当前位置:首页 >> 生物学 >>

Reactions of ferrous neuroglobin and cytoglobin with nitrite under anaerobic conditions


Journal of Inorganic Biochemistry 102 (2008) 1777–1782

Contents lists available at ScienceDirect

Journal of Inorganic Biochemistry
journal homepage: www.elsevier.com/locate/jinorgbio

Reactions of ferrous neuroglobin and cytoglobin with nitrite under anaerobic conditions
Morten Gjerning Petersen a, Sylvia Dewilde b, Angela Fago a,*
a b

Department of Biological Sciences, University of Aarhus, Building 1131, Universitetsparken, DK-8000 Aarhus C, Denmark Department of Biomedical Sciences, University of Antwerp, Universiteitsplein 1, B-2610 Antwerp, Belgium

a r t i c l e

i n f o

a b s t r a c t
Recent evidence suggests that the reaction of nitrite with deoxygenated hemoglobin and myoglobin contributes to the generation of nitric oxide and S-nitrosothiols in vivo under conditions of low oxygen availability. We have investigated whether ferrous neuroglobin and cytoglobin, the two hexacoordinate globins from vertebrates expressed in brain and in a variety of tissues, respectively, also react with nitrite under anaerobic conditions. Using absorption spectroscopy, we ?nd that ferrous neuroglobin and nitrite react with a second-order rate constant similar to that of myoglobin, whereas the ferrous heme of cytoglobin does not react with nitrite. Deconvolution of absorbance spectra shows that, in the course of the reaction of neuroglobin with nitrite, ferric Fe(III) heme is generated in excess of nitrosyl Fe(II)–NO heme as due to the low af?nity of ferrous neuroglobin for nitric oxide. By using ferrous myoglobin as scavenger for nitric oxide, we ?nd that nitric oxide dissociates from ferrous neuroglobin much faster than previously appreciated, consistently with the decay of the Fe(II)–NO product during the reaction. Both neuroglobin and cytoglobin are S-nitrosated when reacting with nitrite, with neuroglobin showing higher levels of S-nitrosation. The possible biological signi?cance of the reaction between nitrite and neuroglobin in vivo under brain hypoxia is discussed. ? 2008 Elsevier Inc. All rights reserved.

Article history: Received 21 January 2008 Received in revised form 1 May 2008 Accepted 19 May 2008 Available online 29 May 2008 Keywords: Nitrite Nitric oxide Neuroglobin Cytoglobin Hypoxia

1. Introduction Since their recent discovery in the brain [1] and in various tissues of vertebrate animals [2,3], respectively, neuroglobin (Ngb) and cytoglobin (Cygb) have been the focus of numerous studies aimed at unravelling their biological functions as well as the functional and structural differences between these proteins and the other members of the globin family, hemoglobin (Hb) and myoglobin (Mb). In contrast to the highly abundant Hb and Mb, the oxygen carriers of the blood and of heart and skeletal muscle, respectively, Ngb and Cygb are both expressed at low levels (low l-molar range) and are hexacoordinate in the ferrous Fe(II) and ferric (met) Fe(III) forms, with the proximal and distal His coordinating the heme iron. Both proteins react slowly with external ligands, as the distal His must dissociate to give access to the heme [4]. Ngb, the phylogenetically most ancient and most conserved among vertebrate globins [5], is intrinsically unstable in the presence of oxygen when in the ferrous form and tends rapidly to oxidize to the ferric form [6,7]. Although the functional role of Cygb remains still largely unsolved [5], in vitro and in vivo studies have clearly shown that Ngb is able to protect neurons against oxygen deprivation during
* Corresponding author. Tel.: +45 8942 2591; fax: +45 8942 2586. E-mail address: angela.fago@biology.au.dk (A. Fago). 0162-0134/$ - see front matter ? 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.jinorgbio.2008.05.008

hypoxia or brain ischemia [8–10]. As reasoned in previous papers [7,11] because of the low protein concentration, the autoxidation of oxygenated Ngb and the fairly low oxygen af?nity of Ngb (7.5 Torr) under physiological conditions [7], this protective function of Ngb is unlikely to re?ect oxygen storage in the brain, as originally presumed [4], or scavenging of excess nitric oxide (NO) by the oxygenated protein [12], but is most likely due to some redox reactions that may protect highly-metabolizing nerve cells from hypoxic damage. So far we have identi?ed few such redox reactions of Ngb that may be of potential significance in vivo, including the rapid scavenging of reactive oxygen and nitrogen species, i.e. toxic peroxynitrite [13] and hydrogen peroxide [14], which are generated at high rate upon brain ischemia, and the extremely fast reaction with cytochrome c that may act to prevent the onset of cellular death, or apoptosis [15]. Among the redox reactions catalyzed by globin proteins, the anaerobic reduction of nitrite to nitric oxide (NO) has received considerable attention in recent years. Nitrite is normally present at fairly high concentrations within mammalian tissues (0.1–10 lM) [16] where it is a major end-product of the cellular messenger NO, which controls vasodilation, neurotransmission and mitochondrial respiration. Nitrite has a crucial role in the cellular protection against ischemia and reperfusion, as seen in heart and liver [17]. Nitrite reacts with thiols under acidic conditions to form S-nitrosothiols

