Interactions between a subset of substrate side chains and AAA+ motor pore loops determine grip during protein unfolding

Abstract

Most AAA+ remodeling motors denature proteins by pulling on the peptide termini of folded substrates, but it is not well-understood how motors produce grip when resisting a folded domain. Here, at single amino-acid resolution, we identify the determinants of grip by measuring how substrate tail sequences alter the unfolding activity of the unfoldase-protease ClpXP. The seven amino acids abutting a stable substrate domain are key, with residues 2-6 forming a core that contributes most significantly to grip. ClpX grips large hydrophobic and aromatic side chains strongly and small, polar, or charged side chains weakly. Multiple side chains interact with pore loops synergistically to strengthen grip. In combination with recent structures, our results support a mechanism in which unfolding grip is primarily mediated by non-specific van der Waal's interactions between core side chains of the substrate tail and a subset of YVG loops at the top of the ClpX axial pore.

Keywords: AAA+ enzymes; ATP-dependent protein unfolding; ClpXP; E. coli; biochemistry; chemical biology; molecular biophysics; protein degradation; structural biology.

Figure 1.. Effects of cassette sequence on GFP unfolding and degradation.

(A) Starting at the N terminus, substrates contained residues 1–229 of A.victoria GFP (PDB 1GFL, Yang et al., 1996), a cassette with 12 variable residues, and a partial ssrA degron. (B) Method for measuring intracellular degradation of substrates by ClpX^ΔN/ClpP. (C) Cellular fluorescence depends upon ClpX^ΔN/ClpP expression and cassette sequence (listed in Table 1). (D) Fraction intracellular degradation for substrates bearing different cassettes. (E) Fits of the substrate dependence of degradation in vitro to a hyperbolic Michaelis-Menten equation. (F) V_max values for different substrates. In panels, C–F, values represent averages (± S.D.) of three biological replicates.

Figure 2.. A small subset of tail residues mediate grip during GFP unfolding.

(A) Fraction intracellular degradation for substrates with tails containing LYV tripeptides in otherwise all-glycine cassettes. Gly₁₂ and GA substrates were included as internal controls. (B) Fraction intracellular degradation for substrates with tails containing one tyrosine (Y) in otherwise all-glycine cassettes. Gly₁₂ and GA substrates were included as internal controls. (C) V_max values from Michaelis-Menten analysis of degradation of purified substrates with single-tyrosine cassettes. (D) Rates of ATP hydrolysis by ClpX^ΔN (0.1 μM hexamer) in the presence of ClpP (0.3 μM 14-mer) in the absence (–) or presence of different substrates (15 μM monomer). (E) ATP cost of degrading substrates with single-tyrosine cassettes. Note that the Y-axis is logarithmic. In all panels, values represent averages (± S.D.) of three biological replicates.

Figure 2—figure supplement 1.. Comparison of K_M values for substrates tested in vitro; comparison of fitted values for K_M for substrate degradation.

Values are the average of three biological replicates ± S.D. None of the substrates exhibited a substantial increase in K_M, indicating that differences in degradation rates result from differences in grip rather than in initial substrate recognition.

Figure 2—figure supplement 2.. Stimulation of ClpXP ATP hydrolysis by purified substrates.

(A) Rates of ATP hydrolysis by ClpX^ΔN (0.1 μM hexamer) in the presence of ClpP (0.3 μM 14-mer) in the absence (–) or presence of different substrates (15 μM monomer). (B) ATP cost of degrading substrates. In both panels, values represent averages (± S.D.) of three biological replicates.

Figure 3.. Side-chain grip effects at tail-position 4.

(A) In substrates with otherwise all-glycine cassettes, fraction intracellular degradation depends on side-chain identity at tail-position 4. (B) Comparison of degradation in vivo for substrates with Thr or Val at tail-position four or Glu or Gln at tail-position 4 (Student’s two-tailed t-test significance; Val/Thr: t = 6.37, df = 4; Glu/Gln: t = 5.47, df = 4). (C) V_max values from Michaelis-Menten analysis of degradation of purified substrates. (D) Effects of position-4 residues, color-coded by side-chain properties, on V_max. (E) Comparison of degradation in vitro between substrates with Ala, Ser, Cys, Thr, or Val at tail-position four or Glu or Gln at tail-position 4 (Student’s two-tailed t-test significance; Val/Thr: t = 13.3, df = 4; Glu/Gln: t = 5.49, df = 4). (F) ATP cost of degrading substrates with Ala, Cys, Thr, Val, Glu, or Gln at tail-position 4. With the exception of panel A, where Gly₁₂ and GA values represent averages (± S.D.) of nine biological replicates, all values represent three biological replicates.

Figure 4.. Multiple substrate residues contribute synergistically to grip.

(A) GA and Ala-4 cassette sequences. A heatmap of V_max values from Figure 2C is overlaid to show contribution of single tyrosine residues as each tail position. (B) Fraction intracellular degradation of substrates with one alanine at tail-position 4 and a second alanine at a variable position in otherwise all-glycine cassettes. (C) Comparison of intracellular degradation for a subset of substrates, including Ala-1. (D) V_max values from Michaelis-Menten analysis of degradation of purified substrates. (E and F) Michaelis-Menten V_max values for purified substrates with one tyrosine (E) or valine (F) at tail-position four and a second tyrosine (E) or valine (F) at each tail position in otherwise all-glycine cassettes. Overlaid dashed lines indicate degradation rate for the parental Tyr-4 (E) or Val-4 (F) substrates. In all panels, values represent averages (± S.D.) of three biological replicates.

Figure 4—figure supplement 1.. Degradation of Dual-Tyr substrates centered at tail position 3.

V_max values for degradation from Michaelis-Menten analysis of purified substrates with one tyrosine at tail-position three and a second tyrosine at a variable position in otherwise all-glycine cassettes. Relative degradation for substrate tails with a single Tyr residue at position 3 or 4 indicated by dashed lines. Values represent averages (± S.D.) of three biological replicates.

Figure 5.. Only a subset of pore-1 loops in ClpX appear to mediate substrate grip.

(A) Model of an extended poly-alanine substrate in the axial pore of ClpX and its interactions with different pore-1 loops based on cryo-EM structures of ClpXP (X.Fei, T.A. Bell, B.M. Stinson, S. Jenni, T.A. Baker, S.C. Harrison, and R.T. Sauer, in preparation). Similar loop-substrate interactions are observed in the yeast AAA+ protease Yme1 (Puchades et al., 2017). On the right, a heatmap of V_max values from Figure 2C is shown. The substrate tail residues are numbered relative to where a folded domain would be expected to sit at the apical surface of the AAA+ ring during unfolding. Tail residues 2–6, which promote strong grip in ClpX, are positioned to interact with the three pore-1 loops at the top of the axial pore. (B) Two models for asymmetric contribution of pore-1 loops to substrate grip.