The top view 20S proteasomes were detected with AFM either in open or closed conformation. (A) Left and center, the shape of 20S proteasome based on the yeast crystal structure (1ryp). From the left: side view with subunits marked with shades of grey and rings labeled, contour of the side view particle, contour of the side view section with the central channel, antechambers and catalytic chamber colored in grey, with active sites marked with stars. The gate area is marked. Center: top view of the α-ring (1ryp), contour of the α-ring with closed gate, contour of the α-ring with the gate open (marked as grey circle). Right: tilted view images of particles classified as closed (top) and open (bottom). (B) Classification of top view wild type (top), inhibitor treated (middle), and mutated (bottom) particles into closed and open conformers. From the left: selections of proteasome fields with 5 or 6 top view particles; diagrams presenting the classification of individual proteasomes; zoomed-in images of the particles with corresponding sections through the topmost part of their a rings; diagrams demonstrating how central sections in four directions were run through the zoomed-in images of particles and (below) which part of the a ring (shaded) is sectioned. A particle is classified as open when all four sections of its a ring are showing a dip. The gray-scale bar represents the height of the particles from the baseline (black) to the top (white). The same height scale applies to images in A and C. (C) Close-ups of individual top-view WT, ligand-treated, or mutated particles. Note the presence or absence of a hole or dip (a dark central circle) in the a rings. The WT proteasomes had either an open (with a hole) or closed (rounded apical part of the α-ring) entrance to the central channel. The treatment with a ligand shifted the conformational equilibrium toward more open (substrate) or nearly all open (0.5 µM ZL₃B(OH)₂) particles. The mutated proteasomes shown (β1ΔLS; MHY1377; Arendt and Hochstrasser, 1999) were always in the open-channel state.

The chemical state of the active center determines the conformation of the gate.
(A) A functional active site is essential to trigger the substrate-induced shift in the conformational equilibrium of 20S proteasomes. Proteasomes were isolated from yeast strains in which the indicated catalytic threonine (Thr1) was substituted with alanine. The mutants were specifically devoid of one of the following activities: PGPH (post-acidic) (β1T1A, MHY1157), T-L (trypsin like, β2T1A, MHY1073) or the ChT-L activity (chymotrypsin-like, β5ΔLS-T1A, MHY973) (Arendt and Hochstrasser, 1997). The proteasomes from the mutant and corresponding parent strains (MHY1156, MHY1066, MHY952) were imaged after addition of a buffer with solvent (DMSO; control) or after a separate addition of the substrate specific for each of the three active centers. Bars represent means ± SD calculated for n = 8 to 20 fields with 200 to 500 top-view particles. The differences in abundance of the conformers were statistically significant at p<0.001 between the control and working proteasomes, with an exception of the respective T1A proteasomes in the presence of a substrate specific for the affected site; in these cases the ratios in working and control particles were identical.
(B) Exposing proteasomes to bortezomib, an inhibitor that forms a tetrahedral transition state analogue with the ChT-L active center, induced the opening of the proteasome gate. Wild-type 20S proteasomes were imaged before and after treatment with the specified concentrations of the inhibitor. The percentage of open particles (bars with white circles) increased with the increasing concentration of the inhibitor, while the ChT-L activity (white bars) decreased. The degradation of the SucLLVY-MCA substrate (releaser of free aminomethylcoumarin; AMC) was determined spectrofluorometrically with the same sample of proteasomes, which were used for AFM imaging. Black bars represent means ± SD (n = 9 to 17 fields with 300 to 600 top-view particles).
(C) N^α-acetylation of catalytic subunits was accompanied by gate opening. The particles with an open gate were refractory to enhancement of their ChT-L activity by addition of 5-fold molar excess of human PA28 activator. The following 20S proteasomes were analyzed: a – β1ΔLS (MHY1377); b - β1ΔLSnat1Δ (MHY1373); c - β1ΔLS-T1A (MHY2267); d - β1ΔLS-T1A nat1Δ (MHY2271); e - nat1Δ (MHY1372). Bars patterned with black circles represent the percentage of closed particles without ligand addition, with means ± SD (n = 10 to 20 fields with 150 to 350 top-view proteasomes). Bars with a diagonal pattern represent the percent of ChT-L activity change upon addition of PA28 (0% change for unliganded CP). The activities of unliganded samples expressed in nanomoles of released AMC product per mg of protein per second were: WT – 0.37, a – 0.45, b – 0.33, c – 0.41, d – 0.35, e – 0.43.

Catalytic mechanism of the proteasome.
The four major steps of catalytic cycle are shown: (1) the nucleophilic attack by the Thr1 hydroxyl on the carbonyl of the peptide bond, (2) a rearrangement of a tetrahedral intermediate and release of the first product, (3) the second nucleophilic attack by a water molecule leading to a formation of a second tetrahedral complex, (4) a rearrangement of the complex and release of the final product (Groll et al., 2006; Groll et al., 1997; Osmulski et al., 2008).

Engagement of the catalytic Thr1 α amine and the open conformation of the gate are correlated. The shift toward open gate is noted when the a amine is blocked (secondary) or ionized.
Left column: schematic representation of catalytic Thr1 under distinct liganding conditions.
Middle column: the ligand used.
Right column: the prevailing conformation of the gate computed from AFM images.