Article Text

Original research
Inhibitory Fcγ receptor deletion enhances CD8 T cell stemness increasing anti-PD-1 therapy responsiveness against glioblastoma
  1. Keun Bon Ku1,2,
  2. Chae Won Kim3,4,
  3. Yumin Kim3,
  4. Byeong Hoon Kang1,3,
  5. Jeongwoo La1,3,
  6. In Kang1,3,
  7. Won Hyung Park1,3,
  8. Stephen Ahn5,
  9. Sung Ki Lee6 and
  10. Heung Kyu Lee3,7
  1. 1Graduate School of Medical Science and Engineering, Korea Advanced Institute of Science and Technology (KAIST), Daejeon 34141, Republic of Korea
  2. 2Center for Infectious Disease Vaccine and Diagnosis Innovation, Korea Research Institute of Chemical Technology, Daejeon 34114, Republic of Korea
  3. 3Laboratory of Host Defenses, Department of Biological Sciences, KAIST, Daejeon 34141, Republic of Korea
  4. 4Life Science Institute, KAIST, Daejeon 34141, Republic of Korea
  5. 5Department of Neurosurgery, Seoul St. Mary’s Hospital, College of Medicine, The Catholic University of Korea, Seoul 06591, Republic of Korea
  6. 6Department of Obstetrics and Gynecology, College of Medicine, Myunggok Medical Research Center, Konyang University, Daejeon 35365, Republic of Korea
  7. 7KAIST Institute of Health Science and Technology, KAIST, Daejeon 34141, Republic of Korea
  1. Correspondence to Professor Heung Kyu Lee; heungkyu.lee{at}kaist.ac.kr

Abstract

Background Certain cancers present challenges for treatment because they are resistant to immune checkpoint blockade (ICB), attributed to low tumor mutational burden and the absence of T cell-inflamed features. Among these, glioblastoma (GBM) is notoriously resistant to ICB. To overcome this resistance, the identification of T cells with heightened stemness marked by T-cell factor 1 (TCF1) expression has gained attention. Several studies have explored ways to preserve stem-like T cells and prevent terminal exhaustion. In this study, we investigate a target that triggers stem-like properties in CD8 T cells to enhance the response to ICB in a murine GBM model.

Methods Using Fcgr2b−/− mice and a murine GL261 GBM model, we confirmed the efficacy of anti-programmed cell death protein-1 (PD-1) immunotherapy, observing improved survival. Analysis of immune cells using fluorescence-activated cell sorting and single-cell RNA sequencing delineated distinct subsets of tumor-infiltrating CD8 T cells in Fcgr2b−/− mice. The crucial role of the stem-like feature in the response to anti-PD-1 treatment for reinvigorating CD8 T cells was analyzed. Adoptive transfer of OT-I cells into OVA-expressing GL261 models and CD8 T cell depletion in Fcgr2b−/− mice confirmed the significance of Fcgr2b−/− CD8 T cells in enhancing the antitumor response. Last, S1P1 inhibitor treatment confirmed that the main source of tumor antigen-specific Fcgr2b−/− CD8 T cells is the tumor-draining lymph nodes (TdLNs).

Results In a murine GBM model, anti-PD-1 monotherapy and single-Fc fragment of IgG receptor IIb (FcγRIIB) deletion exhibit limited efficacy. However, their combination substantially improves survival by enhancing cytotoxicity and proliferative capacity in tumor-infiltrating Fcgr2b−/− CD8 T cells. The improved response to anti-PD-1 treatment is associated with the tumor-specific memory T cells (Ttsms) exhibiting high stemness characteristics within the tumor microenvironment (TME). Ttsms in the TdLN thrives in a protective environment, maintaining stem-like characteristics and serving as a secure source for tumor infiltration. This underscores the significance of FcγRIIB ablation in triggering Ttsms and enhancing ICB therapy against GBM.

Conclusions Deletion of FcγRIIB on CD8 T cells leads to the generation of a Ttsms, which is localized in TdLN and protected from the immunosuppressive TME. Incorporating these highly stemness-equipped Ttsms enhances the response to anti-PD-1 therapy in immune-suppressed brain tumors.

  • Central Nervous System Cancer
  • Immune Checkpoint Inhibitor
  • T Cell
  • Tumor Infiltrating Lymphocyte - TIL
  • Tumor Microenvironment - TME

Data availability statement

Data are available upon reasonable request.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See http://creativecommons.org/licenses/by-nc/4.0/.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

WHAT IS ALREADY KNOWN ON THIS TOPIC

  • Glioblastomas (GBM) is a widely recognized tumor type that exhibits high resistance to immune checkpoint blockade (ICB), mainly attributed to the significant involvement of cytotoxic CD8 T cells in the mechanisms of resistance. CD8 T cells that infiltrate GBM are commonly characterized as being fully exhausted, with a diminished presence of potentially responsive cell subsets. Hence, it is imperative to comprehend and focus on CD8 T cells in GBM.

WHAT THIS STUDY ADDS

  • An inhibitory receptor, Fc fragment of IgG receptor IIb (FcγRIIB), which has been recently discovered to be present in CD8 T cells, impacting the effector functions of CD8 T cells. The deletion of FcγRIIB resulted in an improved therapeutic response to anti-programmed cell death protein-1 (PD-1) treatment in a murine GBM model. Moreover, stem-like tumor-specific CD8 T cells were found to have originated from the tumor-draining lymph node (TdLN). The elimination of FcγRIIB resulted in the augmentation of stem-like tumor-specific CD8 T cells in the TdLN, consequently enhancing the response to anti-PD-1 treatment.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

  • A novel target has been identified for enhancing the efficacy of anti-PD-1 therapy, specifically in traditionally ICB-resistant tumor types such as GBM. The underlying mechanism involves the robust induction of highly stem-like CD8 T cells with specific residency. We propose the synergistic effect of anti-PD-1 and modulation of inhibitory Fcγ receptor.

Introduction

Glioblastoma (GBM), an aggressive brain tumor in the central nervous system (CNS), has a median survival of only 14.6 months.1 2 Despite the efficacy of immune checkpoint blockade (ICB) therapy in treating different types of cancer, clinical trials targeting GBM encounter substantial resistance.3–5 This challenge is primarily associated with the unique tumor microenvironment (TME) of GBM, which is influenced by the characteristics of the CNS, such as the blood-brain barrier and diminished antigen presentation to cytotoxic T lymphocytes (CTLs).6 Consequently, GBM features an immune-suppressed microenvironment, which limits the presence of tumor-infiltrating CTLs. Even when infiltration occurs, T cells encounter various immunosuppressive factors from the TME that result in a dysfunctional state of “exhaustion”.7 8 Exhausted T cells in the GBM demonstrate heightened levels of inhibitory receptors, reduced functionality, and proliferative capacity due to brain TME.

Recent research has underscored the diversity within exhausted T cell populations, highlighting progenitor-exhausted T cells (Tpexs) as promising candidates for ICB.9–11 Tpexs, characterized by stem-like traits with T-cell factor 1 (TCF1) expression, represent a promising target for ICB effectiveness.12 13 Chronic exposure to antigens results in the development of terminally exhausted T cells (Ttexs) and reduces the effectiveness of ICB. GBM is closely associated with elevated Ttex levels. To overcome ICB resistance, there is a growing interest in preserving stem-like features and identifying distinct T cell subsets, such as tumor-specific memory T cells (Ttsms).14–17

The Fc fragment of IgG receptor IIb (FcγRIIB) has gained attention as an inhibitory receptor on CD8 T cells.18–20 The expression of FcγRIIB in circulating memory CD8 T subsets was identified through comprehensive investigations by Morris et al18 and Buquicchio et al.21 Operating as a low-affinity inhibitory Fcγ receptor, FcγRIIB binds to the Fc region of IgG and engages with ligands such as fibrinogen-like protein 2 and C-reactive protein.22–24 The intracellular domain of FcγRIIB contains an immunoreceptor tyrosine-based inhibitory motif (ITIM), that is, phosphorylated by Lyn. The recruitment of SH2 domain-containing inositol phosphatase initiates downstream signaling pathways.19 25 Recognized as an inhibitory receptor on CD8 T cells, the precise role of FcγRIIB in tumor immune responses requires further investigation.

