Nitrosonium-catalyzed decomposition of S-nitrosothiols in solution: A theoretical and experimental study

Article abstract of DOI:10.1021/ja050018f

The decomposition of S-nitrosothiols (RSNO) in solution under oxidative conditions is significantly faster than can be accounted for by homolysis of the S-N bond. Here we propose a cationic chain mechanism in which nitrosation of nitrosothiol produces a nitrosated cation that, in turn, reacts with a second nitrosothiol to produce nitrosated disulfide and the NO dimer. The nitrosated disulfide acts as a source of nitrosonium for nitrosothiol nitrosation, completing the catalytic cycle. The mechanism accounts for several unexplained facets of nitrosothiol chemistry in solution, including the observation that the decomposition of an RSNO is accelerated by O₂, mixtures of O ₂ and NO, and other oxidants, that decomposition is inhibited by thiols and other antioxidants, that decomposition is dependent on sulfur substitution, and that decomposition often shows nonintegral kinetic orders.

Full text of DOI:10.1021/ja050018f

Published on Web 07/15/2005
Nitrosonium-Catalyzed Decomposition of S-Nitrosothiols in
Solution: A Theoretical and Experimental Study
Yi-Lei Zhao,† Patrick R. McCarren,† K. N. Houk,*,† Bo Yoon Choi,‡ and
Eric J. Toone*,‡
Contribution from the Department of Chemistry and Biochemistry, UniVersity of California,
Los Angeles, California 90095-1569, and Department of Chemistry and Biochemistry,
Duke UniVersity, Durham, North Carolina 27708
Received January 3, 2005; E-mail: houk@chem.ucla.edu; eric.toone@duke.edu
Abstract: The decomposition of S-nitrosothiols (RSNO) in solution under oxidative conditions is significantly
faster than can be accounted for by homolysis of the S-N bond. Here we propose a cationic chain
mechanism in which nitrosation of nitrosothiol produces a nitrosated cation that, in turn, reacts with a second
nitrosothiol to produce nitrosated disulfide and the NO dimer. The nitrosated disulfide acts as a source of
nitrosonium for nitrosothiol nitrosation, completing the catalytic cycle. The mechanism accounts for several
unexplained facets of nitrosothiol chemistry in solution, including the observation that the decomposition of
an RSNO is accelerated by O2, mixtures of O2 and NO, and other oxidants, that decomposition is inhibited
by thiols and other antioxidants, that decomposition is dependent on sulfur substitution, and that
decomposition often shows nonintegral kinetic orders.
Introduction
S-Nitrosothiol decomposition produces a range of organic and
inorganic products, but in many cases decomposition liberates
Nitric oxide is an ubiquitous biological species with myriad
activities. Although nitric oxide is synthesized enzymatically
as the neutral +2 oxide of nitrogen, several other nitrogen
oxides, including the +3 (NO+) and +1 (NO-/HNO) oxides,
are known and have been implicated in various biological
processes. The +3 oxide exists in vivo primarily as S-
nitrosothiol (RSNO); these species, especially S-nitrosoglu-
tathione, represent the largest circulating pool of nitric oxide.1
S-Nitrosation of protein thiols may also represent an important
post-translational modification involved in both regulation and
signaling.
S-Nitrosothiols are typically regarded as unstable in solution,
and under some conditions many decompose with half-lives of
seconds to minutes.2 Although primary and secondary S-
nitrosothiols are often regarded as less stable than the corre-
sponding tertiary compounds, wide variations in stability are
the rule and conflicting values are often reported.3 The time
course of S-nitrosothiol decomposition is complex and frequently
shows nonintegral kinetic orders.4
neutral nitric oxide and disulfide. Activation energies for the
decomposition of S-nitrosothiol have been reported from 20 to
31 kcal mol-1 by different groups.5,6 Although NO is formally
the product of S-N bond scission, homolytic bond dissociation
energies for variously substituted alkyl S-nitrosothiols in the
gas phase are uniformly near 31 kcal mol-1, values that predict
apparent half-lives of years near room temperature.6 Addition-
ally, the uniformity of bond dissociation energies as a function
of sulfur substitution is inconsistent with frequently observed
variation in nitrosothiol stability as a function of substitution.
Accordingly, we continue our efforts to elucidate important
pathways of nitrosothiol decomposition.
The stability of S-nitrosothiols is highly dependent on reaction
conditions. A reductive decomposition mediated by cuprous ion
has been well established by Williams and co-workers.7 Oxida-
tive conditions are also known to induce S-nitrosothiol de-
composition, and Grossi et al. recently proposed a catalytic
(5) The BDE of tertiary RSNO is somewhat lower than that of primary RSNO;
thermal decomposition of tertiary RSNO is faster than that of primary
RSNO. (a) Grossi, L.; Montevecchi, P. C. Chem.sEur. J. 2002, 8, 380-
387. (b) Lu, J.-M.; Wittbrodt, J. M.; Wang, K.; Wen, Z.; Schlegel, H. B.;
Wang, P. G.; Cheng, J.-P. J. Am. Chem. Soc. 2001, 123, 2903-2904.
(6) (a) Bartberger, M. D.; Mannion, J. D.; Powell, S. C.; Stamler, J. S.; Houk,
K. N.; Toone, E. J. J. Am. Chem. Soc. 2001, 123, 8868-8869. (b)
Bartberger, M. D.; Houk, K. N.; Power, S. C.; Mannion, J. D.; Lo, K. Y.;
Stamler, J. S.; Toone, E. J. J. Am. Chem. Soc. 2000, 122, 5889-5890. (c)
Baciu, C.; Gauld, J. W. J. Phys. Chem. A 2003, 107, 9946-9952.
(7) (a) Williams, D. L. H. Chem. Commun. 1996, 1085-1090. (b) Dicks, A.
P.; Swift, H. R.; Williams, D. L. H.; Butler, A. R.; Al-S’adoni, H. H.;
Cox, B. G. J. Chem. Soc., Perkin Trans. 2 1996, 481-487. (c) Stubauer,
G.; Giuffre, A.; Sarti, P. J. Biol. Chem. 1999, 274, 28128-28133. (d)
Ramirez, J.; Yu, L.; Li, J.; Braunschweiger, P. G.; Wang, P. G. Bioorg.
Med. Chem. Lett. 1996, 6 (21), 2575-2580. (e) Megson, I. L.; Greig, I.
R.; Gray, G. A.; Webb, D. J.; Butler, A. R. Br. J. Pharmacol. 1997, 122
(8), 1617-1624.
† University of California.
‡ Duke University.
(1) (a) Stamler, J. S. Cell 1994, 78, 931-936. (b) Stamler, J. S. Curr. Top.
Microbiol. Immunol. 1995, 196, 19-36. (c) Williams, D. L. H. Acc. Chem.
Res. 1999, 32, 869-876.
(2) (a) Singh, S. P.; Wishnok, J. S.; Sechive, M.; Deen, W. M.; Tannenbaum,
S. R. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 14428-14433. (b) Munro,
A. P.; Williams, D. L. H. Can. J. Chem. 1999, 77, 550-556. (c) Singh, R.
J.; Hogg, N.; Joseph, J.; Kalyanaraman, B. J. Biol. Chem. 1996, 271,
18596-18603.
(3) (a) Oae, E.; Fukushima, D.; Kim, Y. H. J. Chem. Soc., Chem. Commun.
1977, 407-408. (b) Ignarro, L. J.; Lippton, H.; Edwards, J. C.; Baricos,
W. H.; Hymann, A. L.; Kadowitz, P. J.; Gruetter, C. A. J. Pharmacol.
Exp. Ther. 1981, 218, 739-749.
(4) De Oliveria, M. G.; Shishido, S. M.; Seabra, A. B.; Morgon, N. H. J. Phys.
Chem. A 2002, 106, 8963-8970 and references therein.
9
10.1021/ja050018f CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 10917-10924
10917

A R T I C L E S
Zhao et al.
decomposition pathway initiated by nitrosothiol homolysis and
mediated by N2O3.8
Both 4 and 8, the two lowest-energy structures of the N- and
S-nitrosated species, exist as resonance hybrids (Scheme 2).
Mulliken charges on 4 indicate that half of the positive charge
is borne by the sulfur atom and the remaining half by the
appended NO group. The central NO moiety bears essentially
no charge. In the symmetrical structure, 8, the largest portion
of the charge (32%) is borne by the sulfur, with the remaining
charge distributed between the nitrogens (23%) and oxygens
(11%) of the equivalent NO moieties.
Here we report theoretical and experimental studies on a
catalytic decomposition pathway for nitrosothiol mediated by
nitrosonium (Scheme 1). In this process, nitrosonium binds
nitrosothiol, leading to a stable electrophilic complex that reacts
with a second equivalent of nitrosothiol, releasing the nitric
oxide dimer and nitrosated disulfide. This latter species acts as
a powerful source of nitrosonium, completing the catalytic cycle.
The proposed mechanism has been tested by evaluation of the
decomposition of S-nitroso-n-hexylthiol in the presence of
various sources of nitrosonium.
