Skip Navigation


Carcinogenesis Advance Access originally published online on January 18, 2007
Carcinogenesis 2007 28(6):1314-1322; doi:10.1093/carcin/bgm003
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrowOA All Versions of this Article:
28/6/1314    most recent
bgm003v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (2)
Google Scholar
Right arrow Articles by Huang, J.
Right arrow Articles by Xie, X.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Huang, J.
Right arrow Articles by Xie, X.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2007. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org
The online version of this article has been published under an open access model. Users are entitled to use, reproduce, disseminate, or display the open access version of this article for non-commercial purposes provided that: the original authorship is properly and fully attributed; the Journal and Oxford University Press are attributed as the original place of publication with the correct citation details given; if an article is subsequently reproduced or disseminated not in its entirety but only in part or as a derivative work this must be clearly indicated. For commercial re-use, please contact journals.permissions@oxfordjournals.org

Amino acid substitution polymorphisms of the DNA repair gene MGMT and the susceptibility to cervical carcinoma

Jinyang Huang, Feng Ye1, Huaizeng Chen, Weiguo Lu1 and Xing Xie1,*

Women's Reproductive Health Laboratory of Zhejiang Province
1 Department of Gynecologic Oncology, Women's Hospital, School of Medicine, Zhejiang University, Hangzhou, 310006, China

* To whom correspondence should be addressed. Fax: +86 571 87061878; Tel: +86 571 87061501; Email: xiex{at}mail.hz.zj.cn


   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results
 Discussion
 References
 
In the current study, we examined the association between polymorphisms in the O6-methylguanine-DNA methyltransferase gene (MGMT) and the risk for cervical carcinoma. We prospectively selected 1012 patients, including 539 with carcinoma and 473 with cervical intra-epithelial neoplasia and 800 healthy women from five hospitals in Zhejiang Province, China. Three single-nucleotide polymorphisms (Leu84Phe, Ile143Val and Lys178Arg) were genotyped, and their association with other epidemiological risk factors was examined. Compared with the MGMT Lys178Lys (AA) or Ile143Ile (AA) genotypes, women homozygous for the Arg178Arg (GG) or Val 143Val (GG) genotypes had a significantly increased risk for cervical carcinoma both in the overall carcinoma group and in the high-risk human papillomavirus-positive group. Compared with using Leu84Leu (CC), Phe84Phe (TT) and Leu84Phe (CT) which did not increase the risk for cervical carcinoma. In addition, using 84Leu (C)-143Ile (A)-178Lys (A) as reference, women carrying 84Phe (T)-143Val (G)-178Arg (G) had a 1.87-fold higher risk for cervical carcinoma (95% confidence interval 1.07–3.27). Similar results were observed for squamous cell carcinomas. The effect of the combination of Arg178Arg (GG) and Lys178Arg (AG) genotypes and the 84Phe (T)-143Val (G)-178Arg (G) haplotype was more pronounced in women infected with high-risk human papillomavirus, an early onset of sexual activity, multiple sexual partners, an early age of the first full-term pregnancy and high parity. These findings suggest that polymorphism in MGMT increases the susceptibility of women to cervical carcinoma, especially in those with high-risk sexual and reproductive histories.

Abbreviations: CI, confidence interval; CIN, cervical intra-epithelial neoplasia; HR-HPV, high-risk human papillomavirus; MGMT, O6-methylguanine-DNA-methyltransferase gene; OR, odds ratio; PCR, polymerase chain reaction; SNP, single-nucleotide polymorphism


   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results
 Discussion
 References
 
Cervical carcinoma is closely linked to infection with types of high-risk human papillomavirus (HR-HPV) (1). In ~90% of cervical carcinomas, HR-HPV DNA is integrated into the host cell genome. E6 and E7 are the major proteins responsible for the transformation and immortalization of cervical epithelial cells by HR-HPV types 16 and 18 (2). The transforming functions of E6 and E7 are thought to be due to their binding and inactivation of p53 and retinoblastoma tumor suppressor gene product, respectively (3–5). The interaction of E6 and p53 results in rapid ubiquitin-dependent proteolytic degradation of p53, which prevents cells from undergoing p53-mediated apoptosis (6). Although necessary, deregulation of E6 and E7 expression is not sufficient for the development of a malignant phenotype. Genomic instability resulting in specific genomic aberrations also participates in the development and progression of cervical carcinomas (7), and E6 has been shown to increase mutagenesis and genomic instability (8,9).

Although HR-HPVs are known to be closely linked to the development of cervical cancer, they are also found in nearly 30% of asymptomatic controls (10), indicating that HPV infection alone is not sufficient for the development of cervical cancer. Furthermore, it is clear that only a small minority of persistent HR-HPV infections progress to cervical carcinoma or its precursor lesion, high-grade cervical intra-epithelial neoplasia (CIN). Consequently, other factors, including environmental agents, such as hormonal exposure, smoking, alcohol, immunological status, sexual activity, number of sexual partners, parities and age at the first full-term pregnancy and the host genetic background, may play crucial roles in the development of cervical cancer (11–15).

Current research is focused on identifying cofactors that influence the acquisition of HPV infection or the progression of HPV infection to cancer. Most of the sexual behavior parameters that are associated with cervical cancer have been identified in the past few decades, and the role of multiple sexual partners has been of particular interest. Having had multiple partners is linked with an increased probability of HPV exposure and acquisition and disruption of local immunoreactions in the cervix. It has also been hypothesized that the age at first coitus correlates with the number of sexual partners; hence, attempts have also been made to study their independent effects (16,17). For example, in younger women, the cervix may be more vulnerable to HPV infection because of inadequate production of cervical mucus, which acts as a protective barrier against laceration and infectious agents. As a result, younger women with sexual intercourse could be at higher risk for the development of cervical cancer (18). An independent effect of early age at marriage and multiple sexual partners has been found in two studies (16,17). In this regard, the study of Brinton et al. (17) is of particular significance because it included five geographical areas.

