Reduction to homozygosity of genes on chromosome 11 in human breast neoplasia.

The somatic loss of heterozygosity for normal alleles occurring in human tumors has suggested the presence of recessive oncogenes. The results presented here demonstrate a loss of heterozygosity of several genes on chromosome 11 in primary breast tumors. Restriction fragment length polymorphism analysis of these DNAs further suggests that the most frequent loss of sequences in breast tumors occurs between the beta-globin and parathyroid hormone loci on the short arm of chromosome 11. The loss of heterozygosity for chromosome 11 loci has a significant association with tumors that lack estrogen and progesterone receptors, grade III tumors, and distal metastasis.

Slug/Pcad pathway controls epithelial cell dynamics in mammary gland and breast carcinoma.

Mammary gland morphogenesis results from the coordination of proliferation, cohort migration, apoptosis and stem/progenitor cell dynamics. We showed earlier that the transcription repressor Slug is involved in these functions during mammary tubulogenesis. Slug is expressed by a subpopulation of basal epithelial cells, co-expressed with P-cadherin (Pcad). Slug-knockout mammary glands showed excessive branching, similarly to Pcad-knockout. Here, we found that Slug unexpectedly binds and activates Pcad promoter through E-boxes, inducing Pcad expression. We determined that Pcad can mediate several functions of Slug: Pcad promoted clonal mammosphere growth, basal epithelial differentiation, cell-cell dissociation and cell migration, rescuing Slug depletion. Pcad also promoted cell migration in isolated cells, in association with Src activation, focal adhesion reorganization and cell polarization. Pcad, similarly to Slug, was required for in vitro 3D tubulogenesis. Therefore, Pcad appears to be responsible for epithelial-mesenchymal transition-linked plasticity in mammary epithelial cells. In addition, we found that genes from the Slug/Pcad pathway components were co-expressed and specifically correlated in human breast carcinomas subtypes, carrying pathophysiological significance.

Chordin is underexpressed in ovarian tumors and reduces tumor cell motility.

Ovarian cancers mostly derive from the monolayer epithelium that covers the ovary. There are currently very few molecular clues to the etiology of this cancer. Bone morphogenetic proteins (BMPs) are required for follicular development and female fertility and are expressed in the ovarian surface epithelium (OSE). We previously reported the expression of human chordin (CHRD), a BMP extracellular regulator, in the ovary. Here we show that CHRD is underexpressed in epithelium ovary cancer and epithelial cancer cell lines as compared with normal tissues and OSE, respectively. Besides, we detected BMP expression in all ovarian cell lines analyzed. To determine the functional relevance of the absence of CHRD mRNA in tumors and cancer cell lines, we studied the effects of CHRD on two cancer cell lines, BG1 and PEO14. Migratory and invasive properties were greatly reduced, whereas cell adhesion to the support was enhanced. In addition, we detected chordin (Chrd) expression in OSE of rat ovaries in a pattern similar to that of BMP4. Altogether, these results suggest that CHRD could participate in regulating BMP activity in normal OSE physiology, and that its mis-expression in OSE may facilitate cancer incidence and/or progression.

Transcription Factor Networks derived from Breast Cancer Stem Cells control the immune response in the Basal subtype.

Breast cancer is the most common cancer in women worldwide and metastatic dissemination is the principal factor related to death by this disease. Breast cancer stem cells (bCSC) are thought to be responsible for metastasis and chemoresistance. In this study, based on whole transcriptome analysis from putative bCSC and reverse engineering of transcription control networks, we identified two networks associated with this phenotype. One controlled by SNAI2, TWIST1, BNC2, PRRX1 and TBX5 drives a mesenchymal or CSC-like phenotype. The second network is controlled by the SCML4, ZNF831, SP140 and IKZF3 transcription factors which correspond to immune response modulators. Immune response network expression is correlated with pathological response to chemotherapy, and in the Basal subtype is related to better recurrence-free survival. In patient-derived xenografts, the expression of these networks in patient tumours is predictive of engraftment success. Our findings point out a potential molecular mechanism underlying the balance between immune surveillance and EMT activation in breast cancer. This molecular mechanism may be useful to the development of new target therapies.

