Cancer Cell Biology
Allelic imbalance at 11p15.5–15.4 correlated with c-Ha-ras mutation during radiation-induced neoplastic transformation of human breast epithelial cells
Article first published online: 12 DEC 2002
Copyright © 2002 Wiley-Liss, Inc.
International Journal of Cancer
Volume 103, Issue 6, pages 730–737, 1 March 2003
How to Cite
Roy, D., Calaf, G. and Hei, T. K. (2003), Allelic imbalance at 11p15.5–15.4 correlated with c-Ha-ras mutation during radiation-induced neoplastic transformation of human breast epithelial cells. Int. J. Cancer, 103: 730–737. doi: 10.1002/ijc.10895
- Issue published online: 3 JAN 2003
- Article first published online: 12 DEC 2002
- Manuscript Accepted: 21 OCT 2002
- Manuscript Revised: 16 OCT 2002
- Manuscript Received: 24 MAY 2002
- National Institute of Health. Grant Numbers: CA 49062, ES 07890
- Environmental Health Center. Grant Number: P30 ES 09089
- AVON Products Foundation Breast Cancer Center. Grant Number: CU 51470301
- microsatellite instability;
- loss of heterozygosity;
- high-LET radiation;
- polymerase chain reaction;
- single strand conformation polymorphism
Breast cancer is the most frequent malignancy in women throughout much of the developed world and is associated with a multistage process involving a number of genetic mutations and their corresponding cellular phenotypic alterations. It has already been shown that neoplastic transformation of a spontaneously immortalized human breast epithelial (MCF-10F) cell line by radiation, in combination with estrogen, represents a successful model in studying the molecular and biological alterations that may contribute to the tumorigenic process. In the present study, the incidence of allelic alterations (microsatellite instability/loss of heterozygosity) on chromosome 11 in different radiation-induced primary and secondary tumorigenic cell lines, relative to the control MCF-10F cells was investigated. We identified 3 regions of the chromosome 11 (11p15–p15.5, 11q13 and 11q23) that showed high incidence of LOH among these tumor cell lines and suggested a potential role for these chromosomal regions in breast carcinogenesis. Among them, locus 11p15.5, where c-Ha-ras oncogene is located, had incidence of allelic imbalance between 25–40%. Furthermore, direct sequencing analysis of codons 12 and 61 of the c-Ha-ras oncogene identified various point mutations. These data highlight the importance of chromosome 11 in radiation induced malignant transformation of human breast epithelial cells and suggest the usefulness of the model in uncovering specific derangements during breast cancer progression. © 2002 Wiley-Liss, Inc.
Breast cancer is the most frequent malignancy in women with a cumulative lifetime risk that has been estimated to be about 10–12% among the general populations. Hereditary factors, on the other hand, account for about 5–10% of all cases.1 It is a heterogeneous disease and a variety of molecular alterations have been identified during the progression of this disease.2 A better understanding of the mechanisms related to initiation and progression of this disease may ultimately help to control it.
