The Central Coupler of ClpXP Controls Intersubunit Communication and Couples Intrasubunit Chemical Energy with Force Generation

Key results

1. ShownWith up to four Q208A subunits, in any arrangement tested, ClpXP translocates unfolded polypeptide at close to wild-type speed with an unchanged step of about 2 nm. Hexamers with one or no wild-type subunit grip the ssrA-tag but do not translocate.

   Mean instantaneous velocity changed only slightly vs WWWWWW across the mutant layouts (e.g. WWWWWW N = 59 traces, WWMMMM N = 47). Mean step stayed at about 2 nm, and longer dwells account for the velocity differences. WMMMMM and MMMMMM formed tethers but did not move even at ~1 pN, and degraded Titin^CM-ssrA poorly in bulk. Pause and backtracking probabilities rose from single to quadruple mutants. · A flexible central coupler does not affect…; The central coupler enables intersubunit…; Figure legends

2. ShownQ208A speeds up ATP hydrolysis without changing ATP binding, so more ATP is spent per step. The extra cost depends on both how many subunits are mutant and where they sit in the hexamer.

   kcat was similar for WWWWWW and WWWWWM, ~1.5-fold higher for WWWWMM, WWWMWM, WWWMMM and WWMMMM, and ~4-fold higher for WMWMWM. Km scaled linearly with kcat in most constructs, and mant-ATP binding curves were identical for WWWWWW and MMMMMM. ATP per step: ~2 (WWWWWW, WWWWWM), ~3 (WWWMWM, WWWWMM), ~4 (WWWMMM, WWMMMM), ~7 (WMWMWM). · The central coupler enables intersubunit…; Mant-ATP binding and SDS-PAGE degradation essays; Measurement of ATP hydrolysis rate

3. ShownFull force output needs wild-type central couplers in all six subunits. Hexamers with two or more Q208A subunits unfold GFP less often, even a single mutant unfolds it more slowly, and mutants tolerate less opposing load, while their unloaded speed changes little.

   GFP unfolding probability was ~75% for WWWWWW vs ~45% (WWWWMM, WWWMWM), ~25% (WWMMMM, WWWMMM) and 0 of 54 traces (WMWMWM). The pause before unfolding was ~9 s for WWWWWW vs ~15 s (WWWWWM) and ~21 s (WWWWMM, WWWMWM). F1/2 fell from ~19 pN (WWWWWW) to ~14 pN (WWWWWM) and ~9 pN (WWMMMM), while zero-force speed stayed near ~9 to ~11 nm/s. · The rigidity of the central coupler is essential…; A rigid central coupler determines the maximum…; Figure legends

4. ShownIn cryo-EM of the alternating WMWMWM mutant, Q208A leaves the substrate-free hexamer unchanged. With substrate bound, the Arg-finger helix, sensor and Walker motifs shift at the A/B, E/F and F/A interfaces, consistent with loss of a hydrogen-bond network anchored by Q208 that links neighbouring subunits.

   Class-I (3.7 Å, no substrate) matches wild-type PDB 8E91 (RMSD 0.83 Å). Class-III (3.4 Å, substrate-bound, ATP regeneration only) resembles PDB 6PP7 (RMSD 1.92 Å), but motifs at those interfaces move 2 to 5 Å. In 6PP7, Q208 of subunit B hydrogen-bonds its own Arg-finger helix, which contacts the Arginine Sensor of subunit A. · Cryo-EM structure of WMWMWM ClpXP complexes; Repositioning of Arg-finger, sensor-I, and…; Cryo-EM data processing; Model Building and Refinement

5. SuggestedA second substrate-bound class (Class-II), seen only with ATPγS, has ADP in the top subunit and its pore-1 loop extended toward the substrate. The authors read it as a pre-power-stroke state that supports the probabilistic anticlockwise long-step (PA/LS) model: one ATP resets the loop and a second drives the stroke.

   According to the authors, Class-II (3.7 Å in Results, 3.8 Å in Methods) resembles no published ClpXP structure. Its top-subunit Tyr153 sits 5.4 Å from the nearest substrate residue, vs 26.3 Å in Class-III and 10.9 Å in PDB 9C88. Its nucleotide pattern matches PDB 6PP7. The fit to PA/LS is argued by comparison with earlier PDB entries, not tested directly. · Cryo-EM structure of WMWMWM ClpXP complexes; A new ClpXP structure with an extended…; DISCUSSION; Cryo-EM data processing