Figures and data
Structural position and conservation of two acidic residues in the GHKL ATPase family.
(A) Crystal structure of the aqMutL N-terminal domain (NTD) in complex with the non-hydrolyzable ATP analog AMPPCP. The boxed region indicates the ATP-binding site and is shown enlarged on the right. AMPPCP and the side chains of Glu29 and Glu32 are shown as stick models. Oxygen atoms are shown in red, nitrogen atoms in blue, and phosphorus atoms in orange. The magnesium ion and the water molecule putatively acting as the nucleophile are depicted as green and red spheres, respectively. Hydrogen bonds are indicated by dashed lines, with distances shown. The blue mesh represents an omit Fo–Fc electron density map contoured at 3σ for AMPPNP, the magnesium ion, and the water molecule. (B) Sequence alignment of representative GHKL ATPases. The Bergerat ATP-binding fold is underlined, and the conserved acidic residues corresponding to aqMutL Glu29 and Glu32 are marked with asterisks. Amino acid sequence alignment was performed by CLUSTAL W program 57 and visualized by ESPript3 58.
Effects of mutations on the overall structure and the ATPase activity of the aqMutL NTD.
(A) Far-UV CD spectra of the wildtype and mutant forms of the aqMutL NTD were collected and averaged over 10 consecutive scans. (B, C) ATPase activity of the aqMutL NTD was measured under steady-state conditions using a colorimetric assay that detects the release of inorganic phosphate. Apparent rate constants (kapp) were calculated from the measured phosphate concentrations and plotted as a function of ATP concentration. Kinetic parameters were determined by fitting Michaelis–Menten equation to the data. Each data point represents the mean of three independent experiments, with error bars indicating standard deviation. Theoretical curves calculated from the fitted kinetic parameters are overlaid. Data for 0–1 mM substrate concentrations are shown to compare the wildtype from with mutant forms (B). Activities for higher substrate concentrations were measured for some of the mutant forms (C).
Kinetic parameters for the ATPase activity of the aqMutL NTDs, aqGyrB NTDs, ProS2-tagged human PMS2 NTDs, and histidine-tagged human MLH1 NTDs.
Structure of the AMPPNP-bound dimer of the wildtype aqGyrB NTD.
The boxed region highlights the ATP-binding site and is enlarged in the left panel. The same region is further enlarged and rotated 60° to provide an alternative view of the active site geometry. AMPPNP and the side chains of Glu48 and Asp51 are shown as stick models. Oxygen atoms are shown in red, nitrogen atoms in blue, and phosphorus atoms in orange. The magnesium ion and the water molecule presumed to act as the nucleophile are depicted as green and red spheres, respectively. Hydrogen bonds are indicated by dashed lines, with distances shown. The blue mesh represents an omit Fo–Fc electron density map contoured at 3σ for AMPPNP, the magnesium ion, and the water molecule.
Effects of mutations on the overall structure and the ATPase activity of the aqGyrB NTD.
(A) Far-UV CD spectra of the wildtype and mutant forms of the aqGyrB NTD were recorded and averaged over 10 accumulations. (B) ATPase activity of the aqGyrB NTD was measured under steady-state conditions as described in Fig. 2. The kapp values were plotted as a function of ATP concentration. Michaelis–Menten equation was fitted to the data. Data points represent the mean ± standard deviations from three independent experiments.
Structural basis for the loss and preservation of ATP binding in aqGyrB NTD mutant forms.
The wildtype and all mutant forms were crystallized in the presence of AMPPNP; however, the E48A mutant form did not bind AMPPNP, whereas the E48Q, D51A, and D51N mutant forms were observed in complex with AMPPNP. (A) Overall structures of the wildtype and E48A mutant aqGyrB NTD monomers. The ATPase active site is indicated by a boxed region. The ATP-lid (residues 105–124) in the wildtype form and the corresponding region in the E48A mutant form are shown in magenta and olive, respectively. (B–F) Enlarged views of the ATPase active site in the wildtype (B), E48A (C), E48Q (D), D51A (E), and D51N (F) forms. Side chains of the residues implicated in the ATP-induced conformational changes (e.g. His122, Gln340, and Lys342), as well as residues 48 and 51, are shown as stick models. AMPPNP is depicted as sticks, and the Mg2+ ion is shown as a green sphere. Hydrogen bonds and ionic interactions are indicated by dashed lines.
