Abstract
PD-L1 expression levels in tumors do not consistently predict cancer patients’ response to PD-(L)1 inhibitors. We therefore evaluated how tumor PD-L1 levels affect the anti-PD-(L)1 efficacy and T cell function. We used MART-1-specific TCR-T cells (TCR-TMART-1) stimulated with MART-127-35 peptide-loaded MEL-526 tumor cells with different proportions of them expressing PD-L1 to perform cellular assays and high-throughput single-cell RNA sequencing. Compared to control T cells, TCR-TMART-1 were more sensitive to exhaustion and secreted lower pro-inflammatory but higher anti-inflammatory cytokines with increasing proportions of PD-L1+ tumor cells. The colocalization of T cells and tumor cells in gene clusters correlated negatively with the proportion of PD-L1+ tumor cells and positively with immune cell cytotoxicity. Moreover, elevated proportion of PD-L1+ tumor cells increased PD-L1 expression and decreased PD-1 expression on T cells and enhanced T cell death. The expression of PD-1 and PD-L1 in T cells and macrophages also correlated positively with COVID-19 severity.
Introduction
Programmed cell death-ligand-1 (PD-L1) is the ligand of programmed death-1 (PD-1), which are encoded by CD274 and PDCD1, respectively. PD-L1 is expressed in many cancer tissues, including melanoma [1], a widely recognized immunogenic neoplasm. Expression of PD-L1 is undetectable in most normal tissues, but can be induced by inflammatory cytokines, especially interferon-γ (IFN-γ) in various cell types [2-4]. As a strategy to evade immune responses, PD-L1 is often up-regulated on tumor cells and induces T cell anergy, exhaustion or apoptosis upon engagement with PD-1 expressed on tumor infiltrating lymphocytes (TILs) to impair T cell responses[1, 5]. Expression of PD-L1 is not restricted to tumor cells, PD-L1 is also expressed in TILs and its expression by TILs correlates with aggressive tumors, demonstrating the immunosuppressive role of PD-L1 [6, 7]. Binding of PD-1 and PD-L1 impairs T cell activation by interfering with Ras-Raf-MEK-ERK and PI3K-AKT signaling pathways that promote T cell proliferation and differentiation [8]. In addition to binding PD-1, PD-L1 has been reported to interact with CD80 in cis to modulate T cell function and tumor microenvironment [9, 10].
The PD-1/PD-L1 signaling pathway plays an important role in tumor evasion from host immune responses [11]. Inhibitors of PD-1 and PD-L1 have been studied in various tumor types and have now been approved for treating many malignancies, including melanoma, non–small-cell lung cancer (NSCLC), and bladder cancer. [12-16] PD-L1 expression on tumor cells and tumor infiltrating antigen presenting cells (APCs) has been approved as a companion biomarker for the treatment with some of these inhibitors [17-22]. Positive correlation between higher level of PD-L1 expression and higher response rate in melanoma has also been demonstrated [23-25]. However, some studies showed that PD-L1 expression is insufficient to predict a benefit from immune checkpoint inhibitor (CPI) therapy and PD-L1 expression level alone is a poor predictive biomarker of overall survival [26, 27].
The PD-L1 expression level has different predictive values for response to PD-(L)1 blockade in different types of tumors, many tumors that express PD-L1 do not respond to PD-1 or PD-L1 inhibitors. The overall low response rates of PD-1 and PD-L1 inhibitors limit their clinical application. Thus, it is important to know how PD-L1 and its expression level on tumor cells affect the efficacy of immunotherapy and T cell function. The role of PD-L1 has been studied for many years [4-6, 22, 28], but only from the bulk T cell level, which is hard to elucidate the exact relationship between PD-L1 expression and T cell function.
In this study, we used high-throughput single-cell mRNA sequencing (scRNA-seq), multiplex cytokine secretion assay, and cell cytotoxicity assays to investigate the immunoregulatory effect of tumor PD-L1 on responding TCR-T cells. Our research is the first to dissect at the single-cell level transcriptional features as well as cytokine and cytotoxic signatures of antigen-specific TCR-T cells responding to different tumor PD-L1 ratios. Furthermore, single-cell immune profiling was explored in COVID-19 patients, which is essential for understanding the potential mechanisms underlying COVID-19 pathogenesis.
Results
Increased tumor PD-L1 expression suppressed cytotoxicity and cytokine secretion of TCR-TMART-1
We used cytotoxicity and cytokine secretion assays together with scRNA-seq to interrogate TCR-T cells stimulated by MEL-526 melanoma cells with different proportions of them expressing PD-L1 (Fig. 1A). This approach made it possible to quantitatively dissect the T-cell activation state in relation to their subtypes, gene expression and cell differentiation. HLA-A*0201/Melan-A-specific TCR sequence (designed as TCRMART-1) was attained from T cells stimulated with Melan-A (aa27-35, LAGIGILTV) peptide (data unpublished). Melan-A, also known as MART-1, is a melanocytic marker [29]. Human TCRα and TCRβ sequences fused with murine TCR constant region were synthesized and cloned into a lentiviral vector (Fig. S1A). T cells that expressed or did not express TCRMART-1 are designated as TCR-TMART-1 and Tnull, respectively. Ttotal represents the entire T cell population that includes both TCR-TMART-1 and Tnull. After lentiviral transduction into CD8+ T cells, 17.5% of Ttotal was TCR-TMART-1, which reached 97.2% after fluorescence-activated cell sorting (FACS) (Fig. S1B). To verify the cytolytic capacity, TCR-TMART-1 were stimulated by peptide-loaded MEL-526 cells or a mock control at the effector:target (E:T) ratio of 1:1. Compared to Tnull, TCR-TMART-1 killed MEL-526 cells efficiently when MEL-526 cells were loaded with the MART-127-35 peptide (Fig. 1B). TCR-TMART-1 similarly killed T2 cells, another target cell line (Fig. S1C).