Scheme 2. Canonical Contributors to N- and S-Nitrosonitrosothiol
Resonance Hybrids
Scheme 1. Proposed Cationic Catalytic Cycle for Decomposition
of RSNO (R ) Me)
The energetics of the nitrosation of S-nitrosothiol by various
nitrosating agents are shown in Table 1. For ionic NO+ donors,
such as NO+‚BF4-, NO+‚PCl6-, and NO+‚SbCl6-, the addition
of NO+ to RSNO is thermodynamically favorable. Since
solvation energies of NO+ are greater than that of [RS(NO)-
NO+], the free energy of reaction becomes less favorable as
solvent polarity increases: ∆Grxn for the nitrosation of MeSNO
increases from -24 kcal mol-1 in the gas phase to -10 kcal
mol-1 in acetonitrile. Although NO+ has only a fleeting
existence in water because of rapid hydrolysis (NO+ + 2H2O
f HNO2 + H3O+), the dinitrosothiol cation, [RS(NO)NO+],
could survive in methanol, acetonitrile, and n-heptane solutions
even with traces of water (Table 1).11 Nitrosation by weaker
sources of nitrosonium, such as O2 + NO, N2O3, N2O4, CH3-
ONO, or CH3SNO, is considerably less favorable than by
nitrosonium salts, and nitrosations by alkyl nitrite or nitrosothiol
are predicted to be endothermic.
Results and Discussion
Nitrosation of S-Nitrosothiol. In the gas phase, nitrosonium
can nitrosate both cis- and trans-S-nitrosomethanethiol on either
nitrogen or sulfur to form the novel cation-molecule species
N-nitrosonium-S-nitrosomethanethiol (4-7) and S-nitrosonium-
S-nitrosomethanethiol (8-10) (Figure 1). Formation of these
complexes is barrierless and exothermic by 24 to 34 kcal mol-1
in the gas phase. The C-S-N-O functionality is essentially
planar, with dihedral angles near 0° (cis) or 180° (trans). Five
of the complexes (4-6, 8, 10) are Cs symmetric; the O-N-
N-O dihedral angle in cation 7 is 139°, and cation 9 has two
different cis- and trans-NO substituents. The S-N bond lengths
of the N-nitrosonium-S-nitrosomethanethiols (4-7) range from
1.68 to 1.73 Å. The nitrosonium moiety is bound only loosely,
and the positive charge is borne partially by the sulfur atom. In
contrast to the S-N bonds of N-nitrosonium-S-nitroso-
methanethiols, the S-N bonds of the S-nitrosated species are
elongated by 0.4 Å, with two NO groups bound loosely to sulfur.
A similar lengthening of the S-N bond (0.25 Å) has been
reported in calculations of a CuI-nitrosothiol complex.9 The
lowest energy conformer of the dinitrosothiol cation is the
cis,cis-N-nitrosonium-S-nitrosothiol species 4 with a N-N
distance of 2.14 Å. The isomeric complexes 5-10 show energies
3.1, 4.9, 10.2, 1.4, 1.8, and 2.8 kcal mol-1 above that of 4,
respectively, in the gas phase. The preference for the cis-C-
S-N-O dihedral arises from repulsion between lone pairs on
oxygen and sulfur. A similar preference was previously noted
in a study of anionic nitroxyl disulfide.10
Nitric oxide and dioxygen, when dioxygen is present in a
considerable concentration, can react rapidly to form various
oxides of nitrogen, in particular N2O3 and N2O4.12,13 Grossi and
co-workers have proposed a mechanism of nitrosothiol decom-
position that involves a one-electron oxidation of RSNO by
N2O3 to generate the corresponding nitrosothiol radical cation.8
Our calculations suggest this reaction is highly unfavorable, with
a ∆Grxn,gas of 164 kcal mol-1. The reaction is somewhat less
unfavorable in polar solvents, with predicted free energies of
99 kcal mol-1 and 31 kcal mol-1 in n-heptane and acetonitrile,
respectively. These calculated values are for the reaction of
separated ions, and reaction through ionic pairs could further
lower these endothermicities. In contrast, the radical recombina-
tion of RSNO•+ with NO to form [RS(NO)NO+] (R ) Me) is
quite exothermic with ∆Grxn values of -20, -18, and -16 kcal
mol-1 in the gas phase, n-heptane, and acetonitrile, respectively
(Table 1). Together, these calculations strongly suggest that the
reaction of RSNO with N2O3 is more likely to produce [RS-
(NO)NO+] than RSNO•+.
