Advertisement

Breast Cancer Research and Treatment

, Volume 117, Issue 1, pp 183–191 | Cite as

Expression profiling of familial breast cancers demonstrates higher expression of FGFR2 in BRCA2-associated tumors

  • Anita L. Bane
  • Dushanthi Pinnaduwage
  • Sarah Colby
  • Shelley B. Bull
  • Frances P. O’Malley
  • Irene L. AndrulisEmail author
Open Access
Brief Report

Abstract

Background BRCA1- and BRCA2-associated tumors appear to have distinct molecular signatures. BRCA1-associated tumors are predominantly basal-like cancers, whereas BRCA2-associated tumors have a predominant luminal-like phenotype. These two molecular signatures reflect in part the two cell types found in the terminal duct lobular unit of the breast. To elucidate novel genes involved in these two spectra of breast tumorigenesis we performed global gene expression analysis on breast tumors from germline BRCA1 and BRCA2 mutation carriers. Methodology Breast tumor RNAs from 7 BRCA1 and 6 BRCA2 mutation carriers were profiled using UHN human 19K cDNA microarrays. Supervised univariate analyses were conducted to identify genes differentially expressed between BRCA1 and BRCA2-associated tumors. Selected discriminatory genes were validated using real time reverse transcription polymerase chain reaction in the tumor RNAs, and/or by immunohistochemistry (IHC) or by in situ hybridization (ISH) on tissue microarrays (TMAs) containing an independent set of 58 BRCA1 and 64 BRCA2-associated tumors. Results Genes more highly expressed in BRCA1-associated tumors included stathmin, osteopontin, TGFβ2 and Jagged 1 in addition to genes previously identified as characteristic of basal-like breast cancers. BRCA2-associated cancers were characterized by the higher relative expression of FGF1 and FGFR2. FGFR2 protein was also more highly expressed in BRCA2-associated cancers (P = 0.004). Significance BRCA1-associated tumours demonstrated increased expression of component genes of the Notch and TGFβ pathways whereas the higher expression of FGFR2 and FGF1 in BRCA2-associated cancers suggests the existence of an autocrine stimulatory loop.

Keywords

BRCA1 BRCA2 Expression profiling Familial breast cancer 

Introduction

BRCA1- and BRCA2-associated tumors have many morphologic features in common. These include ductal histology, high histologic grade, pushing tumor margins and a notable host lymphocytic response [1, 2]. Despite these similarities they have distinct molecular signatures; BRCA1-associated tumors have been shown to be predominantly estrogen receptor negative and to have a basal phenotype [3, 4, 5], whereas we have shown previously that BRCA2-associated tumors have a luminal phenotype characterized by the expression of estrogen receptor and luminal-type cytokeratins [1]. These two molecular signatures reflect in part the two cell types, basal/myoepithelial and luminal found in the terminal duct lobular unit of the normal breast.

Basal-like human breast cancers have been shown to exhibit aberrations in the epidermal growth factor receptor (EGFR) signaling pathway [6]. With the exception of Cyclin D1, genes responsible for promoting the growth and survival of BRCA2-associated cancers have yet to be identified [7]. Our objective in performing global gene expression analysis on BRCA1- and BRCA2-associated tumors was to distinguish these distinct pathways of carcinogenesis and to elucidate novel genes necessary for the transformation and survival of BRCA2-associated tumors.

Materials and methods

Breast cancer cases

Flash frozen tumor tissue was available from 7 germline BRCA1 carriers and 6 germline BRCA2 carriers participating in the Ontario Familial Breast Cancer Registry (OFBCR) [8, 9] and the Ontario Cancer Genetics Network (OCGN). Following pathologic confirmation of invasive disease, the tumor tissue was frozen and stored in liquid nitrogen. Testing for germline mutations in BRCA1 and BRCA2 was performed using an RNA/DNA-based protein truncation test with complementary 5′ sequencing as previously described [10]. All mutations were confirmed by DNA sequencing. Mutations were classified as deleterious if they were protein-truncating, missense mutations (rare), or splice-site mutations as defined by the Breast Informatics Consortium (http://research.nhgri.nih.gov/bic/).

The clinicopathologic characteristics and mutation status (for the BRCA1-associated and BRCA2-associated tumors) are summarized in Table 1. Tumors were classified according to the WHO histologic classification of breast tumors [11] and graded by central pathology review using the Nottingham histologic grading system [12]. The pathologist was blinded to the mutational status or family history of the participants. Receptor status was obtained by either biochemical or immunohistochemical methods.
Table 1

