mTOR signalling

FLCN and AMPK interaction, mediated by FNIP1 and FNIP2, has been shown to regulate mTOR signalling (Baba et al., 2006Hasumi et al., 2008). The mTOR pathway is a key regulator of cell growth, proliferation and metabolism (Harris and Lawrence, 2003; Hay and Sonenberg, 2004; Wullschleger et al., 2006) and an increasing amount of evidence suggests that its deregulation is associated with human diseases, including cancer (Sarbassov et al., 2005; Landau et al., 2009; Montero et al., 2014).

Early studies in fission yeast, Schizosaccharomyces pombe (S. pombe), show that FLCN homologue, BHD1 activates Tor2 (van Slegtenhorst et al., 2007). However, subsequent studies in mammalian models show that FLCN inhibits mTOR signalling (discussed below). This discrepancy may be explained by the fact that BHD1 only corresponds to the N-terminal region of vertebrate FLCN (see Section 6 for alignment). Furthermore, deletion of the Drosophila homologue of FLCN, DBHD, led to a reduction of Tor signalling in Drosophila larvae, suggesting that DBHD activates Tor signalling, but the resultant starvation phenotype was partially rescued by human FLCN, indicating that there is some functional overlap between the two genes, with respect to mTOR signalling (Liu et al., 2013).

The functional role of FLCN in mTOR signalling in mammals is unclear since several publications have reported contradictory effects of FLCN loss on phosphorylated ribosomal protein S6 (p-S6R), an indicator of mTOR activation. Three studies reported that transient downregulation of FLCN by siRNA in human cell lines results in reduction of phosphorylation of p-S6R (Takagi et al., 2008Hartman et al., 2009; Bastola et al., 2013). Reduction of p-S6R was also observed in renal cysts developing in mice heterozygous for FLCN Hartman et al., 2009).

In contrast, kidney-specific homozygous knockout of FLCN resulted in an increase in phosphorylated p-S6R, which contributed to the development of polycystic kidneys (Baba et al., 2008Chen et al., 2008). Additionally, activation of mTOR signalling in the kidney tumours of mice heterozygous for FLCN was noted by Hasumi et al. (2009).

Furthermore, in 2014 the same group found that biallelic FLCN inactivation in murine heart muscle causes cardiac hypertrophy, cardiac dysfunction and significantly reduces lifespan compared to wild type littermates. FLCN deletion led to overexpression of PGC1A, leading to increased mitochondrial mass and high intracellular ATP levels (Hasumi et al., 2014). This led to a reduction in AMPKα phosphorylation at T172, and subsequent mTORC1 dysregulation. In support of this model, Rapamycin rescued the heart disease phenotype of these mice.

This discrepancy between mouse models could be due to differences in sample preparation and/or the gene targeting strategies used between studies. However, Hudon et al. (2010) noted that a loss of FLCN expression in their heterozygous FLCN knockout mouse resulted in both elevated and reduced levels of p-S6R, depending on the cellular context – which may account for the differing results observed in the earlier mouse models. Furthermore, FLCN loss seems to have highly cell-specific outcomes – for example, knockdown of FLCN increases mTOR signalling in SAEC cells but has no effect on mTOR signalling in HBE cells (Khabibullin et al., 2014).

More recently, two Japanese studies have also reported a strong expression of p-mTOR and p-S6R in cyst-lining epithelial cells from BHD patient lungs, suggesting that heterozygous loss of FLCN leads to a dysregulation of mTOR signalling which, in turn, causes lung cyst formation (Furuya et al. 2012; Nishii et al., 2013). Additionally, there is currently no treatment for BHD syndrome, a Japanese patient with metastatic kidney cancer responded better to the mTOR inhibitor Everolimus, than to Tyrosine Kinase inhibitor treatment, suggesting that mTOR inhibition may prove an effective treatment for BHD-associated kidney tumours (Nakamura et al., 2013).

However, three recent studies have shown that FLCN is recruited to the cytosolic lysosome surface during amino acid depletion, where it interacts with the Rag proteins (Martina et al., 2014, Petit et al., 2013; Tsun et al., 2013). FLCN’s interaction with RagA/B was FNIP1-dependent (Petit et al., 2013), whilst FLCN’s interaction with RagC/D was FNIP2-dependent (Tsun et al., 2013), suggesting that FLCN’s interacting proteins modify its function. FLCN was found to specifically interact with the GTPase domain of RagA, suggesting it may function as a GEF towards RagA/B (Petit et al., 2013), whilst FLCN-FNIP2 complex was found to act as a GAP towards RagC/D (Tsun et al., 2013). In all three studies, FLCN was found to activate mTORC1 signalling in response to amino acid stimulation, as shown by increased S6K1, TFEB and TFE3 phosphorylation (Martina et al., 2014, Petit et al., 2013). In support of a role for FLCN in activating mTOR signalling in some conditions, a clinical trial showed that Rapamycin was not an effective treatment for fibrofolliculomas (Gijezen et al., 2014).

Loss of FNIP1 leads to an increase in mTOR signalling and a block in B cell and iNKT-cell development in FNIP1-null mice, suggesting that FNIP1 can inhibit mTOR signalling.  (Park et al. 2012)