Abstract
Epidemiologic and genetic studies support the considerable effect of heritable factors on prostate tumorigenesis, although to date, no unequivocal susceptibility gene has been identified. The extensive study of RNASEL in prostate cancer patients worldwide has yielded conflicting results. We reevaluated the role of the RNASEL 471delAAAG Ashkenazi founder mutation in 1,642 Ashkenazi patients with prostate, bladder, breast/ovarian, and colon cancers; Ashkenazi controls; and in non-Ashkenazi prostate cancer patients and controls. The entire RNASEL coding sequence was also screened using denaturing high-performance liquid chromatography and multiplex ligation–dependent probe amplification for possible sequence variations or copy number changes in a population of prostate cancer patients. The 471delAAAG mutation was detected in 2.4% of the Ashkenazi prostate cancer patients; in 1.9% of patients with bladder, breast/ovarian, and colon cancers; and in 2.0% of the Ashkenazi controls. Seven additional variants were detected in RNASEL, including a novel potentially pathogenic splice site mutation, IVS5+1delG, although none were associated with increased prostate cancer risk. Multiplex ligation–dependent probe amplification analysis showed two RNASEL gene copies in all 300 prostate cancer patients tested. We estimated that the RNASEL 471delAAAG founder mutation, which was detected in 2% of the Ashkenazi Jews, originated between the 2nd and 5th centuries A.D., compared with the less frequent (1%) BRCA1 185delAG founder mutation, which originated hundreds of years earlier. Taken together, our analysis does not support a role for the RNASEL 471delAAAG Ashkenazi mutation nor for the other alterations detected in RNASEL in prostate cancer risk in Jewish men. (Cancer Epidemiol Biomarkers Prev 2006;15(3):474–9)
Introduction
Prostate cancer is the most commonly diagnosed malignancy in men in the Western world, including Jewish men in Israel, and is one of the leading causes of cancer-related death (1). Substantial evidence from genome-wide linkage analyses in hereditary prostate cancer families and the study of germ line mutations in familial and sporadic prostate cancer patients supports the strong contribution of genetic components (2) to as much as 43% of all early onset disease (age ≤55 years) and 9% of all cases of prostate cancer (3).
HPC1 (MIM 601518) at 1q24-25 was the first prostate cancer susceptibility locus reported (4) and gene mapping placed RNASEL (MIM180435) in this critical region (5). RNASEL encodes the 2′,5′-oligoadenylate-synthetase-dependent RNase L protein (RNASEL), which mediates the antiviral and proapoptotic activities of the IFN-inducible 2-5A system (6, 7). Reported association between RNASEL mutations and prostate cancer predisposition (8-13), loss of the wild-type RNASEL allele in prostate tumor tissue from patients harboring RNASEL germ line mutations (8, 13), and the decreased enzymatic activity (8, 11) and apoptotic ability (14) of the 462Gln RNASEL variant supported a role for this gene in prostate cancer susceptibility.
However, despite these significant data, conflicting evidence regarding disease-associated RNASEL variants (15, 16) challenges the role of RNASEL as a prostate cancer susceptibility gene. Although a number of studies support the association between certain nonsense and missense RNASEL mutations (Glu265X, Arg462Gln, and Asp541Glu) and disease risk (8, 9, 11, 12, 17), others have suggested that these mutations have either an insignificant effect (9, 11, 18, 19) or even decrease the risk for prostate cancer (10, 12, 18).
Our group detected the 471delAAAG frameshift mutation in RNASEL, the first Ashkenazi founder null mutation in a known hereditary prostate cancer candidate gene (13), and suggested a nearly significant association between the RNASEL 471delAAAG mutation and increased risk of prostate cancer in Ashkenazi Israeli men; however, a subsequent North American study (20) did not show an association between this mutation and prostate cancer risk in Ashkenazi Jews. These inconsistencies were unexpected given the evidence that RNASEL 471delAAAG is an Ashkenazi founder mutation and given the allegedly common origin of Ashkenazi Jews. To reassess the possible association between this Ashkenazi founder mutation and tumorigenesis, we screened a cohort of prostate cancer patients and Ashkenazi patients with other common malignancies, including bladder, breast/ovarian, and colon cancers. Furthermore, we analyzed the complete RNASEL coding sequence in a population of Ashkenazi and non-Ashkenazi prostate cancer patients to further evaluate whether other sequence variations or copy number changes contribute to prostate cancer risk.