Clinicopathologic characteristics of BRCA1- and BRCA2-associated breast cancers

LAB IDa

Mutation

Type of invasive cancer

Age at diagnosis

Grade

Lymphatic invasion

Blood vessel invasion

ERb

PRc

BRCA1 mutation

3435

2457C-T

Ductal, NSTd

50

3

Absent

Absent

Nege

Neg

4080

185delAG

Ductal, NST

40

3

Absent

Absent

Equf

Posg

4326

185delAG

Ductal, NST

35

3

Absent

Absent

NAh

Neg

1834

5293delAAAG

Ductal, NST

30

3

Present

Absent

Neg

NA

1693

5382insC

Ductal, NST

44

3

Absent

Absent

Equ

Pos

2528

4603G-T

Ductal, NST

33

3

NA

NA

Neg

Pos

3078

185delAG

Ductal, NST

32

3

Present

Present

Neg

Equ

BRCA2 mutation

4324

8765delAG

Ductal, NST

44

3

Present

Absent

Neg

Pos

4472

IVS16 + 3 A-C

Ductal, NST

44

3

Absent

NA

Neg

Neg

2172

6174delT

Ductal, NST

46

3

Present

NA

Pos

Pos

3627

9132delC

Ductal, NST

34

3

Present

Absent

Pos

Pos

4374

8765delAG

Ductal, NST

58

3

Present

Absent

Neg

Neg

3242

6764insA

Ductal, NST

75

3

Present

Absent

Pos

Equ

aID: Identification number

bER: Estrogen receptor

cPR: Progesterone receptor

dNST: No special type

eNeg: Negative

fEqu: Equivalent

gPos: Positive

hNA: Not available

A validation set of formalin-fixed paraffin-embedded (FFPE) tumors from 58 BRCA1 carriers, 64 BRCA2 carriers and 242 additional individuals from the OFBCR without BRCA1 or BRCA2 mutations was used to construct tissue microarrays (TMAs).

RNA extraction and reverse transcription

Total RNA was isolated from 20 to 50 mg of tumor tissue using TRIzol reagent (Gibco BRL Life Technologies). Reference RNA was a pool of 13 cell lines, modified from the common reference cell line list [13]. Tumor and reference RNAs (5 μg) were reverse transcribed with Superscript III reverse transcriptase (Invitrogen) to yield cDNA. Tumor and reference cDNA (5 μg) were indirectly labeled using aminoallyl nucleotide analogs with Cy3 and Cy5 fluorescent tags, respectively.

Dye labeling and cDNA array hybridisation

The labeled probes were simultaneously hybridized to UHN human 19K cDNA microarray slides (www.microarrays.ca), and incubated overnight at 42°C. These 19K cDNA microarrays are single-spotted and contain 19,008 characterized or unknown human expressed sequence tags (ESTs). The clone set has been sequence-verified at the UHN Microarray Centre, Toronto. Following hybridization the slides were washed and scanned using an Axon scanner. Fluorescent dye swap experiments were also performed for two tumors.

Pre-processing of expression data

The gene expression data were obtained from the original image files as spot intensities by correcting the mean foreground for each spot with the median local background. The array quality was controlled by requiring arrays to have more than 80% of spots with spot intensities higher than their local background and more than 75% of spots with spot intensity higher than 1.2 times their local background in both channels [14]. Spots with foreground intensity lower than background were treated as missing. A relative expression value was obtained for each gene as the log base 2 ratio of the adjusted intensity for the sample channel versus the reference channel. The log2 ratios were normalized by a within-subarray print-tip “loess” adjustment, followed by a between-array scale adjustment [15]. Poor quality spots as flagged by the GenePix image analysis software were excluded from the normalization. Pre-processing and normalization were carried out using the R-Bioconductor package LIMMA [16]; http://www.Bioconductor.org). There were 18,981 genes retained in the final data set for analysis.

TaqMan® assay-based real-time RT-PCR

mRNA expression levels of six selected genes that were statistically significantly differentially expressed between the two tumor groups and/or of biologic interest were measured in representative tumor specimens by real-time reverse transcription-polymerase chain reaction (RT-PCR) using TaqMan® PRISM 7500 Fast Real-Time PCR System (Applied Biosystems, Foster City, CA). Gene expression levels were determined by quantification relative to a control gene, hypoxanthine guanine phosphorbosyl transferase (HPRT). cDNA was generated using the ABI High Capacity cDNA Archiving Kit (Applied Biosystems, Foster City, CA) and RT-PCR reactions were carried out following the manufacturer’s protocol.

Tissue microarrays

Tissue microarrays were constructed as previously described [1] using FFPE tumors from 58 BRCA1 carriers, 64 BRCA2 carriers and 242 individuals without BRCA1 or BRCA2 mutations from the OFBCR. 4 μm thick sections of these blocks were used for immunohistochemical staining with the FGFR2 antibody (polyclonal, Abcam). 5 μm thick sections were used for mRNA in situ hybridization (ISH). ISH analysis of breast cancer has been reported in detail elsewhere previously [17]. Briefly, a cDNA probe for ISH was made to the gene of interest using 33P-UTP-radiolabeled cRNA. Using routine techniques the TMA sections were hybridized with the radiolabeled antisense probe, washed, and treated with Kodak NBT-2 nuclear emulsion. The sections from the TMAs were stored at 4°C for several weeks prior to development in Kodak D-19 solution; they were subsequently fixed in Kodafix and counterstained with 0.1% toludine blue.

Immunoreactivity and mRNA ISH was scored using the Allred method [18], which combined the intensity of staining with the percentage of positive tumor cells observable resulting in a combined score of 0–8. For FGFR2, a score of 7 or above was considered positive; for Jagged 1, a previously determined score of 4 or higher [17] was considered positive.

