Abstract
Functional variants in inducible T cell costimulator (ICOS) gene are predicted to be associated with the susceptibility of colorectal cancer (CRC). In this study, we enrolled 2,606 participants (involving 1,003 CRC cases and 1,303 healthy controls) and conducted a case-control study to explore the potential relationship of ICOS rs4404254 T>C and rs10932029 T>C polymorphisms with the risk of CRC. A custom-by-design 48-Plex SNPscan Kit was used to obtain the genotypes of ICOS rs4404254 T>C and rs10932029 T>C variants. We found that ICOS rs10932029 T>C polymorphism was associated with risk of CRC in several subgroups (female subgroup: CC vs. TT: adjusted OR = 6.49, 95% CI 1.36-30.90, P = 0.019 and CC vs. CT/TT: adjusted OR = 6.38, 95% CI 1.34-30.32, P = 0.020; < 61 years subgroup: CC vs. TT: adjusted OR = 4.23, 95% CI 1.10-16.24, P = 0.036 and CC vs. CT/TT: adjusted OR = 4.20, 95% CI 1.10-16.09, P = 0.036; never smoking subgroup: CC vs. TT: adjusted OR = 2.82, 95% CI 1.04-7.64, P = 0.041 and CC vs. CT/TT: adjusted OR = 2.83, 95% CI 1.05-7.66, P = 0.041 and BMI ≥ 24 subgroup: CC vs. TT: adjusted OR = 6.81, 95% CI 1.39-33.30, P = 0.018 and CC vs. CT/TT: adjusted OR = 6.79, 95% CI 1.39-33.11, P = 0.018). In addition, we found that ICOS rs4404254 T>C polymorphism was associated with the susceptibility of CRC in never smoking subgroup (CC/TC vs. TT: adjusted OR = 1.23, 95% CI 1.01-1.51, P = 0.045). In summary, our findings suggest that ICOS rs10932029 T>C and ICOS rs4404254 T>C polymorphisms may be associated with the risk of CRC. In the future, a fine-mapping study with a functional evaluation is needed to explore the relationship between ICOS polymorphisms and the risk of CRC.
Keywords: ICOS, polymorphism, colorectal cancer, risk
Introduction
Colorectal cancer (CRC) remains a common public health problem, accounting for over 1.4 million newly diagnosed CRC patients in 2012 worldwide [1]. In China alone, there were approximately 376,300 CRC cases and 191,000 people died from CRC in 2015 [2]. The etiology of CRC remains unclear. Individuals with similar lifestyles, for example, only a small proportion of them might progress to CRC. Thus, the incidence of CRC may be influenced by various environmental factors, and an individual’s genetic background. Recently, a number of studies have focused on the potential role of the immune system on the development of CRC (immune responses and individual’s gene regulation or variants) [3-5]. Thus, investigation of host immune response-related polymorphisms could lead to novel insights into host immune responses in developing CRC.
It is reported that the CD28 family plays a vital role in human T-lymphocyte-dependent humoral immunity. Inducible T cell costimulator (ICOS), a member of the CD28 family, is expressed on activated T cells, and then forms homodimers to regulate cell signaling transduction, immune responses, and cell proliferation [6,7]. Due to ICOS sharing homology with CD28, early research sought to characterize the potential role of ICOS in T-lymphocyte proliferation and activation. Interestingly, some previous studies suggested that ICOS-deficient T-cells might lead to a significant proliferation defect in vitro when compared to wild-type CD4+ T-lymphocytes [8,9]. A recent study indicated that expression of ICOS improved prognosis in CRC patients and the percentage of ICOS(+)CD4(+) cells acting as Th1 cells [10]. Thus, ICOS may interact with its ligand (ICOSL) and influence the development of CRC. The ICOS gene is located on chromosome 2 (position: 203936748-203961577) in humans. The ICOS gene is polymorphic. There are more than 5,000 single-nucleotide polymorphisms (SNPs) in ICOS gene which have been identified (https://www.ncbi.nlm.nih.gov/snp/?term=ICOS), such as ICOS rs10932029 T>C, rs10932037 C>T, rs4675379 G>C, rs4404254 T>C and rs10183087 A>C polymorphisms, etc. Among these SNPs, ICOS rs4404254 T>C and rs10932029 T>C SNPs have been extensively investigated for their susceptibility to cancer [11-14]. But these previous studies reported inconsistent findings rather than conclusive evidence. The aim of this case-control study was to assess the association between ICOS polymorphisms and risk of CRC.
