A structural model for the spindle assembly abnormal 6 (SAS-6) homologue BLD12 from Chlamydomonas reinhardtii. SAS-6 proteins form dimers through their coiled-coil domains, and subsequent association through their amino-terminal domains results in the formation of a ninefold symmetrical ring with a diameter similar to that of 'cartwheel' structures observed previously. Coiled-coil domains radiate out from the ring. Image is reproduced, with permission, from Kitagawa, D. et al. © (2011) Elsevier.

Centrioles provide the core structure for generating centrosomes, cilia and flagella. They themselves undergo a complex biogenesis process, which begins with the formation of a 'cartwheel' structure, in which nine 'spokes' radiate from a central ring-like hub. Spindle assembly abnormal 6 (SAS-6), which is crucial for centriole assembly, localizes to the cartwheel in several species, and two groups now find that SAS-6 might drive cartwheel formation through assembly into oligomers.

Kitagawa et al. purified the evolutionarily conserved amino-terminal domain of SAS-6 from Caenorhabditis elegans and determined its crystal structure at 2.1 Å resolution. This revealed the presence of dimers of SAS-6, in which the conserved Ile154 residue localized to the interaction interface. van Breugel et al. determined the X-ray crystallography structure of an analogous N-terminal fragment of Sas6 from Danio rerio (N-Sas61–156) at 1.92 Å. Again, head-to-head dimers formed through conserved motifs, including Phe131, the equivalent residue to Ile154 in C. elegans SAS-6, and both groups went on to show that dimer formation is essential for SAS-6 function.

van Breugel et al. showed that mutation of key residues that localized to the dimer interface, including Phe131, disrupted dimerization. These sites were also important for efficient centrosomal localization of SAS-6 in human U2OS cells when transiently overexpressed and for normal flagellum formation in Chlamydomonas reinhardtii. Kitagawa et al. found that disruption of SAS-6 oligomerization by mutation of Ile154 or Phe131 prevented centriole formation in early worm embryos and human U2OS cells, respectively.

Both groups also solved the crystal structure of the N-terminal domain together with a part of the adjacent coiled coil. Combined with the structure of the N-terminal domain alone, this allowed a structural model to be built in which nine homodimers interact through adjacent N-terminal domains and assemble into a ring from which the coiled-coil domains radiate outwards.

Cryo-electron microscopy and rotary metal-shadowing electron microscopy allowed the two groups to ask how this oligomerization of SAS-6 relates to assembly of the early centriole structure. In both cases, the authors saw that the SAS-6 assemblies formed a ring-like structure with a diameter consistent with that observed in vivo, and radiating spokes that were consistent with the length of the predicted coiled coils. Together, these findings suggest that SAS-6 has an evolutionarily conserved role in establishing the ninefold symmetry of the centriole through its ability to oligomerize.