Abstract
BACKGROUND
Expression of Fhit and Wwox proteins, tumor suppressors encoded by fragile loci FRA3B and FRA16D, are concordantly lost in breast cancers. The current study examined correlations among Fhit, Wwox, transcription factors AP2α and AP2γ, cytokeratins 5/6 (CK5/6), epidermal growth factor receptor (EGFR), estrogen receptor (ER), progesterone receptor (PR), HER2 and their associations with breast cancer phenotypes.
METHODS
Tissue microarrays constructed from 837 breast cancer blocks were immunostained. Expression in >10% of tumor cells was considered positive for cytoplasmic CK5/6, membranous EGFR, nuclear AP2α and AP2γ. Cytoplasmic Fhit and Wwox staining was scored according to staining intensity. ER, PR and HER2 status of tumors was from records. Correlations among immunohistochemical markers and tumor subtypes were assessed by univariate and multivariate statistical methods.
RESULTS
Triple negative tumors showed more frequent expression of EGFR, CK5/6 (p<0.001) and AP2γ (p = 0.003) and more frequent loss of Fhit and Wwox (p<0.001), with inverse correlation between Fhit, Wwox and EGFR, ER, PR expression (p<0.001). Reduced Fhit expression was more common in HER2 and AP2γ positive cases (p<0.001, p=0.002). There was direct correlation between Fhit and Wwox (p<0.001) and a borderline positive relation between AP2α and γ (p=0.054).
CONCLUSIONS
Results suggest that reduced Fhit, Wwox and nuclear AP2γ expression have roles in pathogenesis of basal-like differentiation in breast cancer. Alteration of expression of fragile site genes occurs in most of these cancers and may contribute to defects in DNA repair, as observed in BRCA1-deficient cancers. Thus DNA damage response checkpoint proteins could be targets for treatment.
Keywords: breast cancer subtypes, triple negative breast cancers, common fragile sites, Fhit, Wwox, AP2 transcription factors
INTRODUCTION
Common chromosomal fragile sites are preferential targets of replication stress in preneoplastic lesions, resulting in deletions involving fragile genes encoded at these conserved chromosome regions1,2. It has been suggested that damage to fragile sites, with lost function of genes located at these sites, may play a significant role in carcinogenesis3. Reduced expression of the fragile tumor suppressors, Fhit and Wwox, encoded by fragile loci FRA3B and FRA16D at chromosomes 3p14.2 and 16q23.3, are concordantly lost in about two thirds of breast cancers4. Expression levels of Fhit and Wwox have been associated with important aspects of breast cancer initiation, response to therapy, and clinicopathological features4-7. Their coordinated loss has also been shown in other types of cancers and cells8,9.
Our group and others have shown that Fhit and Wwox loss is more common in breast cancers that are negative for ER and PR4,7 and that Fhit loss is significantly more frequent in HER2 positive DCIS5. In another study, we also observed that expression levels of Wwox, Wwox-interacting protein AP2γ, ErbB410,11, and the AP2γ transcriptional target protein HER212, correlated with tamoxifen resistance in ER-positive invasive breast cancers6. Lost or reduced expression of Wwox, high-level expression of AP2γ and HER2, and ErbB4 loss were found to be independent markers of tamoxifen resistance. Reduced Wwox expression was better than PR in prediction of resistance, and nuclear AP2γ expression was better than HER26.
The recently described, molecularly based classification of breast cancers, has defined three groups with negative ER status: basal-like, HER2 over-expressing, and normal breast-like cancers13,14. Understanding the pathogenic mechanisms of the basal-like and triple-negative (i.e. ER, PR and HER2-negative) breast cancer subtypes is particularly important because, contrary to hormone receptor-positive or HER2-positive disease, no targeted therapies have been yet developed for these tumors and chemotherapy remains the only therapeutic option for these patients15-17. Both triple negative and basal-like cancers (which mostly show a triple negative phenotype), preferentially affect young women and women of African origin, are usually of high histological grade and show more aggressive clinical behavior16. Overlap between biological and clinical characteristics of sporadic triple negative/basal-like cancers and cancers in BRCA1 mutation carriers has been demonstrated18,19 and the basal cell-like molecular breast cancer subtype is an independent predictor of distant metastasis in African American women17.
The current study was designed to determine whether alterations in expression of genes at fragile sites contribute preferentially to the pathogenesis of certain subtypes of breast cancer. We determined the expression levels for tumor suppressor genes FHIT and WWOX (located at the two most active fragile sites of human genome), and AP2α and AP2γ (which physicially interact with Wwox), in a large cohort of breast cancers which we attempted to classify molecularly according to their expression of hormone receptors, HER2, and CK5/6 and EGFR, markers of basal-like differentiation20,21.