The role of high parity has also been explored in the development of cervical cancer (11). Some recent studies have suggested that high parity and young age at first full-term pregnancy are independent risk factors for the progression of HPV infection to neoplastic cervical lesions (11,19). Other cofactors that influence the progression from HPV infection to cancer are smoking, use of oral contraceptives and sexually transmitted diseases. Despite these findings, the relationships between these cofactors and the risk for cervical carcinoma remain uncertain. There is less information about the role of reproductive and sexual history, but the available evidence suggests that these variables do not have a substantial effect on the development of cervical cancer (20). Only limited conclusions can be drawn from the previous studies of risk factors, however, because they do not consider the strong effect of the genetic background, which is of central importance in cervical carcinoma (21).

Genome stability and integrity is maintained by DNA repair genes during mitosis. It is now thought that an individual's capacity to repair DNA is genetically determined and the result of combinations of multiple genes that may have different activities. Single-nucleotide polymorphisms (SNPs) can result in small structural alterations in the repair enzymes and therefore changes in the susceptibility to cancer. More than 130 human DNA repair genes have been described (22), and a number of SNPs in these DNA repair genes have been identified (23). Defects in DNA repair pathways are linked to many different types of diseases including cervical carcinoma.

The product of the DNA repair gene O6-methylguanine-DNA-methyltransferase gene (MGMT), also known as AGT or ATase, reverses O6 alkylation damage in DNA by transferring the alkyl group to the enzyme, which inactivates it and targets it for ubiquitination and proteasome-mediated degradation. The MGMT protein has a wide range of activities in normal tissues, and its evolutionary conservation indicates that it plays a fundamental role in cell physiology and genome maintenance. This is the only protein known to be involved in the cellular defense against alkylating agents within the mammalian DNA direct reversal repair pathway (24–28). In mice, the radiation-induced increase in MGMT activity is dependent on functional p53 (29). This is different from the effect of p53 itself, whose over-expression directly suppresses transcription of the MGMT gene (30,31). Consequently, basal MGMT activity inversely correlates with expression of wild-type p53 in some tumors (32,33).

Methyltransferase activity between human fibroblast cultures from different individuals varies as much as 20-fold, suggesting that genetic polymorphisms in MGMT are risk factors for the development of cancer (34–36). Several SNPs have been reported in MGMT (23,37,38). One of these, Ile143Val (ATC -> GTC), which is located at codon 143 (exon 5), is associated with susceptibility to lung cancer and colorectal cancer (39–41). Amino acid 143 may be important because it is near the codon for the conserved alkyl group acceptor (42). Another SNP in codon 178, Lys178Arg (AAG -> AGG), has been shown to be in linkage disequilibrium with the SNP at codon 143 (43). Furthermore, although some studies have indicated that a codon 84 variant (CTT -> TTT) has similar activity and physicochemical properties as the wild-type protein (44), recent studies show that heavy smoking significantly increases the risk of breast cancer in women with the codon 84 variant T allele and that there is an inverse association between the consumption of fruits and vegetables and the risk for breast cancer in women with the wild-type codon 84 (45). The significance of these SNPs in cervical cancer and their distribution patterns in the Chinese population, however, have not been investigated.

Given that MGMT maintains genome stability, that its activity inversely correlates with the expression of wild-type p53, a key target of HR-HPV integration, and that variants of MGMT have differences in methyltransferase activity, we hypothesized that variations in MGMT contribute to the susceptibility to cervical carcinoma. To investigate this hypothesis, we carried out a large population-based, case–control study in which we determined the genotype of women for three variants of MGMT (Leu84Phe, Ile143Val and Lys178Arg) and then assessed the association of these polymorphisms with the risk for cervical carcinoma and their interaction with HR-HPV infection, sexual and reproductive risk factors.


   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results
 Discussion
 References
 
Study population
All subjects were from Zhejiang Province and were unrelated ethnic Chinese women. Patients at the Women's Hospital, School of Medicine at Zhejiang University (Hangzhou, China) and the Cancer Hospital of Zhejiang Province (Hangzhou, China) with histologically confirmed primary cervical carcinoma were recruited between June 2004 and May 2006. In the area outside of Hangzhou, a collaborative network of gynecological oncology teams in three areas (Lishui, Wenzhou and Shaoxing) of Zhejiang Province was established to identify new cases of cervical cancer at local hospitals. The diagnosis date and tumor histology was recorded for all women, and there were no age, stage, histology or prior therapy restrictions for inclusion in the study. Controls were randomly selected from healthy women seen for gynecologic examinations at the Women's Hospital, School of Medicine at Zhejiang University, during the period when women with cervical cancer and CIN were enrolled. Control selection criteria included no positive findings during the gynecological examination, no history of cancer, age matching to the patients and residence in Zhejiang Province. After providing signed informed consent, participants responded to a standard questionnaire about their sexual and reproductive history, including the number of sexual partners, the age at first intercourse, the age at first full-term pregnancy and parities (including full-term pregnancy and abortion at or after 28 weeks). A total of 1046 patients, including 561 with cervical carcinoma and 485 with CIN patients, were interviewed. Of these, 1012 eligible patients (539 with cervical carcinoma and 473 with CIN) and 800 eligible controls agreed to provide blood samples for genotyping and correctly completed the questionnaire. Of these, 675 patients (260 with cervical carcinoma and 415 with CIN) and 410 controls agreed to provide cervix brush-off samples for detection of HPV. The study was approved by the Medical Ethics Committees of all the participating hospitals. All women included in the study provided written, informed consent.

DNA extraction
White blood cells were separated from red blood cells by washing three times in phosphate-buffered saline. Next, the DNA was extracted with phenol/chloroform and precipitated with cold ethanol. All DNA samples were dissolved in water and stored at –20°C.

Genotyping
The presence of three SNPs of MGMT was determined by modified polymerase chain reaction (PCR)-mismatch amplification as described by Glaab et al. (46). For the Leu84Phe site in MGMT, the two forward primers were 5'-AGAGTTCCCCGTGCCGGCAC-3' and 5'-AGAGTTCCCCGTGCCGGCAT-3', which differ in only the last base, and the reverse primer was 5'-GCTGTTGGCAAGCAGATAGT-3'. These primers produce a 396 bp fragment. For the Ile143Val site, the two forward primers were 5'-CTTCCAGGTCCCCATCCTGA-3' and 5'-CTTCCAGGTCCCCATCCTGG-3' and the reverse primer was 5'-GCTGTTGGCAAGCAGATAGT-3', and they produce a 320 bp fragment. For the Lys178Arg site, the two forward primers were 5'-GAAGGCCACCGGTTGGGGTA-3' and 5'-GAAGGCCACCGGTTGGGGTG-3' and the reverse primer was 5'-CAGGACACTGCCACTTCCTT-3', and they produce a 214 bp fragment.