(8) (a) Grossi, L.; Montevecchi, C.; Strazzari, S. J. Am. Chem. Soc. 2001, 123,
4853-4854. (b) Lipton, S. A.; Choi, Y. B.; Pan, Z. H.; Lei, S. Z.; Chen,
H. S.; Sucher, N. J.; Loscalzo, J.; Singel, D. J.; Stamler, J. S. Nature 1993,
364, 626-632.
(11) MeSdNOH+ is a higher-energy isomer of MeSH(NO)+; at the B3LYP/
6-31+G* level, the former is higher by 1.7 kcal/mol than the latter.
(12) Olson, L. P.; Kuwata, K. T.; Bartberger, M. D.; Houk, K. N. J. Am. Chem.
Soc. 2002, 124, 9469-9475.
(13) Grossi, L.; Strazzari, S. J. Org. Chem. 1999, 64, 8076-8079 and references
therein.
(9) Toubin C.; Yeung, D. Y. H.; English, A. M.; Peslherbe, G. H. J. Am. Chem.
Soc. 2002, 124, 14816-14817.
(10) Houk, K. N.; Hietbrink, B. N.; Bartberger, M. D.; McCarren, P. R.; Choi,
B. Y.; Voyksner, R. D.; Stamler, J. S.; Toone, E. J. J. Am. Chem. Soc.
2003, 125, 6972-6976.
9
10918 J. AM. CHEM. SOC. VOL. 127, NO. 31, 2005

NO+-Catalyzed Decomposition of S-Nitrosothiols
A R T I C L E S
Figure 1. Calculated structures for secondary nitrosation of RSNO with the CBS-QB3 method.
Table 1. Calculated Free Energies (∆G298,rxn in kcal/mol) of Reactions of RSNO with NO+ or Other Donors
CBS-QB3
(CPCM)//B3LYP/CBSB7
entry
reaction
gas phase
acetonitrile
1
2
3
4
5
6
7
8
NO+ + MeSNO f MeS(NO)NO+
-24.2
+13.0
+8.7
+14.7
+37.2
+1.3
-10.2
-0.2
NO+ + 2H2O f HNO2 + H3O+
MeS(NO)NO+ + MeOH f MeS(H)NO+ + MeONO
MeS(NO)NO+ + H2O f MeS(H)NO+ + HONO
MeS(NO)NO+ + 2H2O f MeSNO + HONO + H3O+
MeS(NO)NO+ + MeSH f MeS(H)NO+ + MeSNO
4 NO+ O2 f 2N2O3
+6.7
+11.8
+10.0
-1.6
-22.8
-24.1
-5.3
-15.8
-20.6
+136.9
+125.1
+182.9
+168.7
+163.9
-20.1
+28.8
-1.1
2 NO+ O2 f N2O4
-
9
N2O3 + MeSNO f MeS(NO)NO+ + NO2
N2O4 + MeSNO f MeS(NO)NO++ NO3
-12.9
+37.8
+34.7
+31.3
-16.4
+7.7
-
10
11
12
13
14
15
16
17
MeONO + MeSNO f MeS(NO)NO+ + MeO-
2 MeSNO f MeS(NO)NO+ + MeS-
N2O3 + MeSNO f MeSNO•+ + N2O3
•-
MeSNO•+ + NO f MeS(NO)NO+
MeSSMe(NO)+ f MeSSMe + NO+
MeSSMe(NO)+ f MeSSMe•+ + NO
MeSSMe + MeSSMe(NO)+ f (MeSSMe) 2 (NO)+
+0.8
+9.7
-8.0
Disulfide Formation from RSNO + [RS(NO)NO+]. The
sulfur of nitrosated nitrosothiol is strongly electrophilic, and a
displacement reaction with RSNO produces a nitrosated disulfide
and the NO dimer. Sixteen transition states were located for
the SN2 reaction of N- and S-nitrosonium-nitrosothiols with
RSNO using B3LYP/6-31+G* and B3LYP/CBSB7 methods
(see Supporting Information). All of the N-nitrosonium-S-
nitrosothiol transition states are lower in energy than the
corresponding S-nitrosonium-S-nitrosothiols. The two rotameric
lowest-energy transition states (11, 12) were studied further with
9
J. AM. CHEM. SOC. VOL. 127, NO. 31, 2005 10919

A R T I C L E S
Zhao et al.
Figure 2. Calculated structures of the transition states and products for SN2-like S-S forming reaction with the CBS-QB3 method.
