Genome-based and transcriptome-based molecular classification of breast cancer

Bièche, Ivan; Lidereau, Rosette

doi: 10.1097/CCO.0b013e3283412ee0
Cancer biology: Edited by Pierre Hainaut and Amelie Plymoth

Purpose of review: The highly heterogeneous clinical, histological, biological and genetic nature of breast malignancies is due in part to their extreme molecular complexity.

Recent findings: Many genetic and epigenetic alterations have been described, affecting a relatively small number of signaling pathways (PI3K, NK-κB, FGF, etc.) and thus several molecular subtypes of breast cancer have been identified.

Summary: The next decade will see even more prolific developments, notably with the advent of next-generation sequencing (NGS) technologies capable of providing individual patients' constitutional and somatic genome sequences both rapidly and cheaply. Knowledge of the full catalogue of somatic genetic alterations will pave the way for fully individualized management of breast cancer patients.

Laboratoire d'Oncogénétique, Hôpital René Huguenin, Institut Curie, St Cloud, France

Correspondence to Rosette Lidereau, Hôpital René Huguenin, Institut Curie, 35 rue Dailly, 92210 St Cloud, France Fax: +33 1 47 11 16 96; e-mail:

Article Outline
Back to Top | Article Outline


Breast cancer is the most frequent malignancy in women. Its incidence is rising and the outcome is still often fatal. The average 5-year relapse-free survival rate is currently about 70%. Breast cancer patients have highly variable clinical, histologic, biological and genetic features, owing to the complex etiology of this malignancy. A positive family history is the main risk factor. Two major susceptibility genes, BRCA1 and BRCA2, have been identified. The detection of somatic genetic and epigenetic alterations in breast tumors represented a major step in our understanding of the molecular bases of breast oncogenesis. Some such alterations have potential as diagnostic and prognostic markers, as treatment response criteria, and as sources of novel therapeutic targets.

Back to Top | Article Outline

Somatic genetic alterations

The most frequently observed somatic genetic abnormalities in breast tumors are DNA amplification and deletions. Other classical genetic alterations, such as point mutations, chromosomal translocation and viral insertion, appear to be rare in this setting, but exhaustive analysis of tumor DNAs and RNAs by means of microarray technology and genome-wide sequencing, are needed to show whether these types of genetic defect also play a noteworthy role in the onset of breast cancer.

Back to Top | Article Outline

DNA amplification and deletion

DNA amplification (mainly of proto-oncogenes, growth factors and their receptors) and DNA deletion (inactivating tumor-suppressor genes) are both frequently observed in breast tumors.

Early this decade, comparative genomic hybridization (CGH) of metaphase chromosomes identified about 20 chromosome regions amplified or deleted in breast tumors [1]. Some of these regions, previously detected with cytogenetic methods, appear to be associated with known oncogenes [8p11 (FGFR1), 8q24 (MYC), 11q13 (CCND1, EMSY) and 17q21.1 (ERBB2)] or tumor suppressor genes [13q14 (RB1), 17p13 (TP53)]. Some other alterations, such as amplicons in 17q23 (PPM1D) and 20q13 (ZNF217, NCOA3/AIB1, Aurora-A), pointed to the involvement of other genes in breast oncogenesis. In particular, alterations of Aurora-A, a major spindle checkpoint gene, may contribute to the aneuploidy frequently observed in breast cancer [2].

More recently, CGH arrays with a higher resolution than CGH have revealed smaller amplified or deleted regions. Chin et al. [3] used BAC-CGH arrays and transcriptomics to analyze 101 breast tumors, thus identifying 66 amplified and overexpressed genes, nine of which might serve as therapeutic targets (FGFR1, IKBKB, PROSC, ADAM9, FNTA, ACACA, PNMT, NR1D1 and ERBB2). Note that ERBB2 is specifically targeted by trastuzumab. BAC-CGH arrays were rapidly superseded by oligonucleotide- or SNP-CGH arrays, which provided a 100-fold increase in resolution, from the order of about a million to only 10 000 base pairs. Using SNP-CGH arrays, Leary et al. [4] studied 45 xenografted breast tumors and detected several amplifications and deletions smaller than 250 kb (CCNE1 and MRE11 amplification, and CDKN2A, PTEN and PCDH8 deletion). Involvement of the PCDH8 gene in breast cancer was later confirmed by functional studies [5].