Materials and methods
Study subjects
The protocol of this case-control study was approved by the Medical Ethics Committee at Fujian Medical University and Jiangsu University and conformed to the Declaration of Helsinki. Prior to recruitment, each participant signed an informed consent. Enrollees from Fujian and Jiangsu province of China were accepted after seeking treatment at Fujian Medical University Union Hospital or Affiliated People’s Hospital of Jiangsu University between October 2014 and August 2017. The diagnosis of CRC patients was made based on pathology. We enrolled a total of 1,003 patients with CRC (620 males, 383 females), with a mean age of 61.10 ± 12.17 years. And 1,303 healthy controls (801 males and 502 females) were also recruited, with a mean age of 61.40 ± 9.61 years. The controls were matched with CRC cases with respect to gender, age, geographic origin, and ethnic background. Demographics and risk factors were obtained by using a questionnaire. Overweight and obesity were defined as body mass index (BMI) ≥ 24 [15,16].
DNA extraction and genotyping
Ethylenediamine tetraacetic acid (EDTA)-anticoagulated venous blood donated by all participants was harnessed to isolate genomic DNA by using a Promega DNA Blood Mini Kit (Promega, Madison, USA). A custom-by-design 48-Plex SNPscan Kit (Genesky Biotechnologies Inc., Shanghai, China) was used to obtain the genotypes of ICOS rs4404254 T>C and rs10932029 T>C SNPs as previously described [17]. For quality control, ninety-two DNA samples were randomly selected and analyzed by SNPscan Kit. The obtained genotypes were not changed.
Statistical analysis
The demographic and selected risk factors between CRC cases and controls were compared by using X2 test or Fisher exact tests. An online goodness-of-fit Chi-square test software (http://ihg.gsf.de/cgi-bin/hw/hwa1.pl) was used to assess whether the genotype distributions of ICOS rs4404254 T>C and rs10932029 T>C polymorphisms were in Hardy-Weinberg equilibrium (HWE). Genotype frequencies of ICOS rs4404254 T>C and rs10932029 T>C polymorphisms in CRC cases and controls were also compared by X2 test or Fisher exact tests. The potential association between ICOS rs4404254 T>C and rs10932029 T>C polymorphisms and CRC risk were estimated as odds ratios (ORs) with their 95% intervals (CIs). The interaction of gene and environment in the risk of CRC was determined by conditional logistic regression. P value < 0.05 (two-tailed) was considered as significant. All data analyses were carried out by using the SAS version 9.4 statistical software (SAS Institute, Cary, USA).
Results
Study characteristics
The demographic variables and selected risk factors are listed in Table 1. We found no difference between CRC cases and healthy controls considering sex (P = 0.867), age (P = 0.496), suggesting that the two groups were well matched. When regarding alcohol consumption, BMI and cigarette use, we found there were significant differences (P < 0.001, P < 0.001 and P = 0.002, respectively). The primary information of ICOS rs4404254 T>C and rs10932029 T>C SNPs is shown in Table 2. Genotype distributions of ICOS rs4404254 T>C and rs10932029 T>C in controls were in accordance with HWE (P = 0.295 and 0.538, respectively).
Table 1.
Distribution of selected characteristics in CRC cases and controls
| Variable | Cases (n = 1,003) | Controls (n = 1,303) | P a | ||
|---|---|---|---|---|---|
|
| |||||
| n | % | n | % | ||
| Age (years) | 61.10 (± 12.17) | 61.40 (± 9.61) | 0.496 | ||
| Age (years) | 0.605 | ||||
| < 61 | 451 | 44.97 | 600 | 46.05 | |
| ≥ 61 | 552 | 55.03 | 703 | 53.95 | |
| Sex | 0.867 | ||||
| Male | 620 | 61.81 | 801 | 61.47 | |
| Female | 383 | 38.19 | 502 | 38.53 | |
| Smoking status | 0.002 | ||||
| Never | 744 | 74.18 | 1038 | 79.66 | |
| Ever | 259 | 25.82 | 265 | 20.34 | |
| Alcohol use | < 0.001 | ||||
| Never | 829 | 82.65 | 1,167 | 89.56 | |
| Ever | 174 | 17.35 | 136 | 10.44 | |
| BMI (kg/m2) | |||||
| < 24 | 670 | 66.80 | 688 | 52.80 | < 0.001 |
| ≥ 24 | 333 | 33.20 | 615 | 47.20 | |
| Site of tumor | |||||
| Colon cancer | 431 | 42.97 | |||
| Rectum cancer | 572 | 57.03 | |||
Two-sided X 2 test and student t test; Bold values are statistically significant (P < 0.05).
BMI: body mass index.
Table 2.
