The proteasome is a key component of the ubiquitin-proteasome system (UPS), responsible for removing damaged or unneeded proteins and regulating major cellular processes1. Regulation of the proteasome is key to this function to ensure that only damaged & unnecessary proteins are degraded. Dysregulation of the proteasome has been implicated in several neurodegenerative diseases (NDs), often characterized by impairment of proteasome function2–7. In this study, we elucidate a mechanism of proteasomal activation and design a small molecule to functionally emulate the mechanism of proteasomal activation. Deriving from our findings on the mechanism of proteasomal activation, we report that our small molecule robustly regulates the activity of the archaeal, yeast, mammalian, and human proteasomes, and are even able to overcome the inhibition of the proteasome by protein aggregates implicated in neurodegenerative disease, such as amyloid-β and α-synuclein.
The Core Particle of the proteasome, also referred to as 20S (Fig. 1A), consists of four stacked heteroheptameric rings (α-β-β-α) with a central pore for substrate entry. The β rings consist of seven subunits (β1-7), three of which harbor the protease sites. The two α rings also consist of seven subunits (α1-7). Substrate entry in the 20S is regulated by the gate, which primarily consists of the N-terminus of α 2, 3, and 4 extending over the central pore8. The closed gate conformation blocks the central pore and prevents proteins or peptides from entering the 20S to be degraded. The N-terminus of each α subunit carries a YDR motif that interacts with neighboring N-termini to stabilize the closed state of the gate8. These N-termini extensions can also change their conformation to an “open” state, whereby they point up and outwards from the α ring pore, which is stabilized by an alternative interaction from the YDR motif9. Assessed by NMR, the basal kinetics of archaeal 20S gate-opening has been suggested to undergo open/close fluctuations on a time scale of seconds10,11. Although the kinetics have not yet been measured for the mammalian 20S, the presence of basal protein or peptide hydrolysis activities suggest that the mammalian 20S gate also fluctuates between these states, though likely slower. Additionally, truncation of α3 N-terminus (α3∆N), which act as a central lynchpin to stabilize the closed state, generates a constitutively open (active) 20S that is highly capable of degrading unstructured proteins8,12.
While the 20S gate is typically in the closed state, binding of proteasome regulatory complexes (Fig. 1A) to the α ring can trigger conformational changes that cause gate-opening, allowing the 20S to accept linearized substrates13,14. Two different mechanisms of 20S gate-opening have been described, the HbYX-dependent and the 11S family-dependent mechanisms. Thus far, the majority of studies on the HbYX-dependent mechanism focuses on the 19S, also known as PA700 or the Regulatory Particle (RP), which associates with the 20S to form the 26S complex that degrades ubiquitinated proteins (Fig. 1A). The 19S consists of a base subcomplex, which is primarily composed of a heterohexameric ring of ATPases (Rpt1-6), and a lid subcomplex, which contains ubiquitin binding and processing subunits. The 19S has been shown to interact with the 20S and stimulate gate-opening by the docking of the C-terminal tails of Rpt1-6, some of which contain the HbYX (hydrophobic-tyrosine-almost any C-terminal residue) motif, in the intersubunit pockets of the 20S α ring15. The use of C-terminal tails to associate with the 20S has also been observed in other Proteasome Activators (PAs) including PAN (proteasome-activating nucleotidase; the archaeal homolog of the 19S), PA200/Blm10 (Fig. 1A), and 11S activators.
Not all C-terminal tails of PAs induce gate-opening when bound to the 20S, as evident by the observation that peptides corresponding to the C-terminus of PA26 (a member of the 11S family) cannot induce gate-opening autonomously16. Conversely, peptides corresponding to the C-terminus of Rpt2, Rpt3, Rpt5, PAN, and PA200/Blm10 can induce autonomous gate-opening16,17. All these peptides that induce gate-opening carry the HbYX motif, which has been shown to be essential for allowing these complexes to associate with the 20S15,16. The C-terminal HbYX motif binds into pockets formed by the interface of the α subunits in the 20S, called intersubunit pockets (Fig. 1B&C). Whereas the C-termini of PAN homohexameric ATPases all have a HbYX motif, only the C-terminus of Rpt2, 3, & 5 from the 19S heteromeric ATPases have the HbYX motif and Rpt1 has a partial HbYX motif, lacking the Hb residue. The roles that the C-termini of Rpt4 and Rpt6 (which lack the HbYX motif) play in the association of the 19S-20S and 20S gating regulation are unclear but have been seen bound to intersubunit pockets via cryo-EM18,19. The binding of HbYX peptides to intersubunit pockets, structurally distant from the gating residues, results in gate conformational change. This demonstrates that the HbYX motif functions allosterically, and likely induce substantial conformational changes in the α subunits that in turn, affect the conformation of gating residues16,20.