In vitro neoplastic transformation of human breast epithelial cells (HBEC) provides a useful model in studying the molecular and biological alterations that contribute to the tumorigenic process.3 One type of genetic alteration common to many tumor types is allelic imbalance in the form of either loss of heterozygosity (LOH) or microsatellite instability (MSI), which often seems to unmask recessive mutations in tumor suppressor loci.4 Frequent allelic losses detected on 1p, 3p, 6q, 8p, 11, 13q, 16q, 17 and 18q chromosomal regions in breast cancers have suggested that a variety of tumor suppressor genes may be involved in their developments.5, 6
Allelic imbalance is probably the most common genetic factor associated with cancer.7 It remains unclear whether allelic imbalance is the cause or the result of carcinogenesis. Nevertheless, it is a direct indication of genomic instability of cancer cells. Among the genomic regions commonly undergoing LOH in breast tumors, chromosome 11 is unique because at least 3 separate regions of LOH have been consistently identified (11p15–p15.5, 11q13 and 11q23), pointing to a potentially complicated role of this chromosome in breast carcinogenesis.8, 9
Human chromosome 11p and 11q are the sites of several known and suspected oncogenes and tumor suppressor genes that have been identified in various human solid tumors, including malignant melanoma and carcinomas of the breast, cervix, ovary and lung.10, 11, 12, 13 Among them, chromosome 11p15.5 has attracted considerable attention due to its linkage to various human diseases. Apart from being an important tumor suppressor locus showing LOH in different cancers, 11p15.5 has been shown by linkage analysis to harbor the gene(s) for the Beckwith-Wiedemann (B-W) syndrome.14 Study of normal breast epithelium adjacent to the tumor cells has found LOH and loss of genes at 11p15.5, which is believed to be a type of genetic shunt to invasive disease from morphologically normal breast epithelium.15 Furthermore, the clustering of known imprinted genes in the 11p15.5 region suggests that the target gene(s) for breast cancer may also be imprinted.16 Microcell-mediated transfer of human chromosome 11 into tumor cell lines have provided additional evidence of the presence of tumor suppressor gene on chromosome 11 in melanoma, breast cancer and cervical cancer.17
The critical region of 11p15 in breast tumors lies between the genetic markers TH and HBB at 11p15.5, in a segment that spans a maximum of 3–4 Mb of DNA. Other genetic markers such as D11S1318 (11p15), D11S1323 (11p15.4), D11S1338 (11p15.5), D11S4046 (11p15.5), D11S4088 (11p15.5) and HRAS1 (11p15.5) also show alterations at different stages of breast cancer progression. Frequent LOH has previously been reported on chromosome 11p15.5 in 20–30% of breast carcinomas and correlated with poor prognosis and tumor progression in this disease.18
The c-Ha-ras oncogene, mapped to 11p15.5, acquires transforming capacity either by single point mutation(s) in the codon 12 or 61, resulting in the expression of an aberrant gene product, or by over-expression of the normal c-Ha-ras p21 protein.19, 20 Activated Ki-, c-Ha- and N-ras, all with a point mutation(s) in codon 12, 13 and 61 but 12 and 61 mutation(s) are more pronounced compared to codon 13.19 Approximately 60–70% of primary human breast carcinomas exhibit over-expression whereas about 30% of the cases are heterozygous for the c-Ha-ras oncogene.21 Studies in rats have implicated c-Ha-ras mutations in the etiology of breast adenocarcinomas.22 In addition, DNA transfection studies have suggested that an activated c-Ha-ras oncogene can convert human breast cancer cells to a more aggressive, estrogen-independent phenotype.23 These findings are consistent with the observation that several human breast cancer cell lines have been shown to contain mutationally activated c-Ha-ras oncogenes.24 The mechanisms responsible for the tissue specific c-Ha-ras gene activation are not well understood, although its high frequency in various cancers suggests its crucial role in their pathogenesis.25, 26 Various preclinical studies have also shown that activated c-Ha-ras oncogene increases cellular radioresistance.27, 28
It is therefore important to investigate the involvement of this specific region (11p15) and specific gene (c-Ha-ras) on chromosome 11 during radiation-induced breast cancer progression by utilizing this unique in vitro model system.29 Highly polymorphic (CA)n repeat microsatellite markers were utilized to determine the incidence of LOH / MSI to a more refined position of chromosome 11 and, at the same time, association of c-Ha-ras oncogene mutation was studied by amplifying the gene at codon 12 and 61. This process may be an efficient way to uncover clues to the specific molecular derangements that contribute to cancer progression. It also helps to identify the appropriate cellular and molecular changes associated with radiation-induced breast carcinogenesis.