Primary information for ICOS polymorphisms
| Genotyped SNPs | Chromosome | Chr Pos (NCBI Build 37) | Region | MAFa for Chinese in database | MAF in our controls (n = 782) | P value for HWEb test in our controls | Genotyping method | Genotyping value (%) |
|---|---|---|---|---|---|---|---|---|
| ICOS rs10932029 T>C | 2 | 204801768 | Intron1 | 0.084 | 0.100 | 0.538 | SNPscan | 98.87 |
| ICOS rs4404254 T>C | 2 | 204825286 | 3’UTR | 0.131 | 0.168 | 0.295 | SNPscan | 98.87 |
MAF: minor allele frequency;
HWE: Hardy-Weinberg equilibrium.
Association of ICOS rs4404254 T>C and rs10932029 T>C SNPs with CRC risk
Genotype distributions of ICOS rs4404254 T>C and rs10932029 T>C SNPs are listed in Table 3. The genotype frequencies of ICOS rs4404254 T>C polymorphism were 65.41% (TT), 31.02% (TC) and 3.57% (CC) in CRC cases and 69.62% (TT), 27.15% (TC) and 3.23% (CC) in controls. When the frequency of ICOS rs4404254 TT genotype was considered as reference, individuals carrying the ICOS rs4404254 TC genotype had a tendency of increased risk to CRC (crude OR = 1.18, 95% CI = 0.98-1.41 for TC vs. TT, P = 0.080). When compared with the frequency of ICOS rs4404254 TT genotype, individuals carrying the ICOS rs4404254 CC genotype were not associated with the risk of CRC (crude OR = 1.14, 95% CI = 0.72-1.80 for CC vs. TT, P = 0.578). When the frequency of ICOS rs4404254 TT genotype was used as reference, individuals carrying the ICOS rs4404254 TC/CC genotype significantly increased the susceptibility of CRC (crude OR = 1.21, 95% CI = 1.02-1.45 for TC/CC vs. TT, P = 0.033). When compared with the frequency of ICOS rs4404254 TC/TT genotype, individuals carrying the CC genotype were not associated with the risk of CRC (crude OR = 1.11, 95% CI = 0.70-1.75 for CC vs. TT/TC, P = 0.654). Adjustments for smoking, BMI, age, sex and drinking, a tendency of increased risk to CRC was also found (adjusted OR = 1.15, 95% CI = 0.96-1.39 for TC vs. TT, P = 0.141; adjusted OR = 1.07, 95% CI = 0.67-1.71 for CC vs. TT, P = 0.767, adjusted OR = 1.18, 95% CI = 0.98-1.41 for TC/CC vs. TT, P = 0.074 and adjusted OR = 1.05, 95% CI = 0.66-1.67 for CC vs. TT/TC, P = 0.831; Table 3).
Table 3.
Logistic regression analyses of associations between ICOS polymorphisms and the risk of overall CRC
| Genotype | Cases n = 1,003) | Controls (n = 1,303) | Crude OR (95% CI) | P | Adjusted ORa (95% CI) | P a | ||
|---|---|---|---|---|---|---|---|---|
|
|
|
|||||||
| n | % | n | % | |||||
| rs10932029 T>C | ||||||||
| TT | 770 | 78.57 | 1,061 | 81.62 | 1.00 | 1.00 | ||
| TC | 195 | 19.90 | 229 | 17.62 | 1.14 (0.92-1.41) | 0.218 | 1.11 (0.89-1.37) | 0.354 |
| CC | 15 | 1.53 | 10 | 0.77 | 2.01 (0.90-4.50) | 0.089 | 1.92 (0.85-4.36) | 0.117 |
| TC+CC | 210 | 21.43 | 239 | 18.38 | 1.21 (0.98-1.49) | 0.071 | 1.17 (0.95-1.45) | 0.139 |
| TT+TC | 965 | 98.47 | 1,290 | 99.23 | 1.00 | 1.00 | ||
| CC | 15 | 1.53 | 10 | 0.77 | 2.01 (0.90-4.48) | 0.090 | 1.93 (0.85-4.37) | 0.116 |
| C allele | 225 | 11.48 | 249 | 9.58 | ||||
| rs4404254 T>C | ||||||||
| TT | 641 | 65.41 | 905 | 69.62 | 1.00 | 1.00 | ||
| TC | 304 | 31.02 | 353 | 27.15 | 1.18 (0.98-1.41) | 0.080 | 1.15 (0.96-1.39) | 0.141 |
| CC | 35 | 3.57 | 42 | 3.23 | 1.14 (0.72-1.80) | 0.578 | 1.07 (0.67-1.71) | 0.767 |
| TC+CC | 339 | 34.59 | 395 | 30.38 | 1.21 (1.02-1.45) | 0.033 | 1.18 (0.98-1.41) | 0.074 |
| TT+TC | 945 | 96.43 | 1,258 | 96.77 | 1.00 | 1.00 | ||
| CC | 35 | 3.57 | 42 | 3.23 | 1.11 (0.70-1.75) | 0.654 | 1.05 (0.66-1.67) | 0.831 |
| C allele | 374 | 19.08 | 437 | 16.81 | ||||
Adjusted for age, sex, BMI, smoking and drinking status;
Bold values are statistically significant (P < 0.05).
