IKBKB rs2272736 is Associated with Gastric Cancer Survival

Introduction

Gastric cancer (GC) is the fifth most common cancer and a leading cause of cancer mortality worldwide.¹ Despite the advances in surgery, chemotherapy and radiation therapy, GC is still the most prevalent cancer in East Asia, causing more than 70,000 deaths every year.² Especially, the 5-year survival of patients with GC is only 30% in China, indicating a very poor prognosis.³ GC is a highly heterogeneous genetic disease and the prognosis of most patients with advanced GC is far from satisfactory. Several studies indicate that NF-κB signaling pathway is associated with biological properties of GC.^4–6 NF-κB pathway plays important roles in tumor initiation, promotion and progression.⁷ NF-κB activation induces DNA damage, oncogenic mutations and genomic instability leading to tumor initiation. Chronic inflammation and NF-κB can also cause chromosomal instability and aneuploidy. NF-κB enhances the proliferation of initiated tumor cells by promoting the production of various cytokines, growth factors and cell cycle proteins.^8,9 Most of the altered NF-κB activity discovered in solid tumors is ascribed to enhanced expression of the cytokines activated by the IKK, which comprises the IKBKA and IKBKB catalytic subunits and the IKBKG regulatory subunit.⁸ As a fundamental regulator of NF-κB activity, not surprisingly, the activity of IKBKB is closely related to tumor development and progression, such as breast cancer,¹⁰ skin cancer,¹¹ and GC.¹²

Genetic variations might be closely associated with the clinical outcome of GC and be potential biomarkers.^13,14 SNPs are the most abundant form of DNA variation in the human genome, which can influence the prognosis of GC.^15–17 Single nucleotide polymorphisms (SNPs) located within cancer-related genes and non-coding RNAs may affect their expression levels by increasing the promoter activity and specific nuclear protein-binding affinity and other regulatory mechanisms.^18–22 So far, reports on the association between IKBKB and GC are rare, and only a few of SNPs in IKBKB were clarified.²³

Thus, this study reveals the relationships between genetic variants of IKBKB and the survival of GC in 1210 patients with primary GC in a Han Chinese population.

Materials and Methods

Ethics Statement

All patients gave their written informed consents. The study was approved by the ethical review board of Nanjing medical university.

Study Population

Totally 1210 GC patients were enrolled in this study. All the patients with newly and histologically diagnosed GC between 2005 and 2012 were recruited from Nanjing First Hospital of Nanjing Medical University (Nanjing, Jiangsu Province). Patients with other cancers were excluded from the study. Clinical features on sex, age, alcohol consumption, smoking, tumor site, tumor grade, TNM classification and Lauren classification were collected. Patients were followed up by telephone interviews for survival information from the date of diagnosis to death or the last follow-up in 2014. The median follow-up time was 31 months. Smoking status was assigned as smoker if they smoked at least one cigarette per day for more than 1 year. Drinking status was assigned as drinker if they consumed one or more glasses of alcohol weekly for at least 1 year. All patients provided written informed consent prior to the collection of their information and paraffin-embedded tissue used for research purposes.

SNP Screening

The CHB (Han Chinese in Beijing, China) and CHS (Southern Han Chinese) samples of the 1000 Genomes Project were used to identify the SNPs. Haploview 4.2 was used to select the tag SNPs. RegulomeDB²⁴ and SNPinfo²⁵ were used for SNP function prediction. The cutoff score used for RegulomeDB (old version) was 6. The FuncPred of SNPinfo was used for this study.

SNP Genotyping

Genomic DNA was extracted from the paraffin-embedded tissue using the TGuide FFPE DNA Extraction Kit (TIANGEN, OSR-M405). The SNP genotyping work was accomplished by a custom-by-design 48-Plex SNPscan™ kit. This kit was developed as patented SNP genotyping technology by Genesky Biotechnologies Inc. (Shanghai, China). The kit is based on double ligations and multiplex fluorescence polymerase chain reactions (PCR).²⁶ The genotyping success rate was at least 95% in the samples.

Statistical Analysis

Departures from Hardy–Weinberg equilibrium were tested for each SNPs using the R package SNPassoc 1.9–2. Hazard ratios (HRs) and 95% CIs were calculated to evaluate the relationships between SNPs and GC survival using the unconditional univariate and multivariate Cox proportional hazards model from the R package survival 3.1–8. Kaplan–Meier survival curves were generated by survminer 0.4.6. A P-value equal to or less than 0.05 was considered statistically significant difference. Schoenfeld residuals were used to check the proportional hazards (PH) assumption.²⁷ The results of PH tests are shown in Supplementary Table 1. Given that variable sex is responsible for the seemingly some evidence against the PH assumption (P = 0.022), sex was considered as a stratification variable in the Cox models. All the above statistical analyses were performed using R 3.6.3.