In this study, we targeted FcγRIIB as an inhibitory receptor to modulate the function of CD8 T cells in the highly immunosuppressive brain TME. Through the investigation, we found that the removal of FcγRIIB enhances the response of CD8 T cells to anti-programmed cell death protein-1 (PD-1) treatment in murine GBM. Fcgr2b−/− mice maintained an enhanced stemness-like phenotype, especially within the Ttsms subset and predominantly resided in tumor-draining lymph node (TdLNs). This study suggests that targeting FcγRIIB has the potential to enhance the effectiveness of ICB therapy to overcome resistance in patients with GBM.

Results

FcγRIIB ablation enhances anti-PD-1-mediated GBM rejection

We used the Cistrom public database to investigate inhibitory receptors in human cancers and their prognostic significance.26 Specifically, we focused on prominent inhibitory receptors associated with immunotherapies, including PD-1, cytotoxic T-lymphocyte antigen 4 (CTLA-4), lymphocyte activation gene-3 (LAG-3), T-cell immunoglobulin and mucin-domain containing-3 (TIM-3), T-cell immunoglobulin and ITIM domain (TIGIT), and the recently identified FcγRIIB. Our analysis revealed a significant correlation between the expression of Pdcd1 and Fcgr2b and the risk of GBM, suggesting their potential as prognostic markers for poor outcomes in patients with GBM (figure 1A,B). Conversely, CTLA-4, LAG-3, TIM-3, and TIGIT did not show a significant correlation with GBM prognosis (online supplemental figure 1A). These findings underscore the importance of PD-1 and FcγRIIB in predicting outcomes in GBM compared with other inhibitory receptors.

Supplemental material

Figure 1

Specific effectiveness of anti-PD-1 treatment in Fcgr2b ablation mice with murine glioblastoma. (A–B) Volcano plots depict HRs as z-scores and −log10 (adjusted p values). The selection of cancer types was based on their ranking within the top 20 adjusted p values among a total of 41 cancer types. (C–E) Murine GBM was induced through intracranial administration of 1×105 GL261 glioma cells on day 0, followed by the assessment of survival analyzed using the log-rank (Mantel-Cox) test. (C) The survival rates of wild-type (WT) mice treated with either anti-PD-1 (n=10, median survival (MS)=21 days) or isotype antibody (n=10, MS=22 days). The mice received antibody injections on days 10, 13, and 16, with a dose of 200 µg/mouse. (D) The survival rates of WT (n=10, MS=23.5 days) and Fcgr2b−/− mice (n=10, MS=25 days). (E) The survival rates of WT mice treated with either anti-PD-1 (n=10, MS=22 days) or isotype antibody (n=10, MS=21.5 days) and Fcgr2b−/− mice treated with either anti-PD-1 (n=10, MS=35 days, long-term survival=20%) or isotype antibody (n=10, MS=23.5 days). (F) Mice that survived long-term were subjected to rechallenge with homologous GL261 cells in the hemisphere contralateral to the primary injected tumor on day 90, and their survival rate is shown (n=6). (G–H) Tumor-bearing brain tissues were collected on day 20, and the frequency (CD45.2GFP+) and density of glioma cells were analyzed. All results were pooled from or representative of two independent experiments with n=4–6 mice/group (C–H). Differences in survival were analyzed by the log-rank (Mantel-Cox) test (C–F). Error bars represent the SEM, and p values were calculated by unpaired Student’s t-tests (H). GBM, glioblastoma; GFP, green fluorescent protein; PD-1, programmed cell death protein-1.

In the orthotopic GBM mouse model, we investigated the relationship between PD-1 and FcγRIIB expression and tumor progression. However, administering anti-PD-1 treatment as monotherapy after injecting GL261 cells did not extend median survival (figure 1C), indicating resistance to ICB in murine and human cases of GBM. Moreover, we assessed the impact of FcγRIIB deletion on mouse GBM progression but found that deleting FcγRIIB did not significantly delay tumor progression (figure 1D).

Remarkably, the administration of anti-PD-1 treatment to Fcgr2b−/− mice led to a substantial improvement in survival outcomes, resulting in a 20% long-term survival rate and a 13-day extension of median survival (figure 1E). Furthermore, its effectiveness against CT2A cells, another GBM model,27 was validated (online supplemental figure 1B) and anti-PD-1-treated Fcgr2b−/− mice showed reduced clinical symptoms, such as weight loss (online supplemental figure 1C). Mice that survived 90 days after GL261 cell injection were rechallenged with autologous GL261 cells on the contralateral side of the primary tumor site. Long-term surviving mice displayed complete resistance to rechallenge with the identical tumor model, indicating the establishment of a lasting tumor-specific memory response facilitated by anti-PD-1 therapy in Fcgr2b−/− mice (figure 1F). Quantification of the GL261 cell mass was analyzed using green fluorescent protein (GFP)-expressing GL261 cells to validate the survival advantages and tumor regression in Fcgr2b−/− mice (figure 1G,H, online supplemental figure 1D). The simultaneous absence of FcγRIIB and PD-1 blocking enhanced tumor regression, suggesting a synergistic effect between anti-PD-1 therapy and FcγRIIB deletion in boosting the immune response against GBM. Moreover, the observed response demonstrated enduring memory for tumor antigens.

Tumor-infiltrating FcγRIIB-deleted CD8 T cells enhance the antitumor response when combined with PD-1 blockade

Fcgr2b−/− mice, on receiving anti-PD-1 antibody treatment, exhibited enhanced survival rates (figure 1E). Tumor-infiltrating immune cells were analyzed to evaluate the impact of the therapy on the TME. Remarkably, antibody injection—whether anti-PD-1 or isotype—did not alter the number of tumor-infiltrating immune cells (online supplemental figure 2A). Importantly, the tumor-infiltrating CD8 T cell population, which is targeted by anti-PD-1 blockade, showed an increase in Fcgr2b−/− mice compared with that in mice treated with the isotype control (online supplemental figure 2B). An increase in the quantity of CD8 T cells was observed in wild-type (WT) mice; however, statistical significance was not reached (online supplemental figure 2B). Additionally, FcγRIIB expression was detected on tumor-infiltrating CD8 T cells (online supplemental figure 2C). FcγRIIB+ CD8 T cells demonstrated increased expression of cytokines, specifically interferon-γ (IFN-γ) and tumor necrosis factor-α (TNF-α), along with inhibitory receptors, exhibiting varying levels of TCF1 and thymocyte selection-associated high mobility group box protein (TOX) and TIM-3 (online supplemental figure 2D–I). The analysis of memory subsets (effector and central) of tumor-infiltrating CD8 T cells was comparable between WT and Fcgr2b−/− mice, regardless of PD-1 blocking (online supplemental figure 2J). To confirm the binding specificity of the anti-PD-1 antibody to CD8 T cells, we followed the experimental procedures detailed by Floris Dammeijer et al.28 We found that tumor-infiltrating CD8 T cells were found to specifically bind to the anti-PD-1 antibody. In the isotype antibody groups (including WT and Fcgr2b−/− mice), only about 2% of counterstained antibodies bound non-specifically to CD8 T cells (figure 2A). Conversely, with anti-PD-1 treatment, approximately 80% of tumor-infiltrating CD8 T cells exhibited positive staining in both WT and Fcgr2b−/− mice (figure 2A), indicating selective targeting of PD-1 on tumor-infiltrating CD8 T cells.

