Circular RNA CHST15 Sponges miR-155-5p and miR-194-5p to Promote the Immune Escape of Lung Cancer Cells Mediated by PD-L1

Background The effects of up-regulated CircCHST15 on lung cancer remained unclear. In this study, the role of CircCHST15 in lung cancer was investigated. Methods Dual-luciferase reporter verified the bioinformatics prediction that CircCHST15 targeted miR-155-5p and miR-194-5p. The correlation between CircCHST15 and PD-L1 was analyzed by Pearson analysis. CCK-8 and colony formation was performed to determine the viability and proliferation of lung cancer cells. After the lung cancer (subcutaneous-xenotransplant) model was established in mice, the T cell subtype and related cytokines in mouse tumor tissues were detected by flow cytometry and ELISA. Moreover, the expressions of CircCHST15, miR-155-5p, miR-194-5p, immune-related, and proliferation-related factors of the lung cancer cells or mice tumor tissues were detected by immunohistochemistry, RT-qPCR, or Western blot. Results CircCHST15 and PD-L1 were high-expressed in lung cancer, and the two was positively correlated. CircCHST15 targeted miR-155-5p and miR-194-5p, the later further targeted PD-L1. Lung cancer cell viability and proliferation were increased by miR-155-5p and inhibited by miR-194-5p. CircCHST15 located in the cytoplasm promoted tumor growth, down-regulated the expressions of miR-155-5p and miR-194-5p, and up-regulated the expressions of PD-L1, Ki-67, PCNA, CCL17, CCL22, IFN-γ, TNF-β, and IL-10. Also, CircCHST15 decreased the CD8+ cells in mouse blood and tumor, but increased the Tregs in mouse tumor. PD-L1 inhibitor showed an opposite effect to CircCHST15 on mouse tumors. Conclusion CircCHST15 sponged miR-155-5p and miR-194-5p to promote the PD-L1-mediated immune escape of lung cancer cells.


INTRODUCTION
Lung cancer originates from the bronchial mucosa, glands, or alveolar epithelium (1,2), and is gradually becoming one of the most challenging cancers with high mortality (1,3). According to the statistics, more than 1.8 million new cases of lung cancer and 1.6 million deaths occur annually (1,3). As early-staged lung cancer is not easily detectable, more than 70-80% of lung cancer patients have already missed the optimal chance for surgery by the time of diagnosis (1,3,4). Chemotherapy is now a main treatment for patients with advanced lung cancer, but is accompanied with various adverse side effects (5). Hence, it is necessary to seek new strategies of treatment for patients with lung cancer.
The development of cancers is dependent not only on the characteristics of cancer cells, but also on the interaction of cancer cells with the immune system (5,6). Studies showed that the microenvironment of tumor allow tumor cells to penetrate into lymphatic vessels and blood vessels, evade the immune surveillance of the body, resist the killing effect of antigenspecific T cells, thus realizing distant metastasis and growth (7). However, the progression and mechanism through which tumor cells evade immune surveillance have not been fully elucidated.
Circular RNAs (circRNAs), a category of non-coding RNAs, is featured by a covalently closed-loop structure and do not have the 5′ and 3′ ends normally contained in linear RNAs (8). In recent decades, evidence has increasingly proven that circRNAs play crucial parts in cancer development, including lung cancer (9)(10)(11). For example, CircZNF124 can promote the progression of nonsmall cell lung cancer cells by targeting mIR-337-3p (12); hsa_circ_0011780 inhibits the proliferation and metastasis of lung cancer (13); down-regulated CircABCB10 enhances the sensitivity of lung cancer cells to cisplatin (14); hsa_circ_0072083 could promote the effect of cisplatin on suppressing the growth of lung cancer (15). Furthermore, a recent study discovered that CircCHST15 (hsa_circ_0109320) is differentially expressed in gefitinib-treated lung cancer and might be a biomarker for the treatment of lung cancer in the future (16). However, the specific role of CircCHST15 in lung cancer still remained unknown. In addition, circRNAs function as spongers to miRNAs and modulate the progression of cancers, including the progression of immune escape (10). The miRNAs sponged by CircCHST15 should be further determined.
On the basis of the current backgrounds, this study focused on determining the specific role of CircCHST15 on lung cancer, especially in the progression of immune escape and the potential miRNAs sponged by CircCHST15 in lung cancer.

