Báo cáo khoa học: Reductive nitrosylation and peroxynitrite-mediated oxidation of heme–hemopexin doc

The optical absorption spectrum of HPX–hemeIII obtained by mixing the HPX–hemeII–NO and peroxynitrite solutions e414¼ 1.16· 105m1Æcm1 Fig.. max Lb [30], horse Mb [47], human Ngb [32] and

oxidation of heme–hemopexin

Paolo Ascenzi1,2, Alessio Bocedi2, Giovanni Antonini1, Martino Bolognesi3 and Mauro Fasano4

1 Department of Biology and Interdepartmental Laboratory for Electron Microscopy, University ‘Roma Tre’, Rome, Italy

2 National Institute for Infectious Diseases IRCCS ‘Lazzaro Spallanzani’, Rome, Italy

3 Department of Biomolecular Sciences and Biotechnology and CNR-INFM, University of Milan, Italy

4 Department of Structural and Functional Biology and Center of Neuroscience, University of Insubria, Busto Arsizio, Italy

Heme scavenging by high- and low-density

lipopro-teins, serum albumin and hemopexin (HPX) provides

protection against heme and iron oxidative damage,

limits access by pathogens to heme, and contributes to

iron homeostasis by recycling the heme iron During

the first seconds after the appearance of heme in the

plasma, > 80% of this powerful oxidizer binds to

high- and low-density lipoproteins, and only the

remaining 20% binds to serum albumin and HPX

Serum albumin and HPX then remove the heme from high- and low-density lipoproteins Afterwards, heme transits to HPX, which releases it into hepatic paren-chymal cells only after internalization of HPX–heme

by specific receptor-mediated endocytosis After deliv-ering the heme intracellularly, HPX is released intact into the bloodstream and the heme is degraded [1–15] HPX–heme is formed by two four-bladed b-propel-ler domains, resembling two thick disks that lock

Keywords

nitric oxide; peroxynitrite;

peroxynitrite-mediated oxidation; rabbit hemopexin;

reductive nitrosylation

Correspondence

P Ascenzi, Department of Biology,

University ‘Roma Tre’, I-00146 Rome, Italy

Fax: +39 06 5517 6321

Tel: +39 06 5517 3200⁄ 2

E-mail: ascenzi@bio.uniroma3.it

(Received 14 September 2006, revised 14

November 2006, accepted 21 November

2006)

doi:10.1111/j.1742-4658.2006.05609.x

Hemopexin (HPX), which serves as a scavenger and transporter of toxic plasma heme, has been postulated to play a key role in the homeostasis of

NO In fact, HPX–heme(II) reversibly binds NO and facilitates NO scaven-ging by O2 HPX–heme is formed by two four-bladed b-propeller domains The heme is bound between the two b-propeller domains, residues His213 and His266 coordinate the heme iron atom HPX–heme displays structural features of heme-proteins endowed with (pseudo-)enzymatic activities In this study, the kinetics of rabbit HPX–heme(III) reductive nitrosylation and peroxynitrite-mediated oxidation of HPX–heme(II)–NO are reported

In the presence of excess NO, HPX–heme(III) is converted to HPX– heme(II)–NO by reductive nitrosylation The second-order rate constant for HPX–heme(III) reductive nitrosylation is (1.3 ± 0.1)· 101

pH 7.0 and 10.0C NO binding to HPX–heme(III) is rate limiting In the absence and presence of CO2 (1.2· 10)3m), excess peroxynitrite reacts with HPX–heme(II)–NO (2.6· 10)6m) leading to HPX–heme(III) and

NO, via the transient HPX–heme(III)–NO species Values of the second-order rate constant for HPX–heme(III)–NO formation are (8.6 ± 0.8)· 104 and (1.2 ± 0.2)· 106m)1Æs)1 in the absence and pres-ence of CO2, respectively, at pH 7.0 and 10.0C The CO2-independent value of the first-order rate constant for HPX–heme(III)–NO denitrosyla-tion is (4.3 ± 0.4)· 10)1s)1, at pH 7.0 and 10.0C HPX–heme(III)–NO denitrosylation is rate limiting HPX–heme(II)–NO appears to act as an effi-cient scavenger of peroxynitrite and of strong oxidants and nitrating species following the reaction of peroxynitrite with CO2 (e.g ONOO-C(O)O–, CO3 , and NO2)

Abbreviations

Hb, hemoglobin; HbI, haemoglobin I; HPX, hemopexin; Lb, leghemoglobin; Mb, myoglobin; Ngb, neuroglobin.

together at a 90 angle; the face of the N-terminal

b-propeller domain packs against one edge of the

C-terminal domain Each propeller blade comprises a

four-stranded antiparallel b-sheet, with the first and

the fourth blades joined by disulfide bridges The heme

is bound between the two four-bladed b-propeller

domains in a pocket formed by the interdomain linker

peptide Residues His213 and His266 coordinate the

heme iron atom giving a stable bis-histidyl complex

(Fig 1) Heme binding and release results from the

opening and closing of the heme-binding pocket, via

movement of the two b-propeller domains and⁄ or the

interdomain linker peptide [16]

Evidence suggests that heme-bound plasma proteins

may display ligand binding (kinetic and

thermody-namic) capacity and pseudo-enzymatic properties

Fer-rous high- and low-density heme–lipoproteins bind NO

[17] Ferrous heme–serum albumin has been reported

to bind NO, O2 and CO [14,18–21], and to exhibit

weak catalase and peroxidase activity [22] HPX– heme(II) binds CO and NO, however, O2 induces HPX–heme(II) oxidation [17,23–27] Furthermore, HPX–heme(III) binds cyanide [16] NO appears to modulate heme binding to HPX–heme and, in turn, HPX–heme may play a key role in NO homeostasis [17,25–27] Indeed, O2 has been reported to react with HPX–heme(II)–NO yielding HPX–heme(III) and