1778

M.G. Petersen et al. / Journal of Inorganic Biochemistry 102 (2008) 1777–1782

(involved in cellular redox signalling) [18,19] and, during hypoxia, it can be recycled back to NO radical [20–22]. Although controversy exists on the exact mechanism [23,24] and on its effective role [25,26], the reduction of nitrite to NO catalyzed by deoxygenated Hb [27,28] has been proposed to contribute to the overall hypoxic vasodilatatory response and to increase blood ?ow and oxygen supply to hypoxic tissues [29,30]. Similarly, at low in vivo oxygen levels, the conversion of nitrite into NO catalyzed by deoxy Mb may downregulate mitochondrial respiration by inhibiting cytochrome c oxidase and therefore extend oxygen availability when heart becomes hypoxic, as during ischemic episodes [31]. Interestingly, a recent study reports that nitrite may also protect the brain from ischemia and reperfusion [32], which suggests that similar nitrite-dependent NO-generating reactions, possibly catalyzed by Ngb or Cygb, also take place in the brain. To further explore the origin of the protective role of Ngb and the possible biological functions of Cygb, we have analyzed the reactions between low levels of nitrite and ferrous Ngb and Cygb under anaerobic conditions. This study extends previous work on the reactions of nitrite with deoxy Hb and Mb and allows further insights into the mechanisms of the nitrite reactivity of mammalian globins.

lowed by UV–visible spectroscopy in 1 cm anaerobic cuvettes under magnetic stirring using a diode-array HP 8543 spectrophotometer. The wavelength range was 500–600 nm. Small volumes of degassed nitrite solutions were injected using gastight Hamilton syringes into anaerobic cuvettes containing $10 lM anaerobic solutions of ferrous Ngb or Cygb and measurements were started <3 s from nitrite additions. Initial reaction rates were calculated at 558 nm. Kinetic traces at 558 nm could best be ?tted by a variable n-order decay. Non-linear least-squares ?tting procedures were performed using the Table Curve 2D software (Systat, CA, USA). 2.3. Product analysis Reference absorbance spectra for Ngb in the deoxy Fe(II), met Fe(III) and nitrosyl Fe(II)–NO forms were measured in 0.05 M phosphate buffer containing 0.5 mM EDTA, pH 7.4 and ?tted by least-squares analyses to the observed spectra using the multicomponent analysis software provided with the HP 8543 spectrophotometer [34]. Reference spectra of pure met NgbFe(III) and deoxy NgbFe(II) were obtained as previously described using published extintion coef?cients [7]. Pure nitrosyl NgbFe(II)–NO was obtained under anaerobic conditions after adding NO in excess over heme to ferrous samples or to ferric samples by reductive nitrosylation. Both methods yielded identical NgbFe(II)–NO spectra, as previously shown [13]. At the end of kinetic experiments, concentration of S-nitrosothiols was measured by the Saville reaction after addition of ammonium sulfamate to eliminate unreacted nitrite [35]. 2.4. Dissociation of NO from NgbFe(II) The rate of NO dissociation from ferrous Ngb was measured anaerobically at 25 °C using a HP 8543 spectrophotometer by adding NgbFe(II)–NO (prepared by adding NO gas to ferrous samples and degassing with argon to eliminate excess NO) to a 9-molar excess of deoxy horse heart Mb contained in 1 cm anaerobic cuvette. Both anaerobic samples contained 5 mM sodium dithionite. Sample concentrations were measured using known extintion coef?cient for the deoxy Fe(II) derivatives [7,36]. Due to its very high rate of NO binding (1.7 ? 107 M?1 s?1) and slow NO dissociation (1.2 ? 10?4 s?1) [37] deoxy Mb minimizes rebinding of free NO to Ngb and is thus an ef?cient NO trap. After mixing, the concentration of Ngb and Mb was of 8 and 70 lM, respectively. Buffer was 0.05 M phosphate buffer containing 0.5 mM EDTA, pH 7.4. Changes in the spectral absorbance of Mb associated to the conversion from the ferrous MbFe(II) to the nitrosyl MbFe(II)–NO form were followed over time at 533 nm, which is isosbestic for the NgbFe(II)– NO and NgbFe(II) absorbance spectra.

