c-MYC is a radiosensitive locus in human breast cells

Abstract

Ionising radiation is a potent human carcinogen. Epidemiological studies have shown that adolescent and young women are at increased risk of developing breast cancer following exposure to ionising radiation compared with older women, and that risk is dose-dependent. Although it is well understood which individuals are at risk of radiation-induced breast carcinogenesis, the molecular genetic mechanisms that underlie cell transformation are less clear. To identify genetic alterations potentially responsible for driving radiogenic breast transformation, we exposed the human breast epithelial cell line MCF-10A to fractionated doses of X-rays and examined the copy number and cytogenetic alterations. We identified numerous alterations of c-MYC that included high-level focal amplification associated with increased protein expression. c-MYC amplification was also observed in primary human mammary epithelial cells following exposure to radiation. We also demonstrate that the frequency and magnitude of c-MYC amplification and c-MYC protein expression is significantly higher in breast cancer with antecedent radiation exposure compared with breast cancer without a radiation aetiology. Our data also demonstrate extensive intratumor heterogeneity with respect to c-MYC copy number in radiogenic breast cancer, suggesting continuous evolution at this locus during disease development and progression. Taken together, these data identify c-MYC as a radiosensitive locus, implicating this oncogenic transcription factor in the aetiology of radiogenic breast cancer.

Figure 1

SNP6.0 and Cytoscan array copy number profile of chromosome 8 in parental and irradiated MCF-10A cell populations. Parental MCF-10A and MCF-10A cells irradiated with fractionated X-ray doses of 5 Gy to a cumulative dose of 40, 60 and 80 Gy were assessed by SNP6.0 array (a). Each SNP marker on chromosome 8 is represented and aligned to its position on chromosome 8 as well as its designated copy number state. An ideogram of chromosome 8 is positioned below the SNP marker plots. A 2.5 Mb copy number gain was identified in the 40 Gγ population and spans the c-MYC locus, increasing its copy number state from 3 to 4. The position of c-MYC is highlighted on each SNP marker plot and the chromosome 8 ideogram by a red arrow. An ~59 Mb copy number gain was identified in the 60 Gγ population, which spanned a number of regions with different constitutive copy number states and also encompassed the c-MYC locus, therefore increasing its copy number state further. The ~59 Mb copy number gain first identified in the 60 Gy population was further pronounced in the 80 Gy population. The ~59 Mb region is indicated by the horizontal black line above the SNP marker plots of the 60 and 80 Gy population. Clones from MCF-10A cells irradiated with fractionated X-ray doses of 5 Gy to a cumulative dose of 80 Gy were assessed by Cytoscan array (b). All 15 clones analysed carried the focal 2.5Mb amplification encompassing c-MYC. Twelve of the 15 clones carried the additional ~59 Mb copy number gain discernible from SNP6.0 data and a deletion affecting the p-arm of chromosome 8. Three of the 15 clones did not carry the 59 Mb copy number gain or the p-arm deletion, but carried a novel copy number gain also encompassing the c-MYC gene.

Figure 2

Alterations that affect c-MYC in MCF-10A cells. Partial karyotype analyses of chromosomes 8 and FISH analysis for c-MYC (red probe) and chromosome 8 centromere (green probe) on 4′,6-diamidino-2-phenylindole (DAPI) counterstained metaphase nuclei in parental MCF-10A (a) and 80 Gy cumulative dose cells (b). Chromosome arms are labelled on the ideograms adjacent to the karyotype images. The 46 Mb region of chromosome arm 8q gain present in parental MCF-10A cells is due to a duplication of 8qter-q22 and subsequent translocation to the end of the short arm resulting in a derivative chromosome 8: der(8)t(8;8)(q22;p23). The der(8)t(8;8)(q22;p23) is identified in both the parental and 80 Gy MCF-10A cell populations. The 80 Gy population has a second abnormal chromosome 8, which comprises a tandem duplication of the 8q12–q24 region to the q-telomere of the constitutively normal chromosome 8: dup(8)(q12–q24). FISH analysis confirmed that c-MYC is present in both the constitutive 46 Mb region of gain on der(8)t(8;8)(q22;p23) and the duplicated region on dup(8)(q12–q24) identified in the 80 Gγ population. The magnified view of the acquired dup(8)(q12–q24) chromosome in the 80 Gγ population (inset in b) shows that multiple copies of c-MYC are present at both expected 8q24 chromosome positions.

