Neuroepithelial stem cell protein (Nestin) is an intermediate filament (IF) protein considered to be a marker of neural stem cells. Nestin expression has been confirmed in stem/progenitor cells of the dermis, hair follicles, intestine, pancreas, bone marrow as well as in neural, muscle and other tissues (reviewed in ). Its expression is down regulated in the course of differentiation and subsequently replaced by another type of IF [3, 10]. Nestin expression has also been detected in the endothelium [11, 12], and we have noted the activation of nestin expression in newly formed human blood vessels . The human nestin molecule has a homology with rodent nestin confirming its highly conserved nature (Fig. 2c) .
The myoepithelial cells of the breast form an outer layer of the terminal duct lobular unit. They play an active part in mammary gland branching morphogenesis , and their correct recognition and detection is crucial in the diagnosis of a number of pathological breast lesions [4, 8]. Several markers have been reported for the immunohistochemical detection of breast myoepithelial cells such as α-smooth muscle actin, smooth muscle myosin heavy chain, calponin, h-caldesmon, S100 protein, p63, CD10, maspin and specific cytokeratins 5, 7, 15 and 17 (reviewed in [2, 8]). However, the specificity and sensitivity of these markers vary widely. Of these, maspin  and p63 [1, 2, 13] are generally considered the most promising. In histopathological practice it is very often necessary to confirm a diagnosis using a wide battery of immunohistological stainings. Undoubtedly, the identification of new markers is desirable.
We analysed the expression of nestin by an indirect immunohistochemical method (primary monoclonal anti-nestin antibody, clone 10C2, cat. no. MAB5326, Chemicon International, diluted 1:200; secondary staining system EnVision, Dako; antigen retrieval by microwave generator in citrate buffer pH 7.0; diaminobenzidine as chromogen) in 300 archival formalin-fixed and paraffin-embedded tissue samples of breast cancer and surrounding mammary tissue as well as in tissue microarrays consisting of 109 cases of breast cancer, developed for another project.
Nestin expression of various intensities was seen in the cytoplasm of myoepithelial cells of almost all normal ducts and lobular acini found in the vicinity of tumours (Fig. 1a and b). The intensity of expression varied from strong/marked (32 cases, ∼10%), to moderate/clearly visible (91 cases, ∼29%) to low/visible (113 cases, ∼38%) and to very low/visible in less than 50% cells only (64 cases, ∼21%). In our opinion, the reasons for the lower positive intensity in some cases can be explained by the fixation conditions. The specimens obtained by excision were generally more positive than specimens from the centre of the mastectomy. Pathological lesions like adenosis with atypical hyperplasia and in situ carcinoma exhibited very strong positivity (Fig. 1c and d), and the detection of nestin could also be used for the diagnosis of pseudo-invasion in some controversial cases (Fig. 1e). In invasive carcinomas, we found nestin positivity in a portion of epithelial cells in a minority (∼1/3) of cases. The positivity in the epithelial component was confirmed by simultaneous AE1/AE3 cytokeratin immunostaining and was significantly higher in hormone receptor and c-erbB-2/HER2/neu negative cases, which likely corresponded to basal cell phenotype cancer (p ≤ 0.003). Apart from the epithelial component, nestin positivity in carcinomatous samples was also confirmed in some stromal elements and endothelium of the capillary network (confirmed by simultaneous CD34 immunostaining; Fig. 1f). In another ongoing study, we are examining whether these positivities are related to the presence of tumour stem/progenitor cells (in the case of cells with epithelial/myoepithelial phenotype) and to neoangiogenesis (in the case of cells with endothelial/mesenchymal phenotype) and whether this has any prognostic or predictive value.
Embryological studies consider the mammary gland to be derived from skin adnexa related to sweat glands. In this light, we compared nestin expression in mammary and sweat glands. Interestingly, we found nestin positivity in sweat glands, not in myoepithelium but in luminal cells (Fig. 2a,b). We observed no nestin expression in the epidermis or squamous epithelial mucosa. This result confirms the complexity and tissue-specific uniqueness of the differentiation and developmental pathways.
In summary, we have described the potential utility of nestin immunostaining in breast pathology for the detection of myoepithelial/progenitor cells and suggest its potential involvement in the processes of tumour development and neoangiogenesis.
This work was supported in part by grants MSM 6198959216, MSM 0021620820 and IGA MH NR 7844.