The PCR was carried out in a 25 µl reaction containing ~50 ng of genomic DNA, 5.0 pmol of each primer, 0.2 mM of each deoxynucleoside triphosphate, 1x PCR buffer [50 mM KCl, 10 mM Tris–HCl (pH 9.0 at 25°C) and 0.1% Triton X-100], 2 mM MgCl2 and 1.0 U Taq DNA polymerase (Sangon, Shanghai, China). The PCRs were performed in a Whatman Biometra T-Gradient Thermocycler (Biometra GmbH i. L., Goettingen, Germany). PCR for the Leu84Phe site was carried out under the following conditions: an initial denaturation at 94°C for 5 min; followed by 35 cycles of 94°C for 30 s, 60°C for 30 s and 72°C for 45 s, and a final elongation step of 72°C for 10 min. PCRs for the Ile143Val and Lys178Arg sites were the same except that the annealing temperatures were 65 and 68°C, respectively. The PCR products were checked by electrophoresis on a 1% agarose gel, followed by staining with ethidium bromide and visualization with a TyphoonTM 9410 Imaging System, (GE Healthcare, Chalfont Street Giles, UK). To examine the reproducibility of the results, all samples were tested twice by two technicians and the results agreed for all masked duplicate sets. To confirm the specificity of the modified PCR-mismatch amplification for the three polymorphisms, 5% of the samples were randomly selected (n = 50) and sequenced in a blinded manner.

The forward and reverse primers for PCR sequencing of the Leu84Phe site were 5'-GCTGCACAGCTAGTTGAGAC-3' and 5'-ACTGTGATGTCAGCGATCGT-3', which produce a 358 bp fragment. The forward and reverse primers for sequencing the Ile143Val and Lys178Arg sites were 5'-CCACGTCACCTCAGGACACT-3' and 5'-CAGGACACTGCCACTTCCTT-3', which produce a 559 bp fragment that includes both codons 143 and 178. Thus, we were able to observe linkage disequilibrium between SNPs at codons 143 and 178 for these samples. The PCR sequencing conditions were as described above, except that the annealing temperature was 54 and 56°C, respectively. Dye terminator DNA sequencing was performed in both directions using a BigDye Terminator Kit (Applied Biosystems, Foster City, CA) and an ABI PRISM 3100 Genetic Analyzer (Applied Biosystems). PCR products and dye-terminated products were purified by ethanol/sodium acetate precipitation. Comparison of the two methods of analyzing polymorphisms did not identify any discordant pairs.

HPV detection
HPV infection was identified using the Hybrid Capture II assay (Digene Diagnostics, Gaithersburg, MD) using probe B, which contains a pool of full-length RNA probes specific for HPV 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59 and 68. Cervical sampling for HPV DNA was performed with the Digene Cervical Sampler, and the tests were classified according to the relative light units/positive control ratio, measured as the relative light units of the specimen divided by the mean relative light units of two positive controls.

Statistical analysis
Differences in frequencies of the MGMT genotypes and alleles between the cases of cervical carcinoma or CIN and controls were evaluated by the chi-square ({chi}2) test. We also compared the observed genotype frequencies with those calculated by the Hardy–Weinberg equilibrium theory (p2 + 2pq + q2 = 1, where p is the frequency of the variant allele and q = 1 – p). Univariate and multivariate logistic regressions were applied to calculate the odds ratios (ORs) and 95% confidence intervals (CIs), respectively, for the association between the genotypes and risk of cervical carcinoma. We used multivariable logistic regression to adjust for potential confounding factors. Risk factors for cervical carcinoma and potential confounders were kept in multivariable models if their removal caused ß parameter estimates to change by >10%. The multivariable logistic regression was adjusted for age, number of sexual partners, age at first intercourse, parities (including full-term pregnancy and abortion at or after 28 weeks) and age at first full-term pregnancy. Adjustments for additional suspected confounders, including smoking, alcohol consumption, oral contraceptive use and first-degree family history of cervical carcinoma did not change the results and are therefore not presented here. Pairwise linkage disequilibrium (D' and r2) between sites 143 and 178 was calculated using SHEsis software (47). Haplotype frequencies were estimated from the genotype data using PHASE version 2.1.1 (http://www.stat.washington.edu/stephens/phase.html), which was based on a Bayesian algorithm and the maximum likelihood model linked with the Stochastic-EM algorithm (48,49). The association between the frequencies of MGMT haplotypes and cervical cancer risk was also evaluated by stratification analyses in subgroups of different lifestyles. All statistical analyses were carried out using Statistical Package for Social Science software, version 11.5 (SPSS, Chicago, IL).


   Results
Top
 Abstract
 Introduction
 Materials and methods
 Results
Discussion
 References
 
Of the 539 cancer patients, 473 (87.8%) had squamous cell carcinomas, 48 (8.9%) adenocarcinomas and 18 had other types of carcinoma, including 15 adenosquamous carcinomas, one undifferentiated carcinoma and two carcinosarcomas. Of the 539 carcinoma patients, 53 (9.8%) were stage Ia, 212 (39.3%) were stage Ib, 133 (24.7%) were stage IIa, 82 (15.2%) were stage IIb, 48 (9.0%) were stage III and 11 (2.0%) were stage IV. Of the 473 CIN patients, 244 (51.6%) were classified as CIN III, 160 (33.8%) as CIN II and 69 (14.6%) as CIN I. The median ages of the cancer patients, CIN patients and control women were 42 (range, 25–81), 41 (range, 21–69) and 40 (range, 20–65) years, respectively. In the carcinoma, CIN and control groups, 280 (51.9%), 234 (49.5%) and 400 (50.0%) individuals, respectively, were <40 years old, and 259 (48.1%), 239 (50.5%) and 400 (50.0%) individuals were >40 years old. Thus, there were no significant differences in ages between the three groups, and the ages of the individuals of the control group appeared to adequately match those of the patient groups. The sexual and reproductive histories, including age at first intercourse, number of sexual partners and age at first full-term pregnancy among the patients with carcinomas, patients with CIN and controls were not significantly different, but the parities (including full-term pregnancy and abortion at or after 28 weeks) was significantly different among the three groups ({chi}2 = 18.235, P < 0.001) (Table I). There were five (0.7%), two (0.4%) and three (0.4%) smokers in the carcinoma, CIN and control groups, respectively, which did not represent a significant difference. HPV infection rates in carcinoma and CIN patients were 90.4 and 85.8%, respectively, whereas in control women, the infection rate was only 31.0%. Thus, the infection rate in both patient groups was significantly higher than in the control group ({chi}2 = 366.310, P < 0.001).