Statistical analysis

To identify genes that discriminate between BRCA1- and BRCA2-associated tumors, supervised univariate analyses of array-based log2 gene expression were performed for each clone treating the modified Student t-test as the primary analysis [19] using the SAM procedure implemented in open source software, version 2.1.0 (http://www.r-project.org/). To assess the sensitivity of the results to this analysis method the random variance test was also applied using BRB ArrayTools (http://linus.nci.nih.gov/BRB-ArrayTools.html). For each test, all clones were ranked according to their ability to discriminate BRCA1-associated tumors from BRCA2-associated tumors. Supplementary Table 1 reports discriminating genes with P-values ≤ 0.01 according to the modified Student t-test. P-values were not adjusted for multiple testing.

Log2 transformed relative expression of genes validated by RT-PCR was compared to array-based log2 gene expression via scatter plots. Means of the RT-PCR expressions were compared between the two BRCA tumor groups using the Student’s t-test and Welch unequal variance t-tests. All tests were two-sided. P-values were not adjusted for multiple testing.

The proportion of tumors positive by immunohistochemical staining and mRNA ISH for BRCA1-associated and BRCA2-associated tumors were compared using Fisher’s exact test for association. All tests were two sided. Statistical analysis was performed using SAS 9.1 software (SAS, Institute Inc., Cary, North Carolina). P-values were not adjusted for multiple testing.

Results

Clinicopathologic characteristics of BRCA1- and BRCA2-associated tumors

Thirteen tumors were obtained from individuals known to harbor a BRCA1 or BRCA2 germline mutation. As indicated in Table 1, all of the 13 BRCA-associated tumors were grade III/III invasive ductal, no special type tumors (NST). Both ER positive/equivalent and negative tumors were included among the BRCA1 and BRCA2 carriers. The mean age at diagnosis of the BRCA1 patients was somewhat younger than the mean age for the BRCA2 subjects (38 vs. 43 years).

Identification of genes that distinguish BRCA1- and BRCA2-associated breast tumors

The modified Student t-test was used to identify clones on the arrays that were differentially expressed between the BRCA1- and BRCA2-associated breast tumors. There were 4 distinguishing clones with a P-value < 0.001 and 127 clones with a P-value < 0.01. The 150 top-ranked differentially expressed genes with a P-value of ≤0.01 are listed in supplementary Table 1 and displayed as a heat map, Fig. 1.
Fig. 1

Unsupervised two dimensional cluster analysis of 7 BRCA1-associated and 6 BRCA2-associated breast tumors. Two dimensional presentation of transcript ratios for 13 tumors. The top 150 significant genes differentially expressed between the two tumor groups from SAM Moderated t-test are shown. Each column represents a tumor and each row a gene. As shown in the color bar, red indicates up regulation, green down regulation and black no change

BRCA1-associated tumors were characterized by the higher relative expression of 52 of these clones. Using the gene ontology database (NCBI), these genes was predicted to be involved in diverse cellular functions such as proliferation, angiogenesis, cell motility, cell adhesion, transcription and DNA repair. Involvement in the MAPK, Wnt, EGFR and TGFβ signaling pathways was identified. BRCA1-associated differentially expressed genes included stathmin/oncoprotein 18 (P = 0.0002), osteopontin (P = 0.002) and TGFβ2 (0.011).

BRCA2-associated tumors preferentially expressed 98 of the 150 clones. These genes were found to have functions related to transcription, signal transduction, cell proliferation, cell adhesion and extracellular matrix remodeling. Involvement in the MAPK signaling pathway was common. BRCA2-associated cancers were characterized by the higher relative expression of FGF1 (P = 0.003) and FGFR2 (P = 0.004).

Validation of differentially expressed genes by real time RT-PCR and IHC

To confirm the differences in gene expression, quantitative real-time RT-PCR was performed for several genes including FGF1, FGFR2, stathmin/oncoprotein 18, osteopontin and TGFβ2. Due to limited amounts of RNA, only 5 BRCA1- and 6 BRCA2-associated tumor RNAs were available for RT-PCR. Similar to the microarray data, the levels of expression of osteopontin and TGFβ2 were significantly higher in the BRCA1-associated tumors compared to the BRCA2-associated tumors (P = 0.03 and 0.001, respectively, Fig. 2). BRCA2-associated tumors expressed higher levels of FGF1 compared to BRCA1-associated tumors and BRCA1-associated tumors had higher levels of stathmin/oncoprotein 18 compare to BRCA2-associate tumors, similar to the microarray data; but these differences however, did not reach statistical significance at the 5% level (P = 0.76 and P = 0.28 respectively, data not shown).
Fig. 2

Scatter plots comparing the expression ratios of 6 genes obtained using both cDNA microarrays and RT-PCR in BRCA1-associated and BRCA2-associated tumors. The log2 expression ratios from RT-PCR (y-axis) and log2 expression ratios from cDNA microarrays (x-axis) for 6 genes from BRCA1-associated (red spots) and BRCA2-associated (purple spots) tumors are illustrated in the scatter plots. The relative expression ratios for both modalities for each tumor correlate positively i.e., RT-PCR and expression data go in the same direction (up regulated or down regulated) for the two tumor groups