Figure 2

Improved antitumor reaction observed following PD-1 blockade in Fc fragment of IgG receptor IIb-deficient tumor-infiltrating cytotoxic T lymphocytes. (A–F) Wild-type (WT) and Fcgr2b−/− mice received intracranial injections of GL261 cells. Mice were intraperitoneally injected with either anti-PD-1 or isotype antibodies on days 10, 13, and 16 following tumor cell injection. 20 days after inoculation with GL261 cells, individual cells were isolated, and immune cells from the tumor were examined using flow cytometry. (A) Representative flow plots and a bar graph illustrate the quantification of anti-PD-1 antibody binding on CD8 T cells in tumor tissues. (B–C) Representative histogram and bar graph illustrating CD44 (B) and pAKT (C) expression levels in tumor-infiltrating CD8 T cells following anti-PD-1 antibody treatment. (D–G) The functional alterations of tumor-infiltrating CD8 T cells were assessed by quantifying IFN-γ, TNF-α, CD107a, Ki-67, and TCF1. Representative flow cytometry plots and bar graphs depict the percentages of IFN-γ+ and TNF-α+ (D), IFN-γ+ TNF-α+ CD107a+ (E), Ki-67+ (F), and TCF1+ (G) CD8 T cells. (H) Single-cell RNA sequencing analysis was conducted on day 20. Among the immune cells, tumor-infiltrating T cells were analyzed for cytotoxicity and proliferative capacity scores using the AddModuleScore function within Seurat. (I) Cytotoxic and proliferative scores of tumor-infiltrating CD8 T cells were analyzed and compared between WT and Fcgr2b−/− mice. (J) Schematic diagram of αCD8 antibody (150 µg/mouse) and anti-PD-1 (200 µg/mouse) treatment in GL261 cell-injected mice. (K) Differences in survival were assessed using the log-rank (Mantel-Cox) test. Groups of GL261-bearing WT and Fcgr2b−/− mice were treated with anti-PD-1 therapy with or without CD8 T cell depletion as follows: WT treated with IgG and αPD-1 (n=11, MS: 25 days), WT treated with αCD8 and αPD-1 (n=10, MS: 25 days), Fcgr2b−/− treated with IgG and αPD-1 (n=10, MS: 35.5 days), Fcgr2b−/− treated with αCD8 and αPD-1 (28 days; n=11). The results (A, B) are representative of two independent experiments with n=4–6 mice per group. Error bars represent the SEM, and p values were calculated using two-tailed, unpaired Student’s t-tests (A–I). The survival rate was analyzed by the log-rank (Mantel-Cox) test (K). All results were pooled from or are representative of two or three experiments with n=6–11 mice/group (A), n=4–6 mice/group (B), n=8–10 mice/group (D, E), n=4–6 mice/group (F, G).APC, Allophycocyanin; APCCy7, Allophycocyanin-Cyanine7; Cy5, Cyanine 5; gMFI, geometric mean fluorescence intensity; IFN, interferon; MS, median survival; PerCP-Cy5.5, Peridinin Chlorophyll-A Protein-Cyanine5.5; PD-1, programmed cell death protein-1; pAKT, phosphorylated protein kinase B; TCF1, T-cell factor 1; TNF, tumor necrosis factor; UMAP, Uniform Manifold Approximation and Projection.

We further examined the functional changes of CD8 T cells on interaction with anti-PD-1 antibody and their role in enhancing antitumor efficacy. The effector cell (CD44+ CD62L) population was elevated among CD8 T cells that interacted with the anti-PD-1 antibody compared with those that did not engage in this interaction (online supplemental figure 3A). Moreover, PD-1-bound Fcgr2b−/− CD8 T cells exhibited elevated CD44 expression compared with WT CD8 T cells (figure 2B). This finding indicates that functional variances may be present in WT and Fcgr2b−/− CD8 T cells, despite comparable levels of anti-PD-1 antibody binding. Thus, we also analyzed the intracellular signaling factor phosphorylated protein kinase B (pAKT), which is associated with T cell activation and survival.29 In WT CD8 T cells, pAKT expression remained consistent regardless of PD-1 blockade. Interestingly, in Fcgr2b−/− CD8 T cells, pAKT levels were significantly elevated when PD-1 was blocked compared with non-blocked cells within the same mouse (online supplemental figure 3B). Furthermore, the overall expression of pAKT in tumor-infiltrating CD8 T cells was significantly higher in anti-PD-1 antibody-treated Fcgr2b−/− CD8 T cells than in WT CD8 T cells (figure 2C). Next, we investigated the levels of the effector cytokines IFN-γ and TNF-α in tumor-infiltrating CD8 T cells. Our findings indicated that mice lacking FcγRIIB with PD-1 blocking displayed significant increases in IFN-γ and TNF-α secretion (figure 2D). To validate the existence of polyfunctional CD8 T cells within the TME, we demonstrated that the frequency of triple-positive cells for IFN-γ, TNF-α, and CD107a was markedly higher in anti-PD-1 antibody-treated Fcgr2b−/− CD8 T cells (figure 2E). Furthermore, our data indicated a significant increase in Ki-67+ CD8 T cells in Fcgr2b−/− mice compared with those in WT mice treated with anti-PD-1 antibodies (figure 2F). Additionally, TCF1 was found to be abundantly expressed in Fcgr2b−/− CD8 T cells (figure 2G). These results elucidate the significantly increased proliferative activity in response to anti-PD-1 therapy, along with a notable enhancement in their antitumor effector capabilities. To further support these findings, we conducted an in vitro killing assay using a GL261 cell system expressing ovalbumin and mCherry (GL261-OVA-mCherry). We isolated CD8 T cells from the brains of mice injected with GL261-OVA and co-cultured them with GL261-OVA-mCherry as target cells. The results demonstrated that PD-1-blocked Fcgr2b−/− CD8 T cells were significantly more effective at killing the target cells compared with WT CD8 T cells. This outcome further confirms the enhanced antitumor response observed in PD-1-blocked Fcgr2b−/− CD8 T cells (online supplemental figure 3C). Conversely, there were no observed alterations in the antitumor functions of tumor-infiltrating CD4 T cells. The ratio of anti-PD-1 binding to CD44 expression in CD4 T cells exhibited similarities between WT and Fcgr2b−/− mice (online supplemental figure 3D,E). Additionally, the expression levels of IFN-γ, TNF-α, and CD107a were comparable between the two groups (online supplemental figure 3F,G).

