The authors have declared that no competing interests exist.
Conceived and designed the experiments: DZ YL X-FS Z-GZ. Performed the experiments: DZ BZ. Analyzed the data: DZ M-JW CW. Contributed reagents/materials/analysis tools: DZ BZ YL. Wrote the paper: DZ.
Glucose-regulated protein 78 (GRP78), an endoplasmic reticulum chaperone, up-regulation serves as an efficient mechanism to promote malignant transformation of colorectal cancer (CRC) and protect CRC cells against apoptosis. Recently, the analysis of GRP78 polymorphisms has already determined that GRP78 rs391957 polymorphism could predict clinical outcome in CRC patients. Thus, we tested whether GRP78 polymorphisms are related to the risk of CRC. In this study, we detected two GRP78 polymorphisms (rs391957 (C>T) and rs430397 (G>A)) in 414 CRC cases and 502 hospital-based cancer-free healthy controls in Southwest China using a polymerase chain reaction–restriction fragment length polymorphism technique. Compared with the CC genotype, carriers of CT and TT genotypes of rs391957 polymorphism had higher risks of CRC (odds ratio (OR) = 1.39, 95% confidence interval (CI) = 1.06–1.83 for CT genotype and OR = 2.10, 95% CI = 1.06–4.14 for TT genotype, respectively). In CRC cases, the variant T allele was significantly associated with tumor invasion stage (P = 0.030), but not with status of lymph nodes metastasis (P = 0.052). Compared with the GG genotype, carriers of GA and AA genotypes of rs430397 polymorphism had higher risks of CRC (OR = 1.63, 95% CI = 1.23–2.15 for GA genotype and OR = 2.92, 95% CI = 1.23–6.94 for AA genotype, respectively). The rs430397 polymorphism was not associated with the clinicopathological characteristics of CRC. These data provide the first evidence that GRP78 rs391957 and rs430397 polymorphisms could serve as markers to predict the risk of CRC.
Colorectal cancer (CRC) is the third most common cancer worldwide, accounting for approximately 10% of total cancer cases. Almost 60% of the CRC cases occur in the developed countries
Glucose-regulated protein 78 (GRP78), also referred to as immunoglobulin heavy chain binding protein, is a member of the heat shock protein 70 family which is constitutively expressed and resides primarily in the endoplasmic reticulum (ER)
Cancer cells are often confronted with hypoxia and glucose deprivation, thereby under the condition of ER stress with overexpression of GRP78. Recently, overexpression of GRP78 has been detected in several cancers such as brain cancer, breast cancer, lung cancer, prostate cancer, gastric cancer and CRC. Additionally, GRP78 has been shown to be associated with the development and progression of cancers.
To test the hypothesis, we conducted a case-control study in a Chinese Han population to investigate whether GRP78 polymorphisms were associated with the risk and clinicopathological characteristics of CRC.
A total of 414 CRC cases and 502 cancer-free controls were recruited for this case-control study. The patients were selected sequentially from January 2009 to August 2009 at the department of Gastrointestinal Surgery, West China Hospital, Sichuan University. All the cases had newly diagnosed, untreated primary CRC, who were identified by preoperative colonoscopy and CT scan, intraoperative exploration and postoperative pathological examination. The rectal cancer was regarded as the tumor located within 15 cm distance from the anal, and the non-rectal colon cancer was regarded as the tumor located beyond 15 cm distance from the anal. Patients with other cancer history and previous radiotherapy and chemotherapy were excluded. All the selected cases agreed to participate. The hospital-based controls were randomly selected from the same hospital during routine health checkups, who were cancer-free healthy individuals identified by colonoscopy and CT scan. The control group matched the case group in gender and age. All the participants were unrelated Chinese Hans in Southwest China. Written informed consent, blood samples and clinical data were collected from all the participants according to the protocols approved by the Ethics Committee of West China Hospital of Sichuan University.
