Sage Journals: Discover world-class research

Abstract

BACKGROUND:

Genetic polymorphisms are believed to represent a key aspect of predisposition to gastric cancer (GC). Therefore, considering the important role of Cathepsin B (CTSB) in promoting cancer onset and development, it could be very worthful to explore the function of CTSB-related genetic polymorphisms in GC.

OBJECTIVE:

In this study, we investigated the correlation of CTSB-related polymorphisms (rs9009A>T, rs6731T>C, rs1293303G>C, rs1874547C>T, rs3779659C>T, rs17814426C>T and rs148669985C>T) with GC risk and prognosis in a case-control study of 994 cases and 1000 controls.

METHODS:

All tag single nucleotide polymorphisms (SNPs) were genotyped by polymerase chain reaction-ligase detection reaction (PCR-LDR) sequencing technology.

RESULTS:

The results indicated rs9009, rs6731 and rs17814426 correlated with decreased risks of GC (HR $=$ 0.97, $p<$ 0.001; HR $=$ 0.86, $P=$ 0.019; HR $=$ 0.85, $P=$ 0.017; respectively). Stratification analysis further showed rs17814426 variant genotypes correlated with earlier T stage ( $p=$ 0.044). In addition, GC patients carrying the C allele of rs6371 had better overall prognosis (HR $=$ 0.62, 95%CI $=$ 0.44–0.88).

CONCLUSION:

Our results firstly suggested the importance of CTSB-related polymorphisms on GC which could predict GC risk and prognosis.

Keywords

Gastric cancer Cathepsin B Polymorphism Susceptibility Prognosis

1. Introduction

Despite some therapeutic advances, gastric cancer is still the fifth most commonly diagnosed cancer and the third leading cause of cancer-related deaths worldwide. There were more than 1,000,000 newly diagnosed cases and an estimated 783,000 annual deaths due to GC in 2018 and more than 40% of them occurs in East Asia [1]. Similar to other malignant tumors, the occurrence and development of GC involves a complicated process that is related to gene-environment interactions [2]. Our previous epidemiological studies have indicated that genetic polymorphisms play an important role in the development of GC [3, 4, 5]. Due to the low early diagnostic rate of GC in China, a large proportion of patients are diagnosed in the advanced stage, leading to poor treatment effect and five-year survival rate [6]. Therefore, it is critical to identify novel and reliable biomarkers to distinguish patients with high risk of developing GC and to predict the prognosis of GC patients.

Cathepsin B (CTSB) is a cysteine protease which serves as a critical element of lysosome cascade and plays an important role on cell survival, tumor neovascularization and metastasis [7]. Aberrant expression of CTSB is closely linked to many serious diseases including various tumors, such as nasopharyngeal carcinoma [8], breast cancer [9], prostate cancer [10], hepatocellular carcinoma (HCC) [11], and colorectal cancer [12]. In gastrointestinal cancer, Chan et al. reported that CTSB was specifically expressed in most colon cancers and was significantly associated with poor overall mortality [13]. Meanwhile, CTSB also played an important role in the occurrence and metastasis of colorectal cancer [12]. In the serum of esophageal cancer patients, the expression level of CTSB was highly upregulated compared with controls, and showed a diagnostic significance [14].

Single nucleotide polymorphisms (SNPs) are the simplest form of DNA variations between individuals in a population, which could affect gene expression, mRNA conformation, and subcellular localization of mRNAs and/or proteins [15]. SNPs are genetic markers that can affect disease susceptibility and prognosis. For example, Cui et al. demonstrated that the GA and AA genotypes of CTSB rs12898 were associated with an increased risk of primary hepatic cancer (PHC) [16]. A recent study found that CTSB A4383C and C76G SNP were correlated with high risk and tumor size of liver cancer, respectively [17]. However, the association between CTSB genetic variations and GC susceptibility or prognosis remains unknown.

Given the important function of CTSB in carcinogenesis and neoplasia, we hypothesized that SNPs sited in CTSB may affect the genetic susceptibility and prognosis to GC. Therefore, we performed a study of selected tag SNPs (rs9009, rs6731, rs1293303, rs1874547, rs3779659, rs17814426 and rs148669985) across the entire CTSB locus to examine the correlation of functional CTSB genotypes with GC risk and prognosis in a Chinese population.

2. Materials and methods

2.1 Subjects

This study was approved by the ethics committee of the First Affiliated Hospital of Nanjing Medical University (Nanjing, China) and conforms with the Declaration of Helsinki. All subjects were required to sign an informed consent form before recruitment. This was a hospital-based case-control study, with 994 GC patients and 1,000 cancer-free controls. GC cases were randomly enrolled from the First Affiliated Hospital of Nanjing Medical University between 2009 and 2019. The diagnosis of GC was confirmed in all cases by gastroscopic biopsy or histopathological examination of surgical specimens. The control group consisted of randomly selected cases from the same hospital during the same period, and were matched according to gender and age. All the participants were genetically unrelated and lived in or around Jiangsu Province. Patients with secondary recurrent malignant tumors, hereditary diseases, blood transfusions, and who had undergone chemotherapy or radiotherapy were excluded from the study. Questionnaires or medical records were used to collect basic information from each participant, including age, gender, smoking history, alcohol consumption, place of residence, hypertension, diabetes, personal medical history and clinical data (tumor location, histological type, tumor invasion depth, lymph node metastasis, etc.). A total of 488 GC patients who underwent surgery between January 2012 and May 2017 were followed-up. The survival time was defined from the date of surgery to the date of follow-up or death, and the date of the last follow-up was September 2019.