Ethics Statement
All the experiments involved in clinical samples were authorized by the Ethics Committee of the Second Affiliated Hospital of Chongqing Medical University (Z20190625H). All patients were informed with the use of their samples in this study, and all signed informed consent. All the animal experiments in this study were approved by the Committee of Experimental Animals of the Second Affiliated Hospital of Chongqing Medical University (Z20190821H).

Clinical Tissues
Cancer tissues and adjacent normal tissues were harvested from 90 lung cancer patients who have received surgery in the Second Affiliated Hospital of Chongqing Medical University between December 2017 and May 2019. None of the patients had received chemotherapy or radiotherapy prior to the study. The major clinical features of the patients ( Table 1), including age of patients, grade, tumor size, clinical stage, differentiation, and lymph node metastasis, were also collected from the Second Affiliated Hospital of Chongqing Medical University.

RNase R Treatment
Briefly, 2.5 mg RNAs of H1395 and A549 cells were incubated with 10U RNase R (M1228, BioVision Incorporated, CA, USA) for 35 min at 37°C. The RNAs incubated with RNase R (RNase

Dual-Luciferase Reporter Assay
Before the experiment, the wide-type sequence of CircCHST15

Biotinylated RNA Pull-Down Assays
Briefly, the H1395 and A549 cells were first transfected with overexpression CircCHST15 plasmids for the up-regulation of the expression of CircCHST15 in the cells. Then, the two cells were further transfected with biotinylated miR-155-5p-WT, miR-155-5p-MUT, miR-194-5p-WT, or miR-194-5p-MUT. Next, the H1395 and A549 cells were lysed, sonicated, and then incubated with C-1 magnetic beads (C37488, Invitrogen) overnight at 4°C. The second day, after probe-coated beads were eluted, RT-qPCR was performed on the rest samples to ensure that CircCHST15 was pulled down by miR-155-5p-WT and miR-194-5p-WT.

CCK-8 Assay
After the transfection, the H1395 (1.0 × 10 4 ) and A549 (1.0 × 10 4 ) cells were respectively added into each well of 96-well plate supplemented with 100 ml RPMI-1640 medium with 10% FBS. After the cell growth for 24 and 48 h, 10 ml of CCK-8 (A56097, OKA, Beijing, China) was further added into each well of the plate for further incubation for 4 h. Then the 96-well plate was inserted into the Imark microplate reader (Bio-Rad, CA, USA) to measure the absorbance under 570 nm.

Colony Formation Assay
After the transfection, the H1395 (1.0 × 10 3 ) and A549 (1.0 × 10 3 ) cells were transferred into a six-well plate, with each well containing 2 ml RPMI-1640 medium with 10% FBS. During the 14-day cell culture, the medium was refreshed every 2 days. Subsequently, after the medium in the plate was discarded, the cells in each well were fixed with 4% paraformaldehyde prepared with 100% paraformaldehyde (ALFA11313, OKA) and PBS (10010049, Gibco). After that, the cells were then dyed with purple crystal (D10739, OKA) for 10 min at room temperature. Finally, the stained cells in the well were observed under a DM4M optical microscope (Solms, Germany), and quantified by Image J 1.8.0 software.