2. Materials and methods All reagents, including horse heart Mb, were from Sigma, unless otherwise is speci?ed. Water was milli-Q grade. Ultrapure (>99.998%) nitrogen (N2), NO and argon gas were from Air Liquide. NO was puri?ed from higher nitrogen oxides by passing through argon-equilibrated 5 M NaOH solutions. Recombinant murine Ngb and human Cygb were expressed in Escherichia coli and puri?ed as met Fe(III) as previously described [6,33]. Potassium nitrite was dissolved in 0.05 M phosphate buffer containing 0.5 mM EDTA, pH 7.4, shortly before use. Solutions were degassed with N2 and kept protected from light. Nitrite concentration was established by the Griess reaction (e 540 nm = 50 mM?1 cm?1) using the Griess reagent (Alexis). 2.1. Anaerobic gel-?ltration Deoxy Fe(II) derivatives of both proteins were prepared by anaerobic gel-?ltration according to a procedure used previously [15]. Brie?y, N2-equilibrated samples of ferric protein ($50 ll, >1.5 mM) were injected using a gastight Hamilton syringe into anaerobic vials containing 1 mg solid sodium dithionite and after a few seconds loaded using a gastight syringe onto a 1.5 ml G-25 Sephadex Super?ne (GE Healthcare) gel-?ltration column sealed with a rubber cap and equilibrated with degassed 0.05 M phosphate buffer, 0.5 mM EDTA, pH 7.4. The column outlet was connected to a needle in order to collect the reduced ferrous sample directly into an anaerobic rubber-sealed cuvette. A spectrum was taken to verify full conversion to the ferrous deoxygenated form and to measure protein concentration (e 558 nm = 28.9 mM?1 cm?1) [7]. To ensure complete oxygen removal from the gel-?ltration column, at least 100 ml of degassed buffer and 2 ? 100 ll sodium dithionite solutions (1 mg/ml) followed by extensive washing with degassed buffer were passed through the column before sample loading. Hamilton gastight syringes were extensively purged with degassed buffer before use. 2.2. Kinetic measurements and data analysis Reactions of nitrite with ferrous Ngb and Cygb were studied under anaerobic conditions in 0.05 M phosphate buffer containing 0.5 mM EDTA, pH 7.4 at 25 °C. The kinetics of the reaction were fol-

3. Results The reaction between ferrous murine Ngb and nitrite was investigated under anaerobic conditions at pH 7.4, at a protein concentration of $10 lM and with nitrite concentrations in the range 7–230 lM. As shown in Fig. 1A, upon addition of nitrite the visible absorbance spectrum changed over time from that of pure ferrous deoxy Ngb, with characteristic peaks at 528 and 558 nm, to a spectrum showing a broad peak at 532 nm. The spectra collected over time were analysed by least-squares ?tting analyses using reference spectra for deoxy NgbFe(II), met NgbFe(III) and nitrosyl NgbFe(II)–NO (Fig. 1B). Fitting deconvolution analyses of the absorbance spectra indicated that the decrease in deoxy NgbFe(II) could be accounted for by the increase in the NgbFe(III) and NgbFe(II)–NO forms (Fig. 1D), according to the general scheme for nitrite reduction by deoxygenated globins:

M.G. Petersen et al. / Journal of Inorganic Biochemistry 102 (2008) 1777–1782

1779

A
Absorbance

0.20

B
Absorbance

0.12 0.10

0.15

0.08 0.06 0.04 0.02

observed fitted Fe(II) Fe(III) Fe(II)-NO

0.10

0.05

0.00 500 520 540 560 580 600

0.00 500 520 540 560 580 600

Wavelength (nm)

Wavelength (nm)

C
ΔAbsorbance
0.005 0.000 -0.005 500 520 540 560 580 600

D
Concentration (mM)

0.012 0.010 0.008 0.006 0.004 0.002 0.000 0 1000 2000 3000

Fe(II) Fe(III) Fe(II)-NO

Wavelength (nm)

4000

Time (s)
Fig. 1. Reaction of ferrous neuroglobin with nitrite. Reaction of ferrous Ngb (7.4 lM) with an equimolar amount of nitrite in 0.05 M phosphate buffer, 0.5 mM EDTA, pH 7.4, 25 °C, under anaerobic conditions. A, Visible absorbance spectra measured at 5 min intervals. B, Comparison between the observed spectrum taken at 30 min (dashed line) and the ?tted spectrum (continuous line) obtained by least-squares as described in Materials and Methods. The relative contribution of each reference spectrum is indicated by thick lines (black, Fe(II); grey, Fe(III); dark grey, Fe(II)–NO). C, Residual spectrum after the least-squares ?tting shown in B. D, Time course of the concentrations (mean ± SE) of Fe(II) (black), Fe(III) (grey) and Fe(II)–NO (dark grey) derivatives obtained as shown in B.

NgbFe?II? ? NO? ? H?
NgbFe?III?ANO ? OH? 2
NgbFe?III? ? NO ? OH?

?1?

where the NO produced may then bind to available NgbFe(II) hemes to generate the NgbFe(II)–NO derivative:

NgbFe?II? ? NO
NgbFe?II?ANO

?2?

in the range 0.1–0.7 (Fig. 2). This reveals a complex reaction mechanism involving multiple interrelated steps. Therefore, the dependence of the observed rates on nitrite concentration was analyzed in terms of initial rate, as also done in previous studies on the reaction of myoglobin and hemoglobin with nitrite [38,39]. The apparent second-order rate constant for the reaction of deoxy NgbFe(II)

