Quantitative determination of SH groups using 19F NMR spectroscopy and disulfide of 2,3,5,6-tetrafluoro-4-mercaptobenzoic acid

Article abstract of DOI:10.1002/mrc.1652

A new method of measurement of thiol concentration by ¹⁹F NMR spectroscopy is developed. The method is based on the detection of products of the exchange reaction of thiols with a newly synthesized fluorinated disulfide, 2,3,5,6-tetrafluoro-4-mercaptobenzoic acid (BSSB). A significant broadening of the ¹⁹F NMR signal of BSSB in the presence of thiols was observed and attributed to the exchange reaction between the parent disulfide and 2,3,5,6-tetrafluoro-4-mercaptobenzoic acid. The rate constant for this reaction was found to be equal to (63 ± 11) × 10³ M^-1 s^-1 at pH 7.0. The method was applied for the measurement of concentration of glutathione and albumin in rat blood. Copyright

Full text of DOI:10.1002/mrc.1652

MAGNETIC RESONANCE IN CHEMISTRY
Magn. Reson. Chem. 2005; 43: 902–909
Published online 22 August 2005 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/mrc.1652
Quantitative determination of SH groups using
19F NMR spectroscopy and disulfide of
2,3,5,6-tetrafluoro-4-mercaptobenzoic acid
Dmitrii I. Potapenko,1,2 Elena G. Bagryanskaya,1 Igor A. Grigoriev,3
Aleksander M. Maksimov,3 Vladimir A. Reznikov,3 Vyacheslav E. Platonov,3
Thomas L. Clanton5 and Valery V. Khramtsov4,5∗
1
International Tomography Center, Novosibirsk 630090, Russia
Novosibirsk State University, Novosibirsk 630090, Russia
Institute of Organic Chemistry, Novosibirsk 630090, Russia
Institute of Chemical Kinetics and Combustion, Novosibirsk 630090, Russia
2
3
4
5
Dorothy M. Davis Heart and Lung Research Institute, The Ohio State University, Columbus, OH 43210, USA
Received 23 March 2005; Revised 23 June 2005; Accepted 27 June 2005
A new method of measurement of thiol concentration by 19F NMR spectroscopy is developed. The
method is based on the detection of products of the exchange reaction of thiols with a newly synthesized
fluorinated disulfide, 2,3,5,6-tetrafluoro-4-mercaptobenzoic acid (BSSB). A significant broadening of the
19F NMR signal of BSSB in the presence of thiols was observed and attributed to the exchange reaction
between the parent disulfide and 2,3,5,6-tetrafluoro-4-mercaptobenzoic acid. The rate constant for this
reaction was found to be equal to .63 11/ × 103 M−1 s−1 at pH 7.0. The method was applied for the
measurement of concentration of glutathione and albumin in rat blood. Copyright  2005 John Wiley &
Sons, Ltd.
KEYWORDS: NMR; thiols; thiol measurements; 19F; thiol-disulfide exchange; glutathione; fluorinated disulfides; benzoic acid
Currently existing methods of SH group registra-
tion (including photometric measurements using Ellman’s
reagent6) require optically transparent samples or prior iso-
lation of SH-containing compounds7–9 and, therefore, cannot
be applied in vivo. Several years ago we proposed an elec-
tron paramagnetic resonance (EPR) spectroscopy approach
for thiol detection, which has better potential for in vivo
use.10–12 The method is based on thiol–disulfide exchange
reactions of stable biradical disulfides with thiols, which
result in formation of monoradical products followed by
significant changes in EPR spectrum. The possibility of work-
ing in nontransparent media and the high sensitivity of
EPR are the most important advantages of this noninvasive
technique. However biological reduction of the probes in
EPR-silent products significantly limits the applications of
the method. The development of fluorine-containing disul-
fides and application of the 19F NMR technique for thiol
detection overcomes the latter disadvantage of the EPR
approach. Moreover, the NMR approach could enable imag-
ing of SH-containing compounds in living organisms. In the
present paper a fluorine-containing disulfide of 4-mercapto-
2,3,5,6-tetrafluorobenzoic acid (BSSB) was synthesized. The
mechanism of the reaction of BSSB with a number of thi-
ols of biological importance was studied and applications
of BSSB to measure SH compounds in rat blood is demon-
strated.
INTRODUCTION
The role of thiols in metabolic processes and in support-
ing intracellular and extracellular homeostasis in vivo is
widely appreciated. Among low-molecular-weight thiols,
glutathione, GSH, and cysteine are the most significant.
Particularly GSH, presenting in various cells in mM con-
centration range, is considered to be the most important
player in defining cellular ‘redox state’.1 The role of labile
thiol groups in supporting the structure and function of
proteins is also well documented. The main protein com-
ponents of blood, hemoglobin, and albumin are important
examples of macromolecular thiols, having in their structure
two and one SH groups easily available for modification,
correspondingly.2–5
ŁCorrespondence to: Valery V. Khramtsov, Dorothy M. Davis Heart
& Lung Research Institute, 201 HLRI, 473 W12th Ave, The Ohio
State University, Columbus, OH 43210, USA.
E-mail: khramtsov-1@medctr.osu.edu
Contract/grant sponsor: RFBR; Contract/grant numbers:
05-04-48632, 05-04-48483.
Contract/grant sponsor: NIH; Contract/grant number: K01
EB03519, HL 53333.
Contract/grant sponsor: INTAS; Contract/grant number:
03-55-1740.
Contract/grant sponsor: Ministry of Education; Contract/grant
number: RF A03-2.11-822.
Copyright  2005 John Wiley & Sons, Ltd.

Thiols determination using 19F NMR 903
Blood was collected from the neck after injection of 6 µl of
heparine sodium (5000 I.E./ml). About 5 ml of the blood was
mixed with 250 µl of 0.5 M EDTA and centrifuged for 10 min
EXPERIMENTAL
Chemicals
Disodium- and monopotassium phosphate, ethylenedi-
aminetetraacetic acid (EDTA), trichloroacetic acid (TCA),
Chelex-100, glutathione (GSH), oxidized glutathione (GSSG),
human serum albumin (ASH), and bovine hemoglobin
(Hem) were obtained from Sigma. Deionized water was used
for all experiments. Phosphate buffer was passed through
a column with Chelex-100 to remove trace amounts of the
transition-metal ions.
°
at 700 g at 4 C. Supernatant (plasma) was collected and
usedwithoutfurtherpreparation. Sedimentatederythrocytes
were frozen in liquid nitrogen. After defreezing, the sample
was diluted by a phosphate buffer, pH 7.0, up to initial blood
volume. 1.5 ml of the obtained mixture was added to 0.15 ml
of 50% (w/v) TCA. The samples were centrifuged at 2500 g
for 10 min and the supernatant was adjusted to pH 7.0 and
used to determine glutathione concentration.
Synthesis of BSH and BSSB
4-Mercapto-2,3,5,6-tetrafluorobenzoic acid (BSH) was syn-
thesized according to Ref. 13. 4,40-Dithio-bis(2,3,5,6-tetra-
fluorobenzoic acid) (BSSB) was synthesized as follows:
Solution of 0.51 g of bromine in 0.5 ml of acetic acid was
added dropwise with stirring to a mixture of 1.36 g of
the 4-mercapto-2,3,5,6-tetrafluorobenzoic acid and 4 ml of
acetic acid. The reaction mixture was allowed to stand
overnight. Then the solvent was partly removed (¾80%).
In this process, the product precipitated, and the precip-
itate was filtered off, washed with water, and dried to
give 1.09 g of BSSB. The product was purified by sublima-
RESULTS
Thiol–disulfide exchange reactions of BSSB with
physiologically relevant thiols
Fluorinated disulfide of benzoic acid, BSSB, was synthesized
as shown in Fig. 1. It was expected that reaction of the BSSB
with thiols would result in the splitting of the disulfide bond
and formation of the corresponding monothiol, fluorinated
benzoic acid, BSH.
Figure 2a shows the 19F NMR spectrum of the disulfide,
BSSB, in deuterated dimethylsulfoxide. The spectral pattern
is characteristic for an AA0BB0 system14 in agreement with
the BSSB structure with nonequivalent fluorine nuclei. The
values of spin–spin interaction constants and chemical shifts
obtained by fitting the experimental spectra to calculated
ones are listed in Table 1 and Fig. 2 captions. Note that
the values of chemical shift strongly depend on the solvent
and pH. The multiplet structure of the 19F NMR spectrum
of the BSSB in water was not observed because of line
broadening. For spectra simulations, the parameters of
spin–spin coupling obtained in DMSO were used.
Figure 3 shows the 19F NMR spectra of BSSB in 0.1 M
phosphate buffer, pH 7.0, measured before and after GSH
addition. The appearance of the signals of two new products
at 26.1 and 16.0 ppm (product P1), and 31.2 and 20.43 ppm
(product P2) is clearly seen in Fig. 3b and c. Significant
broadening of BSSB signal is also observed. It should be
mentioned that the initial BSSB signal (Fig. 3a) was already
broad compared with that in Fig. 2a. Linewidth of the
product P1 also depends on concentration of GSH and
significantly decreases under high concentration of added
glutathione (Fig. 3d). The signal of the product P2 consists of
two multiplets with well-resolved structure. Their LWs were
not affected by variation in GSH concentration. In the excess
of GSH, only the product, P2, was observed (Fig. 3d).
°
tion at 170 C (¾1 mmHg), yielding about 80% purity. M.p.
°
211–213 C. Found: m D 449.9254 C14H2F8O4S2. Calculated:
m D 449.9267.
NMR measurements
19F NMR spectra were measured on a Bruker AVANCE
200 spectrometer in 5-mm tubes. All solutions were titrated
to pH 7.0, and bubbled by Ar for 10 min before mixing
and measurements. All samples contained 10% D2O used as
a lock. Standard ꢀ/2 pulse sequence with 1H decoupling
was used. Standard settings were as follows: number
of scans, 32–512; sweep width, 30 ppm; LB D 0–10 Hz;
acquisition time, 1.5 s; delay between scans, 5 s. Linewidths,
LW, of individual 19F NMR spectra lines of BSSB and BSH
compounds were determined by fitting calculated spectra to
the experimental ones using NUTS95 (Acorn NMR) program
package.
Preparation of blood plasma and erythrocytes for
determination of SH compounds
Wistar rats (male, 100 g) were obtained from Institute of
Cytology and Genetics (Novosibirsk, Russia). Samples were
prepared using a protocol described earlier with slight
modifications.11 Rats were anesthetized by diethylether.
Figure 1. Chemical structures of the fluorinated compounds, BSSB and BSH.
Copyright  2005 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2005; 43: 902–909

904
D. L. Potapenko et al.
(a)
(b)
31.0
30.8
30.6
(ppm)
23.5
23.3
23.1
30.9
30.7
23.5
23.3
23.1
(ppm)
Figure 2. (a) 19F NMR spectrum of the solution of the BSSB in DMSO-υ6. (b) Calculated spectrum with parameters Jij listed in
Table 1. Values of chemical shifts of the signals of BSSB in DMSO-υ6 are ꢁ(F2ꢂ D ꢁ(F6ꢂ D 30.8 ppm, ꢁ(F3ꢂ D ꢁ(F5ꢂ D 23.3 ppm
relative to signal of C6H6.
(b)
(a)
40
(c)
35
30
25
(ppm)
20
15
10 40
(d)
35
30
25
(ppm)
20
15
10
40
35
30
25
20
15
10 40
35
30
25
20
15
10
(ppm)
(ppm)
Figure 3. 19F NMR spectra of 5-mM BSSB solution in 0.1 M phosphate buffer pH 7.0 measured before (a) and after addition of
glutathione in final concentration 2 mM (b), 5 mM (c), and 10 mM (d). Number of scans is 32 for each spectrum.
Table 1. 19F NMR spectra parameters of BSSB and BSH in DMSO and phosphate buffer (a)
BSSB
BSH
JFi,Fj, Hz,
F2
F3
25
F5
F6
F2
F3
F5
F6
F2
F3
F5
ꢀ12.6
ꢀ2.9
ꢀ2.9
ꢀ12.6
25.0
26.5a
ꢀ11.2a
ꢀ2.2a
ꢀ5.8a
ꢀ11.2a
26.6a
Chemical shifts
30.8
30.4a
23.3
20.4a
23.3
20.4a
30.8
30.4a
25.8
26.1a
17.2
16.0a
17.2
16.0a
25.8
26.1a
a Data obtained in 0.1 M phosphate buffer, pH 7.0
The comparison of the spectrum shown in Fig. 3d with
the spectrum of separately synthesized fluorinated benzoic
acid, BSH, allows assigning the product P1 to tetrafluoro-
4-mercaptobenzoic acid (Fig. 1). The corresponding spectral
parameters of this compound are listed in Table 1.
Figure 4 shows the dependence of concentration of
the disulfide BSSB and its products, P1 and P2, on GSH
concentration. The measurements were performed in 0.1-
M phosphate buffer, pH 7.0, and concentrations were
calculated using integral intensity of 19F NMR signals of
Copyright  2005 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2005; 43: 902–909

Thiols determination using 19F NMR 905
BSH (P1)
250
200
150
100
50
10.0
8.0
6.0
4.0
2.0
0.0
BSSB
BSSG (P2)
0
0
2
4
6
8
10
12
14
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
C(GSH) (mM)
C (mM)
Figure 4. Dependence of the concentrations of the disulfide,
Figure 5. Dependence of linewidth, LW, of the individual 19
F
BSSB (ꢀ), and its products, BSH (P1, ꢁ), and BSSG (P2,
)
NMR line of BSSB ( ) and BSH () on concentration of added
°
formed in the reaction with glutathione, on GSH concentration.
GSH was added to 5 mM solution of BSSB in 0.1 M phosphate
buffer, pH 7.0, and NMR spectra were registered afterwards.
Number of scans was 32 for each spectrum. The
°
BSH and BSSB. BSH (BSSB) was added to a 5-mM solution of
BSSB (BSH) in 0.1 M phosphate buffer, pH 7.0, and 19F NMR
spectra were measured. LW values were determined by fitting
the calculated spectra using parameters listed in Table 1 to the
experimental spectra yielding LW (Experimental). Solid lines
correspond to linear approximation LWi D Ai C Bi ð C with
parameters ABSH D 9.95 š 1.52 Hz, BBSH D ꢃ123 š 1.5ꢂ
ð 103 Hz ð Mꢀ1; ABSSB D 6.27 š 4.2 Hz,
concentrations were calculated using integral intensities of the
corresponding NMR signals. Solid lines correspond to
numerous calculations using Eqns (3–7) with parameters
K1, K2 > 103; K1/K2 D 11.6.
BBSSB D ꢃ59.2 š 0.85) ð 103 Hz ð Mꢀ1. ( ) – Dependence of
ꢂ
linewidth of whole multiplet signal of BSSB on the
the corresponding compounds. It is clearly seen that the
concentration of formed BSH is equal to the concentration of
added GSH in each experiment. The solid lines correspond
to the best fit of the Eqns (3–7) to the experimental data
(Discussion for details).
concentration of BSH added.
The LW of individual lines of BSSB were determined
by fitting the calculated spectra (Experimental) using spectral
parameters listed in Table 1 to the experimental spectra.
The obtained dependencies of LW of BSSB on concentration
of added proteins is shown in Fig. 7. The data are in a
The line broadening of the NMR signals of BSSB and BSH
might originate from the thiol–disulfide exchange between
these two compounds. To test this hypothesis, the 19F NMR
spectra of a mixture of these compounds in phosphate buffer
were measured. It was found that linewidth of each of
the compounds is proportional to the concentration of other
compound (Fig. 5). Linewidth of the whole multiplet became
proportional to the concentration of added reagent only
when its concentration was higher than 1.5 mM. The twofold
difference in the slopes of the relationships should be noted,
apparently attributed to the difference in stoichiometry
of the corresponding thiol–disulfide exchange reactions
(Discussion).
Figure 6 shows 19F NMR spectra of 5 mM BSSB with
0.67 mM albumin (ASH; Fig. 6a) or 0.85 mM hemoglobin
(Hem; Fig. 6b) in 0.1 M phosphate buffer, pH 7.0. An
appearance of two new signals at 15.9 and 26.1 ppm
is clearly seen. Moreover the additional new signal at
18.0 ppm appears in the case of hemoglobin addition.
Spectra lines of all components are significantly broad-
ened. The ratio of intensities of the signals of BSSB and
product formed can be measured only in the case of
addition of albumin and is equal to 14.4 š 1.5 (Fig. 6a).
An accurate determination of the corresponding ratio
in the experiment with hemoglobin (Fig. 6b) is hardly
possible due to strong line broadening and spectra
overlapping.
good agreement with a linear approximation, LWBSSB
D
Ai C Ai ð Ci (i D ASH, Hem, GSH) with parameters Ai and
Bi being specified in the caption. The slopes of the lines
increase in the series ASH >Hem > GSH.
Using BSSB for the determination of concentration
of SH groups in rat blood
Figure 8a, b shows 19F NMR spectra measured after addition
of BSSB solution to cellular media of erythrocytes and blood
plasma, correspondingly. An 19F NMR spectral multiplet
line of a new product in the region of 31 ppm is clearly
seen in Fig. 8a. The second multiplet line of the product
overlaps with broad line of BSSB at 20.4 ppm. On the basis
of previously discussed data, this product was assigned to
BSSG (Discussion). Broad lines of formed BSH are also clearly
seen. The ratio of integral intensities, BSSG/BSSB, is equal
to 0.097 š 0.024. Linewidth for individual line of BSSB was
found to be equal to 33 Hz.
Addition of blood plasma to the solution of BSSB widens
NMR lines of the latter one to 94 Hz (Fig. 8b). The ratio
of integral intensities of BSH and BSSB on the spectrum
can be estimated as 0.054 š 0.014. Taking into account that
albumin, ASH, is the main SH-containing compound in the
Copyright  2005 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2005; 43: 902–909

906
D. L. Potapenko et al.
(a)
(b)
40
35
30
25
20
15
10 40
35
30
25
20
15
10
(ppm)
(ppm)
Figure 6. 19F NMR spectra of 5-mM solution of BSSB in 0.1 M phosphate buffer, pH 7.0, measured after addition of 0.67 mM
albumin (a) and 0.85 mM hemoglobin (b). Number of scans is 64 in (a) and 256 in (b).
(a)
350
300
250
200
150
100
50
32
31
30
29
90 87 40
(b)
35
30
(ppm)
25
20
15
10
0
0.0
0.5
1.0
C (mM)
1.5
2.0
Figure 7. Dependence of LW of 19F NMR line of BSSB on
concentration of added albumin (ASH, ), hemoglobin (Hem,
°
), and glutathione (GSH, ). ASH and Hem were added to a
ꢂ
5-mM solution of BSSB in 0.1 M phosphate buffer pH 7.0, and
NMR spectra were measured. Experimental data shown in
Fig. 4 were used to plot dependence for glutathione (). Solid
lines correspond to linear approximation of the experimental
data, LWBSSB D Ai C Bi ð Ci where AASH D 18.5 š 4.0 Hz,
BASH D ꢃ412 š 10ꢂ ð 103 Hz ð Mꢀ1; AHem D 9.35 š 3.6 Hz,
BHem D ꢃ231 š 6.7) ð 103 Hz ð Mꢀ1; AGSH D 15.5 š 1.1 Hz,
BGSH D ꢃ60.2 š 0.98) ð 103 Hz ð Mꢀ1. Dotted line
corresponds to straight line LWBSSB D AGSH C k ð [GSH] with
k D 63 ð 103 Hz ð Mꢀ1
.
90 87 40
35
30
25
20
15
10
(ppm)
Figure 8. 19F NMR spectra of BSSB solution measured after
its addition to cellular media of erythrocytes (a), and blood
plasma (b) from wistar rats (Experimental). 500 µl of 5-mM
solution of BSSB in 0.1 M phosphate buffer, pH 7.0, 1.25 mM
CF3COOH, was mixed with 100 µl of D2O and 400 µl of the
cellular media (a) or blood plasma (b). Number of scans was 64
for each spectrum. Signal at 87.65 ddm corresponds to anion
of trifluoroacetic acid and was used as a reference.
blood plasma (5) these data can be used for the determination
of ASH concentration.
Using all obtained data and taking into account that the
initial linewidth of BSSB before addition of thiol is 11 Hz
(data not shown), the concentration of GSH and albumin
can be determined. Independently, it can be done from
integral intensities of the signals and from an analysis of
the broadening of BSSB NMR lines. The obtained results are
listed in the Table 2.
Line broadening of BSSB due to the exchange reaction
(8) (Discussion) with a measured rate constant, k D 63 ð
103 Mꢀ1 sꢀ1, was used for the determination of concentration
of GSH. Data presented in Fig. 7 were used as calibration
Copyright  2005 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2005; 43: 902–909

Thiols determination using 19F NMR 907
Table 2. Concentration of ASH in blood plasma, and GSH in erythrocyte and
whole rat blood determined by 19F NMR approach using BSSB
From integral
intensities
From
linewidth
Literature
data
[GSH] in erythrocytes, mM
[GSH] in blood, mMa
[ASH] in blood plasma, mM
1.86 š 0.26
0.93 š 0.23
0.56 š 0.14
1.74 š 0.28
0.87 š 0.14
0.5 š 0.8
1.9 š 0.311
0.95 š 0.211
0.61 š 0.155
a Numbers were calculated taking into account experimentally determined fraction
of erythrocytes in the blood samples.
K1, K2 > 103. Note, that equimolarity of BSH and BSSG
formed at low concentration of added GSH is explained by
a ¾10-fold higher value of K1 in comparison with K2. Large
values of K1 and K2 are characteristic for thiol–disulfide
exchange reactions of aryldisulfides with low-molecular-
weight thiols (e.g. the corresponding rate constant for 5,50-
dithiobis(2-nitrobenzoic acid), Ellman’s reagent, with GSH
is 1 ð 103)7, and apparently are explained by the influence of
aryl fragment.
curve for the determination of albumin concentration.
Results presented in Table 2 are in good agreement with
literature data. This demonstrates the applicability of the
method for the determination of thiols ex vivo.
DISCUSSION
Reaction mechanism
Thiol–disulfide exchange reaction between the disulfide,
BSSB, and the thiol, GSH, results in the formation of a
monothiol, BSH, mixed disulfide, BSSG, and disulfide GSSG,
according to the reactions:
Line broadening
Linear dependence of linewidth of the signals of BSSB and
BSH on concentrations of added BSH and BSSB (Fig. 5)
supports the existence of exchange reactions between these
compounds.
K1
BSSB C GSH ꢀ! BSSG C BSH
ꢃ1ꢂ
ꢃ2ꢂ
K2
BSSG C GSH ꢀ! GSSG C BSH
k
BSSB C BSH ꢀ! BSH C BSSB
ꢃ8ꢂ
where K1 and K2 are the equilibrium constants at pH 7.0.
According to these reactions, the 19F NMR multiplet signals
of the product P2 at 31.2 and 20.43 ppm (Fig. 3) should be
assigned to the mixed disulfide BSSG.
Theequilibriumconcentrationsofthecompounds[BSSB],
[BSH], [BSSG], [GSH] and [GSSG] are described by the
following equations:
In the case of the intermediate exchange between several
positions, the linewidth of ith NMR line, LWi, can be
calculated in accordance with equation:15
2
T2obs
2
T20
2
ꢄi
LWi D
D
C
ꢃ9ꢂ
where ꢄi – the average lifetime in the ith state, T20 and T2obs
are intrinsic and observed time of spin–spin relaxation,
correspondingly. Taking into account that one molecule of
BSSB has two exchangeable fluorine-containing aryl groups
while BSH has only one group, the following equations for
the LWs of NMR signals of these compounds are obtained:
[BSSG][BSH]
[BSSB][GSH]
K1 D
K2 D
ꢃ3ꢂ
ꢃ4ꢂ
[GSSG][BSH]
[BSSG][GSH]
According to these equations, the equilibrium concentra-
tions of BSSB and BSSG are inversely proportional to the
concentration of glutathione, in qualitative agreement with
the data shown in Fig. 3.
Note that the BSSB molecule has two fluorine-containing
aryl groups while molecules of BSSG and BSH have only one
aryl group. The corresponding balance equations are:
LWBSSB D LW0BSSB C k[BSH]
LWBSH D LW0BSH C 2k[BSSB]
ꢃ10ꢂ
ꢃ11ꢂ
where LWBSSB, LW0BSSB, and LWBSH, LW0BSH are current and
initial LWs of the NMR signal of BSSB and BSH compounds,
correspondingly. These equations explain experimental
linear dependencies for LWBSH and LWBSSB on [BSSB]
and [BSH], correspondingly (Fig. 5) with the value of
rate constant of exchange reaction (8) being equal to k D
ꢃ63 š 11ꢂ ð 103 Mꢀ1 sꢀ1. Moreover, they explain the twofold
larger slope in the dependence of linewidth of BSH compared
to that of BSSB.
2[BSSB]0 D 2[BSSB] C [BSSG] C [BSH]
[GSH]0 D [GSH] C [BSSG] C 2[GSSG]
[BSSB]0 D [BSSB] C [BSSG] C [GSSG]
ꢃ5ꢂ
ꢃ6ꢂ
ꢃ7ꢂ
where [BSSB]0 and [GSH]0 are the initial concentrations of
BSSB and GSH, correspondingly. The system of Eqns (3–7)
cannot be solved analytically. Fitting numerical calculations
using Eqns (3–7) with varying parameters K1 and K2 to
the experimental data (Fig. 4) yields K1/K2 D 11.6 and
Linewidth of 19F NMR lines of BSSG is comparatively
small and does not depend on concentrations of BSSB
and BSH (Fig. 3). It can be explained by the low values
Copyright  2005 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2005; 43: 902–909

908
D. L. Potapenko et al.
of rate constants of thiol–disulfide exchange of BSSG
with thiols. Indeed, it was found that addition of 20 mM
of GSSG to 5 mM of BSH does not significantly change
the shape of NMR 19F spectrum during half an hour.
Nevertheless, the appearance of a low signal of BSSG (5%
from added BSH, data not shown) was observed after five
hours of incubation. Therefore, the value of rate constant
of the reverse reaction (2) can be estimated to be in the
range 1.4 ð 10ꢀ3 –1.4 ð 10ꢀ4 Mꢀ1 cꢀ1. Taking into account the
obtained estimations of the equilibrium constants K1 and K2,
the value of rate constant of the direct reaction (2) should
be about 1.4–14 Mꢀ1 cꢀ1. This latter value is in the range of
literature data for the rate constants for the thiol–disulfide
exchange reactions.7 On the basis of the obtained data, it
should be expected that the NMR lines of BSSB broaden
after addition of GSH, due to exchange reaction (8) with
BSH formed. The calculations of the dependence of the 19F
NMR linewidth of BSSB on GSH concentration using a rate
constant of the reaction (8) k D 63 Mꢀ1cꢀ1 are in excellent
agreement with experimental data (Fig. 7).
Moreover, using the obtained value for the rate constant
of the exchange reaction (8) and Eqn (10) it is easy to
show that linewidth of the BSSB signals under a BSH
concentration of 1.5 mM should be about 100 Hz. This value
is 10 times higher than the linewidth of BSSB before addition
of BSH and twice the width of the whole multiplet signal
of BSSB (¾50 Hz, Fig. 3). These estimations explain the
experimentally observed linear dependence of the width
of whole multiplet signal of BSSB on BSH concentration
when [BSH] >1.5 mM (Fig. 5).
Initial broadening of NMR 19F spectral lines of BSSB
(LW0BSSB) and BSH (LW0BSH) can be explained by the presence
of BSH and BSSB, respectively, in their solutions. The
presence of BSSB in BSH solutions is easily explained by
oxidation of BSH by oxygen in the environment. Slight
concentrations of BSH in the solution of BSSB can originate
from hydrolysis of the latter.
The method also allows the measurement of thiol groups
within proteins or other macromolecules. However, an
interpretation of the NMR data in this case may be more
complex and less informative as was demonstrated for
the case of hemoglobin and albumin. Bovine hemoglobin
is a quaternary-structure protein containing two pairs
of polypeptide chains and four groups of heme. This
structure contains two cysteine residues easily available
for thiol–disulfide exchange.2–4 Therefore, at least three
different products can be formed in the reaction of Hem
with BSSB. However, in the case of complete modification
of the SH groups of hemoglobin, the formation of a single
double disulfide is expected.
the formation of one additional product in the reaction of
ASH with BSSB is expected:
BSSB C ASH ꢀ! BSSA C BSH
ꢃ13ꢂ
The formation of the disulfide of albumin as well as the
disulfide of hemoglobin is sterically hindered because of the
high molecular weights of the reagents.
NMR spectral lines of BSSA should overlap with broad
lines of BSSB (Fig. 6a). Therefore the expected ratio of the
integrated intensities of experimental NMR 19F lines of BSSB
and BSH for addition of 0.67 mM of ASH to 5 mM BSSB
solution under complete modification of albumin (13) should
be (2 ð ꢃ5 ꢀ 0.67ꢂ C 0.67ꢂ/0.67 D 13.9 which is in a good
agreement with experimentally obtained value, 14.4 š 1.5
(Fig. 6a).
From the number of available SH groups, the expected
line broadening of 19F spectral lines of BSSB due to the
exchange reaction (8) after addition of ASH should be
comparable with that observed after addition of GSH and
about two times higher after addition of Hem. However, as
is seen from Fig. 7, the real slopes in these cases are seven
and four times larger, respectively. Moreover, the linewidth
of formed BSH at concentrations of BSSB larger than 4 mM
after addition of ASH to BSSB is equal to 75 Hz instead of
expected 281 Hz (Eqn 11). These facts cannot be explained
by an intermediate exchange reaction model and are the
subjects of ongoing investigation.
The application of the approach for thiols detection in
rat blood shows good agreement with literature data, and,
therefore, demonstrates the applicability of the method for
the determination of thiols ex vivo.
CONCLUSIONS
The technique developed for thiol determination described
here is based on thiol–disulfide exchange between the
newly synthesized disulfide compound, BSSB, and thiols.
Sensitivity of linewidth of BSSB to the concentration of
SH compound allows determination of 50 µmol lꢀ1 of
thiol (corresponding to a linewidth broadening of 3 Hz)
which is slightly lower than the sensitivity of EPR-based
technique using biradical disulfides.10,11 However, the 19F
NMR spectrum allows for an unambiguous conclusion
about the structure of the presented thiol (Fig. 8a). Stability
of the formed diamagnetic compounds in the reducing
cellular environment provides a significant advantage of
NMR approach over EPR detection.16 The NMR technique
also can be performed in nontransparent samples, which
is not possible using standard optical methods.6–9 The
rate constants for the thiol–disulfide exchange reactions
of the BSSB with thiols studied are too high for the
kinetics to be detected. Therefore an application of an excess
of BSSB over the thiols is required for thiols detection
using measurement of the equilibrium, BSSB/BSH. This
significantly limits the number of biological applications
of the proposed technique using this particular fluorinated
disulfide probe. Measurement of the line broadening of
BSSB for the thiol detection is an alternative approach that
Hem(SH)2 C 2BSSB ꢀ! Hem(SSB)2 C 2BSH
ꢃ12ꢂ
NMR 19F lines of this double disulfide should overlap
with broad lines of BSSB (Fig. 6b). While the signal at
18.00 ppm can be attributed to one of the NMR lines of
the product, the exact analysis of the spectrum presented in
Fig. 6b is complicated because of line broadening.
Albumin, ASH, is a protein with molecular weight about
65 kDa, which contains one available SH group.5 Therefore,
Copyright  2005 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2005; 43: 902–909

Thiols determination using 19F NMR 909
4. Schroeder WA, Shelton JR, Shelton JB, Robertson B, Babin DR.
Arch. Biochem. Biophys. 1967; 120: 1.
has certain advantages in the case of line-overlapping and
complexity of integration of the observed signals.
Synthesis of fluorinated aliphatic disulfides may over-
come some of the mentioned problems. These compounds
are characterized by the rate and equilibrium constants
for thiol–disulfide exchange reaction lying in the range
0.3–1.0 Mꢀ1 sꢀ1, respectively.7 It may allow the use of lower
disulfide concentrations and registering the kinetics of the
exchange reactions.
5. Peters T. All About Albumin: Biochemistry, Genetics, and Medical
Applications. Academic Press: San Diego, 1996.
6. Boyne AF, Ellman GL. Anal. Biochem. 1972; 46: 639.
7. Packer L (ed.). Biothiols. Part A: monothiols and dithiols,
protein thiols and thiyl radicals. Methods in Enzymology, vol. 251.
Academic Press: New York, 1995.
8. Jocelyn PC. Biochemistry of the SH Groups. Academic press:
London, 1972.
9. Jocelyn PC. Methods Enzymol. 1987; 143: 44.
10. Khramtsov VV, Yelinova VI, Weiner LM, Berezina TA, Mar-
tin VV, Volodarsky LB. Anal. Biochem. 1989; 182: 58.
11. Khramtsov VV, Yelinova VI, Glazachev YI, Reznikov VA,
Zimmer G. J. Biochem. Biophys. Methods 1997; 35: 115.
12. Khramtsov VV, Grigor’ev IA, Foster MA, Lurie DJ. Antioxidat.
Redox Signal. 2004; 6: 667.
Acknowledgements
The work was partially supported by RFBR grants 05-04-48632
and 05-04-48483 and NIH KO1 EB03519 and HL 53333. D.I.P. is
very grateful to grant INTAS 03-55-1740 and grant of Ministry of
Education RF A03-2.11-822.
13. Fielding HC, Shirley IM. J. Fluorine Chem. 1992; 39: 15.
14. Emsley JW, Feeney J, Sutcliffe LH. High Resolution Nuclear
Magnetic Resonance Spectroscopy, vol. 1. Pergamon Press: London,
1966.
15. Carrington A, McLachlan AD. Introduction to Magnetic Resonance
with Application to Chemistry and Chemical Physics. Happer &
Row: London, 1967.
REFERENCES
1. Schafer FQ, Buettner GR. Free Radical Biol. Med. 2001; 30:
1191.
2. De Bruin SH, Joordens JJ, Rollema HS. Eur. J. Biochem. 1977; 75:
211.
3. Bogner P, Csutora P, Cameron IL, Wheatley DN, Miseta A.
Biophys. J. 1998; 75: 3085.
16. Khramtsov VV, Berliner LJ, Clanton TL. Magn. Reson. Med. 1999;
42: 228.
Copyright  2005 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2005; 43: 902–909