Cancer Research TCM Europe Jordan
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
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 PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zeigler-Johnson, C.
Right arrow Articles by Rebbeck, T. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zeigler-Johnson, C.
Right arrow Articles by Rebbeck, T. R.
[Cancer Research 64, 8461-8467, November 15, 2004]
© 2004 American Association for Cancer Research

Epidemiology and Prevention

CYP3A4, CYP3A5, and CYP3A43 Genotypes and Haplotypes in the Etiology and Severity of Prostate Cancer

Charnita Zeigler-Johnson1, Tara Friebel1, Amy H. Walker1, Yiting Wang1, Elaine Spangler1, Saarene Panossian1, Margerie Patacsil1, Richard Aplenc1,2, Alan J. Wein3, S. Bruce Malkowicz3 and Timothy R. Rebbeck1

1 Department of Biostatistics and Epidemiology, 2 Department of Pediatrics, and 3 Department of Urology, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania


   ABSTRACT
Top
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The CYP3A genes reside on chromosome 7q21 in a multigene cluster. The enzyme products of CYP3A4 and CYP3A43 are involved in testosterone metabolism. CYP3A4 and CYP3A5 have been associated previously with prostate cancer occurrence and severity. To comprehensively examine the effects of these genes on prostate cancer occurrence and severity, we studied 622 incident prostate cancer cases and 396 controls. Substantial and race-specific linkage disequilibrium was observed between CYP3A4 and CYP3A5 in both races but not between other pairs of loci. We found no association of CYP3A5 genotypes with prostate cancer or disease severity. CYP3A43*3 was associated with family history-positive prostate cancer (age- and race-adjusted odds ratio = 5.86, 95% confidence interval, 1.10–31.16). CYP3A4*1B was associated inversely with the probability of having prostate cancer in Caucasians (age-adjusted odds ratio = 0.54, 95% confidence interval, 0.32–0.94). We also observed significant interactions among these loci associated with prostate cancer occurrence and severity. There were statistically significant differences in haplotype frequencies involving these three genes in high-stage cases (P < 0.05) compared with controls. The observation that CYP3A4 and CYP3A43 were associated with prostate cancer, are not in linkage equilibrium, and are both involved in testosterone metabolism, suggest that both CYP3A4*1B and CYP3A43*3 may influence the probability of having prostate cancer and disease severity.


   INTRODUCTION
Top
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Genes involved in androgen metabolism have been implicated in the etiology of prostate cancer. Testosterone is a major determinant of prostate growth and differentiation. There are numerous lines of evidence that support the role of androgen metabolism in prostate cancer etiology. Circulating levels of androgens have been reported to be higher in populations at increased prostate cancer risk, including African American men (1) , and lower in populations at decreased prostate cancer risk, including Chinese men (2) . Although serum levels of testosterone do not correlate well with prostate cancer risk (3, 4) , serum levels of dihydrotestosterone and other testosterone metabolites do correlate with prostate cancer risk (1 , 3, 4) . Clinical evidence exists that androgens are related to the growth and development of prostate cancers, and androgen ablation in men with hormone-sensitive prostate tumors reduces tumor size and decreases the associated disease burden (5) . This evidence suggests that the metabolism of testosterone into the more biologically active forms of the hormone may be important in determining prostate cancer risk.

Testosterone bioavailabilty is determined by a number of enzymes, including CYP3A4 and CYP3A43. CYP3A4 is involved in the oxidation of testosterone to 2ß-, 6ß-, or 15ß-hydroxytestosterone (6) , which is less biologically active than testosterone or dihydrotestosterone. CYP3A43 also exhibits testosterone 6ß-hydroxylation in vitro and is predominantly expressed in the prostate (7) . Variants that affect CYP3A4 activity could therefore alter prostate tumor occurrence or aggressiveness. A variant in the 5'untranslated region of CYP3A4 (denoted CYP3A4*1B) has been associated with prostate cancer in three studies. We reported previously that Caucasian carriers of CYP3A4*1B had a higher Tumor-Node-Metastasis stage and Gleason grade than men who did not carry this variant. The effect on tumor stage was most pronounced in men diagnosed at a relatively old age who reported no family history of prostate cancer. Subsequently, Paris et al. (8) reported the same association with tumor grade and stage, also with stronger effects in older men. Tayeb et al. (9) reported that CYP3A4*1B was associated with prostate cancer in men with a history of benign prostatic hyperplasia. Plummer et al. (10) also reported strong effects of CYP3A4 and CYP3A5 genotypes with prostate cancer. Kittles et al. (11) reported that the association of prostate cancer with CYP3A4 genotypes in African Americans may be the result of population stratification. However, difficulties with the design of this study have raised questions whether population stratification or poor study design could have most strongly influenced this association (12) . In combination, these results are consistent with an effect of CYP3A genes on the natural history, and possibly prognosis, of prostate cancers. The observation that CYP3A4 genotype is associated with higher clinical stage and grade prostate tumors is consistent with the hypothesis that CYP3A4 may be associated with androgen-mediated increases in prostate cell proliferation or growth.

The CYP3A genes lie in a region of chromosome 7q21-q22 as part of a multigene family (13) , including CYP3A4 (14) , CYP3A5 (15) , CYP3A7, CYP3A43 (7 , 16, 17) in addition to psuedogenes (Fig. 1)Citation . Only CYP3A4, CYP3A5, CYP3A7, and CYP3A43 are expressed in adults (18) . These loci seem to be in linkage disequilibrium (LD; refs 15 and 19 ). Therefore, we evaluated genotypes and haplotypes of CYP3A4, CYP3A5, and CYP3A43 by ethnicity and tumor characteristics to better understand whether these genes are related to prostate cancer.



