Abstract
Aim: The present study aimed to identify genes that influence the susceptibility of cancer cells to radiation. Materials and Methods: The sensitivities of eight colorectal cancer cell lines to gamma radiation were tested. Microarray data and cells with stable overexpression were used to identify candidate genes. Candidate genes correlating with radioresistance were validated with the use of 22 clinical specimens obtained before preoperative radiotherapy from patients with rectal cancer. Results: Regenerating islet-derived protein 4 (REG4) gene expression was 12-fold higher in radioresistant cells. REG4-overexpressing cells had higher survival rates and fewer DNA strand breaks after gamma irradiation. Expression of the antiapoptotic gene baculoviral inhibitor of apoptosis repeat containing 5 (BIRC5) and base excision-repair pathway gene nei endonuclease VIII-like 2 (NEIL2) in REG4-overexpressing cells, was also three to four times higher than that of the parental cell lines. REG4, BIRC5 and NEIL2 expression levels were significantly higher in non-responding patients (n=14) than in responders (n=8). Conclusion: The REG4, BIRC5 and NEIL2 genes might be useful predictors of the sensitivity of cancer patients to radiotherapy.
Rectal cancer is one of the leading causes of cancer-related mortality worldwide. Although preoperative radiotherapy reduces local recurrence and improves survival in patients with resectable, advanced rectal carcinoma (1-4), the response to radiotherapy differs among tumors. Screening of patients most likely to respond is therefore recommended. Gene expression profiling has been shown to be useful for predicting clinical outcomes and classifying molecular tumor subtypes (5, 6). Eschrich et al. developed a predictive model of cellular radiosensitivity in 48 colorectal cancer (CRC) cell lines (7). Spitzner et al. identified 4,796 features that correlated with chemoradiosensitivity in 12 CRC cell lines (8). Gene expression profiling has already been used to predict the response to radiotherapy (9-11). However, no valid biomarkers identified in these studies are currently used clinically. One major reason why these microarray profiles have not been applied to clinical practice is the lack of reproducibility on quantitative analysis of gene expression. It is therefore imperative to verify these data using an alternative quantitative gene-expression approach, such as quantitative real-time reverse transcription polymerase chain reaction (RT-PCR).
We used a stepwise approach that used public DNA microarray data to profile eight CRC cell lines and thereby screen candidate genes responsible for in vitro radiation sensitivity. After validation by RT-PCR, colon cancer cells stably expressing candidate genes were established to examine changes in radiosensitivity, to identify potential biomarker genes and to study possible mechanisms for radioresistance, particularly those associated with DNA repair and apoptosis. We also performed a validation study with the identified candidate genes and rectal cancer specimens obtained from patients who received radiotherapy.
Materials and Methods
Cell cultures. Eight human CRC cell lines (Colo201, Colo205, Colo320DM, DLD1, HCT15, HCT116, HT29 and Lovo) were obtained from the American Type Culture Collection (Manassas, VA, USA) and routinely cultured under standard conditions (5% CO2, 37°C) in complete growth medium (RPMI-1640, Dulbecco's modified Eagle's medium, or Ham's F12) (Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FBS). Periodically, cell line cross-contamination was excluded by short tandem-repeat profiling (12).
In vitro X-ray irradiation and survival colony assay. To calculate the respective surviving fractions (SFs) of the adherent cell lines (DLD1, HCT116, HCT15, HT29, and Lovo) after irradiation (IR) a standard colony-forming assay was performed as described previously (13). For the colo201, colo205 and colo320DM cell lines (suspension culture type), standard methylcellulose cultures (14) were employed, with a final methylcellulose concentration of 1.2%. IR was performed at room temperature using a CABINET X-Ray system (Faxitron X-ray Corp., Wheeling, IL, USA) at dose rate of 1 Gy/min for 1-2 min. After IR, surviving cells were allowed to grow for 12-15 days. The number of colonies containing at least 50 cells was then counted, and SFs were calculated. Presented data are the means±SE (SEM) of at least three independent experiments.
Microarray data analysis. GeneChip data for eight colon cancer cell lines were downloaded from the NCBI Gene Expression Omnibus database (Affymetrix; GSE10843, http://www.ncbi.nlm.nih.gov/geo) and analyzed with GeneSpring software version 11 (Agilent Technologies, Santa Clara, CA, USA) (15).
