Abstract

The contribution of BRCA1 and BRCA2 to familial and non-familial forms of breast cancer has been difficult to accurately estimate because of the myriad of potential genetic and epigenetic mechanisms that can ultimately influence their expression and involvement in cellular activities. As one of these potential mechanisms, we investigated whether allelic imbalance (AI) of BRCA1 or BRCA2 expression was associated with an increased risk of developing breast cancer. By developing a quantitative approach utilizing allele-specific real-time PCR, we first evaluated AI caused by nonsense-mediated mRNA decay in patients with frameshift mutations in BRCA1 and BRCA2. We next measured AI for BRCA1 and BRCA2 in lymphocytes from three groups: familial breast cancer patients, non-familial breast cancer patients and age-matched cancer-free females. The AI ratios of BRCA1, but not BRCA2, in the lymphocytes from familial breast cancer patients were found to be significantly increased as compared to cancer-free women (BRCA1: 0.424 versus 0.211, P = 0.00001; BRCA2: 0.206 versus 0.172, P = 0.38). Similarly, the AI ratios were greater for BRCA1 and BRCA2 in the lymphocytes of non-familial breast cancer cases versus controls (BRCA1: 0.353, P = 0.002; BRCA2: 0.267, P = 0.03). Furthermore, the distribution of under-expressed alleles between cancer-free controls and familial cases was significantly different for both BRCA1 and BRCA2 gene expression (P < 0.02 and P < 0.02, respectively). In conclusion, we have found that AI affecting BRCA1 and to a lesser extent BRCA2 may contribute to both familial and non-familial forms of breast cancer.

INTRODUCTION

Breast cancer is the most common cancer affecting women, with a lifetime risk among females ∼10% by the age of 80 years. In the USA, it has been reported that there will be approximately 180 510 new cases of breast cancer, and more than 40 910 breast cancer-related deaths in 2007 (1). Current estimates suggest that family history is associated with 10–20% of breast cancer (2,3). BRCA1 (OMIM: 113705) and BRCA2 (OMIM: 600185) are two of the most prominent breast cancer susceptibility genes and deleterious mutations in these two genes are estimated to account for about 15–30% of familial breast cancer (4–6).

Germline mutations affecting the coding region of BRCA1 and BRCA2 are thought to lead to expression of mutant proteins, which are either inactive or function as dominant negatives. However, these scenarios have not been supported by functional studies (7–9). In fact, Brca1 and Brca2 knockout mouse models have demonstrated that elimination of Brca1 or Brca2 proteins is sufficient for the development of mammary cancer (10,11). Previously, we have reported that mutant BRCA1 mRNAs containing premature stop codons were eliminated or destabilized by nonsense-mediated mRNA decay (NMD) (12) and lead to a state of haploinsufficiency. As a result, the ratios between the expressions from the mutant alleles and the corresponding wild-type alleles were significantly decreased, resulting in what was referred to as allelic imbalance (AI). AI of BRCA1 or BRCA2 expression could decrease the level of both transcripts and proteins and thus contribute to increased susceptibility of developing breast cancer.

There is growing evidence to support this concept. Epigenetic studies have shown that loss of BRCA1 expression due to promoter hypermethylation is associated with ∼10% of sporadic cases of breast and ovarian cancer (13–18). However, screens to evaluate AI have not been applied in depth to study its potential role in the genesis of familial forms of these diseases. A previous study reported that 6 out of 13 human genes, including BRCA1 and p53, were expressed with significant difference between the two alleles, and this difference was transmitted by Mendelian inheritance (19). Furthermore, Yan et al. (20) observed that decreased expression of one of the adenomatous polyposis coli tumor suppressor gene (APC) alleles was associated with the development of familial adenomatous polyposis. Their studies also found that even more modest decreases in the expression of one APC allele could contribute to attenuated forms of polyposis (20). Based on these findings, we hypothesize that a subset of non-BRCA1/2 mutation carriers with a strong family history of breast cancer are at increased risk of developing this disease as a result of AI in BRCA1 and BRCA2 gene expression.

In the present study, we have developed a quantitative approach to measure the allele-specific expression of BRCA1 and BRCA2. We compared BRCA1/2 allelic variation in a cohort of BRCA1/2 mutation-negative familial breast cancer patients, non-familial breast cancer patients and age-matched cancer-free volunteers. Since susceptibility to breast cancer is far from being fully understood, our study may help to further identify genetic factors which contribute to breast cancer susceptibility.

RESULTS

Development of a quantitative allelic imbalance assay

In order to determine if allele-specific real-time PCR is able to quantitatively measure the AI in BRCA1 and BRCA2 gene expression from the individual allele, RNAs were isolated from the blood lymphocytes of two individuals determined by genotype and sequence analysis to be homozygous for either BRCA1-c.4308T/T or BRCA1-c.4308C/C (Fig. 1A). This polymorphism was chosen since it is relatively common, based on NCBI dbSNP data. The samples were then reverse transcribed and the cDNAs were mixed at various ratios (8:1, 4:1, 2:1, 1:1, 1:2, 1:4 and 1:8) as described in the Materials and Methods section. BRCA1-c.4308T/T was detected by the VIC fluorescence signal and BRCA1-c.4308C/C was detected by the FAM fluorescence signal. As shown in Figure 1B, with decreasing cDNA ratios of c.4308T to c.4308C, the VIC curve (detecting the c.4308T allele) shifted to the right with the increasing value of CT-c.4308T (VIC), while the curve of FAM (detecting c.4308C allele) shifted to the left with the decreasing value of CT-c.4308C (VIC). At the same time, the value of ΔCT (CT-c.4308T (VIC)–CT-c.4308C (FAM)) changed from the negative to the positive. By the regression analysis, a linear relationship between Log2 ratio of cDNAs c.4308T to c.4308C and ΔCT was identified: Log2 (c.4308T/C) = −0.0877 + 1.57897 * ΔCT (P < 0.001) (Fig. 1C). The Pearson correlation coefficient (r) between Log2 (c.4308T/c.4308C) and ΔCT was 0.9798. To establish a similar standard curve for BRCA2 allelic expression, cDNAs from two individuals, who were either homozygous for BRCA2-c.3396A/A or BRCA2-c.3396G/G, were mixed at the following ratios: 8:1, 4:1, 2:1, 1:1, 1:2, 1:4 and 1:8 (c.3396A/A allele:c.3396G/G allele). BRCA2-c.3396A was detected by the VIC fluorescence signal and BRCA2-c.3396G was detected by the FAM fluorescence signal. As shown in Figure 1D, with decreasing ratios of c.3396A to c.3396G, the VIC curve (detecting c.3396A allele) shifted to the right while the FAM curve (detecting c.3396G allele) shifted to the left. After regression analysis, a linear relationship between Log2 (c.3396A/c.3396G) and ΔCT was identified: Log2 (c.3396A/G) = 0.11726 + 1.26458 * ΔCT (P < 0.001) (Fig. 1E). The Pearson correlation coefficient (r) between Log2 (c.3396A/G) and ΔCT was 0.9868.

Figure 1.

Standard curves for BRCA1 and BRCA2 allelic imbalance.