In contrast, the family of 11S activators does not have a HbYX motif on their C-terminal tails. Their mechanism of gate-opening is relatively well-known compared to the HbYX-dependent mechanism. They associate with the 20S using their C-termini to dock in the α intersubunit pockets, similar to the HbYX-dependent activators. However, to trigger gate-opening, the 11S family rely on “activation loops” that interface directly with the base of the gating N-termini in the pore of the α ring21,22. These activation loops appear to sterically repel a reverse turn proline (Pro17) at the base of the gating residues, which destabilize the closed state and stabilize the open state. Interestingly, relative to the HbYX-mechanism, minimal conformational changes in the α subunits (excluding gating regions) are necessary for gate-opening by the 11S activators, as shown by the crystal structure of PA26-20S proteasome9. It is evident that the two families of PAs (HbYX-dependent and HbYX-independent) use different strategies to induce 20S gate-opening. Although the location and effect of HbYX-binding has been investigated, the molecular mechanism of HbYX-dependent gate opening appears to be surprisingly complex and remains unsolved.
Structures of the substrate-engaged human 26S (H26S) from Dong et. al 18 suggest that the human 19S (H19S) associates with the human 20S (H20S) through various interactions between the 19S ATPase’s C-termini and the 20S α-ring. These interactions vary in that they change based on the state of the 26S. As the H26S transition towards a more active state (EA1,2 > EB > EC1,2 > ED1,2)18, more C-termini form stable interactions (as observed via cryo-EM), starting with Rpt3, Rpt5 and Rpt2, then Rpt6, and finally Rpt1. The first tails to dock (Rpt3, 5, & 2) all carry the HbYX motif, yet, as visualized by cryo-EM the gate does not appear open. When the last C-terminus of the ATPases binds (Rpt1, which has a partial HbYX motif), a conformational change occurs, resulting in a stably opened gate. Interestingly, another structural study of Saccharomyces cerevisiae 26S (Y26S)23 suggests the same pattern of C-terminal tail binding for gate opening, while other studies on the Y26S19,24 suggest complete gate-opening occurs after the binding of Rpt6, and that the binding of Rpt1 in the α ring is not required for complete gate-opening. Additionally, some Y26S structures19,24 indicate that the binding of Rpt2, 3, & 5 alone is sufficient to induce partial gate-opening. Li et. al. previously demonstrated via a biochemical study that the mammalian 20S gate was “variably modulated” when complexed with the 19S, and could exist in various states of openness. While cryo-EM findings can approximate gate openness, the structures do not precisely reflect the dynamics of the gate within an individual proteasomal state (EA1,2, EB, EC1,2, ED1,2). For example, the un-activated 20S proteasome by itself still has the ability to degrade linearized proteins or peptides, but structurally, it is observed with a closed gate via cryo-EM or X-ray crystallography. Therefore, structural methods have a limited ability to draw conclusions regarding the functional openness of a gate.
The disparity between the HbYX-dependent mechanism of gate opening in the H26S and Y26S raises the question of how the HbYX-motif binding opens the gate. It is possible that HbYX-dependent gate opening requires occupation of a minimum number of intersubunit pockets to trigger an allosteric transition. Alternatively, it’s possible that only a subset of pocket(s) must be occupied by the corresponding C-termini to trigger gate-opening, or some combination of these two scenarios. We expect that since the HbYX motif is conserved from archaea to humans and is conserved in most PA’s (excluding the 11S family), the HbYX-dependent mechanism of gate-opening would also be conserved.
Recently, we showed that ND-associated proteins (i.e. amyloid-beta, α-synuclein, and huntingtin) can fold into a common conformation that inhibits the 20S and 26S proteasomes20. The inhibitory conformation is recognized by the A11 antibody, using a conformational epitope known to be pathological25. We also learned that these A11-positive (A11+) oligomers inhibit the 20S by allosterically stabilizing the closed gate conformation20. This negative allosteric regulation by the A11 + oligomers appear to be mechanistically coupled to the HbYX-dependent mechanism of 20S gate-opening20, suggesting that A11 + oligomers and the HbYX motif are allosteric regulators of the same gating mechanism. This leads us to hypothesize that stimulating 20S activity via the HbYX-dependent mechanism could not only antagonize impairment by A11 + oligomers and restore proteasome function, but it could also stimulate protein degradation, potentially providing a therapeutic approach for ND. Whereas the molecular interactions that stabilize the closed (i.e. 20S-α subunit YDR motif) and open state of the proteasome’s gate are clear8,26, as are the various 26S proteasome opened/closed states18,19,23,24, the molecular mechanisms that regulate the transition between these closed and open states are not understood. A clear understanding of the gate-opening mechanism in the 20S will provide the molecular framework to guide drug-discovery approaches aimed at activating proteasomal degradation to treat ND and is expected to elucidate how ND-associated oligomers impair proteasome function.