Identificação de marcadores moleculares em câncer de mama através da técnica de microarray utilizando uma plataforma de exons tumor-associados / Identification of breast cancer molecular markers through microarray technology using a tumor-associated exons platform

RESUMO

Análises recentes têm mostrado a ocorrência de splicing alternativo (AS) do mRNA em pelo menos 60% dos genes humanos, sendo que 80% desses eventos ocorrem dentro da região codificadora, aumentando a diversidade proteômica. ... Para identificar variantes de splicing diferencialmente reguladas em câncer de mama, 270 exons expressos em tecidos ... Esses exons foram imobilizados em membranas de nylon juntamente com controles positivos e negativos, e hibridizados contra amostras tumorais e normais de mama. ... Para validação técnica dos exons selecionados como superexpressos de acordo com os critérios estabelecidos, foi empregada a técnica de RT-PCR quantitativo (qRT-PCR), usando o mesmo grupo de amostras já utilizadas anteriormente. ... Os resultados mostraram que a razão do nível de expressão entre as 3 VCE e a expressão constitutiva do gene (VCE/EC) foi significativamente maior em amostras tumorais de mama quando comparadas às amostras normais (p<0,05), sugerindo que TRIM37-VCE, MK-STYX-VCE e BRRN1-VCE são, de fato, variantes de splicing superexpressas em câncer de mama. Essas variantes foram também avaliadas em um grupo independente de 40 amostras tumorais de mama para validar biologicamente a superexpressão da VCE em amostras tumorais de mama quando comparadas às amostras normais. Todos os dados foram correlacionados com características clínicas e histopatológicas das amostras.

ABSTRACT

Current analyses have shown that alternative mRNA splicing (AS) appears in at least 60% of human genes and 80% of these events occurs within the coding region, increasing the proteomic diversity. Some AS variants have been preferentially expressed in human tumors and are potential molecular markers, contributing to the development of more accurate diagnostic and prognostic factors as well as therapeutic targets. To identify differentially regulated splicing variants in breast cancer, 270 exons expressed in tumor tissues were selected by a computational analysis, of which 75 were associated with breast, because they are found to be expressed in libraries that originated from breast tumors and they were not found in the corresponding normal libraries. These exons were immobilized on nylon membranes together with positive and negative controls, and hybridized against tumor and normal breast samples. To identify the most highly expressed exons in tumor tissues, 3 comparisons were performed 4 tumor against 2 normal breast cell lines (LTxLN), 27 tumor against 5 non-neoplasic breast tissues (TxN) and 4 matched tumor-normal samples (PTxPN). A Tstudent test was used to select for differentially expressed exons (p<0.05) in each comparison (LTxLN; TxN; PTxPN). Here, 24 were selected as over expressed exons in the LTxLN comparison, 79 exons in the TxN comparison and 195 exons in the PTxPN comparison. For technical validation, those exons having a fold change of ≥ 3 in tumor samples and being present in at least 2 comparisons were selected. Fourteen exons were identified by microarray experiments and evaluated through quantitative RT-PCR (qRTPCR), using the same sample set utilized previously. In order to test whether the exons selected by microarray experiments belonged to a prone over expressed splicing variant, 2 criteria were adopted (1) Validation by qRTPCR of the variant that comprises the selected over expressed exon (VCE) (fold change ≥ 3); (2) For those exons confirmed in criterion (1), evaluation of whole gene expression (through the analysis of constitutive gene expression) is adopted as a secondary criterium. Three of the VCE's were confirmed through qRT-PCR as being over expressed in breast tumor, such as TRIM37-VCE, MK-STYX-VCE and BRRN1-VCE. The constitutive expression of the gene (EC) was analyzed by qRT-PCR, through the primer design within constitutive exons, that is, present in all variants of the gene. The results showed that the expression level ratio between the 3 VCE's and the constitutive expression of genes (VCE/EC) was significantly higher in tumor samples when compared to normal samples (p<0.05), suggesting that TRIM37-VCE, MK-STYX-VCE and BRRN1-VCE are indeed breast tumor associated variants. These variants were also evaluated in an independent set of 40 breast tumor samples to biologically validate the VCE over expression in tumor samples as compared to normal samples. All data were correlated with clinical and histopathological samples features, and some significative associations were found, such as the expression of estrogen (ER+) and progesterone (PgR+) receptors with TRIM37-VCE. Although an increase of the experimental data is required for the complete exploration of this study, the results suggest 3 splicing variants that are over expressed in ductal carcinoma of the breast, and are candidates for molecular markers.