Direct observation of nitrosylated heme in myoglobin and hemoglobin by electrospray ionization mass spectrometry

Article abstract of DOI:10.1021/ja964249l

Using electrospray ionization mass spectrometry (ESI-MS), we demonstrate the direct observation of NO attached to the heme moiety in;horse heart myoglobin (Mb) and in the α- and β-chains of human hemoglobin (Hb). It was found that a narrow range of ESI-MS conditions conspire to make observation of Fe-NO interactions challenging, and this is presumably the reason why earlier attempts by other research groups to detect intact Fe-NO products by mass spectrometry were unsuccessful For Mb and Hb, mass shifts are observed that are consistent with NO modification of the hemoproteins. ESI mass spectra of the apoprotein portions of Mb and Hb in the presence of NO demonstrated the absence of NO modification of the polypeptide backbones. UV/vis spectra of both Mb/NO and Hb/NO solutions, recorded at the time of ESI-MS analysis, demonstrated hemoprotein(II)-NO formation. To test the hypothesis that intact nitrosylated heme groups are observable by ESI-MS, a nitrosylated model metalloporphyrin was studied. The ESI mass spectrum of nitrosyl-α,β,γ,δ-tetraphenylporphinatoiron(II), [Fe(TPP)NO], showed peaks that were ascribed to [Fe(TPP)]⁺ and [Fe(TPP)NO]⁺. To test further our hypothesis that the hemoprotein-NO peaks are due to heme nitrosylation and contain no significant contributions from NO modification of the polypeptide backbone, we determined the ESI-MS conditions necessary for observing S-nitrosation of Cys residues in Hb. Human Hb contains one Cys residue in Hb(α) (Cys 104) and two Cys residues in Hb(β), but only Hb(β) Cys 93 is surface accessible. When metHb was incubated with S-nitroso-N-acetyl-DL-penicillamine (SNAP), the ESI mass spectrum revealed a single SNAP modification in both Hb(β) and apoHb(β). The ESI-MS conditions used for analyzing the Hb/SNAP solution were too harsh for observing intact heme nitrosylation, and thus, we ascribe the SNAP-modified Hb(β) and apoHb(β) peaks to S-nitrosation of Cys 93 in Hb(β). Under appropriate denaturing sample conditions, it proved possible to S-nitrosate all three Cys residues in human apoHb. pur findings demonstrate that (once correct conditions are established) ESI-MS is a powerful tool for the detection of intact Fe-NO interactions in proteins and porphyrins.

Full text of DOI:10.1021/ja964249l

10424
J. Am. Chem. Soc. 1997, 119, 10424-10429
Direct Observation of Nitrosylated Heme in Myoglobin and
Hemoglobin by Electrospray Ionization Mass Spectrometry
,†
†
‡
‡,§
Rita K. Upmacis,* David P. Hajjar, Brian T. Chait, and Urooj A. Mirza
Contribution from the Department of Biochemistry, Cornell UniVersity Medical College,
New York, New York 10021, and Laboratory for Mass Spectrometry and Gaseous Ion Chemistry,
The Rockefeller UniVersity, New York, New York, 10021
X
ReceiVed December 10, 1996. ReVised Manuscript ReceiVed March 3, 1997
Abstract: Using electrospray ionization mass spectrometry (ESI-MS), we demonstrate the direct observation of NO
attached to the heme moiety in horse heart myoglobin (Mb) and in the R- and â-chains of human hemoglobin (Hb).
It was found that a narrow range of ESI-MS conditions conspire to make observation of Fe-NO interactions
challenging, and this is presumably the reason why earlier attempts by other research groups to detect intact Fe-NO
products by mass spectrometry were unsuccessful. For Mb and Hb, mass shifts are observed that are consistent with
NO modification of the hemoproteins. ESI mass spectra of the apoprotein portions of Mb and Hb in the presence
of NO demonstrated the absence of NO modification of the polypeptide backbones. UV/vis spectra of both Mb/NO
II
and Hb/NO solutions, recorded at the time of ESI-MS analysis, demonstrated hemoprotein -NO formation. To test
the hypothesis that intact nitrosylated heme groups are observable by ESI-MS, a nitrosylated model metalloporphyrin
was studied. The ESI mass spectrum of nitrosyl-R,â,γ,δ-tetraphenylporphinatoiron(II), [Fe(TPP)NO], showed peaks
+
+
that were ascribed to [Fe(TPP)] and [Fe(TPP)NO] . To test further our hypothesis that the hemoprotein-NO
peaks are due to heme nitrosylation and contain no significant contributions from NO modification of the polypeptide
backbone, we determined the ESI-MS conditions necessary for observing S-nitrosation of Cys residues in Hb. Human
Hb contains one Cys residue in HbR (Cys 104) and two Cys residues in Hbâ, but only Hbâ Cys 93 is surface accessible.
When metHb was incubated with S-nitroso-N-acetyl-DL-penicillamine (SNAP), the ESI mass spectrum revealed a
single SNAP modification in both Hbâ and apoHbâ. The ESI-MS conditions used for analyzing the Hb/SNAP solution
were too harsh for observing intact heme nitrosylation, and thus, we ascribe the SNAP-modified Hbâ and apoHbâ
peaks to S-nitrosation of Cys 93 in Hbâ. Under appropriate denaturing sample conditions, it proved possible to
S-nitrosate all three Cys residues in human apoHb. Our findings demonstrate that (once correct conditions are
established) ESI-MS is a powerful tool for the detection of intact Fe-NO interactions in proteins and porphyrins.
