Nitrosylation mechanisms of Mycobacterium tuberculosis and Campylobacter jejuni truncated hemoglobins N, O, and P.

Truncated hemoglobins (trHbs) are widely distributed in bacteria and plants and have been found in some unicellular eukaryotes. Phylogenetic analysis based on protein sequences shows that trHbs branch into three groups, designated N (or I), O (or II), and P (or III). Most trHbs are involved in the O2/NO chemistry and/or oxidation/reduction function, permitting the survival of the microorganism in the host. Here, a detailed comparative analysis of kinetics and/or thermodynamics of (i) ferrous Mycobacterium tuberculosis trHbs N and O (Mt-trHbN and Mt-trHbO, respectively), and Campylobacter jejuni trHb (Cj-trHbP) nitrosylation, (ii) nitrite-mediated nitrosylation of ferrous Mt-trHbN, Mt-trHbO, and Cj-trHbP, and (iii) NO-based reductive nitrosylation of ferric Mt-trHbN, Mt-trHbO, and Cj-trHbP is reported. Ferrous and ferric Mt-trHbN and Cj-trHbP display a very high reactivity towards NO; however, the conversion of nitrite to NO is facilitated primarily by ferrous Mt-trHbN. Values of kinetic and/or thermodynamic parameters reflect specific trHb structural features, such as the ligand diffusion pathways to/from the heme, the heme distal pocket structure and polarity, and the ligand stabilization mechanisms. In particular, the high reactivity of Mt-trHbN and Cj-trHbP reflects the great ligand accessibility to the heme center by two protein matrix tunnels and the E7-path, respectively, and the penta-coordination of the heme-Fe atom. In contrast, the heme-Fe atom of Mt-trHbO the ligand accessibility to the heme center of Mt-trHbO needs large conformational readjustments, thus limiting the heme-based reactivity. These results agree with different roles of Mt-trHbN, Mt-trHbO, and Cj-trHbP in vivo.

Introduction

Based on phylogeny, the globin superfamily contains three lineages: flavohemoglobins and single domain globins (lineage 1), protoglobins (Pgb) and globin coupled sensors (lineage 2), and truncated hemoglobins (trHbs; lineage 3). Members of the globin superfamily belong to two structural classes: one showing the classical 3-on-3 α-helical sandwich (lineages 1 and 2) and one having the 2-on-2 α-helical sandwich (lineage 3) (Fig. 1). Although no definitive conclusion can be drawn about the ancestral state of the globin fold, the occurrence of the 2-on-2 fold, but not of an isolated 3-on-3 fold, in all three kingdoms of life suggests that the 2-on-2 is the ancestral fold. In an evolutionary perspective, the predominant function of globins is (pseudo-)enzymatic, with O2 transport and storage being specialized functions associated with the evolution of metazoans [1]–[6].

Figure 1

Consensus phylogenetic tree of major globin groups from the three kingdoms of life.

This phylogenetic tree was based on the alignment of 150 sequences representing the following groups of globins: 10 plant non symbiotic 3/3 Hbs, 5 plant symbiotic 3/3 Hbs, 15 bacterial 3/3 globin-coupled sensors (GCSs), 4 3/3 protoglobins (Pgbs), 9 bacterial 2/2 Hb1s, 19 bacterial 2/2 Hb2s, 10 bacterial 2/2 Hb3s, 2 Chlamydomonas reinhardtii 2/2 Hbs, 2 ciliate 2/2 Hbs, 3 plant 2/2 Hbs, Thalassiosira pseudonana 2/2 Hb, 20 bacterial 3/3 flavohemoglobins (FHbs), 19 bacterial 3/3 single-domain Hbs, 9 eukaryote 3/3 FHbs, 1 diplomonad Giardia lamblia 3/3 Hb, and 1 mycetozoan Dictyostelium discoideum 3/3 Hb, Cyanidioschyzon merolae and Thalassiosira pseudonana 3/3 single-domain globins, and 3 vertebrate (i.e., human, bird, and fish) 3/3 neuroglobins (Ngbs), cytoglobins (Cygbs), α- and β-globins and myoglobins (Mbs), and 2 urochordate 3/3 Hbs. Modified from [1] (Copyright (2005) National Academy of Sciences, U S A).

TrHbs are widely distributed in bacteria and plants and have been found in some unicellular eukaryotes. They are distantly related to the 3-on-3 globins, showing less than 20% overall identity with the latter. Phylogenetic analysis of protein sequences shows that trHbs branch into three groups, designated N (or I), O (or II), and P (or III). TrHbs belonging to groups N and O separate into two and four subgroups, respectively; trHbs belonging to group P display a level of conservation higher than those of groups N and O (Fig. 1). The overall sequence identity between trHbs from different groups is ≤20%, but may be higher than 80% within a given group. Some bacteria display multiple trHbs belonging to different groups, suggesting a scenario for the evolution of the different groups where the group O gene is the ancestor, and group N and P genes are the results of duplications and transfer events [1]–[6].

TrHbs fold as a 2-on-2 α-helical sandwich characterized by a very short or absent A-helix, a brief CE inter-helical region, and most of the F-helix occurring as a loop, with only the B, E, G, and H α-helices surrounding the heme group. Specific residue deletions and substitutions distributed throughout the trHb sequence allow the achievement of the simplified fold, keeping at the same time a high affinity for the heme, a suitable ligand access to the heme-Fe atom, and the proper heme-Fe oxidation state. However, specific features, such as ligand entry/exit mechanisms holding to promote diffusion of ligands to/from the heme, the heme distal pocket structure and polarity, as well as ligand stabilization mechanisms distinguish members of the three trHb groups (Fig. 2) [5].

Figure 2

Three-dimensional structure of Mt-trHbN, Mt-trHbO, and Cj-trHbP.

(Top) Ribbon views of Mt-trHbN, Mt-trHbO, and Cj-trHbP, including the heme-Fe group (red) and the protein matrix cavity/tunnel systems (blue mash). (Bottom) The heme-Fe pocket of Mt-trHbN, Mt-trHbO, and Cj-trHbP. The heme group is colored in red. The heme ligand (a cyanide ion in all the three structures) and the side chains of residues in the B10, CD1, E7 and F8 topological positions are highlighted. Atomic coordinates were taken from the PDB entries 1S56 (Mt-trHbN), 1NGK (Mt-trHbO), and 2IG3 (Cj-trHbP). All pictures have been drawn with the Swiss-PdbViewer [91].

Mycobacterium tuberculosis (M. tuberculosis) possesses two trHbs, namely Mt-trHbN and Mt-trHbO, and a 3-on-3 flavohemoglobin, while Campylobacter jejuni (C. jejuni) possesses only Campylobacter jejuni trHb type P (Cj-trHbP) and a single-domain 3-on-3 globin [4], [5], [7]–[11].