NO3 , by way of the ferric heme-bound peroxynitrite intermediate HPX–heme(III)–N(O)OO Afterwards, peroxynitrite dissociates from HPX–heme(III)– N(O)OO and isomerizes to nitrate HPX–heme(III) may reduce back to HPX–heme(II) and bind heme lig-ands [27] The recommended IUPAC nomenclature for peroxynitrite is oxoperoxonitrate; for peroxynitrous acid, it is hydrogen oxoperoxonitrate The term per-oxynitrite is used in the text to refer generically to both ONOO)and its conjugate acid HOONO [28] Here, the kinetics of reductive nitrosylation of HPX–heme(III) and peroxynitrite-mediated oxidation

of HPX–heme(II)–NO are reported HPX–heme(II)–

NO appears to act as an efficient scavenger of peroxy-nitrite and of strong oxidants and nitrating species following the reaction of peroxynitrite with CO2 [e.g ONOOC(O)O–, CO3 and NO2) Our results have been analyzed in parallel with those of related heme–protein systems

Results and Discussion

Reductive nitrosylation of HPX–heme(III) Addition of NO (either gaseous or dissolved in the buffer solution) to the HPX–heme(III) solution causes

a shift in the maximum of the optical absorption spec-trum in the Soret band from 414 nm, i.e HPX– heme(III), to 419 nm, i.e HPX–heme(II)–NO, and a corresponding change in the extinction coefficient from

e414¼ 1.16 · 105m)1Æcm)1to e419¼ 1.45 · 105m)1Æcm)1 (Fig 2A) The optical absorption spectrum of the reac-tion product HPX–heme(II)–NO (Fig 2A) is identical

to that obtained by adding NO to HPX–heme(II) [26,27] Accordingly, optical absorption spectroscopic changes for the reaction of HPX–heme(III) with NO were not reversible Pumping off gaseous NO causes a shift in the maximum of the optical absorption spec-trum in the Soret band from 419 nm, i.e HPX– heme(II)–NO, to 428 nm, i.e HPX–heme(II), and a corresponding change in the extinction coefficient from

e419¼ 1.45 · 105m)1Æcm)1to e428¼ 1.47 · 105m)1Æcm)1 The optical absorption spectra of HPX–heme(III), HPX–heme(II)–NO and HPX–heme(II) here deter-mined correspond to those reported in the literature

Fig 1 Rabbit ferric HPX–heme structure, including the coordinating

heme-iron residues His213 and His266 (PDB entry: 1QJS) [16] The

N-terminal domain (residues 1–208) is shown at the top The

C-ter-minal domain (residues 228–435) is shown at the bottom The

arrow indicates the interdomain linker peptide (residues 209–227).

Heme group and His213 and His266 residues are shown in black.

The figure was drawn with SWISS - PDB - VIEWER [75] For details, see

text.

[24,26,27,29] As already reported for Glycine max

leg-hemoglobin (Lb) [30], sperm whale myoglobin (Mb)

[31], horse cytochrome c [31], human neuroglobin

(Ngb) [32], Scapharca inaequivalvis hemoglobin I (HbI) [33] and human hemoglobin (Hb) [31], our data indi-cate that the reaction of HPX–heme(III) with an excess

NO leads to reduction of the heme–Fe(III) atom and generation of the HPX–heme(II)–NO species (Table 1) Over the whole NO concentration range explored, the time course of reductive nitrosylation of HPX– heme(III) (2.6· 10)6m) conforms to a single-exponen-tial decay for > 90% of its course between 360 and

460 nm (Fig 2B) The pseudo-first-order rate constant for HPX–heme(III) reductive nitrosylation (i.e k) is wavelength independent The plot of k versus [NO] is linear (Eqn 2) with a y-intercept at 0, indicating that the reverse reaction rate is negligible (koff< 1·

10)4s)1); the slope of the plot of k versus [NO] corres-ponds to kon¼ (1.3 ± 0.1) · 101m)1Æs)1 (Fig 2C, Table 1) The first-order rate constant for HPX– heme(II)–NO+ conversion to HPX–heme(II)* (i.e h) must exceed by at least one order of magnitude the value of k (6.5 · 10)3s)1) obtained at the highest

NO concentration investigated (5.0· 10)4m), i.e

h> 7· 10)2s)1 (Scheme 1, Fig 2C, Table 1), other-wise a hyperbolic plot of k versus [NO] would be observed [31,34,35]

Values of kon for reductive nitrosylation of HPX– heme(III), horse cytochrome c(III) [31,36] and human Ngb(III) [32] are lower than those reported for G max

2Lb(III) [30], sperm whale Mb(III) [31,36], S inaequi-valvis HbI(III) [33] and human Hb(III) [31], possibly reflecting heme–Fe(III) atom hexa-coordination [16,37, 38] (Table 1) Values of koff for NO dissociation from the heme(III)–NO proteins considered range between

< 1· 10)4 and 1.4· 101s)1 (Table 1), reflecting the different stability of the heme–Fe(III)–NO complexes [30–33,36] Values of h for NO+ dissociation from HPX–heme(II)–NO+, human Ngb(II)–NO+ [32] and

S inaequivalvis HbI(III) [33] are larger than those reported for reductive nitrosylation of G max Lb(III) [30], sperm whale Mb(III) [31,36], horse cytochrome c [31,36] and human Hb(III) [31] (Table 1)

Values of k for reductive nitrosylation of HPX– heme(III) (Fig 2C), human Ngb(III) [32], G max Lb(III) [30] and S inaequivalvis HbI [33] depend line-arly on NO concentration over the whole range explored (i.e between 5.0· 10)5and 1.2· 10)3m) In contrast, values of k for reductive nitrosylation of sperm whale Mb(III) [31], horse cytochrome c [31] and human Hb(III) [31] do not increase linearly with the NO concentration but tend to level off at [NO] > 2· 10)5m The transient heme(III)–NO spe-cies was observed during reductive nitrosylation

of G max Lb(III) [30], sperm whale Mb(III) [31], horse cytochrome c(III) [31] and human Hb(III) [31]

