A. Group II chaperonins (herein Cpn) cycle between an open, substrate-accepting state, and an ATP-induced closed state. In each cycle the substrate (in blue) is released in either the folded or unfolded state. Unfolded substrate rebinds Cpn for iterative rounds of folding. Incubation with ATP and AlFx interrupts iterative cycling by locking Cpn in a closed state that encapsulates the substrate. In the absence of ATP, Proteinase K (PK; scissors) digestion interrupts iterative cycling by specifically digesting the substrate (Fig. 1B) and the open lid segments in Cpn.
B. Proteinase K sensitivity of open and closed Cpn states. PK leads to full digestion of the open Cpn lids (Coomassie stain, top panel) and the bound substrate, ³⁵S-Rhodanese (³⁵S-Rho, bottom panel; lane 2). ATP-induced cycling to the closed state protects both the Cpn lids and the substrate (lane 3). Incubation with ATP·AlFx locks the complex closed leading to complete PK protection of both lids and encapsulated ³⁵S-Rhodanese (lane 4). A purified complex of Cpn·Rhodanese at 0.25 μM was incubated in the presence or absence of 1 mM ATP and/or 1 mM AlFx for 10 min at 37 °C and digested with 20 μg/ml PK for 10 min at 25 °C.
C. Native gel analysis of Cpn-substrate complexes. Incubation with ATP·AlFx shifts the mobility of Cpn (top panel Coomassie blue stain), which carries the encapsulated substrate (bottom panel for autoradiography of ³⁵S-Rhodanese). Non-native rhodanese aggregates cannot migrate into the native gel (data not shown).
D. Folding under cycling conditions. ATP (5 mM) was added to initiate Cpn-mediated folding of Rhodanese, measured at the indicated time-points. Addition of PK at the times indicated immediately interrupts the folding reaction indicating that the Cpn is cycling between open and closed states during folding.
E. Folding under non-cycling conditions. Cpn mediated folding as in D. except that folding was initiated by addition of either ATP (cycling allowed); ATP·AlFx (no cycling allowed); AlFx (control) or ATP·AlFx and PK (no cycling allowed, no rebinding of released Rho). The folding yields and rates were identical for all conditions, indicating that cycling is not required for group II chaperonin mediated folding. See also Figure S1.

A. ATP binding pocket of group II Cpn from T. acidophilum (pdb ID 1A6E) highlighting Asp386, essential for ATP hydrolysis.
B. Rhodanese folding for Cpn-WT and Cpn-D386A. ATP hydrolysis is required to support rhodanese folding, data represented as mean +/− SEM, n=3.
C. Single particle Cryo-EM reconstructions of Cpn-WT and Cpn-D386A. Shown are side and top views of Cpn-WT without (left, gold) and with ATP (right, cyan) and Cpn-D386A with ATP (middle, blue).
D. Overlay of EM density maps for Cpn-WT –ATP and Cpn-D386A +ATP highlight the changes induced by ATP-binding.
E. Role of ATP and ATP ·AlFx on the ³⁵S-Rhodandese interaction with Cpn-WT and Cpn-D386A. Cpn complexes were analyzed on 4% native gels and the Rho-containing Cpn visualized by autoradiography.
F. PK digestion of Cpn-WT and Cpn-D386A complexes with ³⁵S-Rhodandese, analyzed by SDS-PAGE followed by Coomassie blue staining (top), and autoradiography (bottom). Cpn-D386A is incapable of closing (compare lanes 3 for WT with 6, 7 for D386A). See also Figure S2.

