Membrane fusion is mainly controlled by proteins from two large families — SNAREs, which are thought to drive fusion, and Rabs, which act as regulators. But tethering factors, which attach the membranes to each other before tight docking and fusion can occur, had so far resisted classification as a protein family. New data from Sean Munro's laboratory now show that they might be distantly related to each other after all.

Whyte and Munro carried out a screen for Saccharomyces cerevisiae mutants that would be synthetically lethal with a mutation in RIC1, a gene that encodes a subunit of the exchange factor for the Golgi Rab GTPase Ypt6. Several genes fitted the bill, among which was the previously uncharacterized DOR1. On further investigation, the authors found that Dor1 is part of a complex that also contains Sec34 and Sec35, two proteins known to be part of a tethering factor complex that functions in the Golgi. The other five subunits of the complex were identified as previously uncharacterized proteins, and were renamed Cod1, Cod2, Cod3, Cod4 and Cod5. So, in one go, Whyte and Munro identified six new proteins and characterized the composition of the Sec34/Sec35 tethering complex.

But, although the eight proteins form a complex, the deletion mutants had different phenotypes. Whereas deletion of COD1 was lethal, mutants lacking COD3, SEC34 and SEC35 grew very slowly and had defects in the organization of their internal membranes and in Golgi processing. By contrast, mutants lacking DOR1, COD2, COD4 and COD5 were viable and had no defect in these processes. However, they did have a defect in recycling from endosomes to the Golgi. So, the Sec34/Sec35 complex probably acts at more than one transport step in the Golgi.

Iterative PSI-BLAST searches showed that several members of the Sec34/Sec35 complex are related to other tethering factors. Dor1 is distantly related to Sec5, a component of the exocyst — a complex that functions in tethering exocytic vesicles to specific sites on the plasma membrane. Cod4, on the other hand, is the yeast homologue of mammalian GTC-90 — the only identified component of the so-called Golgi transport complex (GTC). The obvious question as to whether GTC and the Sec34/Sec35 complex are one and the same remains to be addressed.

The authors noticed a short region of homology in the amino terminus of all eight members of the complex, as well as in the amino termini of several other tethering factors, including all eight components of the exocyst, and Vps53 and Vps54 — members of the Vps52/Vps53/Vps54 complex, which functions in prevacuole-to-Golgi transport. This region consists of two amphipathic helices separated by an extended loop, which probably form a coiled coil. Moreover, the homology between Sec34 and Exo70 (a component of the exocyst), and between Sec3 (another component of the exocyst) and Vps52 (a component of the Vps52/Vps53/Vps54 complex) extend well beyond this region, indicating a closer relationship between these factors.

So, it seems that tethering factors might after all be a family — admittedly, one with loose ties. This new finding should make it easier to study their function, if you believe in the motto 'you know one, you know them all'.