Animal Study
Thirty-three male BALB/c mice (age: 4 weeks old; weight: 18-20 g) were purchased from Cavens (Changzhou, China, http://www. cavens.com.cn/). All the mice were kept in an SPF environment with 12-h light/12-h dark cycle and provided with free diet. The mice were acclimated for 5 days before the experiment. The animal experiments in this study consisted of two sections. In the first section, 12 mice were divided into four groups as follows: the NC group (n = 3), the Control group (n = 3), the mock group (n = 3), and the shCircCHST15 group (n = 3). Before the operation, the CMT167 cells were transfected with lentivirus without any vector (NC lentivirus) or with lentivirus containing shCircCHST15 vector (shCircCHST15 lentivirus). The vectors were synthesized by Genepharma (Shanghai, China). After the transfection, 2.0 × 10 7 normally cultured CMT167 (for the Control group) transfected with NC lentivirus (for the mock group) or shCircCHST15 lentivirus (for the shCircCHST15 group) were injected into the right forelimb subcutaneous armpit of the mice in the Control group, the mock group, and the shCircCHST15 group, respectively. The mice in the NC group were not performed with any injection. Then the mice were normally raised, and 4 days after the injection, when a tumor of 4-5 mm under the armpit of the mouse was touchable, this indicated that lung cancer mouse model was established successfully. Twenty-eight days after the injection, all the mice were sacrificed by cervical dislocation under anesthesia with 0.2% phenobarbital sodium (B005, Nanjing Jiancheng Bio, Jiangsu, China). Then the tumor and blood of all the mice were collected for later use.
In the second section, 21 mice were divided into seven groups: the NC group (n = 3), the Control group (n = 3), the Anti-PD-L1 group (n = 3), the mock group (n = 3), the CircCHST15 group (n = 3), the Anti-PD-L1+CircCHST15 group (n = 3), and the shCircCHST15 group (n = 3). Before the operation, the CMT167 cells were transfected with NC lentivirus, CircCHST15 lentivirus, or shCircCHST15 lentivirus. After the transfection, 2.0 × 10 7 normally cultured CMT167 cells (for the Control group and the Anti-PD-L1 group) or those transfected with NC lentivirus (for the mock group), CircCHST15 lentivirus (for the CircCHST15 group and the Anti-PD-L1+CircCHST15 group), or shCircCHST15 lentivirus (for the shCircCHST15 group) were respectively injected into the right forelimb subcutaneous armpit of mice. The mice in the NC group were normally raised, and intraperitoneally injected with 300 mg saline on days 8, 10, 12. On days 8, 10, and 12 after the injection, the mice in the NC group, the Control group, the mock group, the CircCHST15 group, and the shcircCHST15 group were intraperitoneally injected with normal saline (300 mg/mouse; R21479, OKA), while those in the Anti-PD-L1 group and the Anti-PD-L1+CircCHST15 group were intraperitoneally injected with PD-L1 inhibitor (300 mg/ mouse; P872738, Macklin, Shanghai, China). Twenty-eight days after injection, all the mice were sacrificed by cervical dislocation under anesthesia with 0.2% phenobarbital sodium. Then the tumor and blood of all the mice were collected for later use.

Flow Cytometry
The proportion of CD4 + and CD8 + cells in peripheral blood and tumor supernatant of the mice was detected by flow cytometry. Specifically, after incubation with the CD4 antibody labeled with FITC (ab59474, Abcam) for 1 h, the samples were further incubated with CD8 antibody labeled with PE (ab59474, Abcam) for 1 h. After that, the samples were further incubated with CD3 antibody conjugated with violetFluor 450 (38527, CST, MA, USA) and CD25 antibody conjugated with PE-Cy7 (43212, CST) for 1 h. Then the percentages of CD4 + , CD8 + , CD3 + , CD25 + cells in the samples were quantified by flow cytometer system (FACS-LSR II; Becton-Dickinson, NJ, USA). Finally, the percentage of Tregs + in the samples were further measured based on data of the percentage of CD4 + and CD8 + cells according to the formula: Tregs + (%) = [(CD3 + + CD4 + + CD25 + + total cell number)/CD3 + + total cells number] × 100%.

ELISA
The mouse ELISA kits of IL-10 (ml037873), TNF-b (ml002098), and IFN-g (ml002277) were commercially ordered from Enzyme-linked Biotechnology (Shanghai, China). In brief, tumor tissue supernatant (50 ml) was added into several wells of the 96-well plate provided in the ELISA kit, meanwhile, 50 ml different doses of standard samples from the ELISA kit were added into the rest wells. Soon after, 50 µl target antibody was transferred into the well containing the test sample or standard sample. After 1 h of incubation and washing twice, the well was then added with 80 ml HRP and incubated for 30 min at 37°C. An equal volume of solution A and solution B were mixed together to a total volume of 100 ml and added into the well for 20-min incubation. After stop buffer (50 µl) was added into the well, the absorbance of the 96-well plate was read by the Imark microplate reader (Bio-Rad, CA, USA) under 450 nm.