Because the levels of FGFR2 were too low to quantitate reliably in the tumors using real time RT-PCR we used immunohistochemistry (IHC) to determine the level of expression of FGFR2 in an independent set of tumors. Tissue microarrays were constructed from FFPE tumors from 64 BRCA2 carriers, 58 BRCA1 carriers and 242 individuals without BRCA1 or BRCA2 mutations. A total of 19 out of 64 (30%) BRCA2-associated tumors stained positive for FGFR2 compared to 3 of 50 (6%) of BRCA1-associated tumors (P = 0.004, Fig. 3).
Fig. 3

Immunohistochemical and ISH studies on representative BRCA2-associated and BRCA1-associated tumors. Tumor (a) and (b) are a representative BRCA2-associated and BRCA1-associated tumor stained for FGFR2. The cytoplasmic stain is positive in tumor (a) as evidenced by the intense brown staining and negative in tumor (b). Tumors represented in (c) and (d) are representative BRCA2-associated and BRCA1-associated tumors following ISH for Jagged1 mRNA. Tumor (c) is negative, with a vessel acting as a positive internal control whereas tumor (d) is positive

Furthermore, 25 of 151 (17%) breast tumors from non-mutation carriers were positive for FGFR2, yielding statistically significant differences among the three tumor groups (P = 0.0007). Further analysis of all tumor groups suggested that FGFR2 expression was inversely correlated with the basal phenotype (as defined by the absence of ER, PR and HER2 expression and the presence of CK 5 or EGFR expression), with 9 of 65 (13%) basal-like tumors staining positively for FGFR2 expression compared with 36 of 136 (27%) non-basal-like breast cancers (P = 0.2). More specifically, FGFR2 negative tumors were more likely to be ER negative than FGFR2 positive tumors (49.8 vs. 26.7%, P = 0.004, Table 2). Similarly FGFR2 negative tumors were more likely to be negative for the luminal cytokeratin CK8/18 and p27 (14.9 vs. 2.2%, P = 0.02 and 36.8 vs. 13.8%, P = 0.01 respectively, Table 2) and more likely to be positive for the basal cytokeratin CK 5 and vimentin (32.0 vs. 19.2%, P = 0.08 and 20.3 vs. 5.9%, P = 0.04 respectively, Table 2). FGFR2 expression was also positively associated with PR expression (P = 0.01, Table 2).
Table 2

Association between FGFR2 expression and other IHC markers in all breast tumors (n = 248)

Other marker

FGFR2

Positive

Negative

P-value**

n*

%

n

%

 

ER

    Positive

33

73.3

102

50.3

0.0049

    Negative

12

26.7

101

49.8

 

PR

    Positive

30

63.8

88

43.4

0.0113

    Negative

17

36.2

115

56.7

 

Her2

    Positive

5

10.6

20

9.9

0.7936

    Negative

42

89.4

182

90.1

 

p27

    Positive

25

86.2

110

63.2

0.0152

    Negative

4

13.8

64

36.8

 

p53

    Positive

13

27.7

77

38.9

0.1511

    Negative

34

72.3

121

61.1

 

CK818

    Positive

44

97.8

172

85.2

0.0207

    Negative

1

2.2

30

14.9

 

CK5

    Positive

9

19.2

65

32.0

0.0815

    Negative

38

80.9

138

68.0

 

Vimentin

    Positive

2

5.9

36

20.3

0.0445

    Negative

32

94.1

141

79.7

 

n: number

** Unadjusted P-values are given

Bold values are significant at the 5% level

BRCA1-associated tumors

Since BRCA1-associated tumors cluster with basal-like cancers [5, 6, 20], we evaluated whether genes previously found to be more highly expressed in basal-like tumors were differentially expressed in the BRCA1 compared to the BRCA2 groups. Microarray based values of keratin 17, vimentin and caveolin1 tended to be higher in BRCA1-associated tumors; however, the differences in expression of these genes (P = 0.095, 0.067, and 0.233, respectively) did not meet our global selection cut-off.

Recently we have found that basal-like tumors tend to exhibit higher expression of Jagged 1 [21], a ligand involved in Notch pathway signaling. The Notch pathway has been shown to regulate a number of processes important in cancer including angiogenesis [22], epithelial to mesenchymal transition, and stem-cell like characteristics [23]. Using RT-PCR we found that the mean expression level of Jagged 1 tended to be elevated in BRCA1-compared to BRCA2-associated tumors (2.07 versus 0.50, P = 0.11). Because only 11 tumors were available for RT-PCR, we examined the expression of Jagged 1 by mRNA ISH on the TMAs. Of 48 BRCA1-associated tumors, 17 (35%) were positive for Jagged1 by ISH compared to 5 of 53 (9%) BRCA2-associated cancers (P = 0.02, Fig. 3).

Discussion

In recent years, gene expression profiling of breast cancers has improved our understanding of the heterogeneity of the disease and generated hypotheses concerning the development and progression of these cancers. Molecular signatures for BRCA1-associated tumors have been delineated, but BRCA2-associated tumors have been less well studied [5, 7]. To elucidate novel genes involved in the development and progression of BRCA2-associated tumors and to distinguish distinct pathways of carcinogenesis in BRCA1 and BRCA2-associated tumors, we compared gene expression microarray patterns in tumors from BRCA1 and BRCA2 carriers.