Fig 2 Kinetics of NO-mediated reductive nitrosylation of HPX–

heme(III), at pH 7.0 and 10.0 C (A) Steady-state and kinetic

differ-ence absorption spectra (line and circles, respectively) in the Soret

region of HPX–heme(III) minus HPX–heme(II)–NO (B) Time course

of reductive nitrosylation of HPX–heme(III), k ¼ 420 nm The NO

concentration was 1.0 · 10)4M (trace a), 2.0 · 10)4M (trace b),

and 5.0 · 10)4M (trace c) The time course analysis according to

Eqn (1) allowed us to determine the following values of k:

1.2 · 10)3s)1(trace a), 2.8 · 10)3s)1(trace b), and 6.7 · 10)3s)1

(trace c) (C) Dependence of the pseudo-first-order rate constant for

reductive nitrosylation of HPX–heme(III) (i.e k) on the NO

concen-tration The continuous line was generated from Eqn (2) with

k on ¼ (1.3 ± 0.1) · 10 1

M )1Æs)1 The HPX–heme(III) concentration

was 2.6 · 10)6M

only Furthermore, the intermediate species HPX–

heme(III)*, HPX–heme(III)–NO, HPX–heme(II)–

NO+, HPX–heme(II)* and HPX–heme(II) (Scheme 1)

were not detected This suggests the following

consid-erations (a) Formation of HPX–heme(III)* does not

appear to be rate limiting, as observed for reductive

nitrosylation of Ngb(III) [32] and horse cytochrome c

[31] (b) HPX–heme(III)–NO hydrolyzes very rapidly

under neutral conditions, similarly to human Ngb(III)–

NO [32] and S inaequivalvis HbI [33] and human

Hb(III)–NO [31] By contrast, G max Lb(III)–NO

[30], sperm whale Mb(III)–NO [31] and horse

cyto-chrome c(III)–NO [31] are rather stable under neutral

conditions and are hydrolyzed at a significant extent

under alkaline conditions (c) HPX–heme(II)

nitrosyla-tion [27] is faster than the NO-induced reducnitrosyla-tion of

the heme–Fe(III) atom, as observed for G max Lb(II)

[30,39], sperm whale Mb(II) [31,40], horse

cyto-chrome c(II) [31,36], mouse Ngb(II) [32,41] (highly

homologous to human Ngb [38,42,43]), S inaequivalvis

HbI(II) [33,44] and human Hb(II) [31,39,40,45]

How-ever, nitrosylation of hexa-coordinate HPX–heme(II),

horse cytochrome c(II) [36] and mouse Ngb(II) [41]

(highly homologous to human Ngb [38,42,43]) is

slower than NO binding to penta-coordinate G max

Lb(II) [39], sperm whale Mb(II) [40], S inaequivalvis

HbI(II) [44] and human Hb(II) [39,40,45,46] (Table 1)

As a whole, NO binding to HPX–heme(III), human

Ngb(III) [32] and S inaequivalvis HbI(II) [33] appears

to be rate limiting (i.e k < h; Scheme 1), whereas the

conversion of heme(II)–NO+ to heme(II) is rate

limit-ing for reductive nitrosylation of G max Lb(III) [30],

sperm whale Mb(III) [31], horse cytochrome c [31] and

human Hb(III) [31] (i.e k > h, Scheme 1)

Effect of CO2on peroxynitrite-mediated oxidation

of HPX–heme(II)–NO

In the absence and presence of CO2, mixing of the HPX–heme(II)–NO and peroxynitrite solutions causes

a shift in the optical absorption maximum of the Soret band from 419 nm, i.e HPX–heme(II)–NO, to 420 nm, i.e HPX–heme(III)–NO, and a corresponding change

in the extinction coefficient from e419¼ 1.45 · 105

to e420¼ 1.59 · 105m)1Æcm)1 (Fig 3A) The HPX– heme(III)–NO solution undergoes a shift in the optical absorption maximum of the Soret band from 420 nm, i.e HPX–heme(III)–NO, to 414 nm, i.e HPX– heme(III), and a corresponding change in the extinction coefficient from e420¼ 1.59 · 105m)1Æcm)1 to e414¼ 1.16· 105m)1Æcm)1 (Fig 3A) The optical absorption spectrum of HPX–heme(III) obtained by mixing the HPX–heme(II)–NO and peroxynitrite solutions (e414¼ 1.16· 105m)1Æcm)1) (Fig 3A) corresponds to that reported in the literature [24,26,27] Values for the opti-cal absorption maximum and extinction coefficient of HPX–heme(II)–NO, HPX–heme(III)–NO and HPX– heme(III) are unaffected by CO2 Analogous to G max

Lb [30], horse Mb [47], human Ngb [32] and human

Hb [28,47], our data indicate that the reaction of HPX– heme(II)–NO with an excess peroxynitrite leads to oxi-dation of the heme–Fe(II) atom and generation of the HPX–heme(III) species

Over the whole peroxynitrite concentration range explored (1.5· 10)5)2.5 · 10)4m), the time course for the peroxynitrite-mediated oxidation of HPX– heme(II)–NO (2.6· 10)6m) corresponds to a bipha-sic process, in the absence and presence of CO2 (1.2· 10)3m) (Fig 3B) Values of the pseudo-first-order

Table 1 Kinetic parameters for reductive nitrosylation of ferric heme-proteins (for details, see Scheme 1) ND, not determined.