High frequency of rare alleles of the human c-Ha-ras-1 proto-oncogene in breast cancer patients.

The c-Ha-ras-1 locus in 104 breast cancer patients and 56 unaffected individuals was examined for allelic restriction fragment-length polymorphism. Four common and 16 rare alleles were detected in the combined populations. The distribution of common and rare alleles differed significantly between the two populations. The common restriction fragments represented 91% of the allele pool in the unaffected population. In breast cancer patients, these common alleles represented only 59% of the allele pool (P less than .001). More specifically, the frequency of two of the common fragments, the 6.5- and 8.0-kilobase alleles, was significantly diminished in the breast cancer population (P less than .001 and P less than .02, respectively). The frequency of rare c-Ha-ras-1 alleles and hence genotypes composed of two rare alleles was increased in the breast cancer population (P less than .001). One of the rare alleles had a significant (P less than .05) association with these breast cancer patients. These results suggest that genotype analysis of the c-Ha-ras-1 locus, in combination with other clinical parameters, may be of prognostic value in assessing the potential for cancer.

Comparative genomic hybridization analysis of breast tumors with predetermined profiles of DNA amplification.

In a separate study (F. Courjal et al., Cancer Res., 57: 4360-4367, 1997), we have analyzed by Southern blotting the relationship between DNA amplification and clinicopathological features of breast cancer. Six regions of recurrent amplifications were tested (8p12, 8q24, 11q13, 12q13, 17q12, and 20q13), and the results suggested that there was a relationship between DNA amplification profiles and breast tumor phenotype. We had delineated three subgroups of tumors showing distinct DNA amplification profiles and clinicopathological characteristics: group A, tumors showing amplification at 11q13 and/or 8p12 and/or 20q13; group B, tumors amplified at ERBB2 and/or MYC and/or MDM2/SAS; and group C, tumors with no detectable amplification. The aim of the present work was to characterize extensively the amplification profiles in the different subgroups of tumors. Sixty-one breast tumors distributed in all three subgroups were studied by comparative genomic hybridization (CGH). There was an overall good agreement between Southern blotting results and CGH data. As expected, CGH revealed gains undetected by Southern blotting. Most of these gains occurred in regions for which no adapted probes were available but also revealed nondetected amplifications at 8q24 or 20q13. Tumors showed multiple aberrations with a medium number of 5.6 copy number variations/tumor, whereas, according to Southern blotting results, 38% of the tumors analyzed were devoid of any amplification. This proportion fell to 6.5% after CGH analysis. Recurrent gains were observed in tumors from all three subgroups, albeit at varying incidences, and involved 1q, 8q, 17q23-q24, and 20q13. Gains covered large regions of DNA and could possibly include several cores of amplification. Some events, such as gains at 16p11-p12 and 14q or losses at 22q, showed more restricted distributions, suggesting the existence of additional sets of preferential coamplifications. The complexity of genetic profiles revealed by CGH indicates that breast cancer development depends on a large (yet undetermined) number of genetic events. The description of molecular phenotypes in breast cancer may therefore prove to be complex, and it should be interesting to see how many breast tumor subtypes will be defined in the end.

Loss of a c-H-ras-1 allele and aggressive human primary breast carcinomas.

The human H-ras protooncogene was shown to be expressed in 16 of 22 invasive ductal carcinomas of the breast. The K- and N-ras protooncogenes were either not expressed or expressed at low levels. No amplification or rearrangement of the three ras genes was detected among the 104 breast carcinoma DNAs tested. These results indicate that the overexpression of H-ras in human breast tumors is not correlated with alteration of the protooncogene. In addition, we did not find any point mutation at the codon 12 of the H-ras or K-ras protooncogenes in 32 and 64, respectively, tumor DNAs examined. However, in tumor DNAs from 14 of 51 patients, heterozygous for H-ras-1 related BamHI restriction fragments, one allele was lost. This allele loss did not alter ras Mr 21,000 protein expression. Correlation with clinicopathological data showed, however, that the loss of one H-ras-1 allele in breast carcinoma DNAs is significantly linked to histological Grade III tumors, the lack of estrogen and/or progesterone receptors, and the subsequent occurrence of distal metastasis. Our results thus indicate that the loss of one H-ras-1 allele correlates with the most aggressive primary carcinomas of the breast.

p53 mutations occur in aggressive breast cancer.