(A) Allele-specific real-time PCR amplification plot analyses of BRCA1-c.4308T (VIC) and -c.4308C (FAM) was performed in cDNAs generated by RT–PCR using RNAs from blood lym

phocytes of two individuals homozygous for either the BRCA1-c.4308T/T or BRCA1-c.4308C/C. DNA sequencing chromatograms confirming the genotype are shown in the right panel. (B) Allele-specific real-time PCR amplification plot was analyzed in mixed cDNAs of BRCA1-c.4308T/T (detected by VIC) and BRCA1-c.4308C/C (detected by FAM) at the following ratios: 8:1, 4:1, 2:1, 1:1, 1:2, 1:4 and 1:8, respectively. (C) The standard curve for BRCA1 allelic imbalance: Log2 (c.4308T/C) = −0.0877 + 1.57897 * ΔCT. The Pearson correlation coefficient (r) between Log2 (c.4308T/c.4308C) and ΔCT was 0.9798 (Data expressed as Mean ± SD, n=3; the mean value of ΔCT for c.4308T/C=1 has been adjusted to zero). (D) Allele-specific real-time PCR amplification plot was analyzed in mixed cDNAs of BRCA2-c.3396A/A (detected by VIC) and BRCA2-c.3396G/G (detected by FAM) at the following ratios: 8:1, 4:1, 2:1, 1:1, 1:2, 1:4 and 1:8, respectively. (E) The standard curve for BRCA2 allelic imbalance: Log2 (c.3396A/G) = 0.11726 + 1.26458 * ΔCT. The Pearson correlation coefficient (r) between Log2 (c.3396A/G) and ΔCT was 0.9868 (Data expressed as Mean ± SD, n = 3; the mean value of ΔCT for c.3396A/G = 1 has been adjusted to zero).

Figure 1.

Standard curves for BRCA1 and BRCA2 allelic imbalance.

(A) Allele-specific real-time PCR amplification plot analyses of BRCA1-c.4308T (VIC) and -c.4308C (FAM) was performed in cDNAs generated by RT–PCR using RNAs from blood lym

phocytes of two individuals homozygous for either the BRCA1-c.4308T/T or BRCA1-c.4308C/C. DNA sequencing chromatograms confirming the genotype are shown in the right panel. (B) Allele-specific real-time PCR amplification plot was analyzed in mixed cDNAs of BRCA1-c.4308T/T (detected by VIC) and BRCA1-c.4308C/C (detected by FAM) at the following ratios: 8:1, 4:1, 2:1, 1:1, 1:2, 1:4 and 1:8, respectively. (C) The standard curve for BRCA1 allelic imbalance: Log2 (c.4308T/C) = −0.0877 + 1.57897 * ΔCT. The Pearson correlation coefficient (r) between Log2 (c.4308T/c.4308C) and ΔCT was 0.9798 (Data expressed as Mean ± SD, n=3; the mean value of ΔCT for c.4308T/C=1 has been adjusted to zero). (D) Allele-specific real-time PCR amplification plot was analyzed in mixed cDNAs of BRCA2-c.3396A/A (detected by VIC) and BRCA2-c.3396G/G (detected by FAM) at the following ratios: 8:1, 4:1, 2:1, 1:1, 1:2, 1:4 and 1:8, respectively. (E) The standard curve for BRCA2 allelic imbalance: Log2 (c.3396A/G) = 0.11726 + 1.26458 * ΔCT. The Pearson correlation coefficient (r) between Log2 (c.3396A/G) and ΔCT was 0.9868 (Data expressed as Mean ± SD, n = 3; the mean value of ΔCT for c.3396A/G = 1 has been adjusted to zero).

Detection of allelic imbalance caused by nonsense-mediated mRNA decay

To examine whether the allele-specific real-time PCR assay is able to detect AI of BRCA1 and BRCA2 gene expression in cell lines, we evaluated RNAs isolated from lymphoblastoid cell lines (LCLs) which were derived from deleterious mutation carriers heterozygous for BRCA1-c.3671ins4 or BRCA2-c.796delT. These frame-shift mutations create the premature stop codons, which are predicted to activate the NMD pathway and thus lead to decreased levels of mRNAs from the mutant alleles (12). As shown in Figure 2A and B, the ratios of BRCA1-c.4308T to -c.4308C between wild type and BRCA1-c.3671ins4 heterozygous samples were 0.93 ± 0.04 and 2.07 ± 0.06, respectively (P < 0.01). By subcloning and sequencing the individual transcripts, we found that the under-expressed allele contained both the BRCA1-c.3671ins4 mutation and the BRCA1-c.4308C polymorphism (detected by the FAM signal) (data not shown). To further examine if the loss of BRCA1-c.3671ins4 was associated with NMD, we treated the BRCA1-c.3671ins4 LCLs with puromycin, a translational inhibitor, 14 h prior to RNA isolation. The ratio of BRCA1-c.4308T to -c.4308C in BRCA1-c.3671ins4 heterozygous cells decreased ∼30%, in comparison to the non-treatment group (1.50 ± 0.05 versus 2.07 ± 0.06, P < 0.01) (Fig. 2B). Our data indicated that treatment with puromycin was able to partially recover the AI caused by NMD. Significant AI was also observed for the BRCA2-c.796delT mutant allele. The ratios of BRCA2-c.3396G to -c.3396A between wild-type and BRCA2-c.796delT heterozygous samples were 0.98±0.06 and 6.59 ± 1.31, respectively (P < 0.01). After treating the BRCA2-c.796delT LCLs with puromycin, the ratio of BRCA2-c.3396G to -c.3396A in BRCA2-c.796delT heterozygous cells decreased ∼31%, in comparison to the non-treatment group (4.90 ± 0.87 versus 6.25 ± 1.17) (Fig. 2C and D). Our results suggested that the loss of expression of BRCA1 or BRCA2 mutant alleles via NMD significantly contributed to the observed AI.

Figure 2.

BRCA1 and BRCA2 allelic imbalance caused by NMD. (A) Allele-specific real-time PCR amplification plots of BRCA1-c.4308T (VIC) and -c.4308C (FAM) for non-template control, BRCA1 wild-type lymphoblastoid cells (WT), BRCA1 mutant (heterozygous BRCA1-c.3671ins4) lymphoblastoid cells without [PC (−)] or with [PC (+)] puromycin treatment. (B) Allelic expression ratios of BRCA1-c.4308T to BRCA1-c.4308C (a: versus WT; b: versus PC (+); t-test, P < 0.05). (C) Allele-specific real-time PCR amplification plots of BRCA2-c.3396A (VIC) and -c.3396G (FAM) for non-template control, BRCA2 wild-type lymphoblastoid cells (WT), BRCA2 mutant (heterozygous BRCA2-c.796delT) lymphoblastoid cells without [PC (−)] or with [PC (+)] puromycin treatment. (D) Allelic expression ratios of BRCA2-c.3396G to BRCA2-c.3396A (a: versus WT; b: versus PC (+); t-test, P < 0.05).

Figure 2.

BRCA1 and BRCA2 allelic imbalance caused by NMD. (A) Allele-specific real-time PCR amplification plots of BRCA1-c.4308T (VIC) and -c.4308C (FAM) for non-template control, BRCA1 wild-type lymphoblastoid cells (WT), BRCA1 mutant (heterozygous BRCA1-c.3671ins4) lymphoblastoid cells without [PC (−)] or with [PC (+)] puromycin treatment. (B) Allelic expression ratios of BRCA1-c.4308T to BRCA1-c.4308C (a: versus WT; b: versus PC (+); t-test, P < 0.05). (C) Allele-specific real-time PCR amplification plots of BRCA2-c.3396A (VIC) and -c.3396G (FAM) for non-template control, BRCA2 wild-type lymphoblastoid cells (WT), BRCA2 mutant (heterozygous BRCA2-c.796delT) lymphoblastoid cells without [PC (−)] or with [PC (+)] puromycin treatment. (D) Allelic expression ratios of BRCA2-c.3396G to BRCA2-c.3396A (a: versus WT; b: versus PC (+); t-test, P < 0.05).

BRCA1 and BRCA2 allelic imbalance is associated with breast cancer risk

To evaluate AI of BRCA1 and BRCA2 gene expression, genotype analysis of the two common polymorphisms, BRCA1-c.4308T/C and BRCA2-c.3396A/G, was performed on DNA samples isolated from fresh-frozen peripheral blood lymphocytes from 85 unrelated BRCA1/2 mutation-negative familial breast cancer carriers (median age at sample collection: 47), 112 non-familial breast cancer carriers (median age at sample collection: 52) and 102 age-matched cancer-free females (median age at sample collection: 51) (Table 1). From these analyses, 37 (43.5%), 48 (42.9%) and 41 (40.2%) of the samples evaluated were determined to be heterozygote for the BRCA1-c.4308T/C polymorphism for familial breast cancer patients, non-familial cancer patients and cancer-free controls, respectively (Table 1). Furthermore, 39 (45.9%), 44 (39.3%) and 36 (35.3%) of the samples above were found to be heterozygous for the BRCA2-c.3396A/G polymorphism (Table 1). Since our initial validation studies were preformed using immortalized LCLs, we first compared AI in RNA isolated from 20 fresh-frozen lymphocytes versus 20 established Epstein-Barr Virus (EBV)-lines. No significant differences were detected between these two sample sets [BRCA1: 0.424 ± 0.129 versus 0.409 ± 0.127 (n = 11); BRCA2: 0.212 ± 0.180 versus 0.225 ± 0.209 (n = 10)]. However, to limit any AI variation potentially introduced by EBV transformation, all subsequent AI assays were performed using RNAs isolated from peripheral blood lymphocytes. Next, RNA isolated from BRCA1-c.4308T/C (n = 126) and BRCA2-c.3396A/G (n = 119) heterozygotes, including single heterozygotes and double heterozygotes, were evaluated for integrity and quantity. Those samples demonstrating high quality and the necessary quantities were used in the AI assay, as described in the Materials and Methods section.

Table 1.

Characteristics of the study groups

Characters Study groups 
 Familial Non-familial Cancer-free 
Sample size 85 112 102 
Age (median)    
 At diagnosis 44 49 NA 
 At sample collection 47 52 51 
Family historya    
 2 or more 85 
 1 23 25 
 0 89 77 
Genotypes    
BRCA1-c.4308T/C 37 48 41 
BRCA2-c.3396A/G 39 44 36 
Characters Study groups 
 Familial Non-familial Cancer-free 
Sample size 85 112 102 
Age (median)    
 At diagnosis 44 49 NA 
 At sample collection 47 52 51 
Family historya    
 2 or more 85 
 1 23 25 
 0 89 77 
Genotypes    
BRCA1-c.4308T/C 37 48 41 
BRCA2-c.3396A/G 39 44 36 

aNumber of first and/or second-degree relatives affected with either breast and/or ovarian cancer.

To evaluate the AI, we used the absolute values of Log2 (BRCA1-c.4308T/C) or Log2 (BRCA2-c.3396A/c.3396G). The mean value of Log2 (c.4308T/C) of BRCA1 in the lymphocytes from familial breast cancer carriers was found to be ∼104% higher than that that in the lymphocytes from cancer-free controls [0.424 ± 0.157 (n = 32) versus 0.211 ± 0.169 (n = 40), P = 0.00001; t-test] (Table 2 and Fig. 3A and B). Log2 of BRCA1-c.4308T/C in the lymphocytes from non-familial breast cancer carriers was 73% higher than that in cancer-free controls [0.353 ± 0.209 (n = 32), P = 0.002 versus control] (Table 2 and Fig. 3A and C). In comparison, the mean value of Log2 of BRCA2-c.3396A/G in the lymphocytes from familial breast cancer patients was moderately higher (10%) than that in cancer-free controls [0.206 ± 0.180 (n = 37) versus 0.172 ± 0.123 (n = 31), P = 0.38; t-test] (Table 2 and Fig. 4A and B). A similar result (38% higher) was observed for Log2 (c.3396A/G) of BRCA2 in the lymphocytes of non-familial breast cancer carriers [0.267 ± 0.171 (n = 26), P = 0.03 versus control] (Table 2 and Fig. 4A and C).

Figure 3.

BRCA1 allelic expression ratios in cancer-free controls, familial and non-familial breast cancer patients. The AI assays were performed using specific primer and probe sets targeting BRCA1-c.4308T/C alleles. Log2 ratios of BRCA1-c.4308T allele to -c.4308C allele expression were presented in cancer-free controls (A), familial (B) and non-familial breast cancer patients (C). (Data expressed as Mean ± SD, n = 3; the mean value of allelic expression ratios of total normal samples has been adjusted to zero).

Figure 3.

BRCA1 allelic expression ratios in cancer-free controls, familial and non-familial breast cancer patients. The AI assays were performed using specific primer and probe sets targeting BRCA1-c.4308T/C alleles. Log2 ratios of BRCA1-c.4308T allele to -c.4308C allele expression were presented in cancer-free controls (A), familial (B) and non-familial breast cancer patients (C). (Data expressed as Mean ± SD, n = 3; the mean value of allelic expression ratios of total normal samples has been adjusted to zero).

Figure 4.

BRCA2 allelic expression ratios in cancer-free controls, familial and non-familial breast cancer patients. The AI assays were performed using specific primer and probe sets targeting BRCA2-c.3396A/G. Log2 ratios of BRCA2-c.3396A allele to -c.3396G allele expression were presented in cancer-free controls (A), familial (B) and non-familial cancer patients (C). (Data expressed as Mean ± SD, n = 3; the mean value of allelic expression ratios of total normal samples has been adjusted to zero).

Figure 4.

BRCA2 allelic expression ratios in cancer-free controls, familial and non-familial breast cancer patients. The AI assays were performed using specific primer and probe sets targeting BRCA2-c.3396A/G. Log2 ratios of BRCA2-c.3396A allele to -c.3396G allele expression were presented in cancer-free controls (A), familial (B) and non-familial cancer patients (C). (Data expressed as Mean ± SD, n = 3; the mean value of allelic expression ratios of total normal samples has been adjusted to zero).

Table 2.

