Optimised laser microdissection of the human ocular surface epithelial regions for microarray studies

Background: The most important challenge of performing insitu transcriptional profiling of the human ocular surface epithelial regions is obtaining samples in sufficient amounts, without contamination from adjacent tissue, as the region of interest is microscopic and closely apposed to other tissues regions. We have effectively collected ocular surface (OS) epithelial tissue samples from the Limbal Epithelial Crypt (LEC), limbus, cornea and conjunctiva of post-mortem cadaver eyes with laser microdissection (LMD) technique for gene expression studies with spotted oligonucleotide microarrays and Gene 1.0 ST arrays. Methods: Human donor eyes (4 pairs for spotted oligonucleotide microarrays, 3 pairs for Gene 1.0 ST arrays) consented for research were included in this study with due ethical approval of the Nottingham Research Ethics Committee. Eye retrieval was performed within 36 hours of post-mortem period. The dissected corneoscleral buttons were immersed in OCT media and frozen in liquid nitrogen and stored at -80 degrees C till further use. Microscopic tissue sections of interest were taken on PALM slides and stained with Toluidine Blue for laser microdissection with PALM microbeam systems. Optimisation of the laser microdissection technique was crucial for efficient and cost effective sample collection. Results: The starting concentration of RNA as stipulated by the protocol of microarray platforms was taken as the cut-off concentration of RNA samples in our studies. The area of LMD tissue processed for spotted oligonucleotide microarray study ranged from 86,253 mu m(2) in LEC to 392,887 mu m(2) in LEC stroma. The RNA concentration of the LMD samples ranged from 22 to 92 pg/mu l. The recommended starting concentration of the RNA samples used for Gene 1.0 ST arrays was 6 ng/5 mu l. To achieve the desired RNA concentration the area of ocular surface epithelial tissue sample processed for the Gene 1.0 ST array experiments was approximately 100,0000 mu m(2) to 130,0000 mu m(2). RNA concentration of these samples ranged from 10.88 ng/12 mu l to 25.8 ng/12 mu l, with the RNA integrity numbers (RIN) for these samples from 3.3 to 7.9. RNA samples with RIN values below 2, that had failed to amplify satisfactorily were discarded. Conclusions: The optimised protocol for sample collection and laser microdissection improved the RNA yield of the insitu ocular surface epithelial regions for effective microarray studies on spotted oligonucleotide and affymetrix platforms.