Polymorphisms of co-inhibitory molecules (CTLA-4/PD-1/PD-L1) and the risk of non-small cell lung cancer in a Chinese population

Abstract

Lung cancer is a leading cause of cancer-related death in China, with non-small cell lung cancer (NSCLC) comprises the most common form. Co-inhibitory molecules, such as cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), programmed death 1 (PD-1) and its ligand PD-L1, play a key roles in the physiopathological process of tumorigenesis. To investigate whether genetic variations of co-inhibitory molecules are associated with the risk of NSCLC, we analyzed polymorphisms of CTLA-4 (-318, +49), PD-1 (PD-1.1, PD-1.3, PD-1.5, PD-1.9) and PD-L1 (+8293) in a cohort of 528 NSCLC subjects and 600 healthy controls. By restriction fragment length polymorphism (RFLP) method, we found that the distributions of the CTLA-4 and PD-1 gene polymorphisms were similar between NSCLC patients and healthy controls. However, for the PD-L1 8923 A/C polymorphism, frequencies of the AC genotype and C-allele were significantly higher in NSCLC patients than in healthy controls (odds ratio [OR] =1.55; 95% confidence interval [CI] 1.13-2.13; P=0.006; OR=1.52; 95% CI 1.14-2.04; P=0.004, respectively). Stratification analysis revealed that prevalence of the 8923C allele was significantly increased in NSCLC patients who smoke compared to those non-smoking patients (OR=1.51; 95% CI 1.00-2.28; P<0.05). Moreover, NSCLC patients carrying the C-allele had higher risk of regional lymph node metastasis than those carrying the A-allele (OR=5.65; 95% CI 2.45~13.03; P<0.001). These data suggest that PD-L1+8293A>C polymorphism may play a role in the development and progression of NSCLC.

Introduction

Lung cancer is the leading cause of cancer-related deaths worldwide [1]. Non-small cell lung cancer (NSCLC) accounts for approximately 80 percent of primary lung cancers, and the majority of which are presented with advanced stages [2,3]. Tumorigenesis is associated with the properties of cancer cells and the interaction with the immune system. Because drugs such as Bacillus Calmette-Guerin (BCG) [4], interleukin (IL)-2 [5] and interferon [6] have been ineffective in treating NSCLC, NSCLC was considered as a non-immunogenic tumor in the past. Recently, immunotherapy became a popular tool to treat cancer due to our better understanding of the immune response in the tumor immune surveillance. In particular, co-inhibitory molecules (broadly categorized as “checkpoint molecules”), such as cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), programmed death 1 (PD-1) and its ligand PD-L1, can limit the magnitude and duration of immune responses. Immune checkpoint inhibitors have been widely used in treating various solid tumors (including NSCLC) in the early phase studies [7-9], making them promising candidates in NSCLC treatment. CTLA-4 (CD152) is mainly expressed on activated T cells, but it is also expressed on NSCLC tumor cells [10]. The binding of CTLA-4 and B7 protein (B7.1 and B7.2) can revert activated T cells into inhibitory T cells, promoting tumor escape from immunosurveillance. Indeed, CTLA4-knockout mice display lymphoproliferative disorders (LPDs). Nevertheless, CTLA-4 blockade can enhance the immune response and cure tumors in mice [12]. Clinical studies have shown that CTLA-4 inhibitor (ipilimumab) is effective in advanced stage melanoma and is currently being studied in NSCLC [7].

