Cell-cycle gene expression analysis using real time PCR in locally advanced squamous-cell head and neck cancer

PURPOSE

The analysis of gene expression, especially those involved in cell cycle control, can help to discover mechanisms determining the outcome of radiation treatment. The main purpose of this study was to examine the expression level of genes responsible for cell cycle regulation in samples of the head and neck cancer, obtained during surgery.

METHODS

Postsurgical samples of SCC of head and neck region were collected. Over 80 genes were analysed using cell cycle quantitative real-time RT-PCR Array method. Presence of 14 high-risk HPV types DNA in frozen or paraffin-embedded tumour pathological samples was also assessed. To correlate gene expression with selected pathological features and clinical outcome we used different hierarchical clustering method.

RESULTS

Hierarchical clustering demonstrated the association between gene expression within certain clusters and gender, tumour site, T stage, N stage, grade, pathological subtype or tumour recurrence.

CONCLUSIONS

Despite some limitations we were able to identify gene clusters that allowed to classify patients according to selected clinical features and occurrence of tumour recurrence. The results of the analysis also confirm that the incidence of HPV infection among the patients from Upper Silesia is relatively low, whereas HPV negative tumours, likely associated with smoking, appeared dominant.

Gene expression profiles in three histologic types, clear-cell, endometrioid and serous ovarian carcinomas

Ovarian carcinoma is the most lethal type of gynecologic malignancy in the Western world. Majority of early stage ovarian cancers are asymptomatic and this is the main reason that more than two-thirds of patients are diagnosed with advanced disease. Ovarian tumors are heterogeneous and the different histologic subtypes are further classified as benign, borderline (low-grade) and malignant (high-grade) to reflect their behavior. The aim of the study was to analyze gene expression profiles in three histologic types of ovarian carcinoma in an attempt to find the molecular differences among serous, endometrioid and clear cell subtypes. The analysis of gene expression was performed on 57 samples of ovarian carcinoma. RNA was isolated from the ovarian cancer tissues. The gene expression changes were determined by microarray analysis and quantitative real time polymerase chain reaction (qRT-PCR). Measurement of relative gene expression levels was used to identify molecular differences among three histologic types of ovarian carcinoma (clear-cell, endometrioid and serous). Unsupervised statistical analysis revealed four biological subtypes among three histotypes under study. The endometrioid ovarian carcinoma was divided into two molecular subtypes. The biggest molecular differences were observed between clear-cell and serous carcinomas (1070 genes, FDR 0.05), the smallest between endometrioid and serous carcinomas (81 genes, FDR 0.05). The biggest group of differentially expressed genes was involved in transport and metabolism. This finding can explain the differences in the response to chemotherapy observed among different histologic types of ovarian carcinomas. In conclusion, we found TCF2 (HNF1B) gene as a suitable marker for ovarian clear cell carcinoma. Gene expression profiling also shed light on the molecular mechanisms of different chemoresistance among the analyzed histotypes.

Calculation of reliable transcript levels of annotated genes on the basis of multiple probe-sets in Affymetrix microarrays

Microarray methods have become a basic tool in studies of global gene expression and changes in transcript levels. Affymetrix microarrays from the HGU133 series contain multiple probe-sets complementary to the same gene (4742 genes are represented by more than one probe-set in a microarray HGU133A). Individual probe-sets annotated to the same gene often show different hybridization signals and even opposite trends, which may result from some of them matching transcripts of more than one gene and from the existence of different splice-variant transcripts. Existing methods that redefine probe-sets and develop custom probe-set definitions use mathematical tools such as Matlab or the R statistical environment with the Bioconductor package (Gentleman et al., 2004, Genome Biol. 5: 280) and thus are directed to researchers with a good knowledge of bioinformatics. We propose here a new approach based on the principle that a probe-set which hybridizes to more than one transcript can be recognized because it produces a signal significantly different from others assigned to the particular gene, allowing it to be detected as an outlier in the group and eliminated from subsequent analyses. A simple freeware application has been developed (available at www.bioinformatics.aei.polsl.pl) that detects and removes outlying probe-sets and calculates average signal values for individual genes using the latest annotation database provided by Affymetrix. We illustrate this procedure using microarray data from our experiments aiming to study changes of transcription profile induced by ionizing radiation in human cells.

Calculation of reliable transcript levels of annotated genes on the basis of multiple probe-sets in Affymetrix microarrays

Microarray methods have become a basic tool in studies of global gene expression and changes in transcript levels. Affymetrix microarrays from the HGU133 series contain multiple probe-sets complementary to the same gene (4742 genes are represented by more than one probe-set in a microarray HGU133A). Individual probe-sets annotated to the same gene often show different hybridization signals and even opposite trends, which may result from some of them matching transcripts of more than one gene and from the existence of different splice-variant transcripts. Existing methods that redefine probe-sets and develop custom probe-set definitions use mathematical tools such as Matlab or the R statistical environment with the Bioconductor package (Gentleman et al., 2004, Genome Biol. 5: 280) and thus are directed to researchers with a good knowledge of bioinformatics. We propose here a new approach based on the principle that a probe-set which hybridizes to more than one transcript can be recognized because it produces a signal significantly different from others assigned to the particular gene, allowing it to be detected as an outlier in the group and eliminated from subsequent analyses. A simple freeware application has been developed (available at www.bioinformatics.aei.polsl.pl) that detects and removes outlying probe-sets and calculates average signal values for individual genes using the latest annotation database provided by Affymetrix. We illustrate this procedure using microarray data from our experiments aiming to study changes of transcription profile induced by ionizing radiation in human cells.