSCDuffyCRAllisonJP. Fundamental mechanisms of immune checkpoint blockade therapy. Cancer Discov. (2018) 8:1069–86. doi: 10.1158/2159-8290.CD-18-0367
9
SharmaPGoswamiSRaychaudhuriDSiddiquiBASinghPNagarajanAet al. Immune checkpoint therapy—current perspectives and future directions. Cell. (2023) 186:1652–69. doi: 10.1016/j.cell.2023.03.006
10
KangYWChoiDMoonDLeeKJOhYYangJet al. Enabling immune checkpoint blockade efficacy in t-lymphopenia by restoring cd8 t cell dynamics with il-7 cytokine therapy. Front Immunol. (2024) 15:1477171. doi: 10.3389/fimmu.2024.1477171
11
GuerraLBonettiLBrennerD. Metabolic modulation of immunity: a new concept in cancer immunotherapy. Cell Rep. (2020) 32:107848. doi: 10.1016/j.celrep.2020.107848
12
LiaoLXuHZhaoYZhengX. Metabolic interventions combined with ctla-4 and pd-1/pd-l1 blockade for the treatment of tumors: mechanisms and strategies. Front Med. (2023) 17:805–22. doi: 10.1007/s11684-023-1025-7
13
MizukoshiENakagawaHTamaiTKitaharaMFushimiKNioKet al. Peptide vaccine-treated, long-term surviving cancer patients harbor self-renewing tumor-specific cd8+ t cells. Nat Commun. (2022) 13:3123. doi: 10.1038/s41467-022-30861-z
14
DedioJKönigPWohlfartPSchroederCKummerWMüller-EsterlW. Nosip, a novel modulator of endothelial nitric oxide synthase activity. FASEB J. (2001) 15:79–89. doi: 10.1096/fj.00-0078com
15
DreyerJSchleicherMTappeASchillingKKunerTKusumawidijajaGet al. Nitric oxide synthase (nos)-interacting protein interacts with neuronal nos and regulates its distribution and activity. J Neurosci. (2004) 24:10454–65. doi: 10.1523/JNEUROSCI.2265-04.2004
16
PörschkeMRodríguez-GonzálezIParfentevIUrlaubHKehlenbachRH. Transportin 1 is a major nuclear import receptor of the nitric oxide synthase interacting protein. J Biol Chem. (2023) 299:102932. doi: 10.1016/j.jbc.2023.102932
17
SchleicherMBrundinFGrossSMüller-EsterlWOessS. Cell cycle-regulated inactivation of endothelial no synthase through nosip-dependent targeting to the cytoskeleton. Mol Cell Biol. (2005) 25:8251–8. doi: 10.1128/MCB.25.18.8251-8258.2005
18
KönigPDedioJOessSPapadakisTFischerAMüller-EsterlWet al. Nosip and its interacting protein, enos, in the rat trachea and lung. J Histochem Cytochem. (2005) 53:155–64. doi: 10.1369/jhc.4A6453.2005
19
KummerWKönigPDedioJMüller-EsterlW. Distribution of the novel enos–interacting protein nosip in the liver, pancreas, and gastrointestinal tract of the rat. Gastroenterology. (2002) 123:314–24. doi: 10.1053/gast.2002.34212
20
SadafSNagarkotiSAwasthiDSinghAKSrivastavaRNKumarSet al. Nnos induction and nosip interaction impact granulopoiesis and neutrophil differentiation by modulating nitric oxide generation. Biochim Biophys Acta Mol Cell Res. (2021) 1868:119018. doi: 10.1016/j.bbamcr.2021.119018
21
YuXZhongYZhuZWuTShenAHuangY. Increased expression of nitric oxide synthase interacting protein (nosip) following traumatic spinal cord injury in rats. J Mol Histol. (2012) 43:661–8. doi: 10.1007/s10735-012-9460-9
22
HoffmeisterMPrelleCKüchlerPKovacevicIMoserMMüller-EsterlWet al. The ubiquitin e3 ligase nosip modulates protein phosphatase 2a activity in craniofacial development. PloS One. (2014) 9:e116150. doi: 10.1371/journal.pone.0116150
23
HoffmeisterMKriegJEhrkeASeigfriedFAWischmannLDietmannPet al. Developmental neurogenesis in mouse and xenopus is impaired in the absence of nosip. Dev Biol. (2017) 429:200–12. doi: 10.1016/j.ydbio.2017.06.026
24
StuartTButlerAHoffmanPHafemeisterCPapalexiEMauckWMet al. Comprehensive integration of single-cell data. Cell. (2019) 177:1888–1902.e21. doi: 10.1016/j.cell.2019.05.031
25
McGinnisCSMurrowLMGartnerZJ. Doubletfinder: doublet detection in single-cell rna sequencing data using artificial nearest neighbors. Cell Syst. (2019) 8:329–337.e4. doi: 10.1016/j.cels.2019.03.003
