Skip to main content

The molecular basis of CULLIN E3 ligase regulation by the COP9 signalosome

Periodic Reporting for period 4 - CsnCRL (The molecular basis of CULLIN E3 ligase regulation by the COP9 signalosome)

Reporting period: 2020-07-01 to 2021-02-28

Despite intense efforts over decades, many proteins implicated in human diseases have eluded pharmacological intervention. A novel type of drug targeting methodology has recently emerged that takes advantage of the ubiquitin transferase systems. Cullin-RING E3 ubiquitin ligases (CRLs) are a large family of regulatory protein complexes, responsible for ~20% of intracellular protein degradation in humans. CRLs are divided into six families based on their Cullin constituent (either CUL1, CUL2, CUL3, CUL4A/B or CUL5). Each Cullin binds a RING-domain containing protein (RBX1 or RBX2) and a vast repertoire of adaptor/substrate receptor modules, collectively creating more than 240 distinct CRLs. The structures of CRLs and their substrates vary enormously. All ~240 CRL family members are controlled by the COP9 signalosome complex (CSN). CSN removes NEDD8, an activator protein covalently attached to the Cullin C-terminal domain, and thereby inactivates the ligase. Prior to this work it had been unclear what enables a single complex, CSN, to be a master regulator of all CRLs. We studied how ubiquitin ligases can deal with activating signals ranging from DNA damage to modified proteins and how small-molecule degrader drugs exert their therapeutic benefit via the ubiquitin system.Overall objectives: The aim of the ERC grant is to understand the architecture, working and regulation of Cullin-RING ligases by the Cop9 signalosome. Given recent developments in the field of small molecule degraders, we extended our focus to questions of how small molecule drugs can reprogram CRLs and how these drugs interplay with CSN and other CRL regulators.Key learnings enabled by the ERC project: We were able to solve the first sub-nanometer resolution structure of CSN bound to a CRL substrate. By combining single particle electron microscopy with functional studies, we demonstrated that bulky substrates, irrespective of their chemical nature and protein sequence, are able to sterically displace CSN from their cognate CRL E3 ligase, providing a first universal mechanism for the regulation of diverse CRLs by diverse substrates (Cavadini et al., Nature 2016). The CRL ubiquitin Ligase System and its CSN master regulator, thus respond not to a specific protein epitope, but instead to the bulk of the protein. We found that this mechanism is similarly exploited by small molecules recruiting neo-substrates, for which we determined structures and provided the mode of action (Petzold et al., Nature 2016; Sievers et al., Science 2018; Słabicki et al., Nature 2020). If the proteins recruited by these small molecules have sufficient bulk and size, they are being recognized as true biological substrates, and as a consequence, the ligase remains intact and operational for the degradation of these neo-substrate targets. These principles of CRL action will guide future drug development efforts for novel therapeutics.Importance of our finding for human health: Protein degradation through the ubiquitin proteasome system is involved in virtually all biological processes. Accordingly, mutation of the ubiquitin ligases and their regulators are frequently found in cancer cells, as well as in other diseases ranging from neurodegeneration to developmental defects. Understanding these diseases and coming up with novel therapeutic solutions requires atomic-level understanding of the proteins involved.
Thalidomide and its second-generation derivatives lenalidomide and pomalidomide, together known as immune modulatory drugs (IMiDs), are potent anticancer drugs that are widely used in the treatment of multiple myeloma and 5q-deletion myelodysplastic syndrome (5q-MDS). IMiDs induce degradation of different proteins upon binding to the CRL4CRBN E3 ubiquitin ligase. In multiple myeloma, IMiD efficacy depends on the ability to degrade the Ikaros/Aiolos transcription factors. In 5q-MDS, however, IMiDs are effective through the degradation of casein kinase isoform 1 alpha (CK1A1). How IMiDs induce the degradation of proteins when bound to the CRL4CRBN ligase, and how they interplay with CRL regulators, had remained elusive. In the context of the ERC grant, we reported the crystal structure of the ternary complex between the substrate receptor CRBN and the DDB1 adaptor of the CRL4CRBN E3 