View this table:
[in this window]
[in a new window]

  		Table I. Frequency distribution of select characteristics by case–control status

 
The frequencies of the genotypes in the carcinoma patients, CIN patients and controls were consistent with those predicted by the Hardy–Weinberg equilibrium model. The genotype distributions and the allele frequencies of all three SNPs in the carcinoma, CIN and control groups were not significantly different. Ninety-seven percent of the subjects had the same genotypes for codons 143 and 178, suggesting that the two variant alleles are in strong linkage disequilibrium (pairwise D' = 1.00; r2 = 0.88), which is consistent with the data (D' = 1.00; r2 = 1.00) in the HapMap for the CHB population (Han Chinese in Beijing, China) http://www.hapmap.org/cgi-perl/gbrowse/hapmap_B35/). Figures 1–3 show the sequencing results of the three SNPs.


Figure 1
View larger version (29K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1. Sequencing results of MGMT Leu84Phe variants. (a) CC homozygote, (b) CT heterozygote and (c) TT homozygote.

 


Figure 2
View larger version (33K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2. Sequencing results of MGMT Ile143Val variants. (a) AA homozygote, (b) AG heterozygote and (c) GG homozygote.

 


Figure 3
View larger version (31K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 3. Sequencing results of MGMT Lys178Arg variants. (a) AA homozygote, (b) AG heterozygote and (c) GG homozygote.

 
Logistic regression analysis revealed that, compared with the MGMT Lys178Lys (AA) genotype, women homozygous for the Arg178Arg (GG) genotype had a 2.48-fold increase in the risk for cervical carcinoma (95% CI = 1.02–6.03), and a combination of the Arg178Arg (GG) and Lys178Arg (GA) genotypes increased the risk by 60% (OR = 1.63, 95% CI = 1.01–2.61). Those heterozygous for the Lys178Arg (GA) genotype, however, did not have an increased risk for cervical carcinoma. Similarly, compared with the wild-type Ile143Ile (AA) genotype, those homozygous for Val143Val (GG) had an elevated risk for carcinoma (OR = 2.44, 95% CI = 1.01–5.95), but those heterozygous for Ile143Val (GA) did not. Neither the Phe84Phe (TT) nor the Leu84Phe (CT) genotype elevated the risk for cervical carcinoma compared with the Leu84Leu (CC) genotype (Table II).


View this table:
[in this window]
[in a new window]

  		Table II. Association between MGMT polymorphisms and the risk of cervical carcinoma and CIN in all and HPV-positive subjects

 
Similar risks for the various genotypes were found for the development of squamous cell carcinoma, a subtype of cervical carcinoma. Specifically, there was a significantly increased risk for the development of squamous cell carcinoma for women carrying the Val143Val (GG) or Arg178Arg (GG) genotypes compared with women with a Ile143Ile (AA)/Lys178Lys (AA) genotype (OR = 2.56, 95% CI = 1.04–6.35 and 2.60, 95% CI = 1.03–6.46, respectively). Similarly, neither the Phe84Phe (TT) nor the Leu84Phe (CT) genotype increased the risk for squamous cell carcinoma. None of the SNPs were associated with a significant elevation in the risks for adenocarcinoma or other types of cancer, nor were they associated with a change in the risk for CIN I or CIN II–III.

We next investigated the association between the Ile143Val and Lys178Arg variant genotypes and the risk for cervical carcinoma, stratified by HR-HPV infection status. In the HR-HPV-positive group, compared with individuals carrying Ile143Ile (AA)/Lys178Lys (AA), the risk for cervical cancer was elevated in those homozygous for Val143Val (GG) or Arg178Arg (GG) (OR = 3.33, 95% CI = 1.15–9.68 and 3.38, 95% CI = 1.16–9.81, respectively), and combination of the Arg178Arg (GG) and Lys178Arg (GA) genotypes increased the risk for cervical cancer by 45% (OR = 1.45, 95% CI = 1.21–3.34). We did not calculate the ORs or 95% CIs for the HPV-negative women because they accounted for only a small fraction of those with carcinoma or CIN (10.6 and 14.2%, respectively).

To examine the role of environmental factors and MGMT polymorphism, we stratified the analysis of the effect of MGMT polymorphisms on the risk for cervical carcinoma according to various environmental factors, including HR-HPV infection status and sexual and reproductive histories. Because codons 143 and 178 are in strong linkage disequilibrium and because SNPs in codon 178 are more strongly associated with a risk of cervical carcinoma in all patients and in the HPV-positive group, the Arg178Arg (GG) and Lys178Arg (GA) genotypes were compared together as a single group with the Lys178Lys (AA) genotype. In the HR-HPV-negative subgroup, the combined Arg178Arg (GG) and Lys178Arg (GA) genotypes did not elevate the risk for cervical carcinoma, whereas in the HR-HPV-positive subgroup, the GG or GA genotype significantly increased the risk for cervical carcinoma (adjusted OR = 1.45, 95% CI = 1.21–3.34). Similar results were also observed for the subgroup of women with an early onset of sexual activity (≤20 years), high parity (>3) and early age of first full-term pregnancy (≤22 years) [adjusted OR = 2.58 (95% CI = 1.23–4.41), 2.52 (95% CI = 1.75–3.45) and 3.22 (95% CI = 1.20–4.66)]. The GG or GA genotype did not increase the risk in either the multiple sexual partners subgroup or in the non-multiple sexual partners subgroup. As shown in Table III, similar results were obtained when the risk for squamous cell carcinoma was examined.


