Changes in the microenvironment have been implicated in the transition of pre-invasive ductal carcinoma in-situ (DCIS) to invasive breast cancer. Normal myoepithelial cells have a tumour suppressor phenotype but they are altered in DCIS and ultimately lost with transition to invasive cancer. A consistent change in DCIS is up-regulation of the integrin αvβ6 in myoepithelial cells. Preliminary observations identified a correlation between myopeithelial αvβ6 and an increased peri-ductal inflammatory infiltrate. The hypothesis of this study is that the altered myoepithelial phenotype influences the peri-ductal inflammatory environment, which in turn mediates a pro-apoptotic effect on myoepithelial cells contributing to their loss. To investigate this, the inflammatory infiltrate was characterised in a series of DCIS tissue in relation to αvβ6 status. This demonstrated significantly higher levels of CD4+ve and FOXP3+ve T cells around αvβ6+ve DCIS ducts compared to αvβ6-ve ducts (P=<0.01), suggesting an increase in Treg cells. In-vivo, Matrigel plugs containing injected into the flanks of female C57/Blk6 normal mice generated influx of higher levels of CD4+ve cells (p=0.005) and FOXP3+ T cells (p=0.007) in the presence of αvβ6+ve myoepithelial cells compared to αvβ6-ve cells, supporting the findings in human tissue samples. Since Treg cells produce TRAIL that can induce apoptosis, we investigated the influence of αvβ6 on myoepithelial cells on the levels of TRAIL in T cells and the hypothesis that αvβ6-positive myoepithelial ells may be more susceptible to TRAIL-induced apoptosis, leading to loss of the myoepithelial barrier. Firstly, levels of TRAIL in Jurkat and primary T cell populations co-cultured with β4 (ii) or β6 myoepithelial cells were measured. This demonstrated a higher level of TRAIL in primary T cells co-cultured β6 myoepithelial cells compared to those co-cultured with β4 myoepithelial cells. β6+ve and β6-ve myoepithelial cells were exposed to TRAIL, and this demonstrated that TRAIL enhanced apoptosis, measured by cleaved PARP, in β6+ve cells. Furthermore, these cells showed loss of the anti-apoptotic protein Galectin-7, and knockdown of Galectin-7 in normal β6-ve myoepithelial cells rendered them more susceptible to TRAIL-induced apoptosis. In DCIS tissues, an inverse relationship between αvβ6 and Galectin-7 in myoepithelial cells was demonstrated, and Cytokine Array analysis showed that αvβ6+ve myoepithelial cells express higher levels of IL-16, which has a role in Treg cell recruitment. Taken together these results suggest that expression of αvβ6 by myoepithelial cells in DCIS generates a tumour-promoter peri-ductal inflammatory infiltrate through altered cytokine release, is associated with reduced galectin-7 expression and enhances myoepithelial cell apoptosis in response to TRAIL. This provides a potential mechanism by which myoepithelial cells may be lost during evolution of DCIS and so contribute to progression to invasive disease.

Mammary cancer is a serious health issue in pet rabbits; prognostic factors are unknown. In a normal mammary gland, glandular secretory cells are surrounded by a single continuous layer of myoepithelial cells. In non-invasive mammary carcinomas, tumor cells are delineated by an intact myoepithelial layer, which is gradually lost to invasive carcinomas. The main aim of this study was to determine in rabbit mammary carcinomas (n = 119) the expression of myoepithelial markers that have prognostic significance in human cancer. Results show that all cases contained some retained myoepithelial cells. In 93% of the tumors, neoplastic cells expressed the myoepithelial marker calponin. There was a statistically significant association between higher percentages of calponin-containing cancer cells and histological features indicative of a better tumor differentiation, i.e., a lower proliferation of tumor cells, an increased percentage of tubular growth within the tumor, and a lower tumor grade, respectively. These results suggest that rabbit mammary carcinomas develop from progression of non-invasive cancer forms, and that calponin expression in cancer cells likely represents a favorable prognostic factor. The latter hypothesis has to be confirmed in long-term follow-up studies.

