Figure 1.
Domain architecture of E. coli DNA gyrase and model of DNA gyrase mechanism.
(a) Domain architecture of E. coli DNA gyrase. GyrB is composed of an ATPase (yellow), a transducer (orange) and a toprim (red) domain. GyrA is composed of a winged-helix (WHD, blue), a tower (blue), a coiled-coil (light purple) and a β-pinwheel (β-PW, purple) domain. (b) A double-stranded DNA segment (G-segment) is captured at the DNA-gate, the N-gate is open to allow access of the T-segment. (c) Upon ATP binding, the ATPase domains dimerize and the T-segment gets trapped. (d) This is followed by ATP hydrolysis, G-segment cleavage, DNA-gate opening and T-segment translocation. The domain colored in light-grey represents the transducer domain in the preceding orientation. The mechanistic details of this step are not clear, in particular whether prior to Pi release the enzyme conformation is changed and how ATP hydrolysis and DNA passage are coordinated. For clarity, the coiled-coil and β-pinwheel domains of GyrA are omitted in (b-d). Adopted from references [3], [5].
Table 1.
Data collection statistics.
Table 2.
Refinement statistics.
Figure 2.
Conformational states of GyrB43 along the ATP hydrolysis reaction path.
The various states have been trapped by co-crystallization with the appropriate nucleotides. The dimeric structures are shown in cartoon representation with semi-transparent molecular surface overlaid. Subunits are distinguished by a slight variation in colour hue. View perpendicular to the molecular dyad. (a) AMPPNP complex (PDB entry 1EI1 [13]), (b) ADP⋅Pi complex, (c) ADP⋅BeF3 complex and (d) ADP complex. Note the distinct opening angles defined by the two transducer domains with the AMPPNP and ADP complexes in a "closed" conformation, the ADP⋅BeF3 complex in a "semi-open" and the ADP⋅Pi complex in "open" conformation.
Table 3.
Pair-wise fit of GyrB43 domains after superposition of the ATPase domains (regular) or transducer domains (italics).
Figure 3.
Structure comparison of (a) GyrB43⋅ADP⋅Pi (magenta) and (b) GyrB43⋅ADP⋅BeF3 (violet) with GyrB43⋅AMPPNP (shown as a dimer with yellow/grey colour with molecular dyad in black).
The structures are superimposed on their ATPase domain (residues 20–220). The rotation axes for the domain reorientation of the transducer domain with respect to the ATPase domain are indicated. These form an angle of 21.5° and 14.5° with the molecular dyad for GyrB⋅ADP⋅Pi and GyrB⋅ADP⋅BeF3, respectively. The insets show the QTK loop of the transducer domain that is rotated relative to the ATPase domain in response to the nucleotide state (dashed lines); Q: Q335, K: K337. (c) Stereoview of the overlay of the three aforementioned structures.
Figure 4.
Quantification of nucleotide induced structural changes within dimeric GyrB43.
(a) Inter-subunit distance changes between symmetry related residues of selected transducer domain Cα-atoms. The changes have been calculated relative to the AMPPNP complex structure, green bars indicate an increase in distance, red bar indicate a decrease. (b) Cartoon representation of dimeric GyrB43 (ADP⋅Pi state) with the residues used for the calculations in (a) indicated. View along the symmetry axis from the C-terminal side.
Figure 5.
GyrB43 crystal packing of the P21212 (left, GyrB43⋅AMPPNP complex) and the C2221 (right, GyrB43⋅ADP⋅Pi complex) form.
(a) The molecular packing is shown within a slab perpendicular to b and centered at y = 1/4. In both forms, dimers are oriented with their molecular dyads parallel to b (viewing direction). In the P21212 form (left) the molecular dyad is local (green elliptical symbol), in the C2221 form (right) the molecular dyad is crystallographic (black symbol). In each case, neighbouring dimers are related by horizontal 2-fold screw-axes. Black spheres represent the position of residue D377 at the end of the transducer domain (also depicted in Fig. 4b). (b) Details of the major crystal contact formed in both packings between the transducer domain and the ATPase domain of a symmetry related dimer. (c) Representation of the symmetry elements, same view as in panel (a). Unit cells are indicated by solid line. With respect to the arrangement of symmetry elements, the dashed rectangle of the scheme at the right is equivalent to the unit cell of the left scheme. Local symmetry elements are indicated in green.
Figure 6.
Structures of the GyrB43 nucleotide binding site as determined for (a) the substrate analog complex GyrB43⋅AMPPNP (PDB entry 1EI1 [13]) and (b) the post-hydrolysis complex GyrB43⋅ADP⋅Pi with Fo-Fc omit map shown at a contour level of 3.0 σ.
Note the distinct interaction of the QTK loop (transducer domain) with the 25–26 loop (ATPase domain) in the two states. The rotation axis for the relative domain reorientation is shown as straight line (same in Fig. 3a) (c) Stereoview of the superimposition of the structures shown in (a) and (b) after superposition on their ATPase domain. The exclamation mark indicates the steric clash that would occur between Q335 in the AMPPNP complex conformation (yellow) with the Pi moiety of the post-hydrolysis state (magenta).
Figure 7.
Surface representation of GyrB43 demonstrating the deeply buried nucleotide (stick model).
View along the narrow tunnel leading to nucleotide. (a) GyrB43⋅AMPPNP complex, (b) GyrB43⋅ADP⋅Pi complex. The insets show close-ups of the nucleotide sites with the ATP lid loop (residues 99–120) and the adjacent subunit of the dimer in surface representation (same colour codes as Fig. 2). Note, that upon ATP hydrolysis glutamine 335 cannot escape "downwards" to relieve a clash with the Pi moiety due to the presence of the ATP lid. Instead the entire transducer domain moves to the side.
Figure 8.
Refined mechanistic scheme of DNA gyrase activity, based on reference [2], representation as in Fig. 1.
The N-gate of the GyrA2GyrB2 heterotetramer with bound G-segment at the central DNA-gate (top) closes upon ATP binding thereby trapping a T-segment in the upper chamber (step a). Hydrolysis of the two ATP molecules causes a 12° rotation of the respective transducer domains relative to the ATPase domain (step b). We propose that this conformational change is coupled to DNA gate opening and T-segment translocation. Subsequent Pi release would be coordinated with G-segment re-ligation and DNA-gate closure (step c). Finally, ADP release results in dissociation of the ATPase domains and a reset of the enzyme (step d).