View larger version (6K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 1. Gene cluster on Chromosome 7q22. Variants of interest include the following: *1 on CYP3A5, missense mutation at intron 3; *1B on CYP3A4, missense mutation at the 5'untranslated region; *3 on CYP3A43, alanine to proline mutation that occurs at exon 10.

 

   MATERIALS AND METHODS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Study Subjects and Data Collection.
A sample of 622 incident prostate cancer cases was identified through Urologic Oncology Clinics at the University of Pennsylvania Health System between 1995 and 2002. We confirmed case status by medical records review using a standardized abstraction form. Men were excluded from this study if they reported having exposure to finasteride (Proscar) at the time of their prostate cancer diagnosis. Patients who were nonincident cases (i.e., those diagnosed >12 months before the date of study ascertainment), or had a prior diagnosis of cancer at any site except nonmelanoma skin cancer, were also excluded. The mean age of diagnosis was 62.9 years (SD = 8 years) with a range of 39 to 85 years.

The 396 controls studied here were men attending University of Pennsylvania Health System general medicine clinics. These clinics see a patient population that is demographically similar to those seen in the University of Pennsylvania Health System Urologic Oncology clinics. These men were ascertained concurrently with the prostate cancer cases (i.e., between 1995–2002). Controls were excluded from this study if they ever had an elevated prostate-specific antigen test based on age- or race-specific values (20) , if they had ever had an abnormal digital rectal examination, if they had a previous cancer diagnosis except nonmelanoma skin cancer, or if they reported having had exposure to finasteride (Proscar) at the time of study ascertainment. Analyses adjusted for age and race were also undertaken to account for residual variation because of these factors. The mean age of controls at the time of their clinic visit was 58.6 years (SD = 11.3 years) with a range of 22 to 92 years.

A standardized questionnaire and review of medical records were used to obtain risk factor, medical history, and prostate cancer diagnostic information. Information collected included prostate cancer occurrences in first- and second-degree relatives; personal history of benign prostatic hyperplasia and vasectomy; previous cancer diagnoses; demographic information such as race, educational level, and occupation; and prostate cancer screening history. All study subjects provided informed consent for participation in this research under a protocol approved by the Committee for Studies Involving Human Subjects at the University of Pennsylvania.

Biosample Collection and Genotype Analysis.
Genomic DNA for the present study was self-collected by each study subject using sterile cheek swabs (Cyto-Pak Cytosoft Brush, Medical Packaging Corporation, Camarillo, CA), and processed using either a protocol modified from Richards et al. (21) as described previously (22) or using a Qiagen 9604 robot with the QIAamp 96 DNA Buccal Swab Biorobot kit (Qiagen, Valencia, CA). The resulting biosamples were used for PCR-based genotype analyses.

The methods used to determine CYP3A4*1B genotypes have been reported previously by Rebbeck et al. (23) . The methods used to analyze CYP3A5*1 included a PCR reaction followed by genotype assessment using a Pyrosequencing PSQ 96 system. The primers used in this amplification included a forward primer for the "wild-type," CYP3A5*3 (C3A5*3.F1, ACCACCCAGCTTAACGAATG), a reverse primer (C3A5*3.R1, TGACACACAGCAAGAGTCTCA, which was biotinylated), and a sequencing primer (AGAGCTCTTTTGTCTTTCA). Preparation for the PCR reaction consisted of a buccal swab protocol and a Qiagen protocol. In the buccal swab protocol, we used 20 µL of Eppendorf Master Mix, 10 µL of double-distilled H20, 0.5 µL of each PCR primer, and 10 µL of template DNA. In the Qiagen protocol, we used 20 µL of Eppendorf Master Mix, 17 µL of double-distilled H20, 0.5 µL of each PCR primer, and 3 µL of template DNA. The temperature profile for the PCR reaction was one cycle at 94°C for 5 minutes, followed by 35 cycles of 94°C for 30 seconds, 59°C for 30 seconds, and 72°C for 30 seconds. This procedure was followed by one cycle at 72°C for 7 minutes. We achieved determination of genotypes using the resulting 120 bp PCR product with the Pyrosequencing PSQ 96 system with nucleotide dispensation order CGACTATC.

Analysis of CYP3A43*3 (Pro340Ala in exon 10) also included a PCR reaction followed by genotype assessment with a Pyrosequencing PSQ 96 System (24) . The primers used in this amplification included a forward primer (3A43-F, AGGGATTTGGGAGCTTCACT), a reverse primer (3A43-R, GAGGAGGCATTCTTGCTGAG, which was biotinylated), and a sequencing primer (AGGAGATTGACGCAGTTTTA). The PCR reaction consisted of a buccal swab protocol and a Qiagen protocol. In the buccal swab protocol, we used 23.5 µL of double-distilled H20, 0.5 µL of each PCR primer, 5 µL of 10x PCR buffer, 2 µL of 25 mmol/L MgCl2, 1 µL of 10 mmol/L deoxynucleoside triphosphates, 0.5 µL of Platinum Taq (Life Technologies, Inc., Grand Island, NY), and 8 µL of template DNA. In the Qiagen protocol, we used the same amount of the reagents, with two exceptions. We used 26.5 µL of double-distilled H20 and 5 µL of template DNA. The temperature profile for the PCR reaction was one cycle at 95°C for 5 minutes, followed by 20 cycles of 94°C for 30 seconds, 62°C for 30 seconds (–0.5°C/cycle), and 72°C for 30 seconds. This procedure was followed by 15 cycles of 94°C for 30 seconds, 52°C for 30 seconds, and 72°C for 30 seconds. Finally, there was one cycle at 72°C for 5 minutes. We achieved determination of genotypes using the resulting 317 bp PCR product using the Pyrosequencing PSQ 96 system with nucleotide dispensation order ACGCTATAG.