Materials and Methods
Study Population
A total of 1,763 DNA samples, from 1,478 Ashkenazi and 285 non-Ashkenazi Jewish Israelis, were analyzed. Clinical characteristics, including median and average age at diagnosis, prostate-specific antigen, Gleason score, and stage of the 402 newly recruited unselected prostate cancer patients (251 Ashkenazi and 151 non-Ashkenazi) diagnosed during the years 1993 to 2003 are presented in Table 1. Samples from 881 Ashkenazi Jewish cancer patients were screened for the RNASEL 471delAAAG mutation. This population included 251 samples from prostate cancer patients, together with DNA samples from 53 patients with transitional cell carcinoma of the bladder, 276 patients with breast and/or ovarian cancer, and 301 patients with colon cancer (Table 2). The 121 unselected prostate cancer patients previously described by Rennert et al. (ref. 13; 87 Ashkenazi and 34 non-Ashkenazi; median age at diagnosis 69 years, range 51-91) were screened for additional mutations in the entire RNASEL gene.
. | Ashkenazi, n (%) . | Non-Ashkenazi, n (%) . | Total, n (%) . | |||
---|---|---|---|---|---|---|
Age at diagnosis | ||||||
Average | 69 | 69 | 69 | |||
Median | 70 (range, 49-92) | 69 (range, 51-85) | 70 (range, 49-92) | |||
PSA at diagnosis (ng/ml) | ||||||
≤4 | 20 (8.0) | 7 (4.6) | 27 (6.7) | |||
4.1-10 | 152 (60.5) | 105 (69.6) | 257 (64.0) | |||
10.1-20 | 57 (22.7) | 23 (15.2) | 80 (19.9) | |||
>20 | 8 (3.2) | 11 (7.3) | 19 (4.7) | |||
Unknown | 14 (5.6) | 5 (3.3) | 19 (4.7) | |||
Gleason score | ||||||
≤6 | 189 (75.3) | 129 (85.4) | 318 (79.1) | |||
≥7 | 49 (19.5) | 21 (13.9) | 70 (17.4) | |||
Unknown | 13 (5.2) | 1 (0.7) | 14 (3.5) | |||
Clinical stage | ||||||
T1 | 145 (57.8) | 85 (56.3) | 230 (57.2) | |||
T2 | 77 (30.7) | 45 (29.8) | 122 (30.3) | |||
T3 | 8 (3.2) | 10 (6.7) | 18 (4.5) | |||
T4 | 2 (0.8) | 1 (0.7) | 3 (0.8) | |||
Unknown | 19 (7.5) | 10 (6.7) | 29 (7.2) |
. | Ashkenazi, n (%) . | Non-Ashkenazi, n (%) . | Total, n (%) . | |||
---|---|---|---|---|---|---|
Age at diagnosis | ||||||
Average | 69 | 69 | 69 | |||
Median | 70 (range, 49-92) | 69 (range, 51-85) | 70 (range, 49-92) | |||
PSA at diagnosis (ng/ml) | ||||||
≤4 | 20 (8.0) | 7 (4.6) | 27 (6.7) | |||
4.1-10 | 152 (60.5) | 105 (69.6) | 257 (64.0) | |||
10.1-20 | 57 (22.7) | 23 (15.2) | 80 (19.9) | |||
>20 | 8 (3.2) | 11 (7.3) | 19 (4.7) | |||
Unknown | 14 (5.6) | 5 (3.3) | 19 (4.7) | |||
Gleason score | ||||||
≤6 | 189 (75.3) | 129 (85.4) | 318 (79.1) | |||
≥7 | 49 (19.5) | 21 (13.9) | 70 (17.4) | |||
Unknown | 13 (5.2) | 1 (0.7) | 14 (3.5) | |||
Clinical stage | ||||||
T1 | 145 (57.8) | 85 (56.3) | 230 (57.2) | |||
T2 | 77 (30.7) | 45 (29.8) | 122 (30.3) | |||
T3 | 8 (3.2) | 10 (6.7) | 18 (4.5) | |||
T4 | 2 (0.8) | 1 (0.7) | 3 (0.8) | |||
Unknown | 19 (7.5) | 10 (6.7) | 29 (7.2) |
Abbreviation: PSA, prostate-specific antigen.