ligase, in the presence of lenalidomide and the CK1A1 neo-substrate. The compound together with a surrounding surface patch on CRBN acts as a molecular glue recruiting CK1A1 to the E3 ligase, where it is juxtaposed to the CUL4/Rbx1 moiety of the active E3 ligase (Petzold et al., Nature 2016). We subsequently showed that zinc-fingers IKAROS/AIOLOS, the efficacy target of these drugs in multiple myeloma, are also bound and degraded in this manner and that different small molecules can tune the specificity for different zinc-fingers (Sievers et al., Science 2018). We then focused on other drugs that induce neo-morphic protein-protein interaction with Cullin-RING ligases: protein kinase inhibitors have long been suspected to have a degradation component inherent in their mode of action. We identified a cyclin-Kinase inhibitor that targets a kinase not only for inhibition but also for gluing to and subsequent degradation by a ubiquitin ligase (Słabicki et al., Nature 2020). These neo-substrates and drug-induced degradation processes are compatible with the steric CSN mechanism dependent on the bulk of the CRL bound substrate (Cavadini et al., Nature 2016) This suggests that other inhibitors may be turned into more proficient molecular-glue degraders, thereby opening novel and innovating ways to target proteins that are currently out of reach for traditional drug discovery.Persistent sun irradiation of cells results in the covalent crosslink of neighboring pyrimidine bases in the DNA. These cancer-causing lesions are repaired with the help of CRLs. The DDB2 damage sensor is embedded in a Cullin4-RING E3 ubiquitin ligase (CRLs) complex together with DDB1. In the context of the ERC grant, we revealed the regulation and targeting of this ligase (Cavadini et al., Nature 2016; Matsumoto et al., Nature 2019). We solved the structure of the DDB1-DDB2 complex bound to 6-4PP lesions on chromatin, and bound to CSN; this demonstrated how these ubiquitin ligases are regulated in their natural chromatin habitat. This work provided a rationale for DNA repair defects in cancer-prone xeroderma pigmentosum patients. We then showed that the principles of damage detection in chromatinized DNA, we delineated, also apply to the broad class of transcription factors (Michael et al., Science 2020). These long sought-after structural snapshots and the accompanying mechanistic studies provide a first conceptual framework for how ubiquitin ligases and TFs read-out DNA on nucleosomal DNA templates.
The ERC has been used to establish cutting-edge cryo-electron-microscopy (cryo-EM) capabilities in the laboratory. The CSN-CRL work has laid the foundation and work-flow for ongoing cryo-EM studies, and cryo-EM has now become an indispensable technology in the laboratory. An example for such a project concerns the immune sensor cGAS, where we were able to extend technical learnings in structural biology to another chromatin bound complex. The cGAS-STING pathway is a highly conserved mechanism for the recognition of foreign DNA. Recent studies found that highly aggressive tumours have co-opted this signalling pathway to also drive tumorigenic behaviour. cGAS is found both in the cytoplasm and in the nucleus, hence posing the question of self versus non-self discrimination of this important foreign DNA sensor. Together with the Ablasser group (EPFL, Lausanne), we showed that cGAS is inhibited by genomic self-DNA and provided a structural explanation for the control of cGAS by genomic DNA (Pathare et al., Nature 2020). In the nucleosome bound configuration, foreign DNA binding to cGAS and subsequent activation by dimerization is prevented by steric clashes. The combination of structural and functional studies provided the long-sought answer how cGAS activity is tuned in the nucleus. Our work outlined that instead of focusing on conserved pathogen-specific features, cGAS increases sensitivity, while averting autoreactivity in the nucleus by the suppressive activity of histones, virtually “inbuilt identifiers” of eukaryotic genomes. Taken together our work in the context of the ERC AdG, resulted in multiple visible publications in Nature, Science, Molecular Cell and Nature Molecular and Structural Biology, it furthered our understanding of the basic biology of ubiquitination and illuminated novel ways in which this pathway can be hijacked by small molecules. This progress would not have been possible without the support of the ERC AdG.
Architecture of CSN-CRL ligase complexes (A), dimeric CRL3-CSN complexes (B), and mechanism of subst