Análise in situ de proteínas diferencialmente expressas em carcinomas epidermoides penianos por espectrometria de massas utilizando Matrix-Assisted Laser Desorption/Ionization (MALDI) / In situ analysis of protein differentially expressed in penile squamous cell carcinoma by Matrix-Assisted Laser Desorption/Ionization (MALDI) mass spectrometry

O carcinoma de pênis (CaPe) corresponde a uma doença maligna mutilante do homem. É mais frequente em regiões economicamente desprivilegiadas, como o Norte/Nordeste do Brasil, onde frequentemente é diagnosticado como doença mais avançada. Assim, novos marcadores diagnósticos, prognósticos e preditivos de tratamentos terapêuticos ainda são necessários. Abordagens proteômicas, incluindo o MALDI Imaging, podem contribuir neste sentido. Esta técnica emergente de espectrometria de massas permite a visualização da distribuição espacial de centenas de dados moleculares diretamente da superfície de uma secção tecidual, adquiridos por razão massa/carga (m/z). Neste contexto, nosso principal propósito foi integrar dados de proteômica clássica (gel 2D e Cromatografia Líquida acoplada à Espectrometria de Massas) e de MALDI Imaging, para obter padrões diferenciais de proteínas associados com amostras de Carcinoma Epidermoide Peniano usual (relacionado ou não ao HPV) e espécimes normais, a fim de buscar possíveis biomarcadores da doença. Um total de 45 amostras de CaPe, congeladas, foram inicialmente genotipadas para a presença do HPV. Destas, 60% foram positivas para variantes virais de alto risco. A proteômica clássica (N=24) evidenciou níveis diferenciais de 35 proteínas entre amostras de CaPe e controles, e 29 entre CaPe HPV positivo versus negativo (P<0,05; ANOVA). Redes de interações demonstraram que estes perfis proteicos interagiam com clusters de proteínas relacionadas com a carcinogênese e progressão tumoral. Entre eles, se destacaram aqueles formados por proteínas antioxidantes e de adesão celular, presentes em níveis elevados em tumores HPV negativos. A partir dos interactomas, quatro alvos proteicos foram selecionados para a análise in situ por imageamento: Calreticulina, 14-3-3 sigma, Serpina B5 e Glutationa-s-transferase. A aquisição de dados do MALDI Imaging foi conduzida após a digestão in situ pela tripsina, usando uma resolução de 200 µm e faixa de 700-3500 m/z para peptídeos (N=31). Os dados de identificação do gel 2D foram então integrados aos do imageamento. A identidade proteica dos filtros foi confirmada, in silico, por meio da presença de peptídeos teóricos co-localizados com o peptídeo experimental alvo nas secções de CaPe. Não houve associação significativa entre os parâmetros clinicopatológicos e as intensidades de sinal dos alvos (P>0,05, U de Mann-Whitney). Análises não supervisionadas, realizadas a partir dos dados do MALDI Imaging, evidenciaram mapas de segmentação que coindiciram com as regiões tumorais e margens adjacentes livres de neoplasia. Entre os principais valores de m/z diferenciadores estava o pico 1413 ± 2,5 Da, abundante nas regiões tumorais, e correlacionado ao peptídeo experimental m/z 1410,86 referente à proteína Calreticulina (CRT), o. Análises estatísticas (PCA e Curva ROC) indicaram este valor de m/z como potencial biomarcador da doença. Por conseguinte, a CRT foi selecionada para a etapa de validação por imunoistoquímica em tecidos parafinados de CaPe (N=158). Níveis elevados de imunoreatividade da CRT foram associados com piores tempo de sobrevida global (Razão de Risco 2,3; IC-1,46-3,96; P<0,001) e câncer específica (Razão de Risco 4,37; IC-1,66-11,51; P=0,002) nos casos de CaPe. A presença de metástase em linfonodos foi considerado um fator prognóstico independente para o risco de morte pelo câncer (Razão de Risco ­ 14,18; CI-3,29-61,12; P <0,001). A imunoreatividade da CRT também foi capaz de predizer a presença de metástase em linfonodos (Chance de Risco: 1,006; IC- 1,0001-1,0012; p=0,044). Estes dados, em conjunto, sugerem que a CRT pode ser um potencial biomarcador prognóstico do CaPe. A estratégia de integração da proteômica clássica com o MALDI Imaging, mostrou-se uma ferramenta útil na busca de novos biomarcadores para o CaPe. Além disto, o trabalho adicionou uma visão analítica à histopatologia clássica, o que deverá inserir as técnicas utilizadas neste projeto em estudos de Anatomia Patológica, tanto em nossa instituição, quanto no contexto global.