MATERIALS AND METHODS
Construction of TMAs
The study had the approval of the Ohio State University Institutional Review Board for studies of human subjects. TMAs were prepared from cores of the tissue blocks of 837 female breast cancers from the four major hospitals in Columbus, Ohio. The clinical information for the patients, histological tumor type, grade (as per the modified Bloom-Richardson system) and ER, PR and HER2 status of tumors were taken from clinical records and pathology reports. ER and PR were clinically evaluated by immunohistochemistry and cases with nuclear staining in >1% of cells were regarded as positive. HER2 was clinically evaluated by a standard combined immunohistochemistry plus FISH approach. All cases with intense, complete membranaous staining in more than 10% of invasive neoplastic cells by immunohistochemistry or presence of HER2 gene amplification by FISH test were considered HER2 positive.
Immunohistochemical analysis
Fhit, Wwox, CK5/6, EGFR, AP2α and AP2γ immunostainings were performed on TMA sections. The specific antisera used, staining methods, detection kits and controls are listed in Table 1. Antigen retrieval was the same for all antisera with the exception of EGFR and CK5/6; sections were boiled in pH 6 citrate buffer in a vegetable steamer for 25 min and cooled in solution for 15 min. EGFR antigen retrieval utilized Proteinase K from the PharmDx kit. Antigen retrieval for CK5/6 was in EDTA buffer, pH 9.0, in a vegetable steamer for 25 min with cooling in solution for 15 min. Details of immunohistochemical (IHC) staining methods have been described previously4,5.
TABLE 1.
Primary antisera and detection kits used in immunohistochemical studies
| Primary antiserum | Source | Description | Dilution | Detection Kit |
|---|---|---|---|---|
| Fhit | Huebner lab* | Rabbit polyclonal | 1/ 5000 | Vector Vectastain Elite kit |
| Wwox | Huebner lab* | rabbit polyclonal | 1/7000 | Vector Vectastain Elite kit |
| AP2γ | Santa Cruz | mouse monoclonal (6E4/4) | 1/75 | Dako Dako universal LSAB+ |
| AP2α | Santa Cruz | mouse monoclonal ( 3B5) | 1/75 | Dako Dako universal LSAB+ |
| CK5/6 | Dako | Mouse monoclonal (D5/16 B4) | 1/50 | Dako Dako universal LSAB+ |
| EGFR | Dako PharmDx Kit | Mouse monoclonal (2-18C9) | Neat | Dako envision plus, included in kit |
Expression of Fhit, Wwox and CK5/6 was cytoplasmic, AP2α, AP2γ nuclear and EGFR cytoplasmic and membranous. Cytoplasmic expression in ≥10% of tumor cells for CK5/6, membranous staining in ≥10% of tumor cells for EGFR and nuclear staining in ≥10% for AP2α, AP2γ, was accepted as positive. Fhit and Wwox staining in tumors was scored according to staining intensity: absent, highly reduced, reduced staining or strong expression. Three pathologists scored the slides and were blinded to the subtype of breast cancer.
The 837 breast cancers were divided into subtypes of breast cancer as defined by their IHC profiles using the designated stains15,20,21. These include luminal A (ER and/or PR positive, HER2 negative); luminal B (ER and/or PR positive, HER2 positive); HER2 over-expressing (ER and PR negative, HER2 positive); basal-like triple negative (ER, PR, HER2 negative and CK5/6 and/or EGFR positive); and other triple negative (ER, PR, HER2, CK5/6 and EGFR negative)
Statistics
Associations among specific markers and specific clinical features were analyzed using {chi}2 or Fisher tests, where appropriate, for univariate analyses, and multiple logistic regression for multivariate analysis, as described previously6. Pairwise correlations of biomarkers were assessed using Spearman’s correlation test. P values of 0.05 were considered to indicate statistical significance.
RESULTS
Characteristics of the breast cancers
The demographic characteristics and the various IHC results are described in Table 2. The median age was 55 years (range 20 to 93 years) and 100 (12%) were under age 40 years. Seven hundred and fifty-nine (90%) were white, 59 (7%) black and the remaining of other races/ethinicities. These reflect the catchment areas of the four contributing hospitals. Seven hundred and forty-nine (89.5%) had invasive ductal, NOS, 34 (4.1%) invasive lobular and 24 (2.9%) had mixed ductal and lobular, 12 (1.4%) mucinous and 18 (2.1%) other specific types of invasive breast cancer, such as medullary, tubular or papillary carcinoma.
TABLE 2.
