WHAT IS ALREADY KNOWN ON THIS TOPIC

Programmed cell death protein 1 (PD-1) blockade boosts antigen-specific humoral immunity, as confirmed by human vaccination studies and mouse tumor models. However, the impact of PD-1 blocking antibodies on human humoral immunity still requires further exploration.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

Our findings reveal a CD38+ Tfr cell subset presented in the peripheral blood, regulated by PD-1 signaling, which plays a role in regulating humoral immunity. This finding suggests the regulatory role of anti-PD-1 therapy in humoral immunity and CD38+ circulating T follicular regulatory cell might be a potential indicator of humoral immunity in PD-1 blockade therapies.

Introduction

In recent years, PD-1 has emerged as a critical regulator of immune responses that profoundly affects T-cell function and immune tolerance, and remarkable breakthroughs have been made in cancer therapy with anti-PD-1 antibodies.1–3 Despite the revolutionary success of anti-PD-1 therapy in cancer treatment, much remains to be understood about how PD-1 regulates humoral immunity during cancer immunotherapy. Studies have shown that PD-1 pathway disruption affects immune responses to vaccines, particularly humoral responses. On influenza vaccination, PD-1 blockade significantly induces circulating T follicular helper (cTfh) cells and antigen specific antibodies.4 5 Mouse studies have also elucidated the role of anti-PD-1 antibodies on the host humoral immune response.6 However, the impacts of PD-1 blocking antibodies on human host humoral immunity still require further exploration.

As an immune checkpoint, PD-1 plays an important role in T-cell activation and differentiation. The negative regulation of TCR signaling by PD-1 elevates the activation signal threshold and curbs sustained excessive stimulation, which is critical for suppressing self-antigen induced T-cell activation and preventing overactivation. Besides, PD-1 affects T cell differentiation. Mouse studies have shown that loss of PD-1 in the T-cell priming phase is expected to result in impaired formation of memory T cells.7–9 In addition, PD-1 is involved in the differentiation and function of regulatory T (Treg) cells. The PD-1/programmed death-ligand 1 (PD-L1) axis is critical for peripheral Treg cell differentiation and homeostasis, as well as the conversion of Th cells into Treg cells.10–12 PD-1 also regulates the suppressive function of Treg cells, and PD-1 blockade enhances the immunosuppression of PD-1+ Treg cells, which can be an indicator of the efficacy of PD-1 blockade therapies.13

Recently, a subpopulation of CD4+ Treg cells, termed T follicular regulatory (Tfr) cells, exhibit phenotypic traits resembling both Treg cells and T follicular helper (Tfh) cells. This CD4+ T-cell subset highly expresses CXCR5, Bcl6, Foxp3, and CD25 (IL-2Rα), playing a critical role in the germinal center (GC) reaction and antibody production. Previous reports have suggested that Tfr cells originate from thymus-derived natural Treg cells,14–16 and these CXCR5-expressing Treg cells are further regulated by TCR signaling, IL-2 signaling, and other regulatory signals to become mature and functional Tfr cells.14 17–21 However, recent evidence has also shown that Tfr cells can also be generated from activated Tfh cells,22–24 and some researchers have named these cells iTfr cells. Moreover, there are CD4+CXCR5+FOXP3+ICOSint T cells in peripheral blood, known as circulating Tfr (cTfr) cells. These cTfr cells exhibit memory-like properties25–27 and can be recruited to the follicles and GCs of secondary lymphoid organs following reactivation.27 Single-cell sequencing, combined with TCR sequencing analysis, revealed different evolutionary origins of Tfr cells, indicating the heterogeneity of these cells.23 Most studies investigating the role of Tfr cells have focused mainly on the regulation of GC reactions and autoimmunity due to their unique positioning. The CXCR5-mediated localization of B follicles and GCs enables Tfr cells to suppress Tfh cell activation, B-cell differentiation, and antibody production, thereby controlling the magnitude of the GC response.14–16 28 Similar to Treg cells, Tfr cells function mainly through inhibitory molecules and cytokines, such as CTLA-4 and IL-10, respectively.14 29–31 In addition to their regulatory roles in GC responses, Tfr cells have recently been found to play a non-redundant role in tumor immunity, especially in the tumor microenvironment (TME) and tertiary lymphoid structures (TLS). An increase in the frequency of Tfr cells in the TME indicates an increased risk of metastasis and curtails the efficacy of anti-PD-1 therapy.32–35 However, studies on the role of Tfr cells in tumor and immune therapy, especially those involving the regulation of human Tfr cells, still require further exploration.

