This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Johnson, M. M. Articles by Chen, C. Search for Related Content PubMed PubMed Citation Articles by Johnson, M. M. Articles by Chen, C.

Screening for Deleterious Nonsynonymous Single-Nucleotide Polymorphisms in Genes Involved in Steroid Hormone Metabolism and Response

¹ Program in Epidemiology, Fred Hutchinson Cancer Research Center and ² Department of Epidemiology, University of Washington, Seattle, Washington

Requests for reprints: Chu Chen, Program in Epidemiology, Fred Hutchinson Cancer Research Center, P.O. Box 19024, Mailstop M5-C800, 1100 Fairview Avenue North, Seattle, WA 98109-1024. Phone: 206-667-6644; Fax: 206-667-2537. E-mail: cchen{at}fhcrc.org

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
To facilitate selection of single-nucleotide polymorphisms (SNP) for molecular epidemiologic studies investigating the hormonal carcinogenesis hypothesis, we used two sequence homology–based tools [Sort Intolerant from Tolerant (SIFT) and Polymorphism Phenotype (PolyPhen)] to predict the potential impact a nonsynonymous SNP (nsSNP), which results in an amino acid substitution, may have on the activity of proteins encoded by genes involved in the steroid hormone metabolism and response pathway. We screened 137 variants. Of these, 28% were predicted by SIFT and PolyPhen as having a potentially damaging effect on protein function. Investigation into the association of these variant alleles with hormone-related cancers may prove to be fruitful.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Epidemiologic association studies focus a great amount of effort into identifying single-nucleotide polymorphisms (SNP) in genes that may have an association with disease risk. Often, the SNPs that have an association with disease are those that are known as nonsynonymous SNPs (nsSNP), which result in an amino acid substitution. These types of polymorphisms are believed to be more likely to cause a change in structure and as such compromise the function of a protein. The structure of a protein can change in various ways due to the biochemical differences of the amino acid variant (acidic, basic, or hydrophobic) and by the location of the variant in the protein sequence (by affecting tertiary or quaternary structure or the active site where substrate binds). Many molecular epidemiologic studies focus on studying SNPs found in coding regions in hopes of finding significant association between SNPs and disease susceptibility, but often find little or no association (1-3). Our ability to better select a nsSNP for an association study can be enhanced by first examining the potential impact an amino acid variant may have on the function of the encoded protein with the use of two innovative sequence homology–based programs, Sort Intolerant from Tolerant (SIFT; http://blocks.fhcrc.org/sift/SIFT.html) and Polymorphism Phenotype (PolyPhen; http://www.bork.embl-heidelberg.de/PolyPhen/).

SIFT uses sequence homology among related genes and domains across species to predict the impact of all 20 possible amino acids at a given position, allowing users to determine which nsSNPs would be of most interest to study by sorting variants by this prediction score (1, 2, 4-9). The SIFT algorithm has been shown to predict a phenotype for a nsSNP more accurately than previously used substitution scoring matrices, such as BLOSUM62, as these matrices do not incorporate information specific to the protein of interest (4, 10). Another advantage to using SIFT is the potential to analyze a larger number of nsSNPs than methods that are dependent on the availability of protein structure alone (4, 11).

The PolyPhen algorithm, like SIFT, also takes an evolutionary approach in distinguishing deleterious nsSNPs from functionally neutral ones. PolyPhen differs from SIFT in that it predicts how damaging a particular variant may be by using a set of empirical rules based on sequence, phylogenetic, and structural information characterizing a particular variant. In addition to using sequence alignments, PolyPhen utilizes protein structure databases, such as PDB (Protein Data Bank) or PQS (Protein Quarternary Structure), DSSP (Dictionary of Secondary Structure in Proteins), and three-dimensional structure databases to determine if a variant may have an effect on the protein's secondary structure, interchain contacts, functional sites, and binding sites (3, 7-9).

These two algorithms have great potential in their use to screen for potentially damaging nsSNPs in genes that may be associated with disease risk, such as hormone-related cancers like breast, prostate, and endometrial cancer. Interindividual variation in the biosynthesis and metabolism of sex hormones has been shown in population studies and may be related to risk of these types of cancers. For instance, Dunning et al. (12) found that circulating levels of several steroid hormones, such as estrogen, androgen, and their precursors, are directly related to the risk of breast cancer. There is evidence that variation in circulating levels of steroid hormones may be associated with polymorphisms of genes involved in the steroid hormone metabolism and response pathways (13-15). In this study, we screened 137 nsSNPs identified in genes in these pathways, using the SIFT and PolyPhen algorithms, to predict which of these nsSNPs would most likely have a damaging effect on the function of the encoded proteins.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
nsSNP Screening Process
Two public SNP databases, dbSNP (http://www.ncbi.nlm.nih.gov/SNP/) and SNP500cancer (http://snp500cancer.nci.nih.gov/home.cfm), were used to identify nsSNPs for the 25 genes of interest. From dbSNP, we identified 113 nsSNPs. An additional 24 nsSNPs were identified from the SNP500cancer database.

