BHD is inherited in an autosomal dominant manner, but the mechanism through which loss of a single FLCN allele leads to pathogenesis is not fully elucidated.

Tumour suppression

Knudson’s two-hit hypothesis states that tumour formation is initiated by biallelic inactivation of a tumour suppressor gene, and that both inherited and sporadic cancers can arise as a result of mutations in the same gene (Knudson 1971). Inherited predisposition to cancer results from a heterozygous mutation in a tumour suppressor gene, e.g. FLCN, (the ‘first hit’), but alone is not sufficient for tumour development; inactivation of the wild-type allele by a somatic mutation (the ‘second hit’) is required.

Somatic mutations in the remaining wild-type copy of FLCN or loss of heterozygosity at chromosome 17p11.2 have been identified in BHD-associated renal tumours, supporting Knudson’s “two-hit” hypothesis and a tumour suppressor role for FLCN in the kidney (Vocke et al., 2005).

Haploinsufficiency

Haploinsufficiency describes the situation where a single allele of a gene cannot make a sufficient amount of protein to allow normal cell function or growth.

In a study of five BHD patients, van Steensel et al. (2007) found no evidence of somatic mutations or loss of heterozygosity in fibrofolliculomas, suggesting haploinsufficiency is sufficient to cause benign tumour growth in the skin. This was supported by Bønsdorff et al. (2008), who, in a study of the canine equivalent of BHD syndrome, found ‘second hit’ FLCN mutations in kidney tumours, but not in skin nodules.

Nahorski et al. (2011) found that eight out of ten BHD-associated truncated FLCN proteins were unstable, suggesting that the majority of mutant FLCN proteins are degraded. Furthermore, Benhammou et al. (2011) found 4 families carrying a deletion of FLCN exon 1, which ablated the transcription of this FLCN allele. Both of these studies suggest that FLCN may be haploinsufficient in some contexts.

Dominant negative

Several studies have reported stable expression of truncated FLCN protein (Menko et al., 2013; Laviolette et al., 2013; Luijten et al., 2013). Menko et al. (2013) report the stable expression of FLCN protein in a renal tumour resected from a BHD patient, despite both FLCN alleles carrying a truncating mutation. This suggests that some mutations can produce a truncated FLCN protein, which may have altered or dominant negative function. Furthermore, Luijten et al., (2013). report that ciliogenesis was not rescued in UOK-257-2 cells, which expresses a reconstituted FLCN allele, when compared with the isogenic FLCN-null UOK-257 cell line. This suggests that the mutant form of FLCN present in UOK-257 cells may prevent the reintroduced FLCN protein from functioning in ciliogenesis.

Compound heterozygosity

One study has shown that compound heterozygosity of FLCN and PTEN is the likely cause of oncocytic tumour growth in a Cowden Syndrome patient and a BHD patient (Pradella et al. 2013). The authors found that a BHD-associated parotid tumour carries a somatic PTEN deletion in addition to a germline FLCN mutation, while the Cowden-associated thyroid tumour carries a somatic FLCN deletion in addition to a germline PTEN deletion. The tumours analysed in this study were both oncocytic and displayed characteristic mitochondrial hyperplasia. No mutations in mitochondrial genes, which can cause sporadic oncocytic tumours, were found and array CGH showed that there was no chromosomal instability in the BHD-associated parotid tumour. Although large duplications of chromosomes 5 and 7 were observed in the thyroid tumour, no extensive chromosomal instability was seen in this tumour. Thus, the authors conclude that compound heterozygosity of FLCN and PTEN causes oncocytic tumorgenesis specifically in the context of cancer predisposition syndromes.