Regulatory mechanisms of immune checkpoints PD-L1 and CTLA-4 in cancer

Immunotherapy, mediated by immune checkpoint inhibitor (ICI), represents a turning point in the anti-tumor treatment of various cancer types in recent years. So far, several immune checkpoint targets, like T cell immunoglobulin and mucin-domain containing-3 (TIM-3), lymphocyte activation gene-3 (LAG-3), T cell immunoglobulin and ITIM domain (TIGIT), indoleamine 2, 3-dioxygenase 1 (IDO1), and V-domain immunoglobulin suppressor of T cell activation (VISTA), have been identified [1, 2], for which ICIs were designed and have presented promising results (Table S1). However, only ICIs targeting cytotoxic T lymphocyte antigen 4 (CTLA-4) and programmed cell death protein 1 (PD-1)/ programmed cell death-ligand 1 (PD-L1) have been FDA-approved and widely used. CTLA-4 is a classical immune checkpoint molecule, acting as a CD28 homolog with a stronger binding affinity to its receptor B7-1 (CD80) or B7-2 (CD86). As the first available ICI, the humanized CTLA-4 antibody ipilimumab has revolutionized clinical cancer care and prolonged the 10-year survival for metastatic melanoma [3]. PD-1 is encoded by PDCD1, and its ligands, PD-L1 and PD-L2, are encoded by CD274 and CD273, respectively. Overall, PD-1/ PD-L1 axis has been more widely explored and applied than PD-1/ PD-L2 axis. CTLA-4/B7 and PD-1/ PD-L1 axis regulate physiological immune homeostasis, downregulate inflammatory responses, and presumptively facilitate immune evasion of cancer cells [4‐6]. Blockade of PD-1/ PD-L1 axis and CTLA-4/B7 axis promoted overall survival in multiple cancer types with significantly increased response rates of therapy [3, 7]. Blockade of PD-1/ PD-L1 axis had decreased side effects compared with blockage of CTLA-4 [8, 9]. Notably, antibodies blocking the PD-1/PD-L1 axis have been approved as first-line or second-line treatment modalities in a wide range of cancers [6, 10]. Further, combined anti-CTLA-4/B7 and anti-PD1/PD-L1 therapy has been applied to cancer treatment [11]. Nevertheless, despite promising clinical management results, only a minority of patients receiving ICIs targeting PD-L1 or CTLA-4 experience longer overall survival. Besides, ICIs failed to show significant therapeutic benefit in clinical trials of colorectal carcinoma (CRC) (confirmed responses in 7 of 84 patients), non-small cell lung cancer (NSCLC) (confirmed responses in 5 of 28 patients), melanoma (confirmed responses in 9 of 22 patients), and glioblastoma (confirmed response rate: 23.1%) [12‐14].

CTLA-4, primarily expressed by T cells, binds to CD80/CD86 and prevents the stimulatory signaling of T cell proliferation provided by binding of CD28 with CD80/CD86 during the priming phase [15]. In contrast, the immune evasion process led by PD-1/ PD-L1 axis is primarily attributed to the overexpressed PD-L1 on cancer cells which binds to PD-1 expressed on antigen-stimulated T cells and inhibited the activity of PI3K/AKT and Ras/MEK/ERK signaling pathways. Consequently, this impairs the proliferation, differentiation, and activation of T cells during the effector phase [16]. Besides tumor cells, PD-L1 could be expressed on tumor-infiltrating lymphocytes (TILs), including tumor-associated macrophages (TAMs), dendritic cells (DCs), regulatory T cells (Tregs), cancer-associated fibroblasts (CAFs), mast cells, and tumor stroma cells (astrocytes, vascular endothelial cells, keratinocytes) [4]. Therefore, PD-L1 overexpression on tumor cells, TILs, or tumor stroma cells inevitably triggers T cell exhaustion, thereby creating an immunosuppressive microenvironment and promotes tumor progression [17]. Correspondingly, PD-L1 amplification has commonly been used as a positive predictive biomarker to blockade the PD-1/ PD-L1 axis in malignancies, including melanoma, NSCLC, and bladder cancer [4, 18, 19]. Multiple regulators influence PD-L1 expression at DNA, RNA, protein level, and drug intervention via direct regulation. Further, the expression level of PD-L1 is predicted and indirectly regulated by different biomarkers. Similarly, CTLA-4 expression, specifically regulated by its localization within the cell, regulates anti-cancer responses [6]. Therefore, a comprehensive summary of PD-L1 and CTLA-4 regulators and their common regulatory mechanisms could help identify patients with favorable response to anti-PD-L1 and anti-CTLA-4 treatments and promote their clinical management (Fig. 1).

The gene regulation can be roughly categorized into three levels including transcription level, post-transcription level, and post-translation level. At different levels of gene regulation, there are different classical regulators. Therefore, we first designed a frame structure for PD-L1 and CTLA-4 regulation at three levels. Then, we carefully searched for the relevant references from PubMed illustrating PD-L1 regulation or CTLA-4 regulation from different perspectives using the following keywords such as inflammatory signaling (IL6, TGF-b, IFN-I), oncogenic signaling (EGFR, ERK, STAT), epigenic modifier, and protein modification (ubiquitination) et.al. We also tried to review the novel points of PD-L1 regulation or CTLA-4 regulation using the keywords such as exosome and cell localization. Finally, we systematically searched for the literatures illustrating drugs or compounds that target PD-L1 or CTLA-4 and influence the expression of these two genes.

