Sicai Zhang

Rho-associated protein kinases (ROCK1/2) are critical regulators of fundamental cellular processes and represent promising therapeutic targets for various pathologies, particularly cardiovascular disease. While the terminal globular kinase domain and the PH-C1 tandem (PHn-C1-PHc) are well characterized, the sequence-structure-function relationship of the central amphipathic α-helical segment remains poorly understood. Utilizing an integrative biophysical approach—including multi-angle light scattering (MALS), analytical ultracentrifugation (AUC), small-angle X-ray scattering (SAXS), negative-stain electron microscopy (EM), X-ray crystallography, and 3D molecular modeling—we elucidate the structural organization of the central coiled-coil region of ROCK1. Notably, the 2.2 Å resolution crystal structure of the homology region 1 (HR1) of ROCK1 reveals a canonical parallel coiled-coil dimer. This structural arrangement diverges significantly from the antiparallel coiled-coil HR1 module found in other Rho effector kinases. Combined with the structural reconstitution of full-length ROCK1, our data demonstrates that the central segment of ROCK1 forms a highly elongated coiled-coil dimer. Furthermore, EM analysis of the isolated full coiled-coil domain uncovers a dynamic oligomerization state, transitioning between a dimer and a tetramer. Together, these biophysical insights establish a foundation that opens avenues for future investigation into ROCK1 regulation, activation, and downstream signaling mechanisms.

Botulinum neurotoxins (BoNTs) are potent bacterial toxins. Distinct BoNT-like toxins, including BoNT/X, BoNT/En, and PMP1, form a distinct lineage. Here we characterize two additional members of this group, PG-toxin 1 and 2 (PG1 and PG2). Their light-chain (LC) and heavy-chain (HC) domains are encoded by adjacent genes. PG1 and PG2 LCs cleave Drosophila SNAP-25 at distinct sites (N196-E197 and R191-K193, respectively); PG1 is more potent, while PG2 activity depends on pH and ionic strength. Both LCs enter Sf9 cells via clathrin-mediated endocytosis. PG1 LC is toxic to Drosophila and northern armyworm independently of its HC, whereas PG2 LC toxicity is enhanced by its HC. Cryo-EM shows that PG1 adopts a distinct compact di-chain architecture with a door-hinge-like “belt.” These findings expand the structural and functional diversity of the BoNT family and support the potential of PG-toxins as insecticidal agents and research tools.

Rho-associated protein kinases (ROCK1/2) play a key role in many fundamental cellular processes and are promising therapeutic targets for various diseases, particularly cardiovascular disease. However, unlike the terminal globular kinase domain and the PH-C1 tandem (PHn-C1-PHc), the sequence-structure-function relationship of the central amphipathic α-helical segment of ROCK proteins is not well understood. This study employed multiple biophysical techniques, including multi-angle light scattering (MALS), analytical ultracentrifugation (AUC), small-angle X-ray scattering (SAXS), negative-stain electron microscopy (EM), X-ray crystallography, and 3D molecular modeling to elucidate the structural basis of the entire ROCK1 segment. The study yielded significant new insights into the structural organization of ROCK proteins. Notably, the crystal structure of the homology region 1 (HR1) of ROCK1 (ROCK ^HR1 ), located at the N-terminus of the amphipathic segment and determined at 2.2 Å resolution, provided the first experimental evidence of a canonical parallel coiled-coil dimer. This structure differs significantly from the antiparallel coiled-coil HR1 module found in other Rho effector kinases. Along with the structural reconstitution of full-length ROCK1, this evidence demonstrates that the central segment of ROCK1 forms a highly elongated coiled-coil dimer. EM analysis using purified ROCK ^FCC revealed that this dimer switches between a dimer and a tetramer.

Diverse TcdB variants use a flexible structural platform to tune their specificity toward different GTPase substrates.

Moving targets of neurotoxins: Proteases that cleave protein targets at specific sequences control many biological functions. The ability to reprogram proteases to cleave new sequences of our choosing would enable new therapeutic and biotechnological applications. Blum et al. report a laboratory evolution method to rapidly evolve proteases that cut new protein sequences and lose their ability to cut nontarget sequences (see the Perspective by Stenmark). Using this method, they evolved botulinum neurotoxin proteases, an important class of enzymes used in patients, to selectively cleave new targets, including a protein unrelated to those natively cleaved by these proteases. This work establishes a powerful approach to generate proteases with tailor-made specificities. Science , this issue p. 803 ; see also p. 782