Immunohistochemistry
The tumor tissues of mice were fixed with 4% paraformaldehyde and made transparent by rxylene (ALFL13317, OKA). Next, the tissues were embedded into paraffin (A56132, OKA), sectioned into 4 mm, and dewaxed. After soaking in antigen repair buffer (R20902, OKA) for 8 min, the tissues were further incubated with the antibody of PD-L1 (13684, CST) overnight at 4°C, soaked in goat-anti-rabbit antibody (ab205718, Abcam) for 1 h, and further dyed with DBA (SFQ004, 4A Biotech, Beijing, China). After washing the tissues under tap water, the tissues were then dyed with hematoxylin (D10519, OKA) for 4 min and further made transparent by xylene. Finally, the expression image of PD-L1 in the tumor tissue was observed under a DMLA full automatic microscope (Leica, Solms, Germany).

RNA Extraction
Total RNA from clinical tissues, cultured cells, and mouse tumor tissues were used for determining the expressions of mRNAs and circRNAs. In brief, the samples were lysed by TRIzol (15596, Invitrogen), mixed with chloroform (A53235, OKA) and centrifuged (14,000 × g) for 30 min for the collection of supernatant. The supernatant was then incubated with an equal volume of isopropanol (E15794, OKA) for 15 min. After centrifuging (14,000 × g) supernatant for 5 min, the total RNA sediments were obtained. Cytoplasm and nuclear RNA from H1395 and A549 cells were detected for the expression location of CircCHST15. The reagents used for the extraction of cytoplasm and nuclear RNA were the kit (37400, NORGEN BIOTEK, Thorold, Canada) specially used for the cytoplasm and nuclear RNA extraction. In short, after the cells were lysed by Buffer J and centrifuged (14,000 × g) for 15 min, the supernatant (cytoplasm RNA) and sediment (nuclear RNA) were respectively collected. The extraction of cytoplasm and nuclear RNA was basically the same, briefly, after incubation with Buffer SK (200 ml) for 10 s, the supernatant and sediment were respectively mixed with 200 ml ethanol (ALF045844, OKA), and the mixtures were centrifuged (3500 × g) for 2 min for harvesting the cytoplasm and nuclear RNA.
Cytoplasm and nuclear RNA from H1395 and A549 cells were detected for the expressions of miRNAs. The reagents used for the extraction of miRNA were the kit (DP501, TianGEN, Beijing, China) specially used for miRNA extraction. To be brief, after the cells were lysed by lysis buffer, the cells were mixed with chloroform and centrifuged (13,400 × g) for 30 min. Then the miRNA sediment was washed by 75% ethanol (A171299, Aladdin, Shanghai, China) once and collected for later use.

RT-qPCR
The total RNAs and miRNAs were isolated from all the samples and reverse-transcribed into cDNAs using Omniscript RT Kit (205111, Qiagen, Dusseldorf, Germany, https://www.qiagen. com/cn/). Then the cDNA was amplified using the qPCR Qupermix Kit (AQ601-01, TransGen, Beijing, China) in the QuantStudio6 system (Applied Biosystems, CA, USA). The reaction conditions were as follows: at 94°C (30 s) for 1 cycle, at 94°C (30 s) for another 40 cycles, at 60°C (30 s) for 40 cycles. Finally, the expression of RNA was quantified by 2 −△△CT method. The primers for cDNAs amplification are shown in Table 2.

Statistical Analysis
Student's t-test and one-way ANOVA were applied to analyze all the data in the current study. LSD and Dunnet's served as posthoc tests. The correlation of CircCHST15 expression with PD-L1 expression in the clinical samples was analyzed by Pearson correlation analysis. All the analyses in this study were performed in SPSS 19.0 software. Statistical data were finally presented as Mean ± SD. Statistically significance was defined when P < 0.05.