Obtaining nsSNP Tolerance Scores with SIFT and PolyPhen
For each gene of interest, we analyzed the protein sequence identified from dbSNP using the SIFT-version 2 database. In general, for the peptide sequence, SIFT performs multiple alignments of a number of sequences until a median conservation for the peptide sequence is reached at the default of 3.0 and predicts whether substitution with any of the other amino acids is tolerated or deleterious for every position in the submitted sequence. The SIFT prediction is given as a tolerance index (TI) score ranging from 0.0 to 1.0, which is the normalized probability that the amino acid change is tolerated. A nsSNP with a TI score of 0.05 is considered to be deleterious. The confidence measure of this result is given as the median sequence conservation score (MSCS) for the given codon. Deleterious predictions with a MSCS 3.25 were made with low confidence because the sequences used to determine the score were not diverse enough.

We adjusted the PolyPhen algorithm settings to include all sequences homologous to the peptide sequence of interest for the calculation of the difference in structural parameters instead of using the single most homologous sequence as a default. All other settings of PolyPhen were set as the default. Predictions of how a particular nsSNP may affect protein structure by PolyPhen are assigned as "probably damaging," a score made with high confidence that the nsSNP should affect protein structure and/or function; "possibly damaging," where it may affect protein function and/or structure; and "benign," as most likely having no phenotypic effect (3). It arrives at these predictions by determining the position-specific independent count difference of the two allelic variants in the polymorphic position and also measures the degree of damaging effect a variant may have on structural parameters of a protein. The position-specific independent count is a logarithmic ratio of the likelihood of a given amino acid occurring at a particular position to the likelihood of the same amino acid occurring at any position in the sequence, also known as background frequency. Thus, PolyPhen, as mentioned before, uses a set of empirical rules based on sequence, phylogenetic, and structural information to determine if protein structure may be compromised.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
The screen for nsSNPs of the two online databases, dbSNP and SNP500cancer, yielded 137 nsSNPs in 25 genes associated with sex steroid biosynthesis, catabolism, and response. Of the 25 genes selected, only UGT1A1 lacked any submitted nsSNPs in the two databases. The results of the SIFT and PolyPhen analyses are listed in Table 1. SIFT scores 0.05 are considered to be deleterious (TI 0.05) and SIFT scores with low confidence (MSCS 3.25) are in parentheses. Of the 137 nsSNPs screened with SIFT and PolyPhen, the 111 nsSNPs that were scored with confidence by SIFT (MSCS < 3.25) were used in determining concordance between SIFT and PolyPhen.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Several studies have used SIFT and/or PolyPhen to assess function of nsSNPs in candidate genes for association studies, but none have reported such assessment of the steroid hormone metabolism pathway (2, 9). Based on an evaluation of published association of nsSNPs and cancer risk, Zhu et al. (1) concluded that the risk of a number of cancers were significantly inversely correlated with SIFT predictions. Their results also suggest that variants that occur in conserved sequences are more likely to be associated with cancer susceptibility. Two studies thus far have combined both the SIFT and PolyPhen algorithms to screen for deleterious nsSNPs. In one paper, Xi et al. (7) examined the DNA repair pathways and found that the predictions made by the two programs are highly correlated, with 62% concordance in the predictions of deleterious or potentially damaging nsSNPs. We observed a similar concordance of 73% between SIFT and PolyPhen predictions, but in genes involved in steroid hormone metabolism and response. In a second, more recent, study, Livingston et al. (8) used SIFT and PolyPhen to identify 57 potentially deleterious nsSNPs involved in DNA repair, cell cycle regulation, apoptosis, and drug metabolism.

Whereas useful, SIFT and PolyPhen are nonetheless somewhat limited. This is because both require available sequence data. However, as more whole genome sequences become publicly available, prediction of functionality of more nsSNPs will become possible and the results more reliable (5, 9). When there are too few sequences available on a particular gene for comparison, for example, the PGR gene, or when the sequences seem to be too homologous to one another, a nonfunctional variant may be predicted to be "intolerant" (6, 9). When the MSCS is >3.25, Savas et al. (2) interpreted the SIFT predictions as "possibly affecting" or "possibly tolerated." However, like Ng and Henikoff (4, 5), we chose to disregard nsSNPs that were scored with low confidence by SIFT until more sequence data become publicly available to allow more reliable predictions to be made. Another limitation of these algorithms is that the impacts of a combination of variants are not assessed and the dependence of functional impact of a variant on genotype of other genes or on exposure risk is not addressed (9). One final restriction of SIFT and PolyPhen is that the algorithms are unable to predict the impact of SNPs that occur outside of the coding region, such as promoter and enhancer regions, and splice sites that may affect protein levels or protein function.

To those conducting large-scale population-based epidemiologic studies, the idea of prioritizing nsSNPs in the investigation of association of SNPs with disease risk is of great interest. The use of SIFT and PolyPhen to select potentially intolerant nsSNPs for epidemiology studies can be an efficient way to explore the role of genetic variation in disease risk and to contain cost. Furthermore, predicted impact of these nsSNPs can be tested with the use of animal models and/or cell line systems to determine if functionality of the protein has indeed been altered.

