Análise do perfil de expressão gênica de carcinoma ductal in situ e invasivo em tumores de mama através da técnica de microarray / Gene expression profile analysis of ductal carcinoma in situ and invasive of the breast cancer by Microarray techonology

O câncer de mama está entre as neoplasias de maior incidência e é responsável pela alta taxa de mortalidade entre as mulheres no mundo todo. O carcinoma ductal é o tipo histológico mais freqüente. O carcinoma ductalin situ (DCIS) inclui um grupo de tumores de mama pré-invasivos com potencial maligno distinto, podendo progredir rapidamente para carcinoma invasivo ou não apresentar evolução durante um longo período da doença. ... Neste estudo foram analisados 40 casos de mama, sendo 5 amostras de tecido de mamário não neoplásico(N), 16 casos de amostras pareadas de carcinoma ductal (in situ e invasivo), 5 DCIS puro, 9 DCIS coexistindo com o componente invasor (DCIS/IDC) e 5 carcinomas ductais invasivos (IDC). .... Dois delineamentos experimentais foram usados: comparação entre lesões in situ e invasivo da mesma amostra (DCIS e IDC); e comparação de grupos de lesões que mimetizam a progressão de câncer ductal de mama, utilizando amostras independentes [células epiteliais de mama capturadas de amostra não neoplásica (N), células tumorais capturadas das lesões: DCIS puro, DCIS/IDC e IDC]. ... Na comparação entre as 3 lesões (DCIS puro, DCIS/IDC e IDC), o DCIS puro mostrou maior divergência molecular, contradizendo os aspectos morfológicos. ... Esse resultado confirmou que o padrão de expressão entre essas lesões é semelhante mesmo avaliando genes sabidamente envolvidos no processo. ... Assim, baseado nos dados desse estudo, demonstramos que as células do carcinoma ductal in situ que coexistem com as células do carcinoma invasivo apresentam alterações moleculares antes da modificação morfológica da lesão, e isso pode ser explorado para se identificar genes alterados que possam predizer a capacidade de invasão. Além disso, esse estudo identificou vários genes candidatos a marcadores moleculares de prognóstico e também preditivos do risco de progressão de doença não invasiva.