In addition to examining effector molecules, we conducted single-cell RNA sequencing (scRNA-seq) analysis to comprehensively explore the CD8 T cell response to anti-PD-1 treatment (online supplemental figure 4A). Transcriptome analysis revealed CD8 T cells as the predominant tumor-infiltrating lymphocytes, characterized by elevated expression levels of cytotoxic molecules essential for tumor cell killing (figure 2H, online supplemental figure 4B). In Fcgr2b−/− mice, CD8 T cells exhibited significantly higher general cytotoxic scores compared with WT mice (figure 2I, online supplemental figure 4C). Moreover, Fcgr2b−/− CD8 T cells demonstrated higher proliferation rates than those in WT mice (figure 2I, online supplemental figure 4D). Comprehensive analysis of transcriptome data revealed enhanced functionality of Fcgr2b−/− CD8 T cells, consistent with the results of fluorescence-activated cell sorting (FACS) analysis. To confirm the association between increased survival outcomes and these enhanced CD8 T cells, we depleted CD8 T cells before anti-PD-1 treatment (figure 2J). Subsequently, the survival enhancement effect was reduced following anti-PD-1 treatment due to CD8 T cell depletion (figure 2K). This finding confirms the direct involvement of Fcgr2b−/− CD8 T cells in improving survival rates after anti-PD-1 treatment.

Fcgr2b−/− CD8 T cells exhibit enrichment of stem-like signatures in the TME

Our findings demonstrate a positive correlation between improved survival rates and an increased tumor-infiltrating CD8 T cell response. This response is characterized by enhanced effector function and proliferation in anti-PD-1-treated Fcgr2b−/− mice. We further examined CD8 T cells using scRNA-seq data from tumor-bearing brains, identifying 12 distinct clusters (online supplemental figure 5A). Based on trajectory information and marker gene expression (online supplemental figure 5B,C), 12 subsets were annotated as Ttsm, Tpex, Ttex, intermediate, and proliferating CD8 T cells (figure 3A).30 31 Additionally, the annotated subclusters were confirmed by representative gene signature scores (online supplemental figure 5D).

Figure 3

Tumor-infiltrating Fcgr2b−/− CD8 T cells reveal a relatively heightened stemness signature characterized by elevated Ttsm subsets. (A–F) Single-cell RNA sequencing analysis was performed on tumor-infiltrating CD8 T cells from anti-PD-1-treated wild-type (WT) and Fcgr2b−/− mice. (A) Uniform Manifold Approximation and Projection (UMAP) visualization of single-cell RNA sequencing data of tumor-infiltrating CD8 T cells, separated by genotype and colored based on cluster classification. The classifications are as follows: tumor-specific memory (Ttsm: cluster 4, 5), progenitor-exhausted (Tpex: cluster 7, 8, 10), cytotoxic (cytotoxic: cluster 1, 2), terminally exhausted (Ttex: cluster 0), intermediate (intermediate: cluster 6, 9), and proliferating (proliferating: cluster 3, 11) CD8 T cells. (B) The enumeration of identity counts for CD8 T cell subsets is presented by genotype (WT+αPD-1: Ttsm 168, Tpex 125, cytotoxic 229, Ttex 186, intermediate 58, proliferating 110; Fcgr2b−/−+αPD-1: Ttsm 349, Tpex 196, cytotoxic 522, Ttex 362, intermediate 153, proliferating 266). (C) Bar plots illustrate the log2 fold change in the identity counts of subsets between WT and Fcgr2b−/− mice. (D) Density plots of CD8 T cells illustrate gene set signatures for the “Stem-like”, “Exhausted progenitor1”, and “Tumor antigen-specific memory” categories (online supplemental table S1). (E) Pseudotime analysis of the correlation between the expression levels of Tcf7 and Tox and cell progression. (F) Gene Set Enrichment Analysis was conducted to compare the transcriptomic signatures of Ttsms between WT and Fcgr2b−/− mice. The Biocarta and Hallmark gene set databases were used, and pathways with nominal p values<0.05 were reported along with normalized enrichment scores (NES). (G–H) Following GL261 cell injection with either anti-PD-1 or isotype antibodies on days 10, 13, and 16 following tumor cell injection, subsets of tumor-infiltrating CD8 T cells were analyzed at day 20 based on PD-1 and TCF1 expression (n=10–11 per group). (G) Flow plots of annotated CD8 T cells classified as Ttex (PD-1+TCF1), Tpex (PD-1+TCF1+), and Ttsm (PD-1TCF1+). (H) A representative histogram and quantification of TOX expression levels on Ttexs, Tpexs, and Ttsms based on genotypes (n=6 per group). (I) Quantification of global TOX expression levels on tumor-infiltrating CD8 T cells based on genotypes (n=6 per group). Error bars represent the SEM, and p values were calculated using unpaired Student’s t-tests (G, I) or two-way analysis of variance (H). gMFI, geometric mean fluorescence intensity; PD-1, programmed cell death protein-1; TCF1, T-cell factor 1.

Pseudotime analysis was performed to identify specific subsets and to describe the cell fate of tumor-infiltrating CD8 T cells. Among them, we highlighted cluster 5 as the primary subset with increased expression of gene sets associated with stemness, including Sell, Ccr7, S1pr1, and Tcf7 (online supplemental figure 5B,C).

In Fcgr2b−/− mice, the count of CD8 T cells significantly exceeded that in WT mice, with 1848 and 876 CD8 T cells, respectively (figure 3B). Overall, CD8 T cell subsets showed a significant increase in Fcgr2b−/− mice, particularly in intermediate and proliferating clusters, followed by cytotoxic subsets (figure 3C). Additionally, Ttsms exhibited a remarkable 2.0-fold increase in Fcgr2b−/− mice (figure 3C). Ttsms, characterized by low exhaustion-related inhibitory receptor expression and abundant stemness-related gene sets, exhibited higher Tcf7 expression levels compared with Tpexs, aligning with their crucial role in anti-PD-1 treatment response. It exhibited gene set expression patterns similar to the Tstem-like (Connolly et al),15 Texprog1 (Beltra et al),32 and Ttsm signatures (Huang et al)14 (figure 3D). Notably, these studies highlight the significance of Ttsms subsets of CD8 T cells characterized by TCF1 expression and basal inhibitory receptor levels, indicating a primordial stage within the TME. Additionally, Ttsms, originating in TdLNs, possess tumor antigen-specific memory.

Pseudotime analysis revealed decreased Tcf7 expression in both WT and Fcgr2b−/− CD8 T cells as they progressed in the TME. The expression of TOX relatively decreased in Fcgr2b−/− CD8 T cells in particular, analysis of specific CD8 T cell subsets—such as cytotoxic, intermediate, and proliferating subsets—revealed that the Fcgr2b−/− group exhibited lower TOX expression compared with their WT counterparts. These subsets, derived from Ttsm and Tpex cells, are typically more prone to exhaustion (figure 3E, online supplemental figure 5E,F). This finding led us to investigate the functional changes mediated by stem-like T cells in WT and Fcgr2b−/− mice. Gene Set Enrichment Analysis (GSEA) revealed increased T cell receptor (TCR)-linked intracellular pathway activation within the Fcgr2b−/− Ttsm cluster (figure 3F). This enrichment was positively correlated with the upregulation of the phosphoinositide 3-kinase (PI3K)–AKT–mammalian target of rapamycin (mTOR), nuclear factor of activated T cells, and myelocytomatosis oncogene (MYC) signaling pathways (figure 3F). Notably, the PI3K–AKT–mTOR pathway was enhanced in Fcgr2b−/− CD8 T cells, as indicated by protein levels of pAKT in flow cytometry analysis (figure 2C). Despite enhanced distal TCR signaling in Fcgr2b−/− Ttsms, sustained oxidative phosphorylation metabolism was also noted (figure 3F) indicating reduced susceptibility to metabolic stress. The enrichment of the interleukin (IL)-12 pathway in GSEA emphasizes its role in preventing CD8 T cells from exhaustion.33 These findings suggest that Fcgr2b−/− CD8 T cell-derived Ttsms represent the most potential subsets for the efficacy of anti-PD-1 therapy.