Using a polymerase chain reaction-single strand conformation polymorphism approach we analyzed 96 human primary breast tumors for the presence of mutations in exons 2, 5, 6, 7, 8, and 9 of the p53 gene. These exons have been shown to comprise highly conserved sequences and the portion including exons 5 through 9 is believed to be the target for over 90% of the acquired mutations in human cancer. Eighteen tumors of the 96 (18.7%) tested showed reproducibly a variant band indicative of a mutation. Most (15 tumors) of the mutations were single nucleotide substitutions and G:C to A:T transitions were prevalent (6 tumors), G:C to T:A transversions came next (4 tumors), and guanines were always on the nontranscribed strand. Concomitant loss of the wild type allele and mutation of the other copy was observed in only 3 of 18 mutated cases; this is consistent with the heterogeneous cellular composition of breast tumors. Furthermore p53 mutations were correlated to estrogen and/or progesterone receptor negative tumors, thus indicating their relationships to aggressive breast cancer. No association could be observed with DNA amplification events in these tumors.

In situ c-myc expression and genomic status of the c-myc locus in infiltrating ductal carcinomas of the breast.

We have studied the expression of the c-myc protooncogene and the cycle-dependent histone 4 gene at the cellular level by RNA:RNA in situ hybridization in 18 primary breast ductal adenocarcinomas. These tumors have previously been examined by Southern and Northern blot analysis for the genomic status of c-myc and its expression, respectively (Escot et al., Proc. Natl. Acad. Sci. USA, 83: 4834-4838, 1986). Positive c-myc hybridization signals were associated with carcinoma cells in all cases, including tumors which had no apparent alterations of the c-myc locus. Steady-state levels of c-myc mRNA appeared heterogeneous in carcinomas with similar histology. High levels of hybridization were found in four of seven tumors with strong amplification of the c-myc locus. Similarly high levels of c-myc hybridization were detected in two of nine cases which had an apparently normal c-myc locus but comparatively low cellularity. In addition to carcinoma cells, dense clusters of infiltrating lymphocytes, present in three tumors, exhibited c-myc hybridization. The expression of the histone 4 gene failed to correlate with levels of c-myc expression. We conclude that in infiltrating ductal carcinomas: (a) the c-myc protooncogene is transcriptionally activated; (b) c-myc amplification is probably underestimated due to heterogeneous cellularity; (c) high-level c-myc amplification is related to high-level expression, but other unknown factors also may play a role; (d) differences in levels of c-myc expression may not only be attributed to differences in the growth fractions; and (e) c-myc mRNA in total RNA from biopsy samples may be contributed by infiltrating lymphocytes.

Mapping of DNA amplifications at 15 chromosomal localizations in 1875 breast tumors: definition of phenotypic groups.

DNA amplification is frequent in breast cancer and has been associated with specific clinicopathological parameters and/or worsened course of the disease. In the present work, we were interested in further defining the association linking the occurrence of DNA amplification to the emergence of specific breast tumor phenotype. To this aim, we studied by Southern blotting a total of 1875 breast tumor DNAs with 26 probes mapping at 15 distinct chromosomal localizations. Of the 26 loci tested, 11 loci showed elevated levels of amplification, 9 loci showed occasional and/or low level of DNA copy number increase, and 6 loci showed very rare or no variation. This allowed us to define six amplified domains mapping at 8p12, 8q24, 11q13, 12q13, 17q12, and 20q13.2, respectively. Over 60% of the tumors analyzed presented at least one amplification at one of these localizations. Amplifications often covered large regions of DNA and bore complex patterns involving coamplification of several colocalized markers. Statistical analysis revealed correlations associating DNA amplification with breast tumor phenotype, as well as sets of preferential coamplifications. Based on these correlations, we defined three subsets of breast cancer according to their patterns of DNA amplification. The first subset (group A) was organized around the amplifications at 11q13 and/or 8p12 and was predominantly composed of estrogen receptor-positive tumors and presented a large proportion of lobular cancers. The second subset (group B) was organized around the amplifications of ERBB2 and/or MYC. These tumors were mostly estrogen receptor-negative and of the ductal invasive type. The third subset (group C) corresponded to tumors in which no amplification was detected in the present screen. Tumors in this group were largely diploid and of low histopathological grading.

