The mycobacterial proteasomal ATPase Mpa forms a gapped ring to engage the 20S proteasome

Although many bacterial species do not possess proteasome systems, the actinobacteria, including the human pathogen Mycobacterium tuberculosis, use proteasome systems for targeted protein removal. Previous structural analyses of the mycobacterial proteasome ATPase Mpa revealed a general structural conservation with the archaeal proteasome-activating nucleotidase and eukaryotic proteasomal Rpt1–6 ATPases, such as the N-terminal coiled-coil domain, oligosaccharide-/oligonucleotide-binding domain, and ATPase domain. However, Mpa has a unique β-grasp domain that in the ADP-bound crystal structure appears to interfere with the docking to the 20S proteasome core particle (CP). Thus, it is unclear how Mpa binds to proteasome CPs. In this report, we show by cryo-EM that the Mpa hexamer in the presence of a degradation substrate and ATP forms a gapped ring, with two of its six ATPase domains being highly flexible. We found that the linkers between the oligonucleotide-binding and ATPase domains undergo conformational changes that are important for function, revealing a previously unappreciated role of the linker region in ATP hydrolysis–driven protein unfolding. We propose that this gapped ring configuration is an intermediate state that helps rearrange its β-grasp domains and activating C termini to facilitate engagement with proteasome CPs. This work provides new insights into the crucial process of how an ATPase interacts with a bacterial proteasome protease.