Study group . | No. tested . | No. carriers (%) . | ||
---|---|---|---|---|
Cancer patients | ||||
Prostate cancer | 251 | 6 (2.4%)** | ||
Bladder cancer (transitional cell carcinoma) | 53 | 1 (1.9%) | ||
Breast/ovarian cancer | 276 | 4 (1.5%) | ||
Colon cancer | 301 | 7 (2.3%) | ||
Other cancer patients | 630 | 12 (1.9%)† | ||
Total cancer patients | 881 | 18 (2.0%)‡ | ||
Controls | ||||
Elderly males (59-92 y) | 83 | 2 (2.4%) | ||
Young females (20-45 y) | 250 | 7 (2.8%) | ||
Elderly females (45-80 y) | 177 | 1 (0.6%) | ||
Total controls | 510 | 10 (2.0%)*,†,‡ | ||
Total Ashkenazis | 1391 | 28 (2.0%) |
Study group . | No. tested . | No. carriers (%) . | ||
---|---|---|---|---|
Cancer patients | ||||
Prostate cancer | 251 | 6 (2.4%)** | ||
Bladder cancer (transitional cell carcinoma) | 53 | 1 (1.9%) | ||
Breast/ovarian cancer | 276 | 4 (1.5%) | ||
Colon cancer | 301 | 7 (2.3%) | ||
Other cancer patients | 630 | 12 (1.9%)† | ||
Total cancer patients | 881 | 18 (2.0%)‡ | ||
Controls | ||||
Elderly males (59-92 y) | 83 | 2 (2.4%) | ||
Young females (20-45 y) | 250 | 7 (2.8%) | ||
Elderly females (45-80 y) | 177 | 1 (0.6%) | ||
Total controls | 510 | 10 (2.0%)*,†,‡ | ||
Total Ashkenazis | 1391 | 28 (2.0%) |
Ashkenazi prostate cancer patients versus total controls: odds ratio, 1.22 (95% confidence interval, 0.39-3.71; P = 0.90).
Other Ashkenazi cancer patients versus total controls: odds ratio, 0.97 (95% confidence interval, 0.39-2.44; P = 0.88).
Total Ashkenazi cancer patients versus total controls: odds ratio, 1.04 (95% confidence interval, 0.45-2.44; P = 0.93).
Controls included the previously described (13) 83 elderly Ashkenazi men (ages 59-92 years), 150 healthy young Ashkenazi women (ages 20-45 years), and 100 healthy young non-Ashkenazi women (ages 20-45 years), together with an additional 100 healthy young Ashkenazi women (ages 20-45 years) and 177 healthy middle aged and elderly Ashkenazi women (ages 45-80 years).
DNA samples were blinded and tested in an anonymous manner. The Institutional and National Supreme Helsinki Committees for Genetic Studies approved the study protocols and the informed consents.
Mutation Screening
Ten pairs of primers were used to amplify the coding region, including exon-intron boundaries and the untranslated regions. PCR reactions were done using StartFast Taq Polymerase according to the recommendation of the manufacturer (Roche Diagnostics, Mannheim, Germany) using a Biometra PCR system (Biometra, GmbH Gottingen, Germany). DNA alterations were analyzed using a WAVE denaturing high-performance liquid chromatography (DHPLC) apparatus as described previously (13, 21). Every amplicon with an apparently abnormal chromatogram was sequenced at each DNA strand using an ABI Prism 310 Genetic Analyzer (Applied Biosystems, Foster City, CA). To allow identification of 471delAAAG homozygotes, mixing studies using normal DNA were done on all samples. The primer sequences and detailed DHPLC conditions are available upon request.
Genotyping of the 1623T>G mutation was done using DpnII restriction enzyme (New England Biolabs, Beverly, MA). The uncut wild-type 401 bp allele was digested to 346 and 55 bp fragments in the presence of this mutation. To ensure complete digestion with DpnII, positive and negative controls were routinely included in each assay. In addition to DpnII digestion, all patient samples were analyzed for heteroduplex formation using DHPLC. Because DHPLC does not differentiate between homozygous wild-type 1623T alleles and homozygous 1623G mutant alleles, mixing studies using normal DNA were done on all prostate cancer patient samples to allow identification of 1623G homozygotes. Results from both DpnII digestion and DHPLC analysis yielded identical genotypes, confirming the adequacy of DpnII digestion for the accurate detection of 1623G/1623T alleles.