View this table:
[in this window]
[in a new window]

  		Table III. Association between the MGMT Lys178Arg genotype and the risk for cervical carcinoma stratified by various environmental factors

 
The frequencies of the five most common haplotypes were significantly different between the patient and control groups (Table IV). Using the most common hoplotype 84Leu (C)-143Ile (A)-178Lys (A) as a reference, women carrying the 84Phe (T)-143Val (G)-178Arg (G) haplotype had a significantly increased risk for cervical carcinoma both in the overall carcinoma and HPV-positive groups Table IV (adjusted OR = 1.87, 95% CI = 1.07–3.27 and 3.09, 95% CI = 1.05–9.11, respectively). The risk for cervical carcinoma was not elevated for women with other haplotypes, including 84Leu (C)-143Val (G)-178Arg (G), 84Leu (C)-143Val (G)-178Lys (A) and 84Phe (T)-143Ile (A)-178Lys (A). The other three possible haplotypes, 84Leu (C)-143Ile (A)-178Arg (G), 84Phe (T)-143Ile (A)-178Arg (G) and 84Phe (T)-143Val (G)-178Lys (A), were predicted to be absent in humans, suggesting that these three SNPs are in linkage disequilibrium; in fact, 97.0% of codons 143 and 178 had the same genotypes. Finally, none of the haplotypes were associated with a change in the risk for CIN I or CIN II–III.


View this table:
[in this window]
[in a new window]

  		Table IV. Haplotypes and the risk of cervical carcinoma and CIN in all and HPV-positive subjects

 
The women with different MGMT haplotypes and different sexual and reproductive histories had significantly different risks for cervical carcinoma (Table V). In the HR-HPV-negative subgroup, the 84Phe (T)-143Val (G)-178Arg (G) haplotype was not associated with an elevated risk of cervical carcinoma, whereas in the HR-HPV-positive subgroup, the 84Phe (T)-143Val (G)-178Arg (G) haplotype was associated with a significantly increased risk for cervical carcinoma using the 84Leu (C)-143Ile (A)-178Lys (A) haplotype as reference (adjusted OR = 3.09, 95% CI = 1.05–9.11). In the non-multiple sexual partners subgroup, the 84Phe (T)-143Val (G)-178Arg (G) haplotype did not elevate the risk of cervical carcinoma, whereas in the multiple sexual partners subgroup, the 84Phe (T)-143Val (G)-178Arg (G) haplotype increased the risk for cervical carcinoma using 84Leu (C)-143Ile (A)-178Lys (A) as a reference, although the difference did not reach statistical significance (adjusted OR = 3.23, 95% CI = 0.98–10.66; P = 0.055). In addition, the effect of haplotype 84Phe (T)-143Val (G)-178Arg (G) was more evident in women who had an early onset of sexual activity (≤20 years), high parity (>3) and young age at first full-term pregnancy (≤22 years) (adjusted OR = 3.11, 95% CI = 1.07–9.07; 2.19, 95% CI = 1.03–4.64 and 4.15, 95% CI = 1.13–15.32, respectively). Similar results were also observed for squamous cell carcinoma (adjusted OR = 3.00, 95% CI = 1.01–8.90; 2.36, 95% CI = 1.11–5.00 and 4.04, 95% CI = 1.09–14.89, respectively). In addition, among women who had more than one sexual partner, the risk for squamous cell carcinoma was significantly higher in women with the 84Phe (T)-143Val (G)-178Arg (G) haplotype than in those with the 84Leu (C)-143Ile (A)-178Lys (A) haplotype (OR = 3.37, 95% CI = 1.02–11.12).


View this table:
[in this window]
[in a new window]

  		Table V. Association between MGMT haplotypes and the risk for cervical carcinoma stratified by various environmental factors

 

   Discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
References
 
There are three common non-synonymous SNPs in MGMT: Leu84Phe, Ile143Val and Lys178Arg. Although the biological significance of the genetic polymorphisms in MGMT remains unclear, these amino acid substitutions may change the function of the protein and, therefore, cause diseases such as cancer. Ile143Val, which is linked with Lys178Arg, may modulate the protein's function because residue 143 is close to the conserved alkyl group acceptor, codon 145, in the active site of MGMT (42). Because of their potential biological significance, the effects of polymorphisms in MGMT have been evaluated in epidemiological studies. Several studies have explored the effect of MGMT polymorphisms on risks for lung cancer, melanoma and glioblastoma, but the results have been inconsistent (23,37,39,40,43,44,50). A study in a Chinese population found that the frequency of the MGMT Leu84Phe polymorphism did not significantly differ between lung cancer patients and controls (51). Another study of American Caucasians showed that polymorphisms in MGMT codons 143 and 178 were not associated with an increased risk for lung cancer (43,52). Studies examining the same Caucasian population, however, found that the MGMT Ile143Val polymorphism was associated with an increased risk for lung cancer (39,40) and that there is an increased risk of adenocarcinoma or small cell lung carcinoma in carriers of the variant alleles Lys178Arg and Ile143Val because the MGMT Lys178Arg polymorphism is associated with increased levels of polycyclic aromatic hydrocarbon–DNA adducts (39). Also, a recent study found that the Leu84Phe, Ile143Val and Lys178Arg polymorphisms in MGMT and their haplotypes or diplotypes were not independently associated with an increased risk of breast cancer but that polymorphisms in MGMT modulated the previously observed inverse association between antioxidant, fruit and vegetable consumption and the risk for breast cancer (45). Furthermore, individuals that consume low amount of fruits and green vegetables and carry the MGMT Ile143Val allele have been found to be at increased risk for lung and gastric cancer (39,53).

To our knowledge, the association between MGMT gene polymorphisms and cervical carcinoma has not been examined previously. Here, we found that the MGMT Lys178Arg (GG) and Ile143Val (GG) genotypes are associated with an elevated risk for cervical carcinoma. We also found that the Leu84Phe variant genotype does not alter the risk for cervical carcinoma. Thus, our results suggest that the Lys178Arg and Ile143Val variant genotypes in MGMT are genetic risk cofactors for cervical carcinoma.