MATERIAL AND METHODS
The recently established radiation-induced breast carcinogenesis model based on MCF-10F cell lines was used in these studies.29 The spontaneously immortalized human breast epithelial cell line MCF-10F was derived from mortal human breast epithelial cell line MCF-10M and has a near diploid karyotype and is of luminal epithelial origin.30 These cells retain all the characteristics of normal epithelium in vitro, including anchorage dependence, non-invasiveness and non-tumorigenicity in the nude mice.30, 31 Cell lines were cultured on Dulbecco's modified Eagle's media (DMEM) as described previously.29, 30, 31, 32, 33 From such a model, the following cell lines were used as control: MCF-10F cell line; MCF-10F cell line treated with 17β-estradiol (E) (10−8 M) (Sigma Chemical Co., St. Louis, MO), named MCF-10F + E and MCF-7 (as positive control).29, 33 The experimental cell lines used in our study were as follows: The MCF-10F cell line irradiated with a single dose of 60cGy of α particle, named 60cGy (passage 45–50) that was anchorage-dependent and non-tumorigenic in nude mice.29 This cell line was also grown in presence of estrogen after irradiation, named 60cGy + E (Passage 48). Another MCF-10F cell line exposed to double doses of 60cGy of α particle, named 60cGy/60cGy (Passage 44–48 as an early passage and Passage 90–95 as a late passage) was also examined. It was anchorage-independent but non-tumorigenic in nude mice.29 This cell line was grown in the presence of estrogen after double doses of irradiation, named 60cGy/60cGy + E (Passage 41). A final group of MCF-10F cell line was subjected to a double dose of 60cGy of α particle and treated with E before each radiation exposure; this line was named 60cGy + E/60cGy + E (Passage 45) and was both anchorage-independent and produced tumors in 3 of 6 nude mice injected.29 Tumor-2 (Passage 30),34 1 of the 3 primary tumor cell line derived from the previous tumorigenic cell line was also used for our study.
All cell cultures were treated with 1 ml of lysis buffer (100 mM NaCl, 20 mM Tris-HCl pH 8.0, 25 mM EDTA pH 8.0, 0.5% sodium dodecyl sulfate) with 200 mg/ml of proteinase K and RNAse (100 μg/ml) and incubated overnight at 37°C with constant gentle agitation.35 They were purified in 2 extractions with a phenol:chloroform (1:1) mixture and the aqueous layer was adjusted to 0.75 M ammonium acetate and DNA was spooled from 2 vol of 100% ethanol, dried and dissolved in TE buffer (10 mM Tris-HCl pH 8.0, 1 mM EDTA pH 8.0) as described.36
Microsatellite polymorphic marker selection
Eight polymorphic dinucleotide (CA)n repeat microsatellite markers (Table I) (Research Genetics, Huntsville, AL) from chromosome 11p15.5 (Fig. 1) were selected. They were selected on the basis of their maximum heterozygosity (more than 0.70) and their location near mapped known tumor suppressor genes, oncogenes, or other cancer related genes and in regions or near loci associated with cell-cycle regulation, DNA replication, DNA repair or signal transduction protein genes. The sequences and characteristics of microsatellite oligonucleotide primers were obtained from the GDB database (http://www.gdb.org) (Table I). We also tested D2S123 (2p16, 0.77, Dinucleotide, 197–227 bp), a CA repeat marker linked to the HMSH2 gene, mapped at 2p16, where LOH is rarely encountered (data not shown).
|Locus||Map position||Type of sequence||Size range (base pairs)||Primer sequence Sense (5′ 3)/Antisense (5′ 3′)||Important genes in these markers1|
|D11S4046||11p15.5||Dinucleotide||183–203||ACTCCAGCCTGGGAAAC/TGATAGACACACCCATTGC||LOH in breast cancer|
|TH||11p15.5||Dinucleotide||244–260||CAGCTGCCCTAGTCAGCAC/GCTTCCGAGTGCAGGTCACA||Tsg101 gene, LOH in breast|
|D11S1318||11p15.5||Dinucleotide||123–145||CCCGTATGGCAACAGG/TGTGCATGTACATGAGTG||Tsg in breast, B-W syndrome|
|D11S1338||11p15.5||Dinucleotide||255–265||GACGGTTTAACTGTATATCTAAGAC/TAATGCTACTTATTTGGAGTGTG||Tsg in breast|
|HBB||11p15.4||Dinucleotide||141–188||GTACGGCTGTCATCACTTAGACCTCA/AGCACACAGACCAGCACGTT||Tsg in breast, Gap junctn-communication|
|D11S1323||11p15.4||Dinucleotide||201–207||TGCTGCTTAGAATGAGTAGATGTC/CTCTATGAAGTTGGAGTCTAGGTTG||Tsg in breast, WT2 gene|
|D11S4088||11p15.5-p15.4||Dinucleotide||204–252||GGGCAGAGGCAGTGGAG/GCATGTTTCGGGGGTG||Tsg in breast|
PCR-single strand conformation polymorphism analysis
PCR-single strand conformation polymorphism (SSCP) was carried out in a total volume of 30 μl containing 50–100 ng of genomic DNA, 1.5 mM MgCl2, 50 mM KCl, 10 mM Tris-HCl (pH 8.3), 200 μM of each dNTPs, 0.8 μM of each primer (Research Genetics, Huntsville, AL) and 0.75 U of AmpliTaq polymerase (Perkin-Elmer Corp., Foster City, CA).37 One of the primer was 5′-end-labeled with [γ-32P] ATP at 3,000 Ci/mmol (Amersham Pharmacia Biotech., IL) by T4-polynucleotide kinase (Amersham Life Sc., IL). After a 5-min pre-incubation period at 94°C, DNA was amplified for 35 cycles comprising of 45 sec at 94°C, 45 sec at 55°C and 1 min at 72°C, followed by a 7 min final extension at 72°C using the GeneAmp® PCR System 2400 (Perkin-Elmer Corp.). The PCR products were processed by diluting 1:1 in denaturing loading buffer (95% formamide, 20 mM EDTA, 0.05% xylene cyanol FF and 0.05% bromophenol blue); denatured at 95°C for 5 min and then frozen at 4°C. About 2 μl of aliquot was loaded and electrophoresed in 6% polyacrylamide gels containing 8.3 M urea for 2–3 hr at 40 W. The gel was fixed in 10% methanol-10% acetic acid, dried and exposed to Kodak X-OMAT-AR film (Eastman Kodak Co., Rochester, NY) at −70°C with intensifying screen for 12–16 hr. PCR reaction was always repeated 2–3 times with different adjacent passages of cells to get the consistent results.
Assessment of allelic losses
MSI and LOH were screened by PCR amplification of polymorphic microsatellite markers. MSI was defined as a shift of specific allelic band or a change (increase or decrease) in the broadness of specific allelic band in the autoradiograms, whereas LOH was defined as a total loss (complete deletion) or a 50% or more reduction (in signal density) in one of the heterozygous allele in autoradiograms. It was first scored by visual inspection of autoradiograms and then band intensity was quantitated in a Densitometric scanner (Model 300A, Molecular Dynamics) by using Bio image software Image Quant (version 3.3; Molecular Dynamics). The optical density range of 0.01–4.0 was chosen in O.D. units whereas resolution (spatial) selected at 100 points/cm in both X and Y direction. The resolution (signal) was selected at 4,096 levels (12-bit) of optical density.
c-Ha-Ras Gene Amplification by PCR
To determine whether radiation and estrogen treatment resulted in point mutations in c-Ha-ras codons 12 and 61, MCF-10F control, irradiated, tumorigenic and tumor cell lines were studied by direct sequencing of the amplified c-Ha-ras oncogene. The Exon 1 of the c-Ha-ras oncogene was amplified with forward (5′-CGA TGA CGG AAT ATA AGC TTG TGG TGGT-3′) and reverse (5′-GTT CAC CTG TAC TGG TGG AAT TCC TCA AA-3′) primers (Life Technologies, Inc., Grand Island, NY) to obtain the 545-bp PCR product for codon 12. Similarly, exon 2 was amplified with forward (5′-ACG CCT GTC CTC CTG CAA GCT TCC TAC-3′) and same reverse primers (Life Technologies, Inc.) to produce the 202-bp PCR product for codon 61. DNA amplifications were carried out in a PCR reaction volume of 30 μl containing 50–100 ng of genomic DNA, 1.5 mM MgCl2, 50 mM KCl, 10 mM Tris-HCl (pH 8.3), 200 μM of each dNTPs, 0.8 μM of each primer (Life Technologies, Inc.) and 0.75 U of AmpliTaq polymerase (Perkin-Elmer Corp.). The genomic DNA was initially denatured for a 5-min pre-incubation period at 95°C and thereafter subjected to 35 cycles comprising of denaturation at 94°C for 45 sec, annealing at 60°C for 1 min and extension for 1 min at 65°C, followed by a 7-min final extension at 65°C using the GeneAmp® PCR System 2400 (Perkin-Elmer Corp.).38
Direct sequencing of amplified DNA fragments
Sequencing was carried out on original PCR products amplified by using the same c-Ha-ras codon 12 and 61 primers. PCR products were gel purified by electrophoresis on 1% Agarose-TAE gel (Life Technologies Inc.) and eluted with 100 μl of elution buffer of QIAquick gel extraction kit (Qiagen Inc., CA). Sequencing was done by using automated sequencer ABI PRISM 3100 Genetic Analyzer (Applied Biosystems/Hitachi, Foster City, CA). Each and every fragment was sequenced at least three times to rule out contamination and PCR fidelity artifacts.