CircCHST15 Was High-Expressed in Lung Cancer Tissues and Cells, and CircCHST15 Was Mainly Expressed in the Cytoplasm
The major clinical features of the collected tissues were analyzed, and we found that the clinical stage of the patients with the high-expressed CircCHST15 was more advanced than those with low-expressed CircCHST15 (P < 0.001, Table 1), also the lymph node metastasis of patients with the higher level of CircCHST15 was severer than that in patients with lowexpressed CircCHST15 (P < 0.001, Table 1). To determine the role of CircCHST15 in lung cancer, the expression of CircCHST15 in lung cancer was analyzed. As compared with the adjacent normal tissues (the ANT group) and normal bronchial epithelial cells (16HBE), the expression of CircCHST15 was up-regulated both in cancer tissues (P < 0.001, Figure 1A) and cancer cells (P < 0.001, Figure 1B). The H1395 and A549 cells were used for the in vitro study because the expression of CircCHST15 in the two cells was relatively higher than in other cancer cells. Mechanically, the loop structure of CircCHST15 was verified by treating the RNA isolated from the two cells with RNase R (Figure 1C), and the results demonstrated that the expression of CHST15 was significantly down-regulated (P < 0.001), while no obvious difference was found in CircCHST15 expression after treatment with RNase R (RNase R+) in comparison with the RNase R− group. Also, the expression of CircCHST15 in the cytoplasm of the two cells was apparently higher than that in the nucleus ( Figure 1D), suggesting that CircCHST15 was mainly expressed in the cytoplasm.  (Figure 2A), we found that siCircCHST15 obviously reduced the tumor weight in comparison with the mock group (P < 0.001, Figure 2B).  Figure 2G) and increased by shCircCHST15 both in the blood ( Figure 2G) and tumor supernatant ( Figure  2H) as compared with the mock group (P < 0.01). These results demonstrated that the effect of CircCHST15 on tumor growth was mediated by regulating the activation of CD8 + T cells in the blood and tumor. Therefore, the expressions of key factors (IFN-g, TNF-b, IL-10) specific to CD8 + T were determined. As shown in Figures 2K-M, the expressions of IFN-g, TNF-b, and IL-10 were up-regulated in the blood of the control mice when compared with the NC group (P < 0.001), but down-regulated by shCircCHST15 when compared with the mock group (P < 0.001). As PD-L1 regulated the immune escape in many cancers, we also detected its expression in the tumor tissues ( Figure 2N), and discovered that the PD-L1 expression in the control and mock mice was down-regulated and it could be up- regulated by shCircCHST15. Additionally, we also found that the expression of CircCHST15 was significantly up-regulated in the control mice when compared with the NC group (P < 0.001, Figure 2O), but down-regulated by shCircCHST15 when compared with the mock group (P < 0.001, Figure 2O).

CircCHST15 Regulated PD-L1 Expression in Lung Cancer Cells and It Was Positively Correlated to the Expression of PD-L1
Then the expression of PD-L1 in the H1395 and A549 cells overexpressed with CircCHST15 or knocked down of CircCHST15 ( Figures 3A, B) was measured to examine whether PD-L1 was regulated by CircCHST15. It could be observed that the gene and protein expressions of PD-L1 in

H1395 cells (Figures 3C-E) and A549 cells (Figures 3F-H)
were up-regulated by CircCHST15 (P < 0.001) and downregulated by shCircCHST15 (P < 0.001) when compared with the NC group and shNC group, respectively. As we found that the expression of PD-L1 was high-expressed in lung cancer tissues in comparison with the Ant group (P < 0.001, Figure  3I). The correlation between the expressions of CircCHST15 and PD-L1 was analyzed. As shown in Figure 3J and Figure  3K, a positive correlation was identified between CircCHST15 and PD-L1 both in adjacent normal tissues and in lung cancer tissues.