Clinicopathological features of tumors, biomarker expression levels and associations with Fhit and Wwox expression
| Fhit | Wwox | |||||
|---|---|---|---|---|---|---|
| Lost or highly reduced | Strong or moderately reduced | P | Lost or highly reduced | Strong or moderately reduced | P | |
| Age | ||||||
| ≤40 | 69 (84.1%) | 13 (15.9%) | <0.001 | 79 (79.8%) | 20 (20.2%) | 0.512 |
| 41-50 | 117(70.9%) | 48(29.1%) | - | 155 (79.1%) | 41 (20.9%) | - |
| >50 | 265 (59.0%) | 184(41.0%) | - | 398 (75.8%) | 127 (24.2%) | - |
| Histologic type | ||||||
| Invasive Ductal, NOS | 414 (65.7%) | 216 (34.3%) | 0.010 | 579 (78.5%) | 159 (21.5%) | 0.036 |
| Invasive Lobular | 25 (54.3%) | 21 (45.7%) | - | 38 (64.4%) | 21 (35.6%) | - |
| Other invasive | 6 (54.5%) | 5 ( 45.5%) | - | 7 ( 58.3%) | 5 ( 41.7%) | - |
| Histologic grade | ||||||
| 1 | 33 (36.7%) | 57 (63.3%) | <0.001 | 85 (72.0%) | 33 (28.0%) | <0.001 |
| 2 | 155 (59.4%) | 106 (40.6%) | - | 225 (72.6%) | 85 (27.4%) | - |
| 3 | 241 (78.2%) | 67 (21.8%) | - | 291(84.6%) | 53 (15.4%) | - |
| Race | ||||||
| Caucasians | 402 (63.0%) | 236 (37.0%) | 0.001 | 574 (76.0%) | 181 (24.0%) | 0.036 |
| African Americans | 44 (84.6%) | 8 (15.4%) | - | 52 (88.1%) | 7 (11.9%) | - |
| ER | ||||||
| positive | 297 (57.9%) | 261 (42.1%) | <0.001 | 443 (73.0%) | 164 (27.0%) | <0.001 |
| negative | 154 (84.2%) | 29 ( 15.8%) | - | 187 (88.2%) | 25 (11.8%) | - |
| PR | ||||||
| positive | 237 (55.4%) | 191 (44.6%) | <0.001 | 373 (73.0%) | 138 (27.0%) | 0.001 |
| negative | 214 (79.9%) | 54 ( 20.1%) | - | 258 (83.5%) | 51 (16.5%) | - |
| HER2 | ||||||
| positive | 121 (77.1%) | 36 (22.9%) | <0.001 | 136 (73.5%) | 49 (26.5%) | 0.233 |
| negative | 324 (61.1%) | 206 (38.9%) | - | 489 (78.0%) | 138 (22.8%) | - |
| CK5/6 | ||||||
| positive | 95 (74.2%) | 33 (25.8%) | 0.021 | 110 (82.1%) | 24 (17.9%) | 0.071 |
| negative | 309 (63.1%) | 181 (36.9%) | - | 399 (74.6%) | 136 (25.4%) | - |
| EGFR | ||||||
| positive | 108 (85%) | 19 (15%) | <0.001 | 123 (91.1%) | 12 (8.9%) | <0.001 |
| negative | 312 (59.8%) | 210 (40.2%) | - | 430 (73.0%) | 155 (27.0%) | - |
| AP2α | ||||||
| positive | 368 (64.7%) | 201 (35.3%) | 0.584 | 437 (73.8%) | 155 (26.2%) | 0.607 |
| negative | 55 (70.5%) | 23 (29.5%) | - | 70 (76.9%) | 21 (23.1%) | - |
| AP2γ | ||||||
| positive | 301 (71%) | 123 (29%) | 0.002 | 322 (73.2%) | 18 (26.8%) | 0.171 |
| negative | 135 (58.7%) | 95 (4.3%) | - | 196 (78.1%) | 55 (21.9%) | - |
Expression of Fhit and Wwox assessed by immunohistochemical analyses on TMAs
There was a significant positive relation between Fhit and Wwox expression (p=0.024). In univariate analysis, when we compared Fhit expression with prognostic markers (Table 2), Fhit loss was more frequent in infiltrating ductal carcinoma, 65.7% of which were highly reduced or negative for Fhit expression; 54.3% of infiltrating lobular carcinomas and 33.3% of other types of invasive tumors were highly reduced or negative for Fhit expression (p=0.01). Fhit expression was also related with histological grade of the tumor, 78.3 % grade 3 tumors, 59.4% of grade 2 tumors and 36.7% of grade 1 tumors showed lost or highly reduced Fhit expression (p<0.001). Fhit expression and patient age were significantly related; lost or highly reduced Fhit expression was observed in 84.1% of tumors of patients ≤40 yrs old, 70.9% of cancers of patients 41-50 yrs old and 59% of tumors of patients >50 yrs old (p<0.001). Very low Fhit expression was more frequent in tumors of African American patients (84.6%), compared to 63% of tumors of caucasians (p=0.002).
Wwox expression scores were also compared with the same prognostic features (Table 2): Wwox was related with tumor type with 78.5% of infilrating ductal carcinomas, 64.4% of infilrating lobular carcinomas and 78.3% of other types of invasive tumors showing highly reduced or no Wwox expression (p=0.01); Wwox expression was also associated with histological grade with 84.6% of grade 3 tumors, 72.6% of grade 2 and 72% of grade 1 tumors showing highly reduced or absent Wwox expression (p<0.001); Wwox expression was not related with age of patients; highly reduced or absent Wwox expression was more common in tumors (88.1%) of African American women compared to 76% of tumors of caucasians (p=0.034). Clinicopathological features of the tumors and associations with Fhit and Wwox expression are given in Table 2.