	Footnotes

 
Grant support: National Cancer Institute, NIH, grant R01CA84141.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received 11/ 8/04; revised 12/28/04; accepted 1/ 6/05.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Zhu Y, Spitz MR, Amos CI, Lin J, Schabath MB, Wu X. An evolutionary perspective on single-nucleotide polymorphism screening in molecular cancer epidemiology. Cancer Res 2004;64:2251–7.[Abstract/Free Full Text]
2. Savas S, Kim DY, Ahmad MF, Shariff M, Ozcelik H. Identifying functional genetic variants in DNA repair pathway using protein conservation analysis. Cancer Epidemiol Biomarkers Prev 2004;13:801–7.[Abstract/Free Full Text]
3. Ramensky V, Bork P, Sunyaev S. Human non-synonymous SNPs: server and survey. Nucleic Acids Res 2002;30:3894–900.[Abstract/Free Full Text]
4. Ng PC, Henikoff S. Predicting deleterious amino acid substitutions. Genome Res 2001;11:863–74.[Abstract/Free Full Text]
5. Ng PC, Henikoff S. Accounting for human polymorphisms predicted to affect protein function. Genome Res 2002;12:436–46.[Abstract/Free Full Text]
6. Ng PC, Henikoff S. SIFT: predicting amino acid changes that affect protein function. Nucleic Acids Res 2003;31:3812–4.[Abstract/Free Full Text]
7. Xi T, Jones IM, Mohrenweiser HW. Many amino acid substitution variants identified in DNA repair genes during human population screenings are predicted to impact protein function. Genomics 2004;83:970–9.[CrossRef][Medline]
8. Livingston RJ, von Niederhausern A, Jegga AG, et al. Pattern of sequence variation across 213 environmental response genes. Genome Res 2004;14:1821–31. Epub 2004 Sep 13.[Abstract/Free Full Text]
9. Rebbeck TR, Spitz M, Wu X. Assessing the function of genetic variants in candidate gene association studies. Nat Rev Genet 2004;5:589–97.[Medline]
10. Henikoff S, Henikoff JG. Automated assembly of protein blocks for database searching. Nucleic Acids Res 1991;19:6565–72.[Abstract/Free Full Text]
11. Sunyaev S, Ramensky V, Bork P. Towards a structural basis of human non-synonymous single nucleotide polymorphisms. Trends Genet 2000;16:198–200.[CrossRef][Medline]
12. Dunning AM, Dowsett M, Healey CS, et al. Polymorphisms associated with circulating sex hormone levels in postmenopausal women. J Natl Cancer Inst 2004;96:936–45.[Abstract/Free Full Text]
13. Kristensen VN, Borresen-Dale AL. Molecular epidemiology of breast cancer: genetic variation in steroid hormone metabolism. Mutat Res 2000;462:323–33.[CrossRef][Medline]
14. Henderson BE, Feigelson HS. Hormonal carcinogenesis. Carcinogenesis 2000;21:427–33.[Abstract/Free Full Text]
15. Mitrunen K, Hirvonen A. Molecular epidemiology of sporadic breast cancer. The role of polymorphic genes involved in oestrogen biosynthesis and metabolism. Mutat Res 2003;544:9–41.[CrossRef][Medline]

This article has been cited by other articles:

				S. Savas, I. W. Taylor, J. L. Wrana, and H. Ozcelik Functional nonsynonymous single nucleotide polymorphisms from the TGF-{beta} protein interaction network Physiol Genomics, April 24, 2007; 29(2): 109 - 117. [Abstract] [Full Text] [PDF]

				S. H. Olson, E. V. Bandera, and I. Orlow Variants in Estrogen Biosynthesis Genes, Sex Steroid Hormone Levels, and Endometrial Cancer: A HuGE Review Am. J. Epidemiol., February 1, 2007; 165(3): 235 - 245. [Abstract] [Full Text] [PDF]

				A. G. Jegga, S. Gowrisankar, J. Chen, and B. J. Aronow PolyDoms: a whole genome database for the identification of non-synonymous coding SNPs with the potential to impact disease Nucleic Acids Res., January 12, 2007; 35(suppl_1): D700 - D706. [Abstract] [Full Text] [PDF]

				S. Damaraju, D. Murray, J. Dufour, D. Carandang, S. Myrehaug, G. Fallone, C. Field, R. Greiner, J. Hanson, C. E. Cass, et al. Association of DNA Repair and Steroid Metabolism Gene Polymorphisms with Clinical Late Toxicity in Patients Treated with Conformal Radiotherapy for Prostate Cancer Clin. Cancer Res., April 15, 2006; 12(8): 2545 - 2554. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Johnson, M. M. Articles by Chen, C. Search for Related Content PubMed PubMed Citation Articles by Johnson, M. M. Articles by Chen, C.