Introduction

Colorectal cancer (CRC) is the third most common cancer with a 5-year survival rate of 30–65% [1]. A portion of the inter-patient variability in its clinical outcome is attributed to inherited and somatic genetic factors. The development and progression of CRC involves unregulated epithelial cell proliferation associated with a series of accumulated genetic alterations [2]. There is evidence that prolonged survival of such genetically unstable colorectal epithelial cells, leading eventually to their ultimate malignant transformation, is associated with progressive inhibition of apoptosis. Genetic polymorphisms for genes controlling cell cycle or apoptosis have been found to increase the risk for a number of human malignancies [3, 4]. However, such studies on CRC risk are scarce.

The two main apoptotic pathways in humans, the extrinsic or receptor-mediated pathway and the intrinsic or mitochondrial pathway, both utilize the caspase enzyme cascade. The extrinsic pathway utilizes caspase-8 (CASP8) and caspase-10 (CASP10) as initiator caspases, while the intrinsic pathway employs caspase-9 (CASP9). Both pathways converge to use caspase-3 (CASP3), caspase-6 (CASP6), and caspase-7 (CASP7) as the effector caspases, which lead to cell death by nuclear membrane breakdown, DNA fragmentation, chromatin condensation and the formation of apoptotic bodies [5, 6].

CASP8 is a key regulator of apoptosis, an essential defense mechanism against hyperproliferation and tumorigenesis. Polymorphisms in CASP8 gene have been reported to influence cancer risk. The variant D302H (rs1045485) has been associated with risk of breast cancer in the European population [7]. Additionally, the −652 6N ins/del promoter variant (rs3834129) has been associated with the risk of developing multiple cancer types, including CRC in the Chinese population [8]. It has been postulated that this deletion polymorphism has direct functional effects on cancer risk on the basis that the deletion destroy a stimulatory protein 1 binding site and decreases CASP8 transcription [8]. On the contrary, Haiman et al. [9] reported that the −652 6N ins/del promoter polymorphism in the CASP8 gene is not associated with cancer risk, and, more recently, Gangwar et al. [10] have confirmed this finding on bladder cancer.

CASP9 is a member of the intrinsic pathway, which is activated as a result of mitochondrial damage and cytochrome c release. After cytochrome c is released into the cytoplasm, it complexes with APAF-1 and forms the apoptosome with procaspase-9 and activates the CASP9 cascade. CASP3 is then activated, leading to cell death [11]. Single-nucleotide polymorphisms are the most common human genetic variation and may contribute to an individuals’ susceptibility to cancer. Many studies have demonstrated that some variants affect either the expression or the activities of various enzymes and are therefore associated with the cancer risk. Several candidate polymorphisms in the CASP9 gene have been recently reported in the public databases (http://www.ncbi.nlm.nih.gov/SNP) [12]. Although the functional effects of these polymorphisms have not been elucidated, it has been hypothesized that some of these variants, particularly their haplotypes, can influence CASP9 expression or activity, thereby modulating susceptibility to cancer.

Hence, in order to confirm or refute the purported association between CASP8 −652 6N ins/del and CASP9 −1263 A>G polymorphisms and CRC risk, we have performed a hospital-based case–control study on Greek population.

Materials and methods

Subjects

A hospital-based case–control study was undertaken. The study population consisted of a well-documented series of 402 patients (203 men and 199 women; mean age, 66.8 years; range, 33–93 years), who underwent surgery for CRC at our institutions between January 2004 and December 2009. Patients with hereditary CRC and inflammatory bowel disease-related cancers were excluded from the study. None of the patients underwent preoperative chemoradiotherapy. The tumor was located in the right colon in 91 cases (22.64%), in the left colon in 177 cases (44.03%), and in the rectum in 134 cases (33.33%). The histological grade was assessed according to WHO criteria [13]: 52 tumors (12.93%) were well differentiated, 284 (70.65%) moderately differentiated, and 64 (15.92%) poorly differentiated. According to the International Union Against Cancer classification and TNM staging system [14], 39 of the tumors (9.70%) were stage I, 141 (35.07%) stage II, 195 (48.51%) stage III, and 27 (6.72%) stage IV (Table 1). Family history and microsatellite instability (MSI) status had been evaluated in three young (<50 years of age) patients (all microsatellite-stable, MSS) and no known inherited or familial predisposition pattern had been identified. In 40 stage II patients (all T3N0M0) MSI status had also been assessed by their oncologist in order to make appropriate decisions on adjuvant treatment.

Table 1 Histopathological characteristics of Greek colorectal cancer samples at diagnosis
Full size table

The age and gender matched 480 healthy controls used were randomly selected from a pool of healthy volunteers who visited the hospital during the same period. The study was approved by the hospital review board, and written informed consent was obtained from each participant.