Most trHbs are involved in the O2/NO chemistry and/or oxidation/reduction function, permitting the survival of the microorganism in the host [5], [11]–[13]. Noteworthy, Mt-trHbN binds reversibly isoniazid, a first-line anti-tuberculosis medication in prevention and treatment of tuberculosis, highlighting a direct role of the pro-drug to impair fundamental functions of mycobacteria, e.g. scavenging of reactive nitrogen and oxygen species, and metabolism [14].

Studies performed with Mycobacterium bovis Calmette-Guerin (BCG) demonstrated that the inactivation of the glbN gene impairs the ability of stationary phase cells to protect aerobic respiration from NO inhibition, suggesting that Mt-trHbN may play a vital role in protecting M. tuberculosis from NO toxicity in vivo [12], [15]. This functional assessment is supported by the observation that Mt-trHbN catalyzes the rapid oxidation of NO into nitrate [13], [16]. Heterologous expression of Mt-trHbN has also been shown to protect Escherichia coli against nitrosative stress [17]. On the other hand, Mt-trHbO does not detoxify NO efficiently [18], but displays peroxidase activity, suggesting an oxidation/reduction function [19]. Lastly, Cj-trHbP has been proposed to play a prominent role in C. jejuni respiration rather than in protection against reactive nitrogen and oxygen species. A strain of C. jejuni lacking trHbP turned out to be disadvantaged with respect to wild-type cells when grown under high aeration, achieving lower growth yields and consuming O2 at approximately half the rate displayed by wild-type cells. Although Cj-trHbP mutated cells are equally sensitive as the wild-type to NO and oxidative stress, the actual functional role of Cj-trHbP remains elusive [7], [8], [11].

Here, a detailed comparative analysis of kinetics and/or thermodynamics of (i) ferrous Mt-trHbN, Mt-trHbO, and Cj-trHbP (trHbN(II), Mt-trHbO(II), and Cj-trHbP(II), respectively) nitrosylation, (ii) nitrite-mediated nitrosylation of Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II), and (iii) NO-based reductive nitrosylation of ferric Mt-trHbN, Mt-trHbO, and Cj-trHbP (trHbN(III), Mt-trHbO(III), and Cj-trHbP(III), respectively) is reported. Ferrous and ferric Mt-trHbN and Cj-trHbP display a very high reactivity towards NO, the conversion of nitrite to NO being facilitated primarily by ferrous Mt-trHbN. This reflects the great ligand accessibility to the heme center of Mt-trHbN, and Cj-trHbP by two protein matrix tunnels and the E7-path, respectively, and the penta-coordination of the heme-Fe atom. In contrast, the accessibility to the heme center of Mt-trHbO needs large conformational readjustments, thus limiting the heme-based reactivity. These results agree with different roles of Mt-trHbN, Mt-trHbO, and Cj-trHbP in vivo.

Materials

Mt-TrHbN(III), Mt-trHbO(III), and Cj-trHbP(III) were cloned, expressed and purified as previously reported [20]–[22]. Mt-TrHbN(II), Mt-trHbO(II), and Cj-trHbP(II) were obtained, under anaerobic conditions, by adding sodium dithionite (final concentration, 1.0×10−3 M to 5.0×10−3 M). The concentration of Mt-trHbN, Mt-trHbO, and Cj-trHbP stock solutions was 3.0×10−6 M, 4.6×10−6 M, and 3.2×10−6 M, respectively.

Gaseous NO (from Aldrich Chemical Co. (Milwaukee, WI, USA) was purified by flowing through a NaOH column in order to remove acidic nitrogen oxides. The stock NO solution was prepared anaerobically by keeping in a closed vessel the degassed 5.0×10−3 M phosphate buffer solution (pH = 7.0) under NO at P = 760.0 mm Hg (T = 20.0°C). The solubility of NO in the aqueous buffered solution is 2.05×10−3 M, at P = 760.0 mm Hg and T = 20.0°C [23]. Sodium dithionite (final concentration, 5.0×10−4 M) was added to NO solutions (final concentration, 1.0×10−4 M to 5.0×10−4 M).

All the other chemicals were obtained from Merck AG (Darmstadt, Germany). All products were of analytical grade and used without purification unless stated.

Methods

Nitrosylation of Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II)

Kinetics of Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) nitrosylation were analyzed in the framework of the minimum reaction mechanism represented by Scheme A [23]:

Values of the pseudo-first-order rate constant (i.e., k) for Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) nitrosylation were obtained by mixing the trHb(II) solutions (final concentration, 1.5×10−6 M, 2.3×10−6 M, and 1.6×10−6 M, respectively) with the NO solution (final concentration, 5.0×10−6 M to 1.2×10−4 M) under anaerobic conditions.

No gaseous phase was present. The nitrosylation of Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) was monitored between 360 and 460 nm.

In view of the linear relationship between the absorbance change and the protein concentration change, values of the pseudo-first-order rate constant k were obtained according to Eqn. (1) [23]:

Values of the second-order rate constant for Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) nitrosylation (i.e., k on) were obtained from the dependence of k on the NO concentration (i.e., [NO]), according to Eqn. (2) [23]:where k off is the first-order rate constant for trHb(II)-NO denitrosylation.

Values of k on and k off for Mt-trHbN(II)(-NO), Mt-trHbO(II)(-NO), and Cj-trHbP(II)(-NO) (de)nitrosylation were obtained at pH 7.0 and 9.0 (1.0×10−1 M Bis-Tris propane buffer), and 20.0°C.

Nitrite-mediated nitrosylation of Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II)

Kinetics of the nitrite-mediated nitrosylation of Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) were analyzed in the framework of the minimum reaction mechanism depicted in Scheme B [24]–[35]:

Values of the pseudo-first-order rate constant (i.e., h; reaction (a) in Scheme B) for the nitrite-mediated nitrosylation of Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) were determined spectrophotometrically by mixing the trHb(II) solutions (final concentration, 1.5×10−6 M, 2.3×10−6 M, and 1.6×10−6 M, respectively) with the nitrite solutions (final concentration, 1.0×10−3 M to 1.0×10−2 M) in the presence of sodium dithionite (final concentration, 2.0×10−3 M), under anaerobic conditions. Remarkably, sodium dithionite concentration lower than 1.0×10−2 M does not effectively reduce nitrite to NO [28]. No gaseous phase was present. The nitrite-mediated nitrosylation of Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) was monitored between 360 and 460 nm.