Subsequently, using FACS, we confirmed the presence of distinct CD8 T cell subsets in tumor-bearing brains, classified as Ttsms, Tpexs, and Ttexs based on TCF1 and PD-1 expression levels.34–36 Remarkably, Fcgr2b−/− mice exhibited a higher frequency of Ttsms compared with WT mice, regardless of anti-PD-1 antibody treatment (figure 3G). Although PD-1 blockade reduced the mean frequency of Ttexs and increased Tpexs in both WT and Fcgr2b−/− CD8 T cells, these changes did not reach statistical significance. Furthermore, there was no significant difference in the frequencies of Ttex and Tpex subsets between WT and Fcgr2b−/− CD8 T cells (figure 3G). Ttsms and Tpexs represent CD8 T cell subsets that are particularly responsive to anti-PD-1 therapies. The evaluation of the exhaustion-related transcription factor TOX showed significantly lower levels in Ttsms compared with Tpexs and Ttexs (figure 3H). As previously mentioned, a significant increase in Ttsms among tumor-infiltrating Fcgr2b−/− CD8 T cells suggests reduced susceptibility to exhaustion. This diminished potential for exhaustion is further supported by the lower overall expression of TOX in Fcgr2b−/− CD8 T cells compared with WT CD8 T cells (figure 3I), which correlates with an increased frequency of the Ttsm subset in Fcgr2b−/− CD8 T cells. Therefore, the enhancement in response to anti-PD-1 therapy is critically influenced by the presence of the Ttsm subset among tumor-infiltrating CD8 T cells.

In addition to the significant alterations observed in CD8 T cell subsets, we also investigated the impact of FcγRIIB deletion on microglia, macrophages, and monocytes, which function as tumor-associated macrophages (TAMs) and play crucial roles in the TME of GBM.37 Transcriptome analysis revealed that Fcgr2b−/− TAMs exhibited upregulated levels of major histocompatibility complex (MHC) class II-related genes (histocompatibility 2, class II antigen A, beta 1, histocompatibility 2, class II antigen E beta, and cluster of differentiation 74), as well as the interferon-inducible gene Ifi202b and the chemokine CC motif chemokine ligand 8 (online supplemental figure 6A). These upregulations indicate a shift towards a pro-inflammatory, M1-like phenotype. The GSEA demonstrated consistent results, highlighting the enrichment of pro-inflammatory pathways, including TNF-α signaling and the IFN-γ response (online supplemental figure 6B). These findings suggest that the enhanced expression of TNF-α and IFN-γ cytokines produced by Fcgr2b−/− CD8 T cells promotes the polarization of TAMs towards an antitumor M1 state. Protein-level analysis confirmed that only the combination of FcγRIIB deletion and anti-PD-1 treatment significantly increased MHC II expression in TAMs (online supplemental figure 6C,D), which is consistent with the results of the transcriptome analysis.

Overall, the deletion of FcγRIIB on CD8 T cells fundamentally alters T cell characteristics, leading to an increase in the Ttsm subset, which is equipped with enhanced stemness within the brain TME and contributes to the overall improvement of the antitumor response.

Requirement of the Ttsm signature in FcγRIIB-deleted CD8 T cells for the anti-PD-1 therapy effect

To elucidate the intrinsic attributes of CD8 T cells enhancing the antitumor response following PD-1 blockade in Fcgr2b−/− mice, we employed an adoptive cell transfer model using GL261-OVA (SIINFEKL) cells and Fcgr2b−/− OT-I cells. Congenic naive WT and Fcgr2b−/− OT-I cells, sorted to achieve over 95% purity (online supplemental figure 7A), were combined equally and then co-transferred to WT recipient mice (online supplemental figure 7B). On the next day, GL261-OVA cells were injected into recipient mice. 10 days after the anti-PD-1 therapy, an analysis of tumor-infiltrating OT-I cells was performed (figure 4A). The frequencies of transferred OT-I cells in tumor-bearing brains were similar between the WT and Fcgr2b−/− OT-I cells (figure 4B, online supplemental figure 7C). Fcgr2b−/− OT-I cells exhibited a notable reduction in triple-positive cells expressing exhaustion markers CD39, PD-1, and TIM-3 compared with WT OT-I cells (figure 4C). The tumor-infiltrating Fcgr2b−/− OT-I cells exhibited 1.4% triple positivity for exhaustion marker expression, while 13% of WT OT-I cells demonstrated triple-positive exhaustion makers among the inhibitory receptor-expressing OT-I cells (figure 4D). Additionally, a decrease was observed in the population of PD-1 and 2B4 double-positive cells among tumor-infiltrating Fcgr2b−/− OT-I cells (online supplemental figure 8A). An analysis of Ki-67 expression confirmed a significant upregulation in Fcgr2b−/− OT-I cells, indicating increased proliferative activity (figure 4E). Moreover, Fcgr2b−/− OT-I cells exhibited higher expression levels of TCF1 compared with WT OT-I cells (figure 4F). These findings indicate that the absence of FcγRIIB in tumor-infiltrating OT-I cells enhances resistance against the suppressive effects of the TME. This is accompanied by improved proliferation and stemness, thereby enhancing the antitumor response against GBM.

Figure 4

Necessity of Fc fragment of IgG receptor IIb deletion in CD8 T cells with a Ttsm signature for the efficacy of anti-PD-1 treatment. Naive OT-I cells from CD45.1+ CD45.2+ wild-type (WT) and CD45.1+ CD45.2 Fcgr2b−/− mice were isolated and subsequently co-transferred in equal numbers into CD45.1 CD45.2+ WT recipient mice. On the subsequent day, GL261-OVA (SIINFEKL) cells were administered via injection, followed by the application of anti-PD-1 treatments. Tumor-infiltrating OT-I cells were analyzed using flow cytometry 10 days after the final treatment. (A) Scheme outlining the experimental set-up for (B–G). (B) Representative flow plot and a bar graph illustrate the quantification of tumor-infiltrating OT-I cells (n=4). The frequency of OT-I cells was analyzed after gating on singlet, live CD45.1+ cells. (C) Representative flow plots and a bar graph illustrate the quantification of PD-1+ CD39+ TIM-3+ WT or Fcgr2b−/− OT-I cells. (D) Pie chart of co-expression of inhibitory receptors (PD-1+, CD39+, and TIM-3+) on tumor-infiltrating WT or Fcgr2b−/− OT-I cells. (E−F) Representative histogram and geometric mean fluorescence intensity (gMFI) of Ki-67 (E) and TCF1 (F) in tumor-infiltrating WT or Fcgr2b−/− OT-I cells. (G) Subset analysis of tumor-infiltrating WT or Fcgr2b−/− OT-I cells. OT-I cells were annotated as Ttex (PD-1+TCF1), Tpex (PD-1+TCF1+), and Ttsm (PD-1TCF1+), and their frequencies are depicted. p values were calculated using paired Student’s t-tests (B, C, E, F, and G). PD-1, programmed cell death protein-1; TCF1, T-cell factor 1; TIM-3, T-cell immunoglobulin and mucin-domain containing-3; Tpex, progenitor-exhausted T cell; Ttex, terminally exhausted T cell; Ttsm, tumor-specific memory T cells.