PD-1 (or PDCD 1) belongs to the CD28 Ig superfamily. It is usually expressed on peripheral T cells, natural killer T (NKT) cells, B cells, and monocytes [13]. Interaction between PD-1 and its ligands PD-L1 (CD274) or PD-L2 (CD273) can transduce an inhibitory signal to attenuate T cell activation and proliferation, which can increase peripheral tolerance [14]. PD-L1 as an important member of the B7 family was found in 1999 [15]. Similar as PD-1, PD-L1 can inhibit the proliferation and activation of T cells and dampen immune responses, allowing tumors to escape immunosurveillance. Overexpression of PD-L1 on cancer cells such as lung carcinomas may inhibit T cell activation by enhancing apoptosis of tumor antigen-specific human T cell clones in vitro [16-19]. Additionally, it has been reported that blockage of PD-L1 leads to diabetes and certain cancers in vivo. One study has confirmed that, by using immunohistochemistry, increased PD-L1 expression in NSCLC tissues was related to the magnitude of tumor cell differentiation and TNM stage [20].

Single nucleotide polymorphism (SNP) is one of the most important factors affecting cancer susceptibility. CTLA-4 and PD-1 gene are located closed to each other on the chromosome region 2q33-36 [21], whereas PD-L1 gene is located on the short arm of chromosome 9 (9p24). CTLA-4+49A>G (rs231775) and -318C>T (rs5742909) are widely studied and have been identified as functional polymorphisms [22-24]. Many key polymorphic sites such as PD-1.1 (-538G>A, rs36084323), PD-1.3 (7146G>A, rs11568821), PD-1.5 (7785T>C, rs2227981) and PD-1.9 (7625C>T, rs2227982) have been found in PD-1 gene [25,26]. The association of PD-L1 (8293A>C, rs2890658) gene polymorphism with autoimmune diseases has been widely reported [27,28]. However, these polymorphisms are rarely reported in lung cancer. Therefore, in this study, we investigated the associations between gene polymorphisms of co-inhibitory molecules (CTLA-4, PD-1 and PD-L1) and NSCLC.

Materials and methods

Study population

From 2010 to 2013, a total of 528 patients (352 men and 176 women) with histologically or cytologically confirmed NSCLC and registered at Shandong Cancer Hospital were enrolled in this study. Patients with a history of cancers were excluded from this study. The median age of the patients was 48 (range 27-74) years old. The control group included 600 volunteers who visited the same hospital for health check-up during the same period. The individuals in control group were screened to ensure that none had ever been diagnosed with cancers or other severe diseases. We only allow the Han Chinese into the study to exclude the potential effects of ethnicity. Each participant has provided a written informed consent. The study was approved by the Review Board of Shandong Cancer Hospital. A 2-3 mL peripheral blood sample was collected from each study participant. This procedure was done in accordance with the ethical principles for medical research.

Genotyping

Genomic DNA was extracted from peripheral blood leukocytes using the AxyPrep Blood Genomic DNA Miniprep Kit (Fastagen, China) according to the manufacturer’s instructions. PCR products were digested with Eco91I (New England Biolabs) and separated on 3% polyacrylamide gels (Table 1). A total of 10% of the PCR products were examined by DNA sequencing to confirm the genotyping results. Results between PCR-restriction fragment length polymorphism (RFLP) and DNA sequencing analysis were 100% concordant.

Table 1.

Primers of CTLA4, PD-1 and PD-L1 SNPs used for RFLP analysis

Marker name	SNP	rs number	Location PCR primers
CTLA-4	+49A>G	rs231775	F: 5’-GCTCTACTTCCTGAAGACCT-3’
			R: 5’-AGTCTCACTCACCTTTGCAG-3’
CTLA-4	-318C>T	rs5742909	F: 5’-AAATGAATTGGACTGGATGG-3’
			R: 5’-TTACGAGAAAGGAAGCCGTG-3’
PD-1.1	-538G>A	rs36084323	F: 5’-ATCTGGAACTGTGGCCATGGTG-3’
			R: 5’-ATTCTGTCGGAGCCTCTGGGAG-3’
PD-1.3	7146G>A	rs11568821	F: 5’-CCCCAGGCAGCAACCTCAAT-3’
			R: 5’-GACCGCAGGCAGGCACATAT-3’
PD-1.5	7785T>C	rs2227981	F: 5’-GTGCCTGTGTTCTCTGTGGA -3’
			R: 5’-CCAAGAGCAGTGTCCATCCT-3’
PD-1.9	7625C>T	rs2227982	F: 5’-GGACAGCTCAGGGTAAGCAG-3’
			R: 5’-AGGGTCTGCAGAACACTGGT-3’
PD-L1	8293A>C	rs2890658	F: 5’-AATGGCTTGTTGTCCAGAGATG-3’
			R: 5’-GTACCACATGGAGTGGCTGC-3’