Under the experimental conditions here used, NgbFe(III) was the main product of the reaction, whereas NgbFe(II)–NO formed only transiently and at low levels (<20%). Observed and ?tted spectra were essentially superimposable, with amplitudes of residual spectra of <2% of the total absorbance (Fig. 1C). Interestingly, albeit small, residual spectra showed peaks at $535 and $570 nm, consistent with a small contribution from the transient formation of the NgbFe(III)–NO intermediate (having peaks at 532 and 565 nm) [13] in the course of the reaction (Eq. (1)). Such shortlived intermediate has been identi?ed by stopped-?ow kinetic methods in the reaction between NgbFe(II)–NO and peroxynitrite [13]. No changes of the absorbance spectra were observed when reacting the products NgbFe(II)–NO or NgbFe(III) with nitrite (not shown), as also seen for human Ngb [13]. Remarkably, in contrast to Ngb, deoxy Fe(II) Cygb did not react with nitrite when using equimolar or a molar excess of nitrite over heme. In all cases, the spectrum of the protein remained unchanged in the deoxy Fe(II) form (not shown). At the nitrite concentrations here investigated, the decay in the absorbance at 558 nm over time could be best ?tted (coef?cient of determination r2 > 0.99) according to a variable n-order exponential decay, rather than mono- or double exponential, with n varying

1.0

Relative Amplitude

7μM 49 μM 97 μM 143 μM 230 μM

0.5

0.0 0 500 1000 1500 2000

Time (s)
Fig. 2. Kinetic traces of the reaction of ferrous neuroglobin with nitrite. Absorbance traces at 558 nm for the reactions of ferrous Ngb with varying nitrite concentrations (7–230 lM) measured in 0.05 M phosphate buffer, 0.5 mM EDTA, pH 7.4, 25 °C, under anaerobic conditions. The best ?t to the traces is indicated and corresponds to a variable n-order decay: 7 lM nitrite (7.4 lM heme), n = 0.099 (r2 = 0.993); 49 lM nitrite (10.3 lM heme), n = 0.438 (r2 > 0.999); 97 lM nitrite (10.7 lM heme), n = 0.574 (r2 > 0.999); 143 lM nitrite (8.4 lM heme), n = 0.457 (r2 > 0.999); 230 lM nitrite (9.3 lM heme), n = 0.740 (r2 > 0.999).

1780
0.0020 0.0016 0.0012 0.0008 0.0004 0.0000 0.0000 0.0001

M.G. Petersen et al. / Journal of Inorganic Biochemistry 102 (2008) 1777–1782

A
1.0

Initial Rate (s-1)

Absorbance

0.5

0.0

0.0002

500

550

600

650

Nitrite (M)
Fig. 3. Plot of initial rates versus nitrite concentration for the reaction between ferrous neuroglobin and nitrite. Initial rates were calculated from the absorbance traces at 558 nm for the reactions of ferrous Ngb with varying nitrite concentrations (7–230 lM) measured in 0.05 M phosphate buffer, 0.5 mM EDTA, pH 7.4, 25 °C, under anaerobic conditions. The apparent second-order rate constant obtained from the slope of the linear regression of the data is 5.1 ± 0.4 M?1 s?1.

Wavelength (nm)

B
Absorbance 533 nm

0.86

0.84

0.82

0.80

0.3

Ngb Cygb

0.78 0 500 1000 1500 2000

SNO/heme

Time (s)
0.2

0.1

0.0 0.0000 0.0001 0.0002

Fig. 5. NO dissociation rate from NgbFe(II)–NO measured with deoxy Mb as NO trap. A, Visible absorbance spectra (0–15 min) and B, absorbance trace at 533 nm measured at 1 min intervals obtained under anaerobic conditions by mixing nitrosylated Ngb with deoxy horse heart Mb in excess in 0.05 M phosphate buffer, 0.5 mM EDTA, 5 mM sodium dithionite, at 25 °C. The ?nal concentration of Ngb and Mb was 8 and 70 lM, respectively.

Nitrite (M)
Fig. 4. Fraction of S-nitrosylated protein versus nitrite concentration. The fraction of SNO per heme was measured by the Saville reaction after reactions of ferrous Ngb (closed symbols) and Cygb (open symbols) with varying nitrite concentrations in 0.05 M phosphate buffer, 0.5 mM EDTA, pH 7.4, 25 °C, under anaerobic conditions, as shown in Figs. 2 and 3.

with nitrite derived from the linear ?t of the plot was 5.1 ± 0.4 M?1 s?1 at 25 °C and at pH 7.4 (Fig. 3). At the end of each experiment, the levels of S-nitrosothiols formed in Ngb and Cygb were measured by the Saville reaction. Both proteins were S-nitrosated when reacting with nitrite, with Ngb showing higher levels of S-nitrosation per heme that increased with nitrite concentration (Fig. 4). We measured the NO dissociation rate from Ngb using deoxy horse heart Mb as a NO trap to understand the decay of nitrosylated NgbFe(II)–NO observed in the reaction with nitrite (see Fig. 1D). As shown in Fig. 5, the absorbance spectra measured showed the progressive change from a maximum at 553 nm to two maxima at 548 and 577 nm, indicating the conversion of the deoxy to the nitrosyl form of Mb [37]. Fitting analysis of the absorbance trace at 533 nm indicated that NO dissociates from murine Ngb at a rate of (2.9 ± 0.1) ? 10?3 s?1 at 25 °C and pH 7.4. From this value, the calculated equilibrium constant of Ngb for NO is KNO = (kon/koff)/ (1 + KHis) = 7.7 ? 106 M?1, where kon = 1.5 ? 108 M?1 s?1 [40] and KHis = 6.7 ? 103 at pH 7.4 [41]. 4. Discussion This study shows that of the two hexacoordinate mammalian globins only Ngb but not Cygb may ef?ciently catalyze the conversion of nitrite to NO. When compared to the nitrite reductase activities of Mb [39] and Hb [38], the apparent second-order rate