Genotyping

DNA was isolated from peripheral blood using the Nucleospin blood kit (Macherey-Nagel, Germany). To confirm the integrity of DNAs, initially, a 430-bp sequence of the human glyceraldehydes 3-phosphate dehydrogenase gene (GAPDH) was amplified.

Polymorphisms in CASP8 (−652 6N ins/del) and CASP9 (−1263 A>G) were analyzed by PCR restriction fragment length polymorphism. Details of primers and PCR conditions were described earlier [8, 12].

Reverse-transcriptase PCR

Total RNA was extracted from 46 fresh biopsies from the pathologic area of CRC patients using the NucleoSpin RNA isolation kit (Macherey-Nagel, Germany) according to the manufacturer’s instructions. Hereafter, reverse transcription was performed by incubating 1 μg total RNA for 1 h at 42°C in the presence of 500 μg/mL of Oligo dT 12–18, 10 mM deoxyribonucleotide triphosphates, 5× first-strand buffer, 0.1 M dithiothreitol, and 200 U/ml MMLV reverse transcriptase (Invitrogen). Prior to RT-PCR analysis, all of the RNA samples used had been DNase-treated to eliminate the risk of DNA contamination. Assessment of the CASP9 mRNA levels was performed by employing the GAPDH expression levels as a reference gene. Real-time quantitative RT-PCR was conducted on an ABI Prism 7000 apparatus (Applied Biosystems, Foster City, CA, USA). Each cDNA sample was mixed with specific primer sets and PCR master mix (Applied Biosystems, No. 4312704). The levels of genes expression were normalized after subtracting the Ct value of the GAPDH RNA internal control from that of the CASP9 Ct value for samples (∆Ct = −|Ct CASP9 (samples) − CtGAPDH|). In order to compare the levels of CASP9 expression between the samples tested, the ∆∆Ct value was determined using the formula (∆∆Ct = ∆CtCASP9(sample A) − ∆CtGAPDH(sample B)). Then, the relative level of CASP9 in cancer samples was compared to normal samples by setting the CASP9 expression in normal samples value to 1 and determining the fold change in expression against this value using the following formula 2∆∆Ct.

Statistical analysis

Genotype and allele frequencies were compared with the chi-square with Yate’s correction using S-Plus (v. 6.2, Insightful, Seattle, WA). Odds ratios (ORs) and 95 confidence intervals (CIs) were obtained with GraphPad (v. 3.00, GraphPad Software, San Diego, CA). The P values are all two-sided. Hardy–Weinberg equilibrium was verified by calculation of expected frequencies and numbers, and significance testing was based on the 1 df χ 2. The survival curves were made using the Kaplan–Meier method and comparison was with the log-rank test. Inference for the OR was aided by GraphPad InStat (version 3.00, GraphPad Software, Inc., San Diego, CA). The effect of various variables on outcome was investigated by multivariate analysis using the Cox proportional hazards model. Strong association (significance) was assumed at P < 0.001. A weak but still significant association, meriting attention, was considered for values ranging between 0.01 and 0.05.

Results

A total of 480 controls and 402 CRC cases were recruited for this study. The histological characteristics of the patients are presented in Table 1. Six patients were MSI-H. No association was found with the examined polymorphisms in the limited subgroup of patients with known MSI status. The genotypic distributions of the two gene polymorphisms were in Hardy–Weinberg equilibrium. The genotype and allele frequencies of CASP8 and CASP9 gene polymorphism in healthy controls and CRC patients are shown in Table 2. No statistically significant association was observed in CASP8. The variant −652 6N del/del genotype was 22.08% prevalent in controls compared with 24.38 in CRC cases (P = 0.88). The G allele of CASP9 −1263 A>G polymorphism was higher in controls as compared with CRC cases (50.94% versus 39.03%, P < 0.0001). Individuals with CASP9 −1263 GG genotype were at lower risk of CRC (P < 0.0001; OR = 0.39, 95%CI: 0.27–0.58).

Table 2 Caspase 8–6 N ins/del and Caspase 9–1263 A>G polymorphisms and susceptibility to colorectal cancer
Full size table

Concerning the tumor characteristics, no statistically significant association was observed in case of either CASP8 or CASP9 allelic and genotypic frequencies with any of the examined parameters (i.e., stage, differentiation status, etc.). To analyze the combined effect of these polymorphisms, we conducted gene–gene interaction analysis; however, we did not observe any statistically significant result.

Since the CASP9 −1263 G allele showed the higher association with CRC risk, we further evaluated its effect on CASP9 mRNA expression levels. As illustrated in Fig. 1, in CRC cases, homozygous carriers of CASP9 −1263 GG genotype and heterozygotes for the −1263 AG genotype expressed 6.64- and 3.69-fold higher mRNA levels of CASP9, respectively, compared to cases with the −1263 AA genotype.