Introduction
number of techniques have been developed to measure heme
nitrosylation and/or S-nitrosation. These include chemilumi-
Nitric oxide (NO) has been implicated to play an important
role in a range of biological processes although the mechanisms
of its action are unclear.1 Nevertheless, it has long been known
that one specific target of NO modification is the heme moiety
8
10
nescence, X-ray crystallography, and spectroscopies such as
UV/vis, infrared, and electron spin resonance.13 However,
none of these techniques (with the possible exception of X-ray
crystallography) can normally be used to measure directly both
of these modifications, while simultaneously providing informa-
tion regarding the molecular mass of the modified intact protein.
Electrospray ionization mass spectrometry (ESI-MS) is a
gentle method for transferring protein ions from solution into
the gas phase and for accurately determining their molecular
1
1
12
2-4
of hemoproteins (e.g., myoglobin and hemoglobin).
Heme
5
nitrosylation results in guanylate cyclase activation and may
lead to nitric oxide synthase inhibition.6 NO may also exert its
biological activity by the formation of S-nitrosothiols upon
7
reaction of a thiol with NOx. Certain hemoproteins, such as
8
9
hemoglobin and prostaglandin H2 synthase, have the potential
14
masses. A unique feature of ESI-MS is its ability for detecting
noncovalent interactions between proteins and molecular part-
for undergoing both heme nitrosylation and S-nitrosation. A
1
5
ners, as well as determining highly labile modifications of
*
Author to whom correspondence should be addressed at the follow-
1
6
ing: Department of Biochemistry, Room A626, Cornell University Medical
College, 1300 York Avenue, New York, New York 10021; phone (212)
proteins. We have previously reported the use of ESI-MS to
monitor S-nitrosated cysteine derivatives of peptides and
7
46-6470, fax (212) 746-8789, e-mail rupmacis@mail.med.cornell.edu.
§
Current address: Schering-Plough Research Institute, 2015 Galloping
(9) Hajjar, D. P.; Lander, H. M.; Pearce, S. F.; Upmacis, R. K.;
Pomerantz, K. B. J. Am. Chem. Soc. 1995, 117, 3340-3346.
(10) Edwards, S. L.; Kraut, J.; Poulos, T. L. Biochemistry 1988, 27,
8074-8081.
(11) Antonini, E.; Brunori, M. Hemoglobin and Myoglobin in Their
Reactions with Ligands; North-Holland Publishing Company: Amsterdam,
London, 1971; p 19.
(12) Sampath, V.; Zhao, X.-J.; Caughey, W. S. Biochem. Biophys. Res.
Commun. 1994, 198, 281-287.
(13) Dickinson, L. C.; Chien, J. C. W. J. Am. Chem. Soc. 1971, 93,
5036-5040.
(14) (a) Fenn, J. B.; Mann, C. K.; Meng, S. F.; Whitehouse, C. M. Science
1989, 246, 64-71. (b) Smith, R. D.; Loo, J. A.; Edmonds, C. G.; Baringa,
C. J.; Usdeth, H. R. Anal. Chem. 1990, 62, 882-899. (c) Chait, B. T.;
Kent, S. B. Science 1992, 257, 1885-1894.
(15) Smith, R. D.; Light-Wahl, K. J. Biol. Mass Spectrom. 1993, 22,
493-501 and references therein.
Hill Road, Kenilworth, NJ 07033.
†
Cornell University Medical College.
The Rockefeller University.
Abstract published in AdVance ACS Abstracts, October 1, 1997.
‡
X
(
1) Moncada, S.; Palmer, R. M.; Higgs, E. A. Pharmacol. ReV. 1991,
4
3, 109-142.
(2) Fox, J. B.; Thomson, J. S. Biochemistry 1963, 2, 465-470.
(3) Gibson, Q. H.; Roughton, F. J. W. J. Physiol. 1957, 136, 507-526.
(4) Antonini, E.; Brunori, M.; Wyman, J.; Noble, R. J. Biol. Chem. 1966,
2
41, 3236-3238.
(
5) Ignarro, L. J. Biochem. Soc. Trans. 1992, 20, 465-469.
(6) Wang, J.; Rousseau, D. L.; Abu-Soud, H. M.; Stuehr, D. J. Proc.
Natl. Acad. Sci. U.S.A. 1994, 91, 10512-10516.
(
7) Stamler, J. S. Cell 1994, 78, 931-93.
(8) Jia, L.; Bonaventura, C.; Bonaventura, J.; Stamler, J. S. Nature 1996,
3
80, 221-226.
S0002-7863(96)04249-7 CCC: $14.00 © 1997 American Chemical Society

Nitrosylated Heme in Myoglobin and Hemoglobin
J. Am. Chem. Soc., Vol. 119, No. 43, 1997 10425
treatment with NO. Two peaks are observed, corresponding
to (i) native Mb (measured molecular mass, Mr ) 17 568 ( 2
Da, calculated Mr ) 17 568 Da) and (ii) Mb that has undergone
loss of the heme group during the MS measurement (apoMb;
measured Mr ) 16 950 ( 2 Da, calculated Mr ) 16 951 Da).