constant of Ngb (5.1 M?1 s?1) is much closer to that of Mb (6 M?1 s?1) than to that of Hb (0.23–0.4 M?1 s?1) measured under similar experimental conditions. Within the range of nitrite concentrations here investigated, the observed reaction rates are low (<0.00145 s?1) and thus are not limited by the rapid association/ dissociation of the distal His from the ferrous or ferric heme of Ngb [13,41,42]. In analogy to other reactions involving Ngb [13,41], the reaction between nitrite and deoxy Ngb appears complex, as the absorbance decay followed a non-integral exponential. As discussed below, this complexity can be explained by the nature of the intermediate complex formed in the course of the reaction

NgbFe?III?ANO
NgbFe?II?ANO?

?3?

characterized by a partial charge transfer from NO to the heme iron. This intermediate complex may react with excess nitrite to generate either NO2 or N2O3 [43] and increase the number of elementary reaction steps contributing to the overall mechanism:

NgbFe?III?ANO ? NO?
NgbFe?II?ANO ? NO2 2 NgbFe?III?ANO ? NO?
NgbFe?II? ? N2 O3 2

?4? ?5?

Another factor that increases the complexity of the overall reaction with nitrite is the equilibrium

NgbFe?II?AHis64
NgbFe?II? His64

?6?

between the hexacoordinate and the pentacoordinate forms of NgbFe(II) that is not present in pentacoordinate deoxy Hb or Mb. This equilibrium is strongly shifted to the left, with dissociation equilibrium constants of 1 ? 10?4 or 3 ? 10?4 at neutral pH depending on the protein conformation [41]. The existence of two distinct protein conformations of Ngb, in spite of its monomeric structure, is another factor that adds to the overall complexity of

M.G. Petersen et al. / Journal of Inorganic Biochemistry 102 (2008) 1777–1782

1781

the reaction. Clearly, the existence of a number of possible concurrent reactions results into the variable order of the overall reaction obtained when ?nding the best ?t to the kinetic traces (Fig. 2). The reaction of Ngb with nitrite here studied produced NgbFe(III) in excess of NgbFe(II)–NO. NO has a lower af?nity for ferric than for ferrous globins and dissociates rapidly (at a rate of $1 s?1 in Hb and Mb and 0.12 s?1 in human Ngb) [13]. NO dissociating from the ferric heme of Ngb will primarily bind to the low amount of ferrous hemes in the pentacoordinate state, i.e. those where the iron binding site is not occupied by the distal His64 (or by the substrate nitrite). As shown here, the af?nity for NO of ferrous Ngb is lower than that of Mb or Hb ($1011–1012 M?1) [40] by a factor 104–105, which explains the low yield of nitrosyl Ngb product observed. Moreover, NO dissociates from ferrous Ngb faster (2.9 ? 10?3 s?1) than estimated in a previous study (2 ? 10?4 s?1), where dithionite, a poor scavenger of NO, was used [40]. This NO dissociation from Ngb causes a progressive decrease in the Fe(II)–NO levels towards the end of the reaction (Fig. 1D) that is not observed in Hb or Mb [31,38]. Obviously, in the reaction with nitrite, NO rebinding to the ferrous heme of Ngb causes a slower decay of the nitrosyl complex (t1/2 $ 1500 s; Fig. 1D) than expected from NO dissociation alone (t1/2 240 s). A predominance of met Fe(III) over nitrosyl Fe(II)–NO has also been found, albeit less pronounced, in the reaction of deoxy Hb with low levels of nitrite [23,24,44], similar to those used here, whereas studies made with high (mM range) levels of nitrite have shown equivalent formation of met and nitrosyl products [38]. Although such discrepancy has been ascribed at least in part to oxygen contamination of Hb samples [38], the oxy derivative of Ngb would not be stable under the conditions here used, i.e. without appropriate reducing agents [7], and would convert rapidly (t1/2 $ 120 s for murine Ngb) [6] into met, which does not react with nitrite. Dissociation of nitrosonium ion NO+ from ferrous heme of the intermediate complex (Eq. (3)) will rapidly regenerate nitrite from the hydration of NO+ or nitrosate thiols, as shown in studies made on Hb, where higher yields of S-nitrosated protein were found in the presence of ferric heme [35]. Thus the direct interaction of NO+ with thiols may be a possible pathway of S-nitrosothiols formation. Additionally, excess nitrite may react with the intermediate complex Fe(III)–NO to generate either NO2 or N2O3 [43], with strong S-nitrosating activity (Eqs. (4) and (5)). Such reaction mechanisms has been proposed to explain how nitrite-dependent vasoactivity could effectively escape intact red blood cells and contribute to the hypoxic vasodilatatory response [45]. Cysteine residues in Ngb and Cygb are signi?cantly more reactive than in human Hb [7,46]. In murine Ngb, two Cys residues are present, Cys55, which is internal and at only $15 ? from the heme iron, and Cys120, which is exposed to the solvent [46]. Because of its closer location to the heme iron, Cys55 is likely to be the primary site of heme-dependent S-nitrosation. Moreover, the presence of S-nitrosated thiols in Cygb shows that S-nitrosation may also be heme-independent, which suggests that other S-nitrosating mechanisms, albeit less ef?cient, are also feasible. A large (>200 ?3) hydrophobic cavity present in both Ngb and Cygb and connecting heme distal site to the solvent [33,47,48] may function to stabilize relatively apolar S-nitrosating species, including HNO2 and N2O3, and even in acting as a NO binding site. Interestingly, Cys120 in murine Ngb is located close to the entrance of the internal cavity [47]. Similarly, the two cysteines of human Cygb, Cys38 and Cys83, are located in close proximity to part of the internal cavity, the Xe1 binding site [49]. The lack of reaction of ferrous Cygb with nitrite is intriguing. Although both Ngb and Cygb are hexacoordinate in the ferrous and ferric state, they differ considerably in their heme reactivity. Cygb has a more apolar distal heme environment and a considerably lower tendency to oxidize in the presence of oxygen [50], which con-