Penile cancer (PeCa) corresponds to a mutilating malignant disease in men. It is more frequent in underprivileged socioeconomic regions (e.g., Noth, North-East of Brazil), where it is frequently diagnosed in advanced stages. Thus, new markers are still needed for early diagnosis, prognosis and prediction of therapy. Proteomic approaches, including MALDI Imaging, could assist in this effort. This emerging spatially resolved mass spectrometric technique can obtain topographical distribution of hundreds of molecules directly from the tissue section surface, mensured by mass/charge ratio (m/z). In this context, our mainly propose was to integrate classic proteomic data (2D gel and Liquid Chromatograph coupled with Mass Spectrometry) with MALDI Imaging to obtain diferential patterns of protein associated with Usual Squamous Cell Penile Carcinoma (HPV related or not) and normal specimens, to look for possible biomarkers of the disease. A total of 45 fresh-frozen PeCa samples were initially searched for HPV genotype, 60% of which were positive for high-risk HPV. Classic proteomics (N=24) demonstrated diferential levels of 35 proteins comparing PeCa and control samples, and 29 comparing HPV-positive versus HPV-negative PeCa samples (P<0.05; ANOVA). Protein networks showed that these protein profiles interact with clusters of proteins related with tumorigenesis and tumor progression processes. Among them, antioxidant and cell adhesion proteins play a critical role in HPV negative penile tumors. Based on interactome data, four protein targets were selected for in situ analyses by imaging: Calreticulin, 14-3-3 protein sigma, Serpin B5 and Glutatione-s-transferase. MALDI Imaging data acquisition of peptides was conducted after in situ trypsin digestion using a lateral resolution of 200 µm, covering the range 700- 3500 m/z (N=31). After that, 2D gel based proteomic data was integrated with Imaging data. The filter protein identities were confirmed in silico by the co-localization of theoretical triptic peptides with the experimental peptides in PeCa sections. There was no significant association between the clinical and pathological parameters and the target signal intensities (P>0.05; U de Mann-Whitney). An unsupervised clustering analysis based on MALDI Imaging data reveled segmentation maps that coincide with histological annotation for tumor and adjacent non-neoplasic regions. Among the mainly differentiating m/z values there was 1413 ± 2.5 Da. This peak was especially co-localized with tumoral regions and correlated with Calreticulin (CRT) experimental peptide (m/z 1410,86). Statistical analysis (PCA and ROC Curves) indicated this m/z value as a potencial biomarker of the disease. For this reason, CRT was selected for validation by immunohistochemistry performed on paraffin-embedded PeCa tissues (N=158). As result, CRT hiperexpression in PeCa tissue increased the risk of unfavorable overall survival (Relative Risk ­ 2.3; CI-1.46-3.96; P<0.001) and cancer specific survival (Relative Risk ­ 4.37; CI-1.66-11.51; P=0.002) in these patients. Lymph node metastasis represented an independent prognostic risk factor for death related to cancer in our patients (Relative risk ­ 14.18; CI-3.29-61.12; P <0.001). CRT immunoreactivity was also capable to predict the presence of lymph node metastases (Risk Chance ­ 1,006; CI-1.0001-1.00123; P =0.044). Taken together, our results sugest that CRT may represent a prognostic biomarker of PeCa. The strategy of integrated classic proteomic and MALDI Imaging revealed as usefull tool to search for news biomarkers of the disease. Futhermore, this work added an analytical perspective to the classical histopathology, allowing to include the techniques used in this project in future morphological studies, both in our institution and in the global context.

Paracoccidoides brasiliensis 30 kDa adhesin: identification as a 14-3-3 protein, cloning and subcellular localization in infection models.

Paracoccidoides brasiliensis adhesion to lung epithelial cells is considered an essential event for the establishment of infection and different proteins participate in this process. One of these proteins is a 30 kDa adhesin, pI 4.9 that was described as a laminin ligand in previous studies, and it was more highly expressed in more virulent P. brasiliensis isolates. This protein may contribute to the virulence of this important fungal pathogen. Using Edman degradation and mass spectrometry analysis, this 30 kDa adhesin was identified as a 14-3-3 protein. These proteins are a conserved group of small acidic proteins involved in a variety of processes in eukaryotic organisms. However, the exact function of these proteins in some processes remains unknown. Thus, the goal of the present study was to characterize the role of this protein during the interaction between the fungus and its host. To achieve this goal, we cloned, expressed the 14-3-3 protein in a heterologous system and determined its subcellular localization in in vitro and in vivo infection models. Immunocytochemical analysis revealed the ubiquitous distribution of this protein in the yeast form of P. brasiliensis, with some concentration in the cytoplasm. Additionally, this 14-3-3 protein was also present in P. brasiliensis cells at the sites of infection in C57BL/6 mice intratracheally infected with P. brasiliensis yeast cells for 72 h (acute infections) and 30 days (chronic infection). An apparent increase in the levels of the 14-3-3 protein in the cell wall of the fungus was also noted during the interaction between P. brasiliensis and A549 cells, suggesting that this protein may be involved in host-parasite interactions, since inhibition assays with the protein and this antibody decreased P. brasiliensis adhesion to A549 epithelial cells. Our data may lead to a better understanding of P. brasiliensis interactions with host tissues and paracoccidioidomycosis pathogenesis.