CBS-QB3 calculations (Figure 2). Both transition states show
typical SN2 characteristics: a nearly linear N-S-S arrangement,
a lengthened N-S bond (2.06 Å; 2.18 Å), and a forming S-S
bond (2.48 Å; 2.44 Å). The relative gas-phase energies of the
two transition states are 7.9 and 9.4 kcal mol-1, respectively
(CBS-QB3).14
cantly lower than that for homolytic scission of nitrosothiol to
free NO and the thiyl radical which, at 27 kcal mol-1, makes
this decomposition manifold of only minor importance near
physiologically relevant temperatures.15 The activation free
energy for nucleophilic displacement of NO dimer from
N-nitrosonium-S-nitrosothiol by nitrosothiol is 8 kcal mol-1
,
reflecting the entropy loss during bimolecular reaction. Both
the reactive complex and the transition state are solvated less
effectively than are the isolated reactants. As a result, the
activation free energies (∆Gsolq) increase by 4 and 10 kcal mol-1
in n-heptane and acetonitrile, respectively (Figure 3). Nonethe-
less, the overall resulting free energy of activation in acetonitrile
(18 kcal mol-1) is still significantly more favorable than
homolytic scission of the nitrosothiol S-N bond.
Nitrosonium Transfer. Completion of the catalytic cycle
requires transfer of a nitrosonium equivalent from the nitrosated
disulfide. While this process is predicted to be endothermic by
4 kcal mol-1 in the gas phase, the same process is calculated to
be exothermic by 2 kcal mol-1 in acetonitrile solution, due to
differential solvation of the N-nitrosonium-S-nitrosothiol (Figure
4).
All transition states resulting from nitrosothiol attack on
N-nitrosonium-S-nitrosothiol lead to identical products: the NO
dimer and a nitrosonium disulfide cation (14, Figure 2). The
novel cation has a normal S-S bond length of 2.07 Å and an
elongated S-N bond length of 2.32 Å, consistent with a
nitrosonium ion weakly coordinated to disulfide. Frequency
analysis confirms that this structure is a true energy minimum
bound by 38 kcal mol-1 (enthalpy) relative to CH3SSCH3 +
NO+ in the gas phase (Table 1). The 1.12 Å N-O bond length
of 14 is consistent with N-O bonds of S-dinitrosothiols, but it
is somewhat longer than the 1.06 Å length calculated for NO+
at the same theory level. Rapid dissociation of the NO dimer
into monomer provides a driving force for the overall reaction.
The lowest enthalpy reaction pathway passes through transi-
tion state 11, with a gas-phase activation enthalpy of 7.9 kcal
mol-1 above complex 16 and 4.9 kcal mol-1 below dissociated
RSNO and RS(NO)NO+ (CBS-QB3). This barrier is signifi-
Transfer could conceivably proceed via either an associative
(addition/elimination) or a dissociative pathway. Dissociation
(14) The intramolecular transition state for rotation around the S-N bond can
interconvert 4 and 5 (see Supporting Information). The activation free
energies are +20.4 and +18.1 kcal/mol in gas phase and acetonitrile,
respectively.
(15) The activation energy of 27 kcal/mol is the free-energy maxima in the
homolysis of RSNO. We also located a concerted bimolecular transition
structure with a barrier of 36 kcal/mol (∆Hq) and 45 kcal/mol (∆Gq) in the
gas phase.
9
10920 J. AM. CHEM. SOC. VOL. 127, NO. 31, 2005

NO+-Catalyzed Decomposition of S-Nitrosothiols
A R T I C L E S
Figure 4. Energetic profile for nitrosonium transfer.
(4 and 5). Conceptually, the simplest interconversion of the
isomers involves rotation around the S-N bond. This process,
however, shows an activation barrier of roughly 20 kcal mol-1
both in the gas phase and in solvents of all polarities.
Nitrosonium transfer by addition-elimination offers an alterna-
tive process for isomerization and is likely important at high
nitrosothiol concentrations.17 The intermediate in this transfer
(16, Figure 2) is formed in a barrierless process that is
exothermic in the gas phase (-4.5 kcal mol-1) and slightly
endothermic in the condensed phase (4 and 9 kcal mol-1 in
hexane and acetonitrile, respectively). The unusual cation, which
formally bonds three NO equivalents, places two N-nitrosonium-
S-nitrosothiol planes in an almost orthogonal orientation, with
a N-N-N angle of about 96°. As was the case for the
nitrosodisulfide cation, a dissociative mechanism proceeding
through free nitrosonium is significantly more endothermic than
the associative route: dissociation of 4 to NO+ is endothermic
by 24, 17, and 10 kcal mol-1 in the gas phase, n-heptane, and
acetonitrile, respectively.
