A single nucleotide polymorphism CTSB rs12898 is associated with primary hepatic cancer in a Chinese population

Abstract

Background & Aims: Primary hepatic cancer (PHC) is a common malignant tumor and the third most frequent cause of cancer-related death worldwide. However, the molecular mechanisms underlying hepatic cancer remain unknown. CTSB is considered a biomarker of cancer as it can facilitate tumor progression. We aimed to investigate the association between genetic polymorphisms of potential regulatory SNPs in the CTSB gene and PHC. Methods: The relationship between CTSB rs12898 and PHC was analyzed in a case-control study with a Chinese population of 608 PHC patients and 608 healthy individuals using SPSS 21.0. Results: PHC was significantly associated with alcohol consumption (P < 0.001), history of hepatitis (P < 0.001), and liver cirrhosis (P < 0.001), but not with smoking (P = 0.168), age (P = 0.175), or sex (P = 0.051). Distribution of three genotypes (GG, GA, and AA) of CTSB rs12898 significantly differed between the cases and controls (P < 0.001). Compared with the GG genotype, the GA and AA genotype was associated with a significantly increased risk of PHC (OR = 1.425, 95% CI = 1.099-1.848, P = 0.007; and OR = 2.220, 95% CI = 1.574-3.132, P < 0.001, respectively). CTSB rs12898 was associated with a significantly increased risk of PHC under a dominant model (OR = 1.592, 95% CI = 1.243-2.040, P < 0.001), and under a recessive model (OR = 1.771, 95% CI = 1.311-2.393, P < 0.001) for the variant A allele. Conclusion: Results suggest that CTSB rs12898G > A may play a role in the pathogenesis of PHC, and may be a marker for susceptibility to PHC.

Introduction

Primary hepatic cancer (PHC) is a common malignant tumor correlated with multiple genetic, viral and environmental factors [1]. Hepatic cancer is commonly diagnosed in China and was identified as one of the leading causes of cancer death from 1974 to 2015 [1,2]. The GOLOBOCAN database shows the incidence of PHC in China is spreading further [3,4]. A strong correlation between hepatic cancer and underlying chronic liver disease is well known, particularly for hepatitis B virus (HBV) and hepatitis C virus (HCV) [5]. However, the molecular mechanisms underlying liver carcinogenesis remain poorly understood, and traditional treatment approaches for hepatic cancer, such as liver transplantation, percutaneous ablation, chemoembolization, and molecular-targeted therapies all have limitations [6]. Thus, the best chance for long-term, disease-free survival can be achieved by early diagnosis before symptoms develop [7]. Therefore, facilitating early diagnosis of hepatic cancer, and improving access to and practicability of optimal treatments will relieve the existing burden of hepatic cancer in China. To achieve this goal, genetic biomarkers for identifying cancer risk, predicting the prognosis, and enabling prevention are indispensable to reduce the mortality rate of PHC.

Single nucleotide polymorphisms (SNPs) are a type of DNA sequence polymorphism. SNPs are not only genetic markers but can also affect disease susceptibility and prognosis. There have been a number of studies on genetic susceptibility for PHC. A recent study showed that rs3754093 and five other SNPs in hEXO1 are related to hepatic cancer [8]. Another study suggested that rs1143633, rs1143627 and rs3917356 of the interleukin-1 (IL1) family were found to be associated with hepatocellular carcinoma (HCC) [9]. An Egyptian study demonstrated that polymorphisms C626G rs20575 and A683C rs20576 in DR4 had roles as potential risk factors for HCC development [10]. In China, it was found that rs2267029, rs28382575, rs738791, rs738792, and rs131451 of the MMP2 gene may help predict early-stage HCC and may be biomarkers for HCC progression [11].