Statistical Methods.
We computed allele and genotype frequencies using gene counting methods, and evaluated Hardy-Weinberg equilibrium assuming random mating using the method of Levene (25) . In addition, we evaluated differences in allele and genotype frequencies across ethnic groups by undertaking Fisher’s exact tests. Pairwise composite disequilibrium coefficients {Delta}, rather than D', were estimated for the three alleles of interest by using the following: {Delta} = DAB + DA/B for alleles A and B at two loci (26) . D' was not computed because we are studying population genotypic data and do not have information regarding phase in double heterozygotes. On the basis of the methods of Weir (26) , both DAB and DA/B are constrained by the same maximum/minimum values, because gametic and nongametic frequencies cannot be negative or greater than corresponding allele frequencies. The maximum/minimum values of {Delta} are, in fact, twice those of DAB and DA/B. Accordingly, we defined a normalized measure of {Delta} called {Delta}', by dividing {Delta} by its maximum/minimum values. This measure is analogous to D', therefore, its usage should be subjected to the same caution as D' (26, 27, 28) . Similar to D[pime], {Delta}' will have a range of values from -1 to +1 and is not independent of allele frequencies.

Because our initial analyses suggested deviations from Hardy-Weinberg proportions for some loci, pairwise disequilibrium computations were made without the assumption of Hardy-Weinberg equilibrium. This method is appropriate for situations in which linkage phase is unknown (29) . Then, we inferred haplotype frequencies, and we imputed missing data using gene counting methods that assumed random mating. We undertook all computations using the GDA software version 1.1 (30) and STATA v.8.0.

For genotype-disease associations, we considered univariate and genotype effects and pairwise interactions involving CYP3A4, CYP3A5, and CYP3A43. For all genes, we combined putative risk alleles based on functional or allele frequency information into binary genotype classes. We also considered these loci together by comparing pairwise multiplicative interactions of all genotype combinations. We undertook genotype-disease associations using unconditional logistic regression, and we computed odds ratios (ORs) for the entire sample, as well as stratified by race (i.e., African American or Caucasian). We further evaluated whether there were differences in the association of genotypes in organ-confined tumors (i.e., stages T1 and T2) compared with tumors diagnosed with extracapsular extension or metastasis (i.e., stages T3 and T4). Finally, we attempted to identify associations in individuals with a family history of prostate cancer (i.e., at least one first or second degree relative with prostate cancer) or were younger than age 60 at diagnosis or time of ascertainment. All analyses were adjusted for age at the time of diagnosis in cases or time of study ascertainment in controls and race (i.e., African American or Caucasian), except for race-specific analyses, which were adjusted for age only. We undertook all statistical analyses for associations using SAS v. 8.01 and STATA v. 8.0. All P values were based on two-sided hypothesis tests.

We undertook haplotype analyses using the hapipf routine as implemented in STATA v.6.0. The frequencies of the observed haplotypes were calculated by race (African American and Caucasian) in controls, cases, low-stage cases, and high-stage cases. Using the likelihoods (L) obtained for the haplotype distribution in each case or control group, we then compared the haplotype distributions for controls versus all cases, low-stage cases, and high-stage cases separately by race using a {chi}2 test of the form 2 {log L(controls) + log L(cases) – log L(total sample) (31) . The distribution of this statistic was evaluated with a {chi}2 test with degrees of freedom equal to the number of haplotype classes considered.


   RESULTS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Allele and Genotype Frequency Distribution.
Table 1Citation presents the variant allele frequencies for CYP3A4*1B, CYP3A5*1, and CYP3A43*3 by ethnicity and case-control status. As has been reported previously (22) , we observed significant differences in the distribution of alleles at CYP3A4 by race, with higher frequencies of the CYP3A4*1B allele in African Americans compared with Caucasians for both cases and controls (P < 0.0001). In addition, we report significant differences in the frequency of alleles at CYP3A5 (P < 0.0001) and CYP3A43 (P < 0.0001), with CYP3A5*1 and CYP3A43*3 allele frequencies being significantly higher in African Americans than in Caucasians. Also, we observed significant deviations from Hardy-Weinberg equilibrium in Caucasians for CYP3A4.


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

  		Table 1 CYP3A4, CYP3A5, and CYP3A43 allele and genotype frequencies by race and case-control status

 
We observed significant LD between CYP3A4*1B and CYP3A5*1 in cases and controls, for African Americans and Caucasians (Table 2)Citation . Borderline significant disequilibrium between CYP3A4*1B and CYP3A43*3 was observed only in African-Americans. We did not observe statistically significant LD in any other group. The estimate of LD between CYP3A4*1B and CYP3A5*1 was higher in African Americans than Caucasians for both cases and controls. However, the magnitude of LD between these two loci was similar for cases and controls within each ethnic group.


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

  		Table 2 Linkage disequilibrium of CYP3A genes in African and European Americans: composite disequilibrium coefficients and exact significance level not assuming HWE

 
Case-Control Associations.
As shown in Table 3Citation , we observed no effect of CYP3A5 alone, and we observed no pattern that suggested a putative association in our data.