Termination of the NO+-Catalyzed RSNO Decomposition:
The chain mechanism proposed here effectively relies on
nitrosonium as the chain-carrying species: any species that
scavenges nitrosonium will terminate the chain. Such species
include oxygen and sulfur nucleophiles, and a nitrosonium-
scavenging chain-termination mechanism is consistent with the
well-known stabilizing effect of excess thiol. Dissociation of
the nitrosodisulfide cation to neutral NO and the disulfide radical
cation also terminates chains. The reaction is endothermic by
10 kcal mol-1 but has a favorable free energy.18 The resulting
disulfide radical cation populates a planar structure (C-S-S-
Figure 3. Energetic profile of the SN2 reaction between RS(NO)NO+ and
RSNO (R ) Me).
of nitrosonium from the nitrosodisulfide cation is endothermic
by 29, 17, and 8 kcal mol-1 in the gas phase, n-heptane, and ace-
tonitrile, respectively. The association free energies (∆Gassn) for
RSS(NO)R+ and RSNO into complex [RSSR(NO)RSNO]+ (17,
Figure 5) are -4, 3, and 8 kcal/mol in the gas phase, n-heptane
(ꢀ ) 1.92), and acetonitrile (ꢀ ) 36.64), respectively (CBS-QB3,
CPCM). Overall, an associative pathway is the strongly preferred
route for nitrosonium transfer in low polarity solvents, while the
dissociative mechanism becomes competitive in polar solvents.
The associative mechanism of nitrosonium transfer passes
through intermediate 17. The calculated gas-phase structure of
this intermediate shows a central N-S distance of 2.48 Å and
a N-N bond length of 2.29 Å. The S-N bond of the RSNO
moiety is slightly shortened from 1.87 Å to 1.71 Å, similar to
that observed in N-nitrosonium-S-nitrosothiol. To confirm that
17 is a true minimum in solution, the structure was reoptimized
in acetonitrile (ꢀ ) 36.64) at the B3LYP/6-31+G* level. The
reoptimized structure is virtually identical to that derived in gas-
phase calculations.16
(16) Because of the flatness of the potential energy surface, no transition state
is located for breaking and forming intermediates 16 and 17. The estimated
activation energy, even in acetonitrile, is less than 3 kcal/mol with respect
to the intermediate based on scanning calculations.
(17) The energetic profiles of the elimination-addition, addition-elimination,
and intramolecular rotation mechanisms for the isomerization between 4
and 5 are shown in the Supporting Information.
Isomerization between 4 and 5. The initial N-nitrosated-S-
nitrosothiol cation exists as an equilibrating mixture of rotamers
9
J. AM. CHEM. SOC. VOL. 127, NO. 31, 2005 10921

A R T I C L E S
Zhao et al.
Figure 5. Calculated structure of disulfide, nitrosothiol-coordinated nitrosonium with the CBS-QB3 method.
C: 180°), in contrast to the neutral disulfide which adopts the
typical C-S-S-C dihedral of (90°.19 The radical cation also
shows a slightly shorter S-S bond (0.05 Å) than the neutral
disulfide.
Experimental Study of NO+-Catalyzed Decomposition:
The stability of n-hexyl nitrosothiol was examined in the
presence of various sources of nitrosonium. As suggested by
calculations, nitrosothiol decomposition is initiated catalytically
by strong sources of nitrosonium (NOBF4, NOPF6), but not by
alkyl nitrites. In the absence of NO+, a 25 mM solution of C6H13-
SNO in degassed acetonitrile decomposes only very slowly at
room temperature (Figure 6a). The addition of 2 mM (8 mol
%) NOBF4 to a sealed reaction vessel resulted in a rapid initial
decomposition: roughly half of RSNO is lost in 20 min. The
reaction does not, however, proceed to completion, and fol-
lowing this initial burst, further decomposition proceeds at a
rate roughly equivalent to that observed in the absence of
nitrosonium catalyst. The chain reaction proposed in Scheme 1
predicts that excess neutral nitric oxide will ultimately inhibit
nitrosothiol decomposition. In good agreement with this mech-
anism, a similar reaction continually sparged with helium to
remove NO results in clean and complete decomposition (Figure
6c). The time course of the decomposition was modeled
according to the catalytic decomposition mechanism of Scheme
1 using the REACT software. Eyring theory predicts second-
order rate constants of 0.40 M-1 s-1 and 8.5 × 106 M-1 s-1 for
displacement of the NO dimer from nitrosated nitrosothiol and
nitrosation of nitrosothiol by nitrosated disulfide, given calcu-
lated activation free energies of 18 and 8 kcal mol-1, respec-
Figure 6. Kinetics of nitrosonium-catalyzed decomposition of RSNO (R:
hexyl, 25 mM) in degassed acetonitrile: (a) in the absence of NO+, (b) in
the presence of 2 mM NOBF4, no special effort to remove NO during the
decomposition, (c) 2 mM NOBF4, continually sparging to remove NO, and
(d) in the presence of 2 mM NO+, using simulation program (REACT),
based on the current mechanism (Figure 6). Note: the difference between
(c) and (d) may be due to monitor-starting-time error.
tively. The calculated and observed decomposition profiles show
remarkable agreement (Figure 6).
