Conformational Regulation of Vertebrate γ‐Tubulin Ring Complexes by <scp>CM1</scp> Proteins

Microtubules are essential cytoskeletal structures formed by the polymerization of α/β-tubulin subunits. The primary nucleating factor in cells is γ-tubulin, which forms γ-tubulin ring complexes (γ-TuRCs) that act as templates for the assembly of 13-protofilament microtubules (reviewed in Kollman et al. 2011). In vertebrates, γ-TuRCs are formed from γ-tubulin complex proteins (GCPs) 2–6 and mitotic spindle organizing proteins (MZT1–2). Together, they create a cone-shaped structure that organizes γ-tubulin into a semi-helical ring (Wieczorek, Urnavicius et al. 2020; Wieczorek, Huang et al. 2020). γ-TuRCs also contain a lumenal bridge, in which a monomer of actin binds to the inside of the cone. Additional factors such as NEDD1, chTOG/XMAP215, and CM1 proteins (e.g., CDK5RAP2) regulate γ-TuRCs by influencing localization to microtubule organizing centers (MTOCs), and/or by modulating the microtubule nucleating activity of the complex (reviewed in Kollman et al. 2011; Tovey and Conduit 2018). Purified γ-TuRCs exhibit surprisingly low nucleation activity in vitro (Zheng et al. 1995; Consolati et al. 2020; Wieczorek et al. 2021). Cryo-electron microscopy (cryo-EM) studies have revealed a structural mismatch between the γ-TuRC and the microtubule lattice (Kollman et al. 2010; Wieczorek, Urnavicius et al. 2020; Liu et al. 2019; Zimmermann et al. 2020; Consolati et al. 2020). Engineering γ-TuRC subcomplexes to correct these structural mismatches has been shown to enhance microtubule nucleation (Kollman et al. 2015; Serna et al. 2024), suggesting that structural alignment of the γ-tubulin ring may regulate γ-TuRC function in cells. However, the precise mechanisms driving γ-tubulin “ring closure” in the vertebrate holocomplex have not been clear. The centrosomin motif 1 (CM1) is a conserved stretch of ~60–70 amino acids found in proteins like CDK5RAP2 that is implicated in γ-TuRC activation. Overexpression of a fragment of CDK5RAP2's CM1 motif induces ectopic microtubule nucleation, and this fragment can bind to γ-TuRCs and enhance nucleation in vitro (Choi et al. 2010). However, these effects have not always been reproducible (Liu et al. 2019; Thawani et al. 2020), raising questions about how the CM1 fragment truly influences γ-TuRC function. In recent work involving γ-TuRCs purified from porcine brain tissue using CM1 fragment-affinity (Xu et al. 2024), we identified a subset of γ-TuRCs that contained CDK5RAP2 CM1 fragments interacting with GCP2 and MZT2, forming what we call CDK5RAP2-GCP2-MZT2 (CMG) modules. CMG modules dock onto GCP2 subunits and induce structural changes that align the γ-tubulin ring more closely with the microtubule lattice. These modules induce rotations in GCP2 and associated γ-tubulin subunits, which propagate to neighboring γ-tubulins and lead to a conformational change in the γ-tubulin ring. We also confirmed that CDK5RAP2 fragments significantly enhance the ability of both porcine and recombinant human γ-TuRCs to nucleate microtubules. Conversely, disrupting CMG modules—either by mutating CDK5RAP2 or truncating GCP2—abolished this enhancement, indicating that complete CMG modules are essential for promoting γ-TuRC activity. Interestingly, we found that CMG modules are added onto the γ-TuRC in a sequential manner (Xu et al. 2024). Moreover, CM1 fragments do not always form CMG modules when bound to the γ-TuRC (Serna et al. 2024), and their effects on γ-TuRC-mediated nucleation follow a hyperbolic trend (Xu et al. 2024). Such apparent anti-cooperativity could allow cells to fine-tune microtubule nucleation in response to environmental signals, or help protect the microtubule minus end from becoming uncapped from the γ-TuRC by microtubule-associated proteins (Rai et al. 2024). Importantly, we found that CMG-decorated porcine γ-TuRCs display a partially-closed γ-tubulin ring, representing an intermediate state between the open conformation of free γ-TuRCs (Xu et al. 2024) and the closed conformation seen in microtubule end-capped γ-TuRC structures (Aher et al. 2024; Dendooven et al. 2024; Brito et al. 2024). Interestingly, full γ-tubulin ring closure may not be necessary for effective nucleation. For example, bulk microtubule nucleation assays have suggested that partial alignment of the γ-tubulin ring is enough to enhance nucleation (Rice, Moritz, and Agard 2021). Structural snapshots of engineered nucleation intermediates have also revealed that γ-TuRC-associated protofilaments remain flexible during early stages of nucleation, resulting in another “type” of partially-closed γ-TuRC that forms without CMG modules (Brito et al. 2024). It is worth noting that, in contrast with our porcine structures, we identified only one complete CMG module in our reconstruction of the recombinant human γ-TuRC. This CMG module is positioned at a site between GCP2 and GCP6, which is similar to a CMG module we observed previously in CM1 fragment-purified HeLa cell-derived γ-TuRCs (Wieczorek, Huang et al. 2020), as well as in a recombinant complex from another group (Serna et al. 2024). Density for actin is well resolved in the lumenal bridge of both recombinant and native human γ-TuRCs, but is quite weak in the porcine γ-TuRC reconstructions, suggesting a link between the presence of actin in the γ-TuRC and its ability to undergo CMG-deconformational ring closure. Consistent with this, some structures of microtubule end-capped γ-TuRCs also reported diminished actin densities (Brito et al. 2024; Vermeulen et al. 2024), and a parallel study to ours suggested that destabilizing actin within the γ-TuRC coincides with γ-tubulin ring closure (Serna et al. 2024). There is therefore growing consensus that actin plays an important regulatory role in microtubule nucleation from γ-TuRCs, possibly by clashing with GCP subunits in the closed γ-TuRC state (Serna et al. 2024); however, the presence of well-resolved actin in one microtubule end-capped γ-TuRC structure (Aher et al. 2024), as well as the complete loss of the entire lumenal bridge in another structure (Brito et al. 2024), raises some questions about these claims. Overall, recent studies support a model where cytoplasmic γ-TuRCs are structurally inhibited and require activation at favorable sites for nucleation (Kollman et al. 2011). However, key questions remain regarding, for example: the relevance of complete γ-TuRC ring closure; the function of γ-TuRC-bound actin; rate-limiting intermediate nucleation states; and structural discrepancies between in vitro and native-like structures of γ-TuRC end-capped microtubules (Vermeulen et al. 2024). Future studies of nucleation intermediates from γ-TuRCs decorated by CM1 proteins and other accessory factors may help unravel the complex mechanism of microtubule nucleation. The author declares no conflicts of interest.