Beroukhim et al. [6] identified 76 amplifications and 82 deletions in 243 breast tumors, in regions containing new possible culprit genes, such as MCL1 and BCL2L1 (apoptosis), and IRAK1, TRAF6, IKBKG and IKBKB (NK-κB signaling pathway).

Back to Top | Article Outline

Point mutations

Significant frequencies of point mutations have so far been detected in very few cancer genes in breast tumors (Table 1) [7,8]. Initial analyses of breast tumor exomes failed to identify any new genes exhibiting a high frequency of mutations [9,10]. Studies of large numbers of complete breast tumor genomes are needed to confirm the apparent rarity of point mutations in breast cancer.

Alterations of the tumor-suppressor gene TP53 represent the most frequent somatic genetic event observed in human tumors. TP53 mutations are found in about 30–35% of cases, along with losses of genetic material in the 17p13 region, in keeping with bi-allelic inactivation of tumor-suppressor genes.

PIK3CA, the gene encoding the catalytic subunit of phosphatidyl inositol 3-kinase (PI3K), is mutated in about 20–30% of breast tumors [11], mainly in exons 9 and 20, corresponding respectively to the helical (mutations E542K and E545K) and kinase (H1047R) domains of the protein. PIK3CA mutations appear to be more frequent in ERα-positive tumors (30–40%) than in ERα-negative tumors (10–20%) [11,12], and could be responsible for tumor resistance to membrane receptor tyrosine kinase inhibitors used in breast cancer (trastuzumab and lapatinib) [13,14].

AKT1 is mutated at a single codon (E17K mutation) in about 5% of breast tumors overall [15] but in about one-third of papillary tumors [16]. This mutation appears to have lower oncogenic potential than PIK3CA mutations [17]. PIK3CA mutations appear to be mutually exclusive with underexpression of the tumor-suppressor gene PTEN [18]. This suggests that, when two genes belong to the same signaling pathway, mutation of one of the two is sufficient to alter the signaling pathway.

CDH1, the gene encoding cadherin E, appears to be mutated most frequently in invasive lobular tumors [19]. The GATA3 gene is mutated in about 5% of breast tumors overall [20] but in about 22% of hereditary tumors harboring the wild-type BRCA1 and BRCA2 genes [19].

Interestingly, with the exception of the AKT1 gene, gain-of-function mutations of other oncogenes (EGFR, KRAS, HRAS, NRAS, BRAF, etc.) that lie downstream of tyrosine kinase receptors and are frequently mutated in other malignancies (lung, colon, etc.) are rare or absent in breast cancer.

Several partial or whole breast tumor exomes have been published [20–22]. Stephens et al. [20] sequenced the 518 genes encoding all kinase proteins (the ‘kinome’) of the human genome in 16 primary breast tumors. Four tumors (25%) exhibited one or more mutations. A similar analysis of 238 oncogenes in 60 breast tumors revealed mutations in 20 tumors (33%), mostly in the PIK3CA gene [21]. The first true exome-wide analysis was done in 2006 on 13 023 genes and 11 ERα-negative tumors, then extended in 2007 to 18 191 genes [9,10]. TP53 and PIK3CA were the genes most frequently mutated, the other genes being mutated in less than 20% of cases (OBSCN, SPTAN1, GRIN2D, GLI1, CUBN, TECTA and FLNB) [9]. These genes are preferentially involved in the PI3K and NK-κB signaling pathways. Interestingly, Wood et al. [10] showed that seven of the 33 genes identified by Theodorou et al. [22] by insertional mutagenesis with MMTV virus were mutated in one of the 11 breast tumors tested.