Currently, most anti-PD-1 antibodies used in the clinic, including nivolumab (Opdivo), pembrolizumab (Keytruda), sintilimab (Tyvyt), and durvalumab (Imfinzi), are applied intravenously. However, the effects of these PD-1-blocking antibodies on the humoral immune response in the peripheral circulation remain to be further investigated. Here, we found that anti-PD-1 therapy significantly increased the serum immunoglobulins and plasmablasts in the peripheral blood, along with a significant increase in the proportion and number of cTfr cells. In addition, PD-1 blockade affected the proliferation and function of cTfr cells. PD-1 signaling is involved in the differentiation of CD38+ cTfr cells, a subset that promotes plasmablasts formation, and is correlated with altered antibody production. Together, these data reveal the immunomodulatory role of PD-1 on host baseline humoral immunity, which may be partially due to its limitation on the differentiation and function of human cTfr cells.

Materials and methodsStudy design

The main purpose of this study was to evaluate the roles of PD-1 in human humoral immunity and immune cells in the peripheral circulation. This study was conducted by analyzing immune cells in the peripheral blood from healthy donors, untreated patients with cancer and patients with cancer undergoing anti-PD-1 therapy.

Human samples

All individuals enrolled provided informed consent for inclusion of their blood samples in the present study. The general information about the patients is provided in online supplemental table 1. Peripheral blood from healthy donors (n=25), new-onset hepatocellular carcinoma (HCC) patients prior to systemic treatment (n=25), and HCC patients receiving anti-PD-1 therapy (n=25) from December 2021 to October 2023 were collected in this study. All HCC patients were precisely diagnosed according to histopathological examinations following the American Association for Study of Liver Disease guidelines.36 The new-onset HCC patients with no prior anticancer therapy history. Patients receiving anti-PD-1 therapy were intravenously injected with camrelizumab (SHR-1210, HengRui Pharmaceuticals Co., 3 mg/kg every 3 weeks), sintilimab (Tyvyt, Innovent, 200 mg every 3 weeks), or tislelizumab (BGB-A317, Novartis, 200 mg every 3 weeks). Blood samples from patients receiving 2–3 courses of anti-PD-1 therapy were collected within 3 weeks after the last injection and before surgery or other traumatic treatments or systemic treatments. Peripheral blood mononuclear cells (PBMCs) were isolated and the serum was collected for further analysis.

PBMC isolation and flow cytometry

PBMCs were isolated from peripheral blood by Ficoll (Cytiva) density gradient centrifugation. The intermediate-layer cells were collected for further analysis. The samples were refiltered to generate a single-cell suspension before use in further analyses. For surface marker staining, PBMCs were incubated on ice for 30 min while the Tfh/Tfr cell staining were conducted at room temperature for 30 min37 with fluorescence-tagged antibodies (online supplemental table 1). For transcription factor staining, cells were fixed and permeabilized with the Foxp3/transcription factor staining buffer set (Thermo Fisher Scientific) according to the manufacturer’s instructions. Transcription factors including FOXP3, BCL6, and CTLA-4 were incubated for 45 min on ice. For intracellular cytokine staining, cells were stimulated with phorbol 12-myristate 13-acetate (20 ng/mL, Invitrogen) and ionomycin (1 µg/mL; Invitrogen) in the presence of Golgi-Plug (1 µg/mL, Invitrogen) for 4 hours. Permeabilization was performed using the intracellular fixation/permeabilization concentrate and diluent kit (Thermo Fisher Scientific) for 20 min on ice, followed by staining with intracellular markers. All FACS analyses were performed on a CytoFLEX S (Beckman Coulter), and the data were analyzed by using CytExpert for CytoFLEX S (V.2.5) software and FlowJo software (V.10.6.1).