The ESI mass spectrum in Figure 1a was obtained using a
2
1
transport capillary temperature of 125 °C and a declustering
22
voltage (the voltage difference between the transport capillary
and tube lens) of 80 V. Under these ESI-MS conditions, Mb
is the dominant species and very little apoMb is observed.
Figure 1b shows the deconvoluted ESI mass spectrum of Mb
after incubation with NO under anaerobic conditions (i.e.,
conditions under which NO is expected to react exclusively with
the heme group of hemoproteins). The ESI-MS conditions are
exactly identical to those used in Figure 1a. Again, two peaks
that correspond to Mb and apoMb are observed, although with
much different intensities to those seen in Figure 1a. No peak
due to NO-modified Mb was observed under these ESI-MS
conditions, even though UV/vis spectral analysis of the Mb/
NO solution indicated complete conversion to the heme-
nitrosylated form of Mb (Vide infra). Thus, we deduce that the
presence of both unmodified apoMb and Mb in the mass
spectrum shown in Figure 1b is the result of the decomposition
of the heme-nitrosylated Mb during the ESI-MS measurement.
In contrast to Figure 1a, the dominant species in Figure 1b is
apoMb. This increased level of apoMb in Figure 1b provides
evidence that the heme NO group in NO-modified Mb has a
destabilizing effect on the heme-globin complex.
Figure 1. Comparison of deconvoluted ESI mass spectra of (a) Mb in
the absence of NO with a transport capillary temperature of 125 °C,
Figure 1c shows details of the deconvoluted ESI mass
spectrum of the same Mb/NO solution used in Figure 1b, but
under gentler ESI-MS conditions. A control ESI mass spectrum
of Mb in the absence of NO under these new ESI-MS conditions
showed a very similar spectrum to that given in Figure 1a (data
not shown). Specifically, the transport capillary temperature
was lowered by 10 °C from 125 °C (Figures 1a and 1b) to 115
(b) Mb in the presence of NO with a transport capillary temperature of
1
25 °C, and (c) Mb in the presence of NO with a transport capillary
temperature of 115 °C. Sample preparation conditions are given in the
Experimental Section. The ESI-MS conditions: (a) declustering voltage
8
0 V, scan range 1800-2600 m/z, sample infusion rate 5 µL/min,
number of scans 5 (3 s/scan); (b) declustering voltage 80 V, scan range
800-2600 m/z, sample infusion rate 5 µL/min, number of scans 5 (3
1
°
C (Figure 1c), while a declustering voltage of 80 V was
s/scan); (c) declustering voltage 80 V, scan range 2000-2300 m/z,
sample infusion rate 5 µL/min, number of scans 20 (3 s/scan). The
asterisk in (a) denotes a potassium adduct of Mb.
maintained. Under these slightly gentler conditions, three peaks
are observed for the Mb/NO solution, i.e., the peaks ascribed
to apoMb, native Mb, and a new peak at 17 599 Da. This latter
peak is 31 ( 2 Da higher than the measured Mr of native Mb,
consistent with NO modification of the protein. In addition to
the observation of NO-modified Mb, the slightly gentler
conditions used in Figure 1c lead to a lower production of
apoMb than is observed in Figure 1b. These observations
provide further evidence that the apoMb peak is to a large extent
formed in the mass spectrometer and that heme nitrosylation
destabilizes the heme globin interaction. It is important to notice
that there is no evidence of a peak in Figure 1c that can be
attributed to NO-modified apoMb, providing evidence that the
polypeptide backbone is not a site of NO modification.
proteins,16,17 and more recently, others have reported the
1
8
characterization of S-nitrosohemoglobin by ESI-MS. How-
ever, earlier attempts by others to detect heme-NO adducts in
proteins by ESI-MS proved unsuccessful.19 Here, we demon-
strate, for the first time, the direct ESI-MS observation of NO
attached to the heme moiety in myoglobin (Mb) and in the R-
and â-chains of hemoglobin (Hb). In addition, we demonstrate
that ESI-MS can be used to detect S-nitrosation of Hb. By
varying the ESI-MS conditions employed, it is possible to
distinguish between heme nitrosylation (which can only be
observed using extremely gentle ESI-MS conditions) and
S-nitrosation (which still requires moderately gentle ESI-MS
conditions).
Heme Nitrosylation of Hemoglobin. Figure 2b shows a
portion of the deconvoluted ESI mass spectrum of Hb after
incubation with NO under anaerobic conditions (i.e., under
conditions that are expected to nitrosylate the heme group, but
leave the globin thiols unmodified). Since it was not possible
to measure the ESI mass spectrum of the intact Hb tetramer,
we instead monitored those regions containing peaks corre-
sponding to the R- and â-chain of Hb. The portion shown in
Figure 2b includes the region encompassing the R-chain of Hb
Results
Heme Nitrosylation of Myoglobin. Figure 1a shows the
2
0
deconvoluted ESI mass spectrum of native Mb prior to
(
16) Mirza, U. A.; Chait, B. T.; Lander, H. M. J. Biol. Chem. 1995,
70, 17185-17188.