Multiplex Ligation–Dependent Probe Amplification
Multiplex ligation–dependent probe amplification (MLPA) analysis was done in 300 unselected prostate cancer patients and 10 healthy controls based on a previously described procedure (22) that was modified for the use of synthetic primers. MLPA probes were designed for exons 1 to 6, the putative promoter, and part of the 5′ and 3′ untranslated regions, and for three additional control genes on chromosomes 1, 17, and X (DPYN, p53, and HPRT1, respectively; Fig. 1B). Each probe consisted of one short and one long synthetic oligonucleotide. The short oligonucleotide contained a target-specific sequence (25 nucleotides) at the 3′ end and a common 19-nucleotide sequence, identical to the labeled PCR primer, at the 5′ end. The long MLPA oligonucleotide contained a target-specific oligonucleotide of 35 nucleotides at the 5′ phosphorylated end, a 23-nucleotide sequence complementing the unlabeled PCR primer and common to all probes at the 3′ end, and a stuffer sequence of variable length in between. The synthetic oligonucleotides are available upon request. The statistical analysis of MLPA output was modified based on the Dosage Quotients algorithm described in detail in the National Genetics Reference Laboratories website (www.ngrl.org.uk/Manchester/Downloads/Health%20Technology%20Assessment/Theory%20of%20the%20analysis%20method.pdf).
Genotyping
Thirty-two Ashkenazi individuals, cancer patients, and controls, carrying the 471delAAAG mutation in the RNASEL gene, and 50 noncarrier Ashkenazi prostate cancer patients were genotyped for four microsatelite markers. These markers, D1S240, D1S158, D1S2818, and D1S466 (ordered from telomere to centromere), span 2.88 Mb around the RNASEL gene, which is located between D1S158 and D1S2818. The distances of the STS markers from D1S240 were determined according to Sequence Map (http://www.ncbi.nih.gov/genome/sts), and were 1.05, 2.85, and 2.88 cM, respectively (assuming 1 cM = 1 Mb). The length of each amplified product was determined using ABI Prism 310 Genetic Analyzer and analyzed using GeneScan Analysis Software version 3.1.2 (Applied Biosystems).
Estimating Mutation Age
Models used to calculate the age of the RNASEL 471delAAAG mutation were based on a priori estimation of the population growth rate using the following formula: Nt = Noert, where No is the initial population size, Nt is the current population size, t is the time in generations since emergence of the initial population, and r is the population growth rate (23). Assuming that the Jewish Ashkenazi population numbered ∼15,000 individuals in the year 1500, and numbered 6,000,000 at the end of the 19th century (24), the population growth rate in this period was determined to be 0.345. Calculation of the mutation age was done using Markov Chain Monte Carlo algorithm, implemented in the DMLE + V2.2 program (http://dmle.org) as described previously (25).
Statistical Analysis
Analyses were done using the EpiInfo 2000 software (http://www.cdc.gov/epiinfo). The odds ratios and confidence intervals were calculated as an estimation of risk among mutation carriers. χ2 with Yates correction was used, when appropriate, to determine significant differences in the frequency of genetic variations between patients and controls.
Results
Screening for the RNASEL 471delAAAG Mutation in Ashkenazi Cancer Patients
The RNASEL 471delAAAG mutation was detected in 6 of the 251 Ashkenazi patients (2.4%) and in 10 of the 510 Ashkenazi controls (2.0%; Table 2). This mutation was not detected in non-Ashkenazi patients or controls. The frequency of the 471delAAAG mutation in Ashkenazi prostate cancer patients did not differ statistically from the carrier frequency in the control population (P = 0.9).
To assess whether this founder mutation is possibly associated with other common malignancies in Jews of Ashkenazi origin, we tested cohorts of patients with bladder, breast/ovarian, and colon cancers (Table 2). The RNASEL 471delAAAG mutation was detected in 1 of the 53 bladder cancer patients (1.9%), in 4 of the 276 women with breast and/or ovarian cancer (1.5%), and in 7 of 301 colon cancer patients (2.3%). Altogether, we detected the 471delAAAG mutation in 2.04% (18 of 881) of the Ashkenazi patients with prostate, bladder, breast/ovarian, and colon cancers, which did not differ statistically from the frequency of this mutation in the 510 Ashkenazi controls (P = 0.93). The cumulative 471delAAAG carrier frequency among all Ashkenazi samples tested, which included both cancer patients and healthy controls, was 2.01% (28 of 1,391).