Statistical analysis

The SPSS statistical software package version 17.0 was used for all statistical analyses. Genotype and allele frequencies of co-inhibitory molecules polymorphism were obtained by direct counting and were compared by chi-square test. Odds ratios (OR) and 95% confidence intervals (CIs) were calculated to assess the relative risk. Demographic and clinical data between groups were compared by chi-square test and Student’s t test. P value less than 0.05 was considered statistically significant.

Results

Clinical characteristics of the study subjects

Clinical characteristics of the subjects are presented in Table 2. There were no significant differences in age (P=0.473) and gender (P=0.723) between cases and controls. The smoking rate, as a determinant risk factor for NSCLC, was significantly higher in patients group compared with the control group (60.2 versus 39.8%, P<0.001) (Table 2). As for the regional lymph node status, 78.8% of the patients were diagnosed with lymph node metastasis.

Table 2.

Characteristics of NSCLC patients and controls

Characteristic	Cases (n=528)	Controls (n=600)	OR (95% CI)	P value
Age (years)
(means ± SD)	27-74 (48±14)	25-76 (50±15)		0.473
Sex
Male	352 (66.7)	394 (65.7)	1
Female	176 (33.3)	206 (34.3)	0.96 (0.75-1.22)	0.723
Smoking status
Nonsmokers	210 (39.8)	328 (54.7)	1
Smokers	318 (60.2)	272 (45.3)	1.83 (1.44-2.32)	<0.001
Nodal stage
N0	112 (21.2)
N1~N3	416 (78.8)
Histological cell type
Adenocarcima	268 (50.7)
Squamous cell carcinoma	192 (36.4)
Others	68 (12.9)

Association between co-inhibitory molecules gene polymorphism and NSCLC susceptibility

Genotype and allele frequencies of the-polymorphisms in NSCLC cases and healthy controls are summarized in Table 3. The genotype distributions for each SNP were in agreement with Hardy-Weinberg equilibrium (HWE). Data showed that the distributions of the allele and genotype CTLA-4 and PD-1 gene polymorphisms were similar between patients and controls. However, for the PD-L1 8923 A/C polymorphism, frequencies of the AC genotype and C-allele were significantly higher in NSCLC patients than in controls (20.1 versus 14.0%, OR=1.55; 95% CI 1.13-2.13; P=0.006; 11.2 versus 7.7%, OR=1.52; 95% CI 1.14-2.04; P=0.004, respectively) (Table 3).

Table 3.