In view of the linear relationship between the absorbance change and the protein concentration change, values of the pseudo-first-order rate constant for the nitrite-mediated nitrosylation of Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) (i.e., h) were obtained according to Eqn. (3) [24], [31], [32], [34], [35]:

Values of the second order rate constant for the nitrite-mediated nitrosylation of Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) (i.e., h on) were obtained from the dependence of h on the nitrite concentration (i.e., [NO2 −]), according to Eqn. (4) [24], [31], [32], [34], [35]:

Values of h on for the nitrite-mediated nitrosylation of Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) (reaction (a) in Scheme B) were obtained between pH 6.4 and 7.8 (1.0×10−1 M Bis-Tris propane buffer), at 20.0°C.

Reductive nitrosylation of Mt-trHbN(III), Mt-trHbO(III), and Cj-trHbP(III)

Kinetics and thermodynamics of the reductive nitrosylation of Mt-trHbN(III), Mt-trHbO(III), and Cj-trHbP(III) were analyzed in the framework of the minimum reaction mechanism represented by Scheme C [36]–[43]:

Values of the pseudo-first-order rate constants (i.e., l and b; reactions (a) and (c) in Scheme C, respectively) and of the dissociation equilibrium constant (i.e., L ( = l off/l on); reaction (a) in Scheme C) for Mt-trHbN(III), Mt-trHbO(III), and Cj-trHbP(III) reductive nitrosylation were obtained by mixing the trHb(III) solutions (final concentration, 1.5×10−6 M, 2.3×10−6 M, and 1.6×10−6 M, respectively) with the NO solution (final concentration, 2.5×10−5 M to 5.0×10−4 M) under anaerobic conditions. No gaseous phase was present. The reductive nitrosylation of Mt-trHbN(III), Mt-trHbO(III), and Cj-trHbP(III) was monitored between 360 and 460 nm.

In view of the linear relationship between the absorbance change and the protein concentration change, values of the pseudo-first-order rate constants l and b were obtained according to Eqns (5a)–(5c) [36]–[44]:

Values of l on and l off (reaction (a) in Scheme C) were determined from the dependence of l on [NO], according to Eqn. (6) [23]:

Values of L (reaction (a) in Scheme C) were obtained according to Eqn. (7) [23], [41], [42]:

where α is the molar fraction of the NO-bound Mt-trHbN(III), Mt-trHbO(III), and Cj-trHbP(III). Moreover, values of L were estimated from the l off/l on ratio [23].

Values of the second-order rate constant b OH− and of the first-order rate constant b H2O for the OH−- and H2O-catalyzed conversion of trHb(II)-NO+ to trHb(II), respectively (reaction (c) in Scheme C), were determined from the dependence of b on [OH−] according to Eqn. (8) [36]–[43]:

Values of L, l on, l off, b OH−, and b H2O for the reductive nitrosylation of Mt-trHbN(III), Mt-trHbO(III), and Cj-trHbP(III) (reactions (3a) and (3c) in Scheme C) were obtained between pH 8.4 and pH 9.4 (1.0×10−1 M Bis-Tris propane buffer), at 20.0°C. Values of k on and k off for Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) nitrosylation (Scheme A, and reaction (Cd) in Scheme C) were obtained at pH 9.0 and 20.0°C; see above).

The reductive nitrosylation of Mt-trHbN(III), Mt-trHbO(III), and Cj-trHbP(III) was also obtained anaerobically by keeping the trHb(III) solutions under purified gaseous NO (760 mmHg), between pH 8.4 and 9.4 (1.0×10−1 M Bis-Tris propane buffer) and 20.0°C.

Stopped-flow apparatus

Kinetic experiments have been carried out spectrophotometrically with the BioLogic SFM 2000 (Claix, France) rapid-mixing stopped-flow apparatus at single wavelength between 360 nm and 460 nm; the dead-time of the stopped-flow apparatus was ∼1 ms and the observation chamber was 1 cm.

Data analysis

Kinetic data obtained at different wavelengths have been normalized each other on the basis of the total absorbance change at the specific wavelength. The results are given as mean values of at least four experiments plus or minus the corresponding standard deviation. All data were analyzed using the Matlab program (The Math Works Inc., Natick, MA, USA).

Results and Discussion

Mixing Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) with NO solutions (pH 7.0 and 9.0, 20.0°C) induces a pH-independent shift of the optical absorption maximum of the Soret band from 430–433 nm (i.e., trHb(II)) to 416–418 nm (i.e., trHb(II)-NO) (Fig. 3, and Figs S1 and S2 in File S1, panel A, and Table 1). pH-independent absorbance spectra of Mt-trHbN(II)-NO, Mt-trHbO(II)-NO, and Cj-trHbP(II)-NO obtained mixing Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) with NO solutions correspond to those obtained by adding gaseous NO to the Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) solutions. Moreover, values of λmax and ε of the absorption spectra in the Soret region of Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) agree with those reported in the literature [8], [20], [21], [45], [46].

Figure 3

Mt-trHbN(II) nitrosylation at 20.0°C.

(A) Difference absorbance spectrum of Mt-trHbN(II) minus Mt-trHbN(II)-NO, at pH 7.0. (B) Difference absorbance spectrum of Mt-trHbN(II) minus Mt-trHbN(II)-NO, at pH 9.0. (C) Normalized averaged time courses of Mt-trHbN(II) nitrosylation, at pH 7.0. The NO concentration was 5.0×10−6 M (trace a) and 1.2×10−5 M (trace b). The time course analysis according to Eqn. (1) allowed the determination of the following values of k = 9.2×101 s−1 (trace a) and 2.8×102 s−1 (trace b). (D) Normalized averaged time courses of Mt-trHbN(II) nitrosylation, at pH 9.0. The NO concentration was 5.0×10−6 M (trace a) and 1.2×10−5 M (trace b). The time course analysis according to Eqn. (1) allowed the determination of the following values of k = 7.4×101 s−1 (trace a) and 1.9×102 s−1 (trace b). (E) Dependence of the pseudo-first-order rate-constant k for Mt-trHbN(II) nitrosylation on the NO concentration, at pH 7.0. The analysis of data according to Eqn. (2) allowed the determination of k on = (2.1±0.3)×107 M−1 s−1. (F) Dependence of the pseudo-first-order rate-constant k for Mt-trHbN(II) nitrosylation on the NO concentration, at pH 9.0. The analysis of data according to Eqn. (2) allowed the determination of k on = (1.6±0.3)×107 M−1 s−1. The Mt-trHbN(II) concentration was 1.5×10−6 M. The NO concentration was 1.0×10−4 M (panels A and B). Where not shown, the standard deviation is smaller than the symbol. For details, see text.

Table 1

Values of λmax and ε of the absorption spectra in the Soret region of ferric and ferrous derivatives of Mt-trHbN, Mt-trHbO, and Cj-trHbP, 20.0°C.