RNA isolation and quantitative RT-PCR analysis. Total RNA was extracted using an RNeasy Mini Kit (Qiagen Inc., Valencia, CA, USA) according to the manufacturer's instructions from frozen samples of human cancer cell lines and CRC samples resected from patients before preoperative radiotherapy. Gene expression levels were determined using TaqMan RT-PCR (Applied Biosystems, Foster City, CA, USA), as described previously (16). First-strand complementary DNA (cDNA) was synthesized from total RNA using a High Capacity cDNA Reverse Transcription kit (Applied Biosystems). Each cDNA sample (2 ng/μl) was added to 3 μl RNase-free water and 4 μl of 2× TaqMan Universal PCR Master Mix (Applied Biosystems). Final primer and probe concentrations were 800 μM and 400 μM, respectively. PCR amplification was performed using an Applied Biosystems Prism 7900HT Sequence Detection System under the following thermal cycler conditions: 2 min at 50°C and 10 min at 94.5°C, then 40 cycles of 30 s at 97°C and 1 min at 60°C. Beta-actin (ACTB) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were used as reference genes. Relative gene expression was calculated by comparing the difference in cycle threshold values between the gene of interest and the endogenous control (ΔCt) for target genes and reference genes.
Establishment of a stable REG4-expressing cell line. For constitutive expression of the regenerating islet-derived protein 4 (REG4) gene, its cDNA was PCR amplified and subcloned into pIRESneo3 (Invitrogen, Carlsbad, CA, USA). The pIRES-REG4 expression vector was transfected into HCT116 and Colo320DM cells with lipofectamine LTX (Invitrogen) according to the manufacturer's instructions. Stable transfectants were selected after two weeks of culture in the presence of 200 μg/ml G418 (Wako, Osaka, Japan) for selection.
Neutral comet assay. DNA damage after IR was assessed using a whole single-cell, gel-electrophoresis assay (comet assay) according to the manufacturer's protocol (Trevigen Inc., Gaithersburg, MD) (17, 18). One hour after 1 Gy IR, cells were embedded in a thin layer of agarose spread on glass microscope slides which were immediately transferred to ice-cold lysis buffer. DNA fragmentation was quantified as the percentage of DNA in the comet tail, as analyzed by the CometScore v1.5 program (TriTek Corporation, Sumerduck, VA, USA). At least 50 randomly selected cells were analyzed per sample.
Patient samples and response to radiotherapy. Informed consent was obtained for the collection of tissue specimens from patients with rectal cancer. The study protocol was approved by the Ethics Committee of Teikyo University School of Medicine, Tokyo, Japan. Twenty-two patients with rectal cancer who gave informed consent for preoperative radiotherapy were studied. We prospectively collected biopsy specimens during colonoscopic examination before preoperative radiotherapy. Samples were snap-frozen in liquid nitrogen and stored at −80°C until use. All patients received a total radiation dose of 50.4 Gy and underwent standardized curative resection four weeks after radiotherapy. The response to radiotherapy was determined by histopathological examination of surgically resected specimens and classified according to a semiquantitative system, as described previously (10). A regression grade of 2 or 3 was classified as ‘responder’, and a grade of 0 or 1 was classified as ‘nonresponder’.
Statistical analysis. All data were statistically analyzed using JMP software, version 8.0.1 (SAS Institute, Cary, NC). All p-values were calculated with two-sided Welch's t-tests of significance; p-values <0.05 were considered to indicate statistical significance.
Results
Radiosensitivity of human CRC cell lines. We initially analyzed cellular susceptibility to IR using a colony-formation assay. Human CRC cells exposed to increasing doses of IR (0, 1 and 2 Gy) exhibited a dose-dependent decrease in surviving colonies (Figure 1). We classified these cell lines into two groups according to the SF after 1 Gy exposure: Colo201, Colo205, Colo320DM and HCT116 were highly radiosensitive (SF<50%), whereas Lovo, HCT15, HT29, and DLD1 showed low-level radiosensitivity (SF≥50%).
Microarray data mining. We identified 44 genes (using 65 probes) that were differentially expressed between high- and low-level radiosensitive cell lines (fold change value >12, t- test with false-discovery rate p<0.002). Hierarchical cluster analysis of the 65 probes categorised the cell lines into two distinct groups (Figure 2). The National Center for Biotechnology Information (NCBI) PubMed database was then searched using a gene symbol for each of the 44 genes and the search terms “colorectal” and “cancer.” Enrichment factors were calculated by dividing the number of hits with all search terms by the number of hits with the gene symbol alone. This mining procedure revealed that among the 44 candidate genes, the REG4 gene most often co-existed with the search term.