Cathepsin B (CTSB) is a protein-coding gene and plays an important role in protein degradation and processing [12]. CTSB is considered a cancer biomarker because it facilitates tumor progression and is upregulated in many tumor types [13]. Overexpression of the encoded protein has been associated with esophageal adenocarcinoma and other tumors [14]. A recent study found that CTSB A4383C (rs13332) is significantly associated with HCC risk, and a significantly higher frequency of large-sized tumors was observed in HCC patients, carrying C76G (rs12338) than those carrying the ancestral genotype of CTSB [15]. CTSB as an oncogene contributes to the development and progression of HCC, thus it may be a valuable prognostic marker for HCC [13].

Regulome-DB is a massive integrated database that contains high-throughput sequencing data from public datasets, genome-regulatory information from ENCODE data, and other vital experimental study data. Regulome-DB has been used to select functional genes and disease-related genes, and ensures the function has great value for accurate diagnosis, treatment, and prognosis. As a useful technology, Regulome-DB has the ability to predict whether certain SNPs have regulatory effects on transcription factor binding and gene expression. It also sorts the functional SNPs by degree of impact, which are divided into categories I-VI (the smaller figure shows a bigger possibility of impact on the transcription factor and gene expression) [16,17].

In this study, we aimed to extend the knowledge of the association between genetic polymorphisms of CTSB gene and PHC by selecting potentially regulatory SNPs using Regulome-DB.

Materials and methods

Subjects

We performed a case-control association study involving 608 PHC patients and 608 healthy individuals. Patients were diagnosed with PHC at Yanbian Hospital and Yanbian Tumor Hospital between June 2011 and August 2017. Control subjects were randomly selected from a pool of healthy volunteers who visited the same hospital for a general health examination.

Patients were diagnosed with PHC according to the Diagnosis and Treatment Standards of Primary Liver Cancer (2011 Edition) published by the Ministry of Health of the People’s Republic of China [18]. There were no age, sex, stage, or histological restrictions, but patients with a prior history of cancer were excluded. Significant pure alcohol consumption was defined as drinking over 41 g/d and less than 61 g/d of alcohol in males, and over 21 g/d and less than 41 g/d in females [19]. All patients gave informed consent and this study was approved by the ethical review committee of Yanbian University Committee for Medical and Health Research Ethics.

DNA extraction

Five milliliters of human venous blood were collected before treatment (radiotherapy, chemotherapy or interventional therapy), and all blood samples were stored in ethylene diamine tetraacetic acid (EDTA) in an anticoagulant tube (10 mL) at -80°C until use. A nucleic acid isolation system was used to extract DNA from white blood cells. The steps of DNA extraction were based on the manufacturer’s instructions of Quick Gene-810 and the matched kit (KURABO, Japan). A Nano Drop 2000 was used to test the purity and concentration of DNA samples; high quality samples were selected (concentration > 1.5 Ng/μL; A260/A280 > 1.8; A260/A230 > 1.5) for testing.

Regulome-SNP selection and genotyping

In this study, we adopted the scoring criteria of the Regulome-DB and selected SNPs in category I among categories I-VI for the credibility of the experimental results. The category I contains 39,432 potential regulatory SNPs. Because CTSB was implicated as a cancer-related biomarker, we selected rs1123140, rs12898, rs1293288, rs1874546, rs2250903, rs2645408, and rs6995787 in this gene based on the Regulome-DB database. Then, by referring the Chinese Han population data Hapmap-SNP (HapMap Data Rel 28 Phase II + III, August 10, on NCBI B36 assembly, dbSNP b126) and National Center for Biotechnology Information (NCBI) database (http://www.ncbi.nlm.nih.gov/), the rs1123140, and rs2250903 were excluded based on linkage diseqilibrium (LD, r² > 0.8) and minor allele frequency (MAF, < 10%). Therefore, five SNPs in CTSB gene (rs12898, rs1293288, rs1874546, rs2645408, and rs6995787) were remained for further test. Moreover, it was found that the over-activation of Akt downregulated CTSB in HepG2/ADM hepatic cancer cells [20]. So we took the SNPs in AKT into considerations and tested rs17726963, rs3786527, and rs7144207 in the Training Set. In addition, because CTSB interacts with Hepatitis B spliced protein (HBSP) to activate MAPK/Akt signaling in hepatoma cells, leading to enhanced migration and invasion of hepatoma cells and angiogenesis [21], rs11647753, rs11863174, and rs7202714 in the MAPK were selected for genotyping in our study. Finally, a total of eleven SNPs were tested in this association study.