Derivative	λmax (nm)	ε (M−1 cm−1)	pH
Mt-trHbN(III)	406	1.41×105	6.4
Mt-trHbN(III)-OH−	410	1.25×105	9.4
Mt-trHbN(III)-NO	421	1.37×105	6.4 to 9.4
Mt-trHbN(II)	432	1.03×105	6.4 to 9.4
Mt-trHbN(II)-NO	417	1.22×105	6.4 to 9.4
Mt-trHbO(III)	409	1.04×105	6.4
Mt-trHbO(III)-OH−	412	1.36×105	9.4
Mt-trHbO(III)-NO	422	1.46×105	6.4 to 9.4
Mt-trHbO(II)	430	9.20×104	6.4 to 9.4
Mt-trHbO(II)-NO	416	1.28×105	6.4 to 9.4
Cj-trHbP(III)	410	1.41×105	6.4
Cj-trHbP(III)-OH−	414	1.09×105	9.4
Cj-trHbP(III)-NO	420	1.28×105	6.4 to 9.4
Cj-trHbP(II)	433	1.19×105	6.4 to 9.4
Cj-trHbP(II)-NO	418	1.41×105	6.4 to 9.4

Under all experimental conditions, the time course of NO binding to Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) corresponds to a mono-molecular process for more than 80% of its course (Fig. 3, and Figs S1 and S2 in File S1, panels B and C). Values of the pseudo-first-order rate constant for Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) nitrosylation (i.e., k; Eqn. (1)) are wavelength-independent at fixed NO concentration (data not shown).

Values of k increase linearly with the NO concentration (Fig. 3, and Figs S1 and S2 in File S1, panels D and E). The analysis of data reported in Figure 3, and Figures S1 and S2 in File S1 (panels D and E), according to Eqn. (2), allowed the determination of values of the second-order rate constant for Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) nitrosylation (i.e., k on; corresponding to the slope of the linear plots), which are essentially pH-independent (Fig. 3, and Figs S1 and S2 in File S1, panels D and E). The y intercept of the linear plots appears very close to zero (Fig. 3, and Figs S1 and S2 in File S1, panels D and E), indicating that values of k off for Mt-trHbN(II)-NO, Mt-trHbO(II)-NO, and Cj-trHbP(II)-NO denitrosylation are lower by at least two-orders of magnitude than values of k obtained at the lowest NO concentration (i.e., k off<1×10−1 s−1).

Values of k on for NO binding to Mt-trHbN(II) and Cj-trHbP(II) vary between 1.1×107 M−1 s−1 and 2.1×107 M−1 s−1 at pH 7.0 and pH 9.0 (Fig. 3, and Figs S1 and S2 in File S1, panels D and E, and Table 2). Values of k on for NO binding to Mt-trHbO(II) here determined at pH 7.0 and 9.0 ( = 1.9×105 M−1 s−1 and 2.3×105 M−1 s−1, respectively; Fig. S1 in File S1, panels D and E, and Table 2) are in agreement with data referring of the slow nitrosylation course, reflecting 80% of the whole process, previously reported at pH 7.5 ( = 1.8×105 M−1 s−1) [18]. Values of k on for NO binding to Mt-trHbO(II) and Cj-trHbP(II) increase changing the pH from 7.0 to 9.0, whereas values of k on for Mt-trHbN(II) nitrosylation decrease. However, the small magnitude of the pH-dependent change, essentially within the error limit, might suggest that these variations are not statistically relevant.

Table 2

Values of the second-order rate constant for nitrosylation of ferrous globins.

Heme-protein		k on (M−1 s−1)
Mt-trHbN		2.1×107 a
		1.6×107 b
Mt-trHbO		1.9×105 a
		2.3×105 b
Cj-trHbP		1.1×107 a
		1.7×107 b
Ma-Pgbc		2.7×107
Arabidopsis thaliana Hb class 1d		2.5×108
Glycine max Lbe		1.2×108
Scapharca inaequivalvis HbIf		1.6×107
Horse heart Mbg		1.6×107
Sperm whale Mbh		1.7×107
Mouse Ngbi		1.5×108
Tetrameric human Hbj	α-chains	2.6×107
	β-chains	2.6×107
Horse heart cytochrome c k		8.3
HSA-heme-Fel		6.3×106
Ibuprofen-HSA-heme-Fe(II)l		4.1×105
Warfarin-HSA-heme-Fe(II)l		4.8×105
Rabbit HPX-heme-Fem		6.3×103

pH 7.0 and 20.0°C. Present study.

pH 9.0 and 20.0°C. Present study.

pH 7.2 and 22.0°C. From [43].

pH 7.0 and 20.0°C. From [53].

pH 7.0 and 20.0°C. From [49].

pH 7.0 and 20.0°C From [50].

pH 9.2 and 20.0°C. From [41].

pH 7.0 and 20.0°C. From [48].

pH 7.5 and room temperature. From [51].

pH 7.0 and 20.0°C. From [47].

pH 6.5 and room temperature. From [36].

pH 7.0 and 20.0°C. From [55].

pH 7.0 and 10.0°C. From [52].

Values of k on for nitrosylation of ferrous heme-proteins span over eight orders of magnitude (Table 2) ([36], [41]–[43], [47]–[53] and present study), mainly reflecting the ligand accessibility to the heme distal pocket and the coordination of the heme-Fe(II) atom. Of note, H-bond interactions locking the heme distal residues TyrCD1 and TrpG8 of Mt-trHbO limits ligand access to the heme distal pocket (Fig. 2) [18]. In contrast, the low values of k on for ferrous horse heart cytochrome c and rabbit hemopexin-heme-Fe (HPX-heme-Fe) nitrosylation reflect the slow rate of hexa- to penta-coordination transition of the heme-Fe(II) atom, which precedes ligand binding ([36], [52] and present study). On the other hand, the fast hexa- to penta-coordination conversion of the heme-Fe(II) atom of Arabidopsis thaliana hemoglobin (Hb) class 1 and Glycine max Lb does not affect NO binding [31], [34]. Accordingly, the decrease of the k on value for the nitrosylation of ferrous human serum heme-Fe-albumin (HSA-heme-Fe) upon binding of warfarin and ibuprofen to the fatty acid binding site 2 (FA2) has been attributed to the allosteric drug-dependent hexa-coordination of the heme-Fe(II) atom [54]–[56].

Nitrite-mediated conversion of Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) to Mt-trHbN(II)-NO, Mt-trHbO(II)-NO, and Cj-trHbP(II)-NO

Mixing Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) with nitrite solutions (between pH 6.4 and 7.8, 20.0°C) induces a pH-independent shift of the optical absorption maximum of the Soret band from 430–433 nm (i.e., trHb(II)) to 416–418 nm (i.e., trHb(II)-NO) (Fig. 4, and Figs S3 and S4 in File S1, panel A, and Table 1). pH-independent absorbance spectra of Mt-trHbN(II)-NO, Mt-trHbO(II)-NO, and Cj-trHbP(II)-NO obtained mixing Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) with nitrite solutions correspond to those obtained by adding either NO solutions or gaseous NO to the Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) solutions. Moreover, values of λmax and ε of the absorption spectra in the Soret region of Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) derivatives agree with those reported in the literature [8], [20], [21], [45], [46].

