Genetics of Endocrine Tumours - Familial Isolated Pituitary Adenoma - FIPA

We wish to find genes which predispose to pituitary tumours, to find out how those genes work and to assess those genes (and similar genes) in other conditions related to the pituitary tumours.

We will study the recently identified new familial pituitary adenoma gene AIP (AhR interacting protein) and its partner molecules for example AhR (aryl hydrocarbon receptor) in familial and sporadic pituitary adenoma cases.

We have 3 main questions:

Is the identified gene (e.g. AIP) involved in the pathogenesis of familial pituitary tumours (acromegaly), are there any mutations in this gene in these families? What is the function of the new gene in pituitary tumorigenesis? Does the new gene or its partners have a role in sporadic pituitary adenoma tumorigenesis?

Pituitary tumours comprise around 15% of all intracranial neoplasms, and present with distinct clinical characteristics, usually in terms of local space-occupying effects, or secondary to tumoral hypersecretion or its consequences. Acromegaly and gigantism are due in more than 99% of cases to a somatotroph adenoma, which has been demonstrated to be monoclonal in the great majority of instances1. It has been suggested that there are 6 major features of oncogenesis which all need to be present in cases of cancer2, but only 3 of these (activation of an oncogene, inactivation of a tumour suppressor gene [TSG], inhibition of apoptosis) appear to be relevant to pituitary tumorigenesis; these tumours are usually benign adenomas, and the formation of new blood vessels and the capacity for metastasis are uncommon, although some way of evading senescence may be important. It has therefore been suggested that a very small number of mutations in oncogenes and/or TSG's may be causally responsible for pituitary adenomas. This makes them an excellent model for the early stages of tumorigenesis. However, while much has been established regarding the molecular pathology of these tumours, including extensive studies from our own laboratory3-7, the initiating mutation or mutations responsible for tumorigenesis have to date defied analysis.

Early work established that some 30% of patients with acromegaly had one of 2 mutations of the alpha-subunit of the receptor-associated G protein, leading to constitutive activation, but it has been difficult to show that this has relevant biological consequences. We have identified a partial failure of the feedback regulation on somatotroph tumours, but no mutations of the relevant genes have been recorded6,8. We and others have also explored the possibility that there are somatic mutations in sporadic somatotroph tumours of genes identified in some hereditary syndromes associated with acromegaly (MEN-I, Carney syndrome), but these appear to be extremely rare9. However, we have worked in collaboration with a Chicago group who have been seeking to identify the gene responsible for familial acromegaly, a very rare dominant condition10-14. Some 46 families have been described worldwide over the last 40 years, and from a cohort of 8 families a region on chromosome 11q13 has been identified: we have sought to use data from our microarray studies15 to pinpoint the abnormal gene in these patients, but so far unsuccessfully. Our recent work has suggested that there are abnormalities of the cell cycle in pituitary adenomas, especially down-regulation of the cycling-dependent kinase inhibitor p27 and activation of cycling E, and that this is secondary to a putative abnormal growth factor receptor(s), but specific mutations of these receptors are absent16.

Very recently, a Finnish group have identified a dominant gene of very low penetrance which appears to segregate with familial somatotroph and prolactin-secreting tumours17. This lies at or near position 11q13, but it remains unclear as to whether this is indeed the same gene identified in our families with a much more strongly penetrant condition. The gene codes for AIP (= aryl hydrocarbon receptor interacting protein, also known as XAP2 hepatitis B virus X-associated protein 2 or ARA9 = AhR (aryl hydrocarbon receptor)-activated protein 9. AIP has 330 amino acids and has a PPIase-like domain FKBP12 and four tetratricopeptide repeats (TPRs), probably important for protein-protein interactions.

AIP is a putative activating partner for the aryl hydrocarbon receptor, probably increasing the function of AhR. AhR has been linked to the induction of hepatic detoxifying gene products in response to environmental toxins such as dioxin18. However, an additional function appears to be regulation of the cell cycle, suppressing cyclin E and increasing expression of p2719. The AhR also has been shown to interact with cyclic AMP. In this latter instance, the interaction with cAMP appears to compete with the dioxin-dependent pathway, such that AhR has enhanced transport into the nucleus with transcriptional effects quite separate to those stimulated by dioxin and related ligands. As cAMP is an important second messenger in somatotroph tumours, this may be the relevant pathway underlying the apparent activity of AIP as a tumour suppressor.

There is therefore a primary reason to believe that AIP may indeed function as a TSG in pituitary adenomas, and loss of heterozygosity for 11q13 as been seen in tumours in familial cases17. However, the initial Finnish data were only based on 3 families, and we now plan to investigate our entire cohort of families with acromegaly and prolactinomas, as well as further families that we are in the process of collecting. We have contacted many endocrinologists throughout the UK, and we have identified a small number of additional families. Finally, in the initial study some of the patients were apparently isolated cases of acromegaly, but presenting at an unusually young age, especially with gigantism. In such cases too mutations of the AIP gene were recorded.

Plan of Investigation: We plan to screen all known families with a history of acromegaly and/or prolactinomas, as well as other pituitary tumour families to be identified in the UK, for mutations of the AIP gene. In addition we would like to screen other genes related to the chaperon AIP molecule, such as AhR, and other genes currently identified (PDE4A5, survivin and Tom20 protein) or may not been identified. As the data suggest that early onset, aggressive, seemingly sporadic cases of acromegaly or prolactinoma could also be caused by germline mutations, we will include patients with clinically sporadic but early-onset aggressive disease.

The patients will be recruited from Endocrinology Departments from the collaborating centres (Barts Hospital London, Newcastle, Oxford, Stroke-on-Trent, Sheffield, Manchester, Aberdeen and Stroke-on-Trent).

Blood samples will be assessed for germline DNA, RNA and protein isolated for peripheral lymphocytes collected in a peripheral blood sample. Affected family members will be identified, as well as first-degree relatives both affected and unaffected. At the same time, all apparent mutations or polymorphisms of AIP will be tracked in a cohort of 100 germline blood samples from normal volunteers for assessment of the background gene frequency in order to assess their relationship to disease status. Patient tumour samples will collected and RNA and protein expression studied.