A. Proposed models for how closure affects substrate interactions with the central Cpn chamber. (i) The substrate remains bound in the closed state or (ii) the substrate is released into the central cavity. The closed Cpn-WT retains substrate in either model (left). Removal of the lid, yielding Cpn-Δlid, allows testing of these models. Cpn-Δlid will lose the substrate if closure weakens the interaction with the chaperonin (model (ii) right). A decrease in substrate affinity might also be revealed using a GroEL-derived trap (Frydman and Hartl, 1996). Cpn substrate binding sites shown as pink lines.
B. Effect of ATP binding and hydrolysis on the Cpn-Δlid-substrate interaction. The indicated Cpn-³⁵S-rhodanese complexes, incubated with or without 1mM ATP for 10 min at 37 °C were analyzed by native gel electrophoresis followed by autoradiography. The amount of ³⁵S-rhodanese that remains Cpn bound in each condition is indicated.
C. Effect of ATP binding and hydrolysis on release of non-native substrate from Cpn complexes. Autoradiography of native gel for reactions carried out as in B., but in the presence of equimolar GroEL-Trap which functions as a scavenger for released non-native proteins. A reaction where denatured rhodanese is added directly to the Trap is included as a control. The amount of Cpn-bound and Trap-bound rhodanese was calculated for each reaction from the native gel analysis.
D. Transition state mimic ATP·AlFx locks the Cpn-Δlid in the symmetrically closed state immediately halting the ATPase cycle [see Fig. S1 and (Zhang et al., 2010)].
E. Proteinase K digestion of ³⁵S-rhodanese complexes with Cpn-WT or Cpn-Δlid in the presence or absence of ATP·AlFx. The ³⁵S-rhodanese is completely digested in the closed Cpn-Δlid.
F. Native gel analysis of ³⁵S-rhodanese-chaperonin complexes incubated as in E. Cpn-Δlid + ATP·AlFx fully releases its substrate (top panel, ³⁵S-rho) even though both Cpns undergo the same conformational change with ATP·AlFx (bottom panel; Coom. Blue).
G. Fluorescence emission spectra of Nile Red-rhodanese (NR-Rho) in the presence and absence of Cpn-Δlid (Kim et al., 2005). Binding to the Cpn causes an increase in fluorescence intensity at 630 nm.
H. Time-dependent changes in the fluorescence intensity of NR-Rho emission at 630 nm. Red trace: NR-Rho-Cpn-Δlid complex in the absence of ATP. Addition of ATP (arrow) causes a decrease in fluorescence (blue trace).
I. Time-dependent changes in the fluorescence intensity of NR-Rho emission at 630 nm as in H. but arrow indicated addition of ATP·AlFx addition, which causes a qualitatively similar decrease in fluorescence intensity (cyan trace). See also Figure S3

A. Order of addition experiment to test availability of substrate binding sites in the open and closed Cpn states. Without nucleotide, both Cpn-WT and Cpn-Δlid are open and bind substrate (Ctrl). Substrate addition prior to incubation with ATP·AlFx allows the substrate to bind first before closure (S–>A); incubation with ATP·AlFx prior to substrate to addition examines if the closed state can bind substrate (A–>S).
B. Native gel analysis of the above incubations. Two Cpn binding substrates of different size were used. (i) Rhodanese (293 aa) and (ii) Peptide PepB (12 aa). The smaller peptide should access more readily the substrate binding sites. ³⁵S-rhodanese was detected by autoradiography; Alexa-488-PepB was detected by fluorescence scan.
C. Conclusions: Ctrl: Both substrates bind chaperonin in the open state. S –>A: When the lid is present (Cpn-WT) closure of the Cpn-substrate complex retains the substrate in the chamber; when the lid is absent (Cpn-Δlid) the substrate escapes the cavity. A–>S: The ATP·AlFx state blocks substrate binding to both Cpn-WT and Cpn-Δlid. Since Cpn-Δlid retains access to the inner chamber in the closed state, this result indicates that the substrate binding sites are hidden in the closed state.

