How do heme-protein sensors exclude oxygen? Lessons learned from cytochrome c', Nostoc puntiforme heme nitric oxide/oxygen-binding domain, and soluble guanylyl cyclase

Abstract

Significance: Ligand selectivity for dioxygen (O(2)), carbon monoxide (CO), and nitric oxide (NO) is critical for signal transduction and is tailored specifically for each heme-protein sensor. Key NO sensors, such as soluble guanylyl cyclase (sGC), specifically recognized low levels of NO and achieve a total O(2) exclusion. Several mechanisms have been proposed to explain the O(2) insensitivity, including lack of a hydrogen bond donor and negative electrostatic fields to selectively destabilize bound O(2), distal steric hindrance of all bound ligands to the heme iron, and restriction of in-plane movements of the iron atom.

Recent advances: Crystallographic structures of the gas sensors, Thermoanaerobacter tengcongensis heme-nitric oxide/oxygen-binding domain (Tt H-NOX(1)) or Nostoc puntiforme (Ns) H-NOX, and measurements of O(2) binding to site-specific mutants of Tt H-NOX and the truncated β subunit of sGC suggest the need for a H-bonding donor to facilitate O(2) binding.

Critical issues: However, the O(2) insensitivity of full length sGC with a site-specific replacement of isoleucine by a tyrosine on residue 145 and the very slow autooxidation of Ns H-NOX and cytochrome c' suggest that more complex mechanisms have evolved to exclude O(2) but retain high affinity NO binding. A combined graphical analysis of ligand binding data for libraries of heme sensors, globins, and model heme shows that the NO sensors dramatically inhibit the formation of six-coordinated NO, CO, and O(2) complexes by direct distal steric hindrance (cyt c'), proximal constraints of in-plane iron movement (sGC), or combinations of both following a sliding scale rule. High affinity NO binding in H-NOX proteins is achieved by multiple NO binding steps that produce a high affinity five-coordinate NO complex, a mechanism that also prevents NO dioxygenation.

Future directions: Knowledge advanced by further extensive test of this "sliding scale rule" hypothesis should be valuable in guiding novel designs for heme based sensors.

FIG. 1.

Structural differences between Thermoanaerobacter tengcongensis heme nitric oxide/oxygen-binding domain (Tt H-NOX; top) and Tt sensor of nitric oxide (SONO; bottom) in the heme binding pocket. Same sensor protein was isolated from Thermoanaerobacter tengcongensis and structure resolved by two different groups as Tt H-NOX (1U4H) and SONO (1XBN). General folding in the vicinity of the heme site is very similar between these two structures, but the W9 and N74 involved in H-bonding with Y140 in 1U4H are absent in 1XBN as the indole and amide nitrogen in W9 and N74 are moved away from the Y140 phenoxyl group. The distance between the phenol oxygen and W9 indole nitrogen is 2.7 and 5.9 Å in 1U4H and 1XBN, respectively; the distance between the phenol oxygen and N74 amide nitrogen in 1U4H is 2.9 Å and to the N74 amide oxygen in 1XBN is 3.8 Å. The dioxygen (O₂) ligand (yellow ball/stick) shows almost opposite orientations relative to the Y140 phenoxyl in these two structures although the distance between the phenol oxygen and the distal O atom of O₂ ligand are similar, 2.5 versus 2.7 Å, respectively. (To see this illustration in color the reader is referred to the web version of this article at www.liebertonline.com/ars).

FIG. 2.

Electronic spectra of α1/β1-sGC-I145Y. Spectra were recorded in the absence or presence of O₂ (∼1 mM), carbon monoxide (CO, ∼1 mM), and nitric oxide (NO, 50 μM).

FIG. 3.

CO-induced structural changes in the heme pocket of cytochrome c′ (cyt c′). CO-free (Top) and CO-bound (Bottom) cyt c′ structures reveal the major movement of the sterically hindered L16 and recruitment of P55 into 5 Å of the heme center upon CO binding. Bound CO is slightly tilted due to the proximity of L16. The view of the bottom structure is 90° rotated from the structural view of the top. All revealed residues are hydrophobic in nature. (To see this illustration in color the reader is referred to the web version of this article at www.liebertonline.com/ars).