2.2 SNP selection

Tag SNPs of CTSB in chromosome 8, position from 11,842,524 to 11,869,448 were selected based on population data of Chinese Han in Beijing from the 1000 Genomes Project (Ensembl GRCh37 release 87-Dec 2016). Pairwise option of Haploview 4.2 software (Cambridge, MA, USA) was used for the selection, with the threshold set as r2 $>$ 0.8. Finally, seven tag SNPs (rs9009, rs6731, rs1293303, rs1874547, rs3779659, rs17814426 and rs148669985) were chosen recapitulating all the common SNPs (minor allele frequency $>$ 0.05) located at the whole CTSB locus.

Table 1
Demographic information

Characteristics	Cases ( $n=$ 994)	Controls ( $n=$ 1000)	$P$ value
Age (y)*	63 $\pm$ 6	62 $\pm$ 9	0.090
Gender, ( $n$ (%))
Female	339 (34.1%)	375 (37.5%)
Male	655 (65.9%)	625 (62.5%)	0.114
Hypertension, ( $n$ (%))
No	661 (66.5%)	624 (62.4%)
Yes	333 (33.5%)	376 (37.6%)	0.056
Diabetes, ( $n$ (%))
No	882 (88.7%)	861 (86.1%)
Yes	112 (11.3%)	139 (13.9%)	0.076
Smoking, ( $n$ )
Never	708 (71.2%)	805 (80.5%)
Ever	286 (28.8%)	195 (19.5%)	0.120
Drinking status
Never	214 (21.5%)	134 (13.4%)
Ever	780 (78.5%)	866 (86.6%)	$<$ 0.001
Residence, ( $n$ (%))
Rural	603 (60.7%)	497 (49.7%)	$<$ 0.001
Urban	391 (39.3%)	503 (50.3%)
Family history
No	942 (94.8%)
Yes	52 (5.2%)
Tumor differentiation ( $n$ (%))
Well $+$ Moderate	553 (55.6%)
Poor	441 (44.4%)
Depth of tumor infiltration ( $n$ (%))
T1	218 (21.9%)
T2	134 (13.5%)
T3	438 (44.1%)
T4	204 (20.5%)
Lymph node metastasis ( $n$ (%))
Negative	362 (36.4%)
Positive	632 (63.6%)
Localization ( $n$ (%))
Cardia	462 (46.5%)
Non-cardia	532 (53.5%)

*, Median (25th–75th percentiles).

Table 2

Association between CTSB gene polymorphisms (rs9009, rs6731 and rs17814426) and risk of gastric cancer

genotype	Cases N	Controls N	Crude OR (95% CI)	$P$ value	Adjusted OR (95% CI) *	$P$ value
Overall	994	1000
rs9009
AA	919	868	1		1
AT	73	123	0.56 (0.41–0.76)	$<$ 0.001	0.55 (0.40–0.75)	$<$ 0.001
TT	2	9	0.21 (0.05–097)	0.046	0.19 (0.04–0.91)	0.038
AT $+$ TT	75	132	0.54 (0.40–0.72)	$<$ 0.001	0.53 (0.39–0.71)	$<$ 0.001
Allelic
A	1911	1859	1
T	77	141	0.97 (0.95–098)	$<$ 0.001
HWE		0.052
rs6731
TT	331	283	1		1
CT	476	505	0.81 (0.66–0.99)	0.036	0.81 (0.66–0.99)	0.045
CC	187	212	0.75 (0.59–0.97)	0.029	0.77 (0.59–0.99)	0.046
CT $+$ TT	663	717	0.79 (0.65–0.96)	0.016	0.79 (0.65–0.96)	0.019
Allelic
T	1138	1071	1
C	850	929	0.86 (0.76–0.98)	0.019
HWE		0.633
rs17814426
CC	344	295
CT	477	507	0.81 (0.67–0.99)	0.040	0.81 (0.66–0.99)	0.037
TT	173	198	0.75 (0.58–0.97)	0.027	0.77 (0.59–0.99)	0.047
CT $+$ TT	650	705	0.79 (0.66–0.96)	0.015	0.79 (0.65–0.95)	0.015
Allelic
C	1165	1097	1
T	823	903	0.86 (0.76–0.97)	0.017
HWE		0.455

*, Adjusted for age, sex, smoking status, drinking status, residence, hypertension, and diabetes in the logistic regression model.

2.3 DNA extraction and genotyping

Genomic DNA was extracted from 2 ml of venous blood from each participant using a standardized protocol as previously described [18]. NanoDrop spectrophotometer (ND-1000) was used to normalize the purity and concentration of all DNA samples, which were diluted to a specific concentration of 10 ng/ $\mu$ l before analysis. The genotypes of the selected tag SNPs (rs9009, rs6731, rs1293303, rs1874547, rs3779659, rs17814426 and rs148669985) were determined by the PCR-LDR sequencing technology [3]. The lengths of DNA fragments and the corresponding primer sequences are listed in Supplementary Table S1.