Breast carcinoma is one of the most incidence neoplasias and is responsible for a high death-rate among women worldwide. The ductal carcinoma is the most frequent histological type. Ductal carcinoma in situ (DCIS) includes a group of preinvasive breast tumors with distinct malignant potentials. DCIS can have different outcomes, progressing rapidly to invasive cancer or slowly changing over a long period of the disease. Lately, one of the most challenges in molecular research in this field is to identify genes able to predict the risk of progression to invasive disease and prognostic marker. In this study were used 40 breast cases, being 5 non-neoplasic mammary tissues samples (N), 16 matched pairs of ductal carcinoma (in situ and invasive), 5 pure DCIS, 9 DCIS coexisting with invasive ductal carcinoma (DCIS/IDC) e 5 IDC. The RNA from epithelium cells laser capture microdissected were amplified and hybridized using reference design with dye swap in two distinct customized cDNA microarrays platforms. One containing 4.608 cDNA that represent human genes (4.8K) and other containing 390 genes belonging to WNT, PI3K signaling pathways and EMT process (Epithelial-mesenchymal transition). Two designing assays were used: comparisons between lesions in situ and invasive from the same patient (DCIS e IDC) and comparison among lesion groups that mimics the progression of breast ductal carcinoma using independent samples (breast epithelium cells microdissected of non-neoplasic tissues (N), tumoral cells microdissected of lesions: DCIS puro, DCIS/IDC e IDC. In the 4.8k platform, 16 matched-pair samples were used to compare the tumor cells expression profile of DCIS and IDC from the same patient. It was identified 33 candidates differentially expressed (t de Student pairwise - Fold >⎜1,5⎥ e pvalue<0,01), being candidates' genes to be involved in transition from DCIS to IDC. To 4 (LUM, RDH-E2, CXCL13 e POSTN) of 8 selected genes was confirmed differential expression through quantitative RT-PCR (qRT-PCR). Two genes (LUM and CRABP2) over expressed in IDC were selected to verify association with others molecular markers and/or clinicopathological parameters through Tissue Microarray. The LUM protein expression showed positive association with CKs 5/6, CK 14, CK8 e 18 and HER2/neu positive, while CRABP2 showed positive association with ER, PR, CK8, CK18, luminal A, p53 and negatively with CK14. Seeking to characterize molecular aspects of ductal carcinoma in situ of the breast progression, the general gene expression profile among 4 groups that mimics the progression was analyzed (N, DCIS pure, CIS/IDC e IDC) performing ANOVA test (pFDR<0,01) and followed by Tukey´s test, being characterized as most diverge the N group, as expected. The comparison among the 3 lesions (DCIS pure, DCIS/IDC e IDC), the pure DCIS showed the most molecular divergence, contradicting the morphological aspects. To identify genes able to predict the potential risk of invasion of DCIS, it was compared the expression profile of two morphologically identical lesions (DCIS pure e DCIS/IDC), identifying 147 genes (ANOVA - Fold >⎜2⎥ e pFDR<0,01). Hierarchical cluster based on expression profile of those genes could segregate the samples into two distinct groups in which 100% of non-neoplasic samples and 60% of pure DCIS remained in the same group and discriminated from DCIS/IDC (100%). Five (C16orf5, SULF-1, LOX, GOSR2 and TXNL2) candidates were confirmed through qRT-PCR as predict markers of DCIS progression. To assess the functional aspects of invasion process, it was used a platform containing genes belonging to WNT, PI3K signaling pathways and EMT process. The same experimental designing was used for 10 matched-pair, resulting in 32 differentially expressed genes between DCIS e IDC (t de Student pairwise - pvalue <0,05). This result confirmed that expression profile between the lesions is similar even assessing genes involved in this process. Between pure DCIS and DCIS/IDC were identified 15 differentially expressed genes (Wilcoxon ­ p<0,05) whose expression profile could segregate the samples in the same way as for the 147 genes. Triple classifiers were built seeking to segregate the DCIS pure and DCIS/IDC samples. One gene triple belonged to each signaling pathways, WNT (CSNK1A1L, LRP3 and SDC2), PI3K (PLCG2, INPP1 and DGKA) and EMT process (HDGF, CDH13 e TWIST1) were able to segregate the samples being predictor candidates of DCIS progression. Hence, based on this study data, we showed that ductal carcinoma in situ coexisting with invasive ductal carcinoma presents molecular alteration before lesion morphological modifications, and this may be exploited to identify altered genes able to predict invasion capacity. Furthermore, this study identified many candidate genes to molecular markers for outcome and also predicting the risk of noninvasive disease progression.

Effects of Oligo dT-T7 RNA Primer in RNA Amplification from Paraffin-Embedded Tissue for Microarray Experiments

Introduction: Formalin-Fixed Paraffin-Embedded Tissue samples (FFPET) represent a valuable source for studies of geneexpression comparisons, since a great number of these samples is available in archive and presents a long time of clinicalfollow-up. However, the quality of total RNA of these samples is known to be inferior to frozen samples, being many timesinadequate for studies of gene expression using conventional methodologies. Objective: This study aims to establish a protocolfor amplification of messenger RNA (mRNA) derived from FFPET samples for using in microarray experiments. Material andMethods: 4 tumoral samples of invasive ductal breast carcinoma FFPET-buffered 10% were used. Total RNA was extracted andthe mRNA was linearly amplified in two rounds based on T7 RNA polymerase methodology using different concentrations ofoligo dT-T7 Primer for first strand cDNA (1st-cDNA) synthesis. Amplified antisense RNA (aRNA) was labeled with cianine-Cy3through reverse transcription in the presence of random primers and co-hybridized with reference RNA (HB4a) labeled withcianine-Cy5 in a customized platform containing 4,608 cDNAs corresponding to human genes. Results: The amplified RNAquality was influenced by the relative amount of oligo dT-T7, showing better results for ratio of 1:0.1 (total RNA : oligo dT-T7).Hybridizations showed value of intensity signals for the most of cDNAs immobilized in the platform. Conclusion: This studyshowed that the control of the relative amounts of RNA derived from FFPET material and oligo dT-T7 is extremely important toobtain high-quality amplified RNA, allowing its use in microarray experiments.