In this study, a three-stage study design was used to select the most significant Regulome-SNPs associated with PHC. In stage one, we built up the Training Set samples, which comprised 382 human venous blood samples from 192 PHC patients and 190 individuals. We analyzed the association between these eleven Regulome-SNPs and PHC, and only CTSB rs12898 was determined as relevant to hepatic cancer (Table S1). In stage two, CTSB rs12898 was tested further in the Testing Set consisting of 190 hepatic cancer patients and 190 healthy controls (Table S2). In the final stage, the Independent Cohort consisted of 226 PHC patients and 228 healthy persons (Table S3). In total, the combined analysis consisted of 608 PHC patients and 608 healthy individuals (Table 4).

Table 4.

Logistic regression analyses for CTSB rs12898 genotypes and alleles in cases and controls

Genotype	CTSB rs12898		P *	Minor allele frequency		P *	OR (95% CI)	P **
	Cases	Controls		Cases	Controls
	n = 608 (%)	n = 608 (%)
GG	153 (25.16)	212 (34.87)	< 0.001	0.48	0.39	< 0.001	1.000 (Reference)
GA	322 (52.96)	313 (51.48)					1.425 (1.099-1.848)	0.007
AA	133 (21.88)	83 (13.65)					2.220 (1.574-3.132)	< 0.001
GA+AA vs. GG							1.592 (1.243-2.040)	< 0.001
AA vs. GG+GA							1.771 (1.311-2.393)	< 0.001

^*Two-sided χ² test for either genotype distribution or allele frequencies between the cases and controls.

^**OR (95% CI) and corresponding P values were calculated by binary logistic regression analysis.

CI, confidence interval; OR, odds ratio.

All samples were sent to Beijing Genomics Institute (BGI) and Sequenom MassARRAY for genotyping. This study was approved by the Institutional Review Boards of the Yanbian Hospital and Yanbian Tumor Hospital.

PCR-RFLP

Polymerase chain reaction restriction fragment length polymorphism (PCR-RFLP) was used for genotypic accuracy inspection. GG is the wild-type allele and AA is the mutant type allele of CTSB rs12898G > A (https://www.ncbi.nlm.nih.gov/pubmed/). Primers were designed using primer design software (http://biotools.nubic.northwestern.edu/OligoCalc.html) and Primer Premier 5.0 (Table 1). Restriction enzymes were identified in Endonucleases design software (http://tools.neb.com/NEBcutter2) and the reference book of New England Biolabs (Table 2). The DNA sequence products were amplified by polymerase chain reaction (PCR) and the length was 160 bp. Then, the PCR products were cut by Hind III and added to the microchip electrophoresis system (MCE-202 MultiNA, SHIMADZU) for DNA analysis. Finally, according to the band sizes, the genotype of CTSB rs12898 was determined (GG genotype, 21 bp, 139 bp; GA genotype, 21 bp, 139 bp, 160 bp; AA genotype, 160 bp). To confirm the genotyping results, approximately 10% of the samples were randomly selected for re-testing by three different investigators, and the results were found to be 100% concordant with the testing genotyping results.

Table 1.

Primer sequences and sequencing conditions of CTSB rs12898

Gene	SNP	Primer sequence	Annealing temperature (°C)	Length of product (bp)
CTSB	rs12898	Forward 5’-GAGGATTCAGCTCATAAAACAAG-3’	54.9	160
		Reverse 5’-CAAACCAGTGGCATACAAATTCA-3’	54.9

Table 2.

