A key function of NEDD8 is to allow cullin-RING ligases to work with other molecular machines – called E2s or ARIH1 – each of which is an intermediary that transfers ubiquitin to target proteins (substrates) recruited to CRLs. Like actual machines, NEDD8-activated CRLs have a lot of different parts, and the parts move. Only when the right inputs (ubiquitin-linked E2 or ARIH-family RBR E3, along with a substrate) are put into a machine, and when the inputs come together, then the product of the machine (ubiquitin linked to substrate, or a chain of ubiquitins linked to each other and then to substrate) is generated in a transient assembly. Before starting the project, we knew what the parts of the NEDD8-activated CRLs and the inputs were. But we lacked a blueprint for how they are put together to connect ubiquitin to the targeted protein. Our project involved devising new methods and new tools using organic chemistry, which allow essentially “freezing” the NEDD8-activated cullin-RING ligases, with either an E2 or with ARIH-family RBR E3 that are themselves linked to ubiquitin, and the target with the moving parts placed essentially as needed to put ubiquitin onto the specific targets. Cullin-RING ligases perform several different reactions, each requiring tailor-made harnesses to "freeze" ethem. With our chemical biology toolkit in-hand, we applied these to our purified complexes and examined the assemblies using state-of-the-art cryo electron microscopy (cryo-EM). Cryo-EM essentially allowed us to see the complexes assembling ubiquitin onto substrates in 3-dimensions. The structures provide a blueprint for how many proteins are turned off.
We also revealed a systemwide mechanism for multiprotein complex formation (for many NEDD8-activated cullin-RING ligases): a limiting component is recycled from idling complexes to fuel mixing-and-matching of parts and transient stabilization of the subset of complexes needed at a given time. This averts supply chain problems, obviates a need for producing new parts, prevents buildup of superfluous and potentially toxic molecular machines, and in the case of CUL1-containing cullin-RING ligase complexes, allows rapidly establishing degradation pathways needed for cellular regulation.
With these structures in-hand, we knew the shape of the active molecular machines. We then made probes - essentially hooks - that allowed us to fish out and identify NEDD8-activated cullin-RING ligases in the cell. The collection of NEDD8-activated cullin-RING ligases - of about 80 we could detect using our method - varies in different cell types, and in different conditions like immune signaling, metabolic changes, and chemotherapy drugs. We are exploring if our method will allow us to determine if some cell types are more responsive to certain types of medications (which depend on NEDD8-activated cullin-RING ligases to work) than others.