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

  		Table 3 Case-control associations of prostate cancer: main effects of CYP3A variants

 
CYP3A4 seemed to be associated (but not significantly) with an overall decreased odds of disease in all subgroups except African Americans. We observed a statistically significant association between CYP3A4 and prostate cancer for Caucasians only [OR = 0.54, 95% confidence interval (CI), 0.32–0.94]. Similar effects were noted for both low- and high-stage tumors, although the association in high-stage tumors did not reach statistical significance.

CYP3A43 showed no obvious pattern of direction across the subgroups. However, a positive association with prostate cancer was observed for CYP3A43*3 in men with a family history of prostate cancer (OR = 5.86, 95% CI, 1.10–31.16). Again, similar magnitudes of effect were observed for low- and high-stage tumors, but neither of these analyses reached statistical significance.

When we considered pairwise interactions of CYP3A4*1A/*1A-CYP3A5*3/*3 versus all other genotypes containing any CYP3A4*1B or CYP3A5*1 alleles (Table 4)Citation , the general effect was protective. We observed statistically significant associations with high-stage cases in the total sample (OR = 0.44, 95% CI, 0.22–0.87) and in the subgroup that was <60 years of age (OR = 0.21, 95% CI, 0.06–0.82). The presence of the CYP3A4*1B-CYP3A5*1 genotypes (i.e., both variants) was positively associated with prostate cancer in African Americans (OR = 2.24, 95% CI, 1.03–4.89). No other general pattern suggesting any other association was observed.


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

  		Table 4 Case-control associations of prostate cancer: pairwise interactions among CYP3A variants

 
We observed no strong interactions involving CYP3A5 and CYP3A43. However, the presence of the CYP3A4*1B and CYP3A43*3 was largely protective. There was an inverse association in men with at least one variant on each gene who happened to be <60 years of age (OR = 0.13, 95% CI, 0.04–0.46 for the total sample and OR = 0.07, 95% CI, 0.01–0.37 for low-stage cases).

Tables 5Citation and 6Citation present haplotype frequencies by case-control status for Caucasians (Table 5)Citation , and African Americans (Table 6)Citation . Eight haplotypes were estimated in our data. Although the same haplotypes were estimated to occur in both African Americans and Caucasians, the frequency distribution of these haplotypes was substantially different in the two ethnic groups. The only statistically significant association observed was between the haplotype distribution in high-stage cases compared with controls. By inspection, differences occur in the frequency of most haplotypes (particularly common haplotypes 3 and 5, and possibly rarer haplotypes 6, 7, and 8). Haplotype 1 was the most prevalent haplotype in Caucasian controls (frequency, 0.84), whereas haplotype 7 was the most common haplotype in African Americans (frequency, 0.28). Haplotypes that contained a CYP3A4*1B allele tended to occur at a lower frequency among cases compared with controls. This suggests that the primary association with disease may be in carriers of CYP3A4*1B overall. By inspection among haplotypes with a frequency >1%, the largest case-control differences were seen for haplotypes 3 and 7 in Caucasians (i.e., when CYP3A4*1B and CYP3A43*3 appear together). These results suggest that a biological interaction of CYP3A4*1B and CYP3A43*3 may exist in Caucasians, because they are not in LD. In African Americans, the greatest case-control differences were observed for haplotypes 3 and 4 (i.e., when CYP3A4*1B on a background of CYP3A5*3). Because CYP3A4*1B and CYP3A5*1 are in LD, it is not clear whether CYP3A4 or CYP3A5 or both genes are causatively associated with prostate cancer.


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

  		Table 5 Comparison of haplotype frequencies by case-control status in Caucasians

 

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

  		Table 6 Comparison of haplotype frequencies by case-control status in African Americans

 

   DISCUSSION
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
The present study confirms previous reports of an association between CYP3A4*1B and prostate cancer occurrence and severity (9, 10 , 23) . A novel aspect of this report is that CYP3A43 is also associated with prostate cancer, particularly in the context of family history-positive disease. CYP3A4 is expressed preferentially in the prostate and is involved in testosterone metabolism (7) . CYP3A43 is alternatively spliced (7 , 16) and can create mRNA hybrids with CYP3A4 (12) . This evidence for a biological interaction between CYP3A4 and CYP3A43, in addition to their overlapping substrate specificity for testosterone is of potential relevance to the observation made here of interactions between CYP3A4 and CYP3A43 in prostate cancer etiology and severity.

This report provides information about the allelic, genotypic, and haplotypic distributions at the CYP3A locus. Previous reports suggested that LD exists at this locus (15) . Our data also indicate that LD exists at the CYP3A locus but that this LD is negligible or nonexistent between some loci. The only consistent LD observed across case-control or racial groups was between CYP3A4*1B and CYP3A5*1. This observation is supported by similar findings of Plummer et al. (10) and Keuhl et al. (15) . Linkage disequilibrium between CYP3A4 and CYP3A5 suggests that associations at one locus could be the result of causative effects at the other locus. However, it is most likely that CYP3A4 or another variant in LD is causatively associated with prostate cancer, as CYP3A5 alone did not confer significant risk or protection in our analysis.