Post-translational regulation at the protein level is the eventual step where the gene expression level is modulated. Among the multiple post-translational modulators, ubiquitination, phosphorylation, glycosylation, and palmitoylation regulate PD-L1 expression.

Besides intracellular interactions between various regulators and PD-L1, PD-L1 expression is also regulated when PD-L1 is secreted as exosomes and soluble proteins at the extracellular level.

Tumor mutation burden (TMB) was positively associated with PD-L1 expression, where patients with high TMB showed improved efficacy of anti-PD-1 therapy [164]. Similarly, microsatellite instability (MSI) positively correlates with PD-L1 expression and predicts a better anti-PD-1 response [165]. Intratumor heterogeneity was also linked to PD-L1 expression [167]. Moreover, TIL abundance was confirmed to correlate with upregulated PD-L1 expression [166].

Notably, CTLA-4 is mainly localized in intracellular compartments with the minority of CTLA-4 detected on the cell surface. Therefore, summarizing the mechanisms by which CTLA-4 could be transported to the cell surface is important. Plasma membrane-associated clathrin adaptor protein 1 (AP-1) binds to the GVYVKM motif of CTLA-4 and subsequently maintains a steady level of intracellular CTLA-4 expression [235]. Likewise, after AP-2 binds to the YVKM motif of CTLA-4 or the FVKM motif of mutant CTLA-4, CTLA-4 was more localized in the cytoplasm and less accumulated on the T cell surface [236, 237]. The strength of the T cell receptor (TCR) signaling pathway determined the upregulated CTLA-4 expression at the immunological synapse [237]. Besides, the adaptor T cell receptor-interacting molecule (TRIM) binds to CTLA-4, forming a CTLA-4/TRIM/LAX/Rab8 effector complex, which in turn mediates the augment of CTLA-4 expression on the surface of T cell [238, 239]. The activity of guanosine triphosphatases (GTPases), phospholipase D (PLD)- 1, PLD-2, and ADP ribosylation factor (ARF)-1 were necessary for increased release of CTLA-4 to the surface of the T cell [240]. Lipopolysaccharide-responsive and beige-like anchor protein (LRBA) loss causes CTLA-4 degradation and suppresses CTLA-4 expression [241]. Additionally, CTLA-4 expression on the cell surface was rapidly upregulated via the upregulated intracellular calcium (Ca²⁺) [237].

As for PD-L1, multiple drugs exhibited efficacy in regulating PD-L1 expression (Fig. 4, Table 3). Metformin attenuated PD-L1 expression by phosphorylating oncogene YAP1 and inhibiting its entry in the nucleus, subsequently inducing the Hippo signaling pathway in CRC [243]. Further, Nanodrug (MS NPs), self-assembled from metformin and anticancer agent 7-ethyl-10-hydroxycamptothecin (SN38), was developed to suppress PD-L1 expression, enhancing the anti-tumor effect of combined immunotherapy and chemotherapy in breast cancer [244]. PARP inhibitor (PARPi) upregulated PD-L1 expression by inactivating GSK3β and enhancing immunosuppression in breast cancer [245]. Temozolomide (TMZ)-treated GBM was observed with upregulated PD-L1 expression [246]. Further, after the TMZ challenge, actinomycin D suppressed PD-L1 expression in GBM [246]. Pemetrexed and sildenafil suppressed PD-L1 expression in NSCLC, where the inhibitory effect was enhanced by histone deacetylase inhibitors AR42 and sodium valproate [247]. Notably, AR42 and valproate alone downregulated PD-L1 expression [247]. Disulfiram combined with Copper (DSF/Cu²⁺) upregulated PD-L1 expression by inactivating PARP1 and phosphorylating GSK3βat Ser9 in HCC [248]. DSF/Cu²⁺ also inhibited T cell infiltration while failed to hamper HCC tumor growth [248]. The Akt inhibitors ipatasertib, capivasertib, uprosertib, and MK-2206 inhibit PD-L1 expression in breast cancer [249]. Trastuzumab deruxtecan (DS-8201a), a HER2-targeting drug, enhanced PD-L1 expression and CD8+ T cell infiltration in breast cancer, which DS-8201a also had synergistic effect with anti-PD-1 therapy [250]. Verteporfin has been explored as a potential inhibitor that hampers IFN-γ-derived PD-L1 expression by disrupting the STAT1-IRF1-TRIM28 signaling cascade T-cell leukemia, B-cell leukemia, ovarian, osteoblastoma, and NSCLC [251]. PI3K inhibitor (LY294002) downregulated PD-L1 expression in gastric cancer [61]. EGFR inhibitors (cetuximab and erlotinib) or MEK inhibitor (selumetinib) suppressed MAPK signaling-mediated EGF-induced and IFN-γ-induced PD-L1 expression in LUAD [57]. In contrast, MEK inhibitor (trametinib), suppressing MAPK signaling, enhanced the IFN-γ-stimulated PD-L1 expression in breast cancer [56]. Besides, the insignificant alterations and IFN-γ-stimulated PD-L1 expression in diverse cancer cell lines after MAPK inhibitor treatment shed light on the individual regulatory role of PD-L1 in various cancer types [62, 260]. Belinostat enhanced IFN-γ production and upregulated PD-L1 expression in HCC, which Belinostat also had a synergistic effect of anti-PD-1 therapy [252]. C108 repressed G3BP2 and downregulated PD-L1 expression in breast cancer and GBM [139].