Associations among breast cancer marker proteins and clinicopathological features
Fhit expression scores were assessed in association with expression of other biomarkers (see Figure 1 for examples of immunohistochemical stains): strong direct correlation of Fhit expression with ER and PR (p<0.001 for both) and inverse correlation with HER2, EGFR, CK5/6, AP2γ (p<0.001, p<0.001, p=0.018 and p=0.002, respectively) were observed (Table 2).
FIGURE 1.
Immunohistochemical staining of protein markers in a triple negative breast cancer. In the top two panels strong expression of Fhit and Wwox is shown in contrast to the negative and very low expression of Fhit and Wwox, repectively, in triple negative tumor 827. Tumor 827 also shows moderate expression of CK5,6, moderate nuclear expression of AP2α and γ and strong membraneous expression of EGFR (all magnifications X200).
Wwox expression level was also positively and highly related with ER and PR (p<0.001 and p=0.001, respectively). An inverse correlation was observed between Wwox and EGFR (p<0.001). A borderline inverse correlation was found between Wwox and CK5/6 (p=0.071). There was no association of Wwox with HER2, AP2α or AP2γ in univariate analyses (p=0.204, p=0.607 and p=0.171; respectively).
We characterized the basal-like phenotype by using 5 biomarkers that included triple negativity for ER, PR, HER2 and positive expression of CK5/6 and/or EGFR13,14. According to this classification, 89 tumors (10.7%) exhibited basal-like differentiation (Table 3). We then assessed the relation of the basal-like phenotype to expression scores for Fhit, Wwox, AP2γ and AP2α, and found a highly significant association of the basal-like phenotype with very low expression of Fhit and Wwox and high expression of AP2γ (p<000.1, p<0.001 and respectively (Table 3); see Figure 1 for examples.
TABLE 3.
The relation of genetically defined breast cancer subtypes with Fhit and Wwox expression
| Fhit | Wwox | |||||
|---|---|---|---|---|---|---|
| Lost or highly reduced | Strong or moderately reduced | P | Lost or highly reduced | Strong or moderately reduced | P | |
| Triple negative | ||||||
| positive | 95 (81.9%) | 21 (18.1%) | <0.001 | 121 (92.4%) | 10 (7.6%) | <0.001 |
| negative | 361 (61.7%) | 224 (38.3%) | - | 514 (74.2%) | 179 (25.8%) | - |
| Basal-like phenotype | ||||||
| positive | 71 (84.5%) | 13 (15.5%) | <0.001 | 86 (96.6%) | 3 (3.4%) | <0.001 |
| negative | 380 (62.1%) | 232 (37.9%) | - | 545 (74.6%) | 186 (25.4%) | - |
| Genetic classification | ||||||
| Luminal A | 229 (55.3%) | 185 (44.7%) | <0.001 | 368 (74.2%) | 128 (25.8%) | <0.001 |
| Luminal B | 69 (70.4%) | 29 (29.6%) | - | 80 (69.6%) | 35 ( 30.4%) | - |
| HER2 over-expressing | 52 (88.1%) | 7 (11.9%) | - | 56 (80.6%) | 14 (20.0%) | - |
| Basal-like | 71 (84.5%) | 13 (15.5%) | - | 86 (96.6%) | 3 (3.4%) | - |
| Triple negative-other | 15 (75%) | 5 (25%) | - | 18 (90%) | 2 (10%) | - |
When we examined the relation of expression of biomarkers with expression of Fhit and Wwox by regression analysis, we found that the three markers most closely associated with Fhit were ER, EGFR and HER2, while PR, EGFR and HER2 were most closely associated with Wwox expression (Tables 4 and 5).
TABLE 4.
Multivariate logistic regression analysis of biomarkers associated with Fhit
| P | Odds Ratio (O.R.) | 95% C.I. for O.R. | ||
|---|---|---|---|---|
| Biomarkers | Lower | Upper | ||
| EGFR | 0.013 | 3.3 | 1.18 | 3.95 |
| ER | 0.0001 | 14.6 | 1.63 | 4.20 |
| HER2 | 0.019 | 30.8 | 1.1 | 3.03 |
TABLE 5.
Multivariate logistic regression analysis of biomarkers associated with Wwox
| P | Odds Ratio (O.R.) | 95% C.I. for O.R. | ||
|---|---|---|---|---|
| Biomarkers | Lower | Upper | ||
| EGFR | 0.008 | 2.64 | 1.29 | 5.41 |
| PR | 0.004 | 2.43 | 1.33 | 4.44 |
| HER2 | 0.026 | 1.69 | 1.06 | 2.68 |
We divided the cases into five groups according to expression patterns of these 5 biomarkers with the aim to find cases matching with the recently described genetic classification11,12: luminal A (ER and/or PR positive, HER2 negative); luminal B (ER and/or PR positive, HER2 positive); HER2 over-expressing (ER and PR negative, HER2 positive); basal-like triple negative (ER, PR, HER2 negative and CK5/6 and/or EGFR positive); others (triple negative but not CK5/6 or EGFR positive) (see Table 3).