Figure 4

Nitrite-mediated nitrosylation of Mt-trHbN(II), at 20.0°C.

(A) Difference absorbance spectrum of Mt-trHbN(II) minus Mt-trHbN(II)-NO, at pH 7.4. (B) Normalized averaged time courses of nitrite-mediated nitrosylation of Mt-trHbN(II), at pH 7.4. The nitrite concentration was 2.5×10−3 M (trace a) and 1.0×10−2 M (trace b). The time course analysis according to Eqn. (3) allowed the determination of the following values of h = 4.0×10−2 s−1 (trace a) and 1.6×10−1 s−1 (trace b). (C) Dependence of h on [NO2 −] for nitrite-mediated nitrosylation of Mt-trHbN(II), at pH 7.4. The continuous line was generated from Eqn. (4) with h on = (1.6±0.2)×101 M−1 s−1. (D) pH-Dependence of h on for the nitrite-mediated nitrosylation of Mt-trHbN(II). The slope of the continuous line was −1.00±0.01. The Mt-trHbN(II) concentration was 1.5×10−6 M. Where not shown, standard deviation is smaller than the symbol. For details, see text.

Under all experimental conditions, the time course of the nitrite-mediated conversion of Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) to Mt-trHbN(II)-NO, Mt-trHbO(II)-NO, and Cj-trHbP(II)-NO, respectively, corresponds to a mono-molecular process for more than 90% of its course (Fig. 4, and Figs S3 and S4 in File S1, panel B). Values of the pseudo-first-order rate constant for the nitrite-mediated conversion of Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) to Mt-trHbN(II)-NO, Mt-trHbO(II)-NO, and Cj-trHbP(II)-NO, respectively, (i.e., h; Eqn. (3)) are wavelength-independent at fixed nitrite concentration (data not shown). In agreement with the literature [28], values of h are independent of the dithionite concentration ranging between 1.0×10−3 M and 5.0×10−3 M (data not shown).

Values of h increase linearly with the nitrite concentration (Fig. 4, and Figs S3 and S4 in File S1, panel C). The analysis of data reported in Figure 4, and Figure S3 and S4 in File S1 (panel C) according to Eqn. (4) allowed the determination of values of the second-order rate constant for the nitrite-mediated conversion of Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) to Mt-trHbN(II)-NO, Mt-trHbO(II)-NO, and Cj-trHbP(II)-NO, respectively, (i.e., h on; corresponding to the slope of the linear plots). The y intercept of the linear plots corresponds to zero, indicating that the nitrite-mediated conversion of Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) to Mt-trHbN(II)-NO, Mt-trHbO(II)-NO, and Cj-trHbP(II)-NO, respectively, can be considered as an irreversible process.

Values of the rate constants for the nitrite-mediated conversion of Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) to Mt-trHbN(II)-NO, Mt-trHbO(II)-NO, and Cj-trHbP(II)-NO, respectively, are lower by several orders of magnitude than those of trHb(II) nitrosylation (Tables 1 and 2). This indicates that the formation of the transient Mt-trHbN(III), Mt-trHbO(III), and Cj-trHbP(III) species (Scheme B), which are quickly converted to Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II), respectively, by reaction with sodium dithionite, represents the rate-limiting step of the nitrite-mediated conversion of trHb(II) to trHb(II)-NO.

As reported for most heme-proteins [24]–[32], [34], [35], the nitrite-mediated conversion of Mt-trHbN(III), Mt-trHbO(II), and Cj-trHbP(II) to Mt-trHbN(II)-NO, Mt-trHbO(II)-NO, and Cj-trHbP(II)-NO, respectively, requires one proton for the NO and OH− formation (Scheme B). Indeed, on increasing the proton concentration by one pH unit, the rate of the nitrite-mediated conversion of trHbN(II) to trHbP(II)-NO (i.e., Log h on) increases by one-order of magnitude (Fig. 4, and Figs S3 and S4 in File S1, panel D). However, the increase of h on for the nitrite-mediated conversion of human cytoglobin on pH decrease has been interpreted accounting for the reversible pH-dependent penta-to-hexa-coordination transition of the heme-Fe(II) atom [33].

Values of h on for nitrite-mediated conversion of ferrous heme-proteins to their ferrous nitrosylated derivatives range between 7.0×10−2 M−1 s−1 and 6.8×101 M−1 s−1 (Table 3) ([25], [26], [29], [31]–[35] and present study), reflecting the different structural and chemical features of the heme site [5], [57]–[67]. Values of h on for NO2 − binding to penta-coordinated ferrous Mt-trHbN, Cj-trHbP, carp myoglobin (Mb)-1, carp Mb-2, horse heart Mb, and sperm whale Mb, ranging between 1.8 M−1 s−1 and 1.6×101 M−1 s−1, are among the fastest observed (Table 3). Of note, the low reactivity of Mt-trHbO reflects the unfavorable ligand accessibility to the heme pocket due to the locked conformation(s) of the heme distal residues TyrCD1 and TrpG8 (Fig. 2) [18]. The very different values of the interconversion rate for the hexa- to penta-coordination of the heme-Fe(II) atom modulates kinetics of nitrite binding to Synechocystis Hb, Arabidopsis thaliana nonsymbiotic Hbs classes 1 and 2, and rice nonsymbiotic Hb class 1, human Cygb, and horse heart cytochrome c [18], [31], [33], [34]. Accordingly, changes of the h on values for the nitrite-mediated conversion of ferrous human neuroglobin (Ngb) to its ferrous nitrosylated derivative (from 1.2×10−2 M−1 s−1 to 1.2×10−1 M−1 s−1) reflect the reversible redox-linked hexa-to-penta-coordination transition of the heme-Fe(II) atom. In fact, under oxidative conditions, the formation of the Cys46-Cys55 bridge stabilizes the high-reactive penta-coordinated heme-Fe(II) atom, thus facilitating the reaction. In contrast, under reductive conditions, the cleavage of the Cys46-Cys55 bridge leads to the formation of the low-reactive hexa-coordinated heme-Fe(II) atom [32]. Similarly, the inhibition of nitrite-dependent conversion of ferrous HSA-heme-Fe to the nitrosylated derivative reflects warfarin binding to the FA2 site with the concomitant penta-to-hexa coordination interconversion of the heme-Fe(II) atom [35], Lastly, the nitrite-mediated conversion of ferrous human Hb to the nitrosylated derivative is modulated allosterically, inositol hexakisphosphate impairing the reactivity of ferrous human Hb stabilizing the low-reactivity T-state [25], [26].