17) Lander, H. M.; Millbank, A. J.; Tauras, J. M.; Hajjar, D. P.;
2
(
(HbR; calculated Mr ) 15 743 Da). Under ESI-MS conditions
Hempstead, B. L.; Schwartz, G. D.; Kraemer, R. T.; Mirza, U. A.; Chait,
in which the transport capillary temperature and declustering
voltage were optimized for the observation of intact NO-
B. T.; Campbell, B.; Quilliam, L. A. Nature 1996, 381, 380-381.
(18) Ferranti, P.; Malorni, A.; Mamone, G.; Sannolo, N.; Marino, G.
FEBS Lett. 1997, 400, 19-24.
19) Osipov, A. N.; Gorbunov, N. V.; Day, B. W.; Elsayed, N. M.; Kagan,
V. E. Methods Enzymol. 1996, 268, 193-203.
20) Mann, M.; Meng, C. K; Fenn, J. B. Anal. Chem. 1989, 61, 1702-
708.
(
(21) Chowdry, S. K.; Katta, V.; Chait, B. T. Rapid Commun. Mass
Spectrom. 1990, 4, 81-87.
(22) Loo, J. A.; Udseth, H. R.; Smith, R. D. Rapid Commun. Mass
Spectrom. 1988, 2, 207-210.
(
1

10426 J. Am. Chem. Soc., Vol. 119, No. 43, 1997
Upmacis et al.
Figure 4. Details of the deconvoluted ESI mass spectrum of Hb in
the presence of NO, showing the portion of the spectrum containing
the Hb peak. Transport capillary temperature 126 °C, declustering
â
voltage 116 V, scan range 2040-2100 m/z, sample infusion rate 3 µL/
min, number of scans 124 (3 s/scan).
observed, which indicates that there is no obvious modification
of the HbR polypeptide backbone and, hence, provides evidence
that HbR Cys 104 has not undergone S-nitrosation under these
conditions. It may be argued that the excitation introduced
during the ESI process that causes loss of the heme group from
HbR may also be sufficient to cause dissociation of NO from
any S-nitrosated thiols that may be present in the protonated
protein. However, we shall demonstrate below that dissociation
of S-nitrosated thiols does not occur under the very gentle ESI
conditions that were used to obtain the spectrum in Figure 3.
Figure 4 displays the portion of the deconvoluted ESI mass
spectrum encompassing the â-chain of Hb (Hbâ), demonstrating
NO modification of Hbâ (even though the peak corresponding
to Hbâ is of very low intensity). The poor response of Hbâ
compared to HbR has been reported previously but has not yet
Figure 2. Comparison of deconvoluted ESI mass spectra of Hb in the
presence of NO with a transport capillary temperature of (a) 135 °C
versus (b) 125 °C. The portion of the spectrum containing the Hb
R
peak is shown. The ESI-MS conditions: (a) declustering voltage 33
V, scan range 1900-2100 m/z, sample infusion rate 5 µL/min, number
of scans 25 (3 s/scan); (b) declustering voltage 33 V, scan range 1900-
2
100 m/z, sample infusion rate 3 µL/min, number of scans 16 (3 s/scan).
The x and y denote, respectively, an unidentified peak and a weak peak
due to apoHb
â
.
2
3
been explained. The ESI-MS conditions used to obtain the
spectrum in Figure 4 are different to those employed for the
observation of NO-modified HbR (Figure 2b). More specifi-
cally, it proved necessary to increase the declustering voltage
from 33 V (Figure 2b) to 116 V (Figure 4) in order to obtain
discernible peaks. The two most intense peaks in Figure 4 are
observed at 16 485 ( 6 and 16 513 ( 6 Da, and likely
correspond to Hbâ (calculated Mr ) 16 484 Da) and NO-
modified Hbâ, respectively. A weak spectrum of apoHbâ was
also obtained in which no evidence of NO modification could
be discerned (data not shown). We thus conclude that, under
anaerobic conditions, the NO modification of Hbâ occurred
exclusively on the heme group.
Figure 3. Deconvoluted ESI mass spectrum of Hb in the presence of
NO, showing the region of the spectrum containing the apoHb peak.
R
Transport capillary temperature 125 °C, declustering voltage 33 V, scan
range 1800-2200 m/z, sample infusion rate 5 µL/min, number of scans
UV/Vis Spectroscopy of Hemoprotein/NO Solutions. UV/
vis spectroscopy was used to monitor heme nitrosylation of both
the Mb/NO and Hb/NO solutions prior to ESI-MS analysis.
Figure 5 shows that in the presence of NO under anaerobic
conditions, the characteristic absorption maxima for metmyo-
1
6 (3 s/scan).
modified HbR (see figure legend), two dominant peaks are
observed (Figure 2b). The first peak corresponds to HbR
(
2
measured Mr ) 15 746 ( 4 Da), while the second has a mass
III
9 ( 2 Da greater than that observed for HbR, indicative of
globin (Mb ) (410 and 500 nm) disappeared within 30 min,
III
NO modification of HbR. Comparison of Figures 2b and 2a
demonstrate that an increase of the transport capillary temper-
ature by as little as 10 °C leads to almost complete loss of the
peak due to NO-modified HbR, a situation analogous to that
observed for NO-modified Mb (Vide supra).
Figure 3 displays a different portion of the deconvoluted ESI
mass spectrum of NO-treated Hb shown in Figure 2b. The
portion shown in Figure 3 includes the region encompassing
HbR after loss of the heme group during the MS measurement
and new bands corresponding to an intermediate Mb -NO
derivative (416, 539, and 575 nm) were formed. Within 2 h,
III
II
the spectrum of Mb -NO gradually changed to that of Mb -
2,11,24
NO (420, 547, and 577 nm).