5

off (s)1) h (s)1) l on ( M )1Æs)1) l

off (s)1)

Human Ngb(III)

Horse cytochrome c 7.2 · 10 2g

Human Hb(III)

2.2 · 10)5l

a pH 7.0 and 10.0 C; this study b pH 7.0 and 10.0 C [27] c pH 7.0 and room temperature [32] d pH 7.0 and 25.0 C [41] e pH 7.5 and 20.0 C [33] f

pH 7.0 and 20.0 C [44] g

pH 6.5 and 20.0 C [36] h

pH < 8.3 and 20.0 C [31] i

pH 7.0 and 20.0 C [74] j

pH 7.0 and 20.0 C [31] k pH 7.0 and 20.0 C [45] l pH 7.0 and 20.0 C [39] m pH 7.0 and 20.0 C [30] n pH 7.0 and 20.0 C [40].

rate constant for the formation of and the first-order

decay of the transient HPX–heme(III)–NO species (i.e b

and d, respectively) are wavelength independent

As shown in Fig 3C, the first step of kinetics for

per-oxynitrite-mediated oxidation of HPX–heme(II)–NO

(bonin Scheme 2) is a bimolecular process as observed

under pseudo-first-order conditions The plot of b

versus [peroxynitrite] is linear (Eqn 6) with a y-intercept

at 0, indicating that the reverse reaction rate is

negli-gible; the slope of the plot of b versus [peroxynitrite]

corresponds to bon¼ (8.6 ± 0.8) · 104m)1Æs)1 and

(1.2 ± 0.2)· 106m)1Æs)1 in the absence and presence

of CO2, respectively (Table 2) By contrast, the second

step (d in Scheme 2) follows a

peroxynitrite-independ-ent monomolecular behavior (Fig 3D), the average

value of d is (4.3 ± 0.4)· 10)1s)1(Table 2)

Fig 3 Kinetics of peroxynitrite-mediated oxidation of HPX–heme(II)–NO in the absence and presence of CO2, at pH 7.0 and 10.0 C (A) Steady-state and kinetic difference absorption spectra (line and symbols, respectively) in the Soret region of HPX–heme(II)–NO minus HPX– heme(III) (line and triangles) and fully populated HPX–heme(III)–NO minus HPX–heme(III) (squares) Steady-state and kinetic difference absorption spectra were independent of CO2 (B) Time course of the peroxynitrite-induced conversion of HPX–heme(II)–NO to HPX–heme(III)

by way transient HPX–heme(III)–NO formation, in the absence (trace a) and presence (trace b) of CO 2 , k ¼ 425 nm The time course analysis according to Eqns (3–5) allowed us to determine the following parameters: b ¼ 4.5 s)1 and d ¼ 4.3 · 10)1s)1 (trace a), and

b ¼ 5.9 · 10 1 s)1and d ¼ 4.5 · 10)1s)1(trace b) The peroxynitrite concentration was 5.0 · 10)5M (C) Dependence of the pseudo-first order rate constant for the peroxynitrite-induced conversion of HPX–heme(II)–NO to HPX–heme(III)–NO (i.e b) on the peroxynitrite concen-tration, in the absence (diamonds) and presence (squares) of CO 2 The continuous line was calculated according to Eqn (6) with

bon¼ (8.6 ± 0.9) · 10 4

M )1Æs)1in the absence of CO

2 (diamonds), and bon¼ (1.2 ± 0.2) · 10 6

M )1Æs)1in the presence of CO

2 (squares) (D) Dependence of the first order rate constant for NO dissociation from HPX–heme(III)–NO (i.e d) on the peroxynitrite concentration, in the absence (diamonds) and presence (squares) of CO 2 The average value of d is (4.3 ± 0.4) · 10)1s)1 The HPX–heme(II)–NO concentration was 2.6 · 10)6M The CO2concentration was 1.2 · 10)3M

Table 2 Kinetic parameters for peroxynitrite-mediated oxidation of ferrous nitrosylated heme-proteins (for details, see Scheme 2) Heme–protein [CO 2 ] ( M ) b on ( M )1Æs)1) d (s)1)

Rabbit HPX–heme(II)–NO a

1.2 · 10)3 1.2 · 10 6 4.3 · 10)1

1.2 · 10)3 5.3 · 10 4  1

G max Lb(II)–NOc

1.0 · 10)3 1.2 · 10 5

2.5 Horse Mb(II)–NO – d 3.1 · 10 4d  1.2 · 10 1d

1.2 · 10)3e 1.7 · 10 5e 1.1 · 10 1e

1.2 · 10)1

a pH 7.0 and 10.0 C; this study b pH 7.2 and 20.0 C [48] c pH 7.3 and 20.0 C [30] d pH 7.5 and 20.0 C [47] e pH 7.0 and 20.0 C [47] f pH 7.2 and 20.0 C [32].

As observed for HPX–heme(II)–NO (Fig 3C),

val-ues of the pseudo-first-order rate constant b (Scheme 2)