Back to Top | Article Outline

Chromosome translocation and fusion genes

Few chromosomal translocations creating fusion genes have so far been identified in breast cancer, mainly owing to the difficulty of studying cancer karyotypes [23]. The t(12;15)(p12;q26.1) translocation fuses the ETV6 and NTRK3 genes, leading to constitutive activation of NTRK3 [24]. This translocation is observed in 85% of secretory breast tumors, which represent a very rare ‘basal-like’ clinical entity accounting for less than 0.15% of all breast tumors [25]. New fusion genes are being actively sought in breast cancer by complete sequencing of tumor DNA or RNA. Two transcriptomes (RNAs) of breast cancer cell lineages have been sequenced (lineages HCC1954 and MCF7), thus identifying a large number of fusion genes, some of which activate genes that have previously been implicated in human oncogenesis [26,27]. This is notably the case of RAD51C, BRIP1, SULF2, MRE11 and NSD1.

Finally, sequencing of 24 breast tumor DNAs (nine cell lines and 15 breast tumors) identified 29 rearrangements that respected the reading frame. However, none of these rearrangements was observed in more than one of the 24 tumor DNAs, or in a further 288 breast tumors [28]. Re-arrangements have nevertheless been identified in known cancer-related genes (BRAF, PAX3, PAX5, NSD1, PBX1, MSI2 and ETV6). Note that this last gene (ETV6) is also involved in the t(12;15)(p12;q26.1) translocation specific to secretory breast tumors (ETV6/NTRK3 fusion gene).

Back to Top | Article Outline

Exogenous and endogenous viruses

The involvement of viruses in human breast oncogenesis remains highly controversial. Analysis of the integration site of the MMTV provirus (mouse mammary tumor virus) identified several oncogenes involved in murine breast oncogenesis and, potentially, in human breast cancer. This is notably the case of Int1/Wnt1, Int2/FGF3 and Int3/Notch4 [29]. More recently, high-throughput screening of the MMTV insertion site identified 33 genes principally involved in the Wnt and Notch developmental pathways and in the FGF (fibroblast growth factor) signaling pathway [22]. Seventeen of these 33 genes have been reported to be overexpressed in breast tumors.

The existence of an MMTV-like retrovirus that can infect humans and cause breast cancer has been suspected for some 50 years. Several independent teams have detected nucleotide sequences of viral origin, very similar to MMTV sequences, in about 40% of breast tumors [30,31]. In addition, an entire viral genome (a candidate ‘human mammary tumor virus’, HMTV) with 96% sequence identity to MMTV has been isolated [32].

More likely is the involvement of endogenous retroviral sequences (HERV, human endogenous retrovirus) corresponding to vestigial exogenous retrovirus infections. Full transcripts and env proteins of the HERV-K family have been detected in breast tumors but not in normal breast tissues [33]. Finally, HERV-K proteins are responsible for humoral and cellular immune responses in some patients with breast cancer [34].

Back to Top | Article Outline

Tumor whole-genome sequence analysis

Whole-genome sequence analysis will provide a full and detailed picture of all genomic changes present in the malignant cell. The cost of genome-wide sequencing continues to fall. Sequencing will eventually replace nearly all hybridization-based high-throughput genotyping methods (FISH, SNP-array, CGH-array, etc).

Two recent articles describe the first complete sequencing of breast tumor genomes. Shah et al. [35••] identified 32 previously undescribed acquired mutations in DNA extracted from a metastasis of an ERα-positive lobular tumor, but none of these mutations was confirmed in an independent series of 192 breast tumors, with the exception of HAUS3 (HAUS augmin-like complex, subunit 3), that was mutated in two other lobular tumors. The authors also identified two genes (COG3 and SRP9) bearing mutations in their RNA but not in their DNA, pointing to an aberrant RNA editing mechanism in the metastasis studied by Shah.