trasts to the high redox potential of Ngb [6]. In addition, access to the heme pocket of Cygb may be partially blocked by a salt bridge between a propionate side chain of the heme and the side chain of Arg84 [33], whereas in Ngb a similar, but weaker bond formed by Lys67 may render the heme pocket more accessible [48]. Taken together, these factors may all contribute to the observed lack of heme reactivity of Cygb with nitrite. The physiological function of Cygb remains to be established, but a role in O2-requiring collagen formation in connective tissue has been proposed [51]. 4.1. Biological signi?cance Both NgbFe(II) and NgbFe(II)–NO forms are produced during overexpression of recombinant Ngb in E. coli under micro-anaerobic conditions in the presence of nitrite [40], which suggests that the reaction here studied also occurs in intact bacterial cells. We have shown that, under physiological conditions of pH and temperature and at the oxygen tension expected in neurons ($1 Torr) [52], Ngb will be almost completely ($90%) in the ferrous deoxy state [7] and thus potentially able to catalyze the conversion of endogenous nitrite to NO. The NO scavenging reaction by oxygenated globins (that converts NO into nitrate and oxygenated heme into ferric heme), proposed as a major in vivo function of Ngb [12] appears unlikely, as it would require high levels of the oxygenated form (the expected $10% of oxygenated Ngb under normal conditions would decrease even further during brain hypoxia), which moreover would form at the continuous expenses of reducing equivalents necessary to reconvert ferric Ngb to the ferrous state [7,11]. Even though mammalian globins convert nitrite into NO less ef?ciently (<6 M?1 s?1) than bacterial nitrite reductases (>108 M?1 s?1) [53], it is important to note that very little NO (n-molar levels) is needed to activate soluble guanylate cyclase or to inhibit cytochrome c oxidase, the two main physiological targets of the biological activity of NO in mammals. As for Mb in heart muscle [31], the NO generated from nitrite may transiently inhibit mitochondrial respiration and have a protective effect in limiting oxygen consumption at the onset of acute hypoxia. However, the low overall concentration of Ngb in the brain would render nitrite reduction less ef?cient than for Mb, which reacts with nitrite with a similar second-order rate constant but is present at a much higher protein concentration. Thus, whereas the protective nitrite reductase activity of Mb in the ischemic heart appears established, further in vivo studies are required to better understand the physiological role of the corresponding reaction of Ngb in the brain and in the retina, where Ngb might be present at higher concentrations. The levels of S-nitrosated Ngb (and in particular of Cygb) that would be formed in vivo in the reaction with physiological concentrations of nitrite (<10 lM) appear too low to function as NO storage but could instead have a potential role in redox signalling pathways, which is another aspect worth of further investigation.

5. Abbreviations Ngb Cygb Hb Mb Neuroglobin Cytoglobin Human adult hemoglobin Myoglobin

Acknowledgements We thank Roy E. Weber (Aarhus) and Luc Moens (Antwerp) for useful discussions and two anonymous referees for constructive comments. We also thank the Danish Natural Science

