Why do we need three dynamins? (Comment on DOI 10.1002/bies.201200033)

Abstract

All living organisms must exchange matter with the environment. Clathrin-mediated endocytosis (CME) is a fundamental process by which eukaryotic cells engulf und internalize substances from the surrounding. After cargo-containing invaginated clathrin-coated pits are formed, dynamin forms collars around the necks of these nascent vesicles and, by acting as a mechanochemical enzyme, uses the energy from GTP hydrolysis to mediate fission from the plasma membrane. Dynamins are large GTPases build up of five domains: the G domain (responsible for GTP substrate binding and hydrolysis), the middle domain, the pleckstrin-homology domain (PHD, which mediates binding to membrane lipids), the GTPase effector domain and the proline-rich domain. Vertebrates encode three dynamins. The neuron-specific dynamin 1 and the ubiquitously expressed dynamin 2 participate in CME. The function of dynamin 3 is poorly understood; it is found in testis and brain. In this issue, Liu et al. 1 discuss the evolution of the dynamins by looking primarily at the PHD. A counterintuitive order of appearance was suggested, in which the neuronal dynamin 1 evolved first, and the ubiquitous dynamin 2 was a later invention. Further, the evolution of dynamin 1 was suggested to have been driven by the development of the central nervous system (CNS). The discussion is based on a phylogenetic analysis presented in the paper and earlier biochemical findings made by the authors 2. They suggest that dynamin 1 is a curvature generator and that dynamin 2 is a curvature sensor, and that these distinct properties are mainly due to Tyr600 and Leu600, respectively. The key residue is part of the variable loop 3 (VL3), one of the three VLs of the PHD forming the phospholipid interaction pocket. Indeed, exchanging the VL3 leucine for tyrosine turned dynamin 2 into a potent curvature generator. In agreement, the curvature-generating potency of dynamin 1 was reduced by a Y600L mutation. Leu600 is only found in dynamin 2; dynamin 1, and dynamin 3 encode a Tyr600. Drosophila and Caenorhabditis elegans, fly and worm, possess only one classical dynamin, which is primarily expressed in the nervous system and has a tyrosine in VL3. Interestingly, in the 1970s, the importance of dynamin for neuronal signal transduction was demonstrated for the first time through a mutation, shibire, in its gene, which caused a reversible temperature-dependent paralysis in fly. Liu et al. identified dynamin-coding sequences in the protists Monosiga brevicollis and Capsaspora owczarzaki, and early metazoans that lacked CNS. In the sequences, a histidine was found in VL3. Remarkably, a flatworm (Schistosoma mansoni) owning a CNS was shown to encode two dynamins, one containing histidine and the other tyrosine in the VL3. It appears likely that the powerful curvature generator, dynamin 1, with a VL3 tyrosine, evolved for the special needs of neuronal signal transmission, where rapid recycling of synaptic vesicles is required. Dynamin 2 could have evolved in vertebrates as a sensor that controls fidelity of CME and releases only mature vesicles from plasma membrane. Could the VL3 residue also contribute to lipid composition specificity, and control dynamin isoform-specific functions in distinct cell types and subcellular compartments? What is dynamin 3 needed for? The dynamin superfamily encompasses proteins – classical dynamins and dynamin-like proteins – participating in deformation of biological membranes, and involved in multiple cellular processes from endocytosis and vesicle trafficking to division of organelles and cells. It is believed that the proteins originally evolved as components of cell and organelle division machineries. Is the existence of an additional dynamin in testis, the place of spermatogenesis and a multitude of cell division processes, just a coincident? Intriguingly, there is a further unusual large GTPase expressed only in testis. Irgc belongs to the immunity-related GTPases (IRGs), a family that consists of more than 20 members in the mouse, but is almost lost in man 3. IRGs participate in the destruction of the membrane surrounding the protozoan parasite Toxoplasma gondii in infected cells. Irgc is highly conserved between mouse and human, and is the only full-length remnant of the IRG family in humans. Furthermore, Irgc is probably the only IRG protein not induced by interferons and without a function in immunity.