Functional analysis of the fat mass and obesity associated (Fto) gene and protein

Genome wide association studies have shown that common variants in the human fat mass and obesity-associated (FTO) gene predispose to obesity and increased fat mass. Mice globally lacking Fto are lean, while mice globally overexpressing Fto have increased body weight due to increased fat mass. FTO protein was shown to localise to the nucleus and demethylate ssDNA and ssRNA. However, the mechanisms by which FTO mediates its effects on body phenotype remain unknown. In this thesis, I found that native FTO can be detected in both the nucleus and the cytoplasm during interphase, and that nuclear FTO was exported through the nuclear membrane during early prophase of the mitotic cell division. I developed co-immunoprecipitation (Co-IP) protocol to pull-down native FTO and identified a large number of new candidate binding partners (CBPs). Computational analysis predicted a role for FTO, and many CBPs, in RNA post-transcriptional modification and processing. I confirmed that the E3 ubiquitin-protein ligase TRIM21 interacts with FTO in multiple mouse tissues and binds FTO through its SPRY domain. Importantly, TRIM21 ubiquitinated FTO which did not lead to its degradation. FTO partially co-localised with TRIM21 and the decapping enzyme DCP2 in mRNA processing bodies (p-bodies). Overexpression of TRIM21 led to the accumulation of FTO outside the nucleus, but was reversed when both proteins were overexpressed. Additionally, I created a muscle specific Fto knock-out mouse model and found that lack of FTO in muscle did not result in the body composition phenotype reported in global Fto knock-out mice. Taken together, FTO can function in both the cytoplasm and the nucleus, where it interacts with TRIM21 which ubiquitinates FTO and potentiates its cytoplasmic localisation. Moreover, function of FTO in muscle does not mediate the obesity phenotype in mice.