Screening for Other Genetic Alterations in the RNASEL Gene
To determine whether known or novel RNASEL mutations other than the 471delAAAG mutation may contribute to prostate cancer risk in the Israeli population, the entire RNASEL gene was screened for genetic alterations in the 121 prostate cancer patients described previously (13). WAVE DHPLC analysis identified seven additional sequence variants in this population, which were confirmed by ABI sequencing (Table 3). Three of the five exonic variants detected were reported previously, the 1385G>A (Arg462Gln; refs. 8-12, 17-19) and 1623T>G (Asp541Glu; refs. 8-12, 18, 19) missense mutations, and the 354C>T silent change (10, 17). Arg462Gln was detected in only 1 of 87 Ashkenazi patients and in 4 of 233 controls (P = 0.88). The 1623T>G allele (Asp541Glu) was the most common single nucleotide substitution detected in our population. No significant differences were detected in its genotype frequencies among the three study groups, Ashkenazi and non-Ashkenazi prostate cancer patients and Ashkenazi elderly controls (Table 3, P = 0.92), and the 1623T>G allele frequencies were also similar (56%, 54%, and 54%, respectively). Due to the high carrier frequency of the 354C>T polymorphism (Table 3), this variant was screened in an expanded population, including a total of 523 prostate cancer patients (338 Ashkenazi and 185 non-Ashkenazi) and 610 controls (510 Ashkenazi and 100 non-Ashkenazi). 354C>T carrier frequency did not differ significantly between Ashkenazi (16 of 338, 3.6%) and non-Ashkenazi (3 of 185, 1.6%) patients (P = 0.12), between Ashkenazi (13 of 510, 2.3%) and non-Ashkenazi (0 of 100) controls (P = 0.22), between Ashkenazi prostate cancer patients and controls (P = 0.13), between non-Ashkenazi prostate cancer patients and controls (P = 0.50), or between all prostate cancer patients tested (19 of 523, 3.6%) and the total control population (13 of 610, 2.1%; P = 0.18).
Exon . | DNA (amino acid) variation . | Ashkenazi patients (%) . | Ashkenazi controls (%) . | Non-Ashkenazi patients (%) . | ||
---|---|---|---|---|---|---|
. | . | No. individuals tested . | . | . | ||
. | . | 87 . | 233* . | 34 . | ||
Exon 1 | 354C>T | 4 (4.6) | 5 (2.1) | — | ||
827C>T (A276V) | — | 1 (0.4) | — | |||
1385G>A (R462Q) | 1 (1.1) | 4 (1.7) | — | |||
Exon 3 | 1596A>G | 1 (1.1) | 1 (1.1)† | — | ||
1623T>G (D541E) | ||||||
TT | 16 (18.4) | 17 (20.5)† | 6 (17.6) | |||
TG | 45 (51.7) | 43 (51.8)† | 19 (55.9) | |||
GG | 26 (29.9) | 23 (27.7)† | 9 (26.5) | |||
Intron 4 | IVS4-33A>G | — | 4 (1.7) | — | ||
Intron 5 | IVS5+1delG | 1 (1.1) | — | — |
Exon . | DNA (amino acid) variation . | Ashkenazi patients (%) . | Ashkenazi controls (%) . | Non-Ashkenazi patients (%) . | ||
---|---|---|---|---|---|---|
. | . | No. individuals tested . | . | . | ||
. | . | 87 . | 233* . | 34 . | ||
Exon 1 | 354C>T | 4 (4.6) | 5 (2.1) | — | ||
827C>T (A276V) | — | 1 (0.4) | — | |||
1385G>A (R462Q) | 1 (1.1) | 4 (1.7) | — | |||
Exon 3 | 1596A>G | 1 (1.1) | 1 (1.1)† | — | ||
1623T>G (D541E) | ||||||
TT | 16 (18.4) | 17 (20.5)† | 6 (17.6) | |||
TG | 45 (51.7) | 43 (51.8)† | 19 (55.9) | |||
GG | 26 (29.9) | 23 (27.7)† | 9 (26.5) | |||
Intron 4 | IVS4-33A>G | — | 4 (1.7) | — | ||
Intron 5 | IVS5+1delG | 1 (1.1) | — | — |
NOTE: Details regarding the 471delAAAG mutation are presented in Table 1. —, not detected.