for peroxynitrite-mediated oxidation of G max Lb(II)–

NO [30], horse Mb [47], human Ngb(II)–NO [32] and

human Hb(II)–NO [28,47] depend linearly over the

whole peroxynitrite concentration range explored, in

the absence and presence of CO2 Values of bonfor

per-oxynitrite-mediated oxidation of HPX–heme(II)–NO

and human Ngb(II)–NO [32] exceed those reported for

G max Lb(II)–NO [30], horse Mb [47] and human

Hb(II)–NO [28,47] (Table 2) CO2 facilitates

peroxy-nitrite-mediated oxidation of HPX–heme(II)–NO

(Fig 3C), G max Lb(II)–NO [30], horse Mb [47] and

human Hb(II)–NO [28,47] increasing values of bon

(Table 2) A similar observation was made for the

reac-tion of G max Lb(II)–O2[48], sperm whale Mb(II)–O2

[49] and human Hb(II)–O2 [50] with peroxynitrite in

the absence and presence of CO2

In the presence of CO2, peroxynitrite changes from

a two- to a one-electron oxidant In fact, CO2 reacts

rapidly with peroxynitrite leading to ONOOC(O)O–

(second-order rate constant is  3 · 104m)1Æs)1),

which in turn decays very rapidly to CO3 and NO2

(first-order rate constant is  5 · 105 s)1) CO3 and

NO2 are stronger oxidant and nitrating agents than

peroxynitrite; NO2 nitrates with preference Tyr and

Trp residues [51,52] Although CO2 facilitates the

nitration of heme–protein aromatic residues by

peroxy-nitrite [53], optical absorbance spectroscopy for HPX

between 230 and 500 nm indicates that no appreciable

aromatic nitration takes place (data not shown) This

suggests that the CO2-induced increase in konfor

per-oxynitrite-mediated oxidation of HPX (

reflects oxidation of the heme–Fe atom by CO3 rather

than conformational transition(s) depending on the

nitration of Tyr and Trp residues by NO2

As observed for HPX–heme(III)–NO (Fig 3D),

val-ues of the first-order rate constant d (Scheme 2) for

NO dissociation from G max Lb(III)–NO [30], horse

Mb [47], human Ngb(III)–NO [32] and human

Hb(III)–NO [28,47] are unaffected by CO2, ranging

between 1· 10)1 and 1.2· 101s)1 (Table 2) The

dis-sociation of heme(III)–NO adducts is facilitated by the

consumption of NO via its reaction with peroxynitrite

excess and⁄ or with the reactive species generated

during peroxynitrite decomposition (e.g NO2) Under

anaerobic conditions, the reaction of NO with

peroxy-nitrite leads to N2O3 and H2O, at pH < 7 At

pH > 7, NO reacts with NO2 leading to N2O3, in

turn, N2O3 reacts with peroxynitrite leading to NO2

and NO2[28]

The transient species heme(III)–NO was observed

during peroxynitrite-mediated oxidation of HPX–

heme(II)–NO (Fig 3B) This intermediate is also seen

in G max Lb(II)–NO [30], horse Mb [47], human Ngb(II)–NO [32] and human Hb(II)–NO [28,47] By contrast, the transient penta-coordinate derivative of HPX–heme(III), i.e HPX–heme(III)*, (Scheme 2) was not observed This transient was also never observed in ligand-binding reaction(s) to human Ngb(III) [32] This suggests the following: (a) NO dissociation from heme(III)–NO represents the rate-limiting step for per-oxynitrite-mediated oxidation of the heme(II)–NO proteins considered [28,30,32,47]; and (b) the HPX– heme(III)*fi HPX–heme(III) reaction (Scheme 2) does not appear to be rate limiting, as reported for peroxy-nitrite-mediated oxidation of human Ngb(II)–NO [32] Under the experimental conditions, kinetic and spectroscopic properties of HPX–heme(II)–NO were unaffected by decomposed peroxynitrite

Conclusions

Our data represent the first evidence for reductive nitrosylation of HPX–heme(III) and for peroxyni-trite-mediated oxidation of HPX–heme(II)–NO As a general remark, the few data available from litera-ture concerning the reductive nitrosylation of ferric heme–proteins and the peroxynitrite-mediated oxida-tion of ferrous nitrosylated heme-proteins (in the absence and presence of CO2) are reported for the purpose of a comparison with those of HPX–heme (Tables 1,2)

Although HPX–heme(III) reduction and nitrosyla-tion occur physiologically and modulate HPX–heme complex (de-)stabilization [27,54–56], reductive nitrosy-lation of HPX–heme(III) appears too slow to occur

in vivo despite NO concentrations > 10)5m under pathological conditions [57–61] The same considera-tions hold also for reductive nitrosylation of human Ngb(III), sperm whale Mb(III), G max Lb(III) and human Hb(III) [30–32] However, as shown in Table 1, heme reduction kinetics are facilitated in bis-histidyl hexa-coordinate heme–proteins (i.e HPX–heme and human Ngb) [62]

The reactivity of peroxynitrite with HPX–heme(II)–

NO (Table 2) is high enough to protect against peroxynitrite-mediated damage, and to impair the for-mation of strong oxidants and nitrating agents (e.g ONOOC(O)O–, CO3 and NO2), in the absence and presence of CO2 [63] In fact, values of the second-order rate constant bonfor peroxynitrite-mediated oxi-dation of HPX–heme(II)–NO are larger than those for the reaction of peroxynitrite with (macro)molecular targets (e.g cysteine residues;  4 · 103m)1Æs)1) and with CO2( 3 · 104m)1Æs)1) [51,52] (Table 2)

The high value of the reaction rate of HPX–

heme(II)–NO with peroxynitrite (i.e bon; Table 2) may

reflect structural features reminiscent those of heme–

proteins endowed with (pseudo-)enzymatic activities

[64] Indeed, the imidazole ring of the proximal His266

residue of HPX–heme is eclipsed with respect to the

heme N–Fe–N coordination bonds [16], as observed

for the proximal His residue in horseradish peroxidase

[65], and in Alcaligenes eutrophus and Escherichia coli

flavohemoglobins [66,67], although His266 is rotated

by 90 with respect to heme propionates in HPX–

heme Furthermore, the negatively charged residue

Glu226 occurs in the neighborhood of the proximal

His266 residue of HPX–heme [16] Similarly,

horserad-ish peroxidase [65], and A eutrophus and E coli

flavo-hemoglobins [66,67] all display Asp or Glu residues in

the neighborhood of the proximal His In horseradish

peroxidase [65], and A eutrophus and E coli

flavoh-emoglobins [66,67], the proximal Asp⁄ Glu residue is

hydrogen-bonded to the proximal His ND1 atom,

partly setting the orientation of the proximal imidazole

[66,67] In HPX–heme the orientation of the proximal

imidazole is defined by a hydrogen bond connecting

the carbonyl O atom of Ser267 to the His266 ND1

atom, whereas the carboxylate of Glu226 falls at

 0.45 nm from the proximal imidazole ring [16]

Although the local stereochemistry makes unlikely the

achievement of a hydrogen bond between His266 and

Glu226 in HPX–heme(III), Glu226 may modulate

heme binding to HPX and HPX–heme ligand binding

capacity by affecting the protonation state of the

prox-imal His266 residue via electrostatic control of the

resi-due pKa[26]

In conclusion, our results describe a curious

situ-ation in which heme binding to a nonheme-protein

(i.e HPX) confers (although transiently) functional

properties (e.g peroxynitrite scavenging) and may be

predictive of (pseudo-enzymatic) function(s) of

heme-carriers (e.g heme–albumin as well as high and low

density heme–lipoproteins) The system studied here

may suggest that the effects arising from heme binding

to HPX might have some role in the regulation of

bio-logical functions Because these effects involve

tran-sient reactive functions, dependent on the interaction

with specific molecules (i.e the heme), they have been

called ‘chronosteric’ effects [68]