The role of CYP3A5 in the metabolism of hormones or other putative prostate carcinogens is not as well understood as that for CYP3A4. CYP3A5*1 is the only CYP3A5 allele to date that produces high levels of full-length CYP3A5 mRNA and expresses CYP3A5 (15) . The more common CYP3A5 polymorphism in Caucasians, CYP3A5*3, produces an aberrantly spliced mRNA with a premature stop codon. Therefore, there is ample reason to believe that the CYP3A5 alleles studied here could have a functionally meaningful effect on disease etiology. CYP3A5*1 has been inversely associated with prostate cancer (10) , which was not replicated here. However, we did observe nonsignificant inverse associations with CYP3A5*1 in family history-positive or early-onset high-stage cancers. Because of the relatively small numbers of observations in that analysis, it is possible that the present study was underpowered to detect associations with CYP3A5*1. Therefore, additional large studies of CYP3A5*1 should be undertaken to confirm the results of Plummer et al. (10) .

In addition to LD with CYP3A5, we observed deviations from Hardy-Weinberg equilibrium for CYP3A4 in Caucasian cases and controls. This information further supports the hypothesis that CYP3A4*1B may be associated with prostate cancer etiology; a number of authors have suggested deviations from Hardy-Weinberg proportions may arise if a deficiency of one allele is seen in cases and/or controls (32, 33, 34) . In our data, there was a statistically significant deficit of CYP3A4*1B alleles in cases and a statistically significant deficit of CYP3A4*1A alleles in controls. These data are consistent with the deviations from expected proportions that may arise if CYP3A4*1B was associated with disease risk.

We also report that CYP3A4, CYP3A5, and CYP3A43 allele frequencies and LD among these genes differ between African Americans and Caucasians. Specifically, there are substantially higher variant CYP3A allele frequencies in Africans or African Americans compared with other ethnicities (35) . Because there is also strong evidence that baseline prostate cancer risks also differ by ethnicity (36) , the conditions for population stratification (i.e., confounding by ethnicity) may be met. Therefore, all analyses were adjusted for major ethnic group (in addition to age) in which these differences exist. Kittles et al. (11) evaluated potential population stratification for an association of CYP3A4 and prostate cancer in an African American sample. Before correction, a strong association was observed with CYP3A4*1B in samples of both Caucasians and African ancestry. However, after correction with random, unlinked markers, the association disappeared in the populations of African ancestry. Our associations were primarily observed in Caucasians, in which population stratification is established not to confer significant bias (37, 38) .

In addition, we report a significant association of CYP3A4 and CYP3A43 with occurrence of prostate cancer. Although CYP3A5*1 had no effect on disease occurrence alone, CYP3A43 increased risk of disease in men with a family history of disease, and CYP3A4*1B had an overall protective effect. It is unclear why CYP3A43 is associated with prostate cancer when examined alone. The CYP3A43*3 allele rarely has been studied, and there is not enough basic science about the gene or this specific variant to suggest why there is a connection to family history of disease. However, one might speculate that this variant is more commonly inherited in men who have a family history of prostate cancer and may be a candidate hereditary gene for prostate cancer.

Our results regarding CYP3A4*1B are consistent with previous reports suggesting significant associations between the variant and the occurrence and severity of prostate cancer (8, 9 , 23) . In concordance with the present results, Plummer et al. (10) observed an inverse association between CYP3A4*1B and prostate cancer risk. Unfortunately, we had limited ability to obtain statistical significance in some subgroups because of small sample size.

In general, our results showing disease associations with CYP3A4 and CYP3A43 are consistent with knowledge of gene and allele function in these genes. CYP3A4 and CYP3A43 are involved in the metabolic deactivation (hydroxylation) of testosterone (6, 7) . CYP3A43 is preferentially expressed in the prostate (7) . However, the function of CYP3A4*1B has been controversial. In addition to epidemiologic evidence that CYP3A4*1B is associated with prostate cancer, the basic science literature has not consistently supported a functionally significant effect. A number of authors have studied the relationship of CYP3A4 expression or function of CYP3A4*1B (39, 40, 41, 42, 43, 44, 45) . Most of these authors concluded that no biologically meaningful effects existed given the small magnitude of effects that were observed. However, almost all studies have reported consistent elevations in expression associated with CYP3A4*1B in the range of 20–200% increase over the consensus CYP3A4*1A. Although it is possible that this magnitude of effects will not confer clinically meaningful effects on drug disposition, it is not clear whether this phenotypic perturbation is sufficient to alter metabolism of exposures (e.g., steroid hormones) that may confer disease risk over the lifetime of an individual. For example, a 20% greater metabolism of testosterone by CYP3A4*1B over the course of a man’s lifetime may be sufficient to increase prostate cancer risk and therefore explain epidemiologic associations between CYP3A4*1B and prostate cancer. If so, the hypothesized direction of the metabolic effect of CYP3A4*1B to increase CYP3A4 expression is consistent with the epidemiologic association reported here: CYP3A4 converts testosterone to 2ß-, 6ß-, or 15ß-hydroxytestosterone and therefore shunts testosterone away from the more biologically active form of dihydrotestosterone (6) . Genetic variants that are associated with increased CYP3A4 activity, such as CYP3A4*1B, would be expected to decrease prostate cancer risk if the effect of the polymorphism is to decrease bioavailability of dihydrotestosterone. Thus, our observation of an inverse association of CYP3A4*1B with prostate cancer is consistent with the expected direction of effect.