The rate-determining step of the catalytic decomposition is
the nucleophilic displacement of the NO dimer from the cationic
nitrosated nitrosothiol. The activation barrier for this conversion
via an associative mechanism is calculated to be solvent-
dependent, with the rate of reaction diminishing with increasing
solvent polarity. To test this hypothesis, the decomposition of
n-hexyl nitrosothiol was initiated with 8 mol % NOBF4 in
toluene (ꢀ ) 2.38), chloroform (ꢀ ) 4.81), and acetonitrile (ꢀ
) 36.7). In contrast to the predicted solvent effect, rates of
decomposition increase with increasing solvent polarity (Figure
(18) Attempts to achieve the diradicaloid transition state failed for the monotonic
uphill potential surface. The estimated barrier for decomposition should
be slightly lower than the reaction enthalpy of 10 kcal/mol because of
favorable entropy changes.
(19) cis-RSSR•+ (R ) Me) is slightly higher energy than trans- (see Supporting
Information).
9
10922 J. AM. CHEM. SOC. VOL. 127, NO. 31, 2005

NO+-Catalyzed Decomposition of S-Nitrosothiols
A R T I C L E S
• At least one of the potentially rate-determining steps
involves a sterically congested nucleophilic displacement reac-
tion, consistent with the general observation that tertiary
nitrosothiols are more stable than their primary and secondary
counterparts.
• Many nitrosothiols show strongly diminished stabilities at
low pH; such conditions could favor the proteolytic release of
nitrosonium, inducing catalytic decomposition.
The chemistry of nitrosothiols is rich and complex, and the
mechanism of nitrosothiol decomposition is almost certainly
highly dependent on reaction conditions. The catalytic pathway
proposed here likely competes with unimolecular homolytic
cleavage and other catalyzed decomposition mechanisms,
including metal-catalyzed and photolytic processes.
Computational Methodology
Figure 7. Solvent effects on n-hexylnitrosothiol decomposition, in the
presence of 2 mM NOBF4, continually sparging to remove NO: (a) in
degassed acetonitrile, (b) in chloroform-acetonitrile mixed solvent (v:v )
14:1), (c) in toluene-acetonitrile mixed solvent (v:v ) 14:1). Note: in (b)
and (c), add stock NOBF4/acetonitrile solution into chloroform and toluene
solutions of nitrosothiol, respectively.
Quantum mechanical calculations were carried out with density func-
tional theory using the B3LYP density functional theory and the 6-31+G*
basis set20 in Gaussian98 and Gaussian03.21 All stationary points were
verified by harmonic vibrational frequency analysis at the B3LYP level.
These stationary points were reoptimized using the CBS-QB3 method
by Petersson et al.,22 known to have a mean error of 1 kcal mol-1 for
the G2 experimental data set. The free energies in solvent shown were
calculated at the B3LYP/CBSB7 level using CPCM with Pauling-
(Merz-Kollman) atomic radii9 and added to the gas-phase CBS-QB3
energies. All free energies are given for 298 K, 1 M concentrations for
solution phase results and at 1 atm pressure for gas-phase calculations.
Kinetics of Nitrosonium-Catalyzed Decomposition of RSNOs: All
reagents (acetonitrile, chloroform, toluene, 1-hexanethiol, tert-butyl-
nitrite, NOBF4) were purchased from Aldrich Chemical Co. and used
without further purification. S-Nitroso-n-hexanethiol was prepared in
situ by nitrosation of n-hexanethiol with tert-butylnitrite.
Kinetic studies were performed using an HP 8453 UV-vis spec-
trophotometer. All reagents and solutions were transferred with gastight
syringes. Solvents were rigorously degassed either by sparging (He)
or by freeze/thaw. The time courses of reactions were followed by
monitoring the disappearance of RSNO absorbance at either 550 or
551 nm. A background correction was made by subtracting the
absorbance at 700 or 800 nm. Sparging experiments were done in 100-
mL round-bottom flasks, while sealed experiments were done in 3.5-
mL screw-top cuvettes lined with PTFE-rubber septa.
