Evidence for roles of the Escherichia coli Hda protein beyond regulatory inactivation of DnaA

Abstract

The ATP-bound form of the Escherichia coli DnaA protein binds 'DnaA boxes' present in the origin of replication (oriC) and operator sites of several genes, including dnaA, to co-ordinate their transcription with initiation of replication. The Hda protein, together with the β sliding clamp, stimulates the ATPase activity of DnaA via a process termed regulatory inactivation of DnaA (RIDA), to regulate the activity of DnaA in DNA replication. Here, we used the mutant dnaN159 strain, which expresses the β159 clamp protein, to gain insight into how the actions of Hda are co-ordinated with replication. Elevated expression of Hda impeded growth of the dnaN159 strain in a Pol II- and Pol IV-dependent manner, suggesting a role for Hda managing the actions of these Pols. In a wild-type strain, elevated levels of Hda conferred sensitivity to nitrofurazone, and suppressed the frequency of -1 frameshift mutations characteristic of Pol IV, while loss of hda conferred cold sensitivity. Using the dnaN159 strain, we identified 24 novel hda alleles, four of which supported E. coli viability despite their RIDA defect. Taken together, these findings suggest that although one or more Hda functions are essential for cell viability, RIDA may be dispensable.

Figure 1. Structural model for the Hda-β clamp interaction

(A) Amino acid alignment of the clamp binding motifs (CBM) of E. coli Pol IV (EcPol IV) and E. coli Hda (EcHda). The Pol IV^LF domain (residues 243-351) bears the CBM at its C-terminus, whereas the Hda CBM resides at its N-terminus. This alignment provides a reference for residue alignment in the structural model, and defines the directionality of the partner proteins relative to the cleft of the β clamp. The E. coli Hda sequence was threaded through the crystal structure of a single molecule of the Shewanella amazonensis Hda (SaHda) protein (PDB ID 3BOS, chain B) (Xu et al., 2009) using the SWISS-MODEL Workspace (Schwede et al., 2003). The resulting model for the EcHda monomer is shown (B), with the CBM colored red. (C) Crystal structure of the Pol IV^LF domain (pink) in complex with the β clamp homodimer (yellow and orange) (PDB ID 1UNN) (Bunting et al., 2003). (D) Overlay of the EcHda monomer structure (blue) on the β clamp (yellow and orange), using the Pol IV^LF (pink) CBM as an alignment guide. Using the pair-fitting function within PyMOL 1.5.0.2 software, α-carbon residues 5-9 (LQLSL) of EcHda were aligned to the α-carbon residues 346-351 of Pol IV^LF (LVLGL), theoretically representing a similar interaction of the CBM for both partner proteins. EcHda clearly extends outward and away from where the Pol IV^LF resides on the β clamp. (E) The modeled structure of EcHda (blue) in complex with the β clamp (yellow and orange) without Pol IV^LF is shown. Structural images were generated using PyMOL v1.5.0.2 software (Schrödinger, 2010).

Figure 2. Overexpression of Hda sensitizes E. coli to NFZ

Sensitivity of isogenic E. coli strains bearing pWSK29 (control plasmid) or pJCB200 (hda⁺) to the indicated concentrations of NFZ was measured as described in Experimental Procedures. Strains were either impaired for polB (Pol II), dinB (Pol IV), or polB and dinB function (A), or impaired for nucleotide excision repair (uvrB) (B), as indicated. (C) Sensitivity to UV irradiation (50 J•m⁻²) of the polB⁺ dinB⁺ uvrB⁺ parent strain bearing pWSK29 or pJCB200 was measured as described in Experimental Procedures. Representative results for each are shown.

Figure 3. Overexpression of Hda suppresses −1 frameshift mutations

The frequency of lacZ⁻′Lac⁺ reversion of strain CC108 bearing pWSK29 (control plasmid) or pJCB200 (hda⁺) are shown. Results represent the average of 12 independent determinations for each strain. Error bars represent the 95% confidence limits.

