In the first period of funding of the ERC grant, we showed that the loss of FH in kidney tubules promotes Atf4 translation and activates the integrated stress response in murine cells, both in vitro and in vivo (Ryan et al., 2021). We also characterised the profound epigenome rewiring that occurs in FH1 loss and are dependent on the activity of the pioneering factor FOXA2 (Rogerson et al., 2023).
During the last funding period, we studied how these co-existing cascades interact and established their hierarchy. To this aim, we generated an inducible mouse model to delete Fh1 in adult animals and investigate the chronology of FH1 loss in vivo in different tissues. Surprisingly, we observed that FH1 loss leads to an early inflammatory response that persists over time and could also be found in HLRCC patient samples (Zecchini et al., 2023). We also found that similar molecular cascades occur in macrophages, in which FH levels decrease upon activation and fumarate accumulates as a consequence (Hooftman et al., 2023). When investigating the molecular mechanisms behind this phenomenon, we unveiled a time-dependent alteration in mitochondrial morphology and ultrastructure. Remarkably, these altered mitochondria release mitochondrial DNA (mtDNA) into the cytosol through the formation of specialised structures known as mitochondria-derived vesicles (MDVs), originating from the inner mitochondrial membrane (IMM). The cytosolic dsDNA sensor cGAS detects the released mtDNA molecules, triggering an interferon response. Crucially, we have demonstrated that adding exogenous, cell-permeable fumarate, but not other TCA cycle metabolites such as succinate, is sufficient to induce this mitochondrial and inflammatory phenotype.
A unique property of fumarate is its ability to modify proteins by succination. Using an antibody to detect protein succination, which we contribute to generate (called thereafter 2SC antibody), we were able to show that succination of mitochondrial proteins is an early event upon incubation of FH1-proficient cells treated with mono-methyl-fumarate (MMF), a cell-permeable derivative of fumarate. Of note, the 2SC-decorated mitochondria exhibit the structural aberrations that coincide with areas of mtDNA release. These results indicate that succination could be responsible for the early structural and ultrastructural changes that poise mtDNA release. Intriguingly, it has been recently shown that in human FH-deficient cancer cells, core components of mtDNA nucleoids, including TFAM and POLG are succinated, possibly affecting the stability of mtDNA (Crooks et al., 2021) and TFAM deficiency has been shown to promote ISG expression via mtDNA release and activation of cGAS/STING (West et al., 2015). It will be crucial to determine whether the changes in mitochondrial membrane lipid composition and cristae structure, along with the succination of components of mitochondrial nucleoids contribute to the formation of MDVs and the ensuing release of mtDNA.
Finally, capitalising on the newly generated conditional Fh1-knockout mouse model, we investigated the acute response to FH1 loss in vivo in multiple tissues. Consistent with what was observed in the epithelial kidney cells, we found a convergent enrichment of inflammatory signatures in the kidney, small intestine, and colon. Although we have not yet investigated the mechanisms underpinning inflammation in the colon and small intestine, it is reasonable to postulate that mtDNA or mtRNA release could govern them, as observed in kidneys and macrophages (Hooftman et al., 2023; Zecchini et al., 2023), respectively. These results indicated that activating an interferon response, mediated by the increased expression of several interferon response genes (ISGs), is an essential and conserved pathway upon FH loss.