Ding et al. [36] fully sequenced DNA extracted from a ‘basal-like’ metastatic tumor, as well as its cerebral metastasis, a xenograft of the primary tumor, and the same patient's lymphocytes. Comparison of the four sequences identified 50 point mutations, 28 large deletions, six inversions, and seven translocations. Only three of these alterations affected genes known to be involved in human oncogenesis, namely a microdeletion of TP53, a large heterozygous deletion of FBXW7, and a homozygous deletion of CTNNA1 (a tumor-suppressor gene encoding alpha-catenin and frequently inactivated in hematologic malignancies) [37]. The mutation profiles of the metastasis and the xenograft were very similar, and the xenograft, despite bearing some additional alterations, remained similar to the primary tumor. These findings endorse the use of xenografts as an in-vivo model for functional studies and for preclinical drug development.

These first analyses of whole-genome sequence identified a multitude of genes altered in breast tumors, yet the number of signaling pathways affected by these mutations appears to be relatively small, at between 10 and 20.

The International Cancer Genome Consortium (ICGC) is intending to sequence the genomes of about 50 cancer types and subtypes in order to better understand the role of genomic alterations in their development [38•]. The consortium will analyze the genomes, but also the epigenomes and transcriptomes, of more than 25 000 tumors. Three projects will specifically focus on breast cancer.

Back to Top | Article Outline

Epigenetic alterations: the epigenome

Many recent articles point to a major role of epigenetic alterations in human tumors [39]. These include DNA methylation, chromatin remodeling, and regulation by noncoding RNA. Here we mainly examine epigenetic processes that can be analyzed at DNA and RNA levels, that is DNA methylation and noncoding RNA (microRNAs and long noncoding RNAs).

Back to Top | Article Outline

DNA methylation: the methylome

Analysis of the 5′ region of several tumor-suppressor genes, particularly in regions rich in CpG dinucleotides (CpG islands), has revealed hypermethylation in some breast tumors [40]. Ordway et al. [41] analyzed nine paired breast tumors and normal tissues and thus identified more than 200 new genes with altered methylation. A subgroup of these genes was validated in a large independent series of 230 tumors. In the case of the GHSR gene (ghrelin receptor gene), promoter hypermethylation correlates with GHSR underexpression.

In the near future the methylome will be analyzed simultaneously with the genome by means of ‘third-generation’ sequencing methods that enable simultaneous analysis of both thousands of nucleotides and the methylation status of cytosine residues [42••].

Back to Top | Article Outline

MicroRNAs: the MicroRNAome

MicroRNAs play a role in the control of gene expression. A given microRNA appears to regulate the expression of a large number of target genes involved in several physiological processes. Some microRNA-encoding genes are located in regions altered in tumors [43].

Table 2 lists the microRNAs most frequently altered in breast cancer, along with their principal target genes and functions [44,45].

MicroRNAs have therapeutic potential. In 2007, Robert A. Weinberg's team showed that microRNA 10b is overexpressed in metastatic breast cancer and that it regulates cell migration and invasion both in vitro (cell model) and in vivo (animal model) [46]. MicroRNA 10b, transcriptional expression of which is regulated by the Twist protein, mainly targets the HOXD10 gene, which is itself a transcriptional repressor of the prometastatic gene RHOC. Three years later, the same team treated BALB/c mice bearing xenografts of a highly metastatic murine cell line (4T1) with microRNA 10b antisense (antagomir) for 3 weeks. Tumor size was unaffected, but expression of Hoxd10 protein was repressed, leading to a very significant decrease (86%) in pulmonary metastases [47].

Back to Top | Article Outline

Long noncoding RNA: the ncRNAome

Long noncoding RNA also regulates gene expression [48,49]. The long ncRNA HOTAIR, located in the HOX gene cluster, is overexpressed in breast tumor metastases [50••]. HOTAIR activation is responsible for aberrant recruitment of the PRC2 complex (polycomb repressive complex 2), altered lysine 27 methylation in histone H3, and increased metastatic potential.

A second example of long ncRNA, transcribed by the ANRIL gene, is a cis-repressive RNA of INK4B/ARF/INK4A locus expression by CBX7 protein recruitment from the PRC1 complex [51]. The ANRIL gene is overexpressed in breast tumors [52], is one major target of GWAS studies, and is thus a putative breast cancer susceptibility locus [53].