2.4 Statistical analysis

All statistical analyses were performed by SPSS22.0 software (IBM, Armonk, NY, USA). All tests were two-sided and $p<$ 0.05 was considered as statistically significant. Distributions of demographic characteristics, genotype and allele frequencies in cases and controls were evaluated by Pearson’s $\chi$ 2 test. Goodness-of- $\chi$ 2 test was used to evaluate the Hardy – Weinberg equilibrium (HWE) of controls. Quantitative variables that deviated from the normal distribution were summarized as intermediate variables and analyzed using the Mann-Whitney rank sum test. Odds ratio (OR) and 95% confidence interval (CI) were used to evaluate the correlation between alleles/genotypes and GC susceptibility. The Kaplan-Meier method was used to assess the survival curve, and the log-rank test was used for the difference between the curves. Since the median survival time could not be established, mean survival time (MST) was provided. The relationships between the SNPs and the overall survival (OS) of GC patients were evaluated by univariate or multivariate Cox regression analyses to calculate the hazard ratio (HR), and adjusted HR with 95% CI.

3. Results

3.1 Demographic information

The detailed characteristics of the 994 GC cases and 1000 controls were listed in Table 1. No obvious associations were found in age, gender, hypertension and diabetes between the case and control groups ( $p>$ 0.05). However, the proportions of smokers, drinkers and rural residents were evidently higher in the GC group than in the control group ( $p<$ 0.05).

Table 3

a. Stratified analyses for CTSB genotypes in cases and controls (rs9009 and rs6731

Variable AT $+$ TT vs. AA for rs9009 Logistic regression for rs9009 CT $+$ CC vs. TT for rs6731 Logistic regression for rs6731

Cases, $n$ Controls, $n$ Adjusted OR (95% CI)* $P$ value Cases, $n$ Controls, $n$ Adjusted OR (95% CI)* $P$ value

Age (y)

$\geqslant$ 62 38/474 75/400 0.40 (0.26–0.61) $<$ 0.001 353/159 343/132 0.88 (0.66–1.16) 0.367

$<$ 62 37/445 57/468 0.71 (0.45–1.10) 0.121 310/172 374/151 0.73 (0.56–0.96) 0.023

Sex

Females 31/308 49/326 0.65 (0.40–1.07) 0.090 212/127 227/98 0.61 (0.44–0.84) 0.003

Males 44/611 83/542 0.46 (0.31–0.68) $<$ 0.001 451/204 440/185 0.93 (0.73–1.18) 0.546

Smoking status

Smokers 18/268 23/172 0.63 (0.33–1.20) 0.161 195/91 139/56 0.88 (0.58–1.33) 0.538

Nonsmokers 57/651 109/696 0.73 (0.54–0.97) 0.032 468/240 578/227 0.77 (0.62–0.96) 0.022

Drinking status

Yes 15/199 17/117 0.59 (0.28–1.27) 0.179 144/70 96/38 0.82 (0.50–1.35) 0.440

No 60/720 115/751 0.53 (0.38–0.74) $<$ 0.001 519/261 621/245 0.79 (0.64–0.97) 0.027

Residence

Rural 50/553 66/431 0.61 (0.41–0.90) 0.013 397/206 353/143 0.76 (0.59–0.99) 0.043

Urban 25/366 66/437 0.45 (0.28–0.73) 0.001 266/125 363/140 0.82 (0.61–1.10) 0.187

b. Stratified analyses for CTSB genotypes in cases and controls (rs17814426)

Variable CT $+$ TT vs. CC for rs17814426 Logistic regression for17814426

Cases, $n$ Controls, $n$ Adjusted OR (95% CI)* $P$ value

Age (y),

$\geqslant$ 62 350/162 332/142 0.93 (0.71–1.23) 0.601

$>$ 62 300/182 373/152 0.68 (0.52–0.89) 0.005

Sex

Females 205/134 272/103 0.58 (0.42–0.81) 0.001

Males 445/210 433/192 0.94 (0.74–1.19) 0.588

Smoking status

Smokers 186/100 136/59 0.80 (0.53–1.20) 0.275

Nonsmokers 464/244 569/236 0.79 (0.63–0.98) 0.032

Drinking status

Yes 142/72 94/40 0.83 (0.51–1.35) 0.452

No 508/272 611/255 0.78 (0.63–0.96) 0.017

Residence

Rural 395/208 347/150 0.81 (0.62–1.05) 0.107

Urban 255/136 358/145 0.76 (0.57–1.01) 0.058

a. Stratified analyses for CTSB genotypes in cases and controls (rs9009 and rs6731
Variable	AT $+$ TT vs. AA for rs9009	Logistic regression for rs9009	CT $+$ CC vs. TT for rs6731	Logistic regression for rs6731
	Cases, $n$	Controls, $n$	Adjusted OR (95% CI)*	$P$ value	Cases, $n$	Controls, $n$	Adjusted OR (95% CI)*	$P$ value
Age (y)
$\geqslant$ 62	38/474	75/400	0.40 (0.26–0.61)	$<$ 0.001	353/159	343/132	0.88 (0.66–1.16)	0.367
$<$ 62	37/445	57/468	0.71 (0.45–1.10)	0.121	310/172	374/151	0.73 (0.56–0.96)	0.023
Sex
Females	31/308	49/326	0.65 (0.40–1.07)	0.090	212/127	227/98	0.61 (0.44–0.84)	0.003
Males	44/611	83/542	0.46 (0.31–0.68)	$<$ 0.001	451/204	440/185	0.93 (0.73–1.18)	0.546
Smoking status
Smokers	18/268	23/172	0.63 (0.33–1.20)	0.161	195/91	139/56	0.88 (0.58–1.33)	0.538
Nonsmokers	57/651	109/696	0.73 (0.54–0.97)	0.032	468/240	578/227	0.77 (0.62–0.96)	0.022
Drinking status
Yes	15/199	17/117	0.59 (0.28–1.27)	0.179	144/70	96/38	0.82 (0.50–1.35)	0.440
No	60/720	115/751	0.53 (0.38–0.74)	$<$ 0.001	519/261	621/245	0.79 (0.64–0.97)	0.027
Residence
Rural	50/553	66/431	0.61 (0.41–0.90)	0.013	397/206	353/143	0.76 (0.59–0.99)	0.043
Urban	25/366	66/437	0.45 (0.28–0.73)	0.001	266/125	363/140	0.82 (0.61–1.10)	0.187