Restriction enzyme design for CTSB rs12898

Enzyme	Sequence length
	GG Type	GA Type	AA Type
Hind III	21 bp, 139 bp	21 bp, 139 bp, 160 bp	160 bp

Statistical analysis

Hardy-Weinberg equilibrium was assessed by a goodness-of-fit χ² test with 1 degree of freedom using SHEsis online software to test the deviations of the genotype frequencies. The demographic and clinical information of cases and controls were compared using SPSS version 21.0. Continuous variables such as age were tested using the Student’s t test. Categorical variables, such as sex, significant alcohol consumption, smoking status, and history of chronic hepatitis and liver cirrhosis were tested using the χ² test.

To test the strength of the associations between cases and controls, binary logistic regression models were used to estimate odds ratios (ORs) and 95% confidence intervals (CIs). All statistical analyses were performed using SPSS version 21.0 (IBM SPSS Statistics 21).

Results

Genotyping results confirmed by PCR-RFLP

The successfully amplified PCR products (160 bp) were selected for further studies, following agarose gel electrophoresis. Samples corresponded to the designed sizes of 21 bp, 139 bp, and 160 bp, which indicated the genotype of CTSB rs12898 GA. The fragment size of 160 bp represents the genotype of CTSB rs12898 AA, and the fragment size of 21 bp and 139 bp indicates the CTSB rs12898 GG genotype.

Patient demographics and characteristics

The demographics and characteristics of the 608 PHC patients and 608 controls enrolled in this study are listed in Table 3. The distribution of age in case and control groups obeys a normal distribution (F = 3.844, P = 0.051). The average age of the patients was 62.62 ± 9.94 years and that of the controls was 63.40 ± 10.67 years; there was no significant difference between the two groups (P = 0.189). There was also no significant difference between two groups for sex (P = 0.051). However, the patient group had a higher prevalence of significant alcohol consumption (P < 0.001), chronic hepatitis (P < 0.001), and liver cirrhosis (P < 0.001). There was no significant difference in smoking between cases and controls (P = 0.168).

Table 3.

Characteristics of the case-control study population

Variable	Cases (%) (n = 608)	Controls (%) (n = 608)	P
Age (years)	62.62 ± 9.94	63.40 ± 10.67	0.189*
Sex
Male	423 (69.57)	391 (64.31)	0.051**
Female	185 (30.43)	217 (35.69)
Smoking status
Smoking	303 (49.84)	279 (45.89)	0.168**
No smoking	305 (50.16)	329 (54.11)
Significant alcohol consumption
Yes	305 (50.16)	216 (35.53)	0.001**
No	303 (49.84)	392 (64.47)
Chronic hepatitis
Yes	477 (78.45)	25 (4.11)	0.001**
No	131 (21.55)	583 (95.89)
Liver cirrhosis
Yes	270 (44.41)	2 (0.33)	0.001**
No	338 (55.59)	606 (99.67)

^*t-test;

^**χ² test.

Genotype frequencies and PHC risk

The genotypic distribution of the control group was in Hardy-Weinberg equilibrium (P = 0.054), demonstrating these data are from the same Mendelian population and had satisfactory genetic equilibrium. The distribution of the CTSB rs12898 genotypes and alleles among the patient group was significantly different from that of the control group (P < 0.001 and P < 0.001, respectively, Table 4). Logistic regression analyses showed that compared with the GG genotype, the AA genotype was associated with a significantly increased risk of PHC (OR = 2.220, 95% CI = 1.574-3.132, P < 0.001). Moreover, compared with the GG genotype, the GA genotype showed an increased risk of PHC (OR = 1.425, 95% CI = 1.099-1.848, P = 0.007). CTSB rs12898 was associated with a significantly increased risk of PHC under a recessive model (OR = 1.771, 95% CI = 1.311-2.393, P < 0.001), and under a dominant model for the variant A allele (OR = 1.592, 95% CI = 1.243-2.040, P < 0.001).