Tumor-infiltrating OT-I cells were classified as Tpexs, Ttexs, and Ttsms based on TCF1 and PD-1 expression. Fcgr2b−/− OT-I cells exhibited comparable proportions of Ttexs and Tpexs subsets to WT OT-I cells, but a significantly higher frequency of Ttsms (figure 4G). These results show that a subset of tumor-infiltrating OT-I cells and the expression levels of inhibitory receptors were replicated in polyclonal CD8 T cells in WT and Fcgr2b−/− mice following GL261 cell injection (figure 3G, online supplemental figure 8B,C). Consequently, Ttsm subsets of both Fcgr2b−/− OT-I and polyclonal CD8 T cells, characterized by low PD-1 and high TCF1 expression, were enriched in the brain TME. Overall, these findings suggest that the absence of FcγRIIB intrinsically alters the characteristics of tumor-infiltrating CD8 T cells.

This modification involves a decrease in inhibitory receptor expression and an increase in TCF1 expression, similar to T cells identified in TdLNs. Recent studies have provided evidence supporting the presence of Ttsms primarily located in TdLNs.14 15 32 38 These cells are considered precursors of the tumor antigen-specific CD8 T cells that persistently migrate to the tumor site and act as the primary responders to anti-PD-1 treatment.

TdLNs are a necessary niche for preserving and providing Ttsms

The absence of FcγRIIB increased Ttsm subset with high expression levels of TCF1 in the TME. These cellular characteristics closely resembled T cells found in TdLNs. Therefore, we investigated the deep cervical lymph nodes as the brain-TdLNs. Intracranial injection of GL261 cells resulted in similar numbers of immune cells in the brains of both WT and Fcgr2b−/− mice (online supplemental figure 2A). However, TdLNs in Fcgr2b−/− mice exhibited a significant increase in the number of immune cells (figure 5A). Additionally, Fcgr2b−/− mice demonstrated a significant increase in both the total count (figure 5B) and frequency of CD8 T cells (online supplemental figure 9A). Moreover, a significant increase was observed in the overall number of Ki-67+ TCF1+ CD8 T cells in the TdLNs of Fcgr2b−/− mice (figure 5C). Analysis of exhaustion-related markers revealed that a significant portion of CD8 T cells in TdLNs exhibited basal levels of TOX and inhibitory receptors (PD-1, CD39, LAG-3, 2B4) expression (figure 5D, online supplemental figure 9B). The CD8 T cell signatures in the TdLNs closely resembled those in tumor-infiltrating Fcgr2b−/− CD8 T cells with lower expression of inhibitory receptors. To further investigate the impact of FcγRIIB deletion on CD8 T cell maintenance in TdLNs, we used the OT-I adoptive cell transfer model. Remarkably, over 80% of Fcgr2b−/− OT-I cells resided in the TdLNs, in contrast to approximately 20% of WT OT-I cells among the adoptively co-transferred population (figure 5E). Furthermore, in TdLNs, the majority of Fcgr2b−/− OT-I cells exhibited high TCF1 expression along with low PD-1 expression (figure 5F).

Figure 5

TdLNs are essential for maintaining responsive tumor-specific CD8 T cells targeted by anti-PD-1 treatment. (A–D) Analysis of deep cervical tumor-draining lymph nodes (TdLNs) in GL261cell-injected wild-type (WT) and Fcgr2b−/− mice under anti-PD-1 antibody treatment. (A) Total immune cell and (B) CD8 T cell numbers in TdLNs were pooled from two experiments (n=4–6 mice/group). (C) Ki67+ TCF1+ CD8 T cell numbers in TdLNs (n=5–6 mice/group). (D) Representative flow plot of TCF1 and TOX expression of CD8 T cells in TdLNs. (E–F) Adoptive co-transfer of equal numbers of OT-I cells and analysis of TdLNs (n=4). (E) Representative flow plots and a bar graph illustrate the quantification of TdLN-residing OT-I cells. The frequency of OT-I cells was analyzed after gating on singlet, live CD45.1+ cells. (F) Subset analysis of TdLN-residing Fcgr2b−/− OT-I cells. OT-I cells were annotated as Tpex (PD-1+TCF1+) and Ttsm (PD-1TCF1+), and their frequencies are depicted. (G–H) To examine the effect of TdLN origin on CD8 T cells, an S1P1 inhibitor, FTY720 (FTY) was administered. (G) Schematic diagram of the experimental set-up for FTY720 treatment and (H) the survival rates were analyzed (WT+αPD-1+PBS: n=5, MS 23, WT+αPD-1+FTY: n=5, MS 21, Fcgr2b−/−+αPD-1+PBS: n=5, MS 26, Fcgr2b−/−+αPD-1+FTY: n=5, MS 21). (I–M) To analyze tumor antigen-specific CD8 T cells, OVA-H-2Kb tetramer+ CD8 T cells (Tet+ CD8 T cells) were quantified in FTY-treated and anti-PD-1 antibody-treated Fcgr2b−/− mice (n=5 per group). (I) Representative flow plots of Tet+ CD8 T cells from the tumor-bearing brains and TdLNs. For the Tet+ CD8 T cells gating of GL261 (without SIINFEKL), a GL261 injected brain was used as a control. The cells were gated as singlet, live CD45.2+ CD11b, CD8a+ CD44+, and Tet+ populations. (J) Quantification of Tet+ CD8 T cells in TdLNs and (K) CD8 T cells in blood (L) analysis of tumor-infiltrating Tet+ and (M) CD39+PD-1+ CD8 T cells. Error bars represent the SEM, and p values were calculated using paired (E) or unpaired Student’s t-tests. The differences in survival were analyzed by the log-rank (Mantel-Cox) test (H). MS, median survival; PBS, phosphate-buffered saline; PD-1, programmed cell death protein-1; TCF1, T-cell factor 1.

To evaluate the impact of CD8 T cells residing in TdLNs on anti-PD-1 therapy, an intervention was performed using an S1P1 inhibitor, FTY720 (figure 5G). Before the initiation of anti-PD-1 therapy, CD8 T cell sequestration was confirmed by the absence of T cells in the blood circulation in mice treated with FTY720. In the survival analysis, FTY720 administration significantly reduced the effectiveness of the anti-PD-1 treatment (figure 5H). In addition, we identified tumor antigen-specific CD8 T cells in TdLNs by analyzing OVA-H-2Kb tetramer+ (Tet+) CD8 T cells following GL261-OVA cell administration. The results showed that FTY720 administration had no impact on the total number of CD8 T cells (online supplemental figure 9C), but it resulted in an increased frequency of Tet+ CD8 T cells (figure 5I,J). Notably, peripheral blood CD8 T cells were undetectable in FTY720-treated groups (figure 5K), suggesting that tumor antigen-primed Fcgr2b−/− CD8 T cells were sequestered in TdLNs by FTY720.

When comparing Tet+ CD8 T cells in the blood, Fcgr2b−/− mice exhibited a higher frequency than WT mice (online supplemental figure 9D,E), indicating that a systemic response may impact the efficacy of anti-PD-1 treatment. Sequentially, FTY720 treatment resulted in a reduction in the number of brain-infiltrating total CD8 T cells and Tet+ CD8 T cells (figure 5I,L, online supplemental figure 9F). These findings suggest that Ttsms, which exhibited effectiveness in anti-PD-1 treatment, are preserved within a specific niche, TdLNs, and their egress from TdLNs to the blood and ultimately to the brain, which contributes to augmenting the antitumor response.

Surprisingly, with the absence of newly recruited antigen-specific Ttsms from TdLNs, tumor-infiltrating CD8 T cells in Fcgr2b−/− mice exhibited an increase in the exhaustion marker (PD-1, CD39) expression (figure 5M). This result underscores the crucial role of TdLNs in enhancing the anti-PD-1 response in Fcgr2b−/− mice by providing a secure environment for Ttsms. The findings illustrate the complex interplay among T cell migration, systemic responses, and TME dynamics in influencing immunotherapeutic outcomes.