b. Stratified analyses for CTSB genotypes in cases and controls (rs17814426)
Variable	CT $+$ TT vs. CC for rs17814426	Logistic regression for17814426
	Cases, $n$	Controls, $n$	Adjusted OR (95% CI)*	$P$ value
Age (y),
$\geqslant$ 62	350/162	332/142	0.93 (0.71–1.23)	0.601
$>$ 62	300/182	373/152	0.68 (0.52–0.89)	0.005
Sex
Females	205/134	272/103	0.58 (0.42–0.81)	0.001
Males	445/210	433/192	0.94 (0.74–1.19)	0.588
Smoking status
Smokers	186/100	136/59	0.80 (0.53–1.20)	0.275
Nonsmokers	464/244	569/236	0.79 (0.63–0.98)	0.032
Drinking status
Yes	142/72	94/40	0.83 (0.51–1.35)	0.452
No	508/272	611/255	0.78 (0.63–0.96)	0.017
Residence
Rural	395/208	347/150	0.81 (0.62–1.05)	0.107
Urban	255/136	358/145	0.76 (0.57–1.01)	0.058

*, Adjusted for age, sex, smoking status, drinking status, residence, hypertension, and diabetes in the logistic regression model.

3.2 Association between CTSB tag SNPs and GC risk

The genotypes and allelic frequencies of selected CTSB tag SNPs and their correlations with GC risk were detailed in Table 2 and Supplementary Table S3. All seven polymorphisms’ distributions were consistent with the HWE ( $p>$ 0.05) (Table 2 and Table S3). For rs9009A $>$ T, the T allele frequency of the controls was dramatically higher than that of the GC group ( $p<$ 0.001). Compared with the AA genotype, the mutated AT (OR $=$ 0.56, 95%CI $=$ 0.41–0.76), TT (OR $=$ 0.21, 95%CI $=$ 0.05–0.97), and AT $+$ TT (OR $=$ 0.54, 95%CI $=$ 0.40–0.72) genotypes of rs9009 were significantly associated with reduced GC risk. The rs6731T $>$ C C allele frequency of the control group was also significantly higher than that of the GC group ( $p=$ 0.003). Considering the TT genotype as a control, the mutated CT (OR $=$ 0.81, 95%CI $=$ 0.66–0.99), CC (OR $=$ 0.75, 95%CI $=$ 0.59–0.97) and CT $+$ CC (OR $=$ 0.79, 95%CI $=$ 0.65–0.96) genotypes were significantly associated with decreased risk of GC either. For rs17814426C $>$ T, variant CT (OR $=$ 0.81, 95%CI $=$ 0.67–0.99), TT (OR $=$ 0.75, 95%CI $=$ 0.58–0.97), CT $+$ TT (OR $=$ 0.79, 0.66–0.96) genotypes and T allele (OR $=$ 0.86, 95%CI $=$ 0.76–0.97) were also associated with a significant reduction. However, there were no significant correlations between the other four CTSB tag SNP genotypes (rs1293303, rs1874547, rs3779659 and rs148669985) and GC risk ( $P>$ 0.05) (Supplementary Table S3).

3.3 Stratification analysis of SNPs and GC risk

To exclude potential GC risk factors, we conducted a stratified analysis of rs9009, rs6731, and rs17814426 based on age (62 years), gender, smoking status, drinking status, and residence. As shown in Tables 3a and b, the reduced risk associated with the rs9009 variant genotype was more pronounced in individuals aged $\geqslant$ 62 years ( $p<$ 0.001, adjusted OR $=$ 0.40, 95% CI $=$ 0.26–0.61), males ( $p<$ 0.001, adjusted OR $=$ 0.46, 95% CI $=$ 0.31–0.68), non-smokers ( $p=$ 0.032, adjusted OR $=$ 0.73, 95% CI $=$ 0.54–0.97), non-drinkers ( $p<$ 0.001, adjusted OR $=$ 0.53, 95% CI $=$ 0.38–0.74). For rs6731 variant genotypes, an obviously reduced risk of GC was observed in younger subjects ( $p=$ 0.023, adjusted OR $=$ 0.73, 95% CI $=$ 0.56–0.96), females ( $p=$ 0.003, adjusted OR $=$ 0.61, 95% CI $=$ 0.44–0.84), non-smokers ( $p=$ 0.022, adjusted OR $=$ 0.77, 95% CI $=$ 0.62–0.96), non-drinkers ( $p=$ 0.027, adjusted OR $=$ 0.79, 95% CI $=$ 0.64–0.97), and individuals living in rural areas ( $p=$ 0.043, adjusted OR $=$ 0.76, 95% CI $=$ 0.59–0.99). For rs17814426, the variant genotypes consisting of the CT and TT variants significantly decreased the risk of GC in subjects aged $<$ 62 years ( $p=$ 0.005, adjusted OR $=$ 0.68, 95% CI $=$ 0.52–0.89), females ( $p=$ 0.001, adjusted OR $=$ 0.58, 95% CI $=$ 0.42–0.81), non-smokers ( $p=$ 0.032, adjusted OR $=$ 0.79, 95% CI $=$ 0.63–0.98), and non-drinkers ( $p=$ 0.017, adjusted OR $=$ 0.78, 95% CI $=$ 0.63–0.96). No associations were found in the opposite subgroups.