Discussion

The human genome project has revealed that only 2% of the human genome contains protein-coding genes, with the vast majority of human genome remained as ‘junk DNA’ [22]. However, it has been proven that so-called junk DNA functions as an essential gene regulator in many biologic processes [23]. Since the existence of ncRNAs has been tested further, the concept of junk DNA is long gone [24,25]. The ENCODE project has demonstrated that 80% of the genome, mostly outside of protein-coding regions, harbors biochemically functional elements (i.e., transcription factor binding), which led to a new insight into the regulatory mechanisms of genome [26]. ENCODE also provided evidence that genetic variation in non-coding DNA affects its regulatory function in gene expression, suggesting possible roles of those variations in carcinogenesis.

CTSB has been implicated as a cancer-related biomarker, necessitating the exploration of CTSB polymorphism in the susceptibility to PHC [27]. CTSB encodes a pre-proprotein to generate multiple protein products. These products include the light and heavy chains of cathepsin B, which dimerize to form the double chain of the enzyme [28]. This enzyme is a type of lysosomal cysteine protease with both endopeptidase and exopeptidase activity and has a great affect on protein turnover and intracellular proteolysis [28]. Cathepsin B is normally associated with the lysosomes involved in autophagy and immune response, but its aberrant expression has been shown to lead to cancer [29]. Numerous studies have shown that abnormal regulation of cathepsin B overexpression is correlated with invasive and metastatic phenotypes in cancers [29]. Researches demonstrated that the regulation of cathepsin B can be altered at multiple levels during tumor expansion, thereby resulting in its overexpression and export outside the cell. This suggested the role of cathepsin B in alterations leading to carcinogenesis [29,30]. It was found that CTSB is a potential pharmacologic target for colorectal tumor therapy, because it was a significant factor in colorectal tumor development, invasion, and metastatic spread [31]. In the present study, CTSB rs12898 was identified as a novel susceptibility marker for hepatic cancer. By replicating the results in two successive validation studies, we provided credibility of CTSB rs12898 as a susceptibility marker of PHC. The results suggested that rs12898, a potential regulatory SNP in CTSB gene, may play important roles in the pathogenesis of PHC. Testing for this SNP may help identifying individuals at high risk for PHC, and screening programs for the high risk individuals may facilitate early diagnosis of PHC. However, the role of CTSB in the pathogenesis of PHC remains unclear. Therefore, further studies are needed to investigate how CTSB participates in the pathogenesis of PHC, and to confirm the association between the SNP and the risk of PHC.

In this study, we observed that alcohol consumption, history of hepatitis (HBV/HCV), and liver cirrhosis are significantly associated with the risk of PHC. Previous studies have shown that hepatic cancer is related to HBV or HCV infection, [32] which is in line with our findings. We also verified again that alcohol consumption and liver cirrhosis were significant risk factors for PHC (P < 0.001, P < 0.001, respectively), which is similar to previous studies [33-36]. All of the above results confirmed the known risk factors for PHC and showed the reliability of our study cohort. However, because this research included only people in the Yanbian area of China, further studies with people from different provinces of China are warranted for validating CTSB rs12898 as a susceptibility SNP for PHC in China.

Conclusion

These findings suggest that the functional polymorphism rs12898 in CTSB may contribute to PHC susceptibility, and that the variant A allele could increase the risk of PHC. High-risk groups who carry the CTSB rs12898 GA or AA genotype are candidates for cancer prevention and early detection.

Disclosure of conflict of interest

None.

Abbreviations

CIs

confidence intervals

EDTA

ethylene diamine tetraacetic acid

HBV

hepatitis B virus

HCC

hepatocellular carcinoma

HCV

hepatitis C virus

LD

linkage diseqilibrium

miRNAs

microRNAs

NAFLD

nonalcoholic fatty liver disease

ORs

Odds ratios

PCR

polymerase chain reaction

PCR-RFLP

polymerase chain reaction restriction fragment length polymorphism

PHC

primary hepatic cancer

SNPs

single nucleotide polymorphisms

3’ UTRs

3’ untranslated regions