Northern blot analysis
Total RNA was extracted by TRIZOL reagent (Life Technologies, Inc.) from all the cultured (control, irradiated, tumorigenic and tumor) cell lines and treated with DNAse I (Boehringer-Mannheim, Indianapolis, IN) according to the manufacturer's instruction.36 Ten microliters of total RNA was electrophoresed in a 1% (w/v) agarose-formaldehyde gel and transferred to a nylon membrane (Hybond-N, Amersham-Pharmacia Biotech, Piscataway, NJ).36 RNA transfer was confirmed by visualization of ethidium bromide stained RNA under UV light. Blots were UV crosslinked and stored at 4°C until hybridization with random primed α-[32P] dATP labeled probe. The probe was generated from a human cDNA clone of putative c-Ha-ras protein mRNA sequence amplified by RT-PCR reaction (ATCC, Manassas, VA). Human β-Actin (Clontech. Inc., CA) was use as a control for this blot. Blot was then exposed to Kodak X-OMAT AR film at −80°C for 24 hr. Intensity was assessed by using computer evaluated densitometric scanning (Molecular Dynamics, NJ).
Assessment of LOH/MSI at various polymorphic markers on 11p15.5–15.4 loci showed that the observed mutations were more pronounced in cell lines exposed to multiple doses of radiation and treatment with estrogen (Fig. 2a–f). We have used a panel of polymorphic microsatellite markers to identify and investigate regions showing aberration on short arm of chromosome 11. A total of 8 microsatellite markers were utilized to detect allelic imbalance in irradiated, tumorigenic and tumor cell lines (Table I). These allelic imbalances were more pronounced as phenotypic characters of cellular transformation progressed from early to late stage when the transformed cell lines became tumorigenic.
Initially, the 60cGy cell line was compared to MCF-10F cell line (Figure 2a), where a total of two loci were altered. Such alterations were in the form of MSI at loci 11p15.5 (TH) and 11p15.4 (D11S1323) whereas no LOH was detected. When the cell lines 60cGy/60cGy (early) and MCF-10F were compared (Fig. 2b), a total of 4 alterations were observed. Loss of heterozygosity was screened at 11p15.4 (D11S1323) whereas MSI at 11p15.5 (HRAS1 and TH) and 11p15.5–15.4 (D11S4088). When cell lines 60cGy/60cGy (late) and MCF-10F were compared (Fig. 2c), however, a total of 6 alterations were detected, 3 for LOH and 3 for MSI. Among those 6 alterations, 4 were similar to early stage and the 2 additional LOHs were found in 11p15.5 (D11S4046) and 11p15.4 (HBB) loci.
Comparisons were also made between the tumorigenic 60cGy+E/60cGy+E and the non-tumorigenic MCF-10F+E control cell lines (Fig. 2d). A total of 6 alterations were identified with LOH in 4 and MSI in 2. Loss of heterozygosity was observed at the 11p15.5 (D11S4046 and TH) and 11p15.4 (HBB and D11S1323) loci, whereas MSI was detected at 11p15.5–15.4 (HRAS1 and D11S4088) locus. Similarly, Tumor-2 cell line showed 8 alterations when compared to the parental MCF-10F cell line (Fig. 2e): 5 for LOH and 3 for MSI. Loci 11p15.5 (D11S4046, TH and D11S1338) and 11p15.4 (HBB and D11S1323) were screened for LOH and MSI was detected at loci 11p15.5 (HRAS1 and D11S1318) and 11p15.5–15.4 (D11S4088). Seven alterations were noticed for MCF-7 when screened and compared to the MCF-10F cell line (Fig. 2f), 4 for LOH and 3 for MSI. The locus 11p15.5 (D11S4046, HRAS1 and TH) and 11p15.4 (HBB) are more susceptible toward LOH whereas loci 11p15.5 (D11S1318), 11p15.4 (D11S1323) and 11p15.5–15.4 (D11S4088) toward MSI.