Table 4
Associations between variant rs17814426 genotypes and clinicopathologic characteristics of gastric cancer

Variable	CT $+$ TT vs. CC for rs17814426		Logistic regression for rs17814426
	CT $+$ TT, N	CC, N	Adjusted OR (95% CI)*	$P$ value
Tumor differentiation
Well	22	9	1
Moderate	111	54	1.36 (0.58–3.17)	0.481
Poor	530	268	1.16 (0.55–2.45)	0.690
Depth of tumor infiltration
T1	151	67	1
T2	96	38	0.89 (0.56–1.43)	0.641
T3	284	154	0.79 (0.55–1.13)	0.195
T4	132	72	0.65 (0.43–098)	0.044
Lymph node metastasis
Negative	245	117	1
Positive	418	214	0.90 (0.69–1.19)	0.464
Localization
Cardia	313	149	1
Non-cardia	350	182	0.99 (0.77–1.30)	0.991

*, Adjusted for age, sex, smoking status, drinking status, residence, hypertension, and diabetes in the logistic regression model.

Furthermore, we investigated the interactions between the mutant genotypes and clinicopathological features in the GC group (Table 4 and Supplementary Table S4). No obvious correlation was observed between the SNPs and clinicopathological features, including the depth of tumor invasion, degree of differentiation, lymph node metastasis, or tumor location. However, for patients with CT and TT variants of rs17814426, it suggested less inclined to deeper tumor infiltration (T4) ( $p=$ 0.044, adjusted OR $=$ 0.65, 95% CI $=$ 0.43–0.98).

3.4 In silico analysis of the function of rs9009, rs6731 and rs17814426

First, in order to further investigate the functions of rs9009, rs6731 and rs17814426 on CTSB RNA secondary structure, we used RNAfold algorithm in silico [19]. It showed the local folding structures of CTSB were slightly different due to the allelic changes of rs9009 and rs17814426, and obviously distinct owing to the changes of the rs6731 allele (Fig. S1). Moreover, based on the GTEx portal database (https://gtexportal.org/), we detected altered expression level of CTSB by rs9009, rs6731 and rs17814426 in gastric tissues. However, only the rs9009 and rs6731 data was acquired from the database. As shown in Fig. S2, people with variant AT/TT genotypes of rs9009 had significantly lower CTSB expression levels in normal gastric tissues than the AA genotype ( $p<$ 0.05). Similarly, for rs6731, compared with the TT genotype, people with the variant CT/CC genotypes had evidently lower CTSB expression levels either.

Table 5
Survival analysis by clinicopathologic features of gastric cancer patients

Characteristics	Patients	Deaths	MST (months)	Log-rank $P$	HR (95%CI)	$P$ *
Overall	488	136
Age (y)*
$<$ 62	235	61	62.9
$\geqslant$ 62	253	75	54.2	0.060	1.38 (0.98–1.95)	0.065
Gender
Male	324	83	55.9		1
Female	164	53	59.0	0.976	1.01 (0.71–1.43)	0.976
Smoking status
Never	353	110	54.8		1
Ever	135	26	72.1	0.010	0.58 (0.38–0.89)	0.013
Drinking status
Never	392	115	57.5
Ever	96	21	63.2	0.108	0.69 (0.43–1.10)	0.116
Residence
Urban	197	46	58.5
Rural	291	90	57.2	0.148	1.29 (0.91–1.85)	0.156
Tumor differentiation
Well $+$ Moderate	112	26	65.5
Poor	376	110	55.8	0.002	1.96 (1.27–3.02)	0.002
T stage
T1 $+$ T2	176	24	66.6
T3 $+$ T4	312	112	55.0	$<$ 0.001	3.14 (2.02–4.88)	$<$ 0.001
Lymph node metastasis
Negative	191	36	63.4
Positive	297	100	57.3	$<$ 0.001	2.33 (1.58–3.42)	$<$ 0.001
Localization ( $n$ (%))
Cardia	234	68	53.2
Non-cardia	254	68	61.8	0.058	0.72 (0.51–1.02)	0.063

$P$ * was calculated by Univariate Cox regression analysis.

Table 6

Association between genotype of rs6731 and overall survival in gastric cancer patients

Genotype	Patients/deaths	MST (months)	Log-rank $P$	Univariate		Multivariate
				HR (95%CI)	$P$	HR (95%CI)*	$P$
rs6731
TT	157/70	53.4		1		1
CT	239/49	63.7	0.022	0.65 (0.44–0.95)	0.025	0.64 (0.44–0.94)	0.021
CC	92/17	61.0	0.021	0.54 (0.32–0.93)	0.026	0.58 (0.34–1.01)	0.055
CT $+$ CC	331/66	64.0	0.003	0.60 (0.42–0.85)	0.004	0.62 (0.44–0.88)	0.007

*, Adjusted for smoking status, tumor differentiation, depth of invasion and lymph nodes metastasis in Cox regression multivariate analysis.