Including 83 elderly males and 150 young females.
Tested only in the 83 elderly male controls.
Screening also identified two novel exonic variants. A missense mutation, Ala276Val (827C>T), was detected in a single control, but in no prostate cancer patients, and a silent change, 1596A>G, was detected in one Ashkenazi patient and one control (Table 3).
Two novel intronic variants were also identified (Table 3). A single nucleotide change (IVS4-33A>G) in intron 4 was detected in 4 of the 233 healthy Ashkenazi controls (1.7%), but in none of the 87 Ashkenazi patients (P = 0.51). This intronic variant was not detected in any of the 34 non-Ashkenazi patients. A novel potentially pathogenic splice-site mutation in intron 5 (IVS5+1delG, Fig. 1A) was detected in 1 of 87 Ashkenazi prostate cancer patients. An additional 251 prostate cancer patients were then screened, although none harbored this mutation nor was it detected in any of the non-Ashkenazi patients (n = 185) or healthy Ashkenazi controls tested (n = 233).
Multiplex Ligation–Dependent Probe Amplification
MLPA is a high-resolution method for detecting copy number changes in genomic sequences (22). DNA samples from 300 prostate cancer patients (188 Ashkenazi and 112 non-Ashkenazi) and 10 controls were screened for either deletions or duplications in the RNASEL gene (Fig. 1B). No such genetic variations were detected in any of these samples.
Estimated Age of the RNASEL 471delAAAG Mutation
Calculating the age of this founder mutation in the Ashkenazi Jewish population is based on the decay of the linkage disequilibrium between the mutation and nearby polymorphic markers on the same chromosome. The allele frequencies of the four microsatellite DNA markers analyzed in 32 Ashkenazi carriers and 50 Ashkenazi noncarriers are detailed in Table 4. The alleles were recoded according to their sizes using SPSS software version 12 (SPSS, Inc., Chicago, IL). Integrating the estimated population growth rate of 0.345, the distances between the relevant markers and the genotypes of the carriers and noncarriers, our calculations suggest that the RNASEL 471delAAAG mutation originated ∼80 generations ago, with a 95% credible set of values of 73 to 92 generations. It is of note that the estimated population growth rate of 0.345, determined for the 16th to 19th centuries, a period of rapid growth of the Ashkenazi Jewish population, is an overestimate regarding the earlier period. Population growth rates before 1500 are difficult to estimate due to the lack of reliable data; therefore, when applying the rate of 0.345 to this period, the estimated mutation age is an underestimate and is likely more accurately represented by the higher values of the estimated range (80-92 generations). Because a generation at that time was estimated as 20 years (26), we calculated that this mutation occurred ∼1,600 to 1,840 years ago, between the 2nd and 5th centuries A.D.
Allele (no. repeats) . | Frequency in 471delAAAG carriers (n = 64 chromosomes) . | Frequency in noncarriers (n = 100 chromosomes) . | ||
---|---|---|---|---|
D1S240 | ||||
8 | — | 0.01 | ||
9 | — | 0.01 | ||
10 | 0.09 | 0.18 | ||
11 | 0.02 | 0.02 | ||
12 | 0.78 | 0.59 | ||
13 | 0.08 | 0.16 | ||
14 | 0.03 | 0.03 | ||
D1S158 | ||||
16 | 0.02 | 0.01 | ||
17 | — | 0.03 | ||
18 | 0.03 | 0.01 | ||
19 | 0.02 | 0.03 | ||
20 | 0.02 | 0.07 | ||