Allelic imbalance in BRCA1 and BRCA2 expression

Genes Population Sample number AI (Mean ± SD)a t-test (P-value) versus cancer-free 
BRCA1 Cancer-free 40 0.211 ± 0.169  
Familial 32 0.424 ± 0.157 0.00001 
Non-familial 32 0.353 ± 0.209 0.002 
BRCA2 Cancer-free 31 0.172 ± 0.123  
Familial 37 0.206 ± 0.180 0.38 
Non-familial 26 0.267 ± 0.171 0.03 
Genes Population Sample number AI (Mean ± SD)a t-test (P-value) versus cancer-free 
BRCA1 Cancer-free 40 0.211 ± 0.169  
Familial 32 0.424 ± 0.157 0.00001 
Non-familial 32 0.353 ± 0.209 0.002 
BRCA2 Cancer-free 31 0.172 ± 0.123  
Familial 37 0.206 ± 0.180 0.38 
Non-familial 26 0.267 ± 0.171 0.03 

a To calculate the mean value of AI, all negative value of Log2 (BRCA1-c.4308T/C) and Log2 (BRCA2-c.3396A/c.3396G) in Figures 3 and 4 were changed to positive values.

Interestingly, the distribution of under-expressed alleles of BRCA1 and BRCA2 was found to be significantly different between cancer-free control and familial breast carriers, but not between cancer-free control and non-familial breast carriers. As shown in Table 3 and Figure 3, under-expressed BRCA1-c.4308T (i.e. Log2 [4308T/C] < 0) and BRCA1-c.4308C (i.e. Log2 [4308T/C] > 0) alleles were found in ∼53% (21 of 40) and ∼47% (19 of 40) of cancer-free controls as compared to ∼28% (9 of 32) and ∼72% (23 of 32) of familial breast cancer carriers, respectively (P < 0.02). In addition, under-expressed BRCA2-c.3396A [i.e. Log2 (3396A/G) < 0] and BRCA2-c.3396G [i.e. Log2 (3396A/G) > 0] alleles were found in ∼45% (14 of 31) and ∼55% (17 of 31) of cancer-free controls as compared to ∼70% (26 of 37) and ∼30% (11 of 37) of familial breast cancer carriers (P < 0.02), respectively (Table 3 and Fig. 4).

Table 3.

Distribution of under-expressed alleles of BRCA1 and BRCA2

Genes Group Under-expressed alleles OR (95% CI) P-valuea 
BRCA1  c.4308T, Log2 [4308T/C]<0 c.4308C, Log2 [4308T/C]>0   
 Cancer-free Controls 21 19   
 Familial 23 2.82 (1.05, 7.60) 0.02 
 Non-Familial 16 16 1.11 (0.44, 2.80) 0.18 
BRCA2  c.3396A, Log2 [3396A/G]<0 c3396G, Log2 [3396A/G]>0   
 Cancer-free Controls 14 17   
 Familial 26 11 0.35 0.02 
 Non-familial 13 13 0.82 0.20 
Genes Group Under-expressed alleles OR (95% CI) P-valuea 
BRCA1  c.4308T, Log2 [4308T/C]<0 c.4308C, Log2 [4308T/C]>0   
 Cancer-free Controls 21 19   
 Familial 23 2.82 (1.05, 7.60) 0.02 
 Non-Familial 16 16 1.11 (0.44, 2.80) 0.18 
BRCA2  c.3396A, Log2 [3396A/G]<0 c3396G, Log2 [3396A/G]>0   
 Cancer-free Controls 14 17   
 Familial 26 11 0.35 0.02 
 Non-familial 13 13 0.82 0.20 

aA χ2 test was used to assess the 2 by 2 tables.

Inheritance effects of AI in BRCA1

A previous study has indicated that AI for several tumor suppressor genes could be transmitted by Mendelian inheritance (19). To test if the AI observed in our study may be inherited, we identified three affected women (i.e. probands) reporting a significant family history of breast and/or ovarian cancer for which we had blood from at least one of their sisters. Furthermore, each sister had to be heterozygous for the BRCA1-c.4308T/C polymorphism. As shown in family A (Table 4), sister Sis-02 displayed a similar AI pattern as compared to the proband, while sister Sis-01 displayed no AI. In the two other families, both the affected probands and their corresponding sisters showed AI (Table 4). We further performed a haplotype analysis to determine whether the alleles showing AI were shared between siblings. As shown in Table 4, sisters with the same AI phenotype shared the same haplotype with their affected sister. Importantly, sister Sis-01 in family A did not share the same haplotype. Her blood sample displayed no AI (0.007 ± 0.147) for BRCA1 gene expressions whereas the AI was detected in her unaffected and affected sisters (Sis-02 and Proband, 0.382 ± 0.176 and 0.375 ± 0.06, respectively) (Table 4). The allele frequencies of the microsatellite markers used for haplotype construction are listed in Supplementary Material, Table S1.

Table 4.

Allelic expression and haplotype analysis of BRCA1 in sisters from three breast cancer-prone families

Family Members Allelic expression [Log2 (BRCA1-c.4308T/C)] Haplotypes D17S855–D17S1322–D17S1325 
Family A Probanda 0.375 ± 0.060 145/155–121/121–193/193 
Sis-01 0.007 ± 0.147 145/153–121/121–195/193 
Sis-02 0.382 ± 0.176 145/155–121/121–193/193 
Family B Probandb 0.477 ± 0.070 145/151–121/124–193/193 
Sis-01 0.232 ± 0.214 145/151–121/124–193/193 
Family C Probandb 0.583 ± 0.243 145/153–121/127–189/189 
Sis-01 0.522 ± 0.156 145/153–121/127–189/189 
Family Members Allelic expression [Log2 (BRCA1-c.4308T/C)] Haplotypes D17S855–D17S1322–D17S1325 
Family A Probanda 0.375 ± 0.060 145/155–121/121–193/193 
Sis-01 0.007 ± 0.147 145/153–121/121–195/193 
Sis-02 0.382 ± 0.176 145/155–121/121–193/193 
Family B Probandb 0.477 ± 0.070 145/151–121/124–193/193 
Sis-01 0.232 ± 0.214 145/151–121/124–193/193 
Family C Probandb 0.583 ± 0.243 145/153–121/127–189/189 
Sis-01 0.522 ± 0.156 145/153–121/127–189/189 

aOvarian cancer carrier.

bBreast cancer carrier.

DISCUSSION

In this study, we developed a quantitative AI assay to examine the expression difference between the alleles of BRCA1 and BRCA2 (Fig. 1). By performing this AI assay with specific primers and probes that target common single nucleotide polymorphisms in BRCA1 and in BRCA2, we were able to detect allelic imbalance associated with NMD in patients carrying frameshift mutations in BRCA1 and BRCA2 (Fig. 2). We next compared AI of BRCA1 and BRCA2 expression among three groups, familial breast cancer patients, non-familial breast cancer patients, and age-matched cancer-free females. AI ratios of BRCA1 in familial breast cancer cases were significantly higher than those from cancer-free controls (P = 0.00001) (Table 2 and Fig. 3). Similar results were observed for AI ratios of BRCA1 in the lymphocytes from non-familial breast cancer patients (P=0.002). AI ratios of BRCA2 in familial or non-familial breast cancer cases were also higher than those from cancer-free controls (P = 0.38 or P = 0.03, respectively). However, the difference was not statistically significant in the ratios of mRNA expressed from the BRCA2 alleles found in familial breast cancer cases when compared to cancer-free controls (Table 2 and Fig. 4). In addition, the distribution of under-expressed alleles between cancer-free controls and familial cases was significantly different for both BRCA1 and BRCA2 gene expression (P < 0.02 and P < 0.02, respectively) (Table 3). Furthermore, we have demonstrated that the AI patterns for BRCA1 expression, albeit in a small number of families, can be transmitted by Mendelian inheritance (Table 4). Although these findings are consistent with a previous study (19), future evaluations will benefit from evaluating AI in large families for evidence of disease segregation.