Expression patterns of the protein biomarkers within specific subtypes were assessed and highly significant associations were found for Fhit, Wwox and AP2γ (p<0.001 for all). AP2α did not show significant association (p=0.406) with specific subtypes. Fhit expression score differences between luminal A and B (p=0.006), luminal A and HER2 over-expression class (p<0.001), luminal A and basal-like (p<0.001), luminal B and HER2 over-expression class (p=0.01), and luminal B and basal-like (0.024), were the most important significant associations. HER2 over-expression and basal-like groups did not show differences in terms of Fhit expression (p=0.629). Lost or highly reduced Fhit expression was seen in 88.1% of HER2 over-expressing and 84.5% of basal-like tumors, whereas 55.5% of luminal A, 70.4% of luminal B and 75% of triple negatives with other than basal-like phenotype showed very low Fhit expression. When we assessed the association of different tumor subtypes in relation to Wwox expression scores, Wwox expression differences between luminal A and basal like (p<0.001), luminal B and basal-like (p<0.001), and HER2 over-expressing and basal-like (p=0.001) were the most important associations. absent or highly reduced Wwox expression occurred in 96.6% of basal-like, 90% of triple negatives other than basal-like, 80% of HER2 over-expressing, 74.2% of luminal A and 69.6% of luminal B tumors (see Table 3).
Coordinate loss of Fhit and Wwox expression
Highly reduced or lost expression of both Fhit and Wwox is observed in 342 (49%) cases. It is likely that the effect of reduced Fhit and Wwox is not very different from complete loss, since it is known that Fhit and Wwox are haploinsufficient tumor suppressors; ie, gene knockout animals with loss of one or both copies are equally highly susceptible to tumor induction22-24. The basal-like phenotype was observed in 20.7% of these cases, which is significantly more compared to cases with high or moderately reduced Fhit and/or Wwox expression (3.9%) (p<0.001). In other words, 83.3% of basal-like cancers had very low or lost expression of both Fhit and Wwox; 31.6% of them versus 7.9% of Fhit and/or Wwox positive cases (p<0.001) expressed EGFR. CK5/6 expression was observed in 25.9% of Fhit and Wwox negative and 15.6% of Fhit and/or Wwox positive cases (p=0.002). HER2 status was not significantly different between the two groups (p=0.174).
DISCUSSION
Our previous studies on the role of fragile genes in breast cancer showed that Fhit and Wwox expression is reduced or lost in >60% of invasive breast cancers and pure DCIS and up to 50% of expression loss is observed in adjacent normal cell areas of breast cancer sections4,5. In these studies, in which Fhit and Wwox expression was examined in invasive breast tumors4, tamoxifen sensitive and resistant breast cancers6, pure DCIS, DCIS adjacent to invasive tumors and normal tissues around invasive or in-situ breast cancer5, Fhit and Wwox expression correlated positively in all tissue compartments. including invasive tumors, DCIS and adjacent normal tissues. Fhit and Wwox loss was more common in ER and PR negative breast tumors1 and Nunez et al7 confirmed this relation of ER and Wwox in breast cancer. The concordant inactivation of Fhit and Wwox is an early event in breast carcinogenesis and especially common in ER and PR negative breast tumors.
Recently described breast cancer subtypes, identified through gene expression profiles or signatures, define 5 breast cancer groups with different prognostic features: luminal A, luminal B, HER2 over-expressing, basal-like and normal breast-like13,14. Basal-like breast cancers are triple negative tumors (ER, PR and HER2 negative) and currently the best method to identify them is by routine immunohistochemistry for positivity for CK5/6 and/or EGFR and negativity for ER, PR and HER220,21. It has been reported that this 5-marker panel has 100% specificity and 76% sensitivity20 in identifying basal-like cancers. The molecular events leading to development of basal-like tumors are especially important; because currently there is no good target for their treatment and these aggressive tumors are seen more commonly in young patients15,16 and patients of African origin. Basal-like cancers, so named because they express genes found in normal basal/myoepithelial cells of the breast such as cytokeratins CK5/6, account for up to 15% of all breast cancers but can account for >50% of breast cancers of some cohorts of patients studied in Africa17. Mutant BRCA1 associated breast cancers show basal-like features and BRCA1 inactivation is a common event in sporadic basal-like breast tumors18,19. In various studies of sporadic breast cancers, 40% or more were moderately to strongly positive for Fhit, whereas only 9% of BRCA1 mutant tumors were Fhit positive, suggesting that the BRCA1 pathway is important in protecting the FRA3B/FHIT locus from damage25; inactivation of the BRCA1 pathway in basal/triple negative cancers may contribute to loss of expression of Fhit in these cancers or possibly vice versa.
The aim of this study was to determine the relation of concordant Fhit and Wwox loss in carcinogenesis of these currently described breast cancer subtypes. This study showed strong correlations between triple negative phenotype or basal-like phenotype, (triple negative, CK5/6 and/or EGFR positive) and reduced Fhit and Wwox expression. The tumors with concordant loss of Fhit and Wwox were much more likely to be basal-like tumors than tumors that strongly expressed Fhit or Wwox. The results show that inactivation of fragile site genes is highly associated with and may have a role in formation of basal-like breast cancers. The molecular events involved in this process need to be investigated in in vivo experiments that may help to identify new targets for treatment. Thorough investigation of non-neoplastic breast lesions with reduced Fhit and Wwox expression may lead to recognition of early preinvasive lesions of basal-like breast cancers or the associated benign breast lesions.