21 | 0.05 | 0.11 | ||
22 | 0.48 | 0.22 | ||
23 | 0.06 | 0.05 | ||
24 | 0.09 | 0.10 | ||
25 | 0.03 | 0.15 | ||
26 | 0.14 | 0.05 | ||
27 | — | 0.04 | ||
28 | 0.05 | 0.01 | ||
29 | — | 0.01 | ||
30 | 0.02 | 0.01 | ||
D1S2818 | ||||
13 | 0.02 | — | ||
14 | 0.75 | 0.53 | ||
15 | 0.08 | 0.15 | ||
16 | — | 0.03 | ||
17 | 0.06 | 0.07 | ||
18 | 0.03 | 0.03 | ||
19 | 0.05 | 0.17 | ||
20 | 0.02 | 0.01 | ||
21 | — | 0.01 | ||
D1S466 | ||||
17 | — | 0.01 | ||
18 | 0.03 | 0.01 | ||
19 | 0.17 | 0.42 | ||
20 | 0.50 | 0.02 | ||
21 | 0.14 | 0.20 | ||
22 | — | 0.03 | ||
23 | 0.09 | 0.14 | ||
24 | 0.02 | 0.10 | ||
25 | 0.02 | 0.06 | ||
28 | 0.02 | — | ||
29 | 0.02 | — |
Allele (no. repeats) . | Frequency in 471delAAAG carriers (n = 64 chromosomes) . | Frequency in noncarriers (n = 100 chromosomes) . | ||
---|---|---|---|---|
D1S240 | ||||
8 | — | 0.01 | ||
9 | — | 0.01 | ||
10 | 0.09 | 0.18 | ||
11 | 0.02 | 0.02 | ||
12 | 0.78 | 0.59 | ||
13 | 0.08 | 0.16 | ||
14 | 0.03 | 0.03 | ||
D1S158 | ||||
16 | 0.02 | 0.01 | ||
17 | — | 0.03 | ||
18 | 0.03 | 0.01 | ||
19 | 0.02 | 0.03 | ||
20 | 0.02 | 0.07 | ||
21 | 0.05 | 0.11 | ||
22 | 0.48 | 0.22 | ||
23 | 0.06 | 0.05 | ||
24 | 0.09 | 0.10 | ||
25 | 0.03 | 0.15 | ||
26 | 0.14 | 0.05 | ||
27 | — | 0.04 | ||
28 | 0.05 | 0.01 | ||
29 | — | 0.01 | ||
30 | 0.02 | 0.01 | ||
D1S2818 | ||||
13 | 0.02 | — | ||
14 | 0.75 | 0.53 | ||
15 | 0.08 | 0.15 | ||
16 | — | 0.03 | ||
17 | 0.06 | 0.07 | ||
18 | 0.03 | 0.03 | ||
19 | 0.05 | 0.17 | ||
20 | 0.02 | 0.01 | ||
21 | — | 0.01 | ||
D1S466 | ||||
17 | — | 0.01 | ||
18 | 0.03 | 0.01 | ||
19 | 0.17 | 0.42 | ||
20 | 0.50 | 0.02 | ||
21 | 0.14 | 0.20 | ||
22 | — | 0.03 | ||
23 | 0.09 | 0.14 | ||
24 | 0.02 | 0.10 | ||
25 | 0.02 | 0.06 | ||
28 | 0.02 | — | ||
29 | 0.02 | — |
NOTE: —, not detected.
Discussion
Nearly a decade has passed since the first genome-wide scan detected linkage to HPC1 (4), yet the role of RNASEL, the hereditary prostate cancer candidate gene identified at this locus, remains equivocal. Substantial data supported the association between RNASEL variants and prostate cancer risk, although replication studies challenge this association (15, 16). In the reassessment of the association between the founder RNASEL 471delAAAG frameshift mutation and prostate cancer risk, carrier frequencies did not differ statistically between Ashkenazi prostate cancer patients and controls. Taking into account the frequency of the 471delAAAG mutation in Ashkenazi prostate cancer patients from our previous (6 of 87; ref. 13) and current (6 of 251) studies, together with the carrier frequency in Canadian Ashkenazi prostate cancer patients (1 of 111; ref. 20), the overall Ashkenazi prostate cancer carrier frequency is 2.9% (13 of 449), which is not statistically different from the combined carrier frequency in control Ashkenazi Israeli (10 of 510) and Canadian (2 of 105) populations (2%, 12 of 615, P = 0.42). Furthermore, the combined carrier frequency in the current prostate cancer population together with the Canadian population (7 of 362, 1.93%) is almost exactly the carrier frequency in the control populations (12 of 615, 1.95%). Additionally, the similar carrier frequencies of RNASEL 471delAAAG in Ashkenazi patients with bladder, breast/ovarian, and colon cancers do not support a role for this variant in the predisposition to these malignancies in Ashkenazi Jews.