We found that FGFR2 and FGF1 were more highly expressed in BRCA2-associated cancers as compared to BRCA1-associated breast cancers, suggesting the presence of an autocrine growth stimulatory loop. FGF1 and FGFR2 belong to a large family of ligands and receptor tyrosine kinases [24]. FGF1 is a mitogen that signals through FGFRs and subsequently activates the MAPK signaling cascade. Overexpression of FGF1 in breast cancer cell lines has been shown to result in increased anchorage independent growth and reduced requirement for estrogen in vitro and to increased tumorigenicity, angiogenesis and metastatic behaviour in vivo [25, 26]. FGF1 is located at 5q31 a region of the genome often subject to loss in BRCA1-associated cancers [27] and basal-like cancers [28].

FGFR2, located at 10q26, encodes at least two receptor isoforms FGFR2-IIIb and FGFR2-IIIc. FGFR2-IIIc is mainly expressed in tissues of mesenchymal origin, whereas FGFR2-IIIb is expressed in the epithelium of many organs including the mammary gland [29, 30]. The FGFRs in general and FGFR2 specifically have roles in embryogenesis, development and carcinogenesis. FGFR2 mRNA is expressed in many carcinoma cell lines [31, 32, 33] and in the breast cancer cell lines examined, FGFR2 expression contributes to their invasive phenotype. Similarly, overexpression of the receptor in normal human mammary epithelial (HME) cells leads to their transformation [34]. These properties are attributable to the activation of the mitogen-activated protein kinase (MAPK) and phosphatidylinositol 3-kinase (PI3K) signaling cascades [34]. Moreover, the gene locus (10q26) is amplified in 2–10% of breast cancers and its expression has been reported in 50–100% of human breast cancers, with high-level expression confined to 4%–12% of cases [35, 36]. It has also been associated with estrogen (ER) and progesterone (PR) receptor expression and improved overall and disease free survival [35]. Moreover in two recent independent genome wide association studies (GWAS) single nucleotide polymorphisms (SNPs) within intron 2 of FGFR2 have been causally associated with increased risk of both familial and sporadic breast cancer [37, 38]. Furthermore this susceptibility locus was associated with younger age of onset and bilateral disease [37]. The authors speculate that the SNPs have functional effects and that the association with breast cancer risk is mediated through regulation of FGFR2 expression possibly through interaction with the ER. In our study we have shown that FGFR2 expression is positively associated with the BRCA2 genotype, a tumor group that we have previously shown to be predominantly ER-positive luminal-type tumors [1]. In addition, we found that FGFR2 expression was positively associated with PR expression, a weak prognostic and predictive factor in breast cancer and often regarded as indicative of a functional ER pathway [39]. Conversely, FGFR2 expression was negatively correlated with basal-like breast cancers which are known to be ER negative and to carry an adverse prognosis [5, 6].

Our expression studies support previous reports indicating that BRCA1-associated breast cancers have a basal-like profile [3, 5]. In the BRCA1-associated tumor group, we detected elevated expression of a number of genes, including keratin 17 [5, 6], vimentin, and caveolin 1 [20], that have previously been linked to the basal-like cancers. Our results also suggest that Y-box binding protein-1 is more highly expressed in BRCA1-associated breast cancers (P = 0.001). This protein has recently been demonstrated to transcriptionally induce EGFR which is commonly overexpressed in basal-like cancers [40]. In addition we found that stathmin/oncoprotein 18 and osteopontin were overexpressed in BRCA1-associated tumors relative to BRCA2-associated tumors. The gene for stathmin, located at 1p36, is one of the 70 genes that compose the ‘70 gene classifier’ that predicts poor prognosis in sporadic breast cancer [41] a classifier that correlates closely with the poor performing subgroups from the intrinsic gene set of which basal-like cancers are one [42]. Stathmin is a microtubule depolymerizing protein involved in cell cycle progression and cell motility. It is highly expressed in a number of human malignancies, including breast cancers where it is negatively correlated with ER expression and positively correlated with grade, aneuploidy, proliferation and mutant p53 [43], all characteristic features of BRCA1-associated and basal-like cancers [3, 5, 6].

Osteopontin, a secreted phosphoprotein, has been shown to interact with a diverse range of factors including integrins, CD44, TGFα, EGFR, Met and VEGF leading to the enhancement of cellular migration, invasion, survival and angiogenesis. [44, 45, 46]. Osteopontin is expressed in a wide variety of human malignancies and its expression has been reported to be correlated with poor prognosis in breast cancer [47].

Novel results from our expression study include the identification of members of the Notch and TGFβ signaling pathways, Jagged 1 and TGFβ2 respectively, as being more highly expressed in BRCA1-associated tumors. Both of these pathways are highly conserved through evolution and play important roles in development, differentiation and tumorigenesis [48, 49]. Moreover, both pathways are known to play a role in mammary stem cell maintenance or renewal [50, 51]. Dontu et al. have shown that Notch signaling can act on mammary stem cells to promote self-renewal, on early progenitor cells to promote proliferation, and on multipotent progenitor cells, facilitating myoepithelial cell lineage specific commitment and proliferation [50]. Furthermore, patients whose breast cancers express Jagged 1 have been found to have a poorer prognosis [17]. Shipitsin and colleagues showed that TGFβ signaling is upregulated in normal breast stem cells and their malignant counterparts and has prognostic effects [51].