Immune cell isolation and culture

CD4+ T cells were first enriched from PBMCs with a Human CD4+ T Cell Enrichment Kit (STEMCELL Technologies). Enriched CD4+ T cells were stained with fluorescence-tagged antibodies, and Treg (CD4+CD127low/−CD25+CXCR5−), Tfh (CD4+CD127+CD25−CXCR5+), and Tfr (CD4+CD127low/−CD25+CXCR5+), CD38+ Tfr (CD4+CD127low/−CD25+CXCR5+CD38+), CD38− Tfr (CD4+CD127low/−CD25+CXCR5+CD38−) cells, naïve B cells (CD19+CD3−IgD+) were isolated using a Beckman Moflo Astrios cell sorter.

The isolated Tfh, Treg and Tfr cells (2×104 cells) were cultured in complete RPMI-1640 medium supplemented with 1% penicillin/streptomycin (Thermo Fisher Scientific), 50 µM β-mercaptoethanol (Thermo Fisher Scientific), and 1% L-glutamine (Thermo Fisher Scientific) and stimulated with anti-CD3/CD28 microbeads (Gibco) at a ratio of 1:1, with additional 100 IU/mL IL-2 (PeproTech) in 96-well round-bottom plates. Sintilimab (1 µg/mL) was added to the culture medium of the indicated groups, and the proliferation, apoptosis and induction of CD38+ Tfr cells were analyzed on day 4. In vitro coculture experiment of Tfh cells, B cells, and cTfr cells were conducted according to previous reports.37–39 In brief, 2×104 Tfh cells, 4×104 B cells, and 1×104 CD38+ or CD38− Tfr cells were sorted and cultured with diluted anti-CD3/CD28 microbeads at a final bead-to-cell-ratio of 1:32, along with anti-IgM (5 µg/mL, Jackson ImmunoResearch) in complete RPMI-1640 medium containing 10% FBS (Gibco), 1% penicillin/streptomycin (Thermo Fisher Scientific), and 1% L-glutamine (Thermo Fisher Scientific) for 6 days. Cells were collected for flow cytometry.

Immunoglobulins analysis

The serum samples were recentrifuged before analysis. Immunoglobulins were analyzed using human IgG/IgA/IgM ELISA kits (MULTI SCIENCES) following the manufacturer’s instructions. Briefly, 100 µL of samples and diluted standard substances were added to the plates and incubated at room temperature for 2 hours. The plates were then washed, and 100 µL of detected antibody was added, followed by another 2-hour incubation. After thorough washing, the substrate was added. The absorbance was measured at 450 nm and 630 nm. The concentrations were calculated by standard curve extrapolation.

RT-qPCR

Total RNA was prepared with TRIzol (Invitrogen), and cDNA was synthesized using PrimeScript RT Master Mix (TaKaRa). Quantitative PCR (qPCR) was performed using Power SYBR Green PCR Master Mix (TaKaRa). The sequences of the gene-specific primers used are listed in online supplemental table 1. All reactions were performed in triplicate, and the results were calculated by the change-in-threshold (2−ΔΔCT) method with GAPDH as a housekeeping reference gene.

Statistical analysis

Statistical analysis was conducted using GraphPad Prism 9 software (GraphPad Software). Data are presented as mean±SEM unless otherwise specified. Student’s t-test (two-tailed) was used for the statistical analysis of differences between two groups when the data fitted a normal distribution, whereas a two-tailed Student’s t-test with Welch’s correction was used when variances were different. One-way analysis of variance with Tukey’s post hoc test was used for the statistical analysis of differences among multiple groups. The correlation between the two parameters was evaluated using the Pearson correlation test. A p value <0.05 was considered statistically significant, with significance levels indicated as *p<0.05 and **p<0.01. P values >0.05 were considered not significant (ns).