3.5 Associations of CTSB SNPs with GC clinical pathological characteristics and prognosis

As shown in Table 5, the results of univariate survival analysis indicated that smoking status (HR $=$ 0.58, 95%CI $=$ 0.38–0.89), depth of tumor invasion (HR $=$ 3.14, 95CI% $=$ 2.02–4.88), degree of differentiation (HR $=$ 1.96, 95%CI $=$ 1.27–3.02), and lymph node metastasis (HR $=$ 2.33, 95%CI $=$ 1.58–3.42) were significantly correlated with poor prognosis of GC patients, while no association was found in variables like age, gender, drinking status, residence and tumor location. Given the important effect of rs9009, rs6731 and rs17814426 on the risk of GC, we further detected the association of the three SNPs on the overall survival rate (OS) of GC patients. As listed in Table 6 and Fig. S3, Kaplan-Meier analysis showed that GC patients with rs6731 CT/CC genotypes had a better OS than those with TT genotype (CT vs. TT, log-rank $p=$ 0.002; CC vs. TT, log-rank $p=$ 0.021; CT $+$ CC vs. TT, log-rank $p=$ 0.003). The results of univariate and multivariate Cox regression analyses indicated that rs6731 was an independent risk factor for OS of GC subjects either. However, the other two SNPs (rs9009 and rs17814426) had no significant correlation with OS in patients with GC (Supplementary Table S5). As for tag SNPs like rs1293303, rs1874547, rs3779659 and rs148669985 which were not correlated with the risk of GC, the results of univariate and multivariate Cox regression analyses also determine there were no correlations with OS in patients with GC (Supplementary Table S6). Therefore, it suggested rs6731 was correlated with better overall survival rate of GC patients.

4. Discussion

Gastric cancer is one of the most fatal cancer worldwide and always diagnosed at advanced stage in China, which suggest the urgent need of surveillance for patients at high risk of GC. Genetic polymorphisms are naturally occurring genetic variations which have been thoroughly investigated in cancers as biomarkers for diagnosis and prognosis [20]. Considering the highly prevalence of gastric cancer in China, we conducted a case-control study to explore some useful polymorphisms for the predication of GC risk and prognosis. Notably, CTSB is a cysteine protease that is involved in tumor growth, migration, invasion, angiogenesis, and metastases development [7]. In gastric cancer, Khan et al. indicated that CTSB expression in GC was negatively correlated with laminin, an extracellular matrix (ECM) protein, suggesting that CTSB is involved in ECM remodeling [21]. Furthermore, it is reported that CTSB rs12898 G $>$ A plays a role in the pathogenesis of primary liver cancer. Chen TP et al. further suggest CTSB A4383C SNP was more frequent in hepatocellular carcinoma (HCC) patients than in healthy controls, which was related to highly risk of HCC [17]. However, the underlying mechanism of CTSB in GC and the correlation of CTSB-related polymorphisms with GC risk and prognosis has not been investigated so far. Herein, in this study, we investigated the relationship between selected CTSB tag SNPs (rs9009, rs6731, rs1293303, rs1874547, rs3779659, rs17814426 and rs148669985) and the risk and prognosis of GC in a Chinese population.

First, the results suggest that variant genotypes of rs9009 (AT/TT), rs6731 (CT/CC) and rs17814426 (CT/TT) were related to significantly reduced risks of GC. The exact mechanism by which these SNPs reduce GC risk is unclear. However, considering the location of rs9009 and rs6731 were in the 3’UTR area of CTSB, so they may affect the expression of CTSB by altering the regulation of miRNA to CTSB [22]. Moreover, given that single base conversion may affect the local secondary structure of RNA, the RNAfold algorithm was used to predict the CTSB folding structure of rs9009T $>$ A, rs6731C $>$ T and rs17814426C $>$ T. The results showed that rs6731 obviously changes the folding patterns, indicating that SNPs may participate in the GC risk or prognosis by modifying the specific structural motifs of CTSB and therefore affecting its stability or interactions with other molecules [23]. However, further studies are needed to explore the exact mechanisms of these interactions.

Then, the stratified analysis in this study showed that the reduced risk of rs9009 variant genotypes were more pronounced among individuals $\geqslant$ 62 years of age, men, non-smokers, non-drinkers, and urban and rural residents. For rs6731 variant genotypes, an obviously decreased risk of GC was observed in younger subjects, non-smokers, and individuals living in rural areas. For rs17814426, the variant genotypes significantly decreased the risk of GC in subjects aged $<$ 62 years, females, non-smokers, and non-drinkers. As reported, the risk of carcinogenesis is considered to increase with the accumulation of genetic events during aging, and GC shows a steep, age-related increase in incidence [24]. Therefore, the effects of rs9009, rs6731 and rs17814426 genotypes tend to be age-specific. Meanwhile, a previous study reported that gastric cancer was more common in males than in females, with a ratio of nearly 2:1, which indicated the interaction between rs9009 or rs17814426 with gender to result in lower GC risk. Besides that, tobacco smoking and alcohol drinking are well-known independent risk factors of GC [25]. The association between these three polymorphisms and GC risk may be obscured by the collective exposure to tobacco carcinogens or alcohol consumption, so the associations between non-smokers or non-drinkers were more obvious [26]. Environmental factors may explain the different effects of urban and rural to individual polymorphisms that genetic differences have a greater impact on low environmental pollution [27].