Several methods have been developed to evaluate allele-specific expression. The first method combines primer extension and capillary electrophoresis (19,21). The second approach utilizes microarray technology to measure allele-specific mRNA expression (22). Compared to the AI assay presented here, the method of primer extension plus capillary electrophoresis is also accurate but relatively time-consuming and expensive. The microarray approach provides a high-throughput and a powerful platform for the simultaneous analysis of large numbers of genes to analyze allele-specific gene expression, but it has less power to define the AI. Like the majority of allelic expression methods (23), our AI assay also requires a transcribed heterozygous variant in the individuals to be evaluated. In the present study, we targeted two common polymorphisms, BRCA1-c.4308T>C and BRCA2-c.3396A>G in the general population. Therefore, a substantial number of subjects homozygous for the polymorphisms had to be excluded. To overcome this limitation of population selection based on genotypes, other primers and probes will need to be developed to target other common polymorphisms in BRCA1 and/or BRCA2. In addition, our approach could easily be applied for studying AI in other cancer susceptibility genes, such as p53, APC, PTEN, etc.

In this study, we have demonstrated AI for both BRCA1 and BRCA2 in breast cancer populations. Interestingly, the increase of AI ratios in familial and non-familial breast cancer patients was more significant for BRCA1 than BRCA2. Loss of BRCA1 expression in breast cancer has been reported to be related to the pathogenesis of breast cancer (13–17). Loss of BRCA2 expression in cancers, in contrast, is still controversial (24,25). These findings indicate that AI in BRCA1 appears to be a more common event in breast cancer development than AI involving BRCA2. However, the mechanism(s) leading to the observed AI is for the most part unknown.

We have demonstrated that both BRCA1 and BRCA2 deleterious mutations can activate the NMD pathway and result in AI [Figure 2, and (12)]. However, all the familial breast cancer patients evaluated in the current study were determined to lack a mutation in BRCA1 and BRCA2 that would trigger NMD. Furthermore, we evaluated the BRCA1 and BRCA2 genes in the sporadic breast cancer patients and cancer-free controls demonstrating AI [i.e. allele expression ratio > 0.25 or < −0.25 (Figs 3 and 4)]. Again, no deleterious germline mutations were detected (data not shown). This is not entirely surprising given that germline mutations in BRCA1 and BRCA2 are rare in women affected with breast cancer without a strong family history of the disease (26–29).

Based on these observations, we conclude that NMD is not likely to be responsible for the observed AI in our case–control comparisons. Therefore, other mechanisms are likely to exist to account for the observed increased AI of BRCA1 and BRCA2 gene expression in female breast cancer patients. For example, the 5′ and 3′ non-coding regions of BRCA1 and BRCA2 are rarely evaluated through genetic testing, even though genetic alterations in these non-coding regions could be important in regulating BRCA1 and BRCA2 expression. For instance, genetic alterations within 5′ DNA or the putative promoter regions are able to disrupt the binding of transcription factors to DNA regulatory elements and hence lead to the loss of allelic gene expression. Several studies have shown that large genomic deletions involving the BRCA1 promoter were associated with hereditary breast cancer (30–32). This concept is further supported by studies of Cowden syndrome (CS) showing that ∼10% of CS-related PTEN mutations occur in the PTEN promoter and lead to a 50% reduction in PTEN expression (33,34). Also, allele-specific hypermethylation of the BRCA1 promoter region and decreased BRCA1 expression is associated with ∼10% of sporadic breast cancer cases (18,30,35). Recent advances have identified a new pathway for gene regulation, i.e. via microRNAs (miRNAs) (36,37). These 21–22 nt RNA molecules are complementary to the 3′-UTR sequence of transcripts and mediate negative post-transcriptional regulation through RNA duplex formation (36,38). By performing in silico analyses in four BRCA1 SNPs and two BRCA2 SNPs (39), we have identified three rare BRCA1 alleles (c.5628G, c.6273T, c.6924A) that could potentially create target sites for selected microRNAs (Supplementary Material, Table S2). Therefore, it is possible that altered mRNA targeting could contribute to AI of BRCA1 gene expression in the absence of frameshift mutations. It will be important in future studies to determine the mechanisms that either disrupt transcription factors binding or alter miRNA binding, leading to constitutively decreased levels of BRCA1 and BRCA2 and an increased risk of developing breast cancer.

In summary, we have developed a quantitative approach to evaluate expression of BRCA1 and BRCA2 from individual alleles, and we have found that AI in BRCA1 and to a lesser extent BRCA2 is associated with increased breast cancer risk. Furthermore, we have demonstrated that the AI patterns for BRCA1 expression could be transmitted by Mendelian inheritance. Since susceptibility to breast cancer is far from being fully understood, our study suggests that alternate mechanisms, other than deleterious coding mutations, may contribute to breast cancer.

MATERIALS AND METHODS

Databases

RefSeqs (GenBank Accession No: NM_007295.2 and NM_000059.1) were used for BRCA1 and BRCA2 mRNA numbering, respectively. The A of ATG translation initiation codon is defined as position +1.

Subjects and genotype analysis

Three populations were used in this study, (i) BRCA1/2 mutation-negative women reporting a personal and family history of breast cancer, i.e. familial; (ii) female breast cancer patients without a significant family history of disease, i.e. non-familial; and (iii) age-matched cancer-free female controls (Table 1). All participants were Caucasian women with European-American ancestry and were from the Delaware Valley, including the greater Philadelphia Metropolitan area in Pennsylvania. For family studies, eligible subjects were women with a personal and family history of cancer (at least two first and/or second-degree relatives affected with either breast and/or ovarian cancer) and were ascertained from the Family Risk Assessment Program (FRAP) at the Fox Chase Cancer Center (FCCC). All relevant institutional review boards approved the study protocol and written informed consent was obtained from all participants. Genotype analyses of the two common polymorphisms, BRCA1-c.4308T/C and BRCA2-c.3396A/G were carried out using ABI PRISM 7900HT Sequence Detection System and Assays-on-Demand SNP Genotyping products for fluorogenic polymerase chain reaction allelic discrimination (Applied Biosystems, Foster City, CA, USA).