Table 3

Values of the second-order rate constant for the nitrite-mediated nitrosylation of ferrous globins.

Heme-protein		h on (M−1 s−1)
Mt-trHbNa		1.6×101
Mt-trHbOa		3.8×10−1
Cj-trHbPa		4.3
Synechocystis Hbb		6.8×101
Arabidopsis thaliana Hb class 1c		2.0×101
Arabidopsis thaliana Hb class 2c		4.9
Rice non symbiotic Hb class 1b		3.3×101
Carp Mb-1d		5.3
Carp Mb-2d		1.8
Horse heart Mbe		2.9
Sperm whale Mbf		6.0
	HisE7Alae	1.8
	HisE7Leue	<0.2
Mouse Ngbg		5.1
Human Cygbh		1.4×10−1
Human Ngb	Cys46-Cys55e , i	1.2×10−1
	Cys46/Cys55e , j	1.2×10−2
Tetrameric human Hb	T statef	1.2×10−1
	R statef	6.0
Horse heart cytochrome c k		7.0×10−2
HSA-heme-Fel		1.3
Warfarin-HSA-heme-Fe(II)l		9.3×10−2

pH 7.4 and 20.0°C. Present study.

pH 7.0; unknown temperature [31].

pH 7.4 and 25.0°C. From [34].

pH 7.6 and 25.0°C. From [90].

pH 7.4 and 25.0°C. From [32].

pH 7.4 and 25.0°C. From [25].

pH 7.4 and 25.0°C. From [29].

pH 7.0 and 25.0°C. From [33].

In “Human Ngb Cys46–Cys55”, the Cys46 and Cys55 residues form an intramolecular disulphide bond.

In “Human Ngb Cys46/Cys55”, the Cys46 and Cys55 residues do not form the intramolecular disulphide bond.

pH 7.4 and 25.0°C. From [33].

pH 7.4 and 20.0°C. From [35].

The reductive nitrosylation of Mt-trHbN(III), Mt-trHbO(III), and Cj-trHbP(III) occurs only at pH≥8.4. Of note, Mt-trHbN(III) binds reversibly NO without converting to Mt-trHbN(II)-NO at pH 7.5 [68].

Mixing Mt-trHbN(III), Mt-trHbO(III), and Cj-trHbP(III) with NO solutions (between pH 8.4 and 9.4, 20.0°C) induces a shift of the optical absorption maximum of the Soret band from 406–414 nm (i.e., trHb(III)) to 420–422 nm (i.e., trHb(III)-NO) (Fig. 5, and Figs S5 and S6 in File S1, panel A, and Table 1). Then, the optical absorption maximum of the Soret band shifts from 420–422 nm (i.e., trHb(III)-NO) to 416–418 nm (i.e., trHb(II)-NO) (Fig. 5, and Figs S5 and S6 in File S1, panel A, and Table 1). Absorbance spectra of Mt-trHbN(II)-NO, Mt-trHbO(II)-NO, and Cj-trHbP(II)-NO obtained mixing Mt-trHbN(III), Mt-trHbO(III), and Cj-trHbP(III) with NO solutions (Fig. 5, and Figs S5 and S6 in File S1, panel A, and Table 1) correspond to those obtained by: (i) adding gaseous NO to the Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) solutions (Fig. 3, and Figs S1 and S2 in File S1, panel A, and Table 1), and (ii) mixing Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) with nitrite solutions in the presence of dithionite (Fig. 4, and Figs S3 and S4 in File S1, panel A, and Table 1). Moreover, values of λmax and ε of the absorption spectra in the Soret region of Mt-trHbN, Mt-trHbO, and Cj-trHbP derivatives agree with those reported in the literature [8], [20], [21], [45], [46], [68]–[70].

Figure 5

Reductive nitrosylation of Mt-trHbN(III), at 20.0°C.

(A) Difference absorbance spectra of Mt-trHbN(III) minus Mt-trHbN(III)-NO and of Mt-trHbN(III)-NO minus Mt-trHbN(II)-NO (open and fillrd circles, respectively), at pH 9.0. (B) Normalized averaged time courses of Mt-trHbN(III) reductive nitrosylation, at pH 9.0. The NO concentration was 2.5×10−5 M (trace a) and 2.0×10−4 M (trace b). The time course analysis according to Eqns (5a)–(5c) allowed the determination of the following values of parameters α, l, and b: trace a - α = 0.61, l = 6.0 s−1, and b = 2.5×10−3 s−1; and trace b - α = 0.89, l = 2.6×101 s−1, and b = 2.4×10−3 s−1. (C) Dependence of l on [NO] for Mt-trHbN(III) reductive nitrosylation, at pH 9.0. The continuous line was generated from Eqn. (6) with l on = (1.4±0.2)×105 M−1 s−1 and l off = 1.6±0.2 s−1. (D) Dependence of α on [NO] for Mt-trHbN(III) reductive nitrosylation, at pH 9.0. The continuous line was generated from Eqn. (7) with L = (1.6±0.2)×10−5 M. (E) Dependence of b on [NO] for Mt-trHbN(III) reductive nitrosylation, at pH 9.0. The average b value is 2.5×10−3 s−1 (dashed line). (F) Dependence of b on [OH−] for Mt-trHbN(III) reductive nitrosylation. The continuous line was generated from Eqn. (8) with b OH− = (1.7±0.2)×102 M−1 s−1 and b H2O = (6.4±0.7)×10−4 s−1. The Mt-trHbN(III) concentration was 1.5×10−6 M. Where not shown, standard deviation is smaller than the symbol. For details, see text.

Over the whole NO concentration range explored, the time course for Mt-trHbN(III), Mt-trHbO(III), and Cj-trHbP(III) reductive nitrosylation corresponds to a biphasic process (Fig. 5, and Figs S5 and S6 in File S1, panel B); values of l and b are wavelength-independent at fixed NO concentration (data not shown).

The first step of kinetics for Mt-trHbN(III), Mt-trHbO(III), and Cj-trHbP(III) reductive nitrosylation (indicated by l in Scheme C) is a bimolecular process as observed under pseudo-first order conditions (Fig. 5, and Figs S5 and S6 in File S1, panel C) at all the pH values investigated. Plots of l versus [NO] are linear (Eqn. (6)), the slope corresponding to l on. Values of the second-order rate constant l on over the pH range explored (i.e., from pH 8.4 to 9.4; Tables S1–S3 in File S1) vary between 1.1×105 M−1 s−1 and 2.3×105 M−1 s−1 (for Mt-trHbN(III)), between 8.5×103 M−1 s−1 and 2.1×104 M−1 s−1 (for Mt-trHbO(III)), and between 8.9×104 M−1 s−1 and 1.7×105 M−1 s−1 (for Cj-trHbP(III)). The y intercept of plots of l versus [NO] corresponds to the dissociation rate constant l off (Eqn. (6)), whose values over the pH range explored (i.e., from pH 8.4 to 9.4; Tables S1–S3 in File S1) vary between 1.5 s−1 and 2.3 s−1 (for Mt-trHbN(III)), between 1.3 s−1 and 2.4 s−1 (for Mt-trHbO(III)), and between 5.9 s−1 and 8.3 s−1 (for Cj-trHbP(III)).