26
KurachiMBarnitzRAYosefNOdorizziPMDiIorioMALemieuxMEet al. The transcription factor batf operates as an essential differentiation checkpoint in early effector cd8+ t cells. Nat Immunol. (2014) 15:373–83. doi: 10.1038/ni.2834
27
BeltraJCAbdel-HakeemMSManneSZhangZHuangHKurachiMet al. Stat5 opposes the transcription factor tox and rewires exhausted cd8+ t cells toward durable effector-like states during chronic antigen exposure. Immunity. (2023) 56:2699–2718.e11. doi: 10.1016/j.immuni.2023.11.005
28
HudsonWHGensheimerJHashimotoMWielandAValanparambilRMLiPet al. Proliferating transitory t cells with an effector-like transcriptional signature emerge from pd-1+ stem-like cd8+ t cells during chronic infection. Immunity. (2019) 51:1043–1058.e4. doi: 10.1016/j.immuni.2019.11.002
29
HenningANRoychoudhuriRRestifoNP. Epigenetic control of cd8+ t cell differentiation. Nat Rev Immunol. (2018) 18:340–56. doi: 10.1038/nri.2017.146
30
BarberDLWherryEJMasopustDZhuBAllisonJPSharpeAHet al. Restoring function in exhausted cd8 t cells during chronic viral infection. Nature. (2006) 439:682–7. doi: 10.1038/nature04444
31
RauschLKalliesA. Molecular mechanisms governing cd8 t cell differentiation and checkpoint inhibitor response in cancer. Annu Rev Immunol. (2025) 43:515–43. doi: 10.1146/annurev-immunol-082223-044122
32
HamannIUnterwalderNCardonaAEMeiselCZippFRansohoffRMet al. Analyses of phenotypic and functional characteristics of cx3cr1-expressing natural killer cells. Immunology. (2011) 133:62–73. doi: 10.1111/j.1365-2567.2011.03409.x
33
ZwijnenburgAJPokharelJVarnaitėRZhengWHofferEShrykiIet al. Graded expression of the chemokine receptor cx3cr1 marks differentiation states of human and murine t cells and enables cross-species interpretation. Immunity. (2023) 56:1955–1974.e10. doi: 10.1016/j.immuni.2023.06.025
34
YamauchiTHokiTObaTSaitoHAttwoodKSabelMSet al. Cx3cr1–cd8+ t cells are critical in antitumor efficacy but functionally suppressed in the tumor microenvironment. JCI Insight. (2020) 5:e133920. doi: 10.1172/jci.insight.133920
35
HeydenLRauschLShannonMHDryburghLMoreiraMLFrolovAet al. Transcriptional regulator satb1 limits cd8+ t cell population expansion and effector differentiation in chronic infection and cancer. Nat Immunol. (2025) 26:2312–27. doi: 10.1038/s41590-025-02316-2
36
RicaRWaldherrMMiyakodaEKutschatAPSchüleinMZhangJet al. Hdac1 controls the generation and maintenance of effector-like cd8+ t cells during chronic viral infection. J Exp Med. (2025) 222:e20240829. doi: 10.1084/jem.20240829
37
GordonCLLeeLNSwadlingLHutchingsCZinserMHightonAJet al. Induction and maintenance of cx3cr1-intermediate peripheral memory cd8+ t cells by persistent viruses and vaccines. Cell Rep. (2018) 23:768–82. doi: 10.1016/j.celrep.2018.03.074
38
GattinoniLKlebanoffCAPalmerDCWrzesinskiCKerstannKYuZet al. Acquisition of full effector function in vitro paradoxically impairs the in vivo antitumor efficacy of adoptively transferred cd8+ t cells. J Clin Invest. (2005) 115:1616–26. doi: 10.1172/JCI24480
39
NeubertENDeRogatisJMLewisSAViramontesKMOrtegaPHenriquezMLet al. Hmgb2 regulates the differentiation and stemness of exhausted cd8+ t cells during chronic viral infection and cancer. Nat Commun. (2023) 14:5631. doi: 10.1038/s41467-023-41352-0
40
ImSJHashimotoMGernerMYLeeJKissickHTBurgerMCet al. Defining cd8+ t cells that provide the proliferative burst after pd-1 therapy. Nature. (2016) 537:417–21. doi: 10.1038/nature19330
41
JiangTGuanXWeiRLvCZanLZhangZet al. Cd8+mt+ effector t cells from enlarged tumor-draining lymph nodes enhanced anti-tumor immunity and improved prognosis in colorectal cancer. Cell Rep. (2025) 44:116625. doi: 10.1016/j.celrep.2025.116625