Allelic imbalance assay

A 1.25 µl of the cDNA synthesized in the RT reaction was used in a real-time PCR reaction (25 µl total volume), performed with ABI PRISM 7900HT Sequence Detection System following methods recommended by the manufacturer. Optimal conditions were as follows: Step 1, 95°C for 10 min; Step 2, 92°C for 15 s, 60°C for 60 s with Optics; repeated for 40 cycles. The primer and probe sets used in real-time PCR reaction to detected BRCA1-c.4308T/C (rs1060915) and BRCA2-c.3396A/G (rs1801406) allelic expression were obtained from Applied Biosystem TaqMan® SNP Assay program (Assay ID: C.3178676 and C.7605673.1 for BRCA1 and BRCA2, respectively). Sequence information for primers and probes is available upon request. Each 96-well PCR plate included negative controls, positive controls and unknown samples. Real-time PCR data were analyzed with ABI SDS 2.2.2 software. In order to produce the BRCA1 allelic expression standard curve, cDNAs from the two samples with homozygous genotypes, BRCA1-c.4308T/T and BRCA2-c.4308C/C, were mixed as the following ratios: 8:1, 4:1, 2:1, 1:1, 1:2, 1:4 and 1:8 (c.4308T/T allele:c.4308C/T allele). For the same purpose, cDNAs from the two samples with homozygous genotypes, BRCA2-c.3396A/A and BRCA2-c.3396G/G, were mixed as the following ratios: 8:1, 4:1, 2:1, 1:1, 1:2, 1:4 and 1:8 (c.3396A/A allele:c.3396G/G allele).

The principles of quantitative real-time PCR provide the basis of this linear relation between Log2 ratio and ΔCT established in our approach to detect AI (40,41). Previous data have shown that AmpliTaq DNA polymerase cleaves the matched and well-hybridized probe and target sequences and produces a fluorescent signal (42). In contrast, mismatches between a probe and target are expected to reduce the efficiency of probe hybridization, and AmpliTaq DNA polymerase is more likely to displace a mismatched probe without cleaving it, which does not produce a fluorescent signal.

Theoretically, Allele 1 gene copy number (detected by FAM):

 
1
formula

and Allele 2 gene copy number (detected by VIC):

 
2
formula

If the fluorescence probes have the same efficiency to hybridize with matched target sequence, that is, A1 = A2 = A, therefore,

 
3
formula

The function (3) was confirmed by two standard curves, Log2 (c.4308T/C) = −0.0877 + 1.57897 * ΔCT and Log2 (c.3396A/G) = 0.11726 + 1.26458 * ΔCT, set up by our experimental data (Fig. 1). Besides using function (3) to calculate the ratio of mRNA expression between the two alleles, function (1) and function (2) are able to be applied for examining the absolute value of each allele mRNA expression. However, the direct analysis of single allele expression is often complicated by the potential variations between individuals with different environmental or physiological background rather than genetic factors. Comparing the relative expression levels of two alleles of the same gene within the same biologic sample will help to minimize these variations.

Peripheral blood lymphocytes and LCLs

Lymphocytes were isolated from peripheral blood and stored at −150°C until needed. None of the blood samples from breast cancer patients were collected at the time of chemo- or radiation therapy. In addition, a subset of cyropreserved lymphocytes from BRCA1 or BRCA2 mutation carriers (e.g. BRCA1-c.3671ins4 and BRCA2-c.796delT) or disease-free individuals were infected with EBV to establish immortal LCLs. LCLs were maintained in RPMI (GIBCO BRL) media supplemented with 20% fetal calf serum and antibiotics at 37°C, 5% CO2 atmospheric condition and 95% humidity. The immortalized LCLs from cancer-free individuals that had been tested negative for mutations in BRCA1 and BRCA2 served as wild-type controls. To prevent potential degradation of unstable transcripts by NMD a translation inhibitor, puromycin (Sigma, St Louis, MO, USA) was added to the LCL cells as described in a previous study (12).

Subcloning the PCR product and sequence analysis

PCR fragments containing a common polymorphism and deleterious mutation were subcloned directly into pCR®4-TOPO vector (Invitrogen, Carlsbad, CA, USA). PCR was then performed to identify bacterial colonies containing appropriate inserts. Plasmid DNA was purified using QIAfilter™ Plasmid Maxi Kit (Qiagen Inc., Valencia, CA, USA) and the insert was sequenced using either the universal M13-primers or the primers for PCR reactions.

RNA isolation and reverse transcription (RT)

Total cellular RNAs were isolated from blood lymphocyte pellets using TRIzol reagent according to the protocols provided by the manufacturer (Invitrogen Corp., Carlsbad, CA, USA). Purified RNAs were further processed to remove any contaminating DNA (DNA-free kit, Ambion, Inc., Houston, TX, USA). After quantification with Bioanalyzer-2100 system using RNA 6000 Nano LabChip kits (Agilent Technologies, Palo Alto, CA, USA), 2 µg of total RNA from each sample was used as a template to be reverse-transcribed (RT) in a 20 μl reaction [containing 5 µM random hexamers, 500 µM deoxynucleoside triphosphate mix, 1× RT (reverse transcriptase) buffer, 5 mM MgCl2, 1.5 units of RNase inhibitor and 7.5 units of MuLV reverse transcriptase]. All reagents were purchased from Applied Biosystems (Branchburg, NJ, USA). The RT reaction conditions were 10 min at 25°C, 1 h at 42°C and 5 min at 94°C.

Haplotype analysis

Haplotypes were constructed for BRCA1 using three polymorphic microsatellite repeat markers located within (D17S855 and D17S1322) or adjacent (D17S1325) to the BRCA1 locus. The sequences of the primer pairs were obtained from the Genome Database (http://www.gdb.org) and PCR reaction was carried out as previously reported (43,44). PCR products with fluorescent dye (HEX) labeled primer were mixed with Hi-Di Formamide and a fluorescent labeled internal size marker. The mixture was subjected to electrophoreseis on an ABI 3100 Automated DNA Sequencer (Applied Biosystems, Foster City, CA, USA) and the data were analyzed by the GeneScan (Version 3.7) and GeneMapper (Version 4.0) software provided by the manufacturer.

Statistical analysis

Allele specific real-time PCR data were analyzed with ABI SDS software v2.2.2 (Applied Biosystems, Foster City, CA, USA). Statistical analysis was conducted using the SAS System (version 9) developed by the SAS Institute, Inc. (Cary, NC, USA). Student’s t-test was employed for continuous data and results were presented as the mean ± SD. We compared the distribution of under-expressed alleles in BRCA1 or BRCA2 between cases and controls using χ2 95% confidence intervals (CI) and the difference in distribution of under-expressed alleles was estimated as odds ratios (OR). A value of P < 0.05 is considered significant.

SUPPLEMENTARY MATERIAL

Supplementary Material is available at HMG Online.

FUNDING

This work is supported in part by Cheryl Herman and the Eileen Stein-Jacoby Fund, grants from the National Cancer Institute, i.e. P50 CA83638 (A.K.G.), U01 CA69631 (M.B.D.) and R25 CA057708 (S.W.); a fellowship from the Department of Defense, W81XWH-04-1-0573 (X.C.); and DOD grants, DAMD17-03-1-0707 (A.K.G.) and DAMD17-03-1-0312 (A.K.G.); and by an appropriation from the Commonwealth of Pennsylvania.

ACKNOWLEDGEMENTS

We would like to acknowledge Dr S. Litwin for help with statistical analyses and Ms K.A. Cattie with haplotype analysis. We also thank the numerous volunteers for providing blood samples for these studies.

Conflict of Interest statement. None of the authors has any conflict of interest.