Genomic DNA was extracted from whole blood of each participant by conventional phenol/chloroform procedure. The concentration and purity of DNA were measured with a spectrophotometer. The isolated DNA was dissolved in TE buffer and stored in the refrigerator at −20°C before analysis. Two GRP78 polymorphisms (rs391957 (C>T) and rs430397 (G>A)) were genotyped using polymerase chain reaction (PCR)–restriction fragment length polymorphism technique. Briefly, forward and reverse primers were used for PCR amplification, and then the amplified products were digested with restriction endonucleases. Furthermore, the lengths of the PCR products and the digested fragments were determined by electrophoresis on 1.5% agarose gel and 3% agarose gel, separately. To genotype GRP78 rs391957 polymorphism, PCR primers were 5′-atctctcctgcgacttctga-3′ (forward) and 5′-gatggaggaagggagaacaa-3′ (reverse), and the size of amplified products was 167 bp. The PCR products were then digested with restriction endonuclease MboII. The variant T allele had the MboII restriction site and two bands (134 bp and 32 bp) were generated after the digestion, whereas the wild C allele lacked the restriction site and a single band (167 bp) was obtained. To genotype GRP78 rs430397 polymorphism, PCR primers were 5′-aattcaggacattgcatcta-3′ (forward) and 5′-tggacagcagcaccatac-3′ (reverse), and the size of amplified products was 271 bp. The PCR products were then digested with restriction endonuclease Hin1III. The variant A allele lacked the Hin1III restriction site and a single band (271 bp) was obtained, whereas the wild G allele had the restriction site and two bands (218 bp and 52 bp) were generated after the digestion. In addition, genotyping results were validated by direct DNA sequencing in a random 5% of samples for quality control. Genotype concordance was 100%.
Calculations for the power and sample size of our case-control design were carried out using PS program software. Results indicated that the number of recruited samples could provide adequate statistical power. Deviation from the Hardy-Weinberg equilibrium for each polymorphism was tested using chi-square test among the controls. The differences in demographic characteristics (e.g. gender and age) between the CRC cases and controls were compared using chi-square and t tests. The associations of genotypes distribution of the GRP78 polymorphisms with clinicopathological characteristics of the CRC cases were evaluated using chi-square test. Risk estimates were calculated for the co-dominant and dominant genetic models using the most common homozygous genotype as the referent category. The effects of genotypes of the GRP78 polymorphisms on the risk of CRC were represented as the odds ratios (ORs) with 95% confidence intervals (CIs) using unconditional logistic regression model adjusted for gender and age. All statistical tests were two sided, and P<0.05 was considered significant. All statistical analyses were performed using the PASW Statistics 18 (SPSS Inc, Chicago, IL).
Of the estimated allele ORs of 1.38 and 1.60, our study with 414 CRC cases and 502 matched controls provided the statistical power of 0.80 and 0.95 at the nominal type I error rate of 0.05, respectively, indicating that our samples could provide adequate power in identifying the association of the risk of CRC with the two GRP78 polymorphisms. The observed genotypes frequency distribution of the two GRP78 polymorphisms in the controls did not show significant deviation from Hardy-Weinberg equilibrium (data not shown).
The demographic and clinical characteristics of the study participants were summarized in
Characteristic | Cases | Controls | P | ||
N = 414 | Frequencies | N = 502 | Frequencies | ||
Mean ± SD | 59.0±13.2 | 58.2±9.1 | 0.295 |
||
Males | 241 | 58.2% | 301 | 60% | |
Females | 173 | 41.8% | 201 | 40% | 0.592 |
I | 73 | 17.7% | |||
II | 149 | 36.2% | |||
III | 139 | 33.7% | |||
IV | 51 | 12.4% | |||
T1 | 34 | 8.3% | |||
T2 | 55 | 13.3% | |||
T3 | 80 | 19.4% | |||
T4 | 243 | 59% | |||
Negative | 236 | 57.3% | |||
Positive | 176 | 42.7% | |||
No | 361 | 87.6% | |||
Yes | 51 | 12.4% | |||
Rectum | 310 | 74.9% | |||
Colon | 104 | 25.1% | |||
Well | 24 | 6% | |||
Moderate | 247 | 61.6% | |||
Poor | 130 | 32.4% | |||
Expansive | 150 | 38.4% | |||
Infiltration | 241 | 61.6% |
P was computed using t test;
P was computed using chi-square test.