We compared the phenotypic alterations of different cell lines with mutation profiles at codon 12 and 61 of the c-Ha-ras oncogene that are within this altered region (Table II). Alteration of the gene at codon 12 of exon 1 and codon 61 of exon 2 produced point mutations in the irradiated, tumorigenic and tumor cell lines when compared to parental MCF-10F cell line (Fig. 3 and Table II). These point mutations were responsible for changes in certain base(s), which ultimately led to changes in amino acid(s). No point mutation was detected when MCF-10F cell line was treated with only estrogen, i.e., in MCF-10F + E cell line. Compare with control MCF-10F cell line, point mutations were also noticed even in the irradiated cell lines exposed to same dose of irradiation but with different passage numbers (data not shown).
|Cell lines1||Passage2||c-Ha-ras mutation||Phenotypic alterations|
|Codon 12||Codon 61||AIA3||IA4||TA5|
|MCF-10F + E||45||−||−||−||−||−|
|10F + 60cGy||46||TCG TAC||CGG CGC||−||−||−|
|10F + 60cGy + E||48||ATC CTC||TCG CCG||−||−||−|
|10F + 60cGy/60cGy||44||AAG AAT||CAC AAC||+||+||−|
|GCA CCA||CCA CCC|
|10F + 60cGy/60cGy + E||41||ATC CTC||CAT CCT||+||+||−|
|TCG TAC||TGC TGG|
|10F + (60cGy + E)/(60cGy + E)||45||AAG AAT||CGG CGC||+||+||+|
|ATC CTC||TGT TGC|
|AGC ATC||CAT CCT|
|Tumor-2||30||ATC CTC||CGG CGC||+||+||+|
|CAT CCT||CAT CCT|
|AGC ATC||GGT GAT|
|GCA GCC||TGC TGG|
The 60cGy cell line (p46), when compared to the parental MCF-10F cell line, showed a (TCGTAC) mutation in codon 12 resulting in the change of Ser to Tyr and a (TCGCCG) transition in codon 61 resulting in the substitution of a Ser to Pro (Table II). There was also a (CGGCGC) base substitution in codon 61 but both the genetic codes belong to the same amino acid Arg. The cell line 60cGy + E (p48) in codon 12 showed a (ATCCTC) transversion resulting in the substitution of Ile with Leu whereas a Ser (TCG) to Pro (CCG) substitution in codon 61 was detected. There was also a base (GCAGCC) substitution on codon 12 but both the genetic codes belong to amino acid Ala.
When 60cGy/60cGy (p44) cell line was compared to MCF-10F cell line, 3 mutations were detected in codon 12: Lys (AAG) to Asn (AAT), Ala (GCA) to Pro (CCA) and His (CAC) to Pro (CCC). Another 2 mutations were detected in codon 61 including a His (CAC) to Asn (AAC) and a base substitution from (CCACCC) in which both genetic codes belong to the same amino acid Pro (Table II). Similarly, 2 mutations were also identified for each of the 2 codon 12 and 61 in 60cGy/60cGy + E (p41) cell line, Ile (ATC) to Leu (CTC) and Ser (TCG) to Tyr (TAC) for codon 12 and His (CAT) to Pro (CCT) and Cys (TGC) to Trp (TGG) for codon 61.