In conclusion, our molecular profiling demonstrates that BRCA1-associated and BRCA2-associated tumors have distinct molecular profiles. Our results confirm the known link between BRCA1-associated tumors and the basal-like signature and highlight a link between the Notch and TGFβ pathways and BRCA1-associated tumorigenesis. In contrast the data suggest that BRCA2-associated tumors express higher levels of FGFR2 and FGF1, suggesting the existence of an autocrine loop leading to downstream pleiotrophic cellular effects.

Notes

Acknowledgements

The authors thank Dr. Brendan Dickson and Hui Zhang for their technical assistance. Grant numbers and sources of support: This work was supported by the Canadian Breast Cancer Foundation and the National Cancer Institute under RFA #CA-95-003 as part of the Breast Cancer Family Registry (BreastCFR).

Open Access

This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.

Supplementary material

10549_2008_87_MOESM1_ESM.pdf (37 kb)
(PDF 38 kb)

References

  1. 1.
    Bane AL, Beck JC et al (2007) BRCA2 mutation-associated breast cancers exhibit a distinguishing phenotype based on morphology and molecular profiles from tissue microarrays. Am J Surg Pathol 31:121–128. doi: 10.1097/01.pas.0000213351.49767.0f PubMedCrossRefGoogle Scholar
  2. 2.
    Lakhani SR, Jacquemier J et al (1998) Multifactorial analysis of differences between sporadic breast cancers and cancers involving BRCA1 and BRCA2 mutations. J Natl Cancer Inst 90:1138–1145. doi: 10.1093/jnci/90.15.1138 PubMedCrossRefGoogle Scholar
  3. 3.
    Lakhani SR, Reis-Filho JS et al (2005) Prediction of BRCA1 status in patients with breast cancer using estrogen receptor and basal phenotype. Clin Cancer Res 11:5175–5180. doi: 10.1158/1078-0432.CCR-04-2424 PubMedCrossRefGoogle Scholar
  4. 4.
    Palacios J, Honrado E et al (2005) Phenotypic characterization of BRCA1 and BRCA2 tumors based in a tissue microarray study with 37 immunohistochemical markers. Breast Cancer Res Treat 90:5–14. doi: 10.1007/s10549-004-1536-0 PubMedCrossRefGoogle Scholar
  5. 5.
    Sorlie T, Tibshirani R et al (2003) Repeated observation of breast tumor subtypes in independent gene expression data sets. Proc Natl Acad Sci USA 100:8418–8423. doi: 10.1073/pnas.0932692100 PubMedCrossRefGoogle Scholar
  6. 6.
    Nielsen TO, Hsu FD et al (2004) Immunohistochemical and clinical characterization of the basal-like subtype of invasive breast carcinoma. Clin Cancer Res 10:5367–5374. doi: 10.1158/1078-0432.CCR-04-0220 PubMedCrossRefGoogle Scholar
  7. 7.
    Hedenfalk I, Duggan D et al (2001) Gene-expression profiles in hereditary breast cancer. N Engl J Med 344:539–548. doi: 10.1056/NEJM200102223440801 PubMedCrossRefGoogle Scholar
  8. 8.
    John EM, Hopper JL et al (2004) The breast cancer family registry: an infrastructure for cooperative multinational, interdisciplinary and translational studies of the genetic epidemiology of breast cancer. Breast Cancer Res 6:R375–R389. doi: 10.1186/bcr801 PubMedCrossRefGoogle Scholar
  9. 9.
    Sutherland HJ, Lacroix J et al (2001) The cooperative familial registry for breast cancer studies: design and first year recruitment rates in Ontario. J Clin Epidemiol 54:93–98. doi: 10.1016/S0895-4356(00)00263-8 PubMedCrossRefGoogle Scholar
  10. 10.
    Andrulis IL, Anton-Culver H et al (2002) Comparison of DNA- and RNA-based methods for detection of truncating BRCA1 mutations. Hum Mutat 20:65–73. doi: 10.1002/humu.10097 PubMedCrossRefGoogle Scholar
  11. 11.
    Tavassoli T, Devilee P (2003) Pathology & Genetics. Tumours of the breast and female genital organs. IARC Press, LyonGoogle Scholar
  12. 12.
    Elston CW, Ellis IO (1991) Pathological prognostic factors in breast cancer. I. The value of histological grade in breast cancer: experience from a large study with long-term follow-up. Histopathology 19:403–410. doi: 10.1111/j.1365-2559.1991.tb00229.x PubMedCrossRefGoogle Scholar
  13. 13.
    Perou CM, Sorlie T et al (2000) Molecular portraits of human breast tumours. Nature 406:747–752. doi: 10.1038/35021093 PubMedCrossRefGoogle Scholar
  14. 14.
    Simon R, Korn E (2003) Design and analysis of DNA microarray investigations. Springer-Verlag, New YorkGoogle Scholar
  15. 15.
    Yang YH, Dudoit S et al (2002) Normalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation. Nucl Acids Res 30:e15. doi: 10.1093/nar/30.4.e15 PubMedCrossRefGoogle Scholar