Among 414 CRC cases, 2 cases had missing data on TNM stage, 13 cases had missing data on tumor differentiation and 23 cases had missing data on tumor growth pattern.
Association analyses between genotypes distribution of the two GRP78 polymorphisms and clinicopathological characteristics of the CRC cases were shown in
Characteristic (No. of cases) | GRP78 rs391957 | P |
GRP78 rs430397 | P |
||
CC (%) | CT+TT (%) | GG (%) | GA+AA (%) | |||
<59 (187) | 102 (45.7) | 85 (44.5) | 116 (47.5) | 71 (41.8) | ||
≥59 (227) | 121 (54.3) | 106 (55.5) | 0.801 | 128 (52.5) | 99 (58.2) | 0.245 |
Males (241) | 131 (58.7) | 110 (57.6) | 139 (57) | 102 (60) | ||
Females (173) | 92 (41.3) | 81 (42.4) | 0.831 | 105 (43) | 68 (40) | 0.538 |
I (73) | 42 (18.8) | 31 (16.4) | 42 (17.4) | 31 (18.2) | ||
II (149) | 71 (31.8) | 78 (41.3) | 91 (37.6) | 58 (34.1) | ||
III (139) | 81 (36.3) | 58 (30.7) | 82 (33.9) | 57 (33.5) | ||
IV (51) | 29 (13.1) | 22 (11.6) | 0.264 | 27 (11.1) | 24 (14.2) | 0.780 |
T1 (34) | 17 (7.6) | 17 (9.0) | 17 (7.0) | 17 (10.0) | ||
T2 (55) | 36 (16.1) | 19 (10.1) | 32 (13.2) | 23 (13.5) | ||
T3 (80) | 33 (14.8) | 47 (24.9) | 50 (20.7) | 30 (17.6) | ||
T4 (243) | 137 (61.5) | 106 (56.0) | 0.030 | 143 (59.1) | 100 (58.8) | 0.672 |
Negative (236) | 118 (52.9) | 118 (62.4) | 143 (59.1) | 93 (54.7) | ||
Positive (176) | 105 (47.1) | 71 (37.6) | 0.052 | 99 (40.9) | 77 (45.3) | 0.376 |
No (361) | 194 (87) | 167 (88.4) | 215 (88.8) | 146 (85.9) | ||
Yes (51) | 29 (13) | 22 (11.6) | 0.675 | 27 (11.2) | 24 (14.1) | 0.369 |
Rectum (310) | 168 (75.3) | 142 (74.3) | 182 (74.5) | 128 (75.3) | ||
Colon (104) | 55 (24.7) | 49 (25.7) | 0.817 | 62 (25.5) | 42 (24.7) | 0.852 |
Well (24) | 11 (5.1) | 13 (7.1) | 14 (6.0) | 10 (6.0) | ||
Moderate (247) | 134 (61.7) | 113 (61.4) | 148 (63.5) | 99 (58.9) | ||
Poor (130) | 72 (33.2) | 58 (31.5 | 0.688 | 71 (30.5) | 59 (35.1) | 0.611 |
Expansive (150) | 83 (39.3) | 67 (37.2) | 88 (36.6) | 67 (40.9) | ||
Infiltration (241) | 128 (60.7) | 113 (62.8) | 0.668 | 144 (63.4) | 97 (59.1) | 0.389 |
P was computed using chi-square test.
Among 414 CRC cases, 2 cases had missing data on TNM stage, 13 cases had missing data on tumor differentiation and 23 cases had missing data on tumor growth pattern.