Relating genotype and phenotype in breast cancer: an analysis of the prognostic significance of amplification at eight different genes or loci and of p53 mutations.

Breast cancer heterogeneity can be related directly to its variability at the genetic level. Thus, tumor genotyping could be a valuable approach to define breast tumor subtypes. It has been shown that it is possible to delineate subgroups of breast tumors according to specific sets of DNA amplifications. The aim of the present work was to study the prognostic significance of these DNA amplifications. We studied DNA amplification at eight genes or loci (AIB1, CCND1, EMS1, ERBB2, FGFR1, MDM2, MYC, and RMC20C001) as well as p53 mutations in a series of 640 breast cancer patients who had not received presurgical therapy and analyzed the correlations with survival DNA amplification was assessed by Southern blotting and was scored positive when exceeding three to five copies. Mutations in the p53 gene were searched by four-color fluorescent single. strand conformational polymorphism, using an automated sequencer. Of the nine genetic alterations tested, four (CCND1, EMS1, FGFR1, and p53 mutations) showed a significant association with reduced disease-free (DFS) and/or overall survival (OVS) in the unselected set of patients by univariate test. Correlations for p53 were found only when selecting mutations in exons 5 or 7. Analysis of node-negative and -positive subgroups of patients showed that MDM2 amplification and p53 mutations bore prognostic significance in node-negative patients, whereas amplification of CCND1, EMS1, and FGFR1 correlated with poor outcome in node-positive patients. Multivariate analysis on an unselected set of patients retained significance for the amplification of EMS1, FGFR1, and MDM2 with DFS, of CCND1 with OVS, and of RMC20C001 with both DFS and OVS. Interestingly, stratified analysis according to nodal status confirmed results obtained in the univariate tests: significance of MDM2 amplification and p53 mutations in node-negative and that of CCND1, EMS1, and FGFR1 in node-positive patients. We also observed an association between the number of genetic alterations observed in a tumor and poor prognosis. Patients with two or more amplified loci had a worsened outcome. Strongly correlating coamplifications such as CCND1 and FGFR1, as well as ERBB2 and MYC, were associated with a significant reduction of patient survival, thus indicating cooperative effects. Our data support the idea that genetic alterations in breast cancer are not only helpful for phenotyping purposes, but can also represent powerful prognostic indicators in the clinical practice.

Consortium study on 1280 breast carcinomas: allelic loss on chromosome 17 targets subregions associated with family history and clinical parameters.

The pattern of loss of heterozygosity (LOH) on chromosome 17 in human breast cancer is complicated and shows many different regions of loss. In an attempt to narrow down the relevant regions of LOH on chromosome 17, we have studied the deletion pattern and its association with clinical parameters in 1280 breast carcinoma-venous blood lymphocyte pairs. In total, 42 different chromosome 17 loci were investigated, and between 25 and 625 cases were analyzed at each locus. The frequency of LOH observed on the p arm was much higher than that observed on the q arm. The opposite effect was observed in 52 ovarian cancer cases investigated, with less LOH on 17p than on 17q. Patterns of loss consistent with interstitial and terminal deletions, as well as loss of either the p or q arm or monosomy 17 were observed. To determine whether loss at particular loci may be associated with biological features of breast tumors, clinical data including age of onset, family history of breast cancer, tumor histopathology, tumor size, estrogen receptor (ER) status, and occurrence of lymph node or distant metastases were collected for each case. Overall, large-sized, ER-negative, lymph node-positive ductal tumors showed the highest frequencies of LOH, with ER-negative and ductal tumors showing LOH for markers along the majority of the chromosome. Eight regions of chromosome 17 appear to be associated with human breast cancer, two on 17p and six on 17q. These regions were not necessarily in the areas exhibiting the highest frequencies of LOH but were defined by interstitial and terminal deletions in multiple independent cases. Seven of these regions showed statistically significant differences in LOH associated with clinical parameters. These data strongly suggest that loci on chromosome 17 may determine aspects of tumor presentation and disease behavior in human breast cancer and pinpoint candidate tumor suppressor gene loci.