To investigate the immunosuppressive role of tumor PD-L1, PD-L1 was overexpressed (OE) on MEL-526 cells (Fig. S1D). Different percentages of PD-L1 positive tumor cells were obtained by mixing OE with wild-type (WT) MEL-526 cells based on the clinical PD-L1 expression ratio [30]. Three tumor cell populations with different percentages of MEL-526 expressing PD-L1 were used in the study: PD-L1low (without exogenous PD-L1, 2.45%), PD-L1int (intermediate, 50.9%), and PD-L1high (high, 100%) (Fig. 1C). The cytolytic activity of TCR-TMART-1 was inhibited by increasing percentages of tumor cells expressing PD-L1 (Fig. 1D), demonstrating a dose-dependent suppression of PD-L1 on TCR-T cell cytotoxicity. PD-L1 also dose-dependently suppressed the secretion of Granzymes (Fig. 1E) and pro-inflammatory cytokines, including TNF α (Fig. 1F), IFNγ and IL2 (Fig. 1G), in Tnull and TCR-TMART-1. Altogether, PD-L1-mediated immune suppression modulated cytotoxicity and cytokine secretion of MART-1-specific TCR-T cells.
Single-cell level analysis of T cells responding to peptide-pulsed MEL-526 cells
Single-cell transcriptome profiling was performed using a negative pressure orchestrated DNBelab C4 system [31]. Transcriptome profiling of a total of 20888 cells from four conditions was obtained after filtering out cells with low quality (Fig. 2A). To investigate the intrinsic T cell heterogeneity, unsupervised clustering was performed (Fig. 2B). T and tumor cells were identified by the expression of classic cell type markers, including PTPRC, CD3D, CD3G, CD3E, CD8A, CD8B, TRAC, TRBC1, and TRBC2 for T cells and MAGEA 4 for MEL-526 cells (Fig. 2C). Based on the expression of signature genes, T cells were composed of clusters 1, 3, 5, 6, 7, 9, 10, 11, 12, 14, 15, 16, and 19 (Fig. 2B). Exogenous TCRMART-1 was detected in cluster 6, 7, 11, 15, 16, and 19, but very little in cluster 10 and 12 (Fig. S1E). Furthermore, differentially expressed genes (DEGs) and known functional markers indicated the clusters of naïve, proliferating, early activated, cytotoxic, and exhausted CD8+ T cells (Fig. 2B).
DEG analysis further identified tumor cell clusters 0, 2, 4, 8, 13, 17, and 18 that showed high expressions of S100A6, MAGEA 4, and HSPB1 as well as chemokines such as CXCL10 and CXCL11 (Fig. 2D). CXCL10 and CXCL11 recruit T cells and promote antitumor activity [32, 33]. For T cell clusters, CXCR4 and early activation marker CD69 were upregulated in cluster 1. Cluster 3 was more similar to cluster 5 and the expression of SELL, IL7R, and TCF7 was upregulated, indicating a naïve phenotype. Expression of classic cytotoxic genes including GZMK, NKG7, CST7, GNLY, and GZMA were increased in cluster 9 and 12 while the expression of cell proliferation gene MKI67 was upregulated in cluster 14. Expression of interleukins IL5, IL2, and IL3 as well as of T cell activation and cytotoxicity genes such as CSF2, XCL2, XCL1, IFNG, and GZMB were upregulated in the remaining T cell clusters (Fig. 2D). According to the above characteristics (Fig. 2D), clusters shared similarities with each other were grouped together (Fig. 2B) and their DEGs were showed in Fig. S2A.
To understand T cell state transitions, an unsupervised inference method Monocle 2 [34] was applied to construct the potential development trajectories of ten T cell clusters (cluster 14, 19, and 16 were excluded due to their distinct expression of MKI67 or mitochondrial genes). Cells from all clusters aggregated according to expression similarities to form a relative process in pseudotime, which began with cluster 3 and 5 (IL7R+RIPOR2, naïve cells), followed by cluster 9 (GNLY+GZMK) and 1 (CD69+GZMA) (Fig. 2E). Cluster 6, 7, 10, 15 (TNFRSF9+GZMB) and 19 (MKI67+TNFRSF9) activated and cytotoxic cells were located in the opposite directions with cluster 12 (GNLY+LAG3) in the pseudotime trajectory plot, demonstrating diverse functions of these cells. According to the trajectory analysis, CD8+ exhausted T cells were more closely linked to intermediate populations cluster 1 and 9 marked by GZMA and GZMK signatures, respectively than to the effector populations (Fig. 2E), consistent with a previous study [35]. Moreover, two main categories of genes were identified in this pseudotime process, one increased and the other one decreased (Fig. 2F). Upregulated genes included T cell activation and cytotoxicity markers, such as XCL2, XCL1, TNFRSF9, and NFKB1, and cell death associated genes, such as FASLG and BCL2L11. Downregulated genes encompassed naïve marker genes such as LEF1, TCF7, IL7R, and SELL (Fig. 2F). Taking together, while T cells differentiated into cytotoxic and exhausted populations, the expression of genes related with T cell activation and cytotoxicity was upregulated and the expression of naïve marker genes was downregulated in this pseudotime axis.