Back to Top | Article Outline

Alterations of protein-coding gene expression

All these genetic alterations can provide aberrant gene expression studied at RNA and at protein levels.

Back to Top | Article Outline

Alterations of gene expression at the messenger RNA level: the transcriptome

Microarray technology has been used to quantify transcripts of the entire genome (transcriptome) in tumor samples. A first analysis of breast tumor transcriptomes [54] identified five main molecular subtypes of infiltrating carcinoma: luminal A, luminal B, ERBB2, basal-like and normal-like (Table 3), associated with predictive and prognostic molecular signatures in breast cancer.

The princeps article by Perou et al. also suggested the involvement in breast tumorigenesis of many unknown genes and signaling pathways. Later studies showed, for example, that basal-like carcinoma, mainly characterized by the absence of nuclear estrogen and progesterone receptor expression and membrane ERBB2 receptor expression, and by somatic inactivation of the BRCA1 gene, is associated with defective double-strand DNA break repair, a higher proportion of cancer stem cells [55], an epithelial-mesenchymal transition [56], a Wnt pathway activation [56,57], and an increased risk of pulmonary metastasis [56–58].

Among a compilation of transcriptome data on 3200 breast tumors analyzed in 31 studies (Oncomine database;, Rhodes et al. [59] identified a new gene (AGTR1; angiotensin II receptor type 1) that is strongly overexpressed in 10–20% of breast tumors. Losartan, an angiotensin II receptor inhibitor, led to a 30% reduction in breast tumor growth in a xenograft model.

Back to Top | Article Outline

Altered gene expression at the protein level: the proteome

Translation of messenger RNAs into proteins is subject to a variety of regulatory mechanisms, and especially post-translational modifications (glycosylation, phosphorylation, acetylation, degradation, etc.). Dysregulation of signal transduction pathways (PI3K/Akt/mTOR/p70S6-kinase and Ras/Raf/MEK/ERK), the main targets of current targeted therapies, is partly due to post-translational modifications. It is not currently possible to analyze all proteins simultaneously (estimated to 1 000 000). Seldi-Tof mass spectrometry of patient sera identified a component of complement fraction C3a from a protein peak with diagnostic value by comparing 176 patients with breast cancer and 46 healthy controls [60].

Hudelist et al. [61] used antibody arrays to analyze the expression level of 378 proteins in paired normal breast and tumor tissue samples. They found that some proteins were underexpressed in the tumors (14–3–3e), whereas others were overexpressed (p53, caseine kinase Ie, annexin XI, CDC25C, eIF-4E, and MAP-kinase 7).

Back to Top | Article Outline


Microarray-based studies have markedly improved breast tumor characterization. Many genetic and epigenetic alterations and a smaller number of altered signaling pathways (PI3K, NK-κB, FGF, etc.) characteristic of breast tumors have thus been identified. However, despite new therapeutic approaches, the use of biomarkers has not yet radically improved the management of patients with breast cancer. Indeed, only four biological markers are routinely used in this setting: estrogen receptor alpha is a predictive marker of the response to estrogen antagonists, the c-erbB2 oncogene is a predictive marker of the response to trastuzumab, and the serum tumor markers CA15.3 and ACE are used to monitor the response to treatment and to detect relapses.

The next decade will see an extraordinary wealth of discoveries in cancer research. Whole-genome sequencing will provide a far more detailed picture of the genomic alterations in malignant cells. Integrated analysis of these data will provide a global vision of individual tumor pathobiology.

Knowledge of the molecular characteristics of a given patient's tumor will give rise to tailored therapies, based on the identification of mutated genes in the primary tumor and in resistant subclones responsible for metastases. Finally, noninvasive analysis of proteins and nucleic acids (DNA and RNA) in blood samples will provide less aggressive approaches to treatment monitoring and early relapse detection.

Back to Top | Article Outline

References and recommended reading

Back to Top | Article Outline

Papers of particular interest, published within the annual period of review, have been highlighted as:

Back to Top | Article Outline

• of special interest

Back to Top | Article Outline

•• of outstanding interest

Back to Top | Article Outline

Additional references related to this topic can also be found in the Current World Literature section in this issue (pp. 127–128).