Stable REG4-expressing cell line and DNA breaks after IR. The gene expression levels of REG4 were confirmed by quantitative RT-PCR to validate the microarray data. REG4 expression showed a significant correlation (r=0.9, p<0.0001) between microarray data and real-time PCR data and differed significantly between radiosensitive and radioresistant cells (p<0.001). To investigate the biological significance of REG4, HCT116 and Colo320DM cells were stably transfected with the pIRES-REG4 expression vector. Two clones (one per cell type) were established, HCT116/REG4 and Colo320DM/REG4, which expressed REG4 messenger RNA (mRNA) at levels 7,500 and 500 times higher than those of the original cell lines, respectively. Up-regulated REG4 protein expression was also confirmed by immunostaining (data not shown). We next examined the susceptibility to IR of cells overexpressing REG4. On the basis of the number of SFs, HCT116/REG4 and Colo320DM/REG4 cells had significantly increased surviving colony counts after 1 Gy IR (p<0.05) (Figure 3A). To elucidate the radioprotective effects of REG4 on colony-forming ability, we evaluated DNA fragmentation in parental cells and those overexpressing REG4 after IR, using the neutral comet assay. As shown in Figure 3B, overexpression of REG4 in HCT116 and Colo320DM cells significantly inhibited radiation-induced DNA fragmentation as compared with parental cells.
DNA repair and apoptosis-related gene expression in stable REG4-overexpressing cell lines. To elucidate the mechanisms downstream of REG4 overexpression, we screened genes related to DNA repair and apoptosis using RT-PCR. Expression levels of only 4 out of the 43 genes tested were more than 2.5 times higher in both REG4-overexpressing cell lines than in the parental cell lines (Table I): nei endonuclease VIII-like 2 (NEIL2), NEIL3, uracil-DNA glycosylase (UNG), and baculoviral inhibitor of apoptosis repeat containing 5 (BIRC5).
REG4, NEIL2, NEIL3, UNG and BIRC5 expression in clinical samples. To determine whether REG4, NEIL2, NEIL3, UNG and BIRC5 were associated with susceptibility of CRC cells to IR, we further validated their expression levels using tissue specimens resected from patients with rectal cancer who received radiotherapy. As shown in Figure 4, expression levels of REG4, NEIL2 and BIRC5, but not NEIL3 and UNG, were significantly higher in non-responders (n=14) when compared to responders (n=8). Expression levels of REG4, NEIL2 and BIRC5 in non-responders were 8.3, 2.4 and 2.4 times higher than the levels observed in responders, respectively.
Discussion
The present study used a stepwise approach to identify candidate genes potentially responsible for the IR sensitivity of rectal cancer cells in vitro and validated the results with clinical samples. Finally, REG4, NEIL2 and BIRC5 were validated as being significantly up-regulated in radiotherapy-resistant clinical cancer samples.
REG4 is a member of the human regenerating (REG) gene family which shares strong structural similarities with proteins of the calcium-dependent lectin superfamily. REG4 has also been recently associated with radiosensitivity. Bishnupuri et al. reported that REG4 is an important modulator of gastrointestinal cell susceptibility to IR and showed that human recombinant REG4 protein increases the mRNA expression of B-cell lymphoma 2, B-cell lymphoma-extra large, and BIRC5 in mouse crypt cells in vivo (19). The unbiased data-mining procedure used in the present study confirmed that REG4 is a gene responsible for radiosensitivity.
Overexpression of REG4 led to the up-regulated expression of NEIL2, NEIL3, UNG and BIRC5 in both of the two independent cell lines. Because REG4 might be a potent activator of the epidermal growth factor receptor (EGFR)/Akt/activator protein-1 signaling pathway (20), expression control of these genes might occur downstream of the EGFR signaling pathway. BIRC5 encodes negative regulatory proteins that prevent apoptotic cell death and has been reported to be a radioresistance factor for rectal cancer (21) and to increase its expression after treatment with recombinant REG4 protein (20); the results of the present study are consistent with those of previous findings. NEIL2 belongs to a class of DNA glycosylases homologous to the bacterial Fpg/Nei family. These glycosylases initiate the first step in base excision repair (BER) by cleaving bases damaged by reactive oxygen species and by introducing a DNA strand break via the associated lyase reaction (22). To date, however, no study has shown a correlation between NEIL2 expression level and the radiosensitivity of colon cancer cells in vitro, or the response of patients with rectal cancer.
A limitation of the present study was the relatively small study group analysed. This prevented us from validating the accuracy of our predictive model in a different set of patients. Nevertheless, the candidate genes identified in this investigation might be useful for selecting patients with CRC who would benefit most from radiotherapy. To confirm the usefulness of these genes, similar studies should be performed in independent cohorts of patients. To our knowledge, this is the first report to demonstrate that gene expression levels of REG4, NEIL2 and BIRC5 correlate with outcomes in patients with CRC who receive radiotherapy.
Footnotes
-
This article is freely accessible online.
- Received September 20, 2011.
- Revision received October 26, 2011.
- Accepted October 26, 2011.
- Copyright© 2011 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved