Dependence of product formation from decomposition of nitroso-dithiols on the degree of nitrosation. Evidence that dinitroso-dithiothreitol acts solely as an nitric oxide releasing compound

View Article Online / Journal Homepage / Table of Contents for this issue

PAPER

www.rsc.org/obc | Organic & Biomolecular Chemistry

Dependence of product formation from decomposition of nitroso-dithiols on

the degree of nitrosation. Evidence that dinitroso-dithiothreitol acts solely as

an nitric oxide releasing compound

Sonja Liebeskind,^aHans-Gert Korth,^bHerbert de Groot^aand Michael Kirsch*^a

Received 29th January 2008, Accepted 31st March 2008

First published as an Advance Article on the web 7th May 2008

DOI: 10.1039/b801583j

Hitherto, the decay mechanisms of nitrosated dithiols as well as formation of related products have not

been conclusively elucidated. In this paper, we demonstrate that nitrosated DL-dithiothreitol (DTT)

decays via two independent pathways, that is, one producing exclusively nitric oxide and one producing

(initially) nitroxyl (HNO/³NO⁻). The importance of the two decomposition pathways depends on the

degree of nitrosation of DTT. Dinitroso-dithiothreitol (NODTTNO) generates quantitatively nitric

oxide, whereas mononitroso-dithiothreitol (NODTT) yields initially nitroxyl. Since NODTT and DTT

are both targets for nitroxyl, their availability governs the HNO-derived formation of nitric oxide (with

NODTT as reactant) or hydroxyl amine and ammonium ion (with DTT as reactant). The formation of

+

NH₄from the HNO–DTT reaction probably proceeds by a stepwise, NH₂OH-independent

mechanism, because DTT-derived sulﬁnamide was identiﬁed by N-15 NMR spectrometry as an

intermediate. Our data are in line with the assumption that triplet nitroxyl (³NO⁻) is formed by a

unimolecular decay of the deprotonated (thiolate) form of NODTT, because CBS-QB3 calculations

predict the existence of a low-lying triplet state of the latter species. The identiﬁed pathways are

proposed to be of general importance for physiological systems because control experiments showed

that the physiological dithiol thioredoxin reacts in a similar manner.

skeletal and cardiac isoforms of the ryanodine receptor, which

Introduction

suggests that ^•NO modulates muscle activity.^10,11

Nitric oxide (^•NO), which is an important mediator in

vasodilatation,¹neurotransmission,²and immune reactions, is a

The trans-nitrosation reaction competes with the thiol-mediated

reduction of the nitroso-function to NH₃at excess thiol

concentrations,¹²reaction (2).

short-lived intermediate in blood because it is rapidly trapped by

the heme iron of hemoglobin. Therefore, it has been suggested

that in vivo nitric oxide uses so-called “transporters” in order to

prolong its life-time.³Thiols have been suggested to be involved in

^•NO transport, and protein-bound cysteine residues may regulate

activities of a variety of enzymes after S-nitrosation in a fashion

that parallels protein phosphorylation and dephosphorylation.^4,5

For instance, experimental evidence has been found that both ex-

ogenously and endogenously produced ^•NO impedes apoptosis via

S-nitrosation of caspases.^6–8One characteristic chemical feature of

S-nitrosothiols is the transfer of the nitroso-function to another

thiol (“transnitrosation”), reaction (1).

2 RSNO + 8 RSH →→→ 5 RSSR + 2 H₂O + 2 NH₃

(2)

In addition, Wong et al.¹³reported that the reaction between

thiols and S-nitrosothiols may yield nitroxyl (HNO), reaction (3),

R¹SNO + R²SH → R¹SSR²+ HNO

(3)

•

which further produces NO by reaction with a second molecule

of S-nitrosothiol, reaction (4).

RSNO + HNO → RSH + 2 ^•NO

(4)

However, the situation is less clear when the nitroso function

is transferred to a dithiol in which the two thiol groups can

interact with each other. The S-nitrosoglutathione (GSNO)-

dependent nitrosation of the physiological dithiols dihydrolipoic

acid (DHLA) and thioredoxin (Trxn) has been studied by several

research groups, as overviewed by Stoyanovski et al.¹⁴However,

conﬂicting reaction mechanisms have been proposed for this

reaction (Scheme 1).

R¹SNO + R²SH ꢀ R¹SH + R²SNO

(1)

By this reaction, low-molecular weight S-nitrosothiols can

nitrosate cysteine residues in proteins. For instance, S-

nitrosocysteine inhibits responses mediated by the N-methyl-D-

aspartate subtype of the glutamate receptor of rat cortical neurons,

presumably via trans-S-nitrosation of the channel.⁹Further, low-

molecular weight RSNOs trans-S-nitrosate vital thiols in both

Arnelle and Stamler¹⁵reported that DHLA and the frequently

used dithiol dithiothreitol (DTT) reduce GSNO to glutathione

(GSH) and nitroxyl (HNO) via the intermediate formation of

mononitroso-dihydrolipoic acid (1 → 2 → 3). This view was

supported by data of Stoyanovsky et al.¹⁴On the contrary,

Petit et al.¹⁶suggested that 2 decomposes to ^•NO and the

^aInstitut fu¨r Physiologische Chemie, Universita¨tsklinikum Essen, Hufeland-

strasse 55, 45122, Essen, Germany. E-mail: michael.kirsch@uni-duisburg-

essen.de; Fax: +49 201-723 5943; Tel: +49 201-723 4108

^bInstitut fu¨r Organische Chemie, Universita¨t Duisburg-Essen, Univer-

sita¨tsstrasse 5, 45117, Essen, Germany

2560 | Org. Biomol. Chem., 2008, 6, 2560–2573

This journal is

The Royal Society of Chemistry 2008

©

View Article Online

dithiothreitol as a model, in the absence and presence of certain

targets, as a degree of nitrosation. These results are presented here.

In contrast to all previous studies we clearly identify a HNO-

•

dependent and a HNO-independent reaction channel of NO

production. The HNO-independent pathway decisively involves

the intermediacy of dinitroso-dithiols. The thermochemistry of all

reactions is computationally established at the CBS-QB3 level of

theory.

Scheme 1 Decomposition of mononitroso-dihydrolipoic acid according

to Stoyanovski et al.¹⁴

Results

Mono- and dinitroso-DTT show very similar electronic ab-

S-centered radical 4. The latter thiyl radical has been shown to

undergo a rapid intramolecular ring closure to the radical anion

5. Such radicals are known to react with oxygen at the diffusion-

sorptions at k_max= 332 and 336 nm, respectively, with molar

absorbtivities (NODTTNO: e₃₃₆= 1685 M ⁻¹cm⁻¹, NODTT: e₃₃₂

=

−1

825 M cm⁻¹) differing by about a factor of two.²³The current

literature does not give a conclusive answer to the question whether

NODTTNO is only generated after complete mono-nitrosation of

DTT because the nitrosation of the two thiol functions appears

to proceed with the same rate constant.²³Since quantum chemical

calculations performed at the CBS-QB3 level of theory predicted

both that the nitrosation of the second thiol group is a little

more exergonic than the nitrosation of the ﬁrst one and that

the trans-nitrosation reaction between two NODTT molecules is

additionally an exergonic reaction (reaction (5), Table 4, entry 5),

•

controlled limit to yield the superoxide radical anion (O₂⁻).^17–19

Similar to the conversion 2 → 4, Niktovic and Holmgren¹⁹

reported that the key physiological dithiol thioredoxin (Trxn)

catalyzes the denitrosation of GSNO with release of ^•NO via the

intermediate formation of mononitroso-thioredoxin (HS-Trxn-

SNO). However, the occurence of reaction 2 → 4 is in question

as S-nitrosothiols such as GSNO do not spontanousely homolyze

in the absence of any Cu²⁺/Cu⁺-ions.^20,21Somewhat surprisingly,

•

the suggestion of Wong et al.¹³that HNO releases NO from S-

nitrosothiols via reaction (4) has not been taken into account

for mononitrosated dithiols, although this reaction apparently

explains the occurrence of both HNO and ^•NO.

In contrast to all the foregoing conclusions, Brock et al.²²

reported that S,S^ꢀ-dinitrosated dithiols, but not mononitrosated

ones, exhibited pharmacological effects typical for nitric oxide.

In physiological systems, clean mono- or dinitrosation of the

available dithiols cannot be expected. From the above, it is obvious

that the way in which the degree of nitrosation of the dithiols

correlates with their chemical activity was unresolved so far.

Therefore, we elucidated the chemical pathways of decomposi-

tion of nitrosated dithiols (see Scheme 2), using primarily DL-

(5)

NODTTNO is expected to be present even in such nitroso-DTT

solutions which were generated at a [NO₂⁻]/[DTT] ratio lower

than 2.

Nitrosation was performed by reacting constant concentrations

of DTT (60 lM) with varying amounts of sodium nitrite at

pH 2 and rapidly adjusting the pH to 7.4 in order to slow

down the spontaneous decomposition of nitrosated DTT. Since

the decomposition at pH 7.4 is still relatively fast (see below),

quantiﬁcationofthe yields of nitrosated species was not attempted,

and the degree of nitrosation was expressed as the [NO₂⁻]/[DTT]

ratio, assuming almost quantitative conversion of the minor

reactant at all [NO₂⁻]/[DTT] ratios.

The validity of this assumption is demonstrated by Fig. 1A,

showing an excellent linear dependence of the optical density at

336 nm on the [NO₂⁻]/[DTT] ratio, with a slope of 0.487 for

[NO₂⁻]/[DTT] values ≤ 2. This would correspond to a nitrosation

yield of 97.5

1%. At [NO₂⁻]/[DTT] values ≥ 2, the optical

density remained constant, clearly demonstrating that both thiol

functions were essentially quantitatively nitrosated under the

applied experimental conditions. The latter ﬁnding is in line with

data of Brock et al.²²Therefore, such solutions were generally

applied for all further experiments.

The oxidation of dihydrorhodamine-123 (DHR-123) is, in the

absence of catalytically active transition metals, a known test for

the intermediacy of oxidizing nitrogen-oxide species.^24,25From

Fig. 1B it is understood that with increasing [NO₂⁻]/[DTT]

ratio the oxidation of DHR-123 increased as well, until at

[NO₂⁻]/[DTT] = 2 a constant yield of rhodamine-123 is reached,

i.e., the same dependence as has been observed for the optical

Scheme 2

This journal is

The Royal Society of Chemistry 2008

Org. Biomol. Chem., 2008, 6, 2560–2573 | 2561

©

View Article Online

Fig. 2 Nitrosation of DAF-2 by in situ generated S-nitroso DTT. Mix-

tures of 60 lM GSNO and the corresponding concentration of DTT were

allowed to react with 10 lM DAF-2 for 180 min at 37 ^◦C in the dark.

In the absence of dithiol, GSNO is rather inefﬁcient in convert-

ing DAF-2 to DAF-2T because only 0.14 0.03 lM DAF-2T was

detected. Since the nitric oxide-releasing decomposition of GSNO

requires catalysis by transition metal ions (preferably copper and

iron),²⁰this low yield can be attributed to efﬁcient complexation

of traces of such metal ions by the added EDTA. Nitrosation of

DAF-2 increased with increasing dithiol concentration, until a

[GSNO]/[DTT] ratio of 2 is reached, that is, at the same ratio as

was found from preformed nitroso-DTT (see Fig. 1). At higher

[GSNO]/[DTT] ratios the ﬂuorescence of DAF-2T decreased

slightly, the reason for which was not further explored.

The foregoing observations strongly suggested that NODTTNO

acts as a general nitrosating agent in the presence of oxygen.

For further corroboration, the nitrosation of N-acetyltryptophan

(NAT) by dinitroso-DTT was studied by H-1 NMR (Fig. 3).

Decomposition of NODTTNO was complete after a reaction

period of 90 min (vide infra). The H-1 NMR spectrum then showed

only the presence of N-nitroso-N-acetyltryptophan (NANT)

and some residual NAT, thus clearly proving the capability of

NODTTNO to nitrosate NAT.

Fig. 1 Degree of DTT nitrosation. A: Dependence of the optical density

at k_max= 336 nm on the degree of DTT nitrosation expressed as

[NO₂⁻]/[DTT] ratio. The absorbance was measured immediately after

mixing the reactants ([DTT] = 60 lM) and adjusting the pH to 7.4 (T =

25 ^◦C). B: Oxidation of dihydrorhodamine-123 as a function of DTT

nitrosation expressed as [NO₂⁻]/[DTT] ratio. Mixtures of solutions of

preformed nitroso-DTT (25 lM) and DHR-12 (50 lM) were allowed to

react for 180 min at 37 ^◦C.

density (Fig. 1A) is followed. This observation was somewhat

unexpected, because hitherto it was generally believed that the

reactive species responsible for oxidation of DHR-123 were

exclusively released from NODTT.^{14–16,19,26}

In order to verify a similar chemical activity of in situ nitrosated

dithiols, the GSNO-mediated nitrosation of the primary aromatic

diamine 4,5-diaminoﬂuorescein (DAF-2) in the presence of DTT

was studied (Fig. 2). Formation of the corresponding ﬂuorescent

triazole product DAF-2T indicates the intermediacy of ^•NO

whose autoxidation in the presence of air produces the nitrosating

species ^•NO₂/N₂O₃.

In order to ﬁnd evidence for the kind of the reactive intermedi-

ates whichareresponsiblefor thenitrosating/oxidizingcapabilities

of NODTTNO, the yields of products from reaction with NAT,

melatonin, DAF-2, DHR-123, aminophenyl ﬂuorescein (APF),

and hydroxyphenyl ﬂuorescein (HPF), respectively, were quanti-

ﬁed (Table 1). Possible reactive intermediates for conversion of

these targets as discussed in the literature are also listed.

Nitrosation of NAT, melatonin, and DAF-2 by NODTTNO

proceeded uniformly with a yield of about 22%. It should be noted

that all experiments were performed in phosphate buffer which is

known to catalyse the hydrolysis of N₂O₃,³⁰and this fact might

Table 1 Nitrosation/oxidation by dinitroso-dithiothreitol

Target

Substrate nitrosation/oxidation (%)

Possible intermediates

References

NAT

23.7 7.5

21.6 1.5

22.5 0.6

6.0 0.3

1.6 0.1

0.0 0.0

N₂O₃, HNO

24,27

25,28

29

Melatonin

DAF-2

DHR-123

APF

N₂O₃, HNO

^•NO₂, HNO, ONOOH

HOCl, ONOOH

•

HO^•, CO₃

29

−

HPF

2562 | Org. Biomol. Chem., 2008, 6, 2560–2573

This journal is

The Royal Society of Chemistry 2008

©

View Article Online

Fig. 3 H-1 NMR analysis. H-1 NMR (500-MHz) detection of nitrosation of N-acetyltryptophan (NAT) by NODTTNO (bottom). The spectrum (region

of aromatic resonances) was recorded immediately after mixing of preformed NODTTNO (10 mM) and NAT (10 mM). Spectra of preformed NODTTNO

(10 mM), N-nitroso-N-acetyltryptophan (NANT) (5 mM) and NAT (10 mM), respectively, are shown for comparison.

have limited the yields of the nitrosation products. Oxidation of

DHR-123 and APF was signiﬁcantly lower, and no oxidation was

observed for HPF. APF and HPF are only converted to ﬂuorescein

by strongly oxidizing species like peroxynitrite and the hydroxyl

radical. Since hydroxyl radicals can be generated from HNO,³¹the

failure of NODTTNO to oxidize HPF disfavours the intermediacy

of nitroxyl. Thus, the product yields of Table 1 strongly suggest

the intermediacy of N₂O₃.

presence of NAT, even though NANT is produced (monitored

photometrically at k_max= 335 nm; data not shown). However, in

the absence of NAT, the decay of NODTTNO does not follow a

clean ﬁrst-order rate law as evident from the logarithmic plot of

the decay data. At longer reaction times, the decay of NODTTNO

appears to be somewhat retarded, probably due to a second-order

component. In fact, the best ﬁt of the data points was obtained

by assuming a superposition of a ﬁrst-order and a second-order

decay process. Noteworthy, in the presence of NAT, the decay

of NODTTNO excellently followed a single ﬁrst-order rate law

In order to discriminate whether NODTTNO nitrosates its

targets via a reactive intermediate (N₂O₃) or directly by a

bimolecular trans-nitrosation process, the inﬂuence of NAT on the

decay of NODTTNO was monitored. The decay of NODTTNO

was followed by reading its optical density photometrically at k =

545 nm in order to avoid interference with the formation of NANT

(k_max= 335 nm).

with a rate constant of k₁= (4.91

0.07) × 10⁻³s⁻¹(298 K),

corresponding to a half-life of 2.3

0.1 min. The fact that in

the presence of NAT, NODTTNO decays by a ﬁrst-order process

clearly indicates that the NODTTNO-mediated nitrosation of

NAT proceeds via a reactive intermediate (vide infra), most

reasonably N₂O₃. The intermediacy of N₂O₃also would explain

the retarded decay in the absence of NAT (Fig. 4) as N₂O₃

is known to catalyze the decay of mono-S-nitrosothiols.³²The

catalysis of decomposition of only one S-nitrosothiol group then

inhibited the alternative (faster) decay, i.e., the simultaneous bond

breaking of the two RS–NO bonds in NODTTNO (see Discus-

sion).

Inspection of Fig. 4 reveals that the decay of NODTTNO,

with a half-life of about 2.4 min, is only slightly affected by the

Since NANT is highly effective in nitrosating thiols,³³the decay

of NODTT ([NO₂⁻]/[DTT] = 1) was only measured in the absence

of NAT. The decay of NODTT has been previously analyzed by

Arnelle and Stamler¹⁵who reported a half-life of ca. 9 min at

pH 7.4 and room temperature. We observed a moderately faster

decay with a half-life of about 7 min at pH 7.4 and 25 ^◦C (Fig. 5A).

However, as evident from the regression line, the decay of NODTT

is not of pure ﬁrst-order. The decay was also dependent on the pH

value, showing a maximum around pH 9 (Fig. 5B). This ﬁnding

indicated that the decomposition of NODTT depended on the

availability of the deprotonated thiol (thiolate) form. The increase

of the half-life time at higher pH values is discussed below on the

basis of a quantum-chemically supported mechanism.

Fig. 4 Kinetics of NODTTNO decay. Decay of NODTTNO (2 mM)

in the absence (closed circles) and in the presence (open circles) of

N-acetyltryptophan (5 mM) at pH 7.4 (T = 25 ^◦C). The optical density

of NODTTNO was read at 545 nm. The solid lines through the left-hand

data points are least-squares ﬁts to a ﬁrst-order rate law (open circles) and

a linear combination of a ﬁrst- and second-order rate law (closed circles).

In order to conﬁrm the proposed release of HNO as the reac-

tive intermediate from NODTT,³we applied a Mn^III–porphyrin

This journal is

The Royal Society of Chemistry 2008

Org. Biomol. Chem., 2008, 6, 2560–2573 | 2563

©

View Article Online

Although NODTTNO does not release any HNO during

decomposition (Fig. 6E and F) but is nevertheless more reactive

than NODTT (compare, e.g., Fig. 1B), an intermediate other than

nitroxyl has to be released. Petit et al.¹⁶claimed ^•NO to be the main

intermediate from decomposition of NODTT. Therefore, ^•NO was

measured polarographically from preformed and in situ generated

nitroso-DTT, prepared at [NO₂⁻]/[DTT] or [GSNO]/[DTT] ratios

of 0.25–2. As the enzyme superoxide dismutase (SOD) is supposed

•

to increase the yield of NO in the presence of HNO³⁷as

•

well as from ^•NO/O₂⁻-releasing systems,³⁸these experiments

were performed in the absence and in the presence of SOD

(100 units ml⁻¹). (Note that the foregoing experiments were carried

out in an open experimental setup, thus only the equilibrium

concentration of ^•NO with respect to diffusion into the gas phase

and autoxidation could be determined.)

As shown in Fig. 7, a maximum equilibrium concentration

•

of NO of 12.4 lM was released from preformed nitroso-DTT

•

synthesized at a [NO₂⁻]/[DTT] ratio = 2. The NO equilibrium

concentration decreased with decreasing [NO₂⁻]/[DTT] ratio to

about 50% at a ratio of 0.25. Noteworthy, at this ratio the presence

of SOD increased the ^•NO equilibrium concentration from 6.2 lM

to 8.3 lM, thus indicating the intermediacy of either HNO or

•

O₂⁻. In order to discriminate between these two intermediates,

the H₂O₂concentration was measured in the absence and in the

presence of SOD, because only a SOD-catalyzed dismutation of

•

O₂⁻would lead to an increased level of H₂O₂. However, the H₂O₂

concentration in the presence of SOD (100 units ml⁻¹) (15.4

1.6 lM) was within the error margins the same as in the absence

of SOD (17.4 0.7 lM). Thus, the SOD-mediated increase in the

^•NO equilibrium concentration most likely derived from reaction

Fig. 5 Kinetics of NODTT decay. Decay of NODTT (2 mM, plus 100 lM

EDTA) at 25 ^◦C. A: Decay trace at pH 7.4. The optical density was

monitored at k_max= 336 nm. Solid line: least-squares ﬁt to a ﬁrst-order

rate law. B: pH dependence of the half-life of NODTT (2 mM). The half-life

values were estimated from simple ﬁrst-order ﬁts.

•

with nitroxyl. The inﬂuence of SOD on the equilibrium NO

concentration was even more pronounced when nitroso-DTT was

in situ generated from the GSNO-DTT reaction, despite the fact

that in this case the equilibrium concentration of nitric oxide was

generally lower than from preformed nitroso-DTT (Fig. 7). These

complex (manganese(III)-tetrakis(1-methyl-4-pyridyl)porphyrin

pentachloride = Mn^IIITMPyp) as a scavenger for nitroxyl as

proposed by Marti et al.,³⁴who used a related Mn^III–porphyrin

derivative, manganese(III)meso-tetrakis-(N-ethylpyridinum-2-yl)-

porphyrin, for the detection of nitroxyl. The HNO-trapping capa-

bility of Mn^IIITMPyp was tested with the HNO donor Angeli’s salt

•

results further conﬁrmed that both NO and HNO are released

when DTT is only partially S-nitrosated.

To clarify whether a physiological dithiol can act in a sim-

ilar manner, a few experiments were performed with nitroso-

thioredoxin (nitroso-Trxn) (56.7 lM). Since preformed nitroso-

Trxn cannot be synthesized at pH values lower than 4, it was

(AS). On addition of AS, the Soret band of Mn^IIITMPyp (e₄₆₃

=

9.2 × 10⁴M⁻¹cm^{−1 35}) decreased instantaneously with concomitant

build-up of the absorption of the nitroso complex Mn^IITMPyp-

NO at k_max= 435 nm (e₄₃₅= 1.1 × 10⁵M⁻¹cm^{−1 36}) (Fig. 6A).

Nitroso-DTT preparations at various [NO₂⁻]/[DTT] ratios were

analyzed.

−

synthesized at pH 5.4 by using an excess amount of NO₂

(133.4 mM). Release of ^•NO was initiated by adjusting the solution

to pH 7.4. From such a solution, an equilibrium nitric oxide

concentration of 2.9 1.5 lM was determined, which corresponds

to about 27% of the level that was achieved from DTT under

the same conditions (^•NO production could not be detected in

the absence of dithiol, thus ruling out inadvertent ^•NO formation

from nitrite via the HNO₂/N₂O₃pathway). A similar relationship

was found when the related nitroso-dithiols were generated

in situ at pH 7.4 from the reaction of GSNO (250 lM) with Trxn

or DTT (55.6 lM each). Here, the Trxn-dependent nitric oxide

equilibrium concentration (1.21 0.1 lM) amounted to 20% of

that produced by DTT. Due to the presence of EDTA (100 lM), a

GSNO-dependent ^•NO release was not detected in the absence of

dithiols. Thus, dinitroso-Trxn also releases nitric oxide at pH 7.4,

although the yields and kinetics differ from that of DTT.

In fact, an identical shift of the Soret band was observed in the

presence of preformed nitroso-DTT synthesized at [NO₂⁻]/[DTT]

ratios of 0.25, 0.5, and 1 (Fig. 6B–D), which indicated the release

of HNO during the decay of nitroso-DTT. Most remarkably,

preformed NODTTNO did not cause any change of the 463-nm

absorption (Fig. 6E). The production of HNO as a function of

nitrosation of DTT (400 lM) was analyzed in more detail after a

reaction period of 300 s (Fig. 6F). A constant amount of about

3.9 lM nitroxyl was trapped from [NO₂⁻]/[DTT] ratios 0.75–1.25.

The amount of trapped nitroxyl decreased to about 50% at a

[NO₂⁻]/[DTT] ratio of 1.5, and the Mn^IIITMPyp assay failed to

detect any nitroxyl at [NO₂⁻]/[DTT] ratios ≥ 1.75. These data

clearly demonstrate that the type of intermediate released from

nitroso-DTT crucially depends on the level of nitrosation of DTT.

With respect to the suggested mechanism of HNO formation

from mononitroso-dithiols (Scheme 1, reaction (2) → (3)), one

2564 | Org. Biomol. Chem., 2008, 6, 2560–2573

This journal is

The Royal Society of Chemistry 2008

©

View Article Online

Fig. 6 Determination of HNO release from Angeli’s salt and nitroso-DTT. A: Addition of AS (400 lM). B, C, D, E: Addition of preformed nitroso-DTT

(500 lM) from various [NO₂⁻]/[DTT] ratios (at a ﬁxed NO₂-concentration). F: Amount of trapped nitroxyl after a reaction period of 90 s from preformed

−

nitroso-DTT (500 lM) that were generated from various [NO₂⁻]/[DTT] ratios (at a ﬁxed NO₂concentration).

would expect that per mole released HNO one mole of thiol

function is consumed. In order to check on this point, the

concentration of the thiol functions of preformed nitroso-DTT

prepared from various [NO₂⁻]/[DTT] ratios (at a ﬁxed amount of

DTT) was quantiﬁed (Fig. 8).

Fig. 7 Release of ^•NO. Equilibrium ^•NO production from preformed

and in situ generated nitroso-DTT in the absence (closed symbols) and

presence (open symbols) of SOD. Circles: Preformed ([NO₂⁻]/[DTT])

nitroso-DTT (at a constant concentration of 250 lM SH groups). Squares:

In situ generated nitroso-DTT from reaction of 250 lM GSNO with varying

concentrations of DTT.

Fig. 8 Yields of thiol oxidation. Consumption of thiol functions of

nitroso-DTT depending on the [NO₂⁻]/[DTT] ratio.

This journal is

The Royal Society of Chemistry 2008

Org. Biomol. Chem., 2008, 6, 2560–2573 | 2565

©

View Article Online

In the absence of NO₂⁻, 980

20 lM thiol groups were

to [N-15]NO₂⁻and [N-15]NO₃⁻. Neither typical HNO-derived de-

composition products (N₂O, N₂) nor NH₂OH, a product expected

from reduction of HNO by DTT,¹⁵could be detected (Fig. 9B). So

far, NH₄⁺has not been reported as a product from decomposition

of nitroso-dithiols, but it has been reported from reaction of

GSNO with GSH.¹²At [NO₂⁻]/[DTT] = 0.5, [N-15]NH₄⁺was the

found from 500 lM DTT 30 min after dissolution of DTT. This

observation clearly demonstrates that DTT is not artiﬁcially

consumed under our reaction conditions. In line with experiments

described below and from literature data, all thiol functions were

oxidized after complete decay of preformed nitroso-DTT prepared

at [NO₂⁻]/[DTT] ratios ≥ 1. Surprisingly, however, about 98%

of the thiol groups were already oxidized after decomposition

of nitroso-DTT prepared at a [NO₂⁻]/[DTT] ratio of 0.5, i.e.,

−

major product. Here, the yields of [N-15]NO₂and [N-15]NO₃

were strongly diminished, but new signals of [N-15]NH₂OH (d =

−277 ppm) and a second, unknown product with d = −300.8 ppm

were detected (Fig. 9C). Sulﬁnamides are characterized by N-15

NMR resonances around 300 ppm,³⁹therefore, it was assumed

that this resonance is due to the monosulﬁnamide of DTT

(DTTSONH₂). In order to support this assumption, the N-15

NMR chemical shifts of [N-15]DTTSONH₂and of the other

products were quantum-chemically computed with the individual

gauges for atoms in molecules (IGAIM) protocol at the B97-

2/aug-cc-pVDZ level of theory. Since the reliability of the applied

model to predict N-15 NMR resonances is not well reported in

the literature, a small set of N-15 NMR resonances was selected

as benchmark for comparing the IGAIM-B97-2/aug-cc-pVDZ

method (Table 2).

986

4 lM thiol were consumed by 250 lM NO₂⁻, whereas a

thiol depletion of only 500 lM had to be expected in the case

of an exclusive HNO-producing pathway. A similar situation

was observed for a [NO₂⁻]/[DTT] ratio of 0.25, because now

125 lM nitrite mediated the oxidation of 368

11 lM thiol,

instead of 250 lM theoretically. These observations made clear

that the reaction mechanism is more complicated than hitherto

assumed.^3,14,16This further suggests that the enhanced oxidation

of thiols may be accompanied by the release of a nitrogen species

other than ^•NO and HNO, respectively.

In order to identify the nitrogen products from preformed

nitroso-DTT, N-15-enriched nitrite was applied at [NO₂⁻]/[DTT]

ratios of 2, 1, 0.5, and 0.25, and the reaction mixtures were ana-

lyzed by N-15 NMR spectrometry after complete decomposition

of nitroso-DTT (Fig. 9A–D).

The applied model predicted N-15 NMR resonances with a

mean absolute deviation of 8.3 ppm and the N-15 chemical shift

of [N-15]DTTSONH₂is calculated to be 295.8 ppm, i.e., 5 ppm

lower than the N-15 NMR resonance of the unkonwn compound

(Table 2). Thus, the calculations were not in conﬂict with the

assignment of the resonance at −300.8 ppm to [N-15]DTTSONH₂.

At a 4-fold excess of DTT, i.e. at [[N-15]NO₂⁻]/[DTT] = 0.25

From fully nitrosated DTT (NODTTNO; [[N-15]NO₂⁻]/

[DTT] = 2) only [N-15]nitrite (d = 229 ppm) and [N-15]nitrate

(d = −4.2 ppm) were detected after decomposition at pH 7.4

(Fig. 9A). At [[N-15]NO₂⁻]/[DTT] = 1, viz. NODTT, a prominent

+

peak of [N-15]NH₄(d = −359.3 ppm) was recorded in addition

(Fig. 9D), only [N-15]NH₂OH and [N-15]NH₄were found as

Fig. 9 N-15 NMR spectrometric product analysis of decomposed nitroso-DTT solutions. Preformed nitroso-DTT at various [[N-15]NO₂⁻]/[DTT]

ratios (100 mM [N-15]nitrite, 100 lM EDTA, ﬁnal pH = 7.4) were allowed to decompose for 6 h at room temperature. Chemical shifts (d) are given in

parts per million (ppm) relative to nitromethane (d = 0) as an external standard.

2566 | Org. Biomol. Chem., 2008, 6, 2560–2573

This journal is

The Royal Society of Chemistry 2008

©

View Article Online

Table 2 Quantum-chemically calculated isotropic absolute shielding constants and N-15 chemical shifts (d, in ppm)

Isotropic shielding constant^a

B97-2

Isotropic chemical shift

B97-2

Compound

Exp.⁴⁰

0.0

228.9

184.0

Deviation

[N-15] CH₃NO₂

−125.8

−356.1

−334.4

−280.5

−120.9

−55.3

1.0

0.0

230.3

208.6

−

[N-15]NO₂

+1.4

+24.6

−0.5

−1.4

−0.9

−24.3

+3.0

[N-15]CH₃ONO

[N-15](CH₃)₂NNO

154.7

155.2

−

[N-15]NO₃

−4.9

−3.5

[N-15]N₂

−70.5

−69.6

[N-15]CN⁻

−126.8

−128.5

−148.2

−145.9

−137.2

−189.3

−234.6

−275.9

−294.2

−293.9

−320.3

−367.3

−102.5

−131.5

−135.8

−147.3

−148.7

−174.1

−235.5

−278.1

−280.6

−290.9

−302.9

−359.6

[N-15][⁻N=N⁺=N⁻]

[N-15]CH₃CN

[N-15]NNO

2.7

22.4

20.2

11.4

−12.4

+1.3

[N-15](CH₃)₂NNO

+11.5

−15.2

+0.9

[N-15]SCN⁻

63.5

[N-15]NNO

108.8

150.1

168.4

168.1

194.5

241.5

[N-15]NH₂OH

+2.2

−

+

−

=

[N-15][ N N =N ]

[N-15]CH₃NCS

[N-15]OCN⁻

−13.6

−3.0

−17.4

−7.7

−3.0

8.3

+

[N-15]NH₄

mean deviation

mean abs. deviation

unknown

−300.8

[N-15]DTTSONH₂

170.0

−295.8

^aIsotropic absolute shielding constants were calculated using the IGAIM protocol at the B97-2/aug-cc-pVDZ//B97-2/aug-cc-pVDZ level of theory.

Solvation corrections (nitromethane for nitromethane, acetonitrile for acetonitrile and CH₃ONO, acetone for (CH₃)₂NNO and CH₃NCS, none for NNO,

H₂O for all others) with the CPCM solvation model were performed at the same level of theory.

products. In conclusion, the N-15 NMR spectra conﬁrmed that

the products from nitrosated DTT strongly depend on the degree

of nitrosation of DTT, or, vice versa, the availability of free SH

goups.

to about 2.1% at a [NO₂⁻]/[DTT] ratio of 0.25. The yield of

NH₄changed inversely to the nitrite yield. About equimolar

+

amounts (96 12 lM) of nitrite and ammonium were produced

+

at [NO₂⁻]/[DTT] = 0.5, until the yield of NH₄approached a

Since product yields are difﬁcult to be quantiﬁed by N-15

NMR spectrometry, the foregoing experiments were repeated at

much lower concentrations and three of the four products, namely

limiting value of about 43% (107 lM) at [NO₂⁻]/[DTT] ratios ≤

0.25. Production of NO₃⁻did not exceed 2% at any [NO₂⁻]/[DTT]

ratios. The observed changes of product yields paralleled those

observed in the N-15 NMR experiments. According to the N-15

NMR spectra, the missing nitrogen content is due to formation

of NH₂OH and DTTSONH₂(see Fig. 9A–D,) which were not

independently quantiﬁed. As NH₄⁺is formed from nitroso-DTT as

long as unnitrosated DTT is available, the question arose whether

+

NO₂⁻, NO₃⁻, and NH₄, were independently quantiﬁed (Fig. 10).

Percentage yields are given relative to the concentration of the

thiol groups.

+

NH₄is formed from reaction of nitroso-DTT with DTT or from

reduction of NH₂OH by DTT. In fact, in text books of inorganic

+

chemistry it is mentioned that NH₂OH can be reduced to NH₄.

+

NH₂OH + 2 e⁻+ 3 H⁺→ NH₄+ H₂O

(6)

As DTT is well known for its reducing capabilities, the yields of

+

NH₄from the NH₂OH–DTT reaction were determined by N-15

NMR spectrometry. Experiments were performed with solutions

of [N-15]NH₂OH (200 mM) in the absence and in the presence

of DTT (400 mM). In both cases, however, only the resonance of

[N-15]NH₂OH was detected after a reaction period of 6 h, proving

Fig. 10 Quantiﬁcation of nitrite, nitrate, and ammonium from de-

+

that the yield of [N-15]NH₄must be lower than 3% with respect

composed nitroso-DTT solutions. Nitroso-DTT prepared from various

NO₂⁻/DTT ratios (250 lM SNO function) was allowed to react for 3 h at

to the applied [N-15]NH₂OH concentration (spectra not shown).

25 ^◦C. Closed circles: NO₂⁻, open circles: NH₄⁺, closed diamonds: NO₃

.

Another possibility of NH₄formation would be the reduction

+

−

+

of HNO by DTT. We therefore quantiﬁed the NH₄production

From NODTTNO ([NO₂⁻]/[DTT] = 2) almost quantitative

amounts of nitrite (248 13 lM) were produced. With decreasing

[NO₂⁻]/[DTT] ratio, the yield of nitrite decreased continuously

from the NH₂OH–DTT reaction as well as from the HNO–DTT

reaction by means of the glutamate dehydrogenase assay (Table 3).

Angeli’s salt (250 lM) was employed as HNO source.

This journal is

The Royal Society of Chemistry 2008

Org. Biomol. Chem., 2008, 6, 2560–2573 | 2567

©

View Article Online

Table 3 Yield of ammonium ion from DTT reactions^a

NH₂OH + DTT^b

AS + DTT^c

[NH₂OH]/[DTT], [AS]/[DTT]

Yield/lM

Yield (%)

Yield/lM

Yield (%)

0.25

0.5

1.0

2.0

No DTT

22.7 8.0

19.0 3.8

40.2 6.5

39.4 6.1

14.3 2.2

0.9

0.8

1.6

0.6

7.8 1.0

9.2 2.1

2.9 1.8

1.3 1.8

1.0 1.2

3.1

3.7

1.2

0.5

0.4

^aConcentrations of NH₄were quantiﬁed after a reaction period of 3 h (100 lM EDTA, T = 37 ^◦C, pH 7.4). ^b2.5 mM NH₂OH. ^c250 lM AS.

+

In agreement with the above data, the production of NH₄

from the NH₂OH–DTT reaction did not exceed 2% with respect

to the applied amount of NH₂OH. A slightly higher yield of

approximately 3.5% was observed from the HNO–DTT reaction.

Control experiments demonstrated that neither NH₂OH nor

the decomposition products of Angeli’s salt inhibited glutamate

dehydrogenase at the applied concentrations (data not shown).

Thus, neither the NH₂OH–DTT reaction nor the HNO–DTT one

can explain the 43% yield of NH₄⁺from decomposition of nitroso-

DTT in the presence of excess DTT, i.e. at [NO₂⁻]/[DTT] = 0.25

volatile product. Consequently, nitrite, formed via autoxidation of

^•NO^41,42and subsequent hydrolysis of N₂O₃,⁴³is essentially the only

stable product (yield > 98%) in the presence of oxygen (Fig. 9).

Although DTTox is the only detectable other product, it is not

•

clear whether formation of NO from NODTTNO proceeds by

simultaneous bond-breaking of the two RS–NO bonds in the

dinitroso-dithiothreitol molecule (reaction (7)) or via a stepwise

process (reaction (8)).

+

(Fig. 9 and 10). Therefore, it must be concluded that NH₄is

formed via a pathway independent of the presence of HNO and

(7)

NH₂OH, respectively.

In order to demonstrate that NH₄may also be formed

NODTTNO → NODTT^•+ ^•NO → DTT + 2 ^•NO

(8)

+

from an in situ generated nitroso-dithiol, additional experiments

were performed with GSNO in the presence of Trxn and DTT,

respectively. Under roughly physiological conditions (100 lM

EDTA, pH 7.4, T = 37 ^◦C), Trxn (170.8 lM) reacted with GSNO

Quantum-chemical calculations performed at the CBS-QB3

level of theory predicted that reaction (7) is exergonic by

−20.9 kcal mol⁻¹(Table 4, entry 1). The stepwise mechanism

(reaction (8)) is predicted to proceed via a moderately endergonic

reaction (21.8 kcal mol⁻¹, Table 4, entry 2) followed by a strongly

exergonic one (−42.8 kcal mol⁻¹, Table 4, entry 3). The Gibbs free

energy for homolytic bond dissociation of the S–NO function in

NODTTNO is predicted to be 31 kcal mol⁻¹. This value is very

similar to data experimentally observed for many simple mono-S-

nitrosothiols.⁴⁴Since the CBS-QB3 methodology is very reliable

in predicting S–NO dissociation energies of S-nitrosothiols⁴⁵and

because such S-nitrosothiols do not spontaneously dissociate,⁴²

we doubt that the decomposition of NODTTNO proceeds via

a stepwise homolytic dissociation mechanism. Nevertheless, a

homolytic route cannot be ruled out with certainty.

+

(341.6 lM) to give 20.7 0.2 lM NH₄, about 20% of the amount

that could be generated from DTT under the same conditions.

+

Thus, NH₄production cannot be ignored on decomposition of

physiological nitroso-dithiols.

Discussion

In the present paper we proved for the ﬁrst time that nitroso-

dithiols decay through various independent pathways, the relative

importance of which depends crucially on the degree of nitro-

sation. A fully dinitrosated dithiol yields solely nitric oxide as

Table 4 Quantum-chemically calculated Gibbs energies and aqueous solvation energies

Entry

Reaction^a

D_rG_g^b/kcal mol⁻¹

D_rG_solv^c/kcal mol⁻¹

D_rG_aq^d/kcal mol⁻¹

1

2

3

4

5

6

7

8

9

NODTTNO

NODTT-S^•

NODTT + HNO

2 NODTT

→

DTTox + 2 ^•NO

NODTT-S^•+ ^•NO

DTTox + ^•NO

−21.5

21.5

−43.0

−17.6

−0.6

−9.0

6.0

29.4

15.4

−24.9

−50.1

−44.5

0.6

0.4

0.2

−20.9

21.8

−42.8

−10.7

−3.3

DTT + 2 ^•NO

6.8

DTT + NODTTNO

DTTox + HNO

³NODTT-S⁻

−2.8

−9.0

7.1

−2.5

−37.8

−9.4

−50.3

30.7

NODTT

−18.0

NODTT-S⁻

13.1

NODTT

³NODTT

26.9

³NODTT-S⁻

DTT + HNO

NODTT + DTT

DTTSONH₂+ DTT

DTTox + ³NO⁻

DTTox + NH₂OH

DTT-SONH₂+ DTTox

2 DTTox + NH₃

−22.4

−34.2

−100.4

−13.7

10

11

12

^aThermodynamic properties were calculated using the complete basis set (CBS-QB3) methodology. ^bGas phase data. ^cSolvation corrections from

UHF/6-31+G(d)//CBS-QB3 single point calculations with the CPCM-UAHF solvation model for water.^{44 d}D_rG_aq= D_rG_g+ D_rG_solv

.

2568 | Org. Biomol. Chem., 2008, 6, 2560–2573

This journal is

The Royal Society of Chemistry 2008

©

View Article Online

In conclusion, the spontaneous release of nitric oxide from

NODTTNO (and most likely of other dinitroso-dithiols) is a

highly efﬁcient, thermodynamically feasible process because the

driving force of the reaction depends on a ring closure of a cyclic

disulﬁde.

10 (Fig. 5B). The pH dependence of Fig. 5B rather points to

the deprotonated, thiolate form of NODTT (NODTT-S⁻) as the

reacting species. This would be in agreement with typical acidity

constants of aliphatic thiols (compare, e.g. mercaptoethanol,

pK_a= 9.5). Interestingly, our CBS-QB3 calculations predicted that

only the thiolate anion of NODTT has a rather low-lying triplet

state, about 13.1 kcal mol⁻¹above the singlet ground state (Table 4,

entry 7). It is therefore proposed that the reaction may take place

via a thermally accessible triplet state (reaction (11)), for which

CBS-QB3 calculation predicts a driving force of −9.3 kcal mol⁻¹

in aqueous solution (Table 4, entries 7 and 9).

•

Now, the question arose why NO release is still fairly good

from nitroso-DTT solutions generated at [NO₂⁻]/[DTT] ratios ≤

1, i.e. where mainly NODTT is formed and thus the release of

HNO has to be expected (Fig. 6B–D). In this case, two processes

•

for NO production can be imagined: First, reaction of HNO

with the parent S-nitrosothiol (reaction (4)) as proposed by Wong

et al.¹³must be taken into account. In fact, the reaction of HNO

with NODTT is predicted to be exergonic by 10.7 kcal mol⁻¹

(Table 4, entry 4). Second, since trans-nitrosation between thiols

and nitrosothiols is a well-known process,⁴²NODTT might be in

•

equilibrium with NODTTNO (reaction (5)), from which NO is

(11)

then released according to reaction (7) (Table 4, entry 1).

1

The possibility that NODTT-S⁻released alternatively NO⁻

The fact that NODTT ([NO₂⁻]/[DTT] = 1) did not decay by

a clean ﬁrst-order process (Fig. 5) would be in correspondence

with reaction (5) because CBS-QB3 calculations predict that this

reaction is slightly exergonic by 3.3 kcal mol⁻¹(Table 4, entry 5).

However, in case of the physiological dithiol-thioredoxin, trans-

nitrosation of its S-mononitrosated form (NOTrxn) analogous to

reaction (5) appears unlikely due to high steric hindrance by the

protein which would hamper exchange of the nitroso function

is unlikely because the experimentally determined singlet–triplet

energy gap of the nitroxyl anion is −16.1 kcal mol⁻¹,⁴⁶so that

reaction (11) would lose its negative driving force. The ³NO⁻

anion comprises a unique conjugate acid–base couple (reaction

(12)) with different ground-state multiplicities, preferring HNO

formation at pH values lower than 11.4, the pK_avalue of HNO.⁴⁷

³NO⁻+ H₃O⁺ꢁ HNO + H₂O

(12)

•

to give S-dinitroso-Trxn. Since, however, ample NO production

from NOTrxn has been found, we conclude that ^•NO is primarily

liberated from S-mononitrosothiols (in the absence of copper ions)

via the HNO pathway (reaction (4)). Nevertheless, for sterically

unhindered dithiols like DTT a contribution by reaction (5) cannot

be ruled out.

The fact that the half-life of NODTT again increased at pH > 10

(Fig. 5B) thus can be explained by the pK_avalue of HNO and the

reversibility of reaction (11). Since the spin inversion of NODTT-

S⁻(reaction (11)) possesses a sizeable barrier of ca. 13 kcal mol⁻¹,

one had to expect that bimolecular reactions (reactions (4) and (5))

could compete with reaction (11). The concentration of NODTT

is governed by its self-reaction (reaction (5)) as well as by the

Since reaction of HNO with the parent S-nitrosothiol is

obviously the dominating pathway of ^•NO generation from

mononitroso-dithiols, the question remained how HNO is released

from NODTT. Although any details about the mechanism are

missing in the literature, the reaction schemes given in those

publications^14,15suggest that generation of HNO from NODTT

proceeds by direct, concerted elimination (reaction (9)).

•

HNO-mediated release of NO (reaction (4)). At excess dithiol

concentration, HNO is trapped by DTT and this reaction—as

ﬁrst discussed by Wong et al.¹³for the GSNO/GSH system—is

the source of the observed products hydroxyl amine, sulﬁnamide,

and ammonium ion. The formation of these compounds via the

Wong et al. mechanism is also strongly supported by our CBS-QB3

calculations (Table 4, entries 10–12).

According to Singh et al.¹², the HNO-independent formation

(9)

+

of NH₄from GSNO via a stepwise mechanism with GSNH₂as

intermediate requires four equivalents of GSH. Such a mechanism

cannot be ruled out with certainty. Nevertheless, the intermediacy

of both HNO and the DTT-derived sulﬁnamide strongly sup-

We were unable to locate a transition structure for reaction (9), so

that the barrier height of the suggested elimination could not be

calculated. The CBS-QB3 calculations predicted that the NODTT-

dependent formation of HNO (reaction (10)) may be a proton-

catalyzed process:

+

ported the view that formation of NH₄from decay of NODTT

should primarily follow the Wong et al. mechanism. In conclusion,

formation of all observed products is proposed to occur via the

reactions summarized in Scheme 3.

Since the chemical properties of nitroso-DTT and nitroso-

dihydrolipoic acid³on the one hand and of nitroso-DHLA and

nitroso-Trxn⁴⁸on the other hand are similar, we believe that the

reactions outlined in Scheme 3 are quite general ones for the

decay of nitroso-dithiols at physiological pH values. It has been

(10)

•

suggested that Trxn reacts in the cytosol to yield either NO¹⁹

However neither direct elimination of HNO from the ni-

trosodithiol (reaction (9)) nor a proton-catalyzed process (reaction

(10)) would explain the observed pH dependence of the decay

rate of NODTT with a maximum (lowest half-life) at pH 9–

or NH₂OH.⁴⁸In this work, experimental evidence is presented

that the formation of the products depends on the degree of

nitrosation of the dithiol. A low level of nitrosation means

that the concentration of dinitroso-dithiol is negligibly low since

This journal is

The Royal Society of Chemistry 2008

Org. Biomol. Chem., 2008, 6, 2560–2573 | 2569

©

View Article Online

(Michigan, US). All other chemicals were of the highest purity

commercially available.

Experimental conditions

All experiments were carried out at 25 ^◦C in 50 mM phosphate

buffer (PB) if not described otherwise. Because nitrosation reac-

tions are sensitive to the presence of metal ions,⁴²solutions were

treated with the chelating resin Chelex 100 prior to use as described

previously,⁵²and further supplemented with 100 lM EDTA.

Synthesis of nitrosated dithiothreitol

As previously described for other thiols,⁵³154.25 mg DTT were

diluted in 50 mM PB containing 100 lM EDTA, and 69.0 mg

NaNO₂or 70.0 mg Na[N-15]NO₂were diluted in doubly distilled

water. Both solutions were mixed at the required ratio, and the

synthesis was started by adding 1 M hydrochloric acid until a pH

of 2 was attained. The reddish solutions were diluted immediately

with 50 mM PB and 100 lM EDTA to give a 10-fold stock solution.

The concentration was photometrically controlled (see Fig. 1).

Samples were prepared as fast as possible because of the low half-

life of nitroso-DTT¹⁵(see Results), and the pH was immediately

adjusted to 7.4 with 1 M NaOH.

Synthesis of S-nitroso-glutathione

As described previously,³³3.07 g GSH and 0.69 mg NaNO₂were

allowed to react in a mixture of 20 ml puriﬁed water and 10 ml 1 M

HCl for 40 min at 4 ^◦C. Then, 20 ml acetone were added and the

solution was stirred for another 10 min. The formed precipitate

was ﬁltered off, washed with 10 ml water, 3 times with 20 ml

diethyl ether, dried in a desiccator and stored at −20 ^◦C. The

concentration of the prepared GSNO was quantiﬁed by reading

Scheme 3

reaction (5) is shifted to the left side. In intracellular milieu,

the concentrations of RSNOs have been determined to be in

the submicromolar range,⁴⁹whereas that of Trxn and GSH are

in the range of 5–50 lM⁴⁸and 3–5 mM,⁵⁰respectively. Under

such conditions, the formation of the nitric oxide-yielding entity

dinitroso-Trxn would be extremely low and only NH₂OH and

−1

the optical density at k_max= 545 nm (e₅₄₅= 20 M cm⁻¹).⁵⁴

Synthesis of 1-nitroso-melatonin

+

1-Nitroso-melatonin (NOMela) was prepared as described by

Kirsch and de Groot.⁵³In brief, melatonin (232 mg) was dissolved

in 4 ml acetone,10 ml water was added and the mixture stirred for

several minutes. Afterwards 207 mg NaNO₂and another 2 ml of

water were added. 1.5 ml of 1 M HCl were added to the yellow

liquid, shaken well and cooled to 1 ^◦C. The mixture was extracted

with 20 ml cold (1 ^◦C) dichloromethane, and the separated organic

phase dried with 750 mg Na₂SO₄. The solvent was evaporated

at −20 ^◦C, and the yellow solid was stored at −20 ^◦C. The

purity was determined photometrically at 345 nm (e₃₄₅= 7070 M⁻¹

cm⁻¹).⁵³To a 100-fold stock solution used for the experiments

20 vol% DMSO were added to prepare an aqueous solution of the

lipophilic substance.

NH₄would be formed. The generation of NH₄is favourable

for biological systems as ammonium can be rapidly detoxiﬁed by

many cell types via the urea cycle.⁵¹Thus, Trxn is expected to

additionally act in vivo as a GSNO reductase but not as a GSNO-

dependent nitric oxide synthase.

Experimental

Chemicals

Thioredoxin (Escherichia coli, purity ≥ 90%), N-acetyltryptophan

(NAT), dithiothreitol (DTT), glutathione (GSH), Chelex 100,

5,5^ꢀ-dithio-bis(2-nitrobenzoic acid) (DTNB) and L-glutamic de-

hydrogenase (GlDH) from bovine liver were obtained from

Sigma (Hamburg, Germany). Na[N-15]NO₂and [N-15]NH₂OH

labled with 99.3% N-15 were purchased from Aldrich/Isotech

Inc (Hamburg, Germany). NaNO₂and (NH₄)₂SO₄were pro-

duced by Merck (Darmstadt, Germany). 4,5-Diaminoﬂuorescein

(DAF-2), diaminoﬂuorescein-2T (DAF-2T), hydroxyphenyl ﬂu-

orescein (HPF) and aminophenyl ﬂuorescein (APF) were from

Alexis (Gru¨nberg, Germany). Angeli’s salt (AS; purity ≥ 99%)

and Mn(III)-tetrakis(1-methyl-4-pyridyl)porphyrin pentachlo-

ride (Mn(III)TMPyp) were obtained from Cayman Chemical

Synthesis of N-nitroso-N-acetyltryptophan (NANT)

The preparation of NANT as described by Bonnett and

Holleyhead⁵⁵was improved by Sonnenschein et al.³³In brief,

526 mg NAT and 162 mg NaNO₂were stirred in 20 ml puriﬁed

water for 2 h at room temperature in the dark. The yellow mixture

was cooled to 1 ^◦C, and 10 ml of cold (1 ^◦C) 1 M HCl were

added. The yellow precipitate was immediately extracted with

60 ml ethyl acetate (1 ^◦C). The organic layer was separated, and the

2570 | Org. Biomol. Chem., 2008, 6, 2560–2573

This journal is

The Royal Society of Chemistry 2008

©

View Article Online

solvent evaporated at room temperature under reduced pressure

(18 Torr) to yield 500 mg of NANT. The concentration of NANT

was photometrically controlled with a SPECTROCORD S 100

nicotinamide-adenine dinucleotide phosphate (NADPH) to nitrite

in the presence of the enzyme nitrate reductase. To the samples,

which were incubated for 20 min at room temperature, pyruvate

and LDH were added, and the samples were incubated for another

20 min at room temperature. The total nitrite-concentration was

determined by using the Griess-reagent as described above. The

base level of nitrite was directly quantiﬁed from a similarly treated

sample in the absence of nitrite reductase. The test was calibrated

with a solution of neat NaNO₃, whose purity was controlled

−1

spetrometer from Analytic Jena (e₃₃₅= 6100 M cm⁻¹).⁵⁵

H-1 NMR spectrometric measurements

Solutions of 10 mM NODTTNO plus 10 mM NAT, 10 mM

NODTTNO, 10 mM NAT, and 5 mM NANT, respectively, were

prepared in 50 mM PB with 10% D₂O and 100 lM EDTA at

pH 7.4. Spectra were recorded by 500 MHz H-1 NMR on a Bruker

AVANCE DRX 500 instrument 90 min after sample preparation.

Chemical shifts (d) are given in parts per million (ppm) relative to

tetramethyl silane (d = 0) as an external standard.

−1

photometrically (e₃₀₂= 10 M cm⁻¹).⁵⁸

Quantiﬁcation of dihydroperoxide

Hydrogen peroxide was determined by the horseradish peroxidase-

catalysed reaction of H₂O₂with 4-amino-antipyrine and 3,5-

dichloro-2-hydroxyl-benzenesulfonic acid. The pink-coloured

product was measured photometrically at 513 nm.⁵⁹Samples of

nitroso-DTT having reacted 10 min in the absence and presence

of SOD (100 units/ml) were measured after another 10 min of

incubation at room temperature. Standard buffer solutuons of

H₂O₂were used for calibration.

N-15 NMR identiﬁcation of adducts

After incubation for 6 h at room temperature, sample aliquots

were supplemented with 10% D₂O and analyzed by N-15 NMR

spectrometry. The products were identiﬁed by 50.67 MHz N-15

NMR on a Bruker AVANCE DRX 500 instrument. Chemical

shifts (d) are given in parts per million (ppm) relative to neat

nitromethane (d = 0) as an external standard.

Quantiﬁcation of ammonium

+

The concentration of NH₄was determined as described by

Bergmeyer et al.⁵⁸Samples were incubated at room temperature for

3 h, then, aliquots were diluted 2.48-fold with 1 M triethanolamine

buffer (TEA, pH = 8.0). To this mixture were added 2.8 mM

a-ketoglutarate, 177 lM NADH, and 0.8% glycerine. After incu-

bation for 20 min at room temperature, optical density of the

samples was read photometrically at 340 nm. Then, 7.2 units ml⁻¹

L-glutamic dehydrogenase (GlDH) were added and the samples

were incubated for another 20 min at room temperature. Finally,

the optical density at 340 nm was read to calculate the decay of

NADH (e₃₄₀= 6200 M⁻¹cm⁻¹),⁵⁸which is directly proportional to

Quantiﬁcation of ^•NO

Release of nitric oxide was detected polarographically with a

graphite nitric oxide-sensing electrode (World Precision Instru-

ments, Berlin, Germany) immediately after sample preparation.

Quantiﬁcation of dithiothreitol

The concentration of thiol functions of DTT was quantiﬁed as

described by Ellmann and improved by us.⁵⁶DTNB (39.6 mg) was

dissolved in 50 mM PB (10 ml, pH = 7.0). The samples were diluted

10-fold with 100 mM PB (pH = 7.0) and to 1 ml of the diluted

sample 6.67 ll stock solution of DTNB were added. The optical

density of the formed p-nitrothiophenol anion (e₄₁₂= 13 600 M⁻¹

cm⁻¹)⁵⁷was read photometrically after incubation for 20 min at

room temperature. The test was calibrated with a solution of neat

DTT.

+

the concentration of NH₄. The test was calibrated with a solution

of neat (NH₄)₂SO₄.

Qualitative detection of nitroxyl

HNO was determined as described by Marti et al.³⁴for a related

Mn(III)–porphyrin complex. To the prepared samples were added

10 lM of Mn(III)TMPyp. Aliquots were immediately photo-

metrically measured by a SPECTROCORD S 100 spetrometer

from Analytic Jena. On reaction with HNO, the Soret-band of

Mn(III)TMPyp at 463 nm (e₄₆₃= 9.2 × 10⁴M⁻¹cm⁻¹;³⁵) decreased

and that of Mn(II)TMPyp-NO at 435 nm (e₄₃₅= 1.1 × 10⁵M⁻¹

cm⁻¹;³⁶) increased.

Quantiﬁcation of nitrite

Nitrite was quantiﬁed by using the Griess-reagent (0.1%

naphthylethylendiamine-dichloride [Sigma] + 1% sulfanil-amide

[Sigma] in 5% H₃PO₄).⁵⁸At the given time of reaction, the samples

were diluted 30-fold with 50 mM PB (pH = 7.5), afterwards

a 2.5-fold volume of Griess-reagent was added. After 10 min

of incubation at room temperature, the optical density was

photometrically read at 542 nm. Calibration was carried out

Determination of reactive intermediates from NODTTNO

NAT or melatonin (10 mM each, 50 mM PB, 100 lM EDTA, 20%

DMSO, pH = 7.4, room temperature) were allowed to react with

preformed NODTTNO (10 mM) for 90 min and the concentration

with a photometrically controlled solution of neat NaNO₂(e₃₅₄

22.9 M⁻¹cm⁻¹).⁵⁸

=

of either NANT (e₃₃₅= 6100 M⁻¹cm⁻¹) or nitroso-melatonin (e₃₄₅

=

7070 M⁻¹cm⁻¹) was calculated.

Quantiﬁcation of nitrate

DHR-123. NODTTNO (50 lM) was allowed to react with

DHR-123 (50 lM) for 180 min at 37 ^◦C, the concentration of

rhodamine-123 was quantiﬁed photometrically at k_max= 500 nm

(e₅₀₀= 78 000 M⁻¹cm⁻¹).

The concentration of nitrate was determined by a commercially

available Nitrate-Test-Kit (Boeringer Mannheim/R-Biopharm,

Germany) improved by us. In brief, nitrate is reduced by

This journal is

The Royal Society of Chemistry 2008

Org. Biomol. Chem., 2008, 6, 2560–2573 | 2571

©

View Article Online

DAF-2, APF, HPF. 15 lM NODTTNO were mixed immedi-

ately with DAF-2, APF or HPF (10 lM each), and the samples were

incubated for 180 min at 37 ^◦C in the dark room. Afterwards, the

concentration of DAF-2T or ﬂuorescein was read ﬂuorometrically.

The APF- and HPF-derived formation of ﬂuorescein or the DAF-

2-mediated formation of DAF-2T were quantiﬁed by reading

their ﬂuorescence with excitation at 495 nm and emission at

515 nm, respectively. Standard calibration curves were prepared

from known amounts of ﬂuorescein and DAF-2T.

14 D. A. Stoyanovsky, Y. Y. Tyurina, V. A. Tyurin, D. Anand, D. N.

Mandavia, D. Gius, J. Ivanova, B. Pitt, T. R. Billiar and V. E. Kagan,

J. Am. Chem. Soc., 2005, 127, 15815–15823.

15 D. R. Arnelle and J. S. Stamler, Arch. Biochem. Biophys., 1995, 318,

279–285.

16 C. Petit, P. Hoffmann, J. P. Souchard, F. Nepveu and S. Labidalle,

C. R. Seances Soc. Biol. Ses Fil., 1996, 190, 641–650.

17 J. P. Barton and J. E. Packer, Int. J. Radiat. Phys. Chem., 1970, 2,

159–166.

18 M. Quintiliani, R. Badiello, M. Tamba, A. Esfandi and G. Gorin,

Int. J. Radiat. Biol., 1977, 32, 195–202.

19 D. Nikitovic and A. Holmgren, J. Biol. Chem., 1996, 271, 19180–19185.

20 A. P. Dicks, H. R. Swift, D. L. H. Williams, A. R. Butler, H. H. Al-

Sa’doni and B. G. Cox, J. Chem. Soc., Perkin Trans. 2, 1996, 481–487.

21 D. L. H. Williams, Chem. Commun., 1996, 1085–1091.

22 M. W. Brock, C. Mathes and W. F. Gilly, J. Gen. Physiol., 2001, 118,

113–133.

23 S. Goldstein and G. Czapski, J. Am. Chem. Soc., 1996, 118, 3419–3425.

24 M. Kirsch and H. de Groot, J. Biol. Chem., 2002, 277, 13379–13388.

25 N. W. Kooy, J. A. Royall, H. Ischiropoulos and J. S. Beckman, Free

Radical Biol. Med., 1994, 16, 149–156.

Quantum chemical calculations

Complete basis set (CBS-QB3) computations were carried out with

the Gaussian 03 (Revision A.11.3) suite of programs.⁶⁰Aqueous

solvation free energies were evaluated on the optimized gas-phase

geometries with the CPCM⁶¹procedure incorporated in Gaussian

03. Both the CPCM/UHF/6-31+G(d) and the CBS-QB3 method-

ology are known to provide estimates within “chemical accuracy”

26 H. Ischiropoulos, D. Duran and J. Horwitz, J. Neurochem., 1995, 65,

2366–2372.

2–3 kcal mol⁻¹). Isotropic absolute shielding constants of

27 M. Kirsch, A. Fuchs and H. de Groot, J. Biol. Chem., 2003, 278,

(

11931–11936.

the nitrogen nucleus were calculated with the individual gauges

for atoms in molecules (IGAIM) protocol⁶²at the B97-2/aug-

cc-pVDZ level of theory. The optimization of the structure and

molecular interactions with the solvent were followed at the same

level of theory as described above.

28 K. M. Miranda, M. G. Espey, N. Ludwick, S. M. Kim, D. Jourd’heuil,

M. B. Grisham, M. Feelisch, J. M. Fukuto and D. A. Wink, J. Biol.

Chem., 2001, 276, 1720–1727.

29 K. Setsukinai, Y. Urano, K. Kakinuma, H. J. Majima and T. Nagano,

J. Biol. Chem., 2003, 278, 3170–3175.

30 R. S. Lewis, S. R. Tannenbaum and W. D. Deen, J. Am. Chem. Soc.,

1995, 117, 3933–3939.

31 D. A. Stoyanovsky, R. Clancy and A. I. Cederbaum, J. Am. Chem. Soc.,

1999, 121, 5093–5094.

Statistics

32 L. Grossi and P. C. Montevecchi, Chem.–Eur. J., 2002, 8, 380–387.

33 K. Sonnenschein, H. de Groot and M. Kirsch, J. Biol. Chem., 2004,

279, 45433–45440.

All experiments were carried out independently at least three times

on different days. The results are expressed as means S.D.

34 M. A. Marti, S. E. Bari, D. A. Estrin and F. Doctorovich, J. Am. Chem.

Soc., 2005, 127, 4680–4684.

35 K. M. Faulkner, S. I. Liochev and I. Fridovich, J. Biol. Chem., 1994,

269, 23471–23476.

Acknowledgements

36 M. Kirsch and H. de Groot, J. Pineal Res., 2008, 44, 244–249.

37 S. I. Liochev and I. Fridovich, J. Biol. Chem., 2001, 276, 35253–35257.

38 M. Kelm, R. Dahmann, D. Wink and M. Feelisch, J. Biol. Chem., 1997,

272, 9922–9932.

Sonja Liebeskind was supported by an IFORES grant of the

Universita¨t Duisburg-Essen.

39 B. Bujnicki, J. Drabowicz, M. Mikolajczyk, A. Kolbe and L. Stefaniak,

J. Org. Chem., 1996, 61, 7593–7596.

40 L. Stefaniak, G. A. Webb and M. Witanowski, Annu. Rep. NMR

Spectrosc., 1993, 25.

41 S. Goldstein and G. Czapski, J. Am. Chem. Soc., 1995, 117, 12078–

12084.

References

1 L. J. Ignarro, H. Lipton, J. C. Edwards, W. H. Baricos, H. L. Hyman,

P. J. Kadowitz and C. A. Gruetter, J. Pharmacol. Exp. Ther., 1981, 218,

739–749.

2 S. Moncada, R. M. J. Palmer and E. A. Higgs, Pharmacol. Rev., 1991,

43, 109–142.

3 J. S. Stamler, Circ. Res., 2004, 94, 414–417.

4 B. M. Gaston, J. Carver, A. Doctor and L. A. Palmer, Mol. Interven-

tions, 2003, 3, 253–263.

5 S. R. Jaffrey, H. Erdjument-Bromage, C. D. Ferris, P. Tempst and S. H.

Snyder, Nat. Cell Biol., 2001, 3, 193–197.

6 G. Melino, F. Bernassola, R. A. Knight, M. T. Corasaniti, G. Nistico

and A. Finazzi-Agro, Nature, 1997, 388, 432–433.

7 J. B. Mannick, X. Q. Miao and J. S. Stamler, J. Biol. Chem., 1997, 272,

24125–24128.

8 Y.-M. Kim, R. V. Talanian and T. R. Billiar, J. Biol. Chem., 1997, 272,

31138–31148.

9 S. Z. Lei, Z.-H. Pan, S. K. Aggarwal, H.-S. V. Chen, J. Hartmann, N. J.

Sucher and S. A. Lipton, Neuron, 1992, 8, 1987–1099.

10 L. Xu, J. P. Eu, G. Meissner and J. S. Stamler, Science, 1998, 279,

234–237.

42 D. L. H. Williams, Nitrosation reactions and the chemistry of nitric

oxide, Elsevier, Amsterdam, 2004.

43 L. J. Ignarro, J. M. Fukuto, J. M. Griscavage, N. E. Rogers and R. E.

Byrns, Proc. Natl. Acad. Sci. USA, 1993, 90, 8103–8107.

44 P. G. Wang, M. Xian, X. Tang, X. Wu, Z. Wen, T. Cai and A. J. Janczuk,

Chem. Rev., 2002, 102, 1091–1134.

45 M. D. Bartberger, J. D. Mannion, S. C. Powell, J. S. Stamler, K. N.

Houk and E. J. Toone, J. Am. Chem. Soc., 2001, 123, 8868–8869.

46 M. Tronc, A. Huetz, M. Landau, F. Pichou and J. Reinhardt, J. Phys.

B: At. Mol. Phys., 1975, 8, 1160–1169 .

47 V. Shaﬁrovich and S. V. Lymar, Proc. Natl. Acad. Sci. USA, 2002, 99,

7340–7345.

48 D. A. Stoyanovsky, J. Am. Chem. Soc., 2005, 127, 15815–15823.

49 Y. Zhang and N. Hogg, Free Radical Biol. Med., 2005, 38, 831–838.

50 A. Meister, J. Biol. Chem., 1988, 263, 17205–17208.

51 R. K. Murray, D. K. Granner, P. A. Mayes and V. W. Rodwell, Harper’s

Illustrated Biochemistry, McGraw-Hill Companies, New York, 24th

edn, 2003.

52 M. Kirsch and H. de Groot, J. Biol. Chem., 2000, 275, 16702–16708.

53 M. Kirsch and H. de Groot, J. Pineal Res., 2006, 40, 10–17.

54 A. J. Holmes and D. L. H. Williams, J. Chem. Soc., Perkin Trans. 2,

2000, 1639–1644.

11 J. Sun, C. Xin, J. P. Eu, J. S. Stamler and G. Meissner, Proc. Natl. Acad.

Sci. USA, 2001, 98, 11158–11162.

12 S. P. Singh, J. S. Wishnok, M. Keshive, W. M. Deen and S. R.

Tannenbaum, Proc. Natl. Acad. Sci. USA, 1996, 93, 14428–14433.

13 P. S.-Y. Wong, J. Hyun, J. M. Fukuto, F. N. Shirota, E. G. deMaster,

D. W. Shoeman and H. T. Nagasawa, Biochemistry, 1998, 37, 5362–

5371.

55 R. Bonnett and R. Holleyhead, J. Chem. Soc., Perkin Trans. 1, 1974,

962–964.

2572 | Org. Biomol. Chem., 2008, 6, 2560–2573

This journal is

The Royal Society of Chemistry 2008

©

View Article Online

56 G. L. Ellmann, Arch. Biochem. Biophys., 1958, 74, 443.

57 G. L. Ellmann, Arch. Biochem. Biophys., 1958, 74, 443–450.

58 H. U. Bergmeyer, M. Grassl and H.-E. Walter, in Methods of Enzymatic

Analysis, ed. H. U. Bergmeyer, Verlag Chemie, Weinheim, 1983, vol.

II, p. 165.

59 I. Ioannidis and H. de Groot, Biochem. J., 1993, 296, 341–345.

60 M. J. Frisch, G. W. Trucks, H. B. Schlegel, P. M. W. Gill, B. G.

Johnson, M. A. Robb, J. R. Cheeseman, T. Keith, G. A. Petersson, J. A.

Montgomery, K. Raghavachari, M. A. Al-Laham, V. G. Zakrzewski,

J. V. Ortiz, J. B. Foresman, J. J. Cioslowski, B. B. Stefanov, A.

Nanayakkara, M. Challacombe, C. Y. Peng, P. Y. Ayala, W. Chen,

M. W. Wong, J. L. Andres, E. S. Replogle, R. Gomperts, R. L. Martin,

D. J. Fox, J. S. Binkley, D. J. Defrees, J. Baker, J. P. Stewart, M. Head-

Gordon, C. Gonzalez and J. A. Pople, Gaussian 98W, Gaussian Inc.,

Pittsburgh, PA, Revision A.9 edn, 2000.

61 V. Barone, M. Cossi and J. Tomasi, J. Chem. Phys., 1997, 107, 3210–

3221.

62 T. Gregor, F. Mauri and R. Car, J. Chem. Phys., 1999, 111, 1815–1822.

This journal is

The Royal Society of Chemistry 2008

Org. Biomol. Chem., 2008, 6, 2560–2573 | 2573

©

Article Doi

Dependence of product formation from decomposition of nitroso-dithiols on the degree of nitrosation. Evidence that dinitroso-dithiothreitol acts solely as an nitric oxide releasing compound

DOI: 10.1039/b801583j

Source and publish data:

Authors:

Article abstract of DOI:10.1039/b801583j

Full text of DOI:10.1039/b801583j

Products guided by the article

R&D Labs maybe for 812697-88-4

Relevant to this article

Hot Product