In addition, we found that the association of Lys178Arg (GG) and Ile143Val (GG) with an elevated risk for cervical carcinoma was specific to squamous cell carcinomas and not relevant for adenocarcinoma. It is possible that different histopathological types may have different biological mechanisms and etiologic cofactors, and the development of squamous cell carcinoma and adenocarcinoma might be influenced by distinct environmental and genetic risk factors (54,55).

We noticed that single-polymorphic alleles have limited effects on the susceptibility to cancer but that the combination of variants may be important for genes with low penetration. A recent study found that Lys178Arg and Ile143Val alone had non-significant main effects on the risk for bladder cancer but that individuals with a combination of variant genotypes or haplotypes were at a significantly increased risk for bladder cancer (56). Similarly, we showed in the current study that women carrying the 84Phe (T)-143Val (G)-178Arg (G) haplotype had a remarkably increased risk for cervical carcinoma compared with women with only a single-variant allele. Therefore, multiple polymorphisms appear to more significant risk factors than single polymorphisms for the development of cervical carcinoma.

We found that women carrying the combined Arg178Arg (GG) and Lys178Arg (GA) genotypes are at a remarkably increased risk for cervical carcinoma both in the overall carcinoma group and in the HR-HPV-positive subgroup. Our results suggest that in women infected with HR-HPV, those carrying codon 178 variants GG or GA are at increased risk for the development of cervical cancer. In addition, high-risk sexual reproductive histories, including multiple sexual partners, early onset of sexual activity, young age at first full-term pregnancy and high parity are associated with a significant increase in the risk for cervical carcinoma and that the risk is even higher for women carrying codon 178 variants GG or GA. We also found that women who carried haplotype 84Phe (T)-143Val (G)-178Arg (G) and had high-risk sexual and reproductive histories had a significantly elevated risk for cervical carcinoma or squamous cell carcinoma compared with those without high-risk sexual and reproductive factors.

Epidemiological studies have shown that over 90% of cervical cancers can be attributed to specific HR-HPV types. Of the proteins encoded by HR-HPV, E6 is the most relevant for the transforming function because it binds to the tumor suppressor p53 and directly increases genomic instability (8,9). High-risk sexual or reproductive history factors probably contribute to carcinogenesis by increasing the risk for infection by HR-HPV, although destruction of the local microenvironment in the cervix could increase the level of mutation and genomic instability.

MGMT may have a fundamental role in cell physiology and genome maintenance. Furthermore, basal MGMT activity inversely correlates with wild-type p53 expression (32,33). These findings suggest that MGMT modulates the susceptibility to cervical carcinoma directly or through the p53 pathway. In vitro studies, however, have been unable to demonstrate that the codon 178/143 variant causes a significant change in enzyme activity. Therefore, additional studies are needed to understand the effect of the polymorphisms on MGMT and its enzyme activity.

Although cervical squamous cell carcinoma is known to develop from precursor CIN lesions through a multi-step process, we failed to observe an association between the three SNPs in MGMT and an increased risk for CIN. This could be due to the fact that most CINs, even CIN III, spontaneously resolve without developing to carcinoma. Those that do not develop to carcinoma could bias and counteract any differences between patients with CIN and controls.

The overall frequencies of the three polymorphisms in Asian populations were similar, suggesting that the sampling in our study was not biased. For instance, the frequencies of the CC and CT genotypes of the Leu84Phe allele among our 800 Chinese controls were 74.0 and 24.8%, respectively, which is similar to the frequencies reported from a study of 60 Chinese individuals by Deng et al. (80 and 20%, respectively) (51). Although the response rate of consent to be tested for HPV in the carcinoma and control groups was slightly different (48.2 and 51.3%, respectively), there should not have been a selection bias because candidates were lost at random.

In summary, our findings suggest that two polymorphisms (143GG and 178GG) in MGMT are associated with an elevated risk for cervical carcinoma. In addition, women who carry a 84Phe (T)-143Val (G)-178Arg (G) haplotype are at significantly increased risk for cervical carcinoma or cervical squamous cell carcinoma, especially those with HR-HPV infection, multiple sexual partners, early onset of sexual activity, high parity and a young age at first full-term pregnancy. Our results support the hypothesis that genetic variations in DNA repair genes contribute to the risk for cervical carcinoma.


   Acknowledgments
 
The authors are grateful to Ms Yaqing Chen (Cancer Hospital of Zhejiang Province), Mr Qian Zhang (First Affiliated Hospital of Wenzhou Medical College), Mr Linzhi Yan (Second Affiliated Hospital of Wenzhou Medical College) and Ms Danli Wang (Women's Hospital, School of Medicine at Zhejiang University) for their assistance in recruiting subjects. In addition, the authors are grateful to Ms Qi Chen (Women's Hospital, School of Medicine at Zhejiang University) for excellent technical support.

Grant support: This work was supported in part by the Science and Technology Project Research Program of Zhejiang Province Grants, China. 2005C23003 (to Huaizeng Chen).

Conflict of Interest Statement: None declared.