When CYP3A4 and CYP3A43 were considered in pairwise interactions to determine the effect of having genotypes with at least one CYP3A4*1B and at least one CYP3A43*3 allele, some effects became stronger, in particular for early-onset prostate cancer. This is an interesting finding, because it suggests an association that may have its greatest impact in African-Americans. Although only 4% of Caucasians carry this allelic combination, it is a more common haplotype observed in our African-American sample (35%; see Tables 5ACitation and 5BCitation ) Therefore, a subset of men who carry the CYP3A4*1B and CYP3A43*3 combination and are likely African-American are significantly less likely to have been diagnosed with prostate cancer before age 60. The reason for this association is not understood, because African-Americans, in general, are at high risk to be diagnosed with prostate cancer before the age of 60 years.

Because CYP3A4*1B and CYP3A43*3 are in linkage equilibrium in our sample, the joint effect of these two genes is unlikely to be explained by LD. Thus, our data suggest that CYP3A4 and CYP3A43 each exert an independent effect on prostate cancer risk that cannot be explained by LD at each locus.

Finally, we also report significant differences in haplotype frequency distributions by race and case-control status. The observation that the combined CYP3A4-CYP3A5-CYP3A43 haplotype may contain additional information about risk prediction beyond that of CYP3A4 genotypes alone suggests that other genes in this region may also be involved in the etiology of prostate cancer or that the consideration of haplotypes in this region provides improved statistical information for studies evaluating prostate cancer risk and progression.

In conclusion, our results confirm the association of CYP3A4*1B and prostate cancer occurrence and severity, suggest a role for CYP3A43*3 in prostate cancer etiology, and further elucidate the relationships of multiple genotypes and haplotypes at the CYP3A locus with prostate cancer etiology. Combined with information about the function of these genes, there is growing evidence that one or more of the genes in the CYP3A locus are involved in prostate cancer etiology.


   ACKNOWLEDGMENTS
 
The authors thank Drs. D. Goldmann, W. Greer, G. W. Crooks, D. A. Horowitz, D. Farhadi, M. D. Cirigliano, M. L. Rusk, V. Weil, S. J. Gluckman, C. Bridges, and C. Guerra for valuable assistance in ascertaining study subjects.


   FOOTNOTES
 
Grant support: Grants from the Public Health Service (R29-ES08031, R01-CA85074, and P50-CA105641 to T. R. Rebbeck) and the University of Pennsylvania Cancer Center.

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.

Requests for reprints: Charnita Zeigler-Johnson, Department of Biostatistics and Epidemiology, University of Pennsylvania School of Medicine, 908 Blockley Hall, 423 Guardian Drive, Philadelphia, PA 19104-6021. E-mail: czeigler{at}cceb.med.upenn.edu

Received 5/13/04. Revised 8/ 9/04. Accepted 9/ 2/04.