  16. 16.
    Smyth GK (2005) Limma: linear models for microarray data. In: Bioinformatic and computional biology solutions using R and bioconductor. Springer, New YorkGoogle Scholar
  17. 17.
    Reedijk M, Odorcic S et al (2005) High-level coexpression of JAG1 and NOTCH1 is observed in human breast cancer and is associated with poor overall survival. Cancer Res 65:8530–8537. doi: 10.1158/0008-5472.CAN-05-1069 PubMedCrossRefGoogle Scholar
  18. 18.
    Allred DC, Harvey JM et al (1998) Prognostic and predictive factors in breast cancer by immunohistochemical analysis. Mod Pathol 11:155–168PubMedGoogle Scholar
  19. 19.
    Tusher VG, Tibshirani R et al (2001) Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci USA 98:5116–5121. doi: 10.1073/pnas.091062498 PubMedCrossRefGoogle Scholar
  20. 20.
    Pinilla SM, Honrado E et al (2006) Caveolin-1 expression is associated with a basal-like phenotype in sporadic and hereditary breast cancer. Breast Cancer Res Treat 99:85–90. doi: 10.1007/s10549-006-9184-1 PubMedCrossRefGoogle Scholar
  21. 21.
    Reedijk M, Pinnaduwage D et al (2007) JAG1 expression is associated with a basal phenotype and recurrence in lymph node-negative breast cancer. Breast Cancer Res Treat, Nov 8 [Epub]Google Scholar
  22. 22.
    Uyttendaele H, Ho J et al (2001) Vascular patterning defects associated with expression of activated Notch4 in embryonic endothelium. Proc Natl Acad Sci USA 98:5643–5648. doi: 10.1073/pnas.091584598 PubMedCrossRefGoogle Scholar
  23. 23.
    Varnum-Finney B, Purton LE et al (1998) The notch ligand, jagged-1, influences the development of primitive hematopoietic precursor cells. Blood 91:4084–4091PubMedGoogle Scholar
  24. 24.
    Dailey L, Ambrosetti D et al (2005) Mechanisms underlying differential responses to FGF signaling. Cytokine Growth Factor Rev 16:233–247. doi: 10.1016/j.cytogfr.2005.01.007 PubMedCrossRefGoogle Scholar
  25. 25.
    McLeskey SW, Ding IY et al (1994) MDA-MB-134 breast carcinoma cells overexpress fibroblast growth factor (FGF) receptors and are growth-inhibited by FGF ligands. Cancer Res 54:523–530PubMedGoogle Scholar
  26. 26.
    Zhang L, Kharbanda S et al (1997) MCF-7 breast carcinoma cells overexpressing FGF-1 form vascularized, metastatic tumors in ovariectomized or tamoxifen-treated nude mice. Oncogene 15:2093–2108. doi: 10.1038/sj.onc.1201386 PubMedCrossRefGoogle Scholar
  27. 27.
    van Beers EH, van Welsem T et al (2005) Comparative genomic hybridization profiles in human BRCA1 and BRCA2 breast tumors highlight differential sets of genomic aberrations. Cancer Res 65:822–827PubMedGoogle Scholar
  28. 28.
    Bergamaschi A, Kim YH et al (2006) Distinct patterns of DNA copy number alteration are associated with different clinicopathological features and gene-expression subtypes of breast cancer. Genes Chromosomes Cancer 45:1033–1040. doi: 10.1002/gcc.20366 PubMedCrossRefGoogle Scholar
  29. 29.
    De Moerlooze L, Spencer-Dene B et al (2000) An important role for the IIIb isoform of fibroblast growth factor receptor 2 (FGFR2) in mesenchymal-epithelial signalling during mouse organogenesis. Development 127:483–492PubMedGoogle Scholar
  30. 30.
    Mailleux AA, Spencer-Dene B et al (2002) Role of FGF10/FGFR2b signaling during mammary gland development in the mouse embryo. Development 129:53–60PubMedGoogle Scholar
  31. 31.
    Bansal GS, Cox HC et al (1997) Expression of keratinocyte growth factor and its receptor in human breast cancer. Br J Cancer 75:1567–1574PubMedGoogle Scholar
  32. 32.
    Carstens RP, Eaton JV et al (1997) Alternative splicing of fibroblast growth factor receptor 2 (FGF-R2) in human prostate cancer. Oncogene 15:3059–3065. doi: 10.1038/sj.onc.1201498 PubMedCrossRefGoogle Scholar
  33. 33.
    Otte JM, Schmitz F et al (2000) Expression of keratinocyte growth factor and its receptor in colorectal cancer. Eur J Clin Invest 30:222–229. doi: 10.1046/j.1365-2362.2000.00617.x PubMedCrossRefGoogle Scholar
  34. 34.
    Moffa AB, Tannheimer SL et al (2004) Transforming potential of alternatively spliced variants of fibroblast growth factor receptor 2 in human mammary epithelial cells. Mol Cancer Res 2:643–652PubMedGoogle Scholar
  35. 35.
    Blanckaert VD, Hebbar M et al (1998) Basic fibroblast growth factor receptors and their prognostic value in human breast cancer. Clin Cancer Res 4:2939–2947PubMedGoogle Scholar
  36. 36.