In a typical sparging experiment, a flask was fitted with stir bar and
rubber septum and then flushed with N2. Solvent was added, followed
by 1-hexanethiol and TBN, and sparging (He) was initiated. An initial
absorption reading was taken, followed by addition of NOBF4 stock.
At various time points, 1-mL aliquots of the reaction mixture were
removed from the flask for UV-vis analysis.
In a typical sealed experiment, a 3.5-mL screw-top cuvette was
flushed with N2. Solvent was added, followed by 1-hexanethiol and
TBN. An initial UV-vis absorption reading was taken, followed by
addition of NOBF4 stock. At various time points, absorptions were
recorded using the same cuvette.
7). This observation is likely the result of opposing solvent
effects on the two chain steps. On one hand, the rate-determining
nucleophilic displacement of the NO dimer from nitrosated
nitrosothiol should be enhanced in low-polarity solvents,
predicting a diminishing rate of reaction over the series toluene,
chloroform, and acetonitrile. On the other hand, nitrosonium
transfer from nitrosated disulfide to nitrosothiol is enhanced in
polar solvent, raising the concentration of the key N-nitroso-
S-nitrosothiol intermediate. These two effects apparently very
nearly cancel, resulting in similar rates of decomposition across
the solvent spectrum. Additionally, stable solvent complexes
may account for deviations from the solvent predicted depen-
dence. Such complexes have been observed in other aromatic
solvents and are presumably important during the toluene
decomposition reaction. Finally, all calculations were conducted
only for the methyl-substituted thiol; corresponding values for
other nitrosothiols may vary slightly. The most significant
changes upon going from our MeSNO model system to
biological RSNOs are expected to stem from varying solvation
energy contributions depending on the substituents.
Conclusion
Despite the phenomenologically well described decomposition
of nitrosothiols to nitric oxide and disulfide, a plausible
molecular mechanism for this transformation under oxidative
conditions is lacking. Here, we have proposed a nitrosonium-
catalyzed process with both a predicted and an experimental
free energy of activation of roughly 18 kcal mol-1. The proposed
mechanism rationalizes at least some of the phenomenology of
nitrosothiol spontaneous decomposition:
• The scavenging effect of free reduced thiol on the
nitrosonium catalyst is consistent with the observation that thiol
stabilizes nitrosothiol solutions.
• The proposed mechanism is strongly dependent on nitro-
sothiol concentration, in keeping with the general observation
that nitrosothiols are less stable in concentrated solution.
• The reaction is dependent on the ratio of concentrations of
nitrosothiol, disulfide, nitric oxide, and nitrosonium present in
the solution, consistent with the frequent observation of non-
integral rate laws for decomposition.
Curve Simulation. The REACT for Windows simulation program
was downloaded free of charge from Alchemy Software (http://
chemicalsoft.com/). Rate constants for the two propagation steps were
derived by Eyring analysis using calculated free energies of activation.
The initial RSNO concentration was set at 25 mM, and the initial RS-
(NO)NO+ concentration was set at 2 mM.
(20) Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652.
(21) Frisch, M. J. et al. Gaussian 98 and 03; Gaussian, Inc., Pittsburgh, PA.
(22) Montgomery, J. A.; Frisch, M. J.; Ochterski, J. W.; Petersson, G. A. J.
Chem. Phys. 1999, 110, 2822-2827. (U)B3LYP/6-31+G*, (U)B3LYP/
CBSQB7, and CBS-QB3 methods result in similar energetic patterns for
the NO+-catalyzed mechanism; the DFT methods underestimated BDE and
activation energy by 2-3 kcal mol-1, comparing to CBS-QB3, which has
been shown to work well for similar species previously.
9
J. AM. CHEM. SOC. VOL. 127, NO. 31, 2005 10923

A R T I C L E S
Zhao et al.
Acknowledgment. We are grateful to the National Institute
of General Medical Sciences, National Institutes of Health, for
financial support of this research.
Supporting Information Available: Optimized structures and
energies for all species discussed, the 16 SN2 transition states
using the B3LYP/6-31+G* and B3LYP/CBSB7 methods, the
intramolecular rotation transition state for conversion between
4 and 5, the experimental n-C3H7SNO decay time in different
solutions, and complete ref 21 (PDF). This material is available
free of charge via the Internet at http://pubs.acs.org.
Note Added after ASAP Publication. After this article was
published ASAP on July 15, 2005, a typographical error
occurring in both Figures 3 and 4 was corrected. The corrected
version was published ASAP on July 19, 2005.
JA050018F
9
10924 J. AM. CHEM. SOC. VOL. 127, NO. 31, 2005