1782

M.G. Petersen et al. / Journal of Inorganic Biochemistry 102 (2008) 1777–1782 [27] K. Cosby, K.S. Partovi, J.H. Crawford, R.P. Patel, C.D. Reiter, S. Martyr, B.K. Yang, M.A. Waclawiw, G. Zalos, X. Xu, K.T. Huang, H. Shields, D.B. Kim-Shapiro, A.N. Schechter, R.O. Cannon, M.T. Gladwin, Nature Med. 9 (2003) 1498–1505. [28] E. Nagababu, S. Ramasamy, D.R. Abernethy, J.M. Rifkind, J. Biol. Chem. 278 (2003) 46349–46356. [29] J.H. Crawford, T.S. Isbell, Z. Huang, S. Shiva, B.K. Chacko, A.N. Schechter, V.M. Darley-Usmar, J.D. Kerby, J.D. Lang Jr., D. Kraus, C. Ho, M.T. Gladwin, R.P. Patel, Blood 107 (2006) 566–574. [30] T.S. Isbell, M.T. Gladwin, R.P. Patel, Am. J. Physiol. Heart Circ. Physiol. 293 (2007) H2565–H2572. [31] S. Shiva, Z. Huang, R. Grubina, J. Sun, L.A. Ringwood, P.H. MacArthur, X. Xu, E. Murphy, V.M. Darley-Usmar, M.T. Gladwin, Circ. Res. 100 (2007) 654–661. [32] K.H. Jung, K. Chu, S.Y. Ko, S.T. Lee, D.I. Sinn, D.K. Park, J.M. Kim, E.C. Song, M. Kim, J.K. Roh, Stroke 37 (2006) 2744–2750. [33] D. de Sanctis, S. Dewilde, A. Pesce, L. Moens, P. Ascenzi, T. Hankeln, T. Burmester, M. Bolognesi, J. Mol. Biol. 336 (2004) 917–927. [34] A. Fago, A.L. Crumbliss, J. Peterson, L.L. Pearce, C. Bonaventura, Proc. Natl. Acad. Sci. USA 100 (2003) 12087–12092. [35] S. Herold, G. Rock, J. Biol. Chem. 278 (2003) 6623–6634. [36] E. Antonini, M. Brunori, Hemoglobin and Myoglobin in their Reactions with Ligands, North-Holland Publishing Company, Amsterdam, 1971. [37] V.G. Kharnitov, J. Bonaventura, V.S. Sharma, in: M. Feelisch, J.S. Stamler (Eds.), Methods in Nitric Oxide Research, John Wiley and Sons, Chichester, UK, 1996, pp. 39–45. [38] K.T. Huang, A. Keszler, N. Patel, R.P. Patel, M.T. Gladwin, D.B. Kim-Shapiro, N. Hogg, J. Biol. Chem. 280 (2005) 31126–31131. [39] Z. Huang, S. Shiva, D.B. Kim-Shapiro, R.P. Patel, L.A. Ringwood, C.E. Irby, K.T. Huang, C. Ho, N. Hogg, A.N. Schechter, M.T. Gladwin, J. Clin. Invest. 115 (2005) 2099–2107. [40] S. Van Doorslaer, S. Dewilde, L. Kiger, S.V. Nistor, E. Goovaerts, M.C. Marden, L. Moens, J. Biol. Chem. 278 (2003) 4919–4925. [41] A. Fago, A.J. Mathews, S. Dewilde, L. Moens, T. Brittain, J. Inorg. Biochem. 100 (2006) 1339–1343. [42] D. Hamdane, L. Kiger, S. Dewilde, J. Uzan, T. Burmester, T. Hankeln, L. Moens, M.C. Marden, FEBS J. 272 (2005) 2076–2084. [43] B.O. Fernandez, I.M. Lorkovic, P.C. Ford, Inorg. Chem. 43 (2004) 5393–5402. [44] M.P. Doyle, R.A. Pickering, T.M. DeWeert, J.W. Hoekstra, D. Pater, J. Biol. Chem. 256 (1981) 12393–12398. [45] J.M. Robinson, J.R. Lancaster Jr., Am. J. Respir. Cell Mol. Biol. 32 (2005) 257–261. [46] D. Hamdane, L. Kiger, S. Dewilde, B.N. Green, A. Pesce, J. Uzan, T. Burmester, T. Hankeln, M. Bolognesi, L. Moens, M.C. Marden, J. Biol. Chem. 278 (2003) 51713–51721. [47] B. Vallone, K. Nienhaus, M. Brunori, G.U. Nienhaus, Proteins 56 (2004) 85–92. [48] A. Pesce, S. Dewilde, M. Nardini, L. Moens, P. Ascenzi, T. Hankeln, T. Burmester, M. Bolognesi, Structure 11 (2003) 1087–1095. [49] D. de Sanctis, S. Dewilde, A. Pesce, L. Moens, P. Ascenzi, T. Hankeln, T. Burmester, M. Bolognesi, Biochem. Biophys. Res. Commun. 316 (2004) 1217– 1221. [50] H. Sawai, N. Kawada, K. Yoshizato, H. Nakajima, S. Aono, Y. Shiro, Biochemistry 42 (2003) 5133–5142. [51] M. Schmidt, F. Gerlach, A. Avivi, T. Laufs, S. Wystub, J.C. Simpson, E. Nevo, S. Saaler-Reinhardt, S. Reuss, T. Hankeln, T. Burmester, J. Biol. Chem. 279 (2004) 8063–8069. [52] P.W. Hochachka, G.N. Somero, Biochemical Adaptation. Mechanism and Process in Physiological Evolution, Oxford University Press, Oxford, 2002. [53] M.C. Silvestrini, M.G. Tordi, G. Musci, M. Brunori, J. Biol. Chem. 265 (1990) 11783–11787.

