Molecular aspects of parathyroid tumorigenesis : with focus on parafibromin and the Wnt pathway

Abstract

Primary hyperparathyroidism (PHPT) denotes the tumorous enlargement of one or several parathyroid glands and constitutes a common disorder, particularly pronounced among postmenopausal women. PHPT patients display hypercalcemia as a consequence of parathyroid hormone hypersecretion, and the symptomatology is habitually based on this metabolic aberrancy. Parathyroid adenomas constitute the vast preponderance of PHPT (80-85%), multiple gland disease/hyperplasia comprises 15% and the malignant parathyroid carcinomas represent about 1% of all cases. Although recent genetic advances have improved our understanding regarding the development of these tumours, a great deal regarding the molecular backdrop to parathyroid tumorigenesis remains obscure.

In paper I, we investigated the frequency of allelic loss at the known parathyroid tumour suppressor MEN1 locus at 11q13 as well as in chromosome regions 1p and 6q in two different patient groups, namely screening detected (all asymptomatic) and routine clinical practice cases (a large fraction with symptomatic disease). Loss at 1p, 6q and 11q were commonly found in similar frequencies suggesting the presence of putative tumour suppressor genes at 1p and 6q. Allelic loss of 1p and 11q was predominantly demonstrated in screening detected patients and appear to confer a more mild-mannered disease, whereas loss at 6q was primarily established among patients recruited from routine clinical practice exhibiting a more clinically explicit hyperparathyroidism.

In paper II, we turned to parafibromin, a protein derived from the HRPT2 tumour suppressor gene recently implicated in the hereditary hyperparathyroidism-jaw tumour (HPT-JT) syndrome as well as in parathyroid malignant disease. We demonstrated that the HRPT2 gene and parafibromin is expressed in various human tissues, and showed that parafibromin is a predominantly nuclear protein. Moreover, three parathyroid adenomas carrying HRPT2 gene mutations were devoid of parafibromin expression, and two cases with wildtype HRPT2 genotype exhibited aberrantly sized parafibromin. All remaining tumours exhibited parafibromin expression, suggesting that loss of parafibromin is a rare event among parathyroid adenomas. In addition, our results support the notion that HRPT2 exhibits tumour suppressor properties in the parathyroid glands.

Parathyroid carcinomas can only be diagnosed if exhibiting an invasive growth pattern or metastases. Since the HRPT2 gene is mutated in the majority of parathyroid malignant tumours, parafibromin has subsequently been proposed as a distinguishing marker for the detection of parathyroid cancer. In paper III, we examined a large group of parathyroid tumours by immunohistochemistry (IHC) using four parafibromin antibodies. The majority of unequivocal parathyroid carcinomas demonstrated reduced parafibromin expression as opposed to retained expression in all benign tumours examined. We conclude that parafibromin IHC can be used as an adjunct, but not a solitary marker when assessing parathyroid tumours, as a positive finding is indicative of benign disease while reduced or negative expression could either signify a parathyroid carcinoma or an HRPT2- mutated parathyroid adenoma. Cases with reduced or absent parafibromin expression should be subjects to HRPT2 gene sequencing to detect possible familial disease.

Following the discovery that parafibromin is associated to the wingless (Wnt) signaling pathway, we assessed a number of Wnt pathway members in paper IV by IHC to investigate whether any of these proteins could be of additional value in separating malignant and benign parathyroid disease. We discovered that adenomatous polyposis coli (APC) expression was completely absent in the majority of carcinomas but retained in all adenomas studied, and propose that APC immunoscreening is a promising tool for the detection of parathyroid malignant tumours. Thus, in paper V we examined a group of atypical parathyroid tumours with unknown malignant potential for parafibromin and APC expression using IHC and found that a subset of atypical tumours displayed concurrent reduced levels of parafibromin as well as absence of APC expression, suggesting that a fraction of atypical tumours display a molecular phenotype similar to malignant tumours which may necessitate intensified follow-up. In addition, as a marker APC confer a higher specificity that parafibromin, however parafibromin might detect familial cases suitable for HRPT2 gene screening, thus making both markers helpful when evaluating parathyroid tumours. Finally, in paper VI we discovered that the APC and RASSF1A promoters exhibited hypermethylation in the vast majority of parathyroid adenomas using bisulphite pyrosequencing. Moreover, parathyroid tumours generally display a genome-wide hypermethylation as compared to normal tissues indicated by LINE-1 pyrosequencing. Hypermethylation of the APC promoter and global hypermethylation was associated to MEN1 and HRPT2 mutational status respectively, indicating that the corresponding proteins menin and parafibromin may exhibit epigenetic regulation properties in parathyroid cells.