ResultsAnti-PD-1 therapy promotes antibody generation

To address the impact of PD-1 blocking antibodies on the humoral immune response in peripheral blood, we collected serum and PBMCs of a cohort including 25 healthy donors (HC group), 25 new-onset HCC patients without systemic treatments (untreated group), and 25 HCC patients receiving 2–3 courses of anti-PD-1 therapy with sintilimab, camrelizumab, or tislelizumab (anti-PD-1 group). We first evaluated the serum immunoglobulin titers among the three groups. Surprisingly, the serum IgG and IgA levels were greater in the anti-PD-1 group than in the HC and untreated groups, but the IgM titer did not significantly differ among the three groups (figure 1A), which indicated changes in humoral immunity under anti-PD-1 therapy. We further explored the cellular immune landscape. Complete blood count reports showed that the number of white blood cells (WBCs) did not differ among the three groups (online supplemental figure S1A), but the percentages and numbers of lymphocytes were slightly lower in the untreated and anti-PD-1 groups than in the HC group (online supplemental figure S1B). We first analyzed B cells (online supplemental figure S1C). The flow cytometry analysis showed that the percentage and number of circulating CD19+ B cells did not significantly differ among the three groups (figure 1B). The percentage of IgD+CD27− population which is composed by naïve B cells and transitional B cells, increased in the anti-PD-1 group, while its number was comparable among the three groups. Besides, both the percentage and number of IgD+CD27+ non-switched memory B cells, IgD−CD27+ switched memory B cells and double-negative B cells were comparable among the three groups (figure 1C). We also observed a significant increase in the percentage and number of plasmablasts in the anti-PD-1 group, consistent with the increase of serum immunoglobulin levels (figure 1D). We also screened CD4+ and CD8+ T cells. The percentages of CD4+ and CD8+ T cells did not show significant difference among the HC, untreated, and anti-PD-1 groups, while the numbers of CD4+ and CD8+ T cells decreased between the HC and tumor groups (online supplemental figure S1D), which seems to be caused by lymphopenia. In the anti-PD-1 group, the percentage of CD45RA+CCR7+ naïve cells in CD4+ T cells and CD8+ T cells significantly decreased, while the CD45RA−CCR7+ central memory (CM) population and the CD45RA−CCR7− effector memory (EM) population in CD4+ T cells increased (online supplemental figure S1E), although the percentage of EM and CM populations of CD8+ T cells did not significantly change (online supplemental figure S1F). These data indicated that PD-1 blockade modulated systemic antibody generation and the cellular immune landscape in the peripheral blood.

Figure 1

Anti-PD-1 therapy promotes antibody generation. (A) Graphs show the serum titers of IgG, IgA, and IgM in the indicated groups (HC n=25, untreated n=25, and anti-PD-1 groups n=25). (B) CD19+ B cell percentage and number in the HC, untreated, and anti-PD-1 groups. (C) Flow cytometry plots showing CD27 and IgD expression in the indicated groups. Numbers adjacent to the outlined areas indicate the percentages of CD27−IgD+, CD27+IgD+, CD27+IgD− and CD27−IgD− B cells. The right graphs show the statistic results and number of indicated B cell subsets. (D) Flow cytometry plots showing expression of CD27+CD38+ plasmablasts. Numbers adjacent to the outlined areas indicate the percentages of plasmablasts in the indicated groups. The right graphs show the statistic results and numbers of plasmablasts. Small horizontal lines indicate the mean; ns, not significant, *p<0.05 and **p<0.01 by one-way analysis of variance. PD-1, programmed cell death protein 1.

PD-1 blockade increases the frequency and number of cTfr cells

Class switching occurs during the GC and is tightly regulated by Tfh and Tfr cells. The increased serum titer of class-switched IgG and IgA (figure 1A), as well as plasmablasts (figure 1D), prompted us to further explore the roles of Tfh and Tfr cells (online supplemental figure S2A). The CD4+FOXP3−CD25−CXCR5+ Tfh population in the circulation seemed to increase in the anti-PD-1 group, although not significantly, compared with that in the HCs and untreated individuals (figure 2A). However, both the percentages and numbers of CD4+FOXP3+CD25+CXCR5+ Tfr cells increased following PD-1 blockade (figure 2B). To exclude the effect of the anti-PD-1 antibody on the CD4+FOXP3+CD25+ regulatory population, we analyzed this subset in the PBMCs. We found no significant difference in the CD4+FOXP3+CD25+ population among the three groups (online supplemental figure S2B). BCL6 is considered the master transcription factor of Tfr cells, although its expression in the periphery is heterogeneous.26 40 41 Therefore, we evaluated BCL6 expression in the CD4+FOXP3+CD25+ Treg cell population and observed an increase in both the percentage of BCL6+CXCR5+ cells and the fluorescence intensity of BCL6 (figure 2C,D). However, PD-1 blockade did not alter the CXCR5 or CD25 fluorescence intensity (figure 2D). The Tfr/Tfh cell ratio, an indicator of suppressive function of Tfr cells, has greater clinical significance than the absolute number of Tfr cells or Tfh cells. We found that the Tfr/Tfh cell ratio increased in the PD-1 blockade group (figure 2E), indicating enhanced suppressive function of Tfr cells under PD-1 blockade conditions. These findings suggest that anti-PD-1 treatment may regulate the homeostasis of circulating Tfr cells.