Similarly, the UV/vis spectra
III
III
monitoring Hb in the presence of NO showed initial Hb -
II
NO production, followed by conversion to Hb -NO (data not
II
shown). A mechanism for hemoprotein -NO formation from
(
23) Li, Y.-T.; Hsieh, Y.-L.; Henion, J. D. J. Am. Soc. Mass Spectrom.
993, 4, 631-637.
24) Hoshino, M.; Maeda, M.; Konishi, R.; Seki, H.; Ford, P. C. J. Am.
Chem. Soc. 1996, 118, 5702-5707.
1
(
(
apoHbR; calculated Mr ) 15 126 Da; measured Mr ) 15 124
(
2 Da). Significantly, only one peak due to apoHbR is

Nitrosylated Heme in Myoglobin and Hemoglobin
J. Am. Chem. Soc., Vol. 119, No. 43, 1997 10427
Figure 7. Deconvoluted ESI mass spectrum of Hb incubated with
SNAP. Transport capillary temperature 150 °C, declustering voltage
9
4 V, scan range 1500-2400 m/z, sample infusion rate 3 µL/min,
number of scans 100 (3 s/scan). The -H denotes loss of hydrogen
from the â-chain of Hb.
Figure 5. Time course of heme nitrosylation of Mb determined by
III
UV/vis spectroscopy. (s) Mb in the absence of NO, (- - -) Mb -NO
capillary temperature 70 °C, declustering voltage 79 V) than
those used to observe nitrosylation of intact heme-containing
proteins. In addition to the peak at m/z 668.2, corresponding
II
formation after 30 min of incubation with NO, and (‚‚‚) Mb -NO
formation after 2 h of incubation with NO.
+
to [Fe(TPP)] (calculated Mr ) 668.6 Da), a second intense
peak is observed at m/z 698.5 (i.e., 30.3 Da higher than that
+
+
measured for [Fe(TPP)] ) corresponding to [Fe(TPP)NO] .
Finally, a weak peak was observed at m/z 729.0 (i.e., 60.8 Da
+
higher than that measured for [Fe(TPP)] ), which we tentatively
+
26
attribute to the dinitrosyl species, [Fe(TPP)(NO)2] .
S-Nitrosation of Hemoglobin. Although NO by itself does
not react with sulfhydryl groups under anaerobic conditions at
2
7
neutral pH to produce S-nitrosothiols, S-nitrosation may be
2
8
possible in the presence of NO and oxygen. To test further
our hypothesis that the hemoprotein-NO peaks in Figures 1c,
2
b, and 4 are due to heme nitrosylation and that the peaks
contain no significant contributions from NO modification of
the polypeptide backbone, we determined the ESI-MS conditions
necessary for observing S-nitrosated Cys residues in Hb. Figure
Figure 6. ESI mass spectrum of Fe(TPP)NO. Transport capillary
temperature 70 °C, declustering voltage 79 V, scan range 650-800
m/z, sample infusion rate 5 µL/min, number of scans 16 (3 s/scan).
7
shows the deconvoluted ESI mass spectrum of methemoglobin
III
24
(metHb) after incubation with S-nitroso-N-acetyl-DL-penicil-
lamine (SNAP) under aerobic conditions. The labelled peaks
shown in Figure 7 were ascribed to (i) apoHb (measured M
hemoprotein /NO solutions has recently been published. In
the present study, the hemoprotein/NO solutions were monitored
II
by UV/vis spectroscopy until Fe -NO formation was evident.
R
r
)
15 130 ( 3 Da), (ii) HbR (measured Mr ) 15 747 ( 2 Da),
These solutions were then analyzed by ESI-MS. UV/vis spectra
of the Mb/NO and Hb/NO solutions at times following mass
spectral analysis showed no additional changes (data not shown).
We thus deduce that the hemoprotein-NO peaks observed in
the mass spectra (Figures 1c, 2b, and 4) are due to hemo-
(
iii) apoHbâ (measured Mr ) 15 875 ( 3 Da), (iv) a reaction
product at 15 905 Da, which is 30 ( 2 Da higher than the
measured Mr of apoHbâ, (v) Hbâ (measured Mr ) 16 503 Da),
and (vi) another reaction product at Mr ) 16 532 Da, which is
II
â
29 ( 2 Da higher than the measured M of Hb . Thus, Figure
protein -NO. Although the UV/vis spectra of both the Mb/
r
7
shows that while apoHbR and HbR remain unchanged in the
NO and Hb/NO solutions showed complete conversion to the
heme-nitrosylated derivatives, the mass spectra showed the
presence of (i) apoprotein, (ii) nitrosylated protein, and (iii)
native protein. We infer that the signals corresponding to the
apo- and native protein arise from the decomposition of the
protonated nitrosylated derivative under ESI-MS conditions.