1 Tirkkonen M, Tanner M, Karhu R, et al. Molecular cytogenetics of primary breast cancer by CGH. Genes Chromosomes Cancer 1998; 21:177–184.
2 Anand S, Penrhyn-Lowe S, Venkitaraman AR. AURORA-A amplification overrides the mitotic spindle assembly checkpoint, inducing resistance to Taxol. Cancer Cell 2003; 3:51–62.
3 Chin K, DeVries S, Fridlyand J, et al. Genomic and transcriptional aberrations linked to breast cancer pathophysiologies. Cancer Cell 2006; 10:529–541.
4 Leary RJ, Lin JC, Cummins J, et al. Integrated analysis of homozygous deletions, focal amplifications, and sequence alterations in breast and colorectal cancers. Proc Natl Acad Sci U S A 2008; 105:16224–16229.
5 Yu JS, Koujak S, Nagase S, et al. PCDH8, the human homolog of PAPC, is a candidate tumor suppressor of breast cancer. Oncogene 2008; 27:4657–4665.
6 Beroukhim R, Mermel CH, Porter D, et al. The landscape of somatic copy-number alteration across human cancers. Nature 2010; 463:899–905.
7 Berx G, Cleton-Jansen AM, Strumane K, et al. E-cadherin is inactivated in a majority of invasive human lobular breast cancers by truncation mutations throughout its extracellular domain. Oncogene 1996; 13:1919–1925.
8 Usary J, Llaca V, Karaca G, et al. Mutation of GATA3 in human breast tumors. Oncogene 2004; 23:7669–7678.
9 Sjöblom T, Jones S, Wood LD, et al. The consensus coding sequences of human breast and colorectal cancers. Science 2006; 314:268–274.
10 Wood LD, Parsons DW, Jones S, et al. The genomic landscapes of human breast and colorectal cancers. Science 2007; 318:1108–1113.
11 Samuels Y, Diaz LA Jr, Schmidt-Kittler O, et al. Mutant PIK3CA promotes cell growth and invasion of human cancer cells. Cancer Cell 2005; 7:561–573.
12 Hennessy BT, Gonzalez-Angulo AM, Stemke-Hale K, et al. Characterization of a naturally occurring breast cancer subset enriched in epithelial-to-mesenchymal transition and stem cell characteristics. Cancer Res 2009; 69:4116–4124.
13 Berns K, Horlings HM, Hennessy BT, et al. A functional genetic approach identifies the PI3K pathway as a major determinant of trastuzumab resistance in breast cancer. Cancer Cell 2007; 12:395–402.
14 Eichhorn PJ, Gili M, Scaltriti M, et al. Phosphatidylinositol 3-kinase hyperactivation results in lapatinib resistance that is reversed by the mTOR/phosphatidylinositol 3-kinase inhibitor NVP-BEZ235. Cancer Res 2008; 68:9221–9230.
15 Carpten JD, Faber AL, Horn C, et al. A transforming mutation in the pleckstrin homology domain of AKT1 in cancer. Nature 2007; 448:439–444.
16 Troxell ML, Levine J, Beadling C, et al. High prevalence of PIK3CA/AKT pathway mutations in papillary neoplasms of the breast. Mod Pathol 2010; 23:27–37.
17 Lauring J, Cosgrove DP, Fontana S, et al. Knock in of the AKT1 E17K mutation in human breast epithelial cells does not recapitulate oncogenic PIK3CA mutations. Oncogene 2010; 29:2337–2345.
18 Saal LH, Holm K, Maurer M, et al. PIK3CA mutations correlate with hormone receptors, node metastasis, and ERBB2, and are mutually exclusive with PTEN loss in human breast carcinoma. Cancer Res 2005; 65:2554–2559.
19 Arnold JM, Choong DY, Thompson ER, et al. Frequent somatic mutations of GATA3 in non-BRCA1/BRCA2 familial breast tumors, but not in BRCA1-, BRCA2- or sporadic breast tumors. Breast Cancer Res Treat 2010; 119:491–496.