Figure 2

PD-1 blockade increases frequency and numbers of circulating Tfr cells. (A) Flow cytometry analysis of CXCR5 expression among CD4+FOXP3−CD25− compartment in the HC group, untreated group, and anti-PD-1 group. Graphs show the percentages and numbers of Tfh cells (CD4+FOXP3−CD25−CXCR5+). (B) Flow cytometry analysis of CXCR5 expression among CD4+FOXP3+CD25+ compartment in the HC group, untreated group, and anti-PD-1 group. Graphs show the percentages and numbers of Tfr cells (CD4+FOXP3+CD25+CXCR5+). (C) Flow cytometry plots showing BCL6+CXCR5+ population among CD4+FOXP3+CD25+ cell compartments in the indicated groups. The right graphs show the percentage and number of BCL6+CXCR5+ cells. (D) The mean fluorescence intensity (MFI) of BCL6, CXCR5, and CD25 in CD25+FOXP3+ cells. (E) The ratio of circulating Tfr to Tfh cells in peripheral blood. Small horizontal lines indicate the mean; ns, not significant, *p<0.05 and **p<0.01 by one-way analysis of variance. PD-1, programmed cell death protein 1; Tfh, T follicular helper cell; Tfr, T follicular regulatory cell.

PD-1 regulates the homeostasis of cTfr cells

To further investigate the regulatory effect of PD-1 on cTfr cells, we examined the proliferation and apoptosis of this cell population. Since FOXP3 staining requires nuclear permeabilization, which disrupts cell viability, we defined the Treg population using the surface markers CD25 and CD127 as CD4+CD127−/lowCD25+ cells.42 43 We observed that cTfr cells in the PD-1 blockade group exhibited greater proliferation, as indicated by Ki-67 staining, compared with those in the HC and untreated groups (figure 3A). However, the proportions of early apoptotic (annexin V+ 7AAD−) and late apoptotic (annexin V+ 7AAD+) cells within the cTfr population did not significantly differ among the three groups (online supplemental figure S3A). To explore the cell-specific effect of PD-1 blockade on cTfr cells, we evaluated the proliferation and apoptosis of Treg cells. Neither Ki-67 expression nor annexin V and PI staining showed significant changes in circulating Treg cells (online supplemental figure S3B–D). To further confirm the direct effects of anti-PD-1 on cTfr cells, we sorted cTfr cells from PBMCs and cultured them ex vivo with or without anti-PD-1 antibody in the presence of IL-2 and anti-CD3/CD28 beads. In the presence of anti-PD-1, cTfr cells exhibited increased Ki-67 expression (figure 3C), while their apoptotic state remained unchanged regardless of the presence of anti-PD-1 (online supplemental figure S3E).

Figure 3

PD-1 regulates the homeostasis of circulating Tfr cells. (A) Representative contour plots of Ki-67 expression and quantification in CD4+CD127−/lowCD25+CXCR5+Tfr population. (B) Graphs show the MFI of Ki-67 in CD4+CD127−/lowCD25+CXCR5+ Tfr cells. (C) CD4+CD127−/lowCD25+CXCR5+Tfr cells were sorted and cultured in the presence of anti-CD3/CD28 beads and 100 IU/mL IL-2. Representative contour plots showed Ki-67 expression and quantification in Tfr cells. (D) Graphs show the mean fluorescence intensity (MFI) of Ki-67 in ex vivo cultured Tfr cells. (E) Relative mRNA levels of indicated genes in circulating Tfr cells. Data were from two independent experiments and represented as fold changes relative to HC group after normalization with GAPDH. Numbers adjacent to black bars indicate fold in each. (F) Flow cytometry plots showing expression of CTLA-4 in CD4+CD127−/lowCD25+CXCR5+Tfr cells from blood. (G) Graphs showing the flow graph and statistics of CTLA-4 MFI in cTfr cells. Each symbol represents one sample; small horizontal lines indicate the mean; ns, not significant; *p<0.05 and **p<0.01. PD-1, programmed cell death protein 1; Tfr, T follicular regulatory cell.