   REFERENCES
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Ross RK, Bernstein L, Lobo RA, et al 5-alpha-Reductase activity and risk of prostate cancer among Japanese and US white and black males. Lancet 1992;339:887-9.[CrossRef][Medline]
  2. Lookingbill DP, Demers LM, Wang C, Leung A, Rittmaster RS, Santen RJ Clinical and biochemical parameters of androgen action in normal healthy Caucasian versus Chinese subjects. J Clin Endo Metab 1991;72:1242-8.[Abstract/Free Full Text]
  3. Vatten LJ, Ursin G, Ross RK, et al Androgens in serum and the risk of prostate cancer: a nested case-control study from the Janus serum bank in Norway. Cancer Epidemiol Biomark Prev 1997;6:967-9.[Abstract]
  4. Gann PH, Hennekens CH, Ma J, Longcope C, Stampfer MJ Prospective study of sex hormone levels and risk of prostate cancer. J Natl Cancer Inst (Bethesda) 1996;88:1118-26.[Abstract/Free Full Text]
  5. Luke MC, Coffey DS The male sex assessory tissues. Structure, androgen action and physiology Knobil E Neill JD eds. . The physiology of reproduction second edition 1994p. 1435-87. New York Raven Press, Ltd.
  6. Waxman DJ, Attisano PF, Guengerich PF, Lapenson DP Human liver microsomal steroid metabolism: identification of the major microsomal steroid 6ß-hydroxylase cytochrome P450 enzyme. Arch Biochem Biophys 1988;263:242-436.
  7. Domanski TL, Finta C, Halpert JR, Zaphiropoulos PG cDNA cloning and initial characterization of CYP3A43, a novel human cytochrome P450. Mol Pharmacol 2001;59:386-92.[Abstract/Free Full Text]
  8. Paris PL, Kupelian PA, Hall JM, et al Association between a CYP3A4CYP3A4 genetic variant and clinical presentation in African-American prostate cancer patients. Cancer Epidemiol Biomark Prev 1999;8:901-5.[Abstract/Free Full Text]
  9. Tayeb MT, Clark C, Sharp L, et al CYP3A4CYP3A4 promoter variant is associated with prostate cancer risk in men with benign prostate hyperplasia. Oncol Rep 2002;9:653-5.[Medline]
  10. Plummer SJ, Conti DV, Paris PL, Curran AP, Casey G, Witte JS CYP3A4CYP3A4 and CYP3A5CYP3A5 genotypes, haplotypes, and risk of prostate cancer. Cancer Epidemiol Biomark Prev 2003;12:928-32.[Abstract/Free Full Text]
  11. Kittles RA, Chen W, Panguluri RK, et al CYP3A4-V and prostate cancer in African Americans: causal or confounding association because of population stratification?. Hum Genet 2002;110:553-60.[CrossRef][Medline]
  12. Cardon LR, Palmer LJ Population stratification and spurious allelic association. Lancet 2003;361:598-604.[CrossRef][Medline]
  13. Finta C, Zaphiropoulos PG The human cytochrome P450 3A locus. Gene evolution by capture of downstream exons. Gene 2000;260:13-23.[CrossRef][Medline]
  14. Hashimoto H, Toide K, Kitamura R, et al Gene structure of CYP3A4, an adult-specific form of cytochrome P450 in human livers, and its transcriptional control. Eur J Biochem 1993;218:585-95.[Medline]
  15. Kuehl P, Zhang J, Lin Y, et al Sequence diversity in CYP3A promoters and characterization of the genetic basis of polymorphic CYP3A5 expression. Nat Genet 2001;27:383-91.[CrossRef][Medline]
  16. Gellner K, Eiselt R, Hustert E, et al Genomic organization of the human CYP3A locus: identification of a new, inducible CYP3A gene. Pharmacogenetics 2001;11:111-21.[CrossRef][Medline]
  17. Westlind A, Malmebo S, Johansson I, et al Cloning and tissue distribution of a novel human cytochrome p450 of the CYP3A subfamily, CYP3A43. Biochem Biophys Res Commun 2001;281:1349-55.[CrossRef][Medline]
  18. Koch I, Weil R, Wolbold R, et al Interindividual variability and tissue-specificity in the expression of cytochrome P450 3A mRNA. Drug Metab Dispos 2002;30:1108-14.[Abstract/Free Full Text]
  19. Wojnowski L, Hustert E, Klein K, et al Re: modification of clinical presentation of prostate tumors by a novel genetic variant in CYP3A4CYP3A4. J Natl Cancer Inst (Bethesda) 2002;94:630-1.[Free Full Text]
  20. Cooney KA, Strawderman MS, Wojno KJ, et al Age-specific distribution of serum prostate-specific antigen in a community-based study of African-American men. Urology 2001;57:91-6.[CrossRef][Medline]
  21. Richards B, Skoletsky J, Shuber AP, et al Multiplex PCR amplification from the CFTR gene using DNA prepared from buccal brushes/swabs. Hum Mol Genet 1993;2:159-163.[Abstract/Free Full Text]
  22. Walker AH, Najarian D, White DL, Jaffe JF, Kanetsky PA, Rebbeck TR Collection of genomic DNA by buccal swabs for polymerase chain reaction-based biomarker assays. Environ Health Perspect 1999;107:517-20.[Medline]
  23. Rebbeck TR, Jaffe JM, Walker AH, Wein AJ, Malkowicz SB Modification of clinical presentation of prostate tumors by a novel genetic variant in CYP3A4. J Natl Cancer Inst (Bethesda) 1998;90:1225-9.[Abstract/Free Full Text]
  24. Cauffiez C, Lo-Guidice JM, Chevalier D, et al First report of a genetic polymorphism of the cytochrome P450 3A43 (CYP3A43) gene: identification of a loss-of-function variant. Hum Mutat 2004;23:101[CrossRef][Medline]
  25. Levene H On a matching problem in genetics. Ann Math Stat 1949;20:91-4.[CrossRef]
  26. Weir BS . Genetic data analysis II 1996 Sinauer Sunderland, MA
  27. Lewontin RC On measures of gametic disequilibrium. Genetics 1988;120:849-52.[Abstract/Free Full Text]
  28. Thomson G, Baur MP Third order linkage disequilibrium. Tissue Antigens 1984;24:250-5.[Medline]
  29. Schaid DJ Linkage disequilibrium testing when linkage phase is unknown. Genetics 2004;166:505-12.[Abstract/Free Full Text]
  30. Lewis PO, Zaykin D. GDA (genetic data analysis): computer program for the analysis of allelic data (version 1.1), 2002. Available from http://lewis.eeb.uconn.edu/lewishome/software.html.
  31. Xie X, Ott J Testing linkage disequilibrium between a disease gene and marker loci. Am J Hum Genet 1993;53:1107[Medline]
  32. Feder JN, Gnirke A, Thomas W, et al A novel MHC class I-like gene is mutated in patients with hereditary haemochromatosis. Nat Genet 1996;13:399-408.[CrossRef][Medline]
  33. Nielsen DM, Ehm MG, Weir BS Detecting marker-disease association by testing for Hardy-Weinberg disequilibrium at a marker locus. Am J Hum Genet 1998;63:1531-40.[CrossRef][Medline]
  34. Hoh J, Wille A, Ott J Trimming, weighting, and grouping SNPs in human case-control association studies. Genome Res 2001;11:2115-9.[Abstract/Free Full Text]
  35. Zeigler-Johnson CM, Walker AH, Mancke B, et al Ethnic differences in the frequency of prostate cancer susceptibility alleles at SRD5A2 and CYP3A4. Hum Hered 2002;54:13-21.[CrossRef][Medline]
  36. Hsing AW, Tsao L, Devesa SS International trends and patterns of prostate cancer incidence and mortality. Int J Cancer 2000;85:60-7.[CrossRef][Medline]
  37. Wacholder S, Rothman N, Caporaso N Population stratification in epidemiologic studies of common genetic variants and cancer: quantification of bias. J Natl Cancer Inst (Bethesda) 2000;92:1151-8.[Abstract/Free Full Text]
  38. Wang Y, Localio R, Rebbeck TR Evaluating bias due to population stratification in case-control association studies of admixed populations. Genet Epidemiol 2004;27:14-20.[CrossRef][Medline]
  39. Westlind A, Lofberg L, Tindberg N, Andersson TB, Ingelman-Sundberg M Interindividual differences in hepatic expression of CYP3A4: relationship to genetic polymorphism in the 5'-upstream regulatory region. Biochem Biophys Res Commun 1999;259:201-5.[CrossRef][Medline]
  40. Amirimani B, Walker AH, Weber BL, Rebbeck TR RESPONSE: re: modification of clinical presentation of prostate tumors by a novel genetic variant in CYP3A4. J Natl Cancer Inst (Bethesda) 1999;91:1588-90.[Free Full Text]
  41. Ando Y, Tateishi T, Sekido Y, Yamamoto T, Satoh T, Hasegawa Y Re: modification of clinical presentation of prostate tumors by a novel genetic variant in CYP3A4. J Natl Cancer Inst (Bethesda) 1999;91:1587-90.[Free Full Text]
  42. Spurdle AB, Goodwin B, Hodgson E, et al The CYP3A4*1B polymorphism has no functional significance and is not associated with risk of breast or ovarian cancer. Pharmacogenetics 2002;12:355-66.[CrossRef][Medline]
  43. Floyd MD, Gervasini G, Masica AL, et al Genotype-phenotype associations for common CYP3A4 and CYP3A5 variants in the basal and induced metabolism of midazolam in European- and African-American men and women. Pharmacogenetics 2003;13:595-606.[CrossRef][Medline]
  44. Amirimani B, Ning B, Deitz AC, Weber BL, Kadlubar F, Rebbeck TR Transcriptional activity effects of a CYP3A4 promoter variant. Environ Mol Mutagen 2003;42:299-305.[CrossRef][Medline]
  45. Hamzeiy H, Bombail V, Plant N, Gibson G, Goldfarb P Transcriptional regulation of cytochrome P4503A4 gene expression: effects of inherited mutations in the 5'-flanking region. Xenobiotica 2003;33:1085-95.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 Int J STD AIDSHome page
T W Mahungu, M A Johnson, A Owen, and D J Back
The impact of pharmacogenetics on HIV therapy
Int J STD AIDS, March 1, 2009; 20(3): 145 - 151.
[Abstract] [Full Text] [PDF]