   References
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 

  1. Walboomers JM, et al. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J. Pathol. (1999) 189:12–19.[CrossRef][ISI][Medline]
  2. Schiffman MH, et al. The epidemiology of cervical carcinogenesis. Cancer (1995) 76:1888–1901.[CrossRef][ISI][Medline]
  3. zur Hausen H. Papillomaviruses causing cancer: evasion from host-cell control in early events in carcinogenesis. J. Natl Cancer Inst. (2000) 92:690–698.[Abstract/Free Full Text]
  4. Scheffner M, et al. The E6 oncoprotein encoded by human papillomavirus types 16 and 18 promotes the degradation of p53. Cell (1990) 63:1129–1136.[CrossRef][ISI][Medline]
  5. Dyson N, et al. The human papilloma virus-16 E7 oncoprotein is able to bind to the retinoblastoma gene product. Science (1989) 243:934–937.[Abstract/Free Full Text]
  6. Thomas M, et al. The role of the E6-p53 interaction in the molecular pathogenesis of HPV. Oncogene (1999) 18:7690–7700.[CrossRef][ISI][Medline]
  7. Ried T, et al. Genomic changes defining the genesis, progression, and malignancy potential in solid human tumors: a phenotype/genotype correlation. Genes Chromosomes Cancer (1999) 25:195–204.[CrossRef][ISI][Medline]
  8. Storey A, et al. Role of a p53 polymorphism in the development of human papillomavirus-associated cancer. Nature (1998) 393:229–234.[CrossRef][Medline]
  9. Kisseljov FL. Virus-associated human tumors: cervical carcinomas and papilloma viruses. Biochemistry (Mosc) (2000) 65:68–77.[Medline]
  10. Villa LL. Human papillomaviruses and cervical cancer. Adv. Cancer Res. (1997) 71:321–341.[ISI][Medline]
  11. Munoz N, et al. Role of parity and human papillomavirus in cervical cancer: the IARC multicentric case-control study. Lancet (2002) 359:1093–1101.[CrossRef][ISI][Medline]
  12. Moreno V, et al. Effect of oral contraceptives on risk of cervical cancer in women with human papillomavirus infection: the IARC multicentric case-control study. Lancet (2002) 359:1085–1092.[CrossRef][ISI][Medline]
  13. Skegg DC. Oral contraceptives, parity, and cervical cancer. Lancet (2002) 359:1080–1081.[CrossRef][ISI][Medline]
  14. Danaei G, et al. Causes of cancer in the world: comparative risk assessment of nine behavioural and environmental risk factors. Lancet (2005) 366:1784–1793.[CrossRef][ISI][Medline]
  15. Vessey M, et al. Mortality in relation to oral contraceptive use and cigarette smoking. Lancet (2003) 362:185–191.[CrossRef][ISI][Medline]
  16. La Vecchia C, et al. Sexual factors, venereal diseases, and the risk of intraepithelial and invasive cervical neoplasia. Cancer (1986) 58:935–941.[CrossRef][ISI][Medline]
  17. Brinton LA, et al. Sexual and reproductive risk factors for invasive squamous cell cervical cancer. J. Natl Cancer Inst. (1987) 79:23–30.[ISI][Medline]
  18. Apter D, et al. Gonadotropin-releasing hormone pulse generator activity during pubertal transition in girls: pulsatile and diurnal patterns of circulating gonadotropins. J. Clin. Endocrinol. Metab. (1993) 76:940–949.[Abstract]
  19. Vaccarella S, et al. Reproductive factors, oral contraceptive use, and human papillomavirus infection: pooled analysis of the IARC HPV prevalence surveys. Cancer Epidemiol. Biomarkers Prev. (2006) 15:2148–2153.[Abstract/Free Full Text]
  20. Parazzini F, et al. Determinants of risk of invasive cervical cancer in young women. Br. J. Cancer (1998) 77:838–841.[ISI][Medline]
  21. Settheetham-Ishida W, et al. Contribution of epigenetic risk factors but not p53 codon 72 polymorphism to the development of cervical cancer in northeastern Thailand. Cancer Lett. (2004) 210:205–211.[CrossRef][ISI][Medline]
  22. Wood RD, et al. Human DNA repair genes. Science (2001) 291:1284–1289.[Abstract/Free Full Text]
  23. Ford BN, et al. Identification of single nucleotide polymorphisms in human DNA repair genes. Carcinogenesis (2000) 21:1977–1981.[Abstract/Free Full Text]
  24. Margison GP, et al. Variability and regulation of O6-alkylguanine-DNA alkyltransferase. Carcinogenesis (2003) 24:625–635.[Abstract/Free Full Text]
  25. Mohrenweiser HW, et al. Challenges and complexities in estimating both the functional impact and the disease risk associated with the extensive genetic variation in human DNA repair genes. Mutat. Res. (2003) 526:93–125.[ISI][Medline]
  26. Margison GP, et al. O6-alkylguanine-DNA alkyltransferase: role in carcinogenesis and chemotherapy. Bioessays (2002) 24:255–266.[CrossRef][ISI][Medline]
  27. Sedgwick B, et al. Recent progress on the Ada response for inducible repair of DNA alkylation damage. Oncogene (2002) 21:8886–8894.[CrossRef][ISI][Medline]
  28. Bronstein SM, et al. Efficient repair of O6-ethylguanine, but not O4-ethylthymine or O2-ethylthymine, is dependent upon O6-alkylguanine-DNA alkyltransferase and nucleotide excision repair activities in human cells. Cancer Res. (1992) 52:2008–2011.[Abstract/Free Full Text]
  29. Rafferty JA, et al. Induction of murine O6-alkylguanine-DNA-alkyltransferase in response to ionising radiation is p53 gene dose dependent. Oncogene (1996) 12:693–697.[ISI][Medline]
  30. Grombacher T, et al. p53 is involved in regulation of the DNA repair gene O6-methylguanine-DNA methyltransferase (MGMT) by DNA damaging agents. Oncogene (1998) 17:845–851.[CrossRef][ISI][Medline]
  31. Harris LC, et al. Wild-type p53 suppresses transcription of the human O6-methylguanine-DNA methyltransferase gene. Cancer Res. (1996) 56:2029–2032.[Abstract/Free Full Text]
  32. Rolhion C, et al. O(6)-methylguanine-DNA methyltransferase gene (MGMT) expression in human glioblastomas in relation to patient characteristics and p53 accumulation. Int. J. Cancer (1999) 84:416–420.[CrossRef][ISI][Medline]
  33. Hengstler JG, et al. Activity of O(6)-methylguanine-DNA methyltransferase in relation to p53 status and therapeutic response in ovarian cancer. Int. J. Cancer (1999) 84:388–395.[CrossRef][ISI][Medline]