Figure 4. The Δhda E. coli strain displays a cold sensitive growth phenotype

The Δhda::cat allele was transduced into strain MG1655 bearing either pWSK29 (control plasmid) or pJCB200 (hda⁺) using P1vir bacteriophage. Representative colonies were selected, grown overnight at 37 °C, serially diluted, spotted onto LB plates, and grown overnight at 30°, 37°, or 42 °C, as indicated. Strain JCB100 (Δhda) bearing pWSK29 (control plasmid) was severely impaired for growth at 30 °C, and displayed a mix of large and small colonies at 37° and 42 °C. Three small (isolates 2–4) and one large (isolate 5) CFU for strain JCB100 bearing pWSK29 were selected, grown overnight at 37° C, serially diluted, spotted onto LB plates, and grown overnight at 30° or 42 °C, as indicated. * indicates distinct isolates from a fresh transduction of Δhda::cat into MG1655 pWSK29.

Figure 5. Positions of mutations identified in Hda

(A) The primary structure of Hda depicting positions of the CBM (green), Walker A and B motifs (blue), Box VI and VII (light blue), sensors 1 and 2 (black line), and the conserved arginine finger (black line) is shown. Residues of Hda that form alpha (α) and 3₁₀ (η) helices or beta strands (β), based on the SaHda structure (PDB ID 3BOS), are indicated. Positions of mutants identified in this study are shown. Mutations are colored according to mutant class: class I is in green, class II is in orange, class III is in red, and class IV is in purple. The new reading frame generated by the D198fs mutant is shown. Positions of each mutation are also represented on the EcHda structural model from Figure 1B. Each panel (B–E) represents a single class of Hda mutations: (B) class I mutations are depicted as green spheres; (C) class II mutations are depicted as orange spheres; (D) class III mutations are depicted as red spheres; and (E) class IV mutations are depicted as purple spheres, while the C-terminal region affected by the D198fs is colored in pink. Residues corresponding to the CBM (blue) or the arginine finger (red) are depicted in stick form for reference. Structural images were generated using PyMOL v1.5.0.2 software (Schrödinger, 2010).

Figure 6. Cell cycle analysis of strains expressing mutant hda alleles

Flow cytometry data was collected with 50,000 cells of the indicated strain labeled with PicoGreen. Fluorescence intensity (abscissa) is presented in logarithmic scale. (A) Chromosome equivalents were determined using strain MG1655. Isogenic strains bearing Δhda, Δdat, ΔseqA, and dnaN (JCB103) or dnaN159 (JCB104) alleles were also analyzed. Representative results for mutants of class I (B), class II (C), class III (D), and class IV (E) are shown.

Figure 7. Mutant hda alleles are unable to confer NFZ sensitivity

E. coli strain MS100 bearing the indicated mutant allele of hda were serially diluted in 0.8% NaCl and spotted on LB plates containing NFZ as described in Experimental Procedures.

Figure 8. hda lethality and NFZ sensitivity of dnaN mutants

Positions P20, G66, and G174 of the β clamp, which bear substitutions in the dnaN159 (G66E, G174A), dnaN780 (G66E), dnaN781 (G174A), dnaN782 (G66A, G174A), or dnaN783 (P20L, G66E, G174A) mutations, are depicted in red on the structural model for the EcHda-clamp complex. Colors are as in Figure 1. This structural representation was generated using PyMOL v1.5.0.2 software. (B) Cultures of E. coli strain MS100 and derivatives bearing the indicated dnaN allele were serially diluted in 0.8% NaCl and spotted on LB plates containing the indicated concentrations of NFZ. Plates were incubated at 30 °C for 16 hours and chilled at 4 °C for 1 hour prior to imaging. (C) Transformation efficiencies of strains bearing the indicated dnaN allele using pWSK29 (control plasmid) or pJCB200 (hda⁺) are shown.