The genotype frequencies of ICOS rs10932029 T>C SNP were 78.57% (TT), 19.90% (TC) and 1.53% (CC) in CRC cases and 81.62% (TT), 17.62% (TC) and 0.77% (CC) in controls. When the frequency of ICOS rs10932029 TT genotype was used as reference, individuals carrying the ICOS rs10932029 TC genotype were not associated with the risk of CRC (crude OR = 1.14, 95% CI = 0.92-1.41 for TC vs. TT, P = 0.218). When compared with the frequency of ICOS rs10932029 TT genotype, individuals carrying the ICOS rs10932029 CC genotype had a tendency of increased risk to CRC (crude OR = 2.01, 95% CI = 0.90-4.50 for CC vs. TT, P = 0.089). When the frequency of ICOS rs10932029 TT genotype was used as reference, individuals carrying the ICOS rs10932029 TC/CC genotype also had a tendency of increased risk to CRC (crude OR = 1.21, 95% CI = 0.98-1.49 for TC/CC vs. TT, P = 0.071). When ICOS rs10932029 TC/TT genotype was considered as reference, individuals carrying the CC genotype also had a tendency of increased susceptibility to CRC (crude OR = 2.01, 95% CI = 0.90-4.48 for CC vs. TT/TC, P = 0.090). However, after adjusting for smoking, BMI, age, sex, and drinking, the association between ICOS rs10932029 T>C polymorphism and risk of CRC was not found (adjusted OR = 1.11, 95% CI = 0.89-1.37 for TC vs. TT, P = 0.354; adjusted OR = 1.92, 95% CI = 0.85-4.36 for CC vs. TT, P = 0.117, adjusted OR = 1.17, 95% CI = 0.95-1.45 for TC/CC vs. TT, P = 0.139 and adjusted OR = 1.93, 95% CI = 0.85-4.37 for CC vs. TT/TC, P = 0.116; Table 3).
Association of ICOS rs4404254 T>C and rs10932029 T>C SNPs with CRC risk in a stratified analysis
To further assess the correlation of ICOS rs4404254 T>C and rs10932029 T>C SNPs with CRC susceptibility, a detailed stratified analysis was carried out by gender, age, BMI, smoking, and drinking. Table 4 summarizes the genotype frequencies of ICOS rs10932029 T>C variants in different subgroups. After adjusting by gender, age, BMI, smoking, and drinking, we found that ICOS rs10932029 T>C polymorphism was associated with development of CRC in several subgroups [female subgroup: CC vs. TT: adjusted OR = 6.49, 95% CI 1.36-30.90, P = 0.019 and CC vs. CT/TT: adjusted OR = 6.38, 95% CI 1.34-30.32, P = 0.020; < 61 years subgroup: CC vs. TT: adjusted OR = 4.23, 95% CI 1.10-16.24, P = 0.036 and CC vs. CT/TT: adjusted OR = 4.20, 95% CI 1.10-16.09, P = 0.036; never smoking subgroup: CC vs. TT: adjusted OR = 2.82, 95% CI 1.04-7.64, P = 0.041 and CC vs. CT/TT: adjusted OR = 2.83, 95% CI 1.05-7.66, P = 0.041 and BMI ≥ 24 subgroup: CC vs. TT: adjusted OR = 6.81, 95% CI 1.39-33.30, P = 0.018 and CC vs. CT/TT: adjusted OR = 6.79, 95% CI 1.39-33.11, P = 0.018 (Table 4)].
Table 4.
Stratified analyses between ICOS rs10932029 T>C polymorphism and CRC risk by sex, age, BMI, smoking status and alcohol consumption
| Variable | ICOS rs10932029 T>C (case/control)a | Adjusted ORb (95% CI); P | |||||
|---|---|---|---|---|---|---|---|
|
|
|
||||||
| TT | TC | CC | Additive model | Homozygote model | Dominant model | Recessive model | |
| Sex | |||||||
| Male | 478/646 | 120/145 | 6/8 | 1.05 (0.80-1.38); P: 0.713 | 0.91 (0.31-2.69); P: 0.859 | 1.08 (0.82-1.41); P: 0.590 | 0.92 (0.31-2.72); P: 0.878 |
| Female | 292/415 | 75/84 | 9/2 | 1.20 (0.85-1.71); P: 0.303 | 6.49 (1.36-30.90); P: 0.019 | 1.35 (0.96-1.90); P: 0.087 | 6.38 (1.34-30.32); P: 0.020 |
| Age | |||||||
| < 61 | 342/482 | 92/113 | 9/3 | 1.12 (0.82-1.53); P: 0.489 | 4.23 (1.10-16.24); P: 0.036 | 1.22 (0.90-1.66); P: 0.209 | 4.20 (1.10-16.09); P: 0.036 |
| ≥ 61 | 428/579 | 103/116 | 6/7 | 1.13 (0.84-1.52); P: 0.423 | 1.10 (0.36-3.33); P: 0.872 | 1.16 (0.87-1.56); P: 0.308 | 1.10 (0.36-3.34); P: 0.869 |
| Smoking status | |||||||
| Never | 575/849 | 141/180 | 12/6 | 1.10 (0.86-1.40); P: 0.472 | 2.82 (1.04-7.64); P: 0.041 | 1.18 (0.93-1.50); P: 0.183 | 2.83 (1.05-7.66); P: 0.041 |
| Ever | 195/212 | 54/49 | 3/4 | 1.15 (0.74-1.78); P: 0.542 | 0.74 (0.16-3.43); P: 0.695 | 1.16 (0.75-1.78); P: 0.509 | 0.74 (0.16-3.42); P: 0.695 |
| Alcohol consumption | |||||||
| Never | 642/952 | 155/204 | 13/8 | 1.08 (0.86-1.37); P: 0.515 | 2.31 (0.94-5.67); P: 0.067 | 1.16 (0.92-1.45); P: 0.214 | 2.33 (0.95-5.69); P: 0.065 |
| Ever | 128/109 | 40/25 | 2/2 | 1.30 (0.74-2.29); P: 0.367 | 0.73 (0.10-5.47); P: 0.756 | 1.29 (0.74-2.25); P: 0.365 | 0.70 (0.09-5.28); P: 0.733 |
| BMI (kg/m2) | |||||||
| < 24 | 514/549 | 134/129 | 8/8 | 1.07 (0.82-1.41); P: 0.608 | 1.01 (0.37-2.72); P: 0.993 | 1.09 (0.84-1.43); P: 0.506 | 1.01 (0.37-2.72); P: 0.987 |
| ≥ 24 | 256/512 | 61/100 | 7/2 | 1.17 (0.82-1.66); P: 0.398 | 6.81 (1.39-33.30); P: 0.018 | 1.32 (0.93-1.86); P: 0.118 | 6.79 (1.39-33.11); P: 0.018 |
The genotyping was successful in 980 (97.71%) CRC cases and 1,300 (99.77%) controls for ICOS rs10932029 T>C;
Adjusted for age, sex, BMI, smoking status and alcohol consumption (besides stratified factors accordingly) in a logistic regression model.
Table 5 presents the genotype frequencies of ICOS rs4404254 T>C variants in different subgroups. Adjustment by conditional logistic regression with gender, age, BMI, smoking, and drinking, we found that ICOS rs4404254 T>C polymorphism was associated with development of CRC in the never smoking subgroup: CC/TC vs. TT: adjusted OR = 1.23, 95% CI 1.01-1.51, P = 0.045 (Table 5).
Table 5.
Stratified analyses between ICOS rs4404254 T>C polymorphism and CRC risk by sex, age, BMI, smoking status and alcohol consumption
| Variable | ICOS rs4404254 T>C (case/control)a | Adjusted ORb (95% CI); P | |||||
|---|---|---|---|---|---|---|---|
|
|
|
||||||
| TT | TC | CC | Additive model | Homozygote model | Dominant model | Recessive model | |
| Sex | |||||||
| Male | 401/564 | 184/207 | 19/28 | 1.19 (0.93-1.51); P: 0.162 | 0.82 (0.45-1.51); P: 0.532 | 1.18 (0.94-1.49); P: 0.155 | 0.80 (0.44-1.47); P: 0.479 |
| Female | 240/341 | 120/146 | 16/14 | 1.11 (0.82-1.49); P: 0.509 | 1.62 (0.76-3.43); P: 0.210 | 1.18 (0.88-1.57); P: 0.267 | 1.59 (0.75-3.35); P: 0.223 |
| Age | |||||||
| < 61 | 277/414 | 148/166 | 18/18 | 1.24 (0.94-1.63); P: 0.128 | 1.38 (0.70-2.73); P: 0.358 | 1.28 (0.98-1.67); P: 0.067 | 1.31 (0.66-2.58); P: 0.438 |
| ≥ 61 | 364/491 | 156/187 | 17/24 | 1.07 (0.83-1.38); P: 0.621 | 0.87 (0.46-1.66); P: 0.671 | 1.08 (0.85-1.39); P: 0.522 | 0.88 (0.46-1.66); P: 0.685 |
| Smoking status | |||||||
| Never | 465/718 | 237/287 | 26/30 | 1.19 (0.97-1.47); P: 0.104 | 1.26 (0.73-2.18); P: 0.399 | 1.23 (1.01-1.51); P: 0.045 | 1.22 (0.71-2.10); P: 0.470 |
| Ever | 176/187 | 67/66 | 9/12 | 1.00 (0.67-1.50); P: 0.994 | 0.69 (0.28-1.70); P: 0.415 | 0.99 (0.67-1.46); P: 0.960 | 0.71 (0.29-1.73); P: 0.447 |
| Alcohol consumption | |||||||
| Never | 532/816 | 251/312 | 27/36 | 1.18 (0.96-1.44); P: 0.116 | 1.08 (0.65-1.81); P: 0.768 | 1.20 (0.99-1.46); P: 0.065 | 1.05 (0.63-1.76); P: 0.848 |
| Ever | 109/89 | 53/41 | 8/6 | 0.99 (0.60-1.64); P: 0.980 | 1.08 (0.36-3.26); P: 0.899 | 1.04 (0.64-1.68); P: 0.870 | 1.10 (0.37-3.30); P: 0.864 |
| BMI (kg/m2) | |||||||
| < 24 | 423/465 | 208/198 | 25/23 | 1.13 (0.89-1.43); P: 0.327 | 1.14 (0.63-2.04); P: 0.665 | 1.16 (0.92-1.46); P: 0.207 | 1.12 (0.63-2.00); P: 0.706 |
| ≥ 24 | 218/440 | 96/155 | 10/19 | 1.19 (0.88-1.62); P: 0.253 | 0.99 (0.45-2.18); P: 0.977 | 1.22 (0.91-1.63); P: 0.195 | 0.96 (0.44-2.11); P: 0.927 |
The genotyping was successful in 980 (97.71%) CRC cases and 1,300 (99.77%) controls for ICOS rs4404254 T>C;
Adjusted for age, sex, BMI, smoking status and alcohol consumption (besides stratified factors accordingly) in a logistic regression model.
Discussion
The etiology of CRC is very complicated, in which the individual’s hereditary factor may play a vital role. Although some susceptibility genes suggest potent associations with the hereditary nonpolyposis CRC, many low-penetrant susceptibility factors involving SNPs predisposing to CRC remain to be elucidated. Recently, a pooled-analysis highlighted that variants of Cytotoxic T lymphocyte antigen-4 (CTLA-4) might influence the risk of CRC, which suggests the important role of costimulatory molecules in the development of CRC [18]. Additionally, ICOS and CTLA-4 are both located on chromosome 2q33. CTLA-4 blockade has been proven to be an active immunotherapeutic strategy in cancer. CTLA-4 blockade results in a higher frequency of CD4+ICOShi T cell and elevated IFN-gamma levels in both nonmalignant and malignant tissues, indicating that ICOS may interact with CTLA-4, and then plays a vital role in tumor immunity [19]. Several studies have focused on the relationship of ICOS rs10932029 T>C and rs4404254 T>C polymorphisms with susceptibility of cancer [11-14]. However, the results remain conflicting. In this study, we found that ICOS rs10932029 T>C polymorphism was associated with the development of CRC in female, < 61 years, never smoking, and BMI ≥ 24 subgroups. In addition, we found that ICOS rs4404254 T>C polymorphism was associated with the risk of CRC in the never smoking subgroup.
In ICOS rs10932029 T>C, we found that CC genotype in ICOS is relevant to an increased risk of CRC among female, < 61 years, never smoking, and BMI ≥ 24 patients. Xu et al. reported that ICOS rs10932029 was associated with the development of breast cancer (BC) in Chinese women, and the C allele may be a susceptibility factor in BC [13], suggesting that ICOS rs10932029 C allele may be related to a decrease activity of T cell. We also found that ICOS rs10932029 C allele was probably increased the risk of CRC, which was very similar to the previous study. ICOS rs10932029 T>C locates on intron region. Although this SNP is a non-coding polymorphism, it is proposed that a T→C substitution may influence expression of ICOS protein by altering gene splicing. ICOS rs10932029 T>C may accordingly confer susceptibility to CRC through these potential mechanisms, and this case-control study suggested that ICOS rs10932029 CC genotype and C allele could have a significant impact on colorectal carcinogenesis. Furthermore, we presumed that our findings could be explained by aberrant expression of ICOS in the presence of a C allele. ICOS rs4404254 T>C locates on 3’UTR region. In the current study, we found a potential association between ICOS rs4404254 T>C polymorphism and the development of CRC in the never smoking subgroup. A previous study indicated that ICOS rs4404254 CC genotype may increase the risk of cervical cancer [12], which is very analogous to our findings. However, Wu et al. reported that ICOS rs4404254 T>C might decrease the risk of CRC [11]. Clearly, these ambiguous results showed that the function of ICOS rs4404254 T>C polymorphism might be altered in different ethnicity or even influence by the environmental factors, which suggested larger case-control or cohort studies in different ethnicities with detailed information of risk factors and lifestyles were needed to extensively explore the potential association.