An interesting area of investigation would be the relationship between basal-like breast cancers and activation of the DNA damage response (DDR) checkpoint. It has been demonstrated that DDR checkpoint proteins are activated in preneoplastic and neoplastic lesions of skin, lung and breast26-28 and proposed that activation of the checkpoint acts as a barrier to cancer progression until mutations in checkpoint genes allow growth of cells through the checkpoint in spite of DNA damage, leading to genome instability and cancer progression. Loss of heterozygosity at the FHIT locus, and probably the WWOX and other fragile loci, is concomitant with the DDR checkpoint activation1,2,26-28. We hypothesize that DNA breakage at the fragile gene loci may actually be the event that triggers the DDR; and paradoxically, loss of expression of the FHIT gene due to this breakage alters the DDR checkpoint, since Fhit is involved in normal responses to DNA damage29-31 and loss of Wwox affects DNA stability (Aqeilan and Huebner, unpublished data). Additionally, it was recently reported that Fhit negative sebaceous gland carcinomas can exhibit impaired homologus recombination repair due to deletions of BRCA1 or 2 genes32. Since basal-like breast cancers show loss or reduction of Fhit protein, these tumors are likely to have defective, activated DDR checkpoints and may be especially sensitive to inhibitors of checkpoint proteins, as are BRCA1 deficient cancers16,18,33-35. Indeed, it has been reported that Chk1 protein is highly expressed in triple negative breast cancers36.
HER2 expression was independently related to both Fhit and Wwox in regression analysis, though the relation with Wwox did not reach significance in univariate analysis. However, concordant loss of Fhit and Wwox was not correlated with HER2 expression. The relation between Fhit, Wwox and HER2 expression will need more detailed examination to determine if fragile gene inactivation also contributes to development of some HER2 over-expressing breast tumors, or if the apparent correlation is due to complex interactions among a number of proteins, including Wwox and Fhit.
The other marker that was strongly and independently inversely correlated with Fhit and Wwox expression is EGFR, a target for basal-like breast cancer treatment that is currently being investigated17. Thus, it will be important to learn more about the correlation between EGFR and Wwox/Fhit expression in vivo and the possible effects of the correlations on responses to experimental EGFR based therapies.
We also examined the expression of AP2α and AP2γ, proteins that physically interact with Wwox10,37. Nuclear AP2γ expression was also more frequent in basal-like tumors, perhaps partially due to Wwox reduction, which would release AP2γ from cytoplasmic binding to Wwox. AP2γ would then be free to act as a transcriptional regulator in the nucleus. However, we do not find a significant association between nuclear AP2α or γ and Wwox expression in specific breast cancer subypes; among the basal-like cancers there were only 3 that strongly expressed Wwox, too few to do statistical analysis. Because there are many known and potential Wwox interactor proteins, we will need to isolate specific interactors in specific breast normal and cancer subtypes in order to define the hierarchy of important Wwox effector proteins. Also, these in vivo analyses suggest that mechanisms of control of levels of expression of the AP2 transcription factors are more complex than illustrated by the in vitro studies.
Recently, a role for BRCA1 as stem cell regulator was described38 and it is possible that Fhit also has a role, albeit undefined, in stem cells, as recently suggested for hematopoietic stem cells39. Thus, it is possible that therapies based on inhibition of DNA damage response checkpoint, even an abnormal checkpoint such as observed for BRCA1 and/or Fhit deficient cells, may be a targeted strategy for treatment of basal-like/triple negative breast cancers and may even target the stem cells of such cancers.
Acknowledgments
Supported by: USPHS National Cancer Institute grant CA120516, Hacettepe University Research Fund grant 05D11101003, and The Stephanie Spielman Fund. We gratefully acknowledge the outstanding technical services of individuals in the Pathology Core Facility of the Ohio State University, especially Susie Jones.