Figure 3

Genotypic and phenotypic alterations of c-MYC in irradiated huMECs and MCF-10A cells. c-MYC interphase FISH copy number analysis of parental and irradiated MCF-10A populations. Three main cell populations were identified by FISH: cells with two copies of chromosome 8 centromere (green probe) and three copies of c-MYC (red probe), cells with two copies of chromosome 8 and four copies of c-MYC and cells with two copies of chromosome 8 and over four copies of c-MYC (a). The proportion of 100 scored nuclei with these three c-MYC genotypes was combined and determined for each population. The proportion of nuclei with ⩾4 copies of c-MYC and therefore any cell population with a c-MYC copy number gain was determined (b). Aliquots of 10 μg of protein extracted from parental and irradiated MCF-10A cell populations were electrophoresed on polyacrylamide gels and analysed for c-MYC and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) expression by western transfer analysis as described in the Materials and methods (c). c-MYC expression was quantified for each cell population by densitometric analysis of western blots from three independent protein samples (d). c-MYC expression is expressed as a percentage of c-MYC expression in parental MCF-10A, which is set at 100% expression. Expression of c-MYC was significantly higher in the 60 Gy population than parental MCF-10A (Turkey's test; P<0.05). c-MYC interphase FISH copy number analysis of parental and irradiated HuMECs (e). Four main cell populations were identified by FISH: diploid cells (two copies of chromosome 8 centromere (aqua probe), two copies of IGH (green probe) and two copies of c-MYC (red probe)); triploid cells (three copies of each locus); tetraploid cells (four copies of each locus); cells with amplification of c-MYC. At least 70 interphase cells were counted at each radiation dose (mock-treated, 2, 3 and 4 Gy) and example images are shown for diploid, tetraploid and c-MYC-amplified cells.

Figure 4

FISH analysis of c-MYC copy number and chromosome 8 copy number in sporadic and radiogenic breast cancers. c-MYC copy number was assessed in tumour cells from radiogenic cancers (n=9) and from sporadic cancers (n=20). c-MYC copy number was higher in the radiogenic cancers (Mann–Whitney U-test; P=0.027) (a). The wider horizontal bars represent the median c-MYC copy number and the narrower horizontal bars represent the 10th and 90th percentiles of the data. The percentage of samples in which at least 10% of the nuclei contained ⩾3, 4, 5 and 6 copies of c-MYC was compared (b) and was significantly higher in the radiogenic cohort than in the sporadic cohort for ⩾6 copies of c-MYC (Fisher's exact test; P=0.022 (*)). The ratio between c-MYC and chromosome 8 centromere copy number was higher in the radiogenic cohort than in the sporadic cohort (Mann–Whitney U-test; P=0.016) (c). The percentage of samples in each cohort, which had a c-MYC to chromosome 8 centromere ratio ⩽1.10 (no or little evidence of c-MYC amplification) or >1.10 (evidence of c-MYC amplification) was also compared (d) (Fisher's exact test; P=0.010).

Figure 5

Dual-stain FISH analysis of samples that had a c-MYC to chromosome 8 centromere ratio >1.25. Representative FISH images of samples SPO2 (ratio=1.90), SPO28 (ratio=1.28), RAD9 (ratio=2.02) and RAD 10 (ratio=2.50) for combined c-MYC (red) and chromosome 8 centromere (green) hybridisation (a). The results show that c-MYC-amplified cell populations from the radiogenic cohort have a higher degree of amplification compared with amplified cell populations from the sporadic cohort. The heterogeneity of c-MYC and chromosome 8 centromere copy number status is shown in the histograms to the right of the images for each sample (b).

Figure 6

c-MYC protein expression in radiogenic and sporadic breast cancers. Formalin-fixed, paraffin-embedded tissue from sporadic (n=33) and radiogenic (n=18) breast cancers were sectioned and analysed by immunohistochemistry with a specific c-MYC antibody as described in the Materials and methods. The level of c-MYC expression was quantified by derivation of a histoscore. c-MYC expression in the sporadic and radiogenic breast cancers was compared for all cases (a) and was higher in radiogenic cases in which more than 10% of the nuclei had detectable c-MYC expression (Mann–Whitney U-test: P<0.001) (b). The proportion of sporadic and radiogenic breast cancer that had c-MYC expression histoscores of 0–50, 51–100 and >100 were compared (χ²: P=0.018) (c). The number of samples in each group in panel (c) is identified above each bar of the histogram. c-MYC expression is shown in tumours known to have a c-MYC copy number of <3 (n=22) or ⩾3 (n=7) (d). The wider horizontal bars represent the median c-MYC copy numbers and the narrower horizontal bars represent the 10th and 90th percentiles of the data (a, b and d). Representative IHC images of samples with no c-MYC expression (RAD13: histoscore=0), moderate expression (RAD9: histoscore=70.4) and high expression (RAD6; histoscore=158.8) are shown in (e).