Endocrine Abstracts

Pituitary tumorigenesis has been analysed from multiple perspectives, yet mRNA expression profiling studies are limited. In this study, microarray analysis was used to identify pathways and networks related to pituitary tumour physiology using key de-regulated genes and bioinformatics analysis. Eight pituitary adenomas (five non-functional tumours, two GH-secreting tumours and a TSH/prolactin-secreting tumour) and a pool of random normal control RNA were profiled for RNA expression using the 29 kb Affymetrix HuGene 1.0 ST chip. Microarray data was analysed using GeneSpring GX 11.0 and network analysis was done using the Ingenuity Pathway Analysis (IPA) software. Data obtained from the microarray was verified on 30 tumours (20 non-functional tumours, six GH-secreting, two prolactinomas and two Cushing’s) using quantitative PCR of key genes involved in the networks identified by the IPA. Different analyses between controls and tumour types were carried out. Among the classical networks discovered known to be involved in pituitary tumorigenesis were the cAMP signalling pathway and the Wnt developmental pathway although both networks were driven by genes not previously described in any other study. Different tumour types were also found to be characterised by variable novel de-regulated molecular pathways, such as the GABA signalling and aryl hydrocarbon receptor-signalling pathways in GH-secreting tumours, and the p53 signalling de-regulation in the non-functional tumours. Cluster analysis from microarray data was also able to distinguish between tumour types, identifying one non-functional tumour as belonging to the functional tumours by its mRNA expression profile. This study validates the use for gene expression profiling for the correct characterisation of pituitary adenoma sub-types and identification of key pathways involved in pituitary tumorigenesis thereby proposing possible therapeutic targets.