REFERENCES
- 1.Durkin SG, Ragland RL, Arlt MF, et al. Replication stress induces tumor-like microdeletions in FHIT/FRA3B. Proc Natl Acad Sci U S A. 2008;105:246–51. doi: 10.1073/pnas.0708097105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Tsantoulis PK, Kotsinas A, Sfikakis PP, et al. Oncogene-induced replication stress preferentially targets common fragile sites in preneoplastic lesions. A genome-wide study. Oncogene. 2008;27:3256–64. doi: 10.1038/sj.onc.1210989. [DOI] [PubMed] [Google Scholar]
- 3.Thavathiru E, Ludes-Meyers JH, MacLeod MC, Aldaz CM. Expression of common chromosomal fragile site genes, WWOX/FRA16D and FHIT/FRA3B is downregulated by exposure to environmental carcinogens, UV, and BPDE but not by IR. Mol Carcinog. 2005;44:174–82. doi: 10.1002/mc.20122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Guler G, Uner A, Guler N, et al. The fragile genes, FHIT and WWOX, are coordinately inactivated in invasive breast cancer: Correlations with clinical features. Cancer. 2004;100:1605–1614. doi: 10.1002/cncr.20137. [DOI] [PubMed] [Google Scholar]
- 5.Guler G, Uner A, Guler N, et al. Concordant loss of fragile genes, Fhit and Wwox, in breast cancer development. Path Internatl. 2005;55:471–478. doi: 10.1111/j.1440-1827.2005.01855.x. [DOI] [PubMed] [Google Scholar]
- 6.Guler G, Iliopoulos D, Guler N, et al. Wwox and AP2γ expression levels predict tamoxifen response. Clin Cancer Res. 2007;13:6115–6121. doi: 10.1158/1078-0432.CCR-07-1282. [DOI] [PubMed] [Google Scholar]
- 7.Nunez MI, Ludes-Meyers J, Abba MC, et al. Frequent loss of WWOX expression in breast cancer: correlation with estrogen receptor status. Breast Cancer Res Treat. 2005;89:99–105. doi: 10.1007/s10549-004-1474-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Sbrana I, Veroni F, Nieri M, et al. Chromosomal fragile sites FRA3B and FRA16D show correlated expression and association with failure of apoptosis in lymphocytes from patients with thyroid cancer. Genes Chrom Cancer. 2006;45:429–36. doi: 10.1002/gcc.20305. [DOI] [PubMed] [Google Scholar]
- 9.Ishii H, Vecchione A, Furukawa Y, et al. Expression of FRA16D/WWOX and FRA3B/FHIT genes in hematopoietic malignancies. Mol Cancer Res. 2003;1:940–7. [PubMed] [Google Scholar]
- 10.Aqeilan RI, Palamarchuk A, Weigel RJ, et al. Physical and functional interactions between the Wwox tumor suppressor protein and the AP-2gamma transcription factor. Cancer Res. 2004;64:8256–61. doi: 10.1158/0008-5472.CAN-04-2055. [DOI] [PubMed] [Google Scholar]
- 11.Aqeilan RI, Donati V, Palamarchuk A, et al. WW domain-containing proteins, WWOX and YAP, compete for interaction with ErbB-4 and modulate its transcriptional function. Cancer Res. 2005;65:6764–72. doi: 10.1158/0008-5472.CAN-05-1150. [DOI] [PubMed] [Google Scholar]
- 12.Bosher JM, Totty NF, Hsuan JJ, et al. A family of AP-2 proteins regulates c-erbB-2 expression in mammary carcinoma. Oncogene. 1996;17:1701–7. [PubMed] [Google Scholar]
- 13.Perou CM, Sørlie T, Eisen MB, et al. Molecular portraits of human breast tumours. Nature. 2000;406:747–52. doi: 10.1038/35021093. [DOI] [PubMed] [Google Scholar]
- 14.Sørlie T, Perou CM, Tibshirani R, et al. Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci USA. 2001;98:10869–74. doi: 10.1073/pnas.191367098. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Reis-Filho JS, Tutt ANJ. Triple negative tumours: a critical review. Histopathol. 2008;52:108–118. doi: 10.1111/j.1365-2559.2007.02889.x. [DOI] [PubMed] [Google Scholar]
- 16.Yehiely F, Moyano JV, Evans JR, et al. Deconstructing the molecular portrait of basal-like breast cancer. Trends in Mol Med. 2006;12:537–544. doi: 10.1016/j.molmed.2006.09.004. [DOI] [PubMed] [Google Scholar]
- 17.Cleator S, Heller W, Coombs RC. Triple-negative breast cancer: therapeutic options. Lancet Oncology. 2007;8:235–244. doi: 10.1016/S1470-2045(07)70074-8. [DOI] [PubMed] [Google Scholar]
- 18.Turner NC, Reis-Filho JS, Russell AM, et al. BRCA1 dysfunction in sporadic basal-like breast cancer. Oncogene. 2006;26:2126–32. doi: 10.1038/sj.onc.1210014. [DOI] [PubMed] [Google Scholar]
- 19.Turner NC, Reis-Filho JS. Basal-like breast cancer and the BRCA1 phenotype. Oncogene. 2006;25:5846–53. doi: 10.1038/sj.onc.1209876. [DOI] [PubMed] [Google Scholar]
- 20.Nielsen TO, Hsu FD, Jensen K, et al. Immunohistochemical and clinical characterization of the basal-like subtype of invasive breast carcinoma. Clin Cancer Res. 2004;10:5367–5374. doi: 10.1158/1078-0432.CCR-04-0220. [DOI] [PubMed] [Google Scholar]
