Full-Length Structure and Heme Binding in the Transcriptional Regulator HcpR

Abstract

HcpR is a heme-dependent transcriptional regulator present in many Gram-negative anaerobic bacteria. In the perio-pathogen Porphyromonas gingivalis, HcpR is crucial for the response to reactive nitrogen species such as nitric oxide (NO). Binding of NO to the heme group of HcpR leads to transcription of redox enzyme Hcp. However, the molecular mechanisms of binding of heme to HcpR remain unknown. In this study, we present the 2.3 Å structure of the P. gingivalis HcpR. Interdomain interactions present in the structure help to form a hydrophobic pocket in the N-terminal sensing domain. A comparison analysis with other CRP/FNR-family members reveals that the molecular mechanisms of HcpR-mediated regulation may be distinct from those of other family members. Using docking studies, we identified a putative heme-binding site in the sensing domain pocket. In vivo complementation and mutagenesis studies verify one pocket residue, Met68, as an important reagent in HcpR-mediated transcriptional activation. Finally, heme-binding studies with purified forms of recombinant HcpR support Met68 as a crucial residue for heme binding as well as coordination.

1

Overview of the crystal structure of full-length HcpR. (A) Ribbon diagram of the HcpR shown in forward and 90° side views showing the orientation of the N-terminal sensing domain and C-terminal DNA-binding domain. Chain A is colored in blue, and Chain B is colored in green. The two chains form a homodimer. (B) Dimerization helices of HcpR. The homodimer is stabilized by hydrophobic interactions among symmetry-related Met, Leu, and Ile residues. Dashes indicate Van der Waals contacts between symmetry-related residues. (C) Highlight of the asymmetry of the DNA-binding domains of chains A and B. Helix E of chain A is broken at the hinge region at residue 152, where Helix E of chain B is extended through residue 170. (D) Superimposing chains A and B highlights the asymmetry of the DNA-binding domains. The DBD of chain A is rotated 166° with respect to chain B. (E) Inset of the boxed region in panel (D). The flap region of chain B is angled backward 39.6° to create a pocket in the SD of chain B.

2

Comparison and conformational differences between HcpR, CooA, and DNR. (A) Overlay of the sensing domains of HcpR chain A (green) and CooA chain A (purplePDB ID 1ft9) depicting the organization of each protein. The N-terminal domain of HcpR (extending from residues 1–127) is slightly larger than that of CooA (extending from residues 2–107). Inset left: The N-terminus of CooA is unstructured to allow the N-terminal Pro residue to coordinate heme. By comparison, the N-terminus of HcpR is well organized. Inset right: The extended loop region represented by residues 89–94 in HcpR occupies the heme-binding site of CooA. (B) View of the sensing domains of HcpR (green) and DNR (tanPDB ID 3dkw). The dimerization of DNR is at a 71° difference when compared to HcpR.

3

Heme docks into the hydrophobic pocket of HcpR. (A) Overview of heme docking in the HcpR pocket. Heme docks at the interface between chain B SD and the middle portion of helix E of chain A. (B) Side chains present in the putative heme pocket. Both Met68 and His149, residues commonly implicated in heme coordination, are optimally located in the pocket to coordinate the heme iron. (C) Side chains present in the heme pocket that form potential contacts with docked heme.

4

(A) Growth of ΔHcpRP. gingivalis complemented with WT and mutant HcpRs (the L156* mutant lacks the DNA-binding domain and serves as a negative control). Complemented strains were grown in 1 mM NO₂ or 2 mM NO₂ for 24 h in Tryptic Soy Broth. After incubation, the optical density at 600 nm was used to assess growth. (B) Complemented ΔHcpRP. gingivalis strains growing in mid-log phase were exposed to 0.2 mM NO₂ for 15 min and collected. The level of the hcp transcript was assessed via qRT-PCR to assess activity of HcpR mutants. Expression levels of hcp mRNA are normalized to the Pg16s rRNA subunit. Asterisks represent t-tests with p values of * <0.05, **<0.01, and ***<0.001.

5

(A) Recombinant tagless and reconstituted WT, M68A, H149A, and M68A/H149A HcpR at 0.25 mg/mLthe amber color of the purified protein indicates the presence of heme in the sample. Amber color in the M68A/H149A mutant is decreased after desalting. The heme occupancy after reconstitution: wildtype-63%; M68A-62%; H149A-60%; M68A/H149A-30% (B) Spectra of the reduced heme-bound WT-HcpR and the M68A/H149A double mutant HcpR at approximately 1 mg/mL. The flattening of the Soret peak in the M68A/H149A HcpR indicates loss of heme binding. (C) Spectra of the reconstituted WT HcpR and M68A HcpR under reduced, anaerobic conditions at 0.2 mg/mL. (D) Spectra of the reconstituted WT HcpR and H149A HcpR under reduced, anaerobic conditions at 0.25 mg/mL. (E) Spectra of the NO-bound WT, M68A, and H149A HcpR under reduced anaerobic conditions. (F) Diagram of the heme coordination before and after the addition of nitric oxide. HcpR is in a 6-coordinate system before the addition of nitric oxide (2-HcpR, 4-N heme). Evidence suggests that Met68 is one of the coordinating ligands. Nitric oxide cleaves both axial bonds from HcpR, and heme exists as a 5-coordinate system (1-NO, 4-N heme). The identity of the 6th axial ligand, possibly His149, is not immediately clear.

6

Structural alignment of full-length HcpR and ΔCHcpR highlights the interdomain interactions around the heme pocket. (A) Structural overlay of the full-length HcpR (green) and the C-terminal truncated ΔCHcpR (light purple). The structures overlay with an RMSD of 0.564 Å across the full length of the truncated HcpR (residues 1–156). (B) Zoomed-in view of the boxed region in (A) of the flap region in chain B of HcpR compared to that of chain B in ΔCHcpR. The flap region of HcpR has shifted out by 13.7 Å, enlarging the hydrophobic pocket. (C) Surface view of the boxed region in (A) of the pocket present in the full-length HcpR and the subsequent position and ΔCHcpR.