The 60cGy+E/60cGy+E (p45) tumorigenic cell line also indicated point mutations with respect to parental MCF-10F cell line. In codon 12, 4 mutations were identified from Lys (AAG) to Asn (AAT), Ile (ATC) to Leu (CTC), Ser (AGC) to Ile (ATC) and Ser (TCG) to Tyr (TAC), whereas in codon 61, 3 mutations were detected that included a amino acid change from His (CAT) to Pro (CCT) and a base substitution from CGG (Arg)CGC (Arg) and TGT (Cys)TGC (Cys) (Table II). Finally, a total of 9 mutations were detected in the Tumor-2 (p30) cell line, among them 5 were identified in codon 12 and the remaining 4 in codon 61. Mutations of Ile (ATC) to Leu (CTC), His (CAT) to Pro (CCT), Ser (AGC) to Ile (ATC), Ser (TCG) to Tyr (TAC) and a base substitution of GCAGCC (both for amino acid Ala) were found in codon 12 whereas mutations resulting in the substitutions of His (CAT) to Pro (CCT), Gly (GGT) to Asp (GAT) and Cys (TGC) to Trp (TGG) and a base substitution of CGGCGC (both for the amino acid Arg) were found in codon 61.
Using Northern blot analysis, all the irradiated, tumorigenic and tumor cell lines were found to show a gradual over-expressions of c-Ha-ras oncogene when compared to the parental MCF-10F cell line (Fig. 4). This gradual increase was pronounced as the cell line(s) moved from the transformed (Lane 4) to the tumorigenic stage (Lane 7). No altered expression was noticed when the parental cell line was treated with estrogen. The Tumor-2 cell line showed about a 4–5-fold increase in expression when compared to control. Expression of human β-actin control remains similar in all the cell lines tested.
In the present study, we have demonstrated that parental MCF-10F cell line exposed to double doses of α-particle radiation and treated with estrogen show a more complex pattern of allelic imbalance compared to cell lines treated with a single dose of radiation without estrogen. Exposure of cell lines to double doses of radiation without estrogen and analyzed at different passages also showed a progressive changes. A single 60cGy dose of α-particles was sufficient to induce persistent genetic changes in 2/8 of the marker loci examined, however, long before the cell reached the tumorigenic stage.
Allelic alterations induced by irradiation with either a single or double doses of α-particles in the presence or absence of estrogen were expressed either in the form of LOH/MSI or by some phenotypic changes, such as anchorage independence or invasive capabilities. These results indicated that the alterations were not only due to intrinsic level of genomic instability in these cell lines but also cell division per se, thereby increasing the risk of genetic errors.39 Consequently, both the doses of radiation and passage number of the cell lines after irradiation directly influenced these alterations and this genetic effect was more deleterious when given in combination with estrogen.
A progressive degree of allelic imbalance (MSI and LOH) was identified at 11p15.5–15.4 in the early transformed stage (60cGy), late transformed stage [60cGy/60cGy (early and late)], tumorigenic (60cGy + E/60cGy + E) and in tumor cell lines established from a tumor nodule (Tumor-2) using specific microsatellite markers belongs to this region. Aberrations in these markers are frequently associated with neoplastic changes due to their location in regions at, or near loci associated with cell-cycle regulation, DNA replication, DNA repair or signal transduction protein genes.40, 41 Our study has identified distinct regions on chromosome 11p15.5–15.4, that are subjected to LOH/MSI during radiation induced breast tumor progression. The critical regions appear to extend between the markers D11S1318 and D11S4088 (i.e., Region 1) and the other between D11S1338 and D11S1323 (i.e., Region 2). Studies from other laboratories have already placed various putative tumor suppressor genes in the larger overlapping area between TH and D11S1323,16, 18 a finding that is consistent with our present observation. LOH involving these regions not only coincides with regions implicated in the pathogenesis of breast cancer but with several childhood and adult tumors including rhabdomyosarcoma, Wilms' tumors (WT), bladder, ovarian and testicular cancers as well.11, 42, 43, 44
There is also an increasing body of evidence indicating the existence of various growth-related imprinted genes in this region. They show genetic or epigenetic alterations in cancer or cancer-predisposing syndromes.45 Several genes (IGF-II, H19, p57KIPS, TSSC5 and BWS, etc.) that map to the LOH on Region 1 are subjected to altered imprinting that may led to tumorigenesis by involving a gene activation hypothesis.46 The second hot-spot of LOH, i.e., Region 2, in breast tumors is centromeric to the putative WT2 gene (Region 1) that is also overlapping with the LOH regions described previously for various breast cancers.47 Our study, therefore, is the first report of a detail analysis of allelic imbalance associated with 11p15.5–15.4 loci that will help to establish the mechanism of action during radiation-induced progression of breast tumorigenesis.