A. Structures of group II chaperonins in the open and closed states highlighting Helix 11 the locus of substrate binding (pink). ATP-induced closing brings together adjacent apical domains, creating a tight interface between Helix 11 of one subunit and loop 327-331 of the neighboring subunit (green). The indicated tetra-alanine substitution in loop 327-331 (herein rls for release loop for substrate) was introduced in both Cpn and Cpn-Δlid (herein Cpn-rls and Cpn-rls-Δlid).
B–C. ATP-induces the compact closed state in Cpn-rls (B) and Cpn-rls-Δlid (C). Top view of each structure, obtained by single particle Cryo-EM reconstructions to 4–6 Å (see also Fig 6, Figs S4 and S5).
D. ATP fails to induce substrate release in Cpn-rls-Δlid. The indicated Cpn-substrate complexes were incubated in the presence and absence of ATP; substrate release assessed using native gel electrophoresis followed by Coomassie staining to visualize the Cpn(s) (top panel) and fluorescence scans to view substrate(s) (middle and bottom panels).
E. ATP·AlFx triggers substrate release in Cpn-rls-Δlid. Incubations and analysis as in (D), except that incubations were carried out in the presence and absence of ATP·AlFx.
F. Nucleotide-induced changes in the environment of NR-Rho bound to Cpn-Δlid (i) or Cpn-rls-Δlid (ii). Experiments performed as in Fig. 3H. Starting from a nucleotide free NR-Rho-Cpn complex (red trace), ATP was added to reaction at time indicated by an arrow (blue trace) and incubation continued. For Cpn-rls-Δlid no drop in fluorescence was observed upon ATP addition; after 5 min AlFx was added to the ATP reaction and incubation continued (cyan trace). See also Figure S4.

A. Use of Cpn-rls to test the role of substrate release in group II chaperonin folding. Incubation with ATP should lead to lid closure without substrate release, whereas addition of ATP·AlFx should release the substrate into the closed cavity.
B. Side views of single particle Cryo-EM reconstructions of ATP·AlFx induced state of Cpn-rls and Cpn-WT highlight the similarity of both closed structures (see also below and Fig. S5).
C–F. Comparative structural analysis of the ATP and ATP·AlFx states of Cpn-WT and Cpn-rls. (i) Top views of overlays for the electron density maps. (ii) Superimposition of apical domain region for a single subunit from the overlaid chaperonin models. Superimposition of structures obtained for Cpn-rls and Cpn-WT reveals that the ATP state of Cpn-WT (purple) is virtually identical to the ATP-AlFx states of both Cpn-WT (blue) and Cpn-rls (cyan). In contrast, ATP induces a different closed state in Cpn-rls (yellow); comparison with ATP·AlFx states reveals major differences in the position of helix 11 (red arrow) and the rls loop (blue arrow, residues 327-331).
G. Cpn-rls binds rhodanese efficiently and encapsulates the substrate upon ATP or ATP·AlFx induced closure. (i) Native gel analysis of ³⁵S-rhodanese bound to Cpn-rls in the presence or absence of ATP or ATP·AlFx. (ii) Proteinase K digestion of incubations from (i). Cpn-rls produces a protease-resistant lid in the presence of ATP or ATP·AlFx (top panel) that fully encapsulates the substrate (bottom panel for ³⁵S-rhodanese). Note similarity with Cpn-WT in Fig. 2E; 2F.
H. Rhodanese folding requires substrate release into the central chamber. Rhodanese complexes with Cpn-WT or Cpn-rls were incubated with the indicated nucleotides and folding was assessed as in Fig. 1, data are mean +/− SEM, n=3. See also Figure S5.

A. ATP regulation of the Cpn substrate cycle. In the absence of ATP chaperonins are open, exposing substrate binding sites (pink). Upon ATP binding, the lid remains open, and the substrate bound, but a subtle conformational change is observed. ATP hydrolysis has a dual function: close the lid and release the substrate by hiding the substrate binding sites. We hypothesize that lid closing may precede substrate release, transiently generating a closed but folding inactive state (in brackets). Substrate release into the closed chamber is required for folding, which occurs within the central chamber.
B. Top view of the chaperonin substrate cycle in (A) highlighting the mechanism of polypeptide release upon ATP hydrolysis. The open, ATP-free and ATP-bound, states expose the substrate binding region (pink). ATP-hydrolysis creates a lateral contact with the rls loop (green) that displaces the substrate into the central cavity.
C. Top views of open and closed crystal structures from Cpn (Pereira et al., 2010) highlighting the substrate binding region (pink) and rls loop (green).