presence of SNAP, both apoHbâ and Hbâ are modified by
SNAP. On the basis of previous studies of S-nitrosation of
8
,18
Hb,
as well as solvent accessibility derived from X-ray
2
9
crystallography data, we ascribe the SNAP-modified apoHbâ
and Hbâ peaks to S-nitrosation of Hbâ Cys 93. Other high mass
peaks that occur on the tail of the Hbâ peaks are too weak to
identify. Similar results were obtained on incubating Hb with
S-nitrosated glutathione (GSNO; data not shown).8 It is
important to note that the ESI mass spectra shown in Figure 7
were recorded under ESI-MS conditions (transport capillary
temperature 150 °C, declustering voltage 94 V) that were much
too harsh for observing intact heme-nitrosylated proteins. In
addition, the UV/vis spectrum of the Hb/SNAP solution showed
Heme Nitrosylation of an Iron Tetraphenylporphyrin. To
test further our hypothesis that the hemoprotein-NO peaks
observed in Figures 1c, 2b, and 4 are due to heme nitrosylation,
we studied a nitrosylated derivative of a heme model com-
pound: nitrosyl-R,â,γ,δ-tetraphenylporphinatoiron(II), [Fe-
(TPP)NO]. To the best of our knowledge, intact nitrosylated
iron porphyrins have not previously been observed by ESI-MS.
Electron impact ionization mass spectrometry was previously
+
used to characterize Fe(TPP)NO, but only the NO and
(
26) Wayland, B. B.; Olson, L. W. J. Am. Chem. Soc. 1974, 96, 6037-
6041.
(27) Pryor, W. A.; Church, D. F.; Govindian, C. K.; Crank, G. J. Org.
Chem. 1982, 47, 159-161.
28) Gow, A. J.; Buerk, D. G.; Ischiropoulos, H. J. Biol. Chem. 1997,
72, 2841-2845.
29) Heidner, E. J.; Ladner, R. C.; Perutz, M. F. J. Mol. Biol. 1976,
104, 707-722.
+
25
[
Fe(TPP)] fragments were detected. The ESI mass spectrum
of Fe(TPP)NO obtained under the present conditions is shown
in Figure 6. It is noteworthy that the conditions required to
observe the intact Fe(TPP)NO are considerably gentler (transport
(
2
(
(25) Scheidt, W. R.; Frisse, M. E. J. Am. Chem. Soc. 1975, 97, 17-21.

10428 J. Am. Chem. Soc., Vol. 119, No. 43, 1997
Upmacis et al.
increase in the capillary temperature (i.e., 10 °C higher than
the temperature used to obtain the mass spectra shown in Figures
1
c and 2b) resulted in the complete disappearance of the peaks
II
II
assigned to Mb -NO and Hb -NO (see Figures 1b and 2a).
If, on the other hand, the capillary temperature and/or declus-
tering voltage were set too low, solvent molecules remained
attached to the ionized protein and a broad tailing parent ion
peak was observed in the spectrum. It is, therefore, evident
that a narrow, but reproducible, range of ESI-MS conditions
exist in which nitrosylated hemoproteins and nitrosylated Fe-
containing porphyrins can be observed.
Appropriate Sample Handling Conditions Are Critical for
the Observation of Heme-NO Interactions. In order to
ensure the observation of NO-modified heme moieties, it was
also necessary to control carefully the sample handling condi-
tions. A mixture of methanol (10% v/v) and ammonium
bicarbonate (5 mM) proved to be an excellent electrospray buffer
for these measurements, because methanol is volatile and
ammonium bicarbonate decomposes to volatile NH3, CO2, and
Figure 8. Deconvoluted ESI mass spectrum of Hb subsequent to
incubation with NO under acidified conditions in the presence of air.
Transport capillary temperature 130 °C, declustering voltage 55 V, scan
range 600-1700 m/z, sample infusion rate 3 µL/min, number of scans
1
3 4
5 (3 s/scan). The asterisk denotes an adduct of H PO and the pound
sign (#) denotes a Na adduct. The -H denotes loss of hydrogen from
the R- and â-chains of Hb.
3
1
H2O under MS conditions.
It was essential to maintain
no evidence of heme nitrosylation (data not shown). Thus, we
deduce that (i) the SNAP-modified Hbâ peak shown in Figure
anaerobic conditions to prevent the formation of nitrous acid
which, if present, could potentially lower the pH and lead to (i)
disruption of the coordinate bond between the heme and the
7
is solely due to SNAP modification of the polypeptide
backbone and (ii) the S-nitrosated Hbâ derivative is considerably
more stable than the heme-NO-globin complex.
32
33
apoprotein and (ii) S-nitrosation of Cys residues. The above-
mentioned factors conspire to make observation of Fe-NO
interactions in proteins and porphyrins challenging and are,
presumably, the reason why earlier attempts to detect heme-
Figure 8 shows that under appropriate denaturing sample
conditions, it is possible to S-nitrosate apoHb completely and
determine the resulting product. The sample was prepared by
withdrawing an aliquot from an Hb/NO solution (prepared under
anaerobic conditions) and acidifying the aliquot with acetic acid
in the presence of air. Under these sample conditions, the heme
group is lost from the protein, the polypeptide backbone is
denatured, and all three Cys residues in human apoHb (Cys
19
NO products by mass spectrometry were unsuccessful.