REFERENCES

1
Jemal
A.
Siegel
R.
Ward
E.
Murray
T.
Xu
J.
Thun
M.J.
Cancer statistics
CA Cancer J. Clin.
 , 
2007
, vol. 
57
 (pg. 
43
-
66
)
2
Collaborative Group on Hormonal Factors in Breast Cancer.
Familial breast cancer: collaborative reanalysis of individual data from 52 epidemiological studies including 58,209 women with breast cancer and 101,986 women without the disease
Lancet
 , 
2001
, vol. 
358
 (pg. 
1389
-
1399
)
3
Dite
G.S.
Jenkins
M.A.
Southey
M.C.
Hocking
J.S.
Giles
G.G.
McCredie
M.R.
Venter
D.J.
Hopper
J.L.
Familial risks, early-onset breast cancer, and BRCA1 and BRCA2 germline mutations
J. Natl Cancer Inst.
 , 
2003
, vol. 
95
 (pg. 
448
-
457
)
4
Bove
B.
Dunbrack
R.L.
Jr
Godwin
A.K.
Pasqualini
J.R.
Breast Cancer: Prognosis, Treatment and Prevention
 , 
2002
Marcel Dekker Inc.
(pg. 
555
-
624
)
5
King
M.C.
Marks
J.H.
Mandell
J.B.
Breast and ovarian cancer risks due to inherited mutations in BRCA1 and BRCA2
Science
 , 
2003
, vol. 
302
 (pg. 
643
-
646
)
6
McClain
M.R.
Palomaki
G.E.
Nathanson
K.L.
Haddow
J.E.
Adjusting the estimated proportion of breast cancer cases associated with BRCA1 and BRCA2 mutations: public health implications
Genet. Med.
 , 
2005
, vol. 
7
 (pg. 
28
-
33
)
7
Fan
S.
Yuan
R.
Ma
Y.X.
Meng
Q.
Goldberg
I.D.
Rosen
E.M.
Mutant BRCA1 genes antagonize phenotype of wild-type BRCA1
Oncogene
 , 
2001
, vol. 
20
 (pg. 
8215
-
8235
)
8
Holt
J.T.
Thompson
M.E.
Szabo
C.
Robinson-Benion
C.
Arteaga
C.L.
King
M.C.
Jensen
R.A.
Growth retardation and tumour inhibition by BRCA1
Nat. Genet.
 , 
1996
, vol. 
12
 (pg. 
298
-
302
)
9
Xia
F.
Taghian
D.G.
DeFrank
J.S.
Zeng
Z.C.
Willers
H.
Iliakis
G.
Powell
S.N.
Deficiency of human BRCA2 leads to impaired homologous recombination but maintains normal nonhomologous end joining
Proc. Natl Acad. Sci. USA
 , 
2001
, vol. 
98
 (pg. 
8644
-
8649
)
10
Xu
X.
Qiao
W.
Linke
S.P.
Cao
L.
Li
W.M.
Furth
P.A.
Harris
C.C.
Deng
C.X.
Genetic interactions between tumor suppressors Brca1 and p53 in apoptosis, cell cycle and tumorigenesis
Nat. Genet.
 , 
2001
, vol. 
28
 (pg. 
266
-
271
)
11
Jonkers
J.
Meuwissen
R.
van der Gulden
H.
Peterse
H.
van der Valk
M.
Berns
A.
Synergistic tumor suppressor activity of BRCA2 and p53 in a conditional mouse model for breast cancer
Nat. Genet.
 , 
2001
, vol. 
29
 (pg. 
418
-
425
)
12
Chen
X.
Truong
T.T.
Weaver
J.
Bove
B.A.
Cattie
K.
Armstrong
B.A.
Daly
M.B.
Godwin
A.K.
Intronic alterations in BRCA1 and BRCA2: effect on mRNA splicing fidelity and expression
Hum. Mutat.
 , 
2006
, vol. 
27
 (pg. 
427
-
435
)
13
Esteller
M.
Silva
J.M.
Dominguez
G.
Bonilla
F.
Matias-Guiu
X.
Lerma
E.
Bussaglia
E.
Prat
J.
Harkes
I.C.
Repasky
E.A.
, et al.  . 
Promoter hypermethylation and BRCA1 inactivation in sporadic breast and ovarian tumors
J. Natl Cancer Inst.
 , 
2000
, vol. 
92
 (pg. 
564
-
569
)
14
Lambie
H.
Miremadi
A.
Pinder
S.E.
Bell
J.A.
Wencyk
P.
Paish
E.C.
Macmillan
R.D.
Ellis
I.O.
Prognostic significance of BRCA1 expression in sporadic breast carcinomas
J. Pathol.
 , 
2003
, vol. 
200
 (pg. 
207
-
213
)
15
Russell
P.A.
Pharoah
P.D.
De Foy
K.
Ramus
S.J.
Symmonds
I.
Wilson
A.
Scott
I.
Ponder
B.A.
Gayther
S.A.
Frequent loss of BRCA1 mRNA and protein expression in sporadic ovarian cancers
Int. J. Cancer
 , 
2000
, vol. 
87
 (pg. 
317
-
321
)
16
Wang
C.
Horiuchi
A.
Imai
T.
Ohira
S.
Itoh
K.
Nikaido
T.
Katsuyama
Y.
Konishi
I.
Expression of BRCA1 protein in benign, borderline, and malignant epithelial ovarian neoplasms and its relationship to methylation and allelic loss of the BRCA1 gene
J. Pathol.
 , 
2004
, vol. 
202
 (pg. 
215
-
223
)
17
Thrall
M.
Gallion
H.H.
Kryscio
R.
Kapali
M.
Armstrong
D.K.
DeLoia
J.A.
BRCA1 expression in a large series of sporadic ovarian carcinomas: a Gynecologic Oncology Group study
Int. J. Gynecol. Cancer
 , 
2006
, vol. 
16
 
Suppl. 1
(pg. 
166
-
171
)
18
Baldwin
R.L.
Nemeth
E.
Tran
H.
Shvartsman
H.
Cass
I.
Narod
S.
Karlan
B.Y.
BRCA1 promoter region hypermethylation in ovarian carcinoma: a population-based study
Cancer Res.
 , 
2000
, vol. 
60
 (pg. 
5329
-
5333
)
19
Yan
H.
Yuan
W.
Velculescu
V.E.
Vogelstein
B.
Kinzler
K.W.
Allelic variation in human gene expression
Science
 , 
2002
, vol. 
297
 pg. 
1143
 
20
Yan
H.
Dobbie
Z.
Gruber
S.B.
Markowitz
S.
Romans
K.
Giardiello
F.M.
Kinzler
K.W.
Vogelstein
B.
Small changes in expression affect predisposition to tumorigenesis
Nat. Genet.
 , 
2002
, vol. 
30
 (pg. 
25
-
26
)
21
Matyas
G.
Giunta
C.
Steinmann
B.
Hossle
J.P.
Hellwig
R.
Quantification of single nucleotide polymorphisms: a novel method that combines primer extension assay and capillary electrophoresis
Hum. Mutat.
 , 
2002
, vol. 
19
 (pg. 
58
-
68
)
22
Lo
H.S.
Wang
Z.
Hu
Y.
Yang
H.H.
Gere
S.
Buetow
K.H.
Lee
M.P.
Allelic variation in gene expression is common in the human genome
Genome. Res.
 , 
2003
, vol. 
13
 (pg. 
1855
-
1862
)
23
Buckland
P.R.
Allele-specific gene expression differences in humans
Hum. Mol. Genet.
 , 
2004
, vol. 
13
 