Among the multiple drugs developed for inhibiting the interaction of PD-L1 with PD-1, Pembrolizumab, Nivolumab, Atezolizumab, Durvalumab, and Avelumab have been approved by FDA. Pembrolizumab showed efficiency in regulating the interaction of PD-L1 with PD-1 in urothelial cancer, melanoma, HNSCC, and NSCLC [253]. Similarly, Nivolumab and Atezolizumab inhibited PD-L1 in NSCLC, urothelial cancer, RCC, and melanoma [253]. Durvalumab and Avelumab inhibited PD-L1 in urothelial cancer [253]. Other PD-L1 inhibitors have also been investigated in many clinical trials. Envafolimab (KN 035 or ASC 22) inhibited the interaction of PD-L1 with PD-1 and exhibited encouraging anti-tumor activity in a phase I clinical trial of advanced solid tumor [254]. Besides, the clinical efficacy and safety of BMS-936559 (MDX-1105) and CS-1001 were explored in a phase I clinical trial of advanced solid tumor [9]. Several drugs are currently undergoing clinical trials. CK-301 showed high binding affinity for PD-L1 and increased IFN-γ in a preclinical research [254]. CBT-502 (TQB-2450) is another inhibitor of PD-L1 with increased IFN-γ in melanoma [254]. Except for inhibiting PD-L1, BGB-A333 inhibited T cell apoptosis and increased T cell proliferation [254].

As for CTLA-4, Ipilimumab, a drug approved by FDA, docked to the MYPPPY motif of CTLA-4 and showed efficacy in inhibiting CTLA-4 in RCC, prostate cancer, cervical cancer, CRC, NSCLC, gastric cancer, pancreatic cancer, ovarian cancer, urothelial cancer, and melanoma [255]. Tremelimumab inhibited CTLA-4 in phase II clinical trials of HCC [256]. A synthetic peptide (p344) was demonstrated to bind to the MYPPPY motif of CTLA-4 and blocked its interaction with B7 ligands [257]. As a HDAC inhibitor (HDACi), suberoylanilide hydroxamic acid (SAHA) synergizes with tacrolimus (FK506) to increase Foxp3 and CTLA-4 expression [258]. Besides, HDAC6 inhibitor, CKD-L, increased CTLA-4 expression in Treg [259].

Immunotherapies represented by targeting PD-1/ PD-L1 and CTLA-4/B7 pathways have shown remarkable clinical efficacies against various cancer types. However, only a small proportion of patients exhibit favorable responses to anti-PD1-PD-L1 or anti-CTLA-4/B7 therapy. Overexpression of PD-L1 and CTLA-4 are recognized as a vital suppressor of anti-tumor immunity and associated with better therapy response and increased clinical benefit. Given the unprecedented predictive values of PD-L1 and CTLA-4 expression in immunotherapy, investigating the diverse regulators on PD-L1 and CTLA-4 expression that could potentially influence immunotherapy efficacy will contribute to the individualized clinical management of cancer patients.

In this review, the mechanisms of PD-L1 regulation were categorized into genetic alteration, epigenetic regulation, inflammatory signaling, and oncogenic signaling at the levels of DNA, regulation at RNA, regulation at protein, extracellular secretion, indirect regulation by biomarkers, and potential drug intervention. In summary, IFN-γ plays a pivotal role in inducing PD-L1 expression through epigenetic regulation, transcriptional regulation, posttranscriptional regulation, and extracellular secretion. In contrast, there is limited literature on the regulatory mechanisms of CTLA-4, with the available literature indicating that the regulation of CTLA-4 mainly depends on its localization within the cell. Targeting IFN-γ and regulating its localization within the cell could perceivably promote the anti-tumor immunotherapy responses. Some potential drugs targeting the PD-L1 regulators, such as IFN-γ, MAPK, and GSK3β exhibited remarkable synergistic anti-tumor effect with ICIs. Notably, the regulation of PD-L1 and CTLA-4 expressions have some similarities in HDAC, methylation of DNA at CpG sites, IL-2, STAT1, c-MYC at DNA level, and miRNA regulation at RNA level. Developing drugs targeting the mutual regulators of CTLA-4 and PD-L1 is expected to improve the therapeutic efficacy.

Further studies should be conducted to investigate more valuable PD-L1 and CTLA-4 regulators to improve the efficacy of immunotherapy and facilitate individualized cancer treatment. Moreover, some regulators with potential synergistic effect with the current ICIs should be expected to be explored in clinical trials.