S-Nitrosated Proteins Can Be Readily Differentiated from
Heme-Nitrosylated Proteins by ESI-MS. To exclude the
possibility that NO may be modifying the polypeptide backbone,
we investigated the ESI-MS conditions under which S-nitrosa-
tion of cysteine residues in Hb can be observed. Human Hb
contains three Cys residues (HbR Cys 104, Hbâ Cys 93, and
Hbâ Cys 112), but only Hbâ Cys 93 is surface accessible and
available for S-nitrosation. Thus, treatment of Hb at neutral
pH with a transnitrosating agent such as SNAP or GSNO should
1
04 in apoHbR, Cys 93 and Cys 112 in apoHbâ) are S-nitrosated
30
by a higher oxide of NO. The spectrum in Figure 8 displays
two peaks at 15 159 and 15 931 Da, which are respectively 28
(
2 and 62 ( 5 Da higher than that measured for apoHbR
15 131 Da) and apoHbâ (15 869 Da), consistent with S-
nitrosation of Cys 104 in apoHbR and of both Cys 93 and Cys
12 in apoHbâ.
(
lead to modification of a single Cys residue in Hbâ, but no
8,34
modification of Cys in HbR.
The results shown in Figure 7
1
confirm this hypothesis. The ESI-MS conditions under which
we observe S-nitrosation of Hbâ (most likely Hbâ Cys 93)8
by SNAP modification of Hbâ (Figure 7) are much harsher than
those used to observe intact Fe-NO interactions in either Mb
,18
Discussion
Observation of Heme-NO Interactions Requires Careful
Adjustment of ESI-MS Parameters. In Figures 1a and 1b,
we see that the measurement of intact Fe-NO interactions in
proteins is challenging because (i) under normal ESI-MS
conditions, heme loss from the protein occurs during the MS
measurement and (ii) heme loss is greater in the presence of
NO. Presumably, the enhanced MS loss of the heme group
after treatment with NO is a result of the destabilizing effect of
heme-NO on the heme-globin complex. In order to observe
intact Fe-NO interactions in proteins and porphyrins by ESI-
MS, it proved critical to control carefully the conditions used
for sample handling and electrospray ionization (see figure
legends). Thus, it was necessary to adjust the source parameters
that control the amount of energy transmitted to the nitrosylated
molecules during ESI. In the Finnigan MAT TSQ-700 elec-
trospray ion source, this energy input into the solvated ion is
(Figure 1c) or Hb (Figures 2b and 4). Thus, the data presented
here demonstrate that heme-nitrosylated proteins are much more
prone to dissociation during ESI-MS than S-nitrosated proteins.
ESI-MS Can Be Used To Determine the Number of NO
Modifications in Proteins. Under appropriate sample condi-
tions, it proved possible to S-nitrosate all three Cys residues in
human apoHb and to observe these modifications by ESI-MS.
In Figure 8, we demonstrate that when an anaerobic Hb/NO
solution is denatured by acidification and is exposed to air, all
three Cys residues in human apoHb are observed to be
S-nitrosated. Under physiological conditions, this ability to
count the number of NO modifications has proved useful in
studies of NO treatment of bovine cytochrome c (Cyt c) under
anaerobic conditions. ESI-MS analysis of a Cyt c/NO solution
revealed that two forms of Cyt c were produced during heme
21
controlled by the transport capillary temperature and the
declustering voltage.22 A slight variation in these parameters
had a large influence on whether species containing Fe-NO
moieties were observed. If the capillary temperature and/or
declustering voltage were set too high, the Fe-NO bond was
broken and heme loss from the protein increased. Thus, a slight
(31) The Merck Index, 10th ed.; Windholz, M., Ed.; Merck & Co., Inc.:
Rahway, NJ, 1983.
(32) Duprat, A. F.; Traylor, T. G.; Wu. G.-Z.; Coletta, M.; Sharma, V.
S.; Walda, K. N. Magde, D. Biochemistry 1995, 34, 2634-2644.
(33) Ignarro, L. J.; Lippton, H.; Edwards, J. C.; Baricos, W. H.; Hyman,
A. L.; Kadowitz, P. J.; Gruetter, C. A. J. Pharm. Exp. Ther. 1981, 218,
739-749.
(34) SNAP modification of horse heart Mb was not examined because
(30) Williams, D. L. H. Chem. Soc. ReV. 1985, 14, 171-196.
Mb contains no Cys residues and, thus, does not undergo S-nitrosation.

Nitrosylated Heme in Myoglobin and Hemoglobin
J. Am. Chem. Soc., Vol. 119, No. 43, 1997 10429
nitrosylation.35 One form corresponds to native Cyt c, while
the second form is a product that is 17 Da lighter in mass than
Cyt c. By subjecting the NO-modified Cyt c proteins to enzyme
digests and subsequent MS analysis, we have discovered that
the Cyt c derivative arises from a deamination reaction (involv-
ing Lys 72) in the presence of NO (data not shown).
Scheidt and Frisse.25 Briefly, R,â,γ,δ-tetraphenylporphinatoiron(III)
chloride, Fe(TPP)Cl, (0.2 g) in chloroform (60 mL) containing dry
pyridine (1 mL) was degassed in a Schlenk tube. NO (1 atm) was
passed above the stirred solution for 20 min. Dry MeOH was added
2
until a precipitate of Fe(TPP)NO appeared. Filtration under N with a
cannula gave a purple powder. For the purposes of electrospray, the
sample (400 µM) was prepared in deaerated chloroform and injected
into the mass spectrometer Via a gas-tight Hamilton syringe.