  34. Rudiger HW, et al. Reduced O6-methylguanine repair in fibroblast cultures from patients with lung cancer. Cancer Res. (1989) 49:5623–5626.[Abstract/Free Full Text]
  35. Pegg AE. Repair of O(6)-alkylguanine by alkyltransferases. Mutat. Res. (2000) 462:83–100.[CrossRef][ISI][Medline]
  36. Esteller M, et al. Generating mutations but providing chemosensitivity: the role of O6-methylguanine DNA methyltransferase in human cancer. Oncogene (2004) 23:1–8.[CrossRef][ISI][Medline]
  37. Egyhazi S, et al. Novel O6-methylguanine-DNA methyltransferase SNPs: a frequency comparison of patients with familial melanoma and healthy individuals in Sweden. Hum. Mutat. (2002) 20:408–409.[Medline]
  38. Inoue R, et al. A genotype of the polymorphic DNA repair gene MGMT is associated with de novo glioblastoma. Neurol. Res. (2003) 25:875–879.[CrossRef][ISI][Medline]
  39. Cohet C, et al. Exon 5 polymorphisms in the O6-alkylguanine DNA alkyltransferase gene and lung cancer risk in non-smokers exposed to second-hand smoke. Cancer Epidemiol. Biomarkers Prev. (2004) 13:320–323.[Abstract/Free Full Text]
  40. Kaur TB, et al. Role of polymorphisms in codons 143 and 160 of the O6-alkylguanine DNA alkyltransferase gene in lung cancer risk. Cancer Epidemiol. Biomarkers Prev. (2000) 9:339–342.[Abstract/Free Full Text]
  41. Tranah GJ, et al. O(6)-methylguanine-DNA methyltransferase Leu84Phe and Ile143Val polymorphisms and risk of colorectal cancer in the Nurses' Health Study and Physicians' Health Study (United States). Cancer Causes Control (2006) 17:721–731.[CrossRef][ISI][Medline]
  42. Chueh LL, et al. Specific amino acid sequences required for O6-methylguanine-DNA methyltransferase activity: analyses of three residues at or near the methyl acceptor site. Carcinogenesis (1992) 13:837–843.[Abstract/Free Full Text]
  43. Yang M, et al. Lack of association between Caucasian lung cancer risk and O6-methylguanine-DNA methyltransferase-codon 178 genetic polymorphism. Lung Cancer (2004) 44:281–286.[CrossRef][ISI][Medline]
  44. Inoue R, et al. Characterization of human polymorphic DNA repair methyltransferase. Pharmacogenetics (2000) 10:59–66.[CrossRef][ISI][Medline]
  45. Shen J, et al. MGMT genotype modulates the associations between cigarette smoking, dietary antioxidants and breast cancer risk. Carcinogenesis (2005) 26:2131–2137.[Abstract/Free Full Text]
  46. Glaab WE, et al. A novel assay for allelic discrimination that combines the fluorogenic 5' nuclease polymerase chain reaction (TaqMan) and mismatch amplification mutation assay. Mutat. Res. (1999) 430:1–12.[ISI][Medline]
  47. Shi YY, et al. SHEsis, a powerful software platform for analyses of linkage disequilibrium, haplotype construction, and genetic association at polymorphism loci. Cell Res. (2005) 15:97–98.[CrossRef][ISI][Medline]
  48. Stephens M, et al. A comparison of bayesian methods for haplotype reconstruction from population genotype data. Am. J. Hum. Genet. (2003) 73:1162–1169.[CrossRef][ISI][Medline]
  49. Stephens M, et al. A new statistical method for haplotype reconstruction from population data. Am. J. Hum. Genet. (2001) 68:978–989.[CrossRef][ISI][Medline]
  50. Ma S, et al. O6-methylguanine-DNA-methyltransferase expression and gene polymorphisms in relation to chemotherapeutic response in metastatic melanoma. Br. J. Cancer (2003) 89:1517–1523.[CrossRef][ISI][Medline]
  51. Deng C, et al. Genetic polymorphism of human O6-alkylguanine-DNA alkyltransferase: identification of a missense variation in the active site region. Pharmacogenetics (1999) 9:81–87.[ISI][Medline]
  52. Krzesniak M, et al. Polymorphisms in TDG and MGMT genes - epidemiological and functional study in lung cancer patients from Poland. Ann. Hum. Genet. (2004) 68:300–312.[CrossRef][ISI][Medline]
  53. Huang WY, et al. Selected DNA repair polymorphisms and gastric cancer in Poland. Carcinogenesis (2005) 26:1354–1359.[Abstract/Free Full Text]
  54. Altekruse SF, et al. Comparison of human papillomavirus genotypes, sexual, and reproductive risk factors of cervical adenocarcinoma and squamous cell carcinoma: Northeastern United States. Am. J. Obstet. Gynecol. (2003) 188:657–663.[CrossRef][ISI][Medline]
  55. Lacey JV Jr, et al. Associations between smoking and adenocarcinomas and squamous cell carcinomas of the uterine cervix (United States). Cancer Causes Control (2001) 12:153–161.[CrossRef][ISI][Medline]
  56. Li C, et al. Exon 3 polymorphisms and haplotypes of O6-methylguanine-DNA methyltransferase and risk of bladder cancer in southern China: a case-control analysis. Cancer Lett. (2005) 227:49–57.[CrossRef][ISI][Medline]
Received July 19, 2006; revised December 6, 2006; accepted January 9, 2007.


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Br Med BullHome page
B. Verbeek, T. D. Southgate, D. E. Gilham, and G. P. Margison
O6-Methylguanine-DNA methyltransferase inactivation and chemotherapy
Br. Med. Bull., March 1, 2008; 85(1): 17 - 33.
[Abstract] [Full Text] [PDF]

Home page
 CarcinogenesisHome page
S. Ogino, A. Hazra, G. J. Tranah, G. J. Kirkner, T. Kawasaki, K. Nosho, M. Ohnishi, Y. Suemoto, J. A. Meyerhardt, D. J. Hunter, et al.
MGMT germline polymorphism is associated with somatic MGMT promoter methylation and gene silencing in colorectal cancer
Carcinogenesis, September 1, 2007; 28(9): 1985 - 1990.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrowOA All Versions of this Article:
28/6/1314    most recent
bgm003v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (2)
Google Scholar
Right arrow Articles by Huang, J.
Right arrow Articles by Xie, X.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Huang, J.
Right arrow Articles by Xie, X.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us