However, there are several limitations in this study. First, all participants were enrolled from two local hospitals, which suggesting the selection bias might have occurred. Second, a replicated investigation focusing on the correlation of ICOS rs10932029 T>C and rs4404254 T>C with the susceptibility of CRC was not performed. Third, for lack of the information of other environmental factors, possibly related to the etiology of CRC, we did not further evaluate the potential interaction of environmental factors with gene variants. Fourth, due to the limited participants in some subgroups, the power might be insufficient in these groups. Finally, we only selected two functional SNPs of ICOS gene and explored the association between these polymorphisms and risk of CRC. In the future, a fine-mapping study should be performed to extensively identify the potential association between ICOS SNPs and risk of CRC.
In summary, our findings suggest that ICOS rs10932029 T>C polymorphism may be associated with development of CRC in females, never smoking, < 61 years and BMI ≥ 24 subgroups. In addition, our findings highlight that ICOS rs4404254 T>C polymorphism is associated with the risk of CRC in the never smoking subgroup. In the future, a fine-mapping study with a functional evaluation is needed to further explore the potential relationship between ICOS polymorphisms and risk of CRC.
Acknowledgements
We appreciate all subjects who participated in this study. We wish to thank Dr. Yan Liu and Mengmeng Jiang (Genesky Biotechnologies Inc., Shanghai, China) for technical support. The project was supported by the Natural Science Foundation of Fujian Province (Grant No. 2015J01435, 2017J01259), the Fujian provincial health and family planning research talent training program (Grant No. 2018-ZQN-13, 2016-1-11), the Joint Funds for the innovation of science and Technology, Fujian province (Grant No. 2017Y9077), and the National Clinical Key Specialty Construction Program.
Disclosure of conflict of interest
None.
References
- 1.Torre LA, Bray F, Siegel RL, Ferlay J, Lortet-Tieulent J, Jemal A. Global cancer statistics, 2012. CA Cancer J Clin. 2015;65:87–108. doi: 10.3322/caac.21262. [DOI] [PubMed] [Google Scholar]
- 2.Chen W, Zheng R, Baade PD, Zhang S, Zeng H, Bray F, Jemal A, Yu XQ, He J. Cancer statistics in China, 2015. CA Cancer J Clin. 2016;66:115–132. doi: 10.3322/caac.21338. [DOI] [PubMed] [Google Scholar]
- 3.Efremova M, Rieder D, Klepsch V, Charoentong P, Finotello F, Hackl H, Hermann-Kleiter N, Lower M, Baier G, Krogsdam A, Trajanoski Z. Targeting immune checkpoints potentiates immunoediting and changes the dynamics of tumor evolution. Nat Commun. 2018;9:32. doi: 10.1038/s41467-017-02424-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Marisa L, Svrcek M, Collura A, Becht E, Cervera P, Wanherdrick K, Buhard O, Goloudina A, Jonchere V, Selves J, Milano G, Guenot D, Cohen R, Colas C, Laurent-Puig P, Olschwang S, Lefevre JH, Parc Y, Boige V, Lepage C, Andre T, Flejou JF, Derangere V, Ghiringhelli F, de Reynies A, Duval A. The balance between cytotoxic T-cell lymphocytes and immune checkpoint expression in the prognosis of colon tumors. J Natl Cancer Inst. 2018:110. doi: 10.1093/jnci/djx136. [DOI] [PubMed] [Google Scholar]