Conclusions
Preparation of S-Nitrosated Hemoglobin. A stock solution of
SNAP (10 mM) was prepared by dissolving SNAP (2.2 mg) in dimethyl
sulfoxide (DMSO, 10 µL) and deionized water (990 µL). A stock
ESI-MS was found to be a versatile tool for Fe-NO detection
in proteins and porphyrins. In the present study, we have
established mass spectrometric conditions to observe such weak
Fe-NO interactions in Mb, Hb, and Fe(TPP)NO. Limitations
of this method include the following: (a) the unavoidable partial
decomposition of the nitrosylated product under certain ESI-
MS conditions, (b) the present inability to measure heme
nitrosylation in intact tetrameric Hb, (c) the relative difficulty
in measuring â-nitrosyl hemes, and (d) the present inability to
measure concentrations <20 µM. Advantages of the present
ESI-MS approach to the detection of Fe-NO interactions
include the ability to obtain, (a) direct measurements of the intact
Fe-NO-containing complex, (b) information concerning the
exact number of NO moieties attached to the molecule, and (c)
information concerning the formation of other byproducts (e.g.,
dinitrosylation of heme in Fe(TPP) and deamination of a Lys
residue in Cyt c),35 as well as S-nitrosation of cysteines.
III
III
solution of human Hb (50 µM) was prepared by dissolving Hb in a
buffer (10% v/v methanol and 90% v/v 5 mM ammonium bicarbonate;
III
pH 8). The stock SNAP and Hb solutions were mixed in appropriate
III
proportions to yield a 10:1 molar ratio of SNAP/Hb . The mixture
was allowed to react for 10 min at room temperature under aerobic
conditions prior to injection into the ESI mass spectrometer. A GSNO/
III
Hb solution was prepared in a similar manner, except that no DMSO
was necessary in the preparation of the stock GSNO solution.
Preparation of S-Nitrosated Apohemoglobin. A solution of Hb/
NO (40 mM in 10% v/v MeOH and 90% v/v 5 mM ammonium
bicarbonate) was prepared as above under anaerobic conditions. An
aliquot was removed from this mixture and exposed to air, followed
by a 4-fold dilution with 29% v/v H
acetic acid.
2
O/68% v/v acetonitrile/3% v/v
ESI-Mass Spectrometry. A Finnigan-MAT TSQ-700 triple quad-
rupole instrument was used to obtain all ESI mass spectra. The
solutions were sprayed through a 100 µm (inner diameter) fused silica
capillary into the mass spectrometer source using a Harvard syringe
pump (model 24000-001). The ESI-MS conditions (capillary temper-
ature, declustering voltage, scan range, sample infusion rate, and number
of scans) are given in the figure legends. The ion signals were recorded
by a Finnigan ICIS data system operated on a DEC station 500/120
system. The reconstructed molecular mass profiles were obtained by
using a deconvolution algorithm (FinniganMAT).
Experimental Section
Materials. Horse heart metmyoglobin and human methemoglobin
were purchased from Sigma; nitric oxide (99.0% min) was obtained
from Matheson; R,â,γ,δ-tetraphenylporphinatoiron(III) chloride was
obtained from Aldrich. All were used as received. S-Nitroso-N-acetyl-
DL-penicillamine (SNAP) and S-nitrosoglutathione (GSNO) were
UV/Vis Spectroscopy. UV/vis spectra were recorded using a Perkin
Elmer Lambda-19 UV/vis/near IR spectrophotometer. All UV/vis
spectra were recorded at room temperature.
36
synthesized according to procedures described by Field et al. and
3
7
Hart, respectively.
Preparation of Heme-Nitrosylated Hemoproteins. The nitrosyl-
ated hemoproteins were prepared by anaerobically adding NO (1 atm)
Acknowledgment. We are very grateful to Professor Gerard
Parkin for his assistance in the preparation of the heme-
nitrosylated derivatives under anaerobic conditions and for the
use of the UV/vis spectrophotometer at Columbia University.
This study was made possible, in part, by funds granted to
R.K.U. through a fellowship program sponsored by The Norman
and Rosita Winston Foundation. R.K.U. was also supported on
a NIH training grant awarded to D.P.H. (HL-07423-17) and an
American Heart Association Scientist Development Award
III
to either horse heart metmyoglobin (Mb ) or human methemoglobin
III
(
Hb ) (20 µM) dissolved in a buffer (10% v/v methanol and 90% v/v
5
mM ammonium bicarbonate; pH 8) in a Thunberg cell. The samples
II
were monitored by UV/vis spectroscopy until Fe -NO formation was
evident. Samples were withdrawn from the cell and injected into the
mass spectrometer Via a gas-tight Hamilton syringe.
Preparation of a Heme-Nitrosylated Tetraphenylporphyrin.
Nitrosyl-R,â,γ,δ-tetraphenylporphinatoiron(II), Fe(TPP)NO, was pre-
pared according to a modified version of a preparation described by
(9630223N). B.T.C. and U.A.M. gratefully acknowledge partial
support from the NIH (grant RR00862). The statements made
and views expressed are solely the responsibility of the authors.
(
35) Manuscript in preparation.
(36) Field, L.; Dilts, R. V.; Ravichandran, R.; Lenhert, P. G.; Carnahan,
G. E. J. Chem. Soc., Chem. Commun. 1978, 249-250.
(
37) Hart, T. W. Tetrahedron Lett. 1985, 26, 2013-2016.
JA964249L