Spec no. 2
(pg. 
R255
-
R260
)
24
Bieche
I.
Nogues
C.
Lidereau
R.
Overexpression of BRCA2 gene in sporadic breast tumours
Oncogene
 , 
1999
, vol. 
18
 (pg. 
5232
-
5238
)
25
Hilton
J.L.
Geisler
J.P.
Rathe
J.A.
Hattermann-Zogg
M.A.
DeYoung
B.
Buller
R.E.
Inactivation of BRCA1 and BRCA2 in ovarian cancer
J. Natl Cancer Inst.
 , 
2002
, vol. 
94
 (pg. 
1396
-
1406
)
26
Futreal
P.A.
Liu
Q.
Shattuck-Eidens
D.
Cochran
C.
Harshman
K.
Tavtigian
S.
Bennett
L.M.
Haugen-Strano
A.
Swensen
J.
Miki
Y.
, et al.  . 
BRCA1 mutations in primary breast and ovarian carcinomas. Science
1994
, vol. 
266
 (pg. 
120
-
122
)
27
Lancaster
J.M.
Wooster
R.
Mangion
J.
Phelan
C.M.
Cochran
C.
Gumbs
C.
Seal
S.
Barfoot
R.
Collins
N.
Bignell
G.
, et al.  . 
BRCA2 mutations in primary breast and ovarian cancers
Nat. Genet.
 , 
1996
, vol. 
13
 (pg. 
238
-
240
)
28
Newman
B.
Mu
H.
Butler
L.M.
Millikan
R.C.
Moorman
P.G.
King
M.C.
Frequency of breast cancer attributable to BRCA1 in a population-based series of American women
JAMA
 , 
1998
, vol. 
279
 (pg. 
915
-
921
)
29
Southey
M.C.
Tesoriero
A.A.
Andersen
C.R.
Jennings
K.M.
Brown
S.M.
Dite
G.S.
Jenkins
M.A.
Osborne
R.H.
Maskiell
J.A.
Porter
L.
, et al.  . 
BRCA1 mutations and other sequence variants in a population-based sample of Australian women with breast cancer
Br. J. Cancer
 , 
1999
, vol. 
79
 (pg. 
34
-
39
)
30
Swensen
J.
Hoffman
M.
Skolnick
M.H.
Neuhausen
S.L.
Identification of a 14 kb deletion involving the promoter region of BRCA1 in a breast cancer family
Hum. Mol. Genet.
 , 
1997
, vol. 
6
 (pg. 
1513
-
1517
)
31
Smith
L.D.
Tesoriero
A.A.
Ramus
S.J.
Dite
G.
Royce
S.G.
Giles
G.G.
McCredie
M.R.
Hopper
J.L.
Southey
M.C.
BRCA1 promoter deletions in young women with breast cancer and a strong family history: a population-based study
Eur. J. Cancer
 , 
2007
, vol. 
43
 (pg. 
823
-
827
)
32
Brown
M.A.
Lo
L.J.
Catteau
A.
Xu
C.F.
Lindeman
G.J.
Hodgson
S.
Solomon
E.
Germline BRCA1 promoter deletions in UK and Australian familial breast cancer patients: identification of a novel deletion consistent with BRCA1:psiBRCA1 recombination
Hum. Mutat.
 , 
2002
, vol. 
19
 (pg. 
435
-
442
)
33
Zhou
X.P.
Waite
K.A.
Pilarski
R.
Hampel
H.
Fernandez
M.J.
Bos
C.
Dasouki
M.
Feldman
G.L.
Greenberg
L.A.
Ivanovich
J.
, et al.  . 
Germline PTEN promoter mutations and deletions in Cowden/Bannayan-Riley-Ruvalcaba syndrome result in aberrant PTEN protein and dysregulation of the phosphoinositol-3-kinase/Akt pathway
Am. J. Hum. Genet.
 , 
2003
, vol. 
73
 (pg. 
404
-
411
)
34
Pezzolesi
M.G.
Zbuk
K.M.
Waite
K.A.
Eng
C.
Comparative genomic and functional analyses reveal a novel cis-acting PTEN regulatory element as a highly conserved functional E-box motif deleted in Cowden syndrome
Hum. Mol. Genet.
 , 
2007
, vol. 
16
 (pg. 
1058
-
1071
)
35
Signori
E.
Bagni
C.
Papa
S.
Primerano
B.
Rinaldi
M.
Amaldi
F.
Fazio
V.M.
A somatic mutation in the 5′-UTR of BRCA1 gene in sporadic breast cancer causes down-modulation of translation efficiency
Oncogene
 , 
2001
, vol. 
20
 (pg. 
4596
-
4600
)
36
Esquela-Kerscher
A.
Slack
F.J.
Oncomirs–microRNAs with a role in cancer
Nat. Rev. Cancer
 , 
2006
, vol. 
6
 (pg. 
259
-
269
)
37
Lai
E.C.
Micro RNAs are complementary to 3′-UTR sequence motifs that mediate negative post-transcriptional regulation
Nat. Genet.
 , 
2002
, vol. 
30
 (pg. 
363
-
364
)
38
Wu
L.
Fan
J.
Belasco
J.G.
MicroRNAs direct rapid deadenylation of mRNA
Proc. Natl Acad. Sci. USA
 , 
2006
, vol. 
103
 (pg. 
4034
-
4039
)
39
Rusinov
V.
Baev
V.
Minkov
I.N.
Tabler
M.
MicroInspector: a web tool for detection of miRNA binding sites in an RNA sequence
Nucleic Acids Res.
 , 
2005
, vol. 
33
 (pg. 
W696
-
W700
)
40
Chiang
P.W.
Song
W.J.
Wu
K.Y.
Korenberg
J.R.
Fogel
E.J.
Van Keuren
M.L.
Lashkari
D.
Kurnit
D.M.
Use of a fluorescent-PCR reaction to detect genomic sequence copy number and transcriptional abundance
Genome Res.
 , 
1996
, vol. 
6
 (pg. 
1013
-
1026
)
41
Heid
C.A.
Stevens
J.
Livak
K.J.
Williams
P.M.
Real time quantitative PCR
Genome Res.
 , 
1996
, vol. 
6
 (pg. 
986
-
994
)
42
Livak
K.J.
Flood
S.J.
Marmaro
J.
Giusti
W.
Deetz
K.
Oligonucleotides with fluorescent dyes at opposite ends provide a quenched probe system useful for detecting PCR product and nucleic acid hybridization
PCR Methods Appl.
 , 
1995
, vol. 
4
 (pg. 
357
-
362
)
43
Weitzel
J.N.
Lagos
V.
Blazer
K.R.
Nelson
R.
Ricker
C.
Herzog
J.
McGuire
C.
Neuhausen
S.
Prevalence of BRCA mutations and founder effect in high-risk Hispanic families
Cancer Epidemiol. Biomarkers Prev.
 , 
2005
, vol. 
14
 (pg. 
1666
-
1671
)
44
Rudkin
T.M.
Hamel
N.
Galvez
M.
Hogervorst
F.
Gille
J.J.
Moller
P.
Apold
J.
Foulkes
W.D.
The frequent BRCA1 mutation 1135insA has multiple origins: a haplotype study in different populations
BMC Med. Genet.
 , 
2006
, vol. 
7
 pg. 
15