To assess whether the function of these expanded cTfr cells was altered after PD-1 blockade, we compared the mRNA abundances of key molecules crucial for Tfr cell function. mRNA from flow-sorted cTfr cells revealed a statistically significant difference only in CTLA4 mRNA abundance in the anti-PD-1 group compared with the HC and untreated groups (figure 3E). We further confirmed the upregulation of CTLA-4 protein expression in CD4+CD127−/lowCD25+CXCR5+ cTfr cells in the anti-PD-1 group (figure 3F,G). Tfr cells may also modulate the immune response through CTLA-4-independent pathways. IL-10, an important suppressive cytokine of Treg and Tfr cells,44–46 did not significantly differ among the HC, untreated, and anti-PD-1 groups (online supplemental figure S3F,G), consistent with the mRNA expression data. These findings suggested that PD-1 mainly influences cTfr cell function through CTLA-4 rather than through IL-10.

Characteristics of CD38+ Tfr cells in the peripheral blood

The increased frequency and enhanced function of cTfr cells seemed contradictory to elevated serum immunoglobulin titers. We notice a previous study identified a CD38-marked subset of GC-resident, Tfh-descended Tfr cells named iTfr cells in the human tonsil, which exhibit suppressive functions while retaining the capacity to help B cells.23 Therefore, we investigated whether PD-1 blockade alters the differentiation and composition of cTfr cells. While CD38+ iTfr cells have been observed in tonsils, the presence of CD38+ Tfr cells in peripheral blood remains unclear. Therefore, we evaluated CD38 expression in cTfr cells from human peripheral blood of healthy donors (online supplemental figure S4A). Surprisingly, we found that cTfr cells expressed higher levels of CD38, compared with circulating Tfh cells and Treg cells (figure 4A,B). To further characterize CD38+ Tfr cells, we reanalyzed high-throughput RNA sequencing data from the published dataset.23 Compared with CD38− Tfr cells, CD38+ Tfr cells showed greater expression of cytokines IL-10 and IL-21 but lower levels of FOXP3 transcripts (online supplemental figure S4B,C). Consistent with previous findings, CD38+ Tfr cells in the peripheral blood exhibited increased Ki-67 expression, indicating enhanced proliferation (figure 4C,D). Additionally, we confirmed that CD38+ cTfr cells expressed elevated levels of IL-21 (figure 4E,F), a critical factor in B cell differentiation and the GC reaction. These data suggested that circulating Tfr cells in the peripheral blood can be categorized based on CD38 expression, with CD38-positive cTfr cells resembling the phenotypic characteristics observed in tonsillar Tfr cells.

Figure 4

Characteristics of CD38+ Tfr in the peripheral blood from healthy donors. (A) Representative contour plots of CD38 expression in Treg, Tfh, and Tfr cell population from healthy donors. Numbers adjacent to the outlined areas indicate the percentages of CD38+ cells. (B) The mean fluorescence intensity (MFI) of CD38 in Treg cells, Tfh cells, and Tfr cells. (C) Representative contour plots of Ki-67 expression and quantification, in CD38+ and CD38− Tfr population. (D) Flow graphs show the MFI of Ki-67 in CD38+ and CD38− Tfr cells. (E) Representative contour plots of IL-21 expression and quantification in CD38+ and CD38− Tfr population. (F) The MFI of IL-21 in CD38+ and CD38− Tfr cells. ns, not significant, *p<0.05 and **p<0.01. PD-1, programmed cell death protein 1; Tfh, T follicular helper cell; Tfr, T follicular regulatory cell; Treg, regulatory T cell.

