Z Gastroenterol 2008; 46 - A96
DOI: 10.1055/s-2008-1079700

Identification of loss of function RARRES1, PTGDR and FOXA1 as potential tumor markers by whole genome microarray analysis from laser capture microdissected colon epithelial cells

S Spisák 1, O Galamb 2, N Solymosi 1, F Sipos 1, K Tóth 1, B Molnár 2, Z Tulassay 2
  • 12nd Department of Medicine, Semmelweis University, Budapest
  • 2Hungarian Academy of Science, Molecular Medicine Research Unit, Budapest

Background: In case of microarray data, the origin of signals (epithelium or stroma) cannot be distinguished. To avoid this obstacle, one should use homogeneous cell populations which can help us to get more exact information about the molecular background of tumors.

Our aims were to identify mRNA expression patterns using LCM samples and to compare four different histological regions of colorectal tumor.

Methods: From 6 Dukes B stage, moderately differentiated, left sided colorectal cancer, malignant and normal specimens were collected and frozen immediately after surgery. Using membrane mounted slides, 6µm thick tissue sections were cut. Using PALM LCM system, 5000 and 10000 epithelial and stromal cells were collected from the healthy and the pathological area, and then total RNA were isolated. RNA quality and quantity were checked by Agilent Bioanalyzer Pico 6000 chip kit. Total RNA was amplified and labelled by two-round IVT reaction and hybridized to HGU133 Plus2.0 array. R statistical software and Bioconductor were used for data analysis. Significantly differentially expressed genes were identified by paired SAM test. Validation was done by TaqMan RT-PCR and TMA.

Results: The sample fixation and RNA isolation method in case of LCM samples from few thousands colonic cells was successfully optimized. We identified the differentially expressed genes between normal and tumor epithelial and stromal cells. In tumor epithelial cells 23 genes (RARRES1, PTGDR, FOXA1) were downregulated, while 6 (PLAGL2, FOXQ1) were upregulated compared to the normal epithelium. The stromal cells appeared unmanageable, required more cells and produced slighter results. In this case 12 differentially expressed genes were identified (SULF1, CTHRC1, SFRP4, THBS2, COL8A1, COL11A1, C1R).

Conclusion: The microarray technology combined with LCM allows analyzing of microenvironment of tumor and examination of different rare cell histological region in order to understand the pathomechanism of the tumor.