It is interesting to note that the c-Ha-ras oncogene, which plays a role in the progression of human breast malignancies, also mapped to this 11p15.5 locus.19 The ras genes acquire their transforming capacity, either by single point mutations in the codon 12 or 61, leading to the expression of an aberrant gene product, or by the over-expression of the normal ras p21.20 Mutational analysis by direct sequencing of these 2 amplified codons identified various single nucleotide changes (Table II) and at the same time, over-expression of this gene in various irradiated, tumorigenic and tumor cell lines were also confirmed by Northern blot hybridization (Fig. 4). LOH at HRAS1 locus at 11p15.5 suggests the possible inactivation of one or more tumor suppressor gene(s) located near this locus that contributes to cellular transformation.48 In light of the present findings, we hypothesized that a precise point mutation at this specific locus by radiation and estrogen treatment to be a contributing factor in the progression of the neoplastic process.
Microcell-mediated chromosome transfer of an intact copy of chromosome 11 into the tumorigenic HeLa cells has provided additional support of the presence of a tumor suppressor gene in this chromosome.49 Based on our findings, it is difficult to rule out the possibility that an activated c-Ha-ras or other unknown oncogenes might have acted jointly to produce the malignant phenotype. It is possible that the activated c-Ha-ras is a surrogate for some yet unidentified genetic defect(s) in human breast cancer.50 Previous studies have also been unsuccessful to make preneoplastic MCF-10A cells malignant by introducing additional genetic anomalies such as over-expression of erbB-2 or TGFα.51 Although, c-Ha-ras mutations are infrequent, occurring only in about 5–10% of human breast cancers, there is considerable evidence to suggest that the pathways regulated by c-Ha-ras are also transcriptionally deregulated in various breast cancer cell lines.52 Mutations and elevated expression c-Ha-ras in this radiation- and estrogen-induced model system is, therefore, consistent with this observation. Thus, it is possible that the aberrant function of c-Ha-ras gene or ras related proteins may contribute to breast cancer development or progression.
The analysis of c-Ha-ras activation in codon 12 and 61 indicate that amino-acid substitutions at either codon 12 or 61 is more pronounced as the cell lines move more toward the tumorigenic stage. Few of the base substitution can only able to change the genetic code which also belong to the same amino acid. These alterations of base substitution are also related with the phenotypic changes of the cell lines (Table II).
One interesting finding of this work is the promotional effect of estrogen, along with radiation, on tumorigenesis. It is known to be a key requirement for the normal development of the mammary gland. Although, some reports have shown estrogen to affect c-Ha-ras expression whereas some others contradict such effects.53, 54, 55 It is difficult, however, to correlate the in vivo effects of estrogen on c-Ha-ras gene expression in case of both ER+ and ER− breast cancers but there is no doubt that the transcriptional regulation of c-Ha-ras by estrogen may play an important role in the progression of breast cancer.56
In conclusion, the complex and diverse nature of breast cancer is a result of many sequential molecular changes with variable efficiency. The array of genetic anomalies favors both chromosomal disintegration and mitotic recombination or oncogenic amplification, which leads to LOH/MSI. The correlation between tumor progression and loss of c-Ha-ras allele suggests that evaluation of this gene could be used as an additional parameter to study the mechanism responsible for the evolution of various adult and childhood cancers related to radiation. It will also help in identifying the appropriate targets for therapeutic intervention that contribute to the radiation-induced breast carcinogenesis.
This work was supported by funding from the National Institute of Health, grants CA 49062, ES 07890 and Environmental Health Center, grant P30 ES 09089 (T.K.H.) and the AVON Products Foundation Breast Cancer Center Institutional grant CU 51470301 to Columbia University (G.C.).
- 2Allelotype of breast cancer: cumulative allele losses promote tumor progression in primary breast cancer. Cancer Res 1990; 50: 7814–9., , , , , , .
- 36Molecular cloning—a laboratory manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press, 1989., , .