20 Stephens P, Edkins S, Davies H, et al. A screen of the complete protein kinase gene family identifies diverse patterns of somatic mutations in human breast cancer. Nat Genet 2005; 37:590–592.
21 Thomas RK, Baker AC, Debiasi RM, et al. High-throughput oncogene mutation profiling in human cancer. Nat Genet 2007; 39:347–351.
22 Theodorou V, Kimm MA, Boer M, et al. MMTV insertional mutagenesis identifies genes, gene families and pathways involved in mammary cancer. Nat Genet 2007; 39:759–769.
23 Rabbitts TH. Commonality but diversity in cancer gene fusions. Cell 2009; 137:391–395.
24 Tognon C, Knezevich SR, Huntsman D, et al. Expression of the ETV6-NTRK3 gene fusion as a primary event in human secretory breast carcinoma. Cancer Cell 2002; 2:367–376.
25 Laé M, Fréneaux P, Sastre-Garau X, et al. Secretory breast carcinomas with ETV6-NTRK3 fusion gene belong to the basal-like carcinoma spectrum. Mod Pathol 2009; 22:291–298.
26 Zhao Q, Caballero OL, Levy S, et al. Transcriptome-guided characterization of genomic rearrangements in a breast cancer cell line. Proc Natl Acad Sci U S A 2009; 106:1886–1891.
27 Hampton OA, Den Hollander P, Miller CA, et al. A sequence-level map of chromosomal breakpoints in the MCF-7 breast cancer cell line yields insights into the evolution of a cancer genome. Genome Res 2009; 19:167–177.
28 Stephens PJ, McBride DJ, Lin ML, et al. Complex landscapes of somatic rearrangement in human breast cancer genomes. Nature 2009; 462:1005–1010.
29 Callahan R, Smith GH. MMTV-induced mammary tumorigenesis: gene discovery, progression to malignancy and cellular pathways. Oncogene 2000; 19:992–1001.
30 Wang Y, Holland JF, Bleiweiss IJ, et al. Detection of mammary tumor virus env gene-like sequences in human breast cancer. Cancer Res 1995; 55:5173–5179.
31 Lawson JS, Glenn WK, Salmons B, et al. Mouse mammary tumor virus-like sequences in human breast cancer. Cancer Res 2010; 70:3576–3585.
32 Liu B, Wang Y, Melana SM, et al. Identification of a proviral structure in human breast cancer. Cancer Res 2001; 61:1754–1759.
33 Wang-Johanning F, Frost AR, Jian B, et al. Quantitation of HERV-K env gene expression and splicing in human breast cancer. Oncogene 2003; 22:1528–1535.
34 Wang-Johanning F, Radvanyi L, Rycaj K, et al. Human endogenous retrovirus K triggers an antigen-specific immune response in breast cancer patients. Cancer Res 2008; 68:5869–5877.
35•• Shah SP, Morin RD, Khattra J, et al. Mutational evolution in a lobular breast tumour profiled at single nucleotide resolution. Nature 2009; 461:809–813.
36 Ding L, Ellis MJ, Li S, et al. Genome remodelling in a basal-like breast cancer metastasis and xenograft. Nature 2010; 464:999–1005.
37 Liu TX, Becker MW, Jelinek J, et al. Chromosome 5q deletion and epigenetic suppression of the gene encoding alpha-catenin (CTNNA1) in myeloid cell transformation. Nat Med 2007; 13:78–83.
38• International Cancer Genome Consortium. International network of cancer genome projects. Nature 2010; 464:993–998.
39 Paro R, Lee JT. Extending the frontiers of epigenetic regulation. Curr Opin Genet Dev 2010; 20:107–109.
40 Lo PK, Sukumar S. Epigenomics and breast cancer. Pharmacogenomics 2008; 9:1879–1902.
41 Ordway JM, Budiman MA, Korshunova Y, et al. Identification of novel high-frequency DNA methylation changes in breast cancer. PLoS One 2007; 2:e1314.
42•• Flusberg BA, Webster DR, Lee JH, et al. Direct detection of DNA methylation during single-molecule, real-time sequencing. Nat Methods 2010; 7:461–465.