Structures surrounding the conserved acidic residues: (A) E48A, (B) E48Q, (C) D51A, and (D) D51N mutant forms of the aqGyrB NTD.
AMPPNP and the side chains at the positions 48 and 51 are shown as stick models. Oxygen atoms are shown in red, nitrogen atoms in blue, and phosphorus atoms in orange. The magnesium ion is shown as a green sphere, and the water molecule proposed to act as the nucleophilic water is shown as a red sphere. Blue mesh represents Fo–Fc omit electron density maps contoured at 3σ, calculated after omission of AMPPNP, the magnesium ion, and the nucleophilic water molecule. Hydrogen bonds are indicated by dashed lines, with the corresponding distances shown.
ATPase activities of human MutL homolog NTDs.
(A) ATPase activity of mutant forms of the human PMS2 NTD. (B) ATPase activity of a mutant form of the human MLH1 NTD. ATPase activities were measured using the same procedures as described for Fig. 2 and Fig. 4. Apparent rate constants (kapp) were plotted as a function of substrate concentration. Data points represent the mean values from three independent measurements, with error bars indicating standard deviations. Solid lines represent the theoretical Michaelis-Menten curves.
Maximum-likelihood phylogeny of the ATPase domain of representative GHKL family members.
The state of the amino acid residue at the position corresponding to aqMutL Glu32 or aqGyrB Asp51 is indicated by colored bars (red, acidic residue; blue, non-acidic residue). The phylogenetic tree was visualized and annotated using the Interactive Tree of Life (iTOL) web server 56. The tree is displayed in a rectangular layout for clarity and should be interpreted as unrooted. Ancestral state reconstruction was performed on the fixed topology, and posterior probabilities are shown for selected internal nodes. The internal node uniting the MutL/GyrB/MORC/Hsp90 clades strongly supports an acidic residue with the posterior probability (PP) of 0.983, whereas the Hsp90 stem ancestor is strongly supported to encode serine at this position with the posterior probability of 1.000. Species abbreviations used in the phylogenetic tree are as follows: Ce, Caenorhabditis elegans; Sc, Saccharomyces cerevisiae; Dr, Danio rerio; Hs, Homo sapiens; Aa, Aquifex aeolicus; Hp, Helicobacter pylori; Tt, Thermus thermophilus; Ec, Escherichia coli; Mt, Mycobacterium tuberculosis; Bs, Bacillus subtilis; Sm, Streptococcus mutans; Dm, Drosophila melanogaster; Clf, Canis lupus familiaris; Mm, Mus musculus; Hc, Hemicordylus capensis; Pv, Patella vulgata; Tf, Tachysurus fulvidraco; Lk, Lepidochelys kempii; Cc, Castor canadensis; Dn, Dasypus novemcinctus.
The ATP-binding ability of the aqMutL NTDs (A) and aqGyrB NTDs (B).
Binding affinities were quantified by equilibrium dialysis. For each protein concentration, the concentration of unbound AMPPNP was determined from the absorbance at 260 nm of the buffer chamber solution after equilibrium had been reached (see Materials and Methods). The concentration of bound AMPPNP was calculated by subtracting the concentration of unbound AMPPNP from the total AMPPNP concentration and was plotted against the protein concentration.
Structure of the ATPase active site of the E. coli GyrB NTD.
Residues corresponding to Glu48, Asp51, His122, Gln340, and Lys342 of aqGyrB are Glu42, Asp45, His116, Gln335, and Lys337, respectively, in E. coli GyrB. Side chains of these residues are shown as stick models together with the bound AMPPNP. Hydrogen bonds and ionic interactions are depicted as dashed lines.
Primers used for site-directed mutagenesis in this study