Finally, the association of these polymorphisms with clinicopathological characteristics were evaluated. The results showed that the patients with CT and TT variants of rs17814426 suffered earlier tumor infiltration, indicating that depth of tumor infiltration may be regulated by genetic variations, leading to different GC development rate. However, further studies are needed to verify this hypothesis. In order to investigate the correlation of these 3 SNPs with GC prognosis, Cox regression analysis and Log rank test were applied. The results showed that only rs6731 was associated with better overall survival rate. In view of that, we think it may be caused by the lack of information of other clinicopathological characteristics such as pathological stages of tumors, postoperative treatments, Lauren classification and the omission of analysis with these deficient clinicopathological information. Meanwhile, the relatively small sample size of rs9009 CC and rs1874547 TT genotype ( $n=$ 1) may also affect the analysis. However, several limitations in this study should not be neglected. First, all cases and controls were consecutively enrolled from the same hospital, during the same period, which may have led to selection bias. Second, due to the lack of relevant clinical data, other independent risk factors of GC, such as Helicobacter pylori, high salt diet, chronic gastric ulcer and high BMI were not examined in the present study. Third, Due to the absence of information on postoperative treatment like chemoradiotherapy, we failed to analyze whether these factors affect our statistical results or whether selected tag SNPs affect the efficacy of above treatments. In addition, survival analyses of progression-free survival or disease-free survival were also not conducted because of the lack of information like progression and recurrence of GC. Finally, this study was conducted on a Chinese Han population, so the generalizability of the findings may be limited in other ethnic groups.

In summary, this study firstly demonstrated that three CTSB SNPs (rs9009, rs6731 and rs17814426) were significantly associated with reduced GC risk in a Chinese population and the variant genotype of rs6731 correlated with better prognosis. Additional studies with larger sample sizes, more clinical information, various populations, as well as mechanistic studies are needed to verify these preliminary findings.

Funding

This work was financially supported by the Liyang City’s 2018 Annual research and development Plan Follows Nanjing Project (LC2019002), the National Nature Science Foundation of China (81874219), the Natural Science Foundation of Jiangsu Province (Grant no. BK20171505).

Author contributions

L. Y. and X. M. designed this study. X. M. and H. F. were responsible for the interpretation of the results. Y. W., C.Z., Z.L. and J. X. processed the data processing and performed the statistical analysis. X. M., Y. W., and H. F. wrote the manuscript. W.C. and P. N. revised and improved the manuscript. Z.X. and L. Y. for the supervision. All authors approved the submission.

Data availability statement

The data used to support the findings of this study are available from the corresponding author upon reasonable request.

Supplementary data

The supplementary files are available to download from http://dx.doi.org/10.3233/CBM-203208.

sj-docx-1-cbm-10.3233_CBM-203208.docx - Supplemental material

Supplemental material, sj-docx-1-cbm-10.3233_CBM-203208.docx

Footnotes

Conflict of interest

The authors declare that there is no conflict of interests.

References

Bray

Ferlay

Soerjomataram

Siegel

R.L.

Torre

L.A.

and Jemal

, Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries, CA Cancer J Clin 68 (2018), 394–424.

M.S.

Chen

C.J.

and Lin

J.T.

, Host-environment interactions: their impact on progression from gastric inflammation to carcinogenesis and on development of new approaches to prevent and treat gastric cancer, Cancer Epidemiol Biomarkers Prev 14 (2005), 1878–82.

Chen

Xiao

Shi

Lin

Jiang

Fan

Yang

and Xu

, Association of TP73-AS1 gene polymorphisms with the risk and survival of gastric cancer in a Chinese Han Population, Artif Cells Nanomed Biotechnol 47 (2019), 3814–3822.

Jiang

Luo

Huan

Wang

Zhang

Yang

and Zhou

, Polymorphisms in lncRNA PTENP1 and the Risk of Gastric Cancer in a Chinese Population, Dis Markers 2017 (2017), 6807452.

Yang

Tang

Zhang

Wang

Zhang

and Yang

L.I.

, Association between LMP2 and LMP7 gene polymorphisms and the risk of gastric cancer: A case-control study, Oncol Lett 10 (2015), 509–517.

Huang

Sun

Cao

Xue

Suo

Tao

Wei

Ying

Chen

Liu

Zheng

Liu

Zhao

Chen

Wang

Xing

and Li

, Morbidity and Mortality of Laparoscopic Versus Open D2 Distal Gastrectomy for Advanced Gastric Cancer: A Randomized Controlled Trial, J Clin Oncol 34 (2016), 1350–7.

Mijanović

Branković

Panin

A.N.

Savchuk

Timashev

Ulasov

and Lesniak

M.S.

, Cathepsin B: A sellsword of cancer progression, Cancer Lett 449 (2019), 207–214.

Tan

Liu

Tang

Kang

Zhao

and Tang

, Diagnostic values of serum cathepsin B and D in patients with nasopharyngeal carcinoma, BMC Cancer 16 (2016), 241.

Bronger

Karge

Dreyer

Zech

Kraeft

Avril

Kiechle

and Schmitt

, Induction of cathepsin B by the CXCR3 chemokines CXCL9 and CXCL10 in human breast cancer cells, Oncol Lett 13 (2017), 4224–4230.

10.

Kumar

Dhar

Campanelli

Butt

N.A.

Schallheim

J.M.

Gomez

C.R.

and Levenson

A.S.

, MTA1 drives malignant progression and bone metastasis in prostate cancer, Mol Oncol 12 (2018), 1596–1607.

11.

Ruan

Zheng

Rong

Zhang

Fang

Zhao

and Luo

, Over-expression of cathepsin B in hepatocellular carcinomas predicts poor prognosis of HCC patients, Mol Cancer 15 (2016), 17.

12.

Bian

Mongrain

Cagnol

Langlois

M.J.