Orientador: Jose Vassalo / Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Ciencias Medicas / Made available in DSpace on 2018-08-07T05:09:42Z (GMT). No. of bitstreams: 1 Schenka_NataliaGuimaraesdeMoraes_D.pdf: 3809756 bytes, checksum: d4d7db8aeacfcd3c285cdee5b7c0a182 (MD5) Previous issue date: 2006 / Resumo: O sistema ductal da mama é revestido por duas camadas de células: uma constituída de células epiteliais luminais e outra de células mioepiteliais (CME). A identificação das CMEs é de grande importância na rotina em patologia cirúrgica, já que estas estão presentes em lesões benignas e geralmente ausentes nas malignas. Contudo, esta identificação morfológica (i.e., à hematoxilina-eosina) pode tornar-se difícil, sendo necessários marcadores imunoistoquímicos mioepiteliais. Até o momento, nenhum marcador mostrou-se, isoladamente, sensível e específico o bastante para auxiliar de maneira fidedigna em dilemas diagnósticos específicos, especialmente em: (1) lesões esclerosantes benignas vs. carcinoma tubular e (2) neoplasias papilíferas. No presente trabalho, dois novos marcadores mioepiteliais (p63 e CD10) foram testados e comparados com marcadores tradicionalmente empregados em patologia cirúrgica (1A4 e calponina). No primeiro dilema diagnóstico, foram estudados 10 casos de adenose esclerosante, 10 casos de cicatriz radiada e 10 casos de carcinoma tubular. Todos os marcadores demonstraram expressão nas CMEs de todas as lesões benignas e foram consistentemente negativos nos casos de carcinoma tubular, porém, p63 e CD10 mostraram maior especificidade. Os marcadores tradicionais mostraram positividade em células estromais em todos os casos de carcinoma tubular, sendo que esta reação-cruzada poderia causar problemas de interpretação ao simular uma camada de CMEs em quatro casos. Contudo, 1A4 mostrou uma positividade mais intensa e reprodutível nas CME. Concluímos, assim, que p63 e CD10 devem ser usados como anticorpos complementares ao 1A4 na distinção entre lesões esclerosantes benignas e carcinoma tubular da mama. No estudo das neoplasias papilíferas, foram avaliados 20 casos incluindo papilomas, papilomas atípicos e carcinomas papilíferos, além do tecido mamário normal adjacente. Todos os marcadores foram difusamente positivos no tecido mamário normal e papilomas, indicando sensibilidades semelhantes na identificação de CME. Uma positividade intensa foi encontrada em 100% dos casos corados para 1A4 e CD10, mas em apenas 76% e 60,5% dos corados para calponina e p63, respectivamente, sendo as diferenças estatisticamente significantes (p<0.05). Este dado sugere que os dois primeiros são tecnicamente mais reprodutíveis. Os marcadores mais específicos foram o p63 e CD10, mostrando reação-cruzada com células estromais em 0 e até 33% dos casos, respectivamente. Já os marcadores 1A4 e calponina mostraram reação-cruzada difusa em todos os casos. O CD10 mostrou uma combinação de uma maior especificidade e reprodutibilidade, com uma boa sensibilidade. Apesar de ser o mais específico, o p63 apresentou a menor sensibilidade e a impressão de uma camada de CMEs variavelmente interrompida (mesmo em tecido normal e lesões benignas), o que poderia causar problemas de interpretação. Além disto, o 1A4 mostrou-se também neste contexto, como o mais reprodutível tecnicamente. Assim sendo, no dilema diagnóstico das neoplasias papilíferas, defendemos o uso combinado do marcador CD10 com o 1A4. Em resumo, apesar de variarem em acurácia diagnóstica quando comparados entre si e com marcadores tradicionais, na dependência do dilema considerado, os novos marcadores testados neste trabalho parecem promissores no diagnóstico diferencial de lesões benignas e malignas da mama / Abstract: The myoepithelial cell (MEC) layer is usually present and continuous in normal breast tissue/benign lesions, and discontinuous to absent in atypical/malignant counterparts. Identifying MECs can be difficult on morphological grounds alone and currently relies on immunomarkers. So far, this detection has been carried out using antibodies such as alpha-smooth muscle actin (1A4) and calponin, but no immunomarker has proved to be accurate enough to identify MECs, particularly in specific diagnostic dilemmas such as: (1) benign sclerosing lesions vs. tubular carcinoma and (2) papillary neoplasms. The specificity of these markers has been questioned because they may be expressed in stromal myofibroblasts and vascular smooth muscle. Two novel myoepithelial markers have been described: the nuclear protein p63 and the surface antigen CD10. In the present study, we assessed the use of p63 and CD10 in comparison to the traditional markers in the specific diagnostic dilemmas specified above. In the first diagnostic problem, we studied 30 cases including sclerosing adenosis, radial scars and tubular carcinomas. All markers were expressed in MECs of all benign lesions and negative in all cases of tubular carcinoma. p63 and CD10 were mostly confined to MECs and thus more specific. Stromal positivity for 1A4 was present in all cases of tubular carcinoma and was misleading in 4 cases, as it simulated a MEC layer around malignant tubules. However, 1A4 was consistently more intensely expressed and thus more technically reproducible than the novel markers. So, we concluded that p63 and CD10 should be used as a complement to 1A4 in distinguishing benign sclerosing lesions from tubular carcinoma of the breast. Concerning the other diagnostic dilemma, we studied 20 cases of papillary neoplasms (including benign papillomas, atypical papillomas and papillary carcinomas), and adjacent normal breast tissue. All markers were diffusely positive in all samples of normal tissue and benign papillomas indicating similar sensitivities in the identification of MECs. Intense positivity was found in 100% of the cases stained with 1A4 and CD10, but in only 76% and 60,5% of those stained with calponin and p63, respectively; the differences were statistically significant (p<0.05), suggesting that the former two rendered more reproducible results. The most specific markers were p63 and CD10 which showed cross-reactivity in 0% and in up to 33% of the cases, respectively. 1A4 and calponin showed diffuse cross-reactivity in all cases. CD10 seems to combine the highest specificity and reproducibility with a good sensitivity. In spite of being the most specific, p63 was the least sensitive, also giving the impression of a discontinuous MEC layer even in normal tissue/benign lesions, which could potentially cause diagnostic problems. Moreover, in this context, 1A4 proved to be the most reproducible. Thus, a minimum panel for immunodetection of MEC to assess papillary lesions should include both markers. In conclusion, in spite of the variable accuracy depending on the diagnostic dilemma considered, the novel markers seem to be promising tools in the differential diagnosis between benign and malignant lesions of the breast / Doutorado / Anatomia Patologica / Doutor em Ciências Médicas

No description available.