43 Iorio MV, Ferracin M, Liu CG, et al. MicroRNA gene expression deregulation in human breast cancer. Cancer Res 2005; 65:7065–7070.
44 O'Day E, Lal A. MicroRNAs and their target gene networks in breast cancer. Breast Cancer Res 2010; 12:201.
45 Calin GA, Croce CM. MicroRNA signatures in human cancers. Nat Rev Cancer 2006; 6:857–866.
46 Ma L, Teruya-Feldstein J, Weinberg RA. Tumour invasion and metastasis initiated by microRNA-10b in breast cancer. Nature 2007; 449:682–688.
47 Ma L, Reinhardt F, Pan E, et al. Therapeutic silencing of miR-10b inhibits metastasis in a mouse mammary tumor model. Nat Biotechnol 2010; 28:341–347.
48 Ponting CP, Oliver PL, Reik W. Evolution and functions of long noncoding RNAs. Cell 2009; 136:629–641.
49 Mercer TR, Dinger ME, Mattick JS. Long noncoding RNAs: insights into functions. Nat Rev Genet 2009; 10:155–159.
50•• Gupta RA, Shah N, Wang KC, et al. Long noncoding RNA HOTAIR reprograms chromatin state to promote cancer metastasis. Nature 2010; 464:1071–1076. The first study of long ncRNA in breast tumors identifying the long ncRNA called HOTAIR, overexpressed in breast tumor metastases.
51 Yap KL, Li S, Muñoz-Cabello AM, et al. Molecular interplay of the noncoding RNA ANRIL and methylated histone H3 lysine 27 by polycomb CBX7 in transcriptional silencing of INK4a. Mol Cell 2010; 38:662–674.
52 Pasmant E, Laurendeau I, Héron D, et al. Characterization of a germ-line deletion, including the entire INK4/ARF locus, in a melanoma-neural system tumor family: identification of ANRIL, an antisense noncoding RNA whose expression coclusters with ARF. Cancer Res 2007; 67:3963–3969.
53 Turnbull C, Ahmed S, Morrison J, et al. Genome-wide association study identifies five new breast cancer susceptibility loci. Nat Genet 2010; 42:504–507.
54 Perou CM, Sørlie T, Eisen MB, et al. Molecular portraits of human breast tumours. Nature 2000; 406:747–752.
55 Charafe-Jauffret E, Ginestier C, Iovino F, et al. Breast cancer cell lines contain functional cancer stem cells with metastatic capacity and a distinct molecular signature. Cancer Res 2009; 69:1302–1313.
56 DiMeo TA, Anderson K, Phadke P, et al. A novel lung metastasis signature links Wnt signaling with cancer cell self-renewal and epithelial-mesenchymal transition in basal-like breast cancer. Cancer Res 2009; 69:5364–5373.
57 Smid M, Wang Y, Zhang Y, et al. Subtypes of breast cancer show preferential site of relapse. Cancer Res 2008; 68:3108–3114.
58 Driouch K, Bonin F, Sin S, et al. A six-gene signature predicting breast cancer lung metastasis. Reply. Cancer Res 2009; 69:9507–9511.
59 Rhodes DR, Ateeq B, Cao Q, et al. AGTR1 overexpression defines a subset of breast cancer and confers sensitivity to losartan, an AGTR1 antagonist. Proc Natl Acad Sci U S A 2009; 106:10284–10289.
60 Li J, Orlandi R, White CN, et al. Independent validation of candidate breast cancer serum biomarkers identified by mass spectrometry. Clin Chem 2005; 51:2229–2235.
61 Hudelist G, Pacher-Zavisin M, Singer CF, et al. Use of high-throughput protein array for profiling of differentially expressed proteins in normal and malignant breast tissue. Breast Cancer Res Treat 2004; 86:281–291.

breast cancer; classification; epigenetic alterations; genetic alterations

© 2011 Lippincott Williams & Wilkins, Inc.