Tumor PD-L1 affected Tnull and TCR-TMART-1 differently at the transcriptional level
To reveal the structure of the overall T cell population, T cells were divided into Tnull and TCR-TMART-1 and their cluster compositions were investigated. Cluster composition of the control (Ctrl) group was different from that of groups stimulated by tumor cells (Fig. 3A). After stimulation, cluster 1 (CD69+GZMA), 6&7&10&15 (TNFRSF9+GZMB), 11 (IL2RA+GZMB), 12 (GNLY+LAG3), and 16 (REL+MT) were increased compared to those of the Ctrl group (Fig. 3A), indicating the percentage of cytotoxic and exhausted T cells were increased upon antigen stimulation. With the increased percentage of tumor cells expressing PD-L1, only cluster 11 was increased in Tnull, whereas clusters 1, 11, and 12 were increased and clusters 6&7&10&15 were decreased in TCR-TMART-1. (Fig. 3A). These results implied that TCR-TMART-1 were more sensitive than Tnull to increasing levels of tumor PD-L1, which also reduced the percentage of activated and cytotoxic TCR-TMART-1.
Results of gene expression (Fig. 3B, S3A) were consistent with the result of cluster composition in each group (Fig. 3A). TCR-TMART-1 were affected more than Tnull by increased tumor PD-L1 (Fig. 3B, S3A, 3C), which inhibited the expression of T cell activation and cytotoxicity genes, including XCL2, XCL1, IL2RA, GZMB, and NKG7 in TCR-TMART-1 (Fig. 3B). Moreover, there was no significant difference in gene expression between TCR-TMART-1 targeting PD-L1int and PD-L1high (Fig. 3C). Enriched signaling pathways were then analyzed. Different signaling were enriched in Tnull and TCR-TMART-1 after encountering tumor cells (Fig. S3B). Compared to TCR-TMART-1 targeting PD-L1int and PD-L1high, TCR-TMART-1 targeting PD-L1low enriched metabolic and vesicle lumen related signaling (Fig. 3D). However, distinct pathways including membrane region, membrane microdomain and raft were enriched in TCR-TMART-1 targeting PD-L1int compared to that targeting PD-L1high (Fig. 3D). In addition, gene set variation analysis (GSVA) revealed that functional subtypes of Tnull and TCR-TMART-1 populations responded to tumor cells differently. Tnull were enriched with cytotoxic and terminally differentiated cells whereas TCR-TMART-1 were enriched with exhausted and anergic cells (Fig. 3E).
Tumor PD-L1 expression resulted in various cellular and molecular responses in T cells
To correlate phenotypes other than cluster composition to T cell cytotoxicity (Fig. 1D), the expression of cytokines, chemokines, cytokine and chemokine receptors, and transcription factors was analyzed. With increased tumor PD-L1, expression of activation and cytotoxicity marker genes including IFNG, TNFSF9, TNFSF14, CSF2, and IL2, were downregulated in Ttotal (Fig. 4A), consistent with results of cytokine secretion assays (Fig. 1F, 1G). In addition, expression of anti-inflammatory cytokines, including IL10, IL13, and IL19, were upregulated in TCR-TMART-1 stimulated with PD-L1high (Fig. 4A, S4A). Although the expression of pro-inflammatory cytokines such as IL12A and IL5 was increased in TCR-TMART-1 targeting PD-L1high (Fig. 4A), the results overall suggested the domination of anti-inflammatory cytokines over pro-inflammatory cytokines, resulting in the inhibition of T cell function. In line with the cytokine expression pattern, the expression of cytokine receptors related with T cell activation, including TNFRSF9, IFNGR1, and IL2RA, was upregulated in TCR-TMART-1 stimulated with PD-L1low while the expression of IL13RA2 and IL13RA1 was increased in TCR-TMART-1 stimulated with PD-L1high (Fig. 4B). Overall, the production of proinflammatory cytokines in TCR-TMART-1 was dose-dependently inhibited by the expression of PD-L1 in tumor cells and function of T cell targeting PD-L1high was inhibited by the production of anti-inflammatory cytokines.
For chemokines (Fig. 4C), more were expressed in TCR-TMART-1 than in Tnull after antigen stimulation. In Ctrl group, CCL25 was highly expressed in TCR-TMART-1 but the expression of the CCL25 receptor gene CCR9 was not detected. When cultured with PD-L1low, CCL8 was highly expressed in TCR-TMART-1 (Fig. 4C) whereas CCR2, CCR3, CCR5, and, especially, CCR1 that encoded CCL8 receptors were upregulated in Tnull (Fig. 4D), indicating that TCR-TMART-1 could recruit Tnull through chemokine secretion. As for PD-L1int, CCL7 was significantly expressed in TCR-TMART-1, but CCR1, CCR2, and CCR3 encoding CCL7 receptors were almost expressed equally low in both Tnull and TCR-TMART-1. In PD-L1high, CXCL2 and CXCL9 were highly expressed in TCR-TMART-1, but CXCR2 or CXCR3 encoding their corresponding receptors were not detected or only weakly expressed (Fig. 4D). In conclusion, TCR-TMART-1 stimulated with PD-L1low effectively attracted and activated Tnull, consistent with its greatest cytotoxicity.
Unique expression pattern of transcription factors (TFs) was also discovered in Tnull and TCR-TMART-1 populations. The expression of ZEB2, RBPJ, NFKB1, GATA3, IRF4 and STAT3, which are important for TCR signaling production and transduction and T cell activation and differentiation [36] [37], were higher in TCR-TMART-1 cultured with PD-L1low (Fig. 4E). Overall, the expression of IRF4, NFKB1, and RBPJ was decreased in TCR-TMART-1 (Fig. S4B) whereas the expression of EOMES in Tnull was progressively downregulated (Fig. S4C) with increasing percentages of PD-L1-expressing tumor cells. The results indicated the expression profile of transcription factors in T cells were also affected by tumor PD-L1 expression.