In line with the observations from the mouse GBM model, we confirmed the expression of FcγRIIB in CD8 T cells derived from the peripheral blood mononuclear cells of patients with GBM (online supplemental figure 10A). Compared with the healthy control group, patients with GBM exhibited a higher frequency of FcγRIIB expression in circulating CD8 T cells (online supplemental figure 10B,C). Consistent with our findings in mice, FcγRIIB expressing CD8 T cells showed elevated in production of IFN-γ and TNF-α (online supplemental figure 10D). Additionally, we assessed stemness characteristics by classifying CD8 T cells based on TCF1 and TOX expression. This analysis revealed reduced TCF1 levels and enrichment of TOX in patients with GBM, indicating a shift towards an exhausted phenotype. These results suggest that peripheral CD8 T cells in patients with GBM exhibit diminished stemness and increased exhaustion compared with the healthy control group. Notably, within the cohort of patients with GBM, exhausted CD8 T cell subsets (TCF1 TOX+) displayed higher levels of FcγRIIB compared with the stemness-associated subset (TCF1+ TOX) (online supplemental figure 10E). This finding suggests that FcγRIIB is significantly upregulated in exhaustion-associated CD8 T cell subsets in patients with GBM.

Discussion

In our study, targeting FcγRIIB shows promise in overcoming ICB therapy resistance in GBM. While PD-1 and FcγRIIB were significant risk factors in patients with GBM, individually targeting PD-1 or deleting FcγRIIB in a murine model did not affect the antitumor response. However, a synergistic effect was observed with FcγRIIB ablation and anti-PD-1 therapy in Fcgr2b−/− mice injected with GL261 cells, resulting in improved survival rates. This survival benefit was accompanied by increased activation of tumor-infiltrating CD8 T cells, leading to enhanced cytotoxicity and proliferation. In the brain TME, we observed high stemness in Fcgr2b−/− CD8 T cells, leading to the formation of Ttsms, recognized as optimal for ICB therapy. Notably, these Ttsms originated from TdLNs and exhibited enhanced responsiveness to anti-PD-1 treatment. The increased presence of Ttsms in Fcgr2b−/− CD8 T cells and the administration of an S1P1 inhibitor highlight the essential role of TdLNs in attracting and maintaining tumor-infiltrating Ttsms. Furthermore, our findings indicate that elevated expression of FcγRIIB in CD8 T cells derived from the PBMCs of patients with GBM is linked to an exhausted phenotype. These findings offer crucial insights for developing effective therapeutic strategies against GBM.

Researches have explored the relationship between Fc receptors and extrinsic antibodies, particularly in ICB, on the functional alterations of CD8 T cells. A study by Moreno-Vicente et al39 emphasized that the Fc region has an impact on extrinsic antibody-mediated functional modifications of CD8 T cells in a hot tumor mouse model, contrasting with diminished effects observed in a cold tumor model. Therefore, the functional alterations of CD8 T cells induced by FcγRIIB deletion may differ depending on the cancer type. A recent study conducted by Bennion et al40 demonstrated that the efficacy of anti-PD-1 therapy was attenuated by the inhibitory signaling pathway of FcγRIIB in a murine melanoma model. The authors demonstrated that the deletion of the Fcgr2b gene or the inhibition of FcγRIIB led to improved efficacy of anti-PD-1 treatment. Our data also exhibited enhanced antitumor response in Fcgr2b−/− mice combined with anti-PD-1 treatment against GBM. Nonetheless, we primarily focused on the characteristics of CD8 T cells to present a novel mechanism explaining the beneficial effect in representative cold tumors, such as GBM. The expression of FcγRIIB in CD8 T cells did not affect the binding affinity of the anti-PD-1 antibody. However, its deletion influenced functional changes mediated by PD-1 blocking. The blockade of PD-1 enhanced the antitumor response against GBM in CD8 T cells and the expansion of Ttsms due to Fcgr2b deletion, emerging as a critical factor contributing to the outcomes. Our study confirmed the increase of Ttsms in polyclonal Fcgr2b−/− CD8 T cells, which corresponds to the changes observed in transferred Fcgr2b−/− OT-I cells. The enhancement of the antitumor response facilitated by anti-PD-1 treatment involves CD8 T cell-intrinsic alterations, achieved by abolishing inhibitory signals from FcγRIIB, as indicated by these findings.

Moreover, the deep cervical lymph nodes, which serve as TdLNs, play a crucial role in sustaining the persistence of tumor-specific CD8 T cells by protecting them from metabolic and epigenetic suppressors in the TME. Recent studies highlight an increasing acknowledgment of systemic CD8 T cell responses in ICB therapy.15 Investigations reveal that TCF1+ CD8 T cells, with reduced PD-1 levels and other exhaustion markers, significantly influence anti-PD-1 therapy.14 15 This characterization implies that CD8 T cells specific to tumor antigens could potentially persist in the TdLN and tertiary lymphoid structures, highlighting the potential advantages of ICB therapy.14 28 38 41 Moreover, Ttsms have been identified as precursors of tumor-infiltrating CD8 T cells. The pseudotime analysis revealed the existence of Ttsms within the TME, underscoring their significance as the principal cluster in the progression of tumor-infiltrating CD8 T cells from Tpexs to Ttexs. The results suggest that Ttsms enhance the effectiveness of anti-PD-1 treatment by targeting a highly responsive cell population within the TME. Ttsms also demonstrate systemic CD8 T cell features, indicating their potential responsiveness to anti-PD-1 therapy.

Previous research has underscored the significance of the brain, deep cervical lymph nodes, and interconnecting network of meningeal lymphatic vessels.42 43 In patients with GBM, the integrity of this link is disrupted as a result of diminished levels of vascular endothelial growth factor-C, resulting in the decreased efficacy of anti-PD-1 treatment in patients.44 45 Given the significance of TdLNs in the efficacy of ICB, the presence of Ttsms of TdLNs in patients with GBM holds particular importance compared with other tumor types.

The GSEA results indicated that the Ttsms subset of Fcgr2b−/− CD8 T cells showed enrichment of the TCR-related pathways, highlighting its significant activation of both proximal and distal TCR signaling. The maintenance of stemness often contradicts TCR-induced effector-like traits because activated CD8 T cells require metabolic changes to execute effector functions, especially when encountering tumor cells that induce exhaustion via metabolic stress. Contrary to expectations, the Fcgr2b−/− Ttsms exhibited more effector-like functions along with an enrichment of the oxidative phosphorylation pathway. This suggests that Fcgr2b−/− Ttsms maintained mitochondrial fitness to modulate the balance between stemness and effector functions.46 47 This phenomenon also supports that Ttsms are initially recruited in the tumor niche, and their metabolic status is stably maintained for TCR stimulation. A recent study introduced a predictive model for combination therapy involving ICB and other drugs.48 The findings revealed that four pathways—TCR signaling, natural killer cell cytotoxicity, T cell cytotoxicity, and antigen processing and presentation—were positively associated with patient response to the synergistic effect of anti-PD-1 therapy. These results align with the GSEA results of Ttsms, indicating that the enhanced TCR and its downstream intracellular pathways correlate with an improved antitumor response in Fcgr2b−/− mice treated with anti-PD-1.

Our study identified FcγRIIB as a new target with the potential to enhance the effectiveness of anti-PD-1 therapy, particularly in ICB resistant tumor model, GBM. Ttsms residing in TdLNs played a crucial role in augmenting the response to anti-PD-1 treatment, and the ablation of FcγRIIB increased the frequency and characteristics of Ttsms. This result offers valuable insights into a promising direction for enhancing immunotherapeutic approaches.

Methods

Mice

Male 8–10-week-old C57BL/6J mice obtained from KAIST and DBL (Eumseong, Korea). B6;129S-Fcgr2btm1Ttk/J (Fcgr2b−/−, stock # 002848), CD45.1 (B6.SJL-PtprcaPepcb/BoyJ, stock # 002014), and OT-I (C57BL/6-Tg(TcraTcrb)1100Mjb/J, stock #003831) mice were acquired from the Jackson Laboratory (Bar Harbor, Maine, USA). Fcgr2b−/− OT-I mice were generated by crossbreeding Fcgr2b−/− and OT-I mice. The mice were housed in a specific pathogen-free facility at KAIST Laboratory Animal Resource Center under controlled conditions: Under a 12-hour light-dark cycle and temperature-controlled and humidity-controlled conditions, with free access to food and water. All experimental procedures were conducted following approved guidelines and protocols (KA2022-065-v1, KA2022-067-v1) by the Institutional Animal Care and Use Committee of KAIST.

Cell lines

The GL261, GL261-GFP (provided by Dr Injune Kim from KAIST), and CT2A (Sigma-Aldrich, St. Louis, Missouri, USA) cell lines were cultured in Dulbecco’s Modified Eagle Medium (DMEM, Corning, New York, USA) supplemented with 10% fetal bovine serum (Corning) and 1% penicillin-streptomycin (GenDEPOT, Barker, Texas, USA). The generation of a GL261-SIINFEKL cell line was described in the online supplemental methods. The cell lines were subcultured using trypsin-EDTA (Corning) and maintained in a humidified CO2 incubator at 37°C.

Syngeneic mouse GBM model

To induce murine GBM in the mouse brain, we used an orthotopic glioma injection model by intracranially inoculating the mouse glioma cell lines GL261 or CT2A. For tumor implantation, the cells were detached using 0.05% trypsin-EDTA, neutralized with DMEM, and subsequently rinsed with Dulbecco’s phosphate-buffered saline (DPBS). For inoculation, cells were ultimately diluted to a concentration of 1×105 cells/2 µL in DPBS. For details on glioma cell line inoculation procedure see online supplemental methods.

Single-cell preparation of immune and tumor cells

In the tumor-infiltrating immune cell analysis, mice were anesthetized and underwent transcardial perfusion with cold DPBS. The right hemisphere of the brain was weighed and dissociated using a Tumor Dissociation Kit (Miltenyi Biotec, Auburn, California, USA) according to the manufacturer’s protocol. For details on the single-cell preparation procedure of immune and tumor cells see online supplemental methods.

Administration of ICB and S1P1 receptor inhibitors

Mice were randomly grouped 10 days after tumor cell injection. The experimental group received 200 µg of αPD-1 antibodies (Bio X Cell, Lebanon, New Hampshire, USA, RMP1-14) intraperitoneally on days 10, 13, and 16. The control group was administered an isotype control injection (Bio X Cell, 2A3). Additionally, to impede T cell egress from lymph nodes, 25 µg of the S1P1 inhibitor (FTY720, Cayman Chemical, Ann Arbor, Michigan, USA) was intraperitoneally administered starting from day 7, and then given every other day to maintain the inhibition.

Adoptive transfer of OT-I cells

To evaluate the immune response of antigen-specific CD8 T cells, OVA-specific transgenic T cells were transferred into naive mice prior to the injection of GL261-SIINFEKL cells. Naive OT-I cells were isolated from the spleens and lymph nodes of CD45.1+ congenic Fcgr2b−/− OT-I mice or CD45.2+ CD45.1+ congenic WT OT-I mice using the Magnisort Mouse CD8 Naive T Cell Enrichment Kit (Invitrogen, Waltham, Massachusetts, USA), resulting in a purity of 93–95% for CD3+CD8+CD44low T cells. Purified naive OT-I cells were quantified using a CellDrop FL device (DeNovix, Wilmington, Delaware, USA). Subsequently, the cells were resuspended in PBS buffer at a 1:1 ratio of WT and Fcgr2b−/− OT-I cells, totaling 1×106 cells in 100 µL, and were intravenously delivered to CD45.2+ recipient mice via the tail vein.

Flow cytometry

Single-cell suspensions were prepared as described above for surface staining, which involved treatment with Fc Block (ATCC, Manassas, Virginia, USA, HB-197, 2.4G2) and subsequent incubation with surface marker antibodies for 30 min in the dark at 4°C. The list of antibodies for staining cell surfaces is provided in online supplemental table S2. To exclude dead cells, a Zombie Aqua Fixable Viability kit (BioLegend, San Diego, California, USA) or Fixable Viability Stain (BD Biosciences, San Jose, California, USA) was used to stain cells for 10 min in the dark at 4°C, followed by washing in FACS buffer.

The detailed methods for staining transcription factors, cytokines, and in vivo-administered anti-PD-1 antibody are provided in the online supplemental methods.

Single-cell RNA sequencing and library preparation

After single brain cells were prepared, cell suspensions were pooled from five samples per group. For immune cell purification, single cells were treated with Fc Block and stained with anti-mouse CD45.2 antibody (eBioscience) and 7-Aminoactinomycin D (7-AAD) viability dye (BioLegend). CD45.2+ 7-AAD live immune cells were sorted using an FACSAria II cell sorter (BD Biosciences). The 10× chromium single cell 3’ library kit (10x Genomics, Pleasanton, California, USA) was used to generate a single-cell library. Samples (10,000 cells) were sequenced using an HiSeq X Ten system (Illumina, San Diego, California, USA). For a detailed analysis process of scRNA-seq, refer to the online supplemental methods.

Statistical analysis

The statistical analysis was performed using GraphPad Prism Software (V.10.2.0). Graphs represent mean values with SEM. Group comparisons were conducted using unpaired two-tailed Student’s t-test or one-way analysis of variance with Tukey’s post hoc test. A two-tailed paired Student’s t-test was used for the OT-I analysis. Kaplan-Meier curves depict survival data analyzed with the log-rank (Mantel-Cox) test. Statistical significance was set at p<0.05, denoted as *p<0.05, **p<0.01, and ***p<0.001.

Data availability statement

Data are available upon reasonable request.

Ethics statements

Patient consent for publication

Ethics approval

This study was approved by the Institutional Review Board of Konyang University Hospital (approval number: KYUH 2020-06-009-008) and Seoul St. Mary’s Hospital (approval number: KC20TISI0251). Informed consent was obtained from all individual participants included in the study.

Acknowledgments

The authors thank Ji Ye Kim at the BioMedical Research Center for technical service and all members of the Host Defenses laboratory for thoughtful advice.

References

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Footnotes

  • X @heungkyulee

  • Contributors KBK, CWK, YK, BHK, JL, IK, WHP, SA, SKL, and HKL designed and performed the experiments. KBK and HKL conceived the study, analyzed the data, and wrote the manuscript. HKL supervised the study. HKL is the guarantor of the study.

  • Funding This work was supported by the National Research Foundation of Korea grant (NRF-2021M3A9D3026428, NRF-2023R1A2C3003825, and RS-2024-00439735). This study was also supported by the Samsung Science and Technology Foundation (SSTF-BA1902-05), Republic of Korea.

  • Competing interests No, there are no competing interests.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data availability All data relevant to the study are included in the article or uploaded as online supplemental information. scRNA-seq data from immune cells in WT and Fcgr2b−/− mice treated with anti-PD-1 have been deposited in NCBI's Gene Expression Omnibus under accession code GSE262592.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.