Final Report Summary - REGAIP (AIP gene regulation and its implications in pituitary tumorigenesis) Background: The aryl hydrocarbon receptor interacting protein (AIP) is a tumour suppressor gene implicated in the pathogenesis of the familial isolated pituitary adenoma syndrome (FIPA), through germline AIP mutations, present in up to 25% of FIPA families. Mutation carriers develop pituitary adenomas at young age, mainly of the growth hormone (somatotrophinoma) and prolactin secreting type, typically large and poorly responsive to therapy. AIP mutations are also found in sporadic pituitary adenomas (2.4-11.7% of young patients with larger tumours). The disease is autosomal dominant, with partial penetrance (~30%). In AIP mutation-negative somatotrophinomas, low AIP expression is a marker of tumour invasiveness and poor treatment response. AIP expression in somatotrophinomas increases following somatostatin analogs treatment and correlates with therapeutic response. Still, the tumour-driving mechanisms of AIP deficiency remain unknown. Aim: The overall objective of project REGAIP (303006) was to characterize the regulation of AIP gene expression, which is expected to improve the understanding of the AIP role in pituitary tumorigenesis and of AIP physiology in general. This could assist in developing new therapies for pituitary adenomas and optimizing the existing ones.Results: We have addressed our objective using a number of complementary approaches.1. Promoter characterization: a) We mapped the AIP gene promoter, using reporter constructs containing sequences of the putative promoter upstream of the translational start and also fragments of coding and non-coding sequence downstream of the start codon, as putative enhancers. We identified minimal sequences required for efficient transcription in transiently transfected rat pituitary tumour GH3 cells and established an enhancer role for an intron 1- exon 2 DNA fragment. b) We described inhibition of promoter construct activity by cAMP pathway activation with forskolin, confirmed by similar changes in endogenous AIP transcript levels. c) Using bioinformatic tools and public ChIP-Seq data from the ENCODE project, we identified putative transcriptional regulators of AIP expression, including the E2F family of cell cycle-regulating transcription factors and demonstrated that cell cycle synchronization reduces AIP levels, in parallel with a reduction of E2F1. d) We confirmed and detailed the effects of a promoter mutation previously described by our group, using mutated promoter reporter constructs and EMSA.2. Study of microRNA regulation of AIP expression: the host laboratory had already developed significant research on this topic since this fellowship’s 2011 grant proposal and effective start (May 2013), involving collaboration with research teams from Hungary and Brazil and resulting in two publications. Therefore, the fellow’s research interest and time were directed towards new avenues of research within the project’s topic, outlined below (paragraphs 4, 5 and 6).3. Study of AIP eQTLs. We have studied the putative expression-modifying role of two AIP variants: a rare non-coding variant c.100-18C>T, a rare variant found in the EXAC exome variant database and observed in several patients with early-onset acromegaly/gigantism - a phenotype consistent with AIP deficiency and c.47G>A, p.R16H a rare polymorphism observed at increased frequency in pituitary adenoma patients. We hypothesised a regulatory role of these variants based on ENCODE ChIP-Seq evidence for transcription factor binding to the variant-containing regions, enhancer database information implicating the c.100-18 area as a tissue-specific AIP regulator and bio-informatic analysis demonstrating loss of putative binding sites for the mutant alleles. We produced luciferase reporter constructs containing a core AIP promoter and either wildtype or mutant allele sequences for each of the two variants, in the vector’s enhancer cloning site. Transfection experiments demonstrated that sequences containing the c.100-18 site significantly increase the reporter activity, compared to a promoter–only control, while the c.47 site containing vectors showed no activity change. The mutated alleles for both variants had no effect on unstimulated or cAMP (forskolin) – stimulated reporter activity, compared to wild-type alleles. We also measured AIP mRNA expression levels in circulating white blood cells from variant carriers and wild-type controls and observed no significant effect of c.100-18C>T on AIP expression levels, whereas transcript levels in c.47G>A carriers were slightly decreased. Our results complement the conclusion of several other experiments performed by Dr. Ferraù (including exploration of splicing defects), showing a lack of effect of the c.100-18C>T variant, thus considered a rare non-pathogenic non-coding variant. This has important consequences for the genetic counselling of variant carriers. 