Between pH 8.4 and pH 9.4, the molar fraction of trHb(III)-NO (i.e., α, Eqn. (7)) increases as a function of free [NO], tending to level off at [NO] >10×L; the analysis of data according to Eqn. (7) allowed to calculate values of L (Fig. 5, and Figs S5 and S6 in File S1, panel D). Values of L over the pH range explored (i.e., from pH 8.4 to 9.4; Tables S1–S3 in File S1) vary between 9.3×10−6 M and 1.9×10−5 M (for Mt-trHbN(III)), between 9.8×10−5 M and 2.1×10−4 M (for Mt-trHbO(III)), and between 3.1×10−5 M and 7.9×10−5 M (for Cj-trHbP(III)). According to the trHb(III):NO 1∶1 stoichiometry of reaction Ca in Scheme C, the Hill coefficient n is 1.00±0.02. As expected for a simple system [23], values of L correspond to those of l off/l on, under all experimental conditions (Table 4 and Tables S1–S3 in File S1).

Table 4

Values of kinetic and thermodynamic parameters for reductive nitrosylation of ferric heme-proteins.

Heme-protein		l on (M−1 s−1)	l off (s−1)	L (M)	l off/l on (M)	b OH− (M−1 s−1)	b H2O (s−1)
Mt-trHbN		1.4×105 a	1.6a	1.6×10−5 a	1.1×10−5 a	1.7×102 b	6.4×10−4 b
Mt-trHbO		9.2×103 a	2.1a	1.9×10−4 a	2.3×10−4 a	2.4×102 b	2.9×10−4 b
Cj-trHbP		1.1×105 a	8.1a	6.5×10−5 a	7.4×10−5 a	9.1×102 b	4.8×10−4 b
Ma-Pgb		4.8×104 c	2.6c	6.1×10−5 c	5.4×10−5 c	2.9×103 d	4.1×10−4 d
Glycine max Lbe		1.4×105 e	3.0e	2.1×10−5 e	2.1×10−5 e	3.3×103 f	3.0×10−4 f
Scapharca inaequivalvis HbI		3.2×101 g	<1×10−3 g	n.d.	3.1×10−5 g	>2×106 h	n.d.
Horse heart Mb		6.8×104 i	5.2i	1.2×10−4 i	7.6×10−5 i	3.9×102 j	n.d
Sperm whale Mb		1.9×105 k	1.4×10l	7.7×10−5 k	n.d.	3.2×102 m	n.d.
Human Ngb	Cys46–Cys55n	2.1×101 o	2.5×10−3 o	n.d.	1.2×10−4 p	≥2×106 q	n.d.
	Cys46/Cys55r	2.9o	2.0×10−3 o	n.d.	1.9×10−4 p	≥5×105 q	n.d.
Tetrameric human Hb		n.d.	n.d.	8.3×10−5 s	n.d.	3.2×103 t	1.1×10−3t
Horse cytochrome c		n.d.	n.d.	7.1×10−5 u	n.d.	1.5×103 m	n.d.
HSA-heme-Fe		2.1×104 v	3.1×10−1 v	1.8×10−5 v	1.5×10−5 v	4.4×103 w	3.5×10−4 w
Rabbit HPX-heme-Fe		1.3×101 x	≤10−4 x	n.d.	≤8×10−6 x	≥7×105 y	n.d.

pH 9.0 and 20.0°C. Present study.

20.0°C. Present study.

pH 7.2 and 22.0°C. From [43].

22.0°C. From [43].

pH 7.3 and 20.0°C. From [39].

20.0°C. From [39].

pH 7.5 and 20.0°C. From [38].

Derived from values of b and [OH−] [38], at 20.0°C.

pH 9.2 and 20.0°C. From [41].

Derived from values of b and [OH−] [41], at 20.0°C.

pH 8.79 and room temperature. From [37].

Derived from values of l on and L [37], at pH 8.79 and room temperature.

Room temperature. From [37].

In “Human Ngb Cys46–Cys55”, the Cys46 and Cys55 residues form an intramolecular disulphide bond.

pH 7.0 and room temperature. From [74].

Derived from values of l on and l off [74], at room temperature.

Derived from values of b and [OH−] [74], at pH 7.0 and room temperature.

In “Human Ngb Cys46/Cys55”, the Cys46 and Cys55 residues do not form the intramolecular disulphide bond.

pH 7.1 and room temperature. From [37].

Room temperature. From [37].

pH 8.35 and room temperature. From [37].

pH 7.5 and 20.0°C. From [42].

20.0°C. From [42].

pH 7.0 and 10.0°C. From [40].

10.0°C. From [40].

n.d., not determined.

The second step of kinetics for Mt-trHbN(III), Mt-trHbO(III), and Cj-trHbP(III) reductive nitrosylation (indicated by Cb-Cd in Scheme C) follows a [NO]-independent monomolecular behavior (Fig. 5, and Figs S5 and S6 in File S1, panel E) at all the pH values investigated. According to Scheme C, the value of b increases linearly on increasing [OH−] (i.e., from pH 8.4 to 9.4; Fig. 5, and Figs S5 and S6 in File S1, panel F, and Tables S1–S3 in File S1). The slope and the y intercept of the plot of b versus [OH−] correspond to b OH− and b H2O values. Values of b OH− and b H2O for for Mt-trHbN(III), Mt-trHbO(III), and Cj-trHbP(III) reductive nitrosylation are 1.7×102 M−1 s−1 and 6.4×10−4 s−1, 2.4×102 M−1 s−1 and 2.9×10−4 s−1, and 9.1×102 M−1 s−1 and 4.8×10−4 s−1, respectively (Table 4).

Over the whole pH range explored (i.e., from pH 8.4 to 9.4), the reductive nitrosylation of Mt-trHbN(III), Mt-trHbO(III), and Cj-trHbP(III) is an irreversible process, as the spectra of Mt-trHbN(II)-NO, Mt-trHbO(II)-NO, and Cj-trHbP(II)-NO revert to Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) instead of Mt-trHbN(III), Mt-trHbO(III), and Cj-trHbP(III) by merely pumping off gaseous NO or bubbling helium through the Mt-trHbN(II)-NO, Mt-trHbO(II)-NO, and Cj-trHbP(II)-NO solutions; however, the denitrosylation process (i.e., the Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) formation) needs about 12 hours to be completed.