42
KimMELeeJS. Advances in the regulation of inflammatory mediators in nitric oxide synthase: implications for disease modulation and therapeutic approaches. Int J Mol Sci. (2025) 26:1204. doi: 10.3390/ijms26031204
43
TaylorELMegsonILHaslettCRossiAG. Nitric oxide: a key regulator of myeloid inflammatory cell apoptosis. Cell Death Differ. (2003) 10:418–30. doi: 10.1038/sj.cdd.4401152
44
Van HoofCGorisJ. Phosphatases in apoptosis: to be or not to be, pp2a is in the heart of the question. Biochim Biophys Acta. (2003) 1640:97–104. doi: 10.1016/S0167-4889(03)00029-6
45
GaoJYangDHuangZPanXCaoRLianCet al. Nosip is a potential therapeutic target in hepatocellular carcinoma cells. iScience. (2023) 26:107353. doi: 10.1016/j.isci.2023.107353
46
de MartiniCCde AndradeJTde AlmeidaSKMOliveira SilvaKLde Rezende EugenioFdos SantosPSPet al. Cellular apoptosis and nitric oxide production in pbmc and spleen from dogs with visceral leishmaniasis. Comp Immunol Microbiol Infect Dis. (2018) 57:1–7. doi: 10.1016/j.cimid.2018.01.003
47
MiglioreLCianfanelliVZevoliniFGesualdoMMarzuoliLPatrussiLet al. An ambra1, ulk1 and pp2a regulatory network regulates cytotoxic t cell differentiation via tfeb activation. Sci Rep. (2024) 14:31838. doi: 10.1038/s41598-024-82957-9
48
García-OrtizASerradorJM. Nitric oxide signaling in t cell-mediated immunity. Trends Mol Med. (2018) 24:412–27. doi: 10.1016/j.molmed.2018.02.002
49
GöösHKinnunenMSalokasKTanZLiuXYadavLet al. Human transcription factor protein interaction networks. Nat Commun. (2022) 13:766. doi: 10.1038/s41467-022-28341-5
50
LiXWangWWangJMalovannayaAXiYLiWet al. Proteomic analyses reveal distinct chromatin-associated and soluble transcription factor complexes. Mol Syst Biol. (2015) 11:775. doi: 10.15252/msb.20145504
51
BeltraJ-CManneSAbdel-HakeemMSKurachiMGilesJRChenZet al. Developmental relationships of four exhausted CD8+ T cell subsets reveals underlying transcriptional and epigenetic landscape control mechanisms. Immunity. (2020) 52:825–841.e8. doi: 10.1016/j.immuni.2020.04.014
52
BlankCUHainingWNHeldWHoganPGKalliesALugliEet al. Defining ‘T cell exhaustion.’. Nat Rev Immunol. (2019) 19:665–74. doi: 10.1038/s41577-019-0221-9
53
LiuZZhangYMaNYangYMaYWangFet al. Progenitor-like exhausted SPRY1+CD8+ T cells potentiate responsiveness to neoadjuvant PD-1 blockade in esophageal squamous cell carcinoma. Cancer Cell. (2023) 41:1852–1870.e9. doi: 10.1016/j.ccell.2023.09.011
54
BogdanC. Nitric oxide synthase in innate and adaptive immunity: an update. Trends Immunol. (2015) 36:161–78. doi: 10.1016/j.it.2015.01.003
Summary
Keywords
CD8+ T cells, chronic antigen stimulation, CX3CR1 subset differentiation, NOSIP, PD-L1 blockade responsiveness
Citation
Li S, Tamai T, Shinzawa Y, Fujisawa S, Tanabe Y, Utsunomiya M, Nakagawa H, Kurachi J, Koura M, Mizukoshi E and Kurachi M (2026) NOSIP overexpression promotes long-term persistence of CD8+ T cells during chronic infection. Front. Immunol. 17:1755657. doi: 10.3389/fimmu.2026.1755657
Received
27 November 2025
Revised
31 May 2026
Accepted
05 June 2026
Published
22 June 2026
Corrected
21 July 2026
Volume
17 - 2026
Edited by
Takanori So, University of Toyama, Japan
Reviewed by
Toshiro Hirai, Osaka University, Japan
Shinya Sakaguchi, Medical University of Vienna, Austria
Updates
Copyright
© 2026 Li, Tamai, Shinzawa, Fujisawa, Tanabe, Utsunomiya, Nakagawa, Kurachi, Koura, Mizukoshi and Kurachi.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Eishiro Mizukoshi, eishirom@m-kanazawa.jp; Makoto Kurachi, kurachi@med.kanazawa-u.ac.jp
†These authors have contributed equally to this work and share first authorship
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.