Results

Characteristics of Study Patients

The characteristics of the 1210 GC cases are shown in Table 1. The average age of these patients was 63 years, ranging from 19 to 91 years. Tumors in 34.6% of patients originated from the cardia. The most common tumor grade was poorly differentiated (55.8%). The percentages of TNM classification Ⅰ, Ⅱ, III and Ⅳ were 20.7, 17.9, 34.5 and 26.9, respectively. There were 22.5% of patients with diffuse type and 77.5% of patients with intestinal type in GC samples. Follow-up information was available for 1210 (86.7%) patients; 185 patients were lost to follow-up. Within the follow-up period, 466 (38.5%) patients died.

Table 1 Clinicopathological Characteristics of GC Patients

Selection of SNPs in IKBKB

In order to clarify the relationships between genetic variants of IKBKB and the survival of GC, tag SNPs selection in IKBKB was needed. A flow chart of SNPs selection is detailed in Figure 1. First, 72 SNPs were identified from the 1000 Genomes Project. The SNPs will require to meet the following criteria: (1) minor allele frequency (MAF) ≥ 0.05; (2) Hardy–Weinberg equilibrium (HWE) ≥ 0.05; (3) call rate > 95%. Then, tag SNPs were selected with linkage disequilibrium (LD) greater than 0.8 using HaploView. Next, RegulomeDB and SNPinfo were used for SNP function prediction and selected SNPs located in the protein coding and promoter regions. Finally, only one SNP rs2272736 were obtained for genotyping in this study. Hence, the subsequent cohort study focused on this SNP.

Figure 1 Flowchart for selecting SNPs in IKBKB. Abbreviations: MAF, minor allele frequency; HWE, Hardy–Weinberg equilibrium; LD, linkage disequilibrium.

Association Between Rs2272736 and GC Survival

The genotype count of IKBKB rs2272736 and its associations with GC survival according to four genetic models (additive, dominant, codominant and recessive model) are summarized in Table 2. No deviation from HWE (P > 0.05) was detected for IKBKB rs2272736. As is shown in Table 2, the count of the GG, GA, AA genotypes were 20, 261, 836 in GC patients. In the additive model, we found that patients with the A allele showed a significantly increased overall survival time compared to that of those with the G allele (adjusted HR = 0.83, 95% CI = 0.68–1.00, P = 0.050). Compared with GG/GA genotypes, the AA genotype showed better GC survival at a borderline significance level (adjusted HR = 0.81, 95% CI = 0.66–1.01, P = 0.055). Kaplan-Meier survival curves for co-dominant, dominant and recessive models showed consistent results with Cox PH model results (Figure 2A–C). By comparing the results of Akaike information criterion (AIC) in four genetic models (additive, dominant, co-dominant and recessive models), recessive model was chosen for stratification analysis of demographic features. As shown in Supplementary Table 2, we observed a significant association between AA genotype and GC survival in subgroup Age > 63 (adjusted HR = 0.73, 95% CI = 0.55–0.95, P = 0.021); subgroup Smoking Ever (adjusted HR = 0.74, 95% CI = 0.59–0.93, P = 0.011); subgroup Drinking Ever (adjusted HR = 0.73, 95% CI = 0.59–0.91, P = 0.005). In addition, a borderline significance level was obtained (adjusted HR = 0.78, 95% CI = 0.61–1.00, P = 0.052) in male patients.

Table 2 Association Between rs2272736 in IKBKB and Gastric Cancer Survival

Figure 2 Kaplan–Meier survival curves of IKBKB rs2272736 for overall survival in GC. (A) IKBKB rs2272736 did not show the relevance for the overall survival in the codominant model. (B) IKBKB rs2272736 did not show the relevance for the overall survival in the dominant model. (C) AA genotype of IKBKB rs2272736 associated with better overall survival in the recessive model.

The associations between IKBKB rs2272736 and clinicopathologic variables related to GC survival stratified by tumor site, tumor grade, TNM classification and Lauren classification are shown in Table 3. We showed a statistically significant correlation between the AA genotype and the GC survival in the cardia; the risk reduction for death was more than 30% (adjusted HR = 0.68, 95% CI = 0.48–0.97, P = 0.032). However, there was no significant association with respect to patients with non-cardia tumors. A subsequent stratification analysis by tumor grade revealed that the AA genotype was correlated with the GC survival in well-differentiated and moderately differentiated GC at a borderline level of significance (adjusted HR = 0.71, 95% CI = 0.49–1.02, P = 0.062), as in TNM Ⅰ and Ⅱ stages (adjusted HR = 0.63, 95% CI = 0.38–1.05, P = 0.076). Further, we found that the AA genotype is significantly associated with better prognosis in the intestinal type of gastric cancer (adjusted HR = 0.65, 95% CI = 0.51–0.84, P = 0.001). No statistically significant correlation was observed in the other subgroups.