The Proliferation of H1395 and A549 Cells Were Increased by miR-155-5p Mimic But Inhibited by miR-194-5p Mimic
The viabilities of H1395 cells ( Figure 6A) and A549 cells ( Figure  6B) were detected after cell cultured for 24 and 48 h, and we discovered that CircCHST15 did not affect the cell viability, which was enhanced by miR-155-5p mimic (P < 0.05) and reduced by miR-194-5p mimic (P < 0.05) as compared with the NC+MC group. Mechanically, the cell proliferation of H1395 cells ( Figures  6C, D) and A549 cells (Figures 6E, F) was assessed, consistent with cell viability, the cell proliferation was increased in the NC+ miR-155-5p group (P < 0.001) and the circ+miR-155-5p group (P < 0.001) when compared with the NC+MC group and circ+MC group, respectively. However, the proliferation was inhibited in the NC+miR-194-5p group (P < 0.001) and in the circ+miR-194-5p group (P < 0.001) as compared with the NC+MC group and circ+ MC group, respectively.  Figure 7B) was decreased by PD-L1 inhibitor (Anti-PD-L1) (P < 0.001) and shCircCHST15 (P < 0.001) when compared with the Control group and the mock group, respectively, but the tumor weight was increased by CircCHST15 (P < 0.001) when compared with the mock group. In the Anti-PD-L1+ CircCHST15, the effects of CircCHST15 and Anti-PD-L1 on tumor weight were weakened by each other (P < 0.05) when compared with the Anti-PD-L1 group and the CircCHST15. As for the T cell subtype ( Figures 7C-H), consistent with the results in Figure 2, the CD4 + T cells in the blood ( Figures 7C, I the CircCHST15 group (P < 0.001), but down-regulated by Anti-PD-L1 (P < 0.001) and shCircCHST15 (P < 0.001). However, in the Anti-PD-L1+CircCHST15, the effect of CircCHST15 on the expression of TNF-b was attenuated by Anti-PD-L1 (P < 0.001) compared with the CircCHST15, moreover, the effects of CircCHST15 and Anti-PD-L1 on the expressions of IFN-g, IL-10, CCL17, and CCL22 were attenuated by each other. Finally, the expressions of PD-L1, CircCHST15, miR-155-5p, miR-194-5p, and proliferation-related factors (Ki67 and PCNA) in the tumor tissues were determined ( Figure 8). As shown in Figures 8A-C, it could be observed that the expressions of PD-L1, Ki67, and PCNA were up-regulated in the Control group (P < 0.001) when compared with the NC group, but when compared with the Control group, these expressions were down-regulated by Anti-PD-L1 (P < 0.001), moreover, compared with the mock group, the above expressions were up-regulated by CircCHST15 (P < 0.001) and down-regulated by shCircCHST15 (P < 0.001). However, in the Anti-PD-L1+CircCHST15 group, the effect of CircCHST15 on the Ki67 expression was attenuated by Anti-PD-L1 (P < 0.001) compared with the CircCHST15 group, and the effects of CircCHST15 and Anti-PD-L1 on the expressions of PD-L1 and PCNA were weakened by each other when compared with the CircCHST15 (P < 0.001) and Anti-PD-L1 group (P < 0.05). As shown in Figure 8D, the expression of CircCHST15 was up-regulated in the Control group (P < 0.001) compared with the NC group, while in the Anti-PD-L1 did not affect the expression of CircCHST15. The expressions of miR-155-5p and miR-194-5p ( Figure 8E) were found down-regulated in the Control group (P < 0.001) when compared with the NC group, similarly Anti-PD-L1 did not affect the expressions of miR-155-5p and miR-194-5p, which, however, were down-regulated by CircCHST15 (P < 0.001) and up-regulated by shCircCHST15 (P < 0.001) when compared with the mock group.