Retraction: miR-661 expression in SNAI1-induced epithelial to mesenchymal transition contributes to breast cancer cell invasion by targeting Nectin-1 and StarD10 messengers.

At the request of the University of Luxembourg and following an external investigation, the Editor and Publisher have agreed to retract this paper owing to unreliable data.

p53 mutations in ovarian cancer: a late event?

Using a combination of polymerase chain reaction and single-strand conformation polymorphism techniques we analyzed 34 ovarian cancer samples (30 primary tumors and four matched metastases) for the presence of mutations in exons 5, 6, 7, 8 and 9 of the p53 gene. Mutations in this portion of the gene are known to lead to the loss of the oncosuppressive potential of p53. Thirty-six percent (11/30) of the ovarian carcinomas tested presented a mutated p53 allele. Mutations were clustered in exons 5 and 7 to the exclusion of the other exons screened. Most mutations (10/11) were point mutations, but no preferential pattern of nucleotide substitution could be observed. In three tumors the mutation of one allele was concomitant with the loss of the wild-type counterpart. Another sample presented both alleles independently mutated. These observations are in agreement with the recessive nature of the p53 mutation. However, analysis of tissue sections from two tumors showed that the portion composed of 100% cancer cells could hold both the mutated and the wild-type form. Moreover analysis of serial sections gave evidence of a heterogeneous cellular content in one of these tumors, suggesting that p53 mutations may, in some cases, occur late during ovarian cancer evolution. It is, moreover, noticeable that, in matched sets of primary tumors and metastases, the same mutation was observed in both tumor samples. Therefore, even as a late event, p53 mutation occurs before metastatic spread.

Amplification of 11q13 DNA sequences in human breast cancer: D11S97 identifies a region tightly linked to BCL1 which can be amplified separately.

In the course of our study on the amplification of 11q13 sequences in human breast cancer, we have investigated the amplification status of the anonymous DNA fragment D11S97 in a series of 125 mammary tumors. Our results indicate that, as with bladder carcinomas, D11S97 can be amplified separately from BCL1. In addition, we have shown that D11597 is physically linked to both D11S146 and BCL1, and is less than 100 kb centromeric to the D115146. These results indicate that, in addition to other 11q13 loci, sequences located approximately 500 kb centromeric from BCL1 could contribute to carcinogenesis of epithelial cells in vivo.

Nucleotide sequence polymorphism in a hotspot mutation region of the p53 gene.

By screening for mutations in the p53 coding sequence by means of single-strand conformation polymorphism (SSCP) in a series of breast tumors we detected a novel polymorphism. This change in the SSCP pattern was detected in 6.2% of the tumor DNAs analysed and implied an A to G substitution at the last base of codon 213, thus representing a neutral change. First suspecting a somatic mutation we confirmed its presence in matched sets of DNAs from normal tissues. Extending our study to a series of 60 ovarian carcinomas and 70 healthy blood donors we noticed that this polymorphism represented only 3% and 2.6% respectively. We wondered if the difference in frequency in the breast cancer population might not be related to familial breast cancer and analysed 26 DNAs from patients showing predisposition to the disease. Two patients presented this polymorphism and one corresponding kindred was analysed, revealing a mendelian mode of transmission but no correlation with the cancer phenotype.

17q21-q25 aberrations in breast cancer: combined allelotyping and CGH analysis reveals 5 regions of allelic imbalance among which two correspond to DNA amplification.