Logistic regression analyses of genotypes of the two GRP78 polymorphisms both showed significant differences between the CRC cases and controls in
Genotype | Controls (%) | Cases (%) | OR |
P |
CC | 315 (63) | 223 (54) | 1 | |
CT | 172 (34) | 169 (41) | 1.39 (1.06–1.83) | 0.018 |
TT | 15 (3) | 22 (5) | 2.10 (1.06–4.14) | 0.033 |
CT+TT | 187 (37) | 191 (46) | 1.45 (1.11–1.89) | 0.006 |
GG | 356 (71) | 244 (59) | 1 | |
GA | 138 (27) | 154 (37) | 1.63 (1.23–2.15) | 0.001 |
AA | 8 (2) | 16 (4) | 2.92 (1.23–6.94) | 0.015 |
GA+AA | 147 (29) | 170 (41) | 1.70 (1.29–2.23) | <0.001 |
Adjusted for age and gender.
To our knowledge, this is the first study which has investigated whether GRP78 rs391957 and rs430397 polymorphisms are associated with the risk of CRC. In this hospital-based case-control study, we found that the CT heterozygous, TT homozygous and combined (CT+TT) genotypes of GRP78 rs391957 polymorphism and the GA heterozygous, AA homozygous and combined (GA+AA) genotypes of GRP78 rs430397 polymorphism were both significantly associated with higher risk of CRC in a Chinese Han population, suggesting that the two GRP78 polymorphisms probably play potential roles in the development of CRC.
In the USA, new cases with rectal cancer accounted for approximately 40% of cases with CRC
It is clear that overexpression of GRP78 occurs in various human cancers and cancer cell lines, correlating with malignancy, metastasis and poor prognosis
Several studies have demonstrated that the polymorphisms in the promoter of gene are likely to enhance or attenuate the expression of gene
We also found that the variant T allele of GRP78 rs391957 polymorphism was associated with local tumor invasion of CRC. However, due to the retrospective design and relative numbers of cases involved, further prospective biomarker embedded clinical trials and cell and tissue experiments will be necessary to validate our result and identify what role the rs391957 polymorphism plays in progression of CRC. In our study, T4 stage was the majority of the cases. Tumor stage is often associated with lack of screening, access to medical care and so on. Thus, our result might also reveal that there is a lack of screening for CRC in southwestern Chinese population. Patients with CRC were unable to get prompt diagnosis, so that the majority of them were at advanced stages when they received proper treatment. It is necessary that southwestern Chinese pay more attention to routine screening for CRC. In addition, the previous study found higher expression of GRP78 contributed to increased lymph node metastasis in gastric cancer patients
The variant T allele frequency is approximately 29% in Asian population, 36% in European population, 29% in the USA and only 9% in African population. Thus the fact that most populations share similar proportions of T carriers reinforces the importance of our findings and the need for further study. However, ethnicity is of major importance in polymorphism analysis, and our findings also need studies about other ethnic populations to confirm it in the future.
The polymorphisms in the intron of gene have received less attention, but are beginning to be recognized for their potential contribution to the development and progression of diseases, including cancers
There have been approximately 200 known polymorphisms within the GRP78 gene
There were significant associations between the risk of CRC and the GRP78 rs391957 and rs430397 polymorphisms. Both the variant T allele of rs391957 polymorphism and the variant A allele of rs430397 polymorphism were associated with higher risk of CRC in southwestern Chinese Han population, though whether the variant T allele of rs391957 polymorphism affected local tumor invasion stages and status of lymph node metastasis needs to be further tested in prospective larger-scale studies in the future. These preliminary findings suggested that the GRP78 polymorphisms could provide clues to forecast susceptibility to CRC. But the precise mechanisms underlying the relationship of the GRP78 polymorphisms and CRC also need further investigation.
We thank Xueping Zhang, Su Peng and Yaling Zhang for their assistance in collecting the blood samples, Keling Chen and Zhaoying Lu for their laboratory assistance.