The possibility that the RNASEL 471delAAAG founder mutation confers an increased risk for carcinogenesis was also examined by estimating its time of coalescence, and comparing it with that of the well-established cancer-causing Ashkenazi founder mutation, BRCA1 185delAG. Our analysis estimated that RNASEL 471delAAAG originated between the 2nd and 5th centuries A.D., coinciding with the period of westward movement of individual Jews, who in the 2nd and 3rd centuries A.D. ventured out of their traditional Mediterranean habitat, following the Roman legions to European territories beyond the Alps (27). The identification of BRCA1 185delAG in both Ashkenazi (28) and Iraqi (29) Jews, in the range of 1% in both of these subpopulations, suggests that it likely originated before the dispersion of the Jews ∼70 A.D. (30). BRCA1 185delAG allele frequency is influenced by the association of dominant mutations with breast and ovarian carcinogenesis that are in many cases lethal, and likely by selection against homozygous carriers. Although the consequences of human BRCA1 mutation homozygosity have, to our knowledge, not been reported in the literature, human homozygotes, like Brca1-deficient mouse models, likely die in utero (31, 32). Homozygosity for RNASEL mutations is apparently compatible with life, as shown by the viability and fertility of RNaseL−/− mice (33) and by the patient homozygous for the 471delAAAG mutation reported in our previous article (13). However, data available to date are not sufficient to determine unequivocally whether this mutation is not associated with lethal disease, and the significance of this possibly pathogenic truncating mutation in ∼2% of the Ashkenazi Jewish population remains to be elucidated.
The substantial and conflicting data from worldwide mutation screening and association studies of RNASEL motivated our comprehensive analysis of RNASEL alterations in an Israeli Jewish prostate cancer cohort. The controversial RNASEL Arg462Gln polymorphism, which is positively associated with disease risk in a Finnish population (9), and suggested to account for up to 13% of prostate cancer susceptibility in the United States (11), was identified in only one (1.1%) Ashkenazi prostate cancer patient, and in 4 (1.7%) Ashkenazi controls. Our findings are consistent with those from Swedish (18) and German (19) studies, and do not support a significant association between the RNASEL Arg462Gln variant and prostate cancer predisposition. Whether discrepancies are due to differences in study design, differences in mutation frequencies in different populations, or possibly, result from the modifying effect of specific RNASEL alterations on other, yet unidentified rare susceptibility genes remains to be explored.
Another example of a population-dependent effect of an RNASEL variant and prostate cancer risk is the Asp541Glu polymorphism, which was associated with increased prostate cancer risk in Japan (12) and decreased prostate cancer risk in Sweden (18). This variant was detected at similar frequencies in our Ashkenazi and non-Ashkenazi patients, and in Ashkenazi controls, and like data from Finnish (9) and United States (10, 11) populations do not support an association between the Asp541Glu polymorphism and disease risk.
Interestingly, the truncating Glu265X (8, 9, 18, 19) and the missense Ile97Leu (8, 10, 19, 34) mutations, previously described in families with prostate cancer, were not identified in any of 121 patients or 83 controls screened for these mutations, further emphasizing the possible influence of ethnic and geographic factors on the frequencies and distribution of different RNASEL alleles.
Of the four novel sequence alterations identified in this study, the most interesting was the potentially pathogenic IVS5+1delG splice site mutation. However, the detection of this RNASEL mutation in only 1 of 523 Ashkenazi and non-Ashkenazi prostate cancer patients challenges the significance of this possibly pathogenic mutation in predisposition to prostate cancer.
Finally, because RNASEL has been suggested as a tumor suppressor gene (35) and the role of chromosomal deletions in tumor suppressor genes has been well established in tumorigenesis (36), we hypothesized that RNASEL germ line deletions or duplications may be associated with prostate cancer risk. MLPA did not detect deletions or duplications in any of the exons as well as the presumptive promoter site and in parts of the 5′ and 3′ untranslated regions of 300 prostate cancer patients.
In summary, whereas the effect of RNASEL on prostate cancer predisposition remains controversial, our comprehensive analysis of RNASEL gene copy number and coding sequence does not support a significant role for RNASEL genetic alterations in prostate cancer predisposition in Israeli Ashkenazi and non-Ashkenazi Jews.
Grant support: M.K. Humanitarian Fund and the Prostate Cancer Foundation, Israel.
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.
Acknowledgments
We thank Amira Stenger, M.Sc., for technical assistance.