Intracellular proteolysis, a process that enables proteome remodeling, protein quality control, regulation of cellular processes and amino acid recycling under starvation conditions, is fundamental for all cells.1, 2 Regulated intracellular proteolysis is typically performed by proteases belonging to the AAA+ (ATPases associated with diverse cellular activities) family of proteins.1, 3, 4 Given how proteolysis is a destructive and irreversible process, AAA+ proteases evolved for the compartmentalized and selective protein degradation they perform, forming barrel-shaped structures that prevent the random interaction of their proteolytic active sites with cellular proteins. Specifically, a proteolytic subunit in AAA+ proteases provides a secluded chamber in which substrate peptide bonds are cleaved. To enter the proteolytic chamber, a substrate must first be recognized by the ring-shaped AAA+ regulatory sub-complex.5 An initial binding step often tethers the substrate to the regulatory sub-complex via interaction with a substrate-bound moiety (i.e., a degradation tag). This tethering step typically determines the specificity of the AAA+ protease for its substrates and allows for further substrate processing.5 Tethering thus increases the local substrate concentration near the protease central pore, thereby facilitating substrate engagement by the AAA+ domain.6 Following such engagement, repeated ATP hydrolysis cycles fuel substrate unfolding and translocation through the central pore into the proteolytic chamber, where substrate degradation ensues.

A single class of AAA+ proteases, namely, the 26S proteasome, is found in the cytoplasm and nucleus of eukaryotic cells, whereas a number of simpler AAA+ proteases, including Lon, FtsH, ClpXP, ClpCP, ClpAP and HslUV, are found in prokaryotes, mitochondria and chloroplasts.1, 2 In addition, archaea and bacterial species belonging to the phyla Actinobacteria and Nitrospira contain proteasomes that are simpler than the eukaryal 26S proteasome.7, 8 In the bacterial proteasomes, the regulatory sub-complex, termed regulatory particle (RP) in proteasomes, comprises a ring-shaped homo-hexamer termed Mpa in mycobacteria and Arc in other actinobacterial species.9, 10, 11 At the same time, the proteolytic sub-complex (core particle (CP)) of bacterial proteasomes, although highly similar in structure and composition to its eukaryotic counterpart, contains only one or two types of alpha and beta subunits, in contrast to the seven different alpha and seven different beta subunits comprising the eukaryotic complex.10 To be degraded by the bacterial proteasome, substrates must first be covalently modified by a protein tag termed Pup (prokaryotic ubiquitin-like protein), a small and natively unfolded protein.12 A single ligase termed PafA conjugates a glutamate side chain at the Pup C-terminus to substrate lysine side chains, thereby tagging a substrate for degradation. Complete understanding of the molecular determinants that dictate target pupylation is still lacking, despite recent studies showing that target proteins are recognized by PafA based on electrostatic interactions.13, 14 Together, Pup and the proteasome form the focal point of the Pup-proteasome system (PPS), a protein tagging and degradation system that was found to be important for Mycobacterium tuberculosis virulence and for survival of Mycobacterium smegmatis, a non-pathogenic species, under nitrogen starvation conditions.15, 16, 17

Pupylated proteins are degraded in a step-wise manner that starts with the binding of Pup by an apical coiled-coil domain of Mpa (Figure 1A, step (i)). This initial tethering stabilizes a helical conformation of Pup residues 21–51, allowing for specific recognition of pupylated proteins by the bacterial proteasome18 This step also facilitates entry of the Pup N-terminal tail into the Mpa pore, leading to ATP-dependent engagement of the tail by the AAA+ domain (Figure 1A, step (ii)). The engagement step allows the AAA+ domain to exert a pulling force on Pup, resulting in Pup dissociation from the coiled-coil domain as it is translocated into the 20S CP (Figure 1A, step (iii)). As Pup traverses the Mpa pore en route into the AAA+ domain, it passes through an inter-domain consisting of two rings of sub-domains that present the conserved oligomer-binding (OB) fold.19, 20 Such OB rings are conserved in all proteasomes,20 suggesting an important role for this domain in substrate processing. In Mpa/Arc, the channel formed by the inter-domain rings is lined with charged loops that could interact with the Pup-tagged substrate. Specifically, loops presenting negatively charged residues gate the entrance into and the exit from the inter-domain, while the core of the channel presents two rings of positively charged loops (Figure 1B). Although it was hypothesized that the Mpa inter-domain loops play a role in Pup binding,21, 22, 23, 24 and despite a recent study showing a structure of Mpa engaged with Pup,25 the contributions of the inter-domain to substrate processing and degradation by the proteasome remained unclear.

Here, we report that the inter-domain does not simply serve as an inert channel that merely mediates substrate passage, but rather acts as a gatekeeper that allows Pup to reach the AAA+ domain for downstream processing. Specifically, we found that the charged inter-domain pore loops are suited for Pup binding, thus facilitating subsequent Pup engagement by the Mpa AAA+ domain. We further observed intramolecular communication between the Mpa inter-domain and the adjacent AAA+ domain. Ultimately, these findings reveal another level of regulation of substrate binding and selectivity in the complex, multi-step mechanism employed by the bacterial proteasome.