Fig. 1
figure 1

Comparison of relative expression levels of CASP9 mRNA as determined by real-time RT-PCR among the different CASP9 −1263 A>G genotypes

Full size image

Follow-up information regarding survival was available for all the patients (Table 3). The median duration of the follow-up was 36 months (range, 6–60 months). A total of 118 patients suffered cancer-related death during the follow-up period. The hazard ratios of the CASP9 genotypes of the patients for overall specific survival are presented in Table 3. The −1263 AA genotype was significantly associated with poor survival (P = 0.04; Fig. 2). At multivariate analysis, the disease stage and −1263AA genotype emerge as independent variables of adverse prognostic significance (Table 4). CASP8 −652 6N ins/del polymorphism was not associated with survival.

Table 3 CASP9 −1263 A>G genotypes in association with CRC overall survival
Full size table
Fig. 2
figure 2

Kaplan–Meier survival curves for CRC patients stratified by CASP9 −1263 A>G genotypes (AA, AG, and GG)

Full size image
Table 4 Cox proportional hazard estimation of overall survival
Full size table

Discussion

Recently, genetic variants in caspase-mediated apoptosis and their role in human cancer susceptibility have been attracting increasing attention. The present study investigated the potential association between CASP8 (−652 6N ins/del) and CASP9 (−1263 A>G) polymorphisms and the risk for CRC.

The CASP8 −652 6N del allele has been found to destroy a binding element for stimulatory protein 1 and to reduce the expression of CASP8, resulting in reduction of the apoptosis reactivity of T lymphocytes upon stimulation by cancer cells or phytohemagglutinin in an ex vivo model [8]. This deletion variant was found to be associated with a reduced risk of lung, esophageal, stomach, breast, and cervical cancer in a Chinese population [8]. However, these finding failed to be replicated in subsequent larger studies [9, 10, 15]. Our data are in agreement with Pittman et al. [15] and Liu et al. [16] who also supported that CASP8 may not contribute to the risk of CRC.

CASP9 is involved in activation through apoptosome-driven intrinsic pathway [17]. The homozygous −1263 GG genotype in our study was observed to be significantly associated with reduced risk of CRC (P < 0.0001, OR = 0.39). Our observations were compatible with a Korean study and an Indian study where CASP9 −1263 A>G polymorphism was observed to be associated with significantly decreased risk of lung and bladder cancer, respectively [10, 18]. Up to now, no related study dealing with the effect of this polymorphism on CRC exist in the international literature.

To investigate if the association between CASP9 −1263 A>G polymorphism and the risk of CRC is due to differences in the transcriptional activity of the CASP9 promoter, we compared the mRNA levels in carriers of the three different genotypes (AA, AG, and GG). The carriers of the G allele exhibited higher mRNA levels than the carriers of the AA genotype. These findings are partially in agreement with Park et al. [18] who have shown that the G-C haplotype of the −1263 A>G and −712 C>T polymorphisms, have significantly higher transcriptional activity than the A–C haplotype. The mechanism by which these polymorphisms lead to a higher promoter activity is unknown. Park et al. [18] suggested that the −1263 A to G transition leads to the creation of an additional simian virus-40 protein 1-binding site. Therefore, it is possible that this change in the putative transcription factor-binding sites owing to the −1263 A>G polymorphism might lead to enhanced transcription levels. Our results suggest that the presence of G allele resulted in a higher CASP9 “production” and may offer protection against the development of CRC.

It is noteworthy that our study provided evidence for the association between CASP9 −1263 A>G polymorphism and overall survival, on the grounds of lack of any association of the CASP9 polymorphism with any of the conventional prognostic parameters, i.e., stage, that were taken into account. The potential association of GG genotype with increased survival merits further evaluation, since its oncologic significance may have important prognostic and therapeutic implications.

Excluding studies focusing on the predictive role of gene polymorphisms to the response to chemotherapy, where MSI may have a principal role, MSI has not been generally included in CRC risk studies analyzing the role of specific polymorphisms [19]. Due to the relatively small number of patients with known MSI status in the current study, no definite conclusions can be derived for a potential relationship with the examined polymorphisms. Nevertheless, the degree of influence of the major CRC molecular pathways (chromosomal instability, MSI) on genetic polymorphisms remains to be further elucidated.

In conclusion, in the current study that comprised a homogeneous Mediterranean population, the CASP9 −1263 A>G promoter polymorphism was significantly associated with the susceptibility to CRC. Since genetic polymorphisms often show ethnic differences, further studies are needed in populations with diverse backgrounds, in order to further elucidate the association between caspase gene polymorphisms and CRC.