Boulanger

Bernatchez

Carrier

J.C.

Boudreau

and Rivard

, Cathepsin B promotes colorectal tumorigenesis, cell invasion, and metastasis, Mol Carcinog 55 (2016), 671–87.

13.

Chan

A.T.

Baba

Shima

Nosho

Chung

D.C.

Hung

K.E.

Mahmood

Madden

Poss

Ranieri

Shue

Kucherlapati

Fuchs

C.S.

and Ogino

, Cathepsin B expression and survival in colon cancer: Implications for molecular detection of neoplasia, Cancer Epidemiol Biomarkers Prev 19 (2010), 2777–85.

14.

Yan

Zhou

Wang

Chen

and Fan

, Clinical significance of serum cathepsin B and cystatin C levels and their ratio in the prognosis of patients with esophageal cancer, Onco Targets Ther 10 (2017), 1947–1954.

15.

Shastry

B.S.

, SNPs: impact on gene function and phenotype, Methods Mol Biol 578 (2009), 3–22.

16.

Cui

Chen

Wang

Sun

Lin

Cui

Jin

Park

J.Y.

Jin

Lee

S.Y.

and Cui

, A single nucleotide polymorphism CTSB rs12898 is associated with primary hepatic cancer in a Chinese population, Int J Clin Exp Pathol 12 (2019), 3063–3069.

17.

Chen

T.P.

Yang

S.F.

Lin

C.W.

Lee

H.L.

Tsai

C.M.

and Weng

C.J.

, A4383C and C76G SNP in Cathepsin B is respectively associated with the high risk and tumor size of hepatocarcinoma, Tumour Biol 35 (2014), 11193–8.

18.

Yang

H.J.

Zhu

H.J.

Sun

Q.M.

Cong

R.H.

Zhou

Tang

N.P.

and Wang

, Tissue inhibitor of metalloproteinase-2 G-418C polymorphism is associated with an increased risk of gastric cancer in a Chinese population, Eur J Surg Oncol 34 (2008), 636–41.

19.

Gruber

A.R.

Lorenz

Bernhart

S.H.

Neuböck

and Hofacker

I.L.

, The Vienna RNA websuite, Nucleic Acids Res 36 (2008), W70–4.

20.

Mocellin

Verdi

Pooley

K.A.

and Nitti

, Genetic variation and gastric cancer risk: a field synopsis and meta-analysis, Gut 64 (2015), 1209–19.

21.

Khan

Krishna

Baker

S.P.

Malhothra

and Banner

B.F.

, Cathepsin B expression and its correlation with tumor-associated laminin and tumor progression in gastric cancer, Arch Pathol Lab Med 122 (1998), 172–7.

22.

Gong

Liu

Zhang

Miao

and Guo

A.Y.

, lncRNASNP: a database of SNPs in lncRNAs and their potential functions in human and mouse, Nucleic Acids Res 43 (2015), D181–6.

23.

Shi

T.Y.

Chen

X.J.

Zhu

M.L.

Wang

M.Y.

K.D.

Shao

Z.M.

Sun

M.H.

Zhou

X.Y.

Cheng

and Wei

, A pri-miR-218 variant and risk of cervical carcinoma in Chinese women, BMC Cancer 13 (2013), 19.

24.

Milne

A.N.

Carvalho

Morsink

F.M.

Musler

A.R.

de Leng

W.W.

Ristimäki

and Offerhaus

G.J.

, Early-onset gastric cancers have a different molecular expression profile than conventional gastric cancers, Mod Pathol 19 (2006), 564–72.

25.

S.H.

Jung

Weiderpass

Jang

Hwang

Ahn

K.P.

Chang

S.H.

Shin

H.R.

Yoo

K.Y.

and Park

S.K.

, Impact of alcohol drinking on gastric cancer development according to Helicobacter pylori infection status, Br J Cancer 113 (2015), 1381–8.

26.

Yang

Sun

Zhou

Tang

Cong

Zeng

and Wang

, Gly82Ser polymorphism of the receptor for advanced glycation end products is associated with an increased risk of gastric cancer in a Chinese population, Clin Cancer Res 14 (2008), 3627–32.

27.

Hung

R.J.

Boffetta

Brennan

Malaveille

Gelatti

Placidi

Carta

Hautefeuille

and Porru

, Genetic polymorphisms of MPO, COMT, MnSOD, NQO1, interactions with environmental exposures and bladder cancer risk, Carcinogenesis 25 (2004), 973–8.

Genetic polymorphisms of Cathepsin B are associated with gastric cancer risk and prognosis in a Chinese population

Abstract

BACKGROUND:

OBJECTIVE:

METHODS:

RESULTS:

CONCLUSION:

Keywords

1. Introduction

2. Materials and methods

2.1 Subjects

2.2 SNP selection

Table 1 Demographic information

2.4 Statistical analysis

3. Results

3.1 Demographic information

3.3 Stratification analysis of SNPs and GC risk

Table 4 Associations between variant rs17814426 genotypes and clinicopathologic characteristics of gastric cancer

Table 5 Survival analysis by clinicopathologic features of gastric cancer patients

4. Discussion

Funding

Author contributions

Data availability statement

Supplementary data

sj-docx-1-cbm-10.3233_CBM-203208.docx - Supplemental material

Footnotes

Conflict of interest

References

Table 1
Demographic information

Table 4
Associations between variant rs17814426 genotypes and clinicopathologic characteristics of gastric cancer

Table 5
Survival analysis by clinicopathologic features of gastric cancer patients