Experimental procedures

Chemicals

Hemin [iron(III)–protoporphyrin(IX)] was obtained from

Sigma Chemical Co (St Louis, MO) Gaseous NO was

purchased from Aldrich Chemical Co (Milwaukee, WI)

NO was purified by flowing through a NaOH column in order to remove acidic nitrogen oxides The NO stock solution was prepared by keeping in a closed vessel the 1.0· 10)1m phosphate buffer solution (pH 7.0) under purified NO, at 760.0 mmHg and 20.0C, anaerobically The solubility of NO in the aqueous buffered solution is 2.05· 10)3m, at 760.0 mmHg and 20.0C The NO stock solution was diluted with degassed 1.0· 10)1m phosphate buffer to reach the desired concentration [26,27,69] The 1.0· 10)1m phosphate buffer solution was kept under helium Peroxynitrite was prepared from

KO2 and NO and from HNO2 and H2O2, under anaer-obic conditions Peroxynitrite was purified by freeze frac-tionation The peroxynitrite concentration was determined

by measuring the optical absorbance at 302 nm (e302¼ 1.67· 103

m)1Æcm)1) The peroxynitrite stock solution was diluted with degassed 1.0· 10)2m NaOH to reach the desired concentration The 1.0· 10)2m NaOH solution was kept under helium Decomposed peroxynitrite was prepared by acidification of the peroxynitrite solution with HCl, then the solution was neutralized with 1.0· 10)1m NaOH [53,70,71] For the experiments car-ried out in the absence of CO2, the 1.0· 10)1m phos-phate buffer and the 1.0· 10)2m NaOH solutions were prepared fresh daily, thoroughly degassed, and kept under helium Experiments in the presence of CO2

(1.2· 10)3m) were carried out by adding to the protein solution the required amount from a freshly prepared 5.0· 10)1m sodium bicarbonate solution The CO2 con-centration is always expressed as the true concon-centration

in equilibrium with HCO3 The value of the constant

of the hydration–dehydration equilibrium

Hþþ HCO3 at pH 7.0 and 10.0C is 6.38 · 10)7 m The bicarbonate concentration present during the reac-tions was 9.5· 10)3m [28,30,32,47,53] All the other chemicals were obtained from Merck AG (Darmstadt, Germany) All products were of analytical or reagent grade and used without purification unless stated

HPX

Rabbit serum HPX was prepared as reported previously [25,72] Protein contaminants were < 3% of the HPX sam-ple as judged by gel electrophoresis and N-terminal amino acid sequence determination The HPX–heme(III) solution (2.0· 10)6)1.5 · 10)4m) was prepared by adding 1.2-molar excess of the HPX solution to the heme(III) solution (1.0· 10)1mphosphate buffer, pH 7.0), at 10.0C [26,27] Under these conditions, no free heme is present in solution [17,25,26,72] In fact, the value of the dissociation equilib-rium constant for heme binding to HPX is < 10)9m [25,72] The HPX–heme(II)–NO solution (2.6· 10)6m) was prepared by reductive nitrosylation of HPX– heme(III) under anaerobic conditions, i.e by adding to

HPX–heme(III) either gaseous NO or the buffered NO

solution (see below)

Reductive nitrosylation of HPX–heme(III)

The value of the second-order rate constant for reductive

nitrosylation of HPX–heme(III) (kon) was determined by

mixing the HPX–heme(III) solution (final concentration

2.6· 10)6m) with the NO solution (final concentration,

1.0· 10)4)5.0 · 10)4m) under anaerobic conditions, at

pH 7.0 (1.0· 10)1mphosphate buffer) and 10.0C [30–33]

No gaseous phase was present Kinetics was monitored

between 360 and 460 nm Under all experimental

condi-tions, final pH measured after mixing ranged always

between 6.9 and 7.1 For an homogeneous comparison

with the available functional data [26,27], kinetics was

obtained at 10.0C

Time courses were fitted to the minimum reaction

mech-anism represented by Scheme 1 [30–33], where HPX–heme*

indicates the transient penta-coordinate HPX–heme species

Values of the NO-dependent pseudo-first-order rate

con-stant for reductive nitrosylation of HPX–heme(III) (i.e k)

have been determined from data analysis, according to

Eqn (1) [30–33]:

½HPX–hemeðIIIÞt¼ ½HPX–hemeðIIIÞi ekt ð1Þ

The value of the second-order rate constant for reductive nitrosylation of HPX–heme(III) (i.e kon) was obtained from the linear dependence of k on the NO concentration (i.e [NO]) according to Eqn (2) [30–33]:

The difference optical absorption spectrum in the Soret region of HPX–heme(III) minus HPX–heme(II)–NO was obtained under steady-state conditions by subtracting the absorbance change in HPX–heme(II)–NO from that of HPX–heme(III)

The kinetic difference optical absorption spectrum in the Soret region of HPX–heme(III) minus HPX–heme(II)–NO was reconstructed from the difference optical absorption spectrum of HPX–heme(II)–NO minus HPX–heme(II)–

NO (De¼ 0.0 m)1Æcm)1) obtained under steady-state conditions plus the total absorbance changes of the HPX– heme(III) reductive nitrosylation process

Peroxynitrite-mediated oxidation of HPX–heme(II)–NO

Values of the second-order rate constant for peroxynitrite-mediated conversion of HPX–heme(II)–NO to HPX– heme(III)–NO (i.e bon) and of the first-order rate constant for NO dissociation from the HPX–heme(III)–NO complex (i.e for the formation of HPX–heme(III); d) were determined by rapid mixing the HPX–heme(II)–NO solution (final concentration 2.6· 10)6m) with the peroxy-nitrite solution (final concentration, 1.5· 10)5m to 2.5· 10)4m) under anaerobic conditions, at pH 7.0 (1.0· 10)1mphosphate buffer) and 10.0C, in the absence and presence of CO2(1.2· 10)3m) [28,30,32,47] The dead time of the SX18MV-R rapid-mixing stopped-flow appar-atus (Applied Photophysiscs Ltd, Leatherhead, UK) was 1.6 ms No gaseous phase was present Kinetics was monit-ored between 360 and 460 nm Under all the experimental conditions, the final pH value measured after mixing ranged always between 6.9 and 7.1 Kinetics was obtained at 10.0C in order to avoid loss of the initial part of the HPX–heme(II)–NOfi HPX–heme(III)–NO reaction especi-ally in the presence of CO2, and for an homogeneous com-parison with the available functional data [26,27]

