Primary research

Gene expression profiling predicts a three-gene expression signature of endometrial adenocarcinoma in a rat model

Sandra Karlsson¹ , Björn Olsson² and Karin Klinga-Levan¹

¹  Systems Biology Research Centre Biomedicine,, School of Life Sciences, University of Skövde, Box 408, SE-541 28 Skövde, Sweden

²  Bioinformatics, School of Life Sciences, University of Skövde, Box 408, SE-541 28 Skövde, Sweden

author email corresponding author email

Cancer Cell International 2009, 9:12doi:10.1186/1475-2867-9-12

Published:	8 May 2009

Abstract

Background

In the Western world, endometrial cancers are the most common gynaecological neoplastic disorders among women. Initial symptoms are often vague and may be confused with several other conditions or disorders. Thus, there is a need for an easy and reliable diagnostic tool. The objective of this work was to identify a gene expression signature specific for endometrial adenocarcinomas to be used for testing potential endometrial biomarkers.

Results

Changes in expression between endometrial adenocarcinomas and non-/pre-malignant endometrium from the BDII EAC rat model were compared in cDNA microarray assays. By employing classification analysis (Weka) on the expression data from approximately 5600 cDNA clones and TDT analysis on genotype data, we identified a three-gene signature (Gpx3, Bgn and Tgfb3). An independent analysis of differential expression, revealed a total of 354 cDNA clones with significant changes in expression. Among the 10 best ranked clones, Gpx3, Bgn and Tgfb3 were found.

Conclusion

Taken together, we present a unique data set of genes with different expression patterns between EACs and non-/pre-malignant endometrium, and specifically we found three genes that were confirmed in two independent analyses. These three genes are candidates for an EAC signature and further evaluations of their involvement in EAC tumorigenesis will be undertaken.