- 21.Cheang MC, Voduc D, Bajdik C, et al. Basal-like breast cancer defined by five biomarkers has superior prognostic value than triple-negative phenotype. Clin Cancer Res. 2008;14:1368–76. doi: 10.1158/1078-0432.CCR-07-1658. [DOI] [PubMed] [Google Scholar]
- 22.Fong LY, Fidanza V, Zanesi N, et al. Muir-Torre-like syndrome in Fhit-deficient mice. Proc Natl Acad Sci USA. 2000;97:4742–7. doi: 10.1073/pnas.080063497. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Zanesi N, Fidanza V, Fong LY, et al. The tumor spectrum in FHIT-deficient mice. Proc Natl Acad Sci USA. 2001;98:10250–5. doi: 10.1073/pnas.191345898. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Aqeilan RI, Trapasso F, Hussain S, et al. Targeted deletion of Wwox reveals a tumor suppressor function. Proc Natl Acad Sci USA. 2007;104:3949–54. doi: 10.1073/pnas.0609783104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Turner BC, Ottey M, Zimonjic DB, et al. The fragile histidine triad/common chromosome fragile site 3B locus and repair-deficient cancers. Cancer Res. 2002;62:4054–60. [PubMed] [Google Scholar]
- 26.Gorgoulis VG, Vassiliou LV, Karakaidos P, et al. Activation of the DNA damage checkpoint and genomic instability in human precancerous lesions. Nature. 2005;434:907–13. doi: 10.1038/nature03485. [DOI] [PubMed] [Google Scholar]
- 27.Bartkova J, Horejsi Z, Koed K, et al. DNA damage response as a candidate anti-cancer barrier in early human tumorigenesis. Nature. 2005;434:864–70. doi: 10.1038/nature03482. [DOI] [PubMed] [Google Scholar]
- 28.Bartek J, Bartkova J, Lukas J. DNA damage signalling guards against activated oncogenes and tumour progression. Oncogene. 2007;26:7773–7779. doi: 10.1038/sj.onc.1210881. [DOI] [PubMed] [Google Scholar]
- 29.Ishii H, Mimori K, Inoue H, et al. Fhit modulates the DNA damage checkpoint response. Cancer Res. 2006;66:11287–92. doi: 10.1158/0008-5472.CAN-06-2503. [DOI] [PubMed] [Google Scholar]
- 30.Ishii H, Wang Y, Huebner K. A Fhit-ing role in the DNA damage checkpoint response. Cell Cycle. 2007;6:1044–8. doi: 10.4161/cc.6.9.4213. [DOI] [PubMed] [Google Scholar]
- 31.Pichiorri F, Ishii H, Okumura H, et al. Molecular parameters of genome instability: Roles of fragile genes at common fragile sites. J Cell Biochem. 2008;104:1525–33. doi: 10.1002/jcb.21560. [DOI] [PubMed] [Google Scholar]
- 32.Becker K, Goldberg M, Helmbold P, et al. Deletions of BRCA1/2 and p53 R248W gain-of-function mutation suggest impaired homologous recombination repair in fragile histidine triad-negative sebaceous gland crcinomas. Brit J Dermat. 2008 Aug 19; doi: 10.1111/j.1365-2133.2008.08783.x. [Epub ahead of print] PMID: 18717684. [DOI] [PubMed] [Google Scholar]
- 33.Bryant HE, Schultz N, Thomas HD, et al. Specific killing of BRCA2-deficient tumours with inhibitors of poly(ADP-ribose) polymerase. Nature. 2005;434:913–917. doi: 10.1038/nature03443. 2005. [DOI] [PubMed] [Google Scholar]
- 34.Farmer H, McCabe R, Lord CJ, et al. Targeting the DNA repair defect in BRCA mutant cells as a therapeutic strategy. Nature. 2005;434:917–921. doi: 10.1038/nature03445. [DOI] [PubMed] [Google Scholar]
- 35.O’Connor MJ, Martin NMB, Smith GCM. Targeted cancer therapies based on the inhibition of DNA strand break repair. Oncogene. 2007;26:7816–7824. doi: 10.1038/sj.onc.1210879. [DOI] [PubMed] [Google Scholar]
- 36.Verlinden L, Vandem Bempt I, Eelen G, et al. The E2F-regulated gene Chk1 is highly expressed in triple-negative estrogen receptor-/progesterone receptor-/her1- breast carcinomas. Cancer Res. 2007;67:6574–6581. doi: 10.1158/0008-5472.CAN-06-3545. [DOI] [PubMed] [Google Scholar]
- 37.Iliopoulos D, Guler G, Han SY, et al. Roles of FHIT and WWOX fragile genes in cancer. Cancer Lett. 2006;232:27–36. doi: 10.1016/j.canlet.2005.06.048. [DOI] [PubMed] [Google Scholar]
- 38.Liu S, Ginestier C, Charafe-Jauffret E, et al. BRCA1 reglates human mammary stem/progenitor cell fate. Proc Natl Acad Sci USA. 2008;105:1680–1685. doi: 10.1073/pnas.0711613105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Ishii H, Mimori K, Ishikawa K, et al. FHIT-deficient hematopoietic stem cells survive hydroquinone treatment long-term, carrying precancerous alterations. Cancer Res. 2008;68:3662–3670. doi: 10.1158/0008-5472.CAN-07-5687. [DOI] [PubMed] [Google Scholar]