The time courses were fitted to two consecutive mono-exponential processes according to the minimum reaction mechanism depicted in Scheme 2 [28,30,33,47]

Values of the (pseudo-)first-order rate constants for the formation of the HPX–heme(III)–NO complex (i.e b) and for NO dissociation from the transient HPX–heme(III)–NO complex (i.e for the formation of HPX–heme(III); d) have been determined from data analysis, according to Eqns (3– 5) [73]:

½HPX–heme(II)–NOt¼ ½HPX–heme(II)–NOi ebt ð3Þ Scheme 1.

ơHPXỜheme(III)ỜNOtỬ ơHPX-heme(II)ỜNOi

 đb  đđebt=đdbỡỡ

ợ đedt=đbdỡỡỡỡ đ4ỡ

ơHPXỜheme(III)tỬ ơHPXỜheme(II)ỜNOi

 đơHPXỜheme(II)ỜNOt

ợ ơHPXỜheme(III)ỜNOtỡ đ5ỡ The value of bonwas obtained from the linear dependence

of b on the peroxynitrite concentration (i.e [peroxynitrite])

according to Eqn (6) [28,30,32,47]:

bỬ bon ơperoxynitrite đ6ỡ The difference optical absorption spectrum in the Soret

region of HPXỜheme(II)ỜNO minus HPXỜheme(III) was

obtained under steady-state conditions by subtracting the

absorbance change of HPXỜheme(III) from that of HPXỜ

heme(II)ỜNO

The kinetic difference optical absorption spectra in the

Soret region of HPXỜheme(II)ỜNO minus HPXỜheme(III)

and of HPXỜheme(III)ỜNO minus HPXỜheme(III) were

reconstructed from the difference optical absorption

spec-trum of HPXỜheme(III) minus HPXỜheme(III) (DeỬ

0.0 m)1ẳcm)1) obtained under steady-state conditions plus

the absorbance changes of the overall process HPXỜ

heme(II)ỜNO + HOONOfi HPXỜheme(III) + NO and

of the partial reaction HPXỜheme(III)ỜNOfi HPXỜ

heme(III) + NO (Scheme 2)

The absolute optical absorption spectrum of HPXỜ

heme(III)ỜNO in the Soret region was reconstructed from

the optical absorption spectrum of HPXỜheme(III)

obtained under steady-state conditions plus the absorbance

changes of the partial reaction HPXỜheme(III)Ờ

NOfi HPXỜheme(III) + NO (Scheme 2)

Nitration of HPX Tyr and Trp residues was investigated

by optical absorbance spectroscopy [53] Briefly, the heme

was removed from HPXỜheme(III), obtained by reacting

HPXỜheme(II)ỜNO (1.5ở 10)4m) with peroxynitrite

(1.0ở 10)3m), by mixing the HPXỜheme(III) solution

(1.5ở 10)4m) with the cold ()20.0 C) acid acetone solu-tion (2Ờ3 mL of 2.0 m HCl per liter of acetone) One vol-ume of HPXỜheme(III) in water was added slowly with vigorous stirring to about 30 vol of the cold acid acetone solution The precipitated HPX was collected by centrifuga-tion and then dissolved in a minimum amount of water The dissolved HPX was dialyzed in the cold against a dilute bicarbonate solution (1.0ở 10)3m) and subsequently against 1.0ở 10)2mphosphate buffer at pH 7.0 The pro-tein precipitate was removed by centrifugation [69] Then, the optical absorbance spectrum of HPX ( 1 ở 10)4m) was recorded between 230 and 500 nm

Data analysis

All experiments were carried out at least in quadruplicate The results are given as mean values plus or minus the standard deviation All data were analyzed using matlab (The Math Works Inc., Natick, MA)

Acknowledgements

This work was partially supported by grants from the Ministry for Education, University, and Research of Italy (University ỔRoma TreỖ, Rome, Italy, ỔCLAR 2005Ỗ to PA) and from the Ministry for Health of Italy (National Institute for Infectious Diseases IRCCS

ỔỔLazzaro SpallanzaniỖỖ, Rome, Italy, ỔRicerca Corrente 2005Ỗ to PA)

References

1 Hunt RC, Handy I & Smith A (1996) Heme-mediated reactive oxygen species toxicity to retinal pigment epithelial cells is reduced by hemopexin J Cell Physiol

168, 81Ờ86

2 Hunt RC, Hunt DM, Gaur N & Smith A (1996) Hemo-pexin in the human retina: protection of the retina against heme-mediated toxicity J Cell Physiol 168, 71Ờ80

3 Peters T Jr, ed (1996) All About Albumin: Biochemistry, Genetics and Medical Applications Academic Press, San Diego, CA

4 Miller YI & Shaklai N (1999) Kinetics of hemin distri-bution in plasma reveals its role in lipoprotein oxida-tion Biochim Biophys Acta 1454, 153Ờ164

5 Conrad ME & Umbreit JN (2000) Iron absorption and transport Ờ an update Am J Hematol 64, 287Ờ 298

6 Delanghe JR & Langlois MR (2001) Hemopexin: a review of biological aspects and the role in laboratory medicine Clin Chim Acta 312, 13Ờ23

7 Conrad ME & Umbreit JN (2002) Pathways of iron absorption Blood Cell Mol Dis 29, 336Ờ355

Scheme 2.