Under all experimental conditions, free Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) was never detected spectrophotometrically because of the very rapid reaction of Mt-trHbN(II), Mt-trHbO(II), and Cj-trHbP(II) with NO (k on≥8.6×105 M−1 s−1; Table 1).

Ferric Mt-trHbN, Cj-trHbP, Methanosarcina acetivorans protoglobin (Ma-Pgb), horse heart Mb, sperm whale Mb, and HSA-heme-Fe react rapidly with NO ([41], [43], [48], [55] and present study), even though the binding of a weak exogenous ligand (i.e., a water molecule) occurs likely at the sixth coordination position of the heme-Fe(III) atom [5], [57], [58], [66], [67], [71]. As reported above, the low reactivity of Mt-trHbO reflects the unfavorable ligand accessibility to the heme distal pocket [18]. Unexpectedly, the very low value of l on for the reductive nitrosylation of penta-coordinated Scapharca inaequivalvis HbI may reflect either the non-occurrence of the heme-Fe(III)-NO intermediate or the concomitant nitrosation of the Cys92 residue affecting the heme-Fe reactivity [38]. On the other hand, the reactivity of ferric hexa-coordinated Glycine max Lb, horse cytochrome c, and rabbit HPX-heme-Fe ([37], [39], [40], [72] and present study) reflects the hexa- to penta-coordination conversion of the heme-Fe atom ([59], [65], [73] and present study) (Table 4). As reported for the nitrite-mediated nitrosylation of ferrous human Ngb [32] (Table 3), NO binding to ferric human Ngb is modulated by the formation and cleavage of the Cys46-Cys55 disulphide bridge [74].

Values of l off for NO dissociation from nitrosylated heme-Fe(III) proteins span over five orders of magnitude (Table 4) ([37]–[43], [74] and present study), reflecting the different stabilization mode of the heme-Fe bound ligand by heme distal residues [5], [54], [58], [59], [66], [71], [73], [75]–[79].

Although values of l on and l off NO binding to heme-Fe(III) proteins are very different (Table 4) ([37]–[43], [74] and present study), values of L ( = l off/l on) are similar (Table 4) ([37]–[43], [74] and present study), indicating the occurrence of kinetic compensation phenomena.

Values of b OH− for reductive nitrosylation of heme-Fe(III) proteins, spanning over four orders of magnitude (Table 4) ([37]–[43], [74] and present study), possibly reflect either the different OH− accessibility to the heme pocket ([5], [23], [58]–[60], [63], [64], [67], [71], [75], [80]–[84]) or the heme-Fe(III) protein reduction potentials [37], [39]. Although the values of b OH− and b H2O cannot be compared directly, OH− ions catalyze reductive nitrosylation of heme-Fe(II)-NO+ much more efficiently than H2O. Moreover, the linear dependence of b on [OH−] indicates that no additional elements appear to be involved in irreversible reductive nitrosylation of Mt-trHbN, Mt-trHbO, and Cj-trHbP (present study), as previously reported for related heme-Fe(III) proteins (reaction (c) in Scheme C; Fig. 5, and Figs S5 and S6 in File S1, panel E) [37]–[43], [74].

Reductive nitrosylation of most ferric heme-proteins is limited by the OH−-mediated reduction of the heme-Fe(II)-NO+ complex to heme-Fe(II) (reaction (c) in Scheme C) [37]–[39], [41]–[43], [74]. By contrast, NO binding to hexa-coordinated horse cytochrome c and ferric rabbit HPX-heme (reaction (a) in Scheme C) represents the rate-limiting step [37], [40] (Tables 2 and 4 for comparison).

Conclusion and Perspectives

The occurrence of different types of trHbs (i.e., trHbN, trHbO and trHbP) in bacteria, plants and some unicellular eukaryotes (Fig. 1) opens the question of their role, also in view of their frequently simultaneous presence in the same organism; this envisages the possibility that different types of trHbs reflect different physiological roles in these organisms [1]–[6], [11].

The high reactivity of Mt-trHbN and Cj-trHbP reflects both the penta-coordination of the heme-Fe-atom and the ligand accessibility to the heme pocket (Fig. 2). Indeed, the ligand access pathway through protein matrix tunnels in penta-coordinated Mt-trHbN (Fig. 2) [5], [83] and the E7-path in penta-coordinated Cj-trHbP (Fig. 2) [5], [22], [66] is characterized by the lowest energy barrier, with values close to those displayed by penta-coordinated sperm whale Mb. Of note, ligand entry to and exit from the heme distal pocket of sperm whale Mb is modulated by gating movement of the HisE7 residue [75]. In contrast, the low reactivity of penta-coordinated Mt-trHbO [21] reflects H-bond interactions that lock the heme distal residues TyrCD1 and TrpG8 into a conformation(s) that limits ligand access to the heme distal pocket (Fig. 2) [18], [70]. Of note, the E7-path appears to sustain ligand diffusion to the heme distal cavity of Mt-trHbO; indeed, Mt-trHbO hosts two protein matrix cavities instead of the protein matrix tunnel occurring in Mt-trHbN (Fig. 2) [84].

M. tuberculosis is massively exposed to NO during its intra-macrophagic life, and Mt-trHbN has been shown to play a prominent role in protection from nitrosative stress [16]. Therefore, Mt-trHbN is most likely to function as NO dioxygenase that converts toxic NO into harmless nitrate in the presence of oxygen, and relieves toxicity due to NO and nitrosative stress. This is also consistent with the NO-inducible response of the gene encoding for trHbN in M. tuberculosis, both in vitro and inside infected macrophages [85]. Of note, also the promoter of the trHbO gene is induced during macrophage infection, though it poorly responds to NO induction [85]. Based on biochemical data, the evidence of trHbO in protection from NO is weak, while it is more likely that this protein is involved in oxygen sensing and aerobic respiration [86]. Similar considerations could hold true for the enteric pathogen C. jejuni, which colonizes the intestinal tract of birds and can cause enteritis in humans. Like other pathogens, C. jejuni is exposed to NO and other nitrosating species during host infection [87], the expression of both bacterial Cj-trHbP and 3-on-3 globin being upregulated by NO [11]. The 3-on-3 globin plays the major role in resistance to nitrosative stress and aerobically converts NO to nitrate [88], whereas the contribution of trHbP is less prominent. In contrast, both globins are devoid of NO-protective activity under oxygen-limited conditions that normally exist in vivo [89]. Therefore, the role of Cj-trHbP is clearly distinct form that of the 3-on-3 globin, being related to O2 metabolism [7], [8], likely performing a peroxidase or P450-type of oxygen chemistry [9].

As a whole, the comparison of the biochemical properties of microbial globins, including trHbs, will allow to shed light on their functional diversification and on the molecular bases of microbe ecology.

Supporting tables and figures.

(DOCX)

Click here for additional data file.