Home page
 AJPHHome page
R. B. Warnecke, A. Oh, N. Breen, S. Gehlert, E. Paskett, K. L. Tucker, N. Lurie, T. Rebbeck, J. Goodwin, J. Flack, et al.
Approaching Health Disparities From a Population Perspective: The National Institutes of Health Centers for Population Health and Health Disparities
Am J Public Health, September 1, 2008; 98(9): 1608 - 1615.
[Abstract] [Full Text] [PDF]

Home page
 Drug Metab. Dispos.Home page
P. J. Roberts, K. D. Rollins, A. D. M. Kashuba, M. F. Paine, A. C. Nelsen, E. E. Williams, C. Moran, J. K. Lamba, E. G. Schuetz, and R. L. Hawke
The Influence of CYP3A5 Genotype on Dexamethasone Induction of CYP3A Activity in African Americans
Drug Metab. Dispos., August 1, 2008; 36(8): 1465 - 1469.
[Abstract] [Full Text] [PDF]

Home page
 Am J EpidemiolHome page
T. R. Rebbeck, A. B. Troxel, S. Norman, G. Bunin, A. DeMichele, R. Schinnar, J. A. Berlin, and B. L. Strom
Pharmacogenetic Modulation of Combined Hormone Replacement Therapy by Progesterone-Metabolism Genotypes in Postmenopausal Breast Cancer Risk
Am. J. Epidemiol., December 15, 2007; 166(12): 1392 - 1399.
[Abstract] [Full Text] [PDF]

Home page
 CarcinogenesisHome page
A.-M. Moilanen, J. Hakkola, M. H. Vaarala, S. Kauppila, P. Hirvikoski, J. T. Vuoristo, R. J. Edwards, and T. K. Paavonen
Characterization of androgen-regulated expression of CYP3A5 in human prostate
Carcinogenesis, May 1, 2007; 28(5): 916 - 921.
[Abstract] [Full Text] [PDF]

Home page
 JNCI J Natl Cancer InstHome page
T. R. Rebbeck, A. B. Troxel, Y. Wang, A. H. Walker, S. Panossian, S. Gallagher, E. G. Shatalova, R. Blanchard, G. Bunin, A. DeMichele, et al.
Estrogen sulfation genes, hormone replacement therapy, and endometrial cancer risk.
J Natl Cancer Inst, September 20, 2006; 98(18): 1311 - 1320.
[Abstract] [Full Text] [PDF]

Home page
 Cancer Epidemiol. Biomarkers Prev.Home page
J. Park, L. Chen, L. Ratnashinge, T. A. Sellers, J.-P. Tanner, J.-H. Lee, N. Dossett, N. Lang, F. F. Kadlubar, C. B. Ambrosone, et al.
Deletion Polymorphism of UDP-Glucuronosyltransferase 2B17 and Risk of Prostate Cancer in African American and Caucasian Men.
Cancer Epidemiol. Biomarkers Prev., August 1, 2006; 15(8): 1473 - 1478.
[Abstract] [Full Text] [PDF]

Home page
 J AndrolHome page
R. Obligacion, M. Murray, and I. Ramzan
Drug-Metabolizing Enzymes and Transporters: Expression in the Human Prostate and Roles in Prostate Drug Disposition
J Androl, March 1, 2006; 27(2): 138 - 150.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
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 PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zeigler-Johnson, C.
Right arrow Articles by Rebbeck, T. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zeigler-Johnson, C.
Right arrow Articles by Rebbeck, T. R.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online