DISCUSSION
CircRNAs were abnormally expressed in different cancers (9,11). In lung cancer, studies found that the expressions of hsa_circ_0072083, CircZNF124, and CircABCB10 are up-regulated (12,14,15) but that of hsa_circ_11780, CircPTK2, and hsa_circ_0078767 are down-regulated (13,17,18). This study discovered the up-regulation of CircCHST15 in lung cancer tissues and cells, which was similar to the findings of Liu et al. (16). We also established a lung cancer model (subcutaneousxenotransplant) in mice, and found for the first time that downregulated CircCHST15 inhibited the tumor growth, suggesting that CircCHST15 can act as an oncogene in lung cancer. Subsequently, our findings further showed that down-regulated CircCHST15 regulated distribution of T cell subtypes and inhibited the expression of PD-L1 in the tumor tissues, thus, we speculated that the effect of CircCHST15 on lung cancer might be associated with the immune system, as PD-L1 has critical functions in the immune response and immune escape in different cancers (19)(20)(21)(22). It is widely acknowledged that circRNAs can sponge some miRNAs to affect the functions of miRNAs in various biological activities (9,11). Hsa_circ_0052112 sponges miR-125-5p to promote breast cancer cell migration (23); has_circ_0078710 participates in the progression of liver cancer by sponging miR-31 (24); hsa_circ_0012919 sponges miR-125a-3p to promote the progression of DNA methylation of CD11a in T cells (25). In this study, we found for the first time that miR-155-5p and miR-194-5p could be sponged by CircCHST15 in lung cancer cells. The expression of miR-155-5p is associated with the progression of lung cancer (26,27). Previously, studies found that miR-155-5p and miR-194-5p regulate the metastasis, autophagy, apoptosis, drug resistance in cancers such as in cervical cancer, ovarian cancer, and breast cancer (28)(29)(30)(31)(32)(33). In this study, the viability and proliferation of lung cancer cells were found to be promoted by miR-155-5p and inhibited by miR-194-5p. The target relationships between miR-155-5p with PD-L1 and between miR-194-5p with PD-L1 were further verified in this study. Furthermore, the observation that CircCHST15 did not affect the viability and proliferation of lung cancer cells further encouraged us to further focus on exploring the specific effect of CircCHST15 on immune escape in the development of lung cancer.
To examine the effect of CircCHST15 on immune escape mediated by regulating PD-L1, a lung cancer model (subcutaneous-xenotransplant) was established with the inhibitor of PD-L1 in mice. We found that PD-L1 inhibitor not only inhibited the growth of the tumor, but also reversed the effect of CircCHST15 on promoting tumor growth. Antibody blockade of PD-L1 can activate an anti-tumor immune response leading to durable remissions in a set of cancer patients (34). Anti-PD-1/PD-L1 therapy has generated significant clinical benefits in patients with non-small cell lung cancer (NSCLC) (35). High expression of PD-L1 in lung cancer may contribute to poor prognosis and tumor cells immune escape through suppressing tumor infiltrating dendritic cells maturation (36). EGFR-TKI resistance promotes immune escape in lung cancer via up-regulating PD-L1 expression (37). Autochthonous EGFR-driven lung tumors inhibit antitumor immunity by activating the PD-1/PD-L1 pathway to suppress Tcell function and increase levels of proinflammatory cytokines (38). During the progression of immune escape in cancers, the upregulated PD-L1 regulates the expression of PD-L on the surface of T cells, which has strong immune functions, and then the binding of PD-L1 to PD-L further inhibits T cell functions (39). T cells are composed of CD4 + T cells and CD8 + T cells, which play different roles in immune responses (40). In this study, the proportion of CD4 + T cells remained stable in model mice, while that of CD8 + T cells was increased by PD-L1 inhibitor and inhibited by CircCHST15 in the mouse tumor, and PD-L1 inhibitor reversed the effect of CircCHST15. This indicated that the effect of CircCHST15 on promoting immune escape was mediated by suppressing the normal functions of CD8 + T cells. It has been reported that exosomal PD-L1 promoted tumor growth through immune escape in non-small cell lung cancer via inhibit CD8 + Tcell activity and induce the apoptosis of CD8 + T cells (41). Additionally, these results were supported by the expressions of key factors (IFN-g, TNF-b, IL-10, CCL17, and CCL22) specific to the immune responses mediated by CD8 + T cells. Noticeably, we further confirmed that after the overexpression CircCHST15, the expressions of miR-155-5p and miR-194-5p were down-regulated and PD-L1 expression was up-regulated in the tumor tissues.
To conclude, our research discovered that CircCHST15 sponged miR-155-5p and miR-194-5p to promote the PD-L1mediated immune escape in lung cancer.