How the hell does dynamin act to fuse membranes?

OK sorry for the lack of updates. I've been rereading lots of mlp literature. Mlp = myosin like proteins. These are found on nucleoplasmic face of the nuclear pore complex (NPC) and seem to filter what goes out of the nucleus, specifically unspliced mRNA precursors. It's old data so I won't go over it here. I recently bumped into the latest instalment of the dynamin in the mitochondria saga.

Jodi Nunnari's group has a paper in Cell about how Mgm1, a dynamin like protein found in the inner mitochondrial matrix, is required for inner-membrane fusion in mitos.

I just don't get the whole thing, can someone propose a reasonable mechanism as to how dynamin regulates membrane fusion?

Background: Dynamin was isolated by the Vallee group as a nucleotide dependent microtubule binding protein. The original thought was that dynamin was another microtubule motor (hence the name). Then a bunch of papers came out, including great papers from Sandy Schmid (then a postdoc in the Rothman lab) describing dynamin as a pinching protein. The idea: as a vesicle buds off from a membrane, dynamin forms a spiral polymer around the neck that connects the vesicle and the membrane.

i-f37e6ec11675b6ab0bcb41d8dd0b016c-dynamin contracting.jpg

The spiral contracts to push the vessicle away from the membrane and to constrict the neck. The result: the vesicle is pinched off. In the Halloween spirit, imagine that dynamin is a torture device around your own neck and it can elongate your neck and constrict it to severe your head. For a great animated illustration of this, click here.

In agreement with this idea, dynamin can spontaneously form spirals (and sometimes rings) in a test-tube. Add on to this the famous shibire mutant and you have a pretty good case that dynamin is a pinch-ase. This fly had a defective dynamin that would paralyze at elevated temperatures. It turns out that the mutation caused dynamin to freeze up and prevent membrane fusion and fission. When the neuro-muscular junctions were visualized up close, people observed unpinched vessicles with giant necks. Around the necks one could see incredibly long spirals ... the pinch-ase seemed broken.

You might think that this is the end of the story. Dynamin promotes membrane fission. But was dynamin really pinching or was it performing some other function? Most vesicles did not require dynamins to pinch off of membranes and no one has realy seen dynamin pinch membranes in vitro. Lots of questions were raised, some by Sandy Schmid herself. Recently the Kirchhausen lab, here at Harvard Medical School isolated a small inhibitor of dynamin, and gave it the name dynasore (as Kirchhausen says, it makes dynamin sore). Treatment with the drug not only prevented the completion of membrane invagination (by pinching) giving a shibire type intermediate, but also inhibited the assembly of invaginations. Thus dynamin may not only act as a pinchase but may help to assemble a budding vesicle.


I'm not exactly sure what people think right now, but with the discovery of dynamin like molecules in mitochondria, things got weird.

It turns out that there are 3 dynamins in yeast mitochondria, Dnm1, Fzo1 and Mgm1. From what I told you from above, you could imagine that a dynamin type molecule could help mitochondrial fission by pinching a mito into two. It turns out that Dnm1 is required for mitochondria to brake into two. Now the strange part, Fzo1 and Mgm1 are required for two mitos to fuse ... more specifically Fzo1 is required to fuse the outer mitochondrial membranes, and Nunnari's group now shows (using genetics and a nice in vitro mito fusion assay) that Mgm1 is required to fuse the mitochondrial inner membrane.

If you believe that dynamins are the fusion machinery, how does this work? It's a real mystery. It certainly does not resemble fusion mediated by SNARES (I'll blog about how the SNARE machinery works in the future). It's a real mystery.

Shelly Meeusen, Rachel DeVay, Jennifer Block, Ann Cassidy-Stone, Sarah Wayson, J. Michael McCaffery and Jodi Nunnari
Mitochondrial Inner-Membrane Fusion and Crista Maintenance Requires the Dynamin-Related GTPase Mgm1
Cell (06) 127:383-395

More like this

Could it be that dynamin molecules interact both with the membrane and with each other, and thereby induce membrane curvature, analogously to what clathrin is thought to do in clathrin-mediated endocytosis?

By PhysioProf (not verified) on 01 Nov 2006 #permalink

That could be part of it. Mgm1 has one transmembrane domain that anchors it to the inner mitochondrial membrane. A fraction of Mgm1 is proteolytically cleaved by a rhomboid protease (these are membrane bound proteases) releasing Mgm1 in to the inner mitochondrial space. According to the intro of the paper, both forms of Mgm1 are necessary genetically.

Schmid did the her dynamin work after she left the Rothman lab.

By Acme Scientist (not verified) on 01 Nov 2006 #permalink

Yeah I was speaking to Tom about it and Schmid was in the Rothman lab in the early 80s where she worked on clathrin. I looked it up and here is the paper that first proposed that dynamin was a "pinch-ase":

Hinshaw JE, Schmid SL.
Dynamin self-assembles into rings suggesting a mechanism for coated vesicle budding.
Nature. 1995 Mar 9;374(6518):190-2.

I guess Schmid now thinks that dynamin has a purely signaling role.

And Physioprof, dynamin could promote the formation of tubes off of the surface of the inner mito membrane, but how would that stimulate membrane fusion?

I do not know much about membrane lipid biophysics, but it seems intuitively that by increasing the curvature of the membrane, you are decreasing the strength of the lateral hydrophobic interactions between aliphatic chains in the plane of the membrane and thus promoting the ability of those chains to engage in interactions with aliphatic chains in another, nearby, membrane. This would then promote fusion.

By PhysioProf (not verified) on 02 Nov 2006 #permalink