Chromosome 17q is frequently rearranged in breast cancer. Allelotyping studies have proposed the existence of at least four regions of allelic imbalance (AI). Here we present a study combining allelotyping using 19 CA repeat markers mapping in the 17q21-25 region and molecular cytogenetics (CGH and FISH). Allelotyping was undertaken on 178 pairs of cognate tumor and normal DNA in order to determine the number of regions of AI and define the shortest overlaps. AI ranged from 34-54% of the informative cases according to the marker and, overall, 66% of the tumors presented AI at one of the markers tested. Analysis of the patterns of imbalances revealed at least five common regions of imbalance respectively defined by markers: D17S855, which is intragenic of BRCA1 (SRO 1), D17S1607 (SRO 2), D17S1855 (SRO 3), between D17S789 and D17S785 (SRO 4) and D17S784 (SRO 5). In order to characterize the nature of the genetic events revealed by allelotyping we performed CGH analysis on a subset of 43 tumors presenting variable patterns of imbalance. CGH showed that AI at 17q could represent four different types of genetic events: loss of chromosome 17, gain of 17q, gain of 17q22-q24, loss of 17q11-q21 and/or 17q25-qter. Some of these anomalies could occur concomitantly within the same tumor. Since 35% of the tumors analysed by CGH presented gains, these data indicated that AI at 17q were not solely indicative of losses of genetic material and could also represent DNA amplification. Gains were most commonly observed in the 17q23-q24 regions. This suggested that AI in SRO 2 and SRO 3 corresponded to DNA amplification. To assess this, we isolated BAC clones by PCR screening for markers D17S1607 and D17S1855 and used these in FISH experiments on six breast tumor cell lines and 14 breast cancer specimens. FISH results showed that both D17S1607 and D17S1855 were frequently involved in DNA amplification (8-30 copies). Altogether, our data show that allelotyping can be efficiently used in amplicon mapping. Clinico-pathological correlations indicated that imbalance at 17q preferentially occurred in high grade, PR- and ERBB2 amplified tumors.

p53 gene mutations in human epithelial skin cancers.

In the present study we analysed 38 epithelial skin cancers, 19 basal cell carcinomas (BCCs), 13 squamous cell carcinomas (SCCs) and six Bowen diseases (BwDs), using a combination of polymerase chain reaction (PCR) and single-stranded conformation polymorphism (SSCP) techniques for the presence of p53 and RAS gene mutations. Whereas 48% (9/19) of the BCCs tested presented a mutated p53 gene, the frequency was lower (15%, 2/13) in our series of SCCs and negative in the BwDs. Nine of the 11 characterized mutations were single-nucleotide substitutions and, interestingly, seven of these involved CC dimers, where a C was changed into a T or a G (three C-->T transitions and four C-->G transversions). This mutational pattern, added to the fact that all the mutated tumors occurred at sun-exposed body sites, implicates UV light in their genesis. Furthermore, we observed two internal deletions of 6 and 24 bp whose flanking sequences contained two or three Cs on either strand. In addition to molecular detection, we searched for p53 protein accumulation, by immunocytochemical staining, in a subset of 23 epithelial skin tumors (nine bearing a mutation, 14 which scored negative in our assay). Three commercially available anti-p53 antibodies (PAb CM1, mAbs DO7 and 1801) were used, and 3/23 (all showing a mutated p53 gene) presented specific nuclear staining. In contrast to other reported data we could not detect any activating RAS gene mutation in our series of human skin cancers.

Relative efficiency of denaturing gradient gel electrophoresis and single strand conformation polymorphism in the detection of mutations in exons 5 to 8 of the p53 gene.

p53 is the most commonly mutated gene in a large variety of human tumors including familial cancers. Because p53 mutations have in a number of human cancer types, been related to a negative outcome of the disease and the importance of pre-symptomatic diagnosis in cancer-prone families, screening for p53 mutations is becoming more and more widely used. In order to avoid sequencing of the complete coding sequence, several pre-screening methods have been developed and applied to the p53 gene. Among them, Single Strand Conformation Polymorphism (SSCP) and Denaturing Gradient Gel Electrophoresis (DGGE) appear to be highly sensitive. In this work, we used 52 different p53 variants to compare the two methods. In our conditions, DGGE is more sensitive than SSCP since 100% of the variants were detected. SSCP detected 90% of the variants, but efficiency of the method can still be improved by additional optimization experiments.