Mitochondrial fusion requires the coordinated fusion of the outer and inner membranes. The whole process relies largely on dynamin-like proteins that hydrolyze GTP [4]. For fusion, mitochondria have mitofusins (Fzo1 in yeast) on the surface of their outer membranes [5]. These molecules allow tethering of two organelles before fusion of the outer membrane itself occurs. Lipid mixing of the outer membrane could be catalyzed by lipid-modifying enzymes, such as mitochondrial phospholipase D (mito-PLD) [6]. The mechanism of mitochondrial inner membrane fusion is less clear. It has been demonstrated, however, that it largely depends on another dynamin-like GTPase, Opa1 (Mgm1 in yeast) [7]. It is not known how fusion of inner and outer membranes is coordinated in mammals, but in yeast a third protein of the outer membrane, Ugo1, which interacts with both Fzo1 and Mgm1, may fulfill the role of a membrane fusion coordinator [8]. Mitochondrial fission relies on the cytosolic dynamin-related protein 1 (Drp1 in mammals, Dnm1 in yeast), which uses the protein Fis1 as a receptor on the mitochondrial outer membrane [9]. Mitochondrial dynamins can be regulated by post-translational modifications, including phosphorylation, sumoylation and ubiquitination [9, 10], which impact on their function and consequently on mitochondrial shape and dynamics.

Our knowledge of how mitochondria fuse and fragment has significantly increased over the past decade, mainly thanks to genetic studies performed in Drosophila or yeast that allowed identification of key players of these processes. However, the picture is incomplete and additional components of the fusion and fission machineries certainly remain to be identified. Moreover, the intracellular cascades that control these machineries are not well characterized yet.

In 2004, Jody Nunnari and colleagues [11] were able to induce, for the first time, fusion of isolated mitochondria in vitro. Mitochondria of yeast expressing either mitochondrially targeted GFP or dsRed were isolated, mixed, centrifuged at 4°C to promote membrane tethering, and resuspended at 37°C. Under these conditions, mitochondrial fusion could be observed by confocal or electron microscopy. This cell-free fusion reaction confirmed the requirement of GTP, ATP, an intact membrane potential, and Fzo1 and Mgm1 for fusion of the outer and inner mitochondrial membranes, respectively, as shown in previous cell fusion assays [12]. However, although useful, this cell-free assay is not optimal to obtain a reliable quantification of mitochondrial fusion, in part because the merge of green and red fluorescent markers is only semi-quantitative. Moreover, the high resolution required to image fused mitochondria with a confocal microscope is difficult to combine with automation and high-throughput screening.

These data suggest that this assay could be used as a reliable readout to identify new factors that are part of, or control, the core mitochondrial fusion machinery. Differences in the activities of the cytosols from multiple sources suggest a tissue specificity of the factors regulating mitochondrial fusion. These factors could be purified by classical biochemical procedures, their specific activities being measured by Renilla luciferase activities in the mitochondrial fusion assay. Moreover, high throughput technologies can be envisaged, in particular screening of large libraries of chemicals. At the moment, only mitochondria from cell lines expressing the split luciferases are available. This may restrict the number of assays performed and may be a limitation for high throughput screening. This limitation could be overcome by expressing these luciferase reporters in transgenic mice and by isolating mitochondria from different tissues of these animals. This would also allow testing whether different mitochondria express various forms of the core fusion machinery, which may respond differently to the signaling cascades. In conclusion, this novel assay should be useful to those who more and more are interested in quantifying mitochondrial fusion and should boost research aiming at understanding the mechanisms that govern mitochondrial dynamics. Ultimately, this could lead to better understanding and treatment of mitochondrial diseases in humans.