Distribution of genotype and allele frequencies in NSCLC patients and controls

Genotypes	Cases, n (%)	Controls, n (%)	OR (95% CI)	P value
CTLA-4+49
AA	172 (32.6)	222 (37.0)	1
GA	282 (53.4)	306 (51.0)	1.19 (0.92-1.54)	0.185
GG	74 (14.0)	72 (12.0)	1.33 (0.91-1.94)	0.146
A	626 (59.3)	750 (62.5)	1
G	430 (40.7)	450 (37.5)	1.15 (0.97-1.36)	0.118
CTLA-4-318
CC	468 (88.6)	510 (85.0)	1
CT	60 (11.4)	90 (15.0)	0.73 (0.51-1.03)	0.074
TT	0	0
C	996 (94.3)	1110 (92.5)	1
T	60 (5.7)	90 (7.5)	0.74 (0.53-1.04)	0.085
PD-1.1
GG	144 (27.3)	156 (26.0)	1
GA	246 (46.6)	296 (49.3)	0.90 (0.68-1.19)	0.467
AA	138 (26.1)	148 (24.7)	1.01 (0.73-1.40)	0.951
G	534 (50.6)	608 (50.7)	1
A	522 (49.4)	592 (49.3)	1.00 (0.85-1.19)	0.963
PD-1.3
GG	426 (80.7)	456 (76.0)	1
GA	102 (19.3)	142 (23.7)	0.77 (0.58-1.02)	0.072
AA	0	2 (0.3)
G	954 (90.3)	1054 (87.8)	1
A	102 (9.7)	146 (12.2)	0.77 (0.59-1.00)	0.058
PD-1.5
CC	244 (46.2)	256 (42.7)	1
CT	216 (40.9)	246 (41.0)	0.92 (0.72-1.19)	0.526
TT	68 (12.9)	98 (16.3)	0.73 (0.51-1.04)	0.080
C	704 (66.7)	758 (63.2)	1
T	352 (33.3)	442 (36.8)	0.86 (0.72-1.02)	0.083
PD-1.9
CC	343 (65.0)	404 (67.3)	1
CT	148 (28.0)	168 (28.0)	1.04 (0.80-1.35)	0.784
TT	37 (7.0)	28 (4.7)	1.56 (0.93-2.60)	0.090
C	834 (79.0)	976 (81.3)	1
T	222 (21.0)	224 (18.7)	1.16 (0.94-1.43)	0.161
PD-L1 8293
AA	416 (78.8)	512 (85.3)	1
AC	106 (20.1)	84 (14)	1.55 (1.13~2.13)	0.006
CC	6 (1.1)	4 (0.7)	1.85 (0.52~6.59)	0.345
A	938 (88.8)	1108 (92.3)	1
C	118 (11.2)	92 (7.7)	1.52 (1.14~2.04)	0.004

Stratification analysis of PD-L1 gene SNP in NSCLC

We further investigated whether the PD-L1 8923 SNP could be correlated with certain clinical characteristics in NSCLC. Data revealed that prevalence of the 8923C allele was significantly increased in NSCLC patients who smoke compared to the non-smoking cases (OR=1.51; 95% CI 1.00-2.28; P<0.05). Moreover, NSCLC patients carrying the C-allele had higher risk of regional lymph node metastasis than those carrying the A-allele (OR=5.65; 95% CI 2.45-13.03; P<0.001) (Table 4).

Table 4.

Allele and genotype frequencies of PD-L1 A8293C in patients with NSCLC according to clinicopathologic characteristics

	Total (n)	AA	AC	CC	A	C
Gender
Male	352	280	70	2
Female	176	136	36	4
Smoking status
Nonsmokers	210	173	37	-	383 (91.2)	37 (8.8)
Smokers	318	243	69	6	555 (87.3)	81 (12.7)*
Nodal stage
N0	112	106	6	-	218 (97.3%)	6 (2.7%)
N1~N3	416	310	100	6	720 (86.5%)	112 (13.5%)**

^*OR=1.51, 95% CI=1.00-2.28, P<0.05;

^**OR=5.65, 95% CI=2.45-13.03, P<0.001.

Discussion

Immune escape can lead to the development of malignant tumor. An effective antitumor immunotherapy depends on the successful activation of T lymphocytes. It has been confirmed that T cells achieving physiological activation threshold require the synergy of dual signals: one is the peptide: major histocompatibility complex (MHC) recognized by the T-cell receptor (TCR); the other is a coordinated stimulus signal, delivered from matured antigen-presenting cells (APCs) to T cells via co-stimulatory molecules. The latter can result in T cell activation, anergy or even apoptosis. The co-stimulatory pathways consist of tumor necrosis factor/tumor necrosis factor receptor (TNF/TNFR) and B7/CD28 superfamily and co-inhibitory molecules.

Emerging evidence has shown that tobacco smoking strongly increases the risk of lung cancer. However, more and more studies have shown that non-smokers can also suffer from lung cancer. According to epidemiological studies, thirty percent of lung cancer patients in Asia are non-smokers [29], suggesting that genetic factors may influence the risk of lung cancer. Therefore, we proposed that certain common genetic polymorphisms in co-inhibitory molecule genes may affect the risk of lung cancer, and investigated previously described polymorphic markers in CTLA-4, PD-1 and PD-L1 genes.

The SNPs located at CTLA-4-318 and +49 are most widely studied. These transitions could influence T-cell activation and play a role in antitumor immunity. The AA genotype of CTLA-4+49, caused by a substitution of alanine to threonine, leads to higher expression of CTLA-4 mRNA and protein than the GG genotype. In addition, it has been shown that the protein product coded by the CTLA-4+49A>G [AA] genotype can improve the inhibitory effect on T-cell activation. The CTLA-4-318C>T [T] allele is known as being associated with higher promoter activity, and may significantly increase the expression of surface CTLA-4 in stimulated and non-stimulated cells. These variations may limit the potency of antitumor immunity. Therefore, we hypothesized that these polymorphisms might influence NSCLC susceptibility in a Chinese population. Interestingly, in our study, the CTLA-4+49A>G and -318C>T polymorphisms were not associated with susceptibility to NSCLC. This is consistent with the results of Karabon L et al. [30] and Song B et al. [31]. However, Antczak A et al found that the presence of G-allele and GG genotype in NSCLC tissues (+49A/G) was significantly associated with the increased risk of NSCLC [32]. This inconsistency could be due to differences in ethnicity and sample size.

Data on PD-1 gene polymorphism and its association with tumor malignancy are limited. In current study, we investigated PD-1 polymorphisms in patients with NSCLC and healthy individuals. Our results showed no differences in the distribution of the PD-1 (PD-1.1, PD-1.3, PD-1.5 and PD-1.9) genotypes and alleles in NSCLC group compared to the control group. Previous reports showed that these polymorphisms may play different roles in different cancers. Recently, Sasaki H et al. reported that PD-1.1 polymorphisms were not associated with susceptibility to NSCLC but were correlated with a poor prognosis [25]. To our knowledge, there were no other reports about PD-1 gene polymorphism and its association with NSCLC susceptibility. Thus, it’s difficult to verify our results by comparing with the published data.

So far, the PD-L1 gene polymorphism has been extensively investigated in many autoimmune diseases, and it has been shown to be closely related to certain diseases. For example, Anna L et al. found that PD-L1 rs1411262C/T gene polymorphism was closely related to Grave’s disease and Addison’s disease [28]. Hayashi M et al. found that PD-L1 8923 A/C gene polymorphism was closely related to Grave’s disease susceptibility and prognosis [27]. Carolina P et al. found that PD-L1 rs2297137A/G and rs4143815C/G gene polymorphism were closely related to type 1 diabetes (T1D) in Chilean population [33]. However, reports about the correlation between PD-L1 gene polymorphism and cancer susceptibility are rare. In our present study, we observed a significant association between the polymorphism of PD-L1 8293 A>C and NSCLC susceptibility. Additionally, we found that NSCLC patients carrying the C-allele had higher risk of regional lymph node metastasis than those carrying the A-allele. Our result is consistent with another study from China [34]. To validate our results, it is necessary to conduct the experiment in a different population, with a larger sample size and the matching of cases-controlled trial. So far, the function of 8293 polymorphism is not clear. It could be that the C-allele (AC genotype) of PD-L1 8293 functions through up-regulation of death/negative ligands, thus dampening an effective immune response, and eventually causing lung cancer. However, further investigation is needed.

In summary, we found that the PD-L1+8293A>C polymorphism plays a role in NSCLC. The C-allele may increase the risk of NSCLC and regional lymph node metastasis. Further research is required to disclose the functional consequence of different PD-L1+8293 genotypes on PD-L1 molecule.

Disclosure of conflict of interest

None.