4. The fellow took a key part in a population study of the R304* AIP allele in Ireland, involving subject recruitment and data collection in Mid Ulster (Northern Ireland) and data collection and analysis from additional patient and control cohorts. The fellow co-wrote as first author the manuscript of the study, currently submitted. We demonstrated that the Irish R304* allele is due to a ~2500 years old founder and has a significant impact on today’s acromegaly / gigantism patients and the general population of Ireland. This has implications for AIP mutation screening of pituitary adenoma patients in Ireland and countries with significant Irish immigrant populations (e.g. Canada and the USA). This study offered a great opportunity for scientific outreach, through interaction with the local Mid Ulster community, patients and their families and local medical colleagues, continued after the screening study. The study has created a significant public health impact for the local population, through detection and testing of asymptomatic mutation carriers and generation of public awareness. Results were disseminated through presentations at the W. Harvey Annual Research Review 2014, Departmental Research Updates and ENDO/ICE (Chicago 2014) international congress and an article in the periodical of the Society for Endocrinology, authored by the fellow (https://www.endocrinology.org/endocrinologist/issue.aspx?issue=108). The project was shortlisted for the Project of the Year by the Times Higher Education Awards 2013, making a significant outreach impact (http://www.qmul.ac.uk/media/news/items/smd/113404.html).5. We observed that AIP mutation positive patients have a relatively increased frequency of pituitary apoplexy (PitApo) (Hernandez-Ramirez et al, submitted manuscript) and wondered if PitApo might be correlated with AIP mutations in large patient populations. However, PitApo is rare and prevalence data is scarce; we therefore initiated a retrospective study of pituitary apoplexy in the large KIMS database of over 15000 adult growth hormone-deficient patients treated with growth hormone (GH). Our preliminary results, presented as a poster at the European Congress of Endocrinology (Dublin, 16-20 May 2015) demonstrate that PitApo patients have similar GH replacement requirements and responses to GH treatment, compared to non-functioning pituitary adenoma patients, in terms of IGF-I status, changes in blood lipid levels and quality-of-life scores. These results are of clinical importance for PitApo patients. The study has enabled the fellow to work in an international team of expert clinicians and statisticians, expanding his teamwork abilities and offering excellent networking opportunities.6. The fellow has initiated research into the AIP effects on aryl hydrocarbon receptor (AHR) in the pituitary. To explore the pituitary role of AHR and consequences of AIP deficiency / mutations on pituitary AHR function, we used GH3 cells as a model. We demonstrated AHR expression and response to treatment with the AHR ligand FICZ in GH3 cells and showed that AIP is required for the AHR transcriptional effects, measured with a luciferase reporter and qRT-PCR of endogenous CYP1A1 transcript (a sensitive AHR-responsive gene). We also studied the effects of known AIP mutations and several variants of unknown significance (VUS) on AHR activity, as a surrogate test of pathogenicity. Compared to wild-type AIP (positive control) and EV / stringent mutants (negative controls), VUS can be categorized in three classes: a) Wild-type-like (i.e. polymorphisms), b) Mutant-like and c) Intermediate. The fellow supervised two undergraduate students who contributed to this project. Research training was comprehensive and included webinars and seminars, monthly Endocrine Club Seminars with invited top speakers, hands-on flow cytometry training, a course for new PhD student supervisors, “writing retreat” interactive sessions. The fellow undertook a two-day laboratory visit to the Exeter Genetic Diagnostic Laboratory (Prof. Sian Ellard, April 2015), an excellent opportunity for networking and learning of key genetic diagnosis concepts and protocols, to be implemented at the fellow’s home institution, in a new molecular genetics diagnostic laboratory. Teaching included evaluation of exam papers for the Barts and the London School of Medicine Endocrinology Postgraduate Master, undergraduate teaching (BMedSci lectures at Barts and the London and laboratory supervision of two students), informal supervision and laboratory guidance of three PhD students and an invited postgraduate lecture at the C.I. Parhon National Institute of Endocrinology Bucharest). New transferable skills were acquired and developed, including: immunofluorescence staining and flow cytometry of cultured cells, quantitative real-time PCR, cell culture techniques and good practice, data analysis and presentation, critical manuscript reading and peer-review of multiple manuscripts, preparation and delivery of lectures. The target groups for our research results are the endocrine clinical and research community, including the emerging Romanian one, pituitary patient organisations (FIPA Patients charity) and specific communities affected by AIP-related pituitary adenomas (Irish ethnics).