- 5.Lee KS, Kwak Y, Ahn S, Shin E, Oh HK, Kim DW, Kang SB, Choe G, Kim WH, Lee HS. Prognostic implication of CD274 (PD-L1) protein expression in tumor-infiltrating immune cells for microsatellite unstable and stable colorectal cancer. Cancer Immunol Immunother. 2017;66:927–939. doi: 10.1007/s00262-017-1999-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Paulos CM, Carpenito C, Plesa G, Suhoski MM, Varela-Rohena A, Golovina TN, Carroll RG, Riley JL, June CH. The inducible costimulator (ICOS) is critical for the development of human T(H)17 cells. Sci Transl Med. 2010;2:55ra78. doi: 10.1126/scitranslmed.3000448. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Wikenheiser DJ, Stumhofer JS. ICOS co-stimulation: friend or foe? Front Immunol. 2016;7:304. doi: 10.3389/fimmu.2016.00304. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Dong C, Juedes AE, Temann UA, Shresta S, Allison JP, Ruddle NH, Flavell RA. ICOS co-stimulatory receptor is essential for T-cell activation and function. Nature. 2001;409:97–101. doi: 10.1038/35051100. [DOI] [PubMed] [Google Scholar]
- 9.Tafuri A, Shahinian A, Bladt F, Yoshinaga SK, Jordana M, Wakeham A, Boucher LM, Bouchard D, Chan VS, Duncan G, Odermatt B, Ho A, Itie A, Horan T, Whoriskey JS, Pawson T, Penninger JM, Ohashi PS, Mak TW. ICOS is essential for effective T-helper-cell responses. Nature. 2001;409:105–109. doi: 10.1038/35051113. [DOI] [PubMed] [Google Scholar]
- 10.Zhang Y, Luo Y, Qin SL, Mu YF, Qi Y, Yu MH, Zhong M. The clinical impact of ICOS signal in colorectal cancer patients. Oncoimmunology. 2016;5:e1141857. doi: 10.1080/2162402X.2016.1141857. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Wu D, Tang R, Qi Q, Zhou X, Zhou H, Mao Y, Li R, Liu C, Wang W, Hua D, Zhang X. Five functional polymorphisms of B7/CD28 co-signaling molecules alter susceptibility to colorectal cancer. Cell Immunol. 2015;293:41–48. doi: 10.1016/j.cellimm.2014.11.006. [DOI] [PubMed] [Google Scholar]
- 12.Ivansson EL, Juko-Pecirep I, Gyllensten UB. Interaction of immunological genes on chromosome 2q33 and IFNG in susceptibility to cervical cancer. Gynecol Oncol. 2010;116:544–548. doi: 10.1016/j.ygyno.2009.10.084. [DOI] [PubMed] [Google Scholar]
- 13.Xu F, Li D, Zhang Q, Fu Z, Zhang J, Yuan W, Chen S, Pang D, Li D. ICOS gene polymorphisms are associated with sporadic breast cancer: a case-control study. BMC Cancer. 2011;11:392. doi: 10.1186/1471-2407-11-392. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Hu J, Li QL, Hou SH, Peng H, Guo JJ. Association of inducible T cell costimulator polymorphisms with susceptibility and outcome of hepatitis B virus infection in a Chinese Han population. Scand J Immunol. 2015;82:275–281. doi: 10.1111/sji.12319. [DOI] [PubMed] [Google Scholar]
- 15.Zhai Y, Zhao WH, Chen CM. [Verification on the cut-offs of waist circumference for defining central obesity in Chinese elderly and tall adults] . Zhonghua Liu Xing Bing Xue Za Zhi. 2010;31:621–625. [PubMed] [Google Scholar]
- 16.Zhang X, Zhang S, Li Y, Detrano RC, Chen K, Li X, Zhao L, Benjamin EJ, Wu Y. Association of obesity and atrial fibrillation among middle-aged and elderly Chinese. Int J Obes (Lond) 2009;33:1318–1325. doi: 10.1038/ijo.2009.157. [DOI] [PubMed] [Google Scholar]
- 17.Ding G, Wang Y, Chen Y, Yin J, Liu C, Fan Y, Qiu H, Tang W, Chen S. Methylenetetrahydrofolate reductase tagging polymorphisms are associated with risk of esophagogastric junction adenocarcinoma: a case-control study involving 2,740 Chinese Han subjects. Oncotarget. 2017;8:111482–111494. doi: 10.18632/oncotarget.22845. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Wang Y, Wang X, Zhao R. The association of CTLA-4 A49G polymorphism with colorectal cancer risk in a Chinese Han population. Int J Immunogenet. 2015;42:93–99. doi: 10.1111/iji.12181. [DOI] [PubMed] [Google Scholar]
- 19.Chen H, Liakou CI, Kamat A, Pettaway C, Ward JF, Tang DN, Sun J, Jungbluth AA, Troncoso P, Logothetis C, Sharma P. Anti-CTLA-4 therapy results in higher CD4+ICOShi T cell frequency and IFN-gamma levels in both nonmalignant and malignant prostate tissues. Proc Natl Acad Sci U S A. 2009;106:2729–2734. doi: 10.1073/pnas.0813175106. [DOI] [PMC free article] [PubMed] [Google Scholar]