Table 3 Stratification Analyses for rs2272736 Genotypes and Gastric Cancer Survival

Prediction of Functional Effects of IKBKB Rs2272736

Based on the online tool of eQTL analysis from GTEx project, we found that rs2272736 was not an eQTL for IKBKB (data not shown). As IKBKB rs2272736 was a missense variant and located close to an intron-exon boundary, we need to predict whether it has an impact on the biological function of a protein or a possible effect on splicing. This variant was predicted to be “benign” by Polyphen-2,²⁸ “tolerated” by SIFT,²⁹ “splice site changes” by Mutation Taster,³⁰ and “splice-altering” by dbscSNV V1.1.³¹ The above results suggested that rs2272736 might affect biological functions by altering splicing rather than the structure and function of a protein.

In addition, we checked whether SNPs in LD with the SNP rs2272736 could have potential biological functions. By using LDlink,³² we identified 329 SNPs in LD with rs2272736 (D’ 0.9–1.0) (data not shown). Unfortunately, we did not find any SNP that was predicted functional.

Discussion

This study set out to gain a better understanding of the association between an IKBKB polymorphism and GC survival in a Han Chinese population. We found that A allele of rs2272736 in IKBKB was significantly associated with a reduced risk of death in patients with developing GC. Rs2272736 was not an eQTL for IKBKB and we did not find any SNP in LD with rs2272736 potentially functional. Further, function prediction indicated that rs2272736 might have an effect on splicing, which could be related to the survival differences of GC patients.

IKBKB, inhibitor of nuclear factor kappa B kinase subunit beta, is a key component of the canonical IKK complex in the conventional pathway of NF-κB activation, which is located at 8p11.21.^33,34 IKBKB phosphorylates the inhibitors of NF-κB on 2 critical residues, followed by the polyubiquitination of the inhibitors and subsequent degradation by the proteasome.^35–37 In addition to involving many diseases associated with chronic or acute inflammation, IKBKB and NF-κB exhibit higher constitutive activity or aberrant regulation reported in multiple types of cancers.^{10–12,32–39} However, IKBKB function in GC initiation and progression is far from being understood. The role of IKBKB was proposed to be the promotion of the proliferation of the GC cells dependent on the activation of β-catenin and Erk pathways,⁴⁰ the regulation of gastric carcinogenesis through anti-apoptotic signaling and cell proliferation,¹² and the regulation of anoikis resistance by DBC1 in GC cells.⁴¹ These studies suggested that IKBKB acts as an important oncogene facilitating GC progression, but further experiments are needed to further confirm the functions of IKBKB.

We evaluated the effect of IKBKB on clinicopathological phenotypes of GC. A previous study showed that IKBKB rs2272733 was associated with gefitinib-induced skin toxicity in East Asian non-small cell lung cancer (NSCLC) patients, which can serve as a potential biomarker for predicting gefitinib-related Adverse Drug Reactions.⁴² Moreover, Schizandrin A can form key hydrophobic interactions with IKBKB, contributing to its potent IKBKB inhibitory effect, which may enhance the efficacy of gefitinib in NSCLC patients.⁴³ In addition, elevated IKBKB protein expression was considered a be a marker for higher nuclear grade tumors and significantly shorter survival in clear cell renal cell carcinoma patients.⁴⁴ Accordingly, there is an unambiguous relationship between IKBKB and clinicopathological variables of various forms of cancer. However, no studies have yet been found on the relationship between genetic variants of IKBKB and the GC survival. In our study, we found that the missense polymorphism rs2272736 (allele A) is associated with reduced risk of death for GC. In addition, results showed that compared with the GG/GA genotypes, the AA genotype of rs2272736 in IKBKB was significantly correlated to longer overall survival in cardiac, well-differentiated and moderately differentiated, intestinal type of and early-stage GC. As mentioned in our results, a possible mechanism underlying the effectiveness of rs2272736 is that it may have an influence on splicing, and further studies are needed to confirm and validate these findings.

This study has examined the impact of IKBKB rs2272736 on the prognosis of GC. Although the results of this study support the view that IKBKB has important implications for gastric tumorigenesis in terms of genetic associations, it has certain limitations due to a Han Chinese population of limited size. Therefore, further studies are needed to clarify the relationship between IKBKB and the development and outcome of GC.