Research Council and the Novo Nordisk Foundation for ?nancial support. References
[1] T. Burmester, B. Weich, S. Reinhardt, T. Hankeln, Nature 407 (2000) 520–523. [2] T. Burmester, B. Ebner, B. Weich, T. Hankeln, Mol. Biol. Evol. 19 (2002) 416– 421. [3] J.T. Trent III, M.S. Hargrove, J. Biol. Chem. 277 (2002) 19538–19545. [4] P. Ascenzi, A. Bocedi, D. de Sanctis, A. Pesce, M. Bolognesi, M.C. Marden, S. Dewilde, L. Moens, T. Hankeln, T. Burmester, Biochem. Molec. Biol. Educ. 32 (2004) 305–313. [5] T. Hankeln, B. Ebner, C. Fuchs, F. Gerlach, M. Haberkamp, T.L. Laufs, A. Roesner, M. Schmidt, B. Weich, S. Wystub, S. Saaler-Reinhardt, S. Reuss, M. Bolognesi, D. de Sanctis, M.C. Marden, L. Kiger, L. Moens, S. Dewilde, E. Nevo, A. Avivi, R.E. Weber, A. Fago, T. Burmester, J. Inorg. Biochem. 99 (2005) 110–119. [6] S. Dewilde, L. Kiger, T. Burmester, T. Hankeln, V. Baudin-Creuza, T. Aerts, M.C. Marden, R. Caubergs, L. Moens, J. Biol. Chem. 276 (2001) 38949–38955. [7] A. Fago, C. Hundahl, S. Dewilde, K. Gilany, L. Moens, R.E. Weber, J. Biol. Chem. 279 (2004) 44417–44426. [8] Y. Sun, K. Jin, X.O. Mao, Y. Zhu, D.A. Greenberg, Proc. Natl. Acad. Sci. USA 98 (2001) 15306–15311. [9] Y. Sun, K. Jin, A. Peel, X.O. Mao, L. Xie, D.A. Greenberg, Proc. Natl. Acad. Sci. USA 100 (2003) 3497–3500. [10] A.A. Khan, Y. Wang, Y. Sun, X.O. Mao, L. Xie, E. Miles, J. Graboski, S. Chen, L.M. Ellerby, K. Jin, D.A. Greenberg, Proc. Natl. Acad. Sci. USA 103 (2006) 17944– 17948. [11] A. Fago, C. Hundahl, H. Malte, R.E. Weber, IUBMB Life 56 (2004) 689–696. [12] M. Brunori, A. Giuffre, K. Nienhaus, G.U. Nienhaus, F.M. Scandurra, B. Vallone, Proc. Natl. Acad. Sci. USA 102 (2005) 8483–8488. [13] S. Herold, A. Fago, R.E. Weber, S. Dewilde, L. Moens, J. Biol. Chem. 279 (2004) 22841–22847. [14] E. Fordel, L. Thijs, L. Moens, S. Dewilde, FEBS J. 274 (2007) 1312–1317. [15] A. Fago, A.J. Mathews, L. Moens, S. Dewilde, T. Brittain, FEBS Lett. 580 (2006) 4884–4888. [16] N.S. Bryan, B.O. Fernandez, S.M. Bauer, M.F. Garcia-Saura, A.B. Milsom, T. Rassaf, R.E. Maloney, A. Bharti, J. Rodriguez, M. Feelisch, Nature Chem. Biol. 1 (2005) 290–297. [17] M.R. Duranski, J.J.M. Greer, A. Dejam, S. Jaganmohan, N. Hogg, W. Langston, R.P. Patel, S.F. Yet, X. Wang, C.G. Kevil, M.T. Gladwin, D.J. Lefer, J. Clin. Invest. 115 (2005) 1232–1240. [18] P.A. Morris, D.L.H. Williams, J. Chem. Soc. Perkin Trans. 2 (1988) 513–516. [19] D.T. Hess, A. Matsumoto, S.O. Kim, H.E. Marshall, J.S. Stamler, Nature Rev. Mol. Cell Biol. 6 (2005) 150–166. [20] V.P. Reutov, E.G. Sorokina, Biochemistry (Mosc.) 63 (1998) 874–884. [21] J.O. Lundberg, E. Weitzberg, Arterioscler. Thromb. Vasc. Biol. 25 (2005) 915– 922. [22] A. Fago, F.B. Jensen, in: B. Tota, B.A. Trimmer (Eds.), Advances in Experimental Biology, Nitric Oxide, vol. 1, Elsevier, Amsterdam, 2007, pp. 199–212. [23] M. Angelo, D.J. Singel, J.S. Stamler, Proc. Natl. Acad. Sci. USA 103 (2006) 8366– 8371. [24] E. Nagababu, S. Ramasamy, J.M. Rifkind, Biochemistry 46 (2007) 11650–11659. [25] T. Dalsgaard, U. Simonsen, A. Fago, Am. J. Physiol. Heart Circ. Physiol. 292 (2007) H3072–H3078. [26] S. Deem, J.H. Min, J.D. Moulding, R. Eveland, E.R. Swenson, Am. J. Physiol. Heart Circ. Physiol. 292 (2007) H963–H970.


相关文章:
更多相关标签: