We found concordant BRCA1 protein staining in frozen and FFPE tissue specimens of 22 randomly selected breast tumors. Using anti-BRCA1 antibodies AP16 and K-18 on frozen sections we found punctate BRCA1 staining and more homogeneous staining with the MS110 antibody in variable numbers of tumor cell nuclei and nucleoli of breast cancer patients
[26, 27]. There was variable homogeneous nuclear staining with the MS110 antibody in contiguous pressure cooker antigen-retrieved FFPE tumor tissue. These results extend our previous studies which showed BRCA1 nuclear and nucleolar localization in frozen tissue sections and estradiol-treated MCF7 cells
[26–28]. However, Bogdani et al.
 using D-20, K-18, I-20, and C-20 anti-BRCA1 antibodies on Bouin’s fixed tumor tissue of young and old patients with breast cancer, found nuclear and cytoplasmic BRCA1 staining in about 50% of sporadic specimens, and less staining in tumor cells in young patients and one with a germline mutation. Perez-Valles et al.
 showed predominantly cytoplasmic staining with the D-20, I-20, and K-18 polyclonal antibodies in FFPE samples of both tumoral and non-tumoral cells, but nuclear and cytoplasmic staining with the I-20 antibody in FFPE samples after microwave pre-treatment in breast tumor tissue. Recently, Milner et al.
 in an extensive evaluation of MS110 antibody staining as a patient selection biomarker found sub-cellular localization to the nucleus. However, they concluded that although MS110 did detect BRCA1 in FFPE tumor tissue samples, BRCA1 expression levels in standard methodologies were not reproducible enough to enable its use as a selection marker.
We found less BRCA1 nuclear staining in the tumor tissue of frozen tissue specimens and in contiguous FFPE tissue with higher histological grade. The inverse correlation between lower BRCA1 protein expression and higher histological grade has been well established
; less nuclear staining has also been described in estrogen receptor/progesterone receptor/HER2-negative (triple negative) breast tumor tissue
In a patient with a BRCA1 mutation (185delAG), we found FFPE tumor tissue stained with the MS110 antibody was mostly negative and the BRCA1 staining in the FFPE normal tissue was weakly positive. These results are consistent with reports showing that heterozygous germline BRCA1 mutations involved in familial breast cancer produce truncated protein, preferential LOH of the wild-type allele in tumor tissue, and loss of MS110 reactivity
[35–37]. However, Perez-Valles et al.
 did not find differences in nuclear BRCA1 protein expression between cases with and without BRCA1 germline mutations by immunohistochemistry.
In our study, we also found diffuse, and irregular BRCA1 protein staining with the K-18 antibody in the frozen section tumor tissue from a patient, with a BRCA1 mutation (185delAG), which is difficult to explain. Although there is no direct evidence that the BRCA1 185delAG mutation results in transcription of a functional truncated protein, Buisson et al.
 have reported that mRNAs from patients with the BRCA1 185delAG mutations are able to elude the non-sense mediated mRNA decay (NMD) pathway and remain stable in the cell. O’Donnell et al.
 proposed that the truncated protein product may modulate chemo-sensitivity. Wilson et al.
 suggest that the 185delAG mutation results in deletion of most of the protein, including the MS110 epitope which maps between aa residues 89–222. However, the 185delAG BRCA1 N-terminus 39 aa may have an epitope within aa 70–89 that is detected by the K-18 antibody.
We have found positive BRCA1 protein nuclear staining in frozen and FFPE tumor and lactating tissue from a patient with a BRCA2 mutation. However, as our samples were tested for all three targeted mutations, there is no evidence that BRCA1 is altered in this patient. Therefore these results are consistent with evidence that although BRCA1 protein is a negative regulator of cell growth, BRCA1 expression is up-regulated, during cell proliferation and lactation
In our FFPE sections we frequently found that normal epithelial tissue surrounding the tumor had strong nuclear staining with the MS110 antibody, although weak staining was observed in the normal tissue in two of our specimens. This generally agrees with data that normal tissue surrounding the tumor has positive nuclear staining with the MS110 antibody
. However, it has also been observed by Bogdani et al.
 using the I-20 polyclonal antibody that positive nuclear staining was found in 75% of normal tissue from both young and old patients, and that nuclear staining was absent in the normal tissue of six samples. Milner et al.
 also found that associated non-neoplastic normal breast epithelium showed moderate to strong nuclear staining with no evidence of cytoplasmic staining in normal breast epithelium.
In summary, our immunohistological studies of frozen and FFPE tissues of breast tumors with monoclonal antibodies MS110 and AP16 and the polyclonal antibody K-18, with immunoperoxidase and immunofluorescence detection, have shown that BRCA1 protein has a nuclear and nucleolar localization. We have used freezing in liquid N2, followed by -20o C methanol fixation of the captured frozen sections in order to minimize translocation of antigen during fixation. Guerra–Rebollo et al.
 also found that after using methanol fixation, BRCA1 was localized in nucleoli, but after γ-irradiation BRCA1 was depleted from the nucleolus, and associated with ionizing-radiation induced foci (IRIF).
We have previously shown co-localization of BRCA1 protein and nucleolin in the nucleolus and speckles of MCF7 and HeLa cell nuclei
[26, 27]. In the present study, we found that in the mutant HCC1937 cell line (5382insC), co-localization of BRCA1 protein and nucleolin occurred in aberrant, patched regions in some nucleoli. HCC1937 extracts revealed low levels of a 210 kDa BRCA1 species consistent with the prediction of a truncated protein product lacking the BRCA1 C terminus
. This result suggests that protein truncation at the C-terminus may yield variable nucleolar localization. The functions of BRCA1 protein are yet to be fully elucidated, however, BRCA1 protein participates in many signaling pathways involved in transcription, checkpoint control, and is recruited for the formation of DNA repair complexes in association with proteins such as Mre11-Nbs1-Rad50, and BRCA2
. The data that BRCA1 is localized in the nucleolus, in addition to speckles, may explain a possible functional participation in processes of ribosomal biosynthesis, cell cycle progression, and as a reservoir for complexes formed in response to cellular stress and DNA repair
The function of BRCA1 protein in tumorigenesis has been found to be complex, and as a tumor suppressor, it is postulated that reduced expression leads to multiple abnormalities, including a defect in the homologous recombination (HR) pathway of DNA repair. These defects are associated with a hypersensitivity to many agents that cause DNA double strand breaks, such as ionizing radiation (IR). In a study testing breast cancer biopsies irradiated ex vivo for the ability to form BRCA1, FANCD2 and RAD 51 foci, Willers et al.
 detected BRCA1 DNA repair foci defects in triple-negative breast cancers, a phenotype associated with BRCA1 deficiency. In addition, Burga et al.
 found an association of increased proliferation and increased BRCA1 immunohistochemical expression in breast cancer epithelial cells from BRCA1 mutation carriers, which they ascribed to epidermal growth receptor pathway activation.
The data presented here support a role for BRCA1 in the pathogenesis of sporadic and inherited breast cancers. The use of well-characterized reagents and possibilities for co-localization experiments in both cell culture, and frozen and FFPE tissues, may lead to further insights into the molecular pathways involved in BRCA1 protein function and possibly the further development of targeted therapeutics.