PD-1 blockade increases the CD38+ cTfr cell population

To determine whether PD-1 regulates the composition of cTfr cells, we next analyzed CD38 expression in cTfr cell populations from the HC, untreated, and anti-PD-1 groups. We observed a significant increase in both the percentage and number of CD38+ cTfr cells following PD-1 blockade (figure 5A), suggesting that PD-1 signaling influences the proportion of these cells and involves in their regulation. KEGG enrichment analysis of the published scRNA-seq data23 revealed enrichment of immune and disease-associated pathways, including the cytokine-cytokine receptor interaction pathway and the MAPK signaling pathway, among the differentially expressed genes of the CD38+ Tfr cells. Moreover, the PD-L1 expression and PD-1 checkpoint pathway in cancer showed high enrichment scores in CD38+ Tfr cells (figure 5B), indicating the involvement of PD-1 signaling in their regulation. To validate these findings, we first checked the expression of PD-1 in CD38+ Tfr cells. Compared with the CD38− cTfr compartment, the CD38+ cTfr cells showed high expression of PD-1 (figure 5C,D). We then used an ex vivo culture system to induce CD38+ Tfr cells in the presence of a PD-1 blocking antibody. CD4+CD127+CD25−CXCR5+ Tfh cells and CD4+CD127low/−CD25+ Treg cells were sorted from the PBMCs of healthy donors and cultured with anti-CD3/CD28, IL-2, and sintilimab. Under ex vivo culture conditions, Tfh cells upregulated FOXP3 expression (online supplemental figure S5A,B). In contrast to the control group without anti-PD-1 antibody treatment, PD-1 blockade induced the differentiation of CD38+ Tfr cells from the CD4+CD127+CD25−CXCR5+ Tfh population (figure 5E), while the proportions of differentiated CD38+ Tfr cells among CD4+CD127low/−CD25+ Treg cells were comparable between the control and PD-1 blockade cultures (online supplemental figure S5C). These data suggested that PD-1 signaling is involved in the differentiation of CD38+ cTfr cells from the Tfh cell subset.

Figure 5

PD-1 blockade increases CD38+ cTfr cell population. (A) Flow cytometry plots showing expression of CD38 in CD4+CD127−/lowCD25+CXCR5+ Tfr cells from blood of indicated groups. Numbers adjacent to the outlined areas indicate the percentages of CD38+ Tfr cells. Graphs show the percentages and numbers of CD38+ Tfr cells in indicated groups. (B) KEGG pathway analysis of differentially expressed genes between CD38+ and CD38− Tfr cells. Advanced bubble chart shows enrichment of differentially expressed genes in signaling pathways. (C) Flow cytometry plots showing the PD-1+ percentage in the CD38+ or CD38− circulating Tfr cells. (D) The MFI of PD-1 in CD38+ and CD38− Tfr cells. (E) CD4+CD127+CD25−CXCR5+ Tfh cells were sorted and culture in the indicated culture condition to induce the iTfr population. Flow cytometry plots showing expression of CD38 in CD4+CD127−/lowCD25+CXCR5+ Tfr cells in the condition of control medium and anti-PD-1 antibody ex vivo. Small horizontal lines indicate the mean; ns, not significant, *p<0.05 and **p<0.01. PD-1, programmed cell death protein 1; Tfh, T follicular helper cell; Tfr, T follicular regulatory cell.

To assess whether PD-1 affects the homeostasis and function of CD38+ cTfr cells, we analyzed the CD38+ Tfr population in the HC, untreated, and anti-PD-1 groups ex vivo. We found no significant difference in Foxp3 expression among the three groups (online supplemental figure S6A,B), indicating that PD-1 blockade did not affect the stability of FOXP3 in the increased CD38+ cTfr cell population. Besides, levels of the inhibitory molecule CTLA-4 in CD38+ cTfr cells did not significantly differ among the groups (online supplemental figure S6C,D). Given that the CD38+ Tfr cell subset is an inducible Tfr cell population characterized by high IL-21 expression, we measured IL-21 levels in CD38+ cTfr cells. Despite the increase in the proportion of CD38+ Tfr cells following anti-PD-1 treatment, IL-21 expression remained unchanged compared with that in the HC group or untreated group (online supplemental figure S6E,F). These findings indicate that PD-1 affects the development and differentiation of CD38+ cTfr cells but not the stability or function of fully differentiated CD38+ cTfr cells.