8 Goswami T, Rolfs A & Hediger MA (2002) Iron

trans-port: emerging roles in health and disease Biochem Cell

Biol 80, 679–689

9 Mattu M, Fasano M, Spallarossa A, Bolognesi M &

Ascenzi P (2002) Hemopexin: the primary specific

car-rier of plasma heme Biochem Mol Biol Educ 30, 332–

335

10 Tolosano E & Altruda F (2002) Hemopexin: structure,

function, and regulation DNA Cell Biol 21, 297–306

11 Baker HM, Anderson BF & Baker EN (2003) Dealing

with iron: common structural principles in proteins that

transport iron and heme Proc Natl Acad Sci USA 100,

3579–3583

12 Ascenzi P, Bocedi A, Visca P, Altruda F, Tolosano E,

Beringhelli T & Fasano M (2005) Hemoglobin and

heme scavenging IUBMB Life 57, 749–759

13 Braun V (2005) Bacterial iron transport related to

viru-lence Contrib Microbiol 12, 210–233

14 Fasano M, Curry S, Terreno E, Galliano M, Fanali G,

Narciso P, Notari S & Ascenzi P (2005) The

extraordin-ary ligand-binding properties of human serum albumin

IUBMB Life 57, 787–796

15 Hvidberg V, Maniecki MB, Jacobsen C, Hojrup P,

Moller HJ & Moestrup SK (2005) Identification of the

receptor scavenging hemopexin–heme complexes Blood

106, 2572–2579

16 Paoli M, Anderson BF, Baker HM, Morgan WT,

Smith A & Baker EN (1999) Crystal structure of

hemo-pexin reveals a novel high-affinity heme site formed

between two b-propeller domains Nat Struct Biol 6,

926–931

17 Fasano M, Mattu M, Coletta M & Ascenzi P (2002)

The heme–iron geometry of ferrous nitrosylated heme–

serum lipoproteins, hemopexin, and albumin: a

compar-ative EPR study J Inorg Biochem 91, 487–490

18 Kharitonov VG, Sharma VS, Magde D & Koesling D

(1997) Kinetics of nitric oxide dissociation from

five-and six-coordinate nitrosyl hemes five-and heme proteins,

including soluble guanylate cyclase Biochemistry 36,

6814–6818

19 Baroni S, Mattu M, Vannini A, Cipollone R, Aime S,

Ascenzi P & Fasano M (2001) Effect of ibuprofen and

warfarin on the allosteric properties of haem–human

serum albumin A spectroscopic study Eur J Biochem

268, 6214–6220

20 Mattu M, Vannini A, Coletta M, Fasano M & Ascenzi

P (2001) Effect of bezafibrate and clofibrate on the

heme–iron geometry of ferrous nitrosylated heme–

human serum albumin: an EPR study J Inorg Biochem

84, 293–296

21 Komatsu T, Ohmichi N, Nakagawa A, Zunszain PA,

Curry S & Tsuchida E (2005) O2and CO binding

prop-erties of artificial hemoproteins formed by complexing

iron protoporphyrin IX with human serum albumin

mutants J Am Chem Soc 127, 15933–15942

22 Monzani E, Bonafe` B, Fallarini A, Redaelli C, Casella

L, Minchiotti L & Galliano M (2001) Enzymatic prop-erties of hemalbumin Biochim Biophys Acta 1547, 302– 312

23 Muller-Eberhard U & Grizzuti K (1971) Conforma-tional studies on human and rabbit plasma hemopexin and on the effect of ligands on the rabbit heme–hemo-pexin complex Biochemistry 10, 2062–2066

24 Shaklai N, Sharma VS, Muller-Eberhard U & Morgan

WT (1981) The interaction of heme–hemopexin with

CO J Biol Chem 256, 1544–1548

25 Shipulina N, Hunt RC, Shaklai N & Smith A (1998) Coordination of nitric oxide by heme–hemopexin

J Protein Chem 17, 255–260

26 Fasano M, Bocedi A, Mattu M, Coletta M & Ascenzi P (2004) Nitrosylation of rabbit ferrous heme–hemopexin

J Biol Inorg Chem 9, 800–806

27 Fasano M, Antonini G & Ascenzi P (2006) O2-Mediated oxidation of hemopexin–heme (II)–NO Biochem Biophys Res Commun 345, 704–712

28 Herold S (2004) The outer-sphere oxidation of nitrosyl-iron(II) hemoglobin by peroxynitrite leads to the release

of nitrogen monoxide Inorg Chem 43, 3783–3785

29 Shipulina N, Smith A & Morgan WT (2000) Heme binding by hemopexin: evidence for multiple modes of binding and functional implications J Protein Chem 19, 239–248

30 Herold S & Puppo A (2005) Kinetics and mechanistic studies of the reactions of metleghemoglobin, ferrylle-ghemoglobin, and nitrosylleghemoglobin with reactive nitrogen species J Biol Inorg Chem 10, 946–957

31 Hoshino M, Maeda M, Konishi R, Seki H & Ford PC (1996) Studies on the reaction mechanism for reductive nitrosylation of ferrihemoproteins in buffer solutions

J Am Chem Soc 118, 5702–5707

32 Herold S, Fago A, Weber RE, Dewilde S & Moens L (2004) Reactivity studies of the Fe(III) and Fe(II)NO forms of human neuroglobin reveal a potential role against oxidative stress J Biol Chem 279, 22841–22847

33 Boffi A, Sarti P, Amiconi G & Chiancone E (2002) The interplay between heme iron and protein sulfhydryls in the reaction of dimeric Scapharca inaequivalvis hemoglo-bin with nitric oxide Biophys Chem 98, 209–216

34 Coletta M, Angeletti M, De Sanctis G, Cerroni L, Giar-dina B, Amiconi G & Ascenzi P (1996) Kinetic evidence for the existence of a rate-limiting step in the reaction

of ferric hemoproteins with anionic ligands Eur J Bio-chem 235, 49–53

35 Couture M, Das TK, Lee HC, Peisach J, Rousseau DL, Wittenberg BA, Wittenberg JB & Guertin M (1999) Chlamydomonaschloroplast ferrous hemoglobin Heme pocket structure and reactions with ligands J Biol Chem

274, 6898–6910

36 Hoshino M, Oawa K, Seki H & Ford PC (1993) Photo-chemistry of nitric oxide adducts of water-soluble iron(III)