Colocalization of tumor and TCR-TMART-1 increased immune cell cytotoxicity
Interestingly, some genes expressed by immune cells were detected in tumor cell clusters, including XCL1, XCL2, GZMB, IL32, IL13, IL2RA, CXCL9, CD2, CSF2, and IL3 that are associated with T cell activation or cytotoxicity (Fig. 4F). This observation indicated that T cells and tumor cells were close enough or in contact when T cells were activated by tumor cells [38], thus were separated in the same droplet for scRNA-seq. To answer if this phenomenon accounted for the difference of T cell cytotoxicity caused by different percentages of PD-L1 expressing tumor cells, the above genes were assessed in tumor populations. Expression of cytotoxic genes XCL1, XCL2, and GZMB was highest in both Tnull and TCR-TMART-1 cultured with PD-L1low (Fig. 4G), in line with the highest cytotoxicity of T cells in PD-L1low (Fig. 1D). Moreover, the expression of T cell marker genes including CD3D, CD3E, CD3G, CD8A, CD8B, PTPRC, TRAC, TRBC1, and TRBC2 was also detected in tumor populations (Fig. 4H), confirming the presence of T cells in tumor populations.
The expression of the tumor cell marker gene MAGEA4 was also detected in T cell populations, while no MAGEA4 expression was detected in Ctrl group (Fig. 4I), suggesting the specificity of MAGEA4 expression in tumor cells. In addition, the expression of MAGEA4 was higher in TCR-TMART-1 than in Tnull in each group and was the highest in TCR-TMART-1 cultured with PD-L1low, consistent with immune cell cytotoxicity (Fig. 1D).
Increased expression of tumor PD-L1 enhanced T cell death
To detect the impact of PD-L1 expression on tumor and immune cell death, gene sets of cell death pathways, including apoptosis, necrosis, autophagy, pyroptosis, and ferroptosis, were used for GSVA analysis. We first analyzed tumor cells after they were cocultured with T cells for 24 h. Cell death pathways, especially necrosis and autophagy, were most enriched in PD-L1int (Fig. 5A), suggesting a non-linear correlation between PD-L1 expression and tumor cell death at the transcriptional level. When tumor populations were seperated into PD-L1-expressing or PD-L1-non-expressing (nonPD-L1) subsets, cell death pathways were most enriched in PD-L1-expressing cells of PD-L1int (Fig. 5B). The expression of the key members of these cell death pathways was further analyzed (Fig. 5C). Apoptotic genes, including TRADD, BID, FAS, FASL, autophagy gene BECN1, and ferroptosis genes GLS2, VDAC3, CARS, GPX4, HSPB1, NFE2L2, were upregulated in PD-L1-expressing tumor cells of PD-L1low (Fig. 5C, S5A, S5B), providing a possible reason for the strongest cytotoxicity observed in PD-L1low (Fig. 1D).
To further assess the difference between tumor populations of each group, GO analysis were performed in tumor cells. PD-L1-expressing populations in each group had similar enriched signaling pathways, including protein localization or targeting to endoplasmic reticulum (ER) and antigen processing and presentation pathways (Fig. 5D). In contrast, pathways enriched in nonPD-L1 populations varied from each other and from PD-L1-expressing subsets (Fig. 5D). To gain insight into whether immune cell death would be affected by tumor PD-L1, cell death pathways (Fig. 5E) and gene expression (Fig. S5C) were analyzed in Tnull and TCR-TMART-1. Cell death pathways were more enriched in TCR-TMART-1 than in Tnull in each group and the enrichment of cell death signaling was positively correlated with the level of tumor PD-L1 while negatively correlated with T cell cytotoxicity (Fig. 1D). These results suggested that tumor PD-L1 enhanced Tnull and TCR-TMART-1 cell death, thus inhibited T cell function.
Tumor PD-L1 levels correlated positively with PD-L1 expression and negatively with PD-1 expression in T cells
PD-L1 has been reported to interact with CD80 in cis to selectively suppress the CD80:CTLA4 interaction but not the CD80:CD28 interaction [9]. To reveal how the PD-L1 network worked here, the expression of PD-L1, PD-1, CD80, CTLA4, and CD28 was assessed. With the increase of tumor PD-L1 (Fig. 6A, S6A), the proportion of PD-L1+ and the level of PD-L1 in Tnull and, especially, in TCR-TMART-1 gradually increased (Fig. 6B, S6B). PD-1 expression was highest in TCR-TMART-1 targeting PD-L1low (Fig. 6C, S6C) and decreased with increasing tumor PD-L1. It implied the strongest T cell activation induced highest PD-1 expression with lowest PD-L1 expression. Since CD80 expression in tumor cells (Fig. 6D) or Ttotal (Fig. 6E) was much lower than PD-L1 expression (Fig. 6A, 6B), CD80 would entirely bind to PD-L1 in cis, rather than to CTLA4 (Fig. 6F) in trans. PD-L1:CD80 cis-heterodimer could then trigger co-stimulatory receptor CD28 in Tnull and TCR-TMART-1 (Fig. 6G). Considering the dramatic difference in expression levels of PD-L1 and CD80, the dominant signaling was the interaction between PD-L1 and PD-1 under the circumstances. Overall, the tumor PD-L1 level positively correlated with PD-L1 expression while negatively correlated with PD-1 expression on T cells, perfectly demonstrating that PD-1 expression is an activation marker for T cells [39].
Blockade of PD-1 has been reported to lead to a compensatory upregulation of other checkpoint pathways[40], thus we analyzed whether increased tumor PD-L1 affected other checkpoint molecules. The expression of inhibitory checkpoint molecules (ICMs), including ADORA2A, BTLA, CD160, and PDCD1, was downregulated while the expression of CD276 and IDO1 was upregulated with the increased tumor PD-L1 (Fig. 6H). Simultaneously, the expression of stimulatory checkpoint molecules (SCMs) such as ICOS and TNFRSF9 was highest in TCR-TMART-1 targeting PD-L1low (Fig. 6I), consistent with its greatest cytotoxicity. Taking together, PD-L1 expression on tumor cells affected the expression of other inhibitory and stimulatory checkpoint molecules, which together impacted T cell function.
Expression of PDCD1 and CD274 correlated with COVID-19 severity
The PD-1/PD-L1 signaling plays an essential role not only in regulating tumor immune responses but also in balancing homeostasis and tolerance in virus infection [41]. The current pandemic coronavirus disease 2019 (COVID-19) is caused by infection of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) [42-44], where the role of PD-1/PD-L1 is currently unclear. Thus it is necessary to investigate how PD-1/PD-L1 signaling works during COVID-19 progress in order to deal with it. Publicly available data of bronchoalveolar cells from three moderate (M1-M3) and six severe (S1-S6) COVID-19 patients, and four healthy controls (HC1-HC4) were collected for analysis (66630 cells, Table S1) [45]. 31 clusters were identified by classical signature genes according to the reference (Fig. 7A) [45]. Expression of PD-1 and PD-L1 was first analyzed at the patient group level in different cell subpopulations, four trends of their expression dynamics were observed (Fig. 7B, 7C). PDCD1 expression was gradually elevated in T cell, B cell, myeloid dendritic cells (mDCs), and macrophages from HC to mild cases then to severe patients. In the 2nd trend, PDCD1 expression was specifically increased in plasma cells and epithelial cells in severe patients but not in mild patients (Fig. 7B). For the 3rd trend, PDCD1 expression was upregulated in mild patients but slightly reduced in severe patients in NK and plasmacytoid dendritic cells (pDCs) (Fig. 7B). No expression of PDCD1 was detected in mast cells and neutrophils in the 4th trend (Fig. 7B). The CD274 expression in macrophages, mast cells, pDC, and T cells (1st trend) correlated well with COVID-19 severity and was specifically increased in plasma cells of severe patients (2nd trend) (Fig. 7C). When analyzed at the individual level, expression of PDCD1 and CD274 was also elevated in mild and severe patients (Fig. S7A, S7B). Overall, PDCD1 expression in T cells, B cells, mDCs, and macrophages and CD274 expression in macrophages, mast cells, pDC, and T cells correlated well with COVID-19 severity. Furthermore, PDCD1 and CD274 expression was specifically increased in epithelial and plasma cells of severe patients. Inflammatory signaling participates in modulating PD-L1 expression, particularly, STAT1, which can be activated by IFNγ or interleukin 6 (IL-6), is a crucial regulator for PD-L1 expression [46, 47]. Furthermore, plasma IFNγ level [43] and the IL-6 level in bronchoalveolar lavage fluid (BALF) [45] were reported to be increased in COVID-19 patients. Consistently, STAT1 was found upregulated in both mild and severe patients (Fig. 7D), suggesting increased CD274 expression might at least partly resulting from increased STAT1 level in COVID-19 patients.
To further elucidate the immune checkpoint landscape of COVID-19 patients, expression of classical inhibitory and stimulatory checkpoint molecules was assessed. For ICMs, expression of CD160, CD244, PD-1, BTLA, TIGIT, LAG3, KLRG1, and ADORA2A were increased in mild patients compared to HC and severe patients while expression of CTLA4, HAVCR2, IDO1, and CD276 were highest in severe patients (Fig. 7E). Regarding SCMs, expression of TNFRSF9, CD28, ICOS, and CD27 were elevated in mild patients in comparison to HC and severe patients while expression of TNFRSF18 and CD40 were highest in severe patients (Fig. 7F). When analyzed at the cell subpopulation level, unique expression patterns of ICMs and SCMs were demonstrated in each cell subpopulation (Fig. S7C, S7D).
Discussion
It has been well documented that the efficacy of CPI in treating tumors is affected by the PD-L1 level, and the relative amounts of PD-1 and its interactors in tumors [9]. However, the molecular mechanism by which different levels of PD-L1 on tumors affect the therapeutic efficacy of TCR-T cell therapy remains unclear.
There are few studies about the effect of PD-L1 expression levels on TCR-T cell function. Our study provides an insight about TCR-T cell response to different proportions of tumor cells expressing PD-L1 at the single-cell level. The results of cell-based assays revealed that higher proportion of PD-L1+ tumor cells more strongly inhibited T-cell function (Fig. 1D-1G). Single-cell transcriptome profiling demonstrated the inhibition from different aspects, including cell differentiation (Fig. 3), secretion of cytokines and chemokines (Fig. 4A, 4C), colocalization of tumor and immune cells (Fig. 4F-4I), tumor and immune cell death (Fig. 5), as well as the whole PD-L1 network (Fig. 6A-6G).
TCR-TMART-1 were more vulnerable than Tnull when targeting increasing proportion of PD-L1-bearing tumor cells (Fig. 3B-3C). It indicates that TCR-T therapy could be co-administrated with PD-L1/PD-1 interference to obtain better anti-tumor efficacy. Clinical trials with TCR-T cells armed with a PD-1 antagonist are ongoing (NCT04139057, NCT03578406). The result also implies that TCR-T cells will benefit from elimination of their PDCD1, such as by using CRISPR-based approaches, to protect themselves against PD-L1-mediated inhibition [48].
In addition to pro-inflammatory cytokines, the expression of anti-inflammatory cytokines, including IL10, was upregulated in TCR-TMART-1 targeting PD-L1high (Fig. 4A, S4A). It was reported that IL-10 levels in serum and in ascites were increased after treatment with PD-1 inhibitor, implicating that IL-10 participated in undermining the efficacy of anti-PD-(L)1 therapies [49]. Thus, combined blockade of PD-L1 and IL10 may further enhance T-cell immunity [50, 51]. Interestingly, colocalization of T and tumor cells was detected and correlated negatively with the expression of tumor PD-L1 (Fig. 4F-4I). Colocalization of T and tumor cells was supported by the imaged interaction of T cells and APCs [38]. Therefore, tumor PD-L1 may inhibit T cell cytotoxicity through hindering the colocalization and interaction of antigen-specific T cells and tumor cells.
Various cell death pathways were involved in tumor and T cell death (Fig. 5, S5) and the enrichment of cell death signaling in T cells correlated positively with tumor PD-L1 expression level. This is consistent with a former study, where CD8+ T-cell apoptosis was promoted by PD-1 and PD-L1 upregulation [52], implying T cell death caused by PD-L1 signaling is one of the tumor evasion pathways.
Furthermore, the elevation of tumor PD-L1 expression dose-dependently increased the expression of PD-L1 in T cells (Fig. 6B, S6B), while PD-1 expression was dose-dependently decreased (Fig. 6C, S6C). PD-1 has been reported to induce apoptosis of antigen-specific T cells [53], but here tumor PD-L1 seems to play a dominant role in promoting effector T cell death. Moreover, consistent with that PD-1+ CD8+ T cells were functional cytotoxic T cells that targeted tumors and experienced exhaustion [54], PD-1 expression in T cells correlated positively with cytotoxicity in our study.
Lastly, since COVID-19 is pandemic and threatening thousands of people’s life, it is urgent and essential to investigate the molecular mechanism of the immune pathogenesis of the disease. Compared to healthy controls, PDCD1 expression in T cells, B cells, mDCs, and macrophages (Fig. 7B) and CD274 expression in macrophages, mast cells, pDCs, and T cells (Fig. 7C) were upregulated in COVID-19 patients, and correlated well with COVID-19 severity. Moreover, expression of PDCD1 and CD274 was specifically increased in plasma cells of severe patients (Fig. 7B, 7C), which could serve as a biomarker for prognosing the severity of COVID-19. Many clinical trials for treating COVID-19 are ongoing. Among them, one clinical trial uses PD-1 monoclonal antibody to block PD-1 in COVID-19 patients (NCT04268537). Based on our results, PDCD1 expression was dramatically upregulated in T cells and macrophages especially in severe patients (Fig. 7B) and its blockade would further increase the secretion of multiple pro-inflammatory cytokines (Fig. 4A), which will enhance the cytokine release syndrome reported in COVID-19 patients and possibly associated with disease severity [42, 43], leading to further tissue damage or even more death especially in severe COVID-19 patients [55, 56]. A current study supports that checkpoint inhibitor immunotherapy is risky for severe outcomes in SARS-CoV-2-infected cancer patients, though these patients were treated with immune checkpoint inhibitors (ICI) before SARS-CoV-2 infection [57]. Furthermore, the expression of IL10 was upregulated with increasing tumor PD-L1 (Fig. 4A), indicating a role of IL-10 in keeping a microenvironmental balance. Thus, the anti-inflammatory cytokine IL-10 might protect COVID-19 patients from T cell hyperactivation, which in turn may cause a dreaded complication characterized by acute respiratory distress syndromes in COVID-19 patients [58]. In addition, lower ratio of IL6 to IL10 serum levels was reported to coincided with the recovery of pneumonia [59].
In conclusion, cell-based cytotoxicity and cytokine secretion assays together with scRNA-seq were applied to interrogate MART-1-specific transgenic T cells upon antigen-specific stimulation with different ratios of tumor PD-L1. This study provides the first comprehensive illustration of tumor PD-L1 inhibition on TCR-T cell function at the single-cell level, and reveals some common mechanisms regarding how different subtypes of TCR-T cells respond to PD-L1 inhibition. It provides valuable information about the inhibition by increased tumor PD-L1 expression on TCR-T cells which are being applied in clinical trials, and about COVID-19, whose severity correlated well with the expression of PDCD1 and CD274.
Author contributions
Q.G. and R.D. designed the project. Q.G. wrote the manuscript. Q.G. and C.C. revised the manuscript. S.L. performed the bioinformatic analysis. S.W., H.C. and Q.G. conducted the experiments related to single-cell sequencing. Q.X. and R.D. performed FACS analysis. F.W. and L.Z. assisted experiments and manuscript revision. R.D., H.C., and Q.G. performed all the other experiments and data analysis. Q.G., C.C., Y.G., and X.D. supervised the project.
Declaration interests
The authors declare no competing financial interest.
Methods
Cell lines and cell culture
HEK293T (ATCC, CRL-11268) and T2 (174×CEM.T2, CRL-1992) cell lines were purchased from ATCC, and MEL-526 (BNCC340404) cell line was purchased from BNCC. HEK293T and MEL-526 cells were cultured in DMEM (Gibco, 21063029) supplemented with 10% fetal bovine serum (Hyclone, SH30084.03HI), penicillin (100 IU/mL), and streptomycin (50 μg/mL). T2 cells were cultured in IMDM (Gibco, 12440053) supplemented with 20% fetal bovine serum (Hyclone, SH30084.03HI), penicillin (100 IU/mL), and streptomycin (50 μg/mL). CD8+ T cells were cultured in HIPP-T009 (Bioengine, RG0101302) supplemented with 2% fetal bovine serum (Hyclone, SH30084.03HI) containing IL-2 (20 ng/ml), IL-7 (10 ng/ml) and IL-15 (10 ng/ml).
Plasmid construction
TCRMART-1 sequence was identified from our previous work (data unpublished), and its constant regions were replaced by mouse TCR constant region α and β, respectively to prevent mispairing with endogenous TCR. TCR α chains and β chains were linked by P2A self-cleaving peptide. The recombinant DNA encoding TCRMART-1 was synthesized by GeneScript (Nanjing, China) and ligated into pRRLSIN.cPPT.PGK vector (Addgene, 12252).
PD-L1 cDNA ORF Clone in Cloning Vector was purchased from Sino Biological (HG10084). PD-L1 cDNA was cloned into pRRLSIN.cPPT.PGK vector (Addgene, 12252) with ClonExpress II One Step Cloning Kit (Vazyme, C112) according to the user manual.
Lentivirus production
293T cells were transfected with a mixture of interested plasmid and packaging constructs (PsPAX2 and PMD2G) as previous [60]. The culture supernatants were collected 72 h after transfection and filtered through a 0.45 uM filter. Subsequently, the supernatants were concentrated by ultracentrifugation at 35,000 rpm for 90 min. The pellet was suspended and stored at -80°C.
Generation of tumor cells expressing PD-L1
After lentivirus infection of PD-L1 lentivirus into MEL-526 cells for 2 days, PD-L1+ cells were sorted out by FACS. Different portions of PD-L1+ tumor cells were obtained by mixing wildtype and PD-L1+ MEL-526 cells.
Generation of MART-1-specific T cells
Human Peripheral Blood Mononuclear Cells (PBMCs) were isolated from the blood of HLA-A*0201-restricted healthy donors with informed consent. CD8+ T cells were purified from PBMC via human CD8 MicroBeads (Miltenyi Biotec, 130-045-201) and activated with T Cell TransAct (Miltenyi Biotec, 130-111-160). After 36-48 h, CD8+ T cells were transduced with TCRMART-1 lentivirus at MOI=25 in a 6-well or 12-well plate. Simultaneously, polybrene was added to the culture at a final concentration of 2 μg/ml to promote infection efficiency. Then the well plate was centrifuge at 800g at room temperature for 30 minutes.
Peptide synthesis
MART-1 originated peptide ELAGIGILTV (HLA-A*0201) was synthesized by GenScript (Nanjing, China) with a purity of ≥ 99.0%. Peptides were dissolved with 100% dimethyl sulfoxide (DMSO; Sigma-Aldrich, D5879-500ML) at the concentration of 10 mg/ml, and were stored at -20□.
TCR-T cell stimulation with target tumor cell
TCR-T cells and MEL-526 cells (5×10^5 cells/ml concentration, in 200 μl) pulsed with peptide (final concentration 10 μg/mL) or not were incubated for 24 h in a round bottom 96-well plate. Afterwards, the co-culture was subjected to scRNA-seq. Unstimulated TCR-T cells (5×10^5 cells/ml) were incubated for 6 h alone before subjected to scRNA-seq.
Intracellular staining
Cells were perforated and fixed using Cytofix/Cytoperm kit (BD Pharmingen, 554715).The antibodies used are as followed: Allophycocyanin(APC)-anti-HLA-A2 antibody (eBioscience, 17-9876-42), Phycoerythrin (PE)-anti-human CD8a antibody (eBioscience, 12-0086-42), APC-anti-human CD274(PD-L1) antibody (BD Pharmingen, 563741), PE-anti-human CD279(PD-1) antibody (Biolegend, 367404), PE anti-mouse TCR β chain Antibody (Biolegend, 109207), APC Anti-human IFN γ (eBioscience, 502512), PE-anti-human Granzyme B (BD Pharmingen, 561142), APC anti-human CD107a (Biolegend, 328620), PE-anti-Ki67 antibody (Abcam, ab270650).
Cell killing assays
Target cells were labeled with Carboxyfluorescein succinimidyl ester (CFSE; Invitrogen) and co-cultured with 50% TCR-T cells at E:T ratio of 1:2. After 24 h, cells were collected and stained with PI and subsequently detected by FACS.
Cytokine secretion measurement
The secretion of TNF-α, granzyme A, and granzyme B by T cell were evaluated using BDTM cytometric bead array (CBA) system. Tnull or TCR-TMART-1 cells were co-cultured with MEL-526 cells pulsed with peptide or not and supernatants were collected 24 h later. CBA assay was performed according to the instruction manual.
Statistical analysis
Data analyzing was preformed using PRISM 6 (GraphPad Software) and RStudio. *P<0.05, **P<0.005, ***P<0.001. Values are presented as mean Standard deviation (SD). Error bars represented the SD.
ScRNA-seq
Single-cell 3’ mRNA transcriptome profiling was performed using a negative pressure orchestrated DNBelab C4 system according to the workflow [31].
ScRNA-seq data preprocessing
For all the samples, the iDrop Software Suite (v.1.0.0) was used to perform sample de-multiplexing, barcode processing and single-cell 3’ unique molecular identifier (UMI) counting with default parameters. Cleaned reads were then aligned onto the complete UCSC hg38 human genome by splicing-aware aligner STAR with default parameters. Valid cells were automatically identified based on the UMI number distribution of each cell. The filtering criteria were used to obtain high-quality single cell: the number of genes in each cell in the range of 400 to 6000, the ratio of mitochondrial genes less than 0.2, and the number of UMI more than 1000.
Unsupervised clustering
The expression matrix obtained in the above steps was used as input to Seurat v. 3 to perform batch effect correction, standardization, dimensionality reduction, and clustering. First, the “LogNormalize” function was applied to normalize the data. Second, the “vst” method in the “FindVariableFeatures” function was used to detect variable genes, and the top 3000 variable genes were selected for downstream analysis. Third, the “FindIntegrationAnchors” and “IntegrateData” functions were used to correct batch effects. Fourth, the top 3000 variable genes were applied for PCA dimensionality reduction. The UMAP was performed on the top 20 principal components for visualizing these cells. At the same time, graph-based clustering was performed on the PCA-reduced data for clustering analysis with Seurat v.3. The resolution was set to 1 to obtain a most representative result.
Differential gene expression analysis
We applied the FindMarkers to differential gene expression analysis. For each cluster of T cells and tumor cells, DEGs were generated relative to all of the other cells. A gene was considered significant with adjusted P< 0.05 and logFC > 0.25. To compare DEGs across CD8+ T cells and tumor cells under different experimental conditions, the limma method was used with the parameters recommended in the user guide for analysis. Then DEGs were identified when met these criteria: FDR adjusted p value of F test < 0.01.
Developmental trajectory inference
The Monocle (version 2) algorithm with the signature genes of different functional clusters was applied to order CD8+ T cells excluding clusters expressing proliferating or mitochondrial genes in pseudo time. UMI value was first converted into normalized mRNA counts by the “relative2abs” function in monocle and created an object with parameter “expressionFamily = negbinomial.size” according to the Monocle tutorial. Then the CD8+ T cell differentiation trajectory was determined by the default parameters of Monocle.
Gene set enrichment analysis
Gene Ontology (GO) enrichment analysis was performed on the differential genes of each cluster, and the results were used for cell type definition. The “enrichGO” function in the “clusterProfiler” package to perform GO analysis using the corresponding default parameters. Pathways with the q value <0.05 corrected by FDR were used for analysis.
GSVA
GSVA was used to identify the molecular phenotype of each cluster with the normalized UMI data. The average normalized expression across T cell clusters was first obtained. Then, GSVA scores of gene sets for different clusters were calculated. GSVA values were plotted as a heatmap using R package “pheatmap”.
Data availability
The data that support the findings of this study have been deposited into CNGB Sequence Archive (CNSA: https://db.cngb.org/cnsa/) of CNGBdb with accession number CNP0001109.
Ethics approval and consent to participate
The study was approved by the Institutional Review Board on Bioethics and Biosafety of BGI. A written information consent was regularly obtained from all donors.
Figure legend
Figure S1. TCRMART-1 construction. (A). Schematic design of TCRMART-1. (B) Expression of TCRMART-1 on CD8+ T cells transfected by lentivirus before (middle) and after (right) cell sorting. (C) Killing of T2 cells by Tnull and TCR-TMART-1 after co-incubation for 6 h at E:T ratio of 1:1. (D) PD-L1 was over expressed in MEL-526 cells. (E) Violin plot showing the expression of TCRMART-1 in different cell clusters.
Figure S2. Volcano plot showing differentially expressed genes across T cell clusters. Each red/blue dot denotes an individual upregulated/downregulated gene (logFC >= 1 and p.value < 0.01).
Figure S3. Gene expression and signaling pathways in T cells responding to different expression of tumor PD-L1. (A) The expression of DEGs in Tnull targeting PD-L1low-, PD-L1int-, PD-L1high-expressing tumor cells (left) and the expression of DEGs in TCR-TMART-1 (right). (B) The bubble plot showing the top 5 pathways in TCR-TMART-1 targeting PDL1low-, PDL1int-, PDL1high tumor cells.
Figure S4. Expression of cytokines and transcription factors in T cells. (A) The bar plot showing the average expression level of IL10, IL13, and IL19 in Tnull and TCR-TMART-1. (D) The bar plot showing the average expression level of IRF4, NFKB1, and RBPJ in Tnull and TCR-TMART-1. (E) EOMES expression in Tnull and TCR-TMART-1.
Figure S5. Expression of cell death associated genes. (A) The proportion of cells expressing TRADD, BID, FAS, FASLG, and BECN1 in tumor cells. (B) The proportion of cells expressing GLS2, VDAC3, CARS, GPX4, HSPB1, and NFE2L2 in tumor cells. (C) Heatmap showing the expression of cell death associated genes in Tnull and TCR-TMART-1.
Figure S6. Transcriptional profiles of CD274 and PDCD1 in T cells and tumor cells. (A) Bar plot showing the average expression level (left) and percentage (right) of CD274 in T null and TCR-TMART-1. (B) Bar plot showing the average expression level (left) and percentage (right) of CD274 in MEL-526 cells. (C) Bar plot showing the average expression level (left) and percentage (right) of PDCD1 in T null and TCR-TMART-1. (D) Bar plot showing the average expression level (left) and percentage (right) of PDCD1 in MEL-526 cells.
Figure S7. Immune profiling of checkpoint molecules in COVID-19 patients. (A) PDCD1 expression was upregulated in mild and severe COVID-19 patients compared to HC. Error bars represent ± standard error. (B) CD274 expression was upregulated in mild and severe COVID-19 patients compared to HC. (C) Heatmap showing the expression pattern of inhibitory checkpoint molecules in different cell subsets from HC, mild and severe COVID-19 patients. (D) Heatmap showing the expression pattern of stimulatory checkpoint molecules in different cell subsets.
Acknowledgments
We sincerely thank the support provided by China National GeneBank and Guangdong Provincial Key Laboratory of Genome Read and Write (No. 2017B030301011). This research was funded by Technology and Innovation Commission of Shenzhen Municipality, grant number JCYJ20170817150015170 and JSGG20180508152912700. This manuscript was edited by Life Science Editors.
Footnotes
↵5 Lead Contact