CD38+ cTfr cells are involved in the humoral immune response

To further investigate the impacts of increased numbers of CD38+ cTfr cells on humoral immunity, we evaluated the functions of these cells. Isolated CD38+ cTfr cells and CD38− cTfr cells were cocultured with homologous Tfh cells and naïve B cells (online supplemental figure S7A). Compared with their CD38− cTfr counterparts, CD38+ cTfr cells significantly induced plasmablast differentiation (figure 6A). Additionally, we found a positive correlation between the number of plasmablasts and the percentage of CD38+ cTfr cells in patients undergoing anti-PD-1 treatment (figure 6B). To explore whether serum immunoglobulin titers exhibited a linear relationship with CD38+ Tfr cell percentages in human samples, we calculated pairwise Pearson correlations between these variables. In the anti-PD-1 group, IgG and IgA titers were positively correlated with CD38+ cTfr cell percentages, while IgM levels were not significantly correlated (figure 6C). No correlation was observed between the percentage of CD38+ cTfr cells and the serum immunoglobulin titers in the HC and untreated groups (online supplemental figure S7B,C). The increase in serum antibody titers coincided with the increase in the percentage of CD38+ cTfr cells indicating a regulatory role of PD-1- in CD38+ cTfr cells on humoral immunity.

Figure 6

CD38+ cTfr cells involved in the humoral immune response. (A) Flow sorted Tfh cells, B cells were cocultured with CD38+ or CD38− cTfr cells. Flow cytometry plots showing expression of CD27+CD38+ plasmablasts in CD3− B cells. Numbers adjacent to the outlined areas indicate the percentages of plasmablasts in indicated groups. The right graphs show the statistic results of plasmablasts. (B) Correlation between the percentage of CD38+ Tfr cells and the number of plasmablasts in HC (left), untreated (middle), and anti-PD-1 (right) groups. (C) Correlation between the percentage of CD38+ Tfr cells and the serum titers of IgG (left), IgA (middle), and IgM (right) in anti-PD-1 group. Small horizontal lines indicate the mean; ns, not significant; *p<0.05 and **p<0.01. cTfr, circulating T follicular regulatory cell; PD-1, programmed cell death protein 1; Tfh, T follicular helper cell.

Discussion

Despite the great success achieved by anti-PD-1 therapy, its effects on host humoral immunity remain to be fully demonstrated. Here, we describe a non-specific role of PD-1 in regulating human circulating Tfr cells and humoral immunity. PD-1 blockade significantly increases the titers of class-switched immunoglobulins and the proportion of plasmablasts in the peripheral blood, indicating an altered GC reaction. However, the number of circulating Tfh cells and Treg cells did not significantly differ, while the proportion of cTfr cells increased significantly. Our data showed that PD-1 blockade increased the percentage and number of cTfr cells and modulated their homeostasis and function. Besides, our data showed that PD-1 signaling may regulate the expression level of BCL6 in the cTfr cell population, although some studies have shown that BCL6 is rarely detected in human cTfr cells.40 41 The increases in the percentage and number of cTfr cells were partially due to the high proliferation of these cells in the anti-PD-1 group, suggesting a role of PD-1 in the proliferation of cTfr cells. We also confirmed that PD-1 blockade promoted the proliferation of cTfr cells ex vivo. Moreover, PD-1 blockade increased the mRNA and protein levels of the suppressive marker CTLA-4 in the cTfr population, indicating that PD-1 blockade enhances the suppressive function of cTfr cells.

It is noteworthy that PD-1 inhibition led to a marked increase in CD38+ cTfr cells. This CD38+ newly defined subset of iTfr cells, identified in the tonsils, is located within GCs and originates from highly activated Tfh cells, retaining the capacity to support B-cell function.23 Our current study addressed the existence of CD38+ Tfr cells in the peripheral blood and undergo expansion at the presence of anti-PD-1 antibodies. While circulating Tfh cells are known to expand in tumor-bearing mice after PD-1 blockade, we did not observe a comparable effect in patients undergoing anti-PD-1 therapy at our analysis time point. We speculated that PD-1 inhibition may promote the differentiation of circulating Tfh cells into CD38+ Tfr cells, potentially contributing to an increase in CD38+ cTfr cells. It is anticipated that anti-PD-1 therapy may induce a transient increase in circulating Tfh cells in patients during the early stages of treatment. However, as treatment progresses, Tfh cells may gradually return to baseline levels as they continue differentiating into Tfr cells. The pronounced expression of CD38 in CD38+ Tfr