    Penault-Llorca F, Bertucci F et al (1995) Expression of FGF and FGF receptor genes in human breast cancer. Int J Cancer 61:170–176. doi: 10.1002/ijc.2910610205 PubMedCrossRefGoogle Scholar
  37. 37.
    Easton DF, Pooley KA et al (2007) Genome-wide association study identifies novel breast cancer susceptibility loci. Nature 447(7148):1087–1093PubMedCrossRefGoogle Scholar
  38. 38.
    Hunter DJ, Kraft P et al (2007) A genome-wide association study identifies alleles in FGFR2 associated with risk of sporadic postmenopausal breast cancer. Nat Genet 39:870–874. doi: 10.1038/ng2075 PubMedCrossRefGoogle Scholar
  39. 39.
    Fitzgibbons PL, Page DL et al (2000) Prognostic factors in breast cancer. College of American pathologists consensus statement 1999. Arch Pathol Lab Med 124:966–978PubMedGoogle Scholar
  40. 40.
    Stratford AL, Habibi G et al (2007) Epidermal growth factor receptor (EGFR) is transcriptionally induced by the Y-box binding protein-1 (YB-1) and can be inhibited with Iressa in basal-like breast cancer providing a potential target for therapy. Breast Cancer Res 9:R61. doi: 10.1186/bcr1767 PubMedCrossRefGoogle Scholar
  41. 41.
    van de Vijver MJ, He YD et al (2002) A gene-expression signature as a predictor of survival in breast cancer. N Engl J Med 347:1999–2009. doi: 10.1056/NEJMoa021967 PubMedCrossRefGoogle Scholar
  42. 42.
    Fan C, Oh DS et al (2006) Concordance among gene-expression-based predictors for breast cancer. N Engl J Med 355:560–569. doi: 10.1056/NEJMoa052933 PubMedCrossRefGoogle Scholar
  43. 43.
    Brattsand G (2000) Correlation of oncoprotein 18/stathmin expression in human breast cancer with established prognostic factors. Br J Cancer 83:311–318. doi: 10.1054/bjoc.2000.1264 PubMedCrossRefGoogle Scholar
  44. 44.
    Shijubo N, Uede T et al (2000) Vascular endothelial growth factor and osteopontin in tumor biology. Crit Rev Oncog 11:135–146PubMedGoogle Scholar
  45. 45.
    Tuck AB, Elliott BE et al (2000) Osteopontin-induced, integrin-dependent migration of human mammary epithelial cells involves activation of the hepatocyte growth factor receptor (met). J Cell Biochem 78:465–475. doi:10.1002/1097-4644(20000901)78:3<465::AID-JCB11>3.0.CO;2-CPubMedCrossRefGoogle Scholar
  46. 46.
    Tuck AB, Hota C et al (2003) Osteopontin-induced migration of human mammary epithelial cells involves activation of EGF receptor and multiple signal transduction pathways. Oncogene 22:1198–1205. doi: 10.1038/sj.onc.1206209 PubMedCrossRefGoogle Scholar
  47. 47.
    Allan AL, George R et al (2006) Role of the integrin-binding protein osteopontin in lymphatic metastasis of breast cancer. Am J Pathol 169:233–246. doi: 10.2353/ajpath.2006.051152 PubMedCrossRefGoogle Scholar
  48. 48.
    Moses HL, Serra R (1996) Regulation of differentiation by TGF-beta. Curr Opin Genet Dev 6:581–586. doi: 10.1016/S0959-437X(96)80087-6 PubMedCrossRefGoogle Scholar
  49. 49.
    Roberts AB, Wakefield LM (2003) The two faces of transforming growth factor beta in carcinogenesis. Proc Natl Acad Sci USA 100:8621–8623. doi: 10.1073/pnas.1633291100 PubMedCrossRefGoogle Scholar
  50. 50.
    Dontu G, Jackson KW et al (2004) Role of notch signaling in cell-fate determination of human mammary stem/progenitor cells. Breast Cancer Res 6:R605–R615. doi: 10.1186/bcr920 PubMedCrossRefGoogle Scholar
  51. 51.
    Shipitsin M, Campbell LL et al (2007) Molecular definition of breast tumor heterogeneity. Cancer Cell 11:259–273. doi: 10.1016/j.ccr.2007.01.013 PubMedCrossRefGoogle Scholar

Copyright information

© The Author(s) 2008

Authors and Affiliations

  • Anita L. Bane
    • 1
    • 2
    • 3
  • Dushanthi Pinnaduwage
    • 4
  • Sarah Colby
    • 4
  • Shelley B. Bull
    • 4
    • 5
  • Frances P. O’Malley
    • 1
    • 3
  • Irene L. Andrulis
    • 1
    • 2
    • 3
    • 6
    Email author
  1. 1.Department of Pathology and Laboratory MedicineMount Sinai HospitalTorontoCanada
  2. 2.Fred A. Litwin Centre for Cancer Genetics, #984 Samuel Lunenfeld Research InstituteMount Sinai HospitalTorontoCanada
  3. 3.Department of Pathology and Laboratory MedicineUniversity of TorontoTorontoCanada
  4. 4.Prosserman Center for Health Research, Samuel Lunenfeld Research InstituteMount Sinai HospitalTorontoCanada
  5. 5.Department of Public Health SciencesUniversity of TorontoTorontoCanada
  6. 6.Department of Molecular and Medical GeneticsUniversity of TorontoTorontoCanada

Personalised recommendations