FIG. 4.

The main residues predicted to be present in the distal heme pocket of soluble guanylyl cyclase (sGC). Six hydrophobic amino acid residues occupy the distal heme pocket modeled against Ns H-NOX crystallographic data to provide an environment where low polarity is expected [adapted from (81)]. (To see this illustration in color the reader is referred to the web version of this article at www.liebertonline.com/ars).

FIG. 5.

Key residues in the heme distal pocket of ligand-free Ns H-NOX (A), and its CO-bound (B), and NO-bound (C) forms. Top view of the ligand-free sensor and side views of the ligand-bound sensor are provided with the key Trp74 and the CO and NO ligands shown in space-filling model to show the influence of Trp74 in steric constraint on ligand binding geometry. Protein Data Bank (PDB) codes for the free, NO- and CO-bound Ns H-NOX are 2O09, 2O0C and 2O0G, respectively. (To see this illustration in color the reader is referred to the web version of this article at www.liebertonline.com/ars).

FIG. 6.

Sliding scale rule principle revealed by graphical analyses of K_D, versus ligand types. K_D values for cyt c′ (yellow squares), I145Y sGC (maroon reversed triangles), sGC (red circles), Ns H-NOX (blue circles), H64V Mb (black triangles), Fe(II)PP(1-MeIm) model heme (black circles), L16A cyt c′ (dark yellow squares), and H61L Lb (green reversed triangle) are plotted against NO, CO, and O₂ ligands on a logarithm scale to show the ∼8-order of magnitude dynamic range of the binding parameter values. Other than the unusually low value of K_D(NO) for sGC (red circles and red lines), the general relationship between the different heme proteins are parallel lines for all three ligands. The estimated K_D(NO), K_D(O₂) values, for sGC, cyt c′, Ns H-NOX based on the sliding scale rule are also highlighted (dashed lines). O₂ concentrations of air-saturated buffer and O₂-saturated buffer are indicted by horizontal blue dashes. (To see this illustration in color the reader is referred to the web version of this article at www.liebertonline.com/ars).

FIG. 7.

Sliding scale rule graphical analyses of k_on and k_off against ligand types and the ring orientation of the proximal His ligand. k_on (a, left) and k_off (a, right) values for the same set of hemeproteins shown in Figure 6 are plotted against NO, CO, and O₂ ligands in logarithm scale to show the V-shaped and reversed L-shaped sliding scale relationship for the k_on and k_off values. The estimated k_off(NO) and k_off(O₂), for cyt c′, I145Y sGC, sGC, and Ns H-NOX based on the sliding scale rule were also highlighted (red dashed lines). The ring orientations of the proximal His ligands, His93 and His 120, respectively, relative to the heme porphyrin macrocycle in H64V Mb and L16A cyt c′ are shown in b. This top down view of heme and proximal histidine ligand reveals the eclipsed and staggered conformations between the imidazole ring (red dashes) relative to the line connecting the nearest pair of two pyrrole nitrogen atoms (blue dashes). PDB codes are 2mgj (H64V Mb) and 2yl0 (L16A cyt c′). (To see this illustration in color the reader is referred to the web version of this article at www.liebertonline.com/ars).

FIG. 8.

Single- and multiple-binding interactions between sGC and CO and NO, respectively. Resting sGC, present as a 5c high-spin species (A), exhibits 1-step reversible binding with CO to form a 6c complex (E). For NO binding to sGC, in addition to a similar 6c complex (B), two 5c-NO complexes are observed (D* and D) at NO stoichiometry ≤1 and >1, respectively. The irreversible steps of NO binding (B→D* and C→D) resulted in an overall enhanced binding affinity. Bis-NO complex (C) formation is supported by the fact of [NO]-dependent 2nd step (B→C) and recent ligand binding study using ¹⁵NO/¹⁴NO and sequential freeze-quench electron paramagnetic resonance (EPR) measurements (132). The rate constants of each chemical step are from our kinetic measurements at 24°C. (To see this illustration in color the reader is referred to the web version of this article at www.liebertonline.com/ars).