The use of hydrogen peroxide for closing disulfide bridges in peptides

Article abstract of DOI:10.1023/B:RUBI.0000023093.05123.31

The use of hydrogen peroxide for the formation of disulfide bridges was studied in 15 peptides of various lengths and structures. The oxidation of peptide thiols by hydrogen peroxide was shown to proceed under mild conditions without noticeable side reactions of Trp, Tyr, and Met residues. Yields of the corresponding cyclic disulfides were high and mostly exceeded those obtained with other oxidative agents, in particular, iodine. It was established that the use of hydrogen peroxide in organic medium also provided sufficiently high yields when large-scale syntheses of oxytocin and octreotide (up to 10 g) were carried out.

Full text of DOI:10.1023/B:RUBI.0000023093.05123.31

Russian Journal of Bioorganic Chemistry, Vol. 30, No. 2, 2004, pp. 101–110. Translated from Bioorganicheskaya Khimiya, Vol. 30, No. 2, 2004, pp. 115–125.
Original Russian Text Copyright © 2004 by Sidorova, Molokoedov, Az’muko, Kudryavtseva, Krause, Ovchinnikov, Bespalova.
The Use of Hydrogen Peroxide for Closing Disulfide Bridges
in Peptides
1
M. V. Sidorova*, A. S. Molokoedov*, A. A. Az’muko*, E. V. Kudryavtseva*,
E. Krause**, M. V. Ovchinnikov*, and Zh. D. Bespalova*
*Cardiological Research and Production Association, Russian Ministry of Health,
ul. Tret’ya Cherepkovskaya 15a, Moscow, 121552 Russia
**Institute of Molecular Pharmacology, Berlin, Germany
Received October 2, 2002; in final form, November 29, 2002
Abstract—The use of hydrogen peroxide for the formation of disulfide bridges was studied in 15 peptides of
various lengths and structures. The oxidation of peptide thiols by hydrogen peroxide was shown to proceed
under mild conditions without noticeable side reactions of Trp, Tyr, and Met residues.Yields of the correspond-
ing cyclic disulfides were high and mostly exceeded those obtained with other oxidative agents, in particular,
iodine. It was established that the use of hydrogen peroxide in organic medium also provided sufficiently high
yields when large-scale syntheses of oxytocin and octreotide (up to 10 g) were carried out.
Key words: chemical synthesis, disulfide bond, hydrogen peroxide, natural peptides and their analogues
1
INTRODUCTION
idues between Cys residues [3, 4]. The cyclization of
free thiols (variant A) by the air oxygen usually leads to
low yields of target products (9–15%) [2]. The applica-
tion of potassium ferricyanide or dimethyl sulfoxide
usually result in homogeneous reaction mixtures and
the yields of cyclic products are considerably higher
(from 20 to 60%, and, in some cases, up to 80%, which
depends on the peptide structure [2, 5]). However, a
multistage purification of product is necessary for the
removal of excess of these oxidative agents [1]. A very
attractive one-step formation of the disulfide bridges by
the action of iodine is often accompanied by side reac-
tions [1, 2, 6–8] and has only a limited use in the case
of Trp- and Met-containing peptides.
The development of peptide chemistry currently
allows the preparation of practically any peptides, their
analogues, and even small proteins; however, the syn-
thesis of peptides with internal disulfide bridges is still
2
a real challenge [1, 2]. Low yields at the stage of for-
mation of SS bonds often reduce to zero all the efforts
made for the preparation of linear precursors. The prep-
aration of cyclic peptide disulfides from the corre-
sponding SH-precursors (variant A) and the direct con-
version of the cysteine-protected derivatives into cyclic
products (deprotection of the SH-groups with simulta-
neous cyclization, variant B) are most widely used
among the diversity of methods known for the synthesis
of disulfide-containing peptides. As a rule, in both
cases, the cyclization is carried out in very dilute solu-
tions, with the peptide concentration being of 10–4–
10−5 M in order to avoid an intermolecular aggregation
and side reactions [1]. The directed formation of S–S
bonds in the highly diluted solutions significantly
depends on structural peculiarities of a peptide, in par-
ticular, on the nature and the number of amino acid res-
Our synthesis of peptides with intramolecular disul-
fide bridges on a scale far exceeding usual laboratory
needs prompted us to search for such oxidation method
for SH groups in various peptides that could provide
sufficiently homogeneous reaction mixtures and a sim-
ple (preferably one-step) isolation of highly pure prod-
uct. Hydrogen peroxide attracted our attention, because
we found separate references [9] to its application for
cyclization of peptide thiols. We compared the effi-
ciency of various oxidative agents—O2, I2,
K3[Fe(CN)6], and H2O2—and demonstrated the advan-
tages of hydrogen peroxide application to the disulfide
bond formation in the fragments of HIV gp41 glyco-
protein from various strains (Z3, ELI, BRU, and ROD)
[6, 10, 11]. Our conclusions about the efficiency of
hydrogen peroxide are in a good agreement with the
results in [12], also devoted to a comparison of different
cyclization methods.
1
Corresponding author; phone: +7 (095) 414-6716; fax: +7 (095)
414-6786; e-mail: peptide@cardio.ru
The abbreviations recommended by the IUPAC–IUB commission
2
in Eur. J. Biochem., 1984, vol. 183, pp. 9–37, are used along with
the following ones: AcOH, acetic acid; Acm, acetamidomethyl;
DIC, N,N'-diisopropylcarbodiimide; dF, D-Phe; dW, D-Trp; ESI
MS, electrospray ionization mass spectrometry; ET, endothelin;
Fmoc, 9-fluorenylmethyloxycarbonyl; MALDI MS, matrix-
assisted laser desorption and ionization mass spectrometry; Pmc,
2,2,5,7,8-pentamethylchroman-6-sulfonyl; β-Mpa, β-mercapto-
propionic acid; NMP, N-methylpyrrolidone; and u-PAR1, uroki-
nase receptor.
1068-1620/04/3002-0101 © 2004 MAIK “Nauka /Interperiodica”

102
SIDOROVA et al.
Gln, and Pmc for Arg. The Fmoc group was chosen for
A220
the temporary protection of the Nα-amino function. Cys
residues were blocked by theAcm group for all the pep-
tides, except for the u-PAR1-(1–16) fragment (Table 1).
In this fragment, Cys3 was blocked by the Acm group,
whereas two remaining Cys residues (Cys6 and
Cys12), which are connected by the disulfide bridge in
the protein, were protected by the Trt group. The cou-
pling was achieved by the DIC/HOBt method. The pep-
tides were cleaved from the support by the treatment
with the reagent K [14]. The linear Acm-precursors of
peptides and the linear SH-precursor of the (1–16)-u-
PAR1 fragment were purified by preparative HPLC to
the purity of 95%. The linear precursors of oxytocin,
octreotid, and [βMpa1,Arg8]-vasopressin were pre-
pared by ZAO Sintez peptidov (Peptide synthesis)
using the conventional methods of peptide synthesis in
solution.
a b
0.5
Mercury acetate (2 equiv per SH function) in acidic
aqueous medium (30–50% AcOH or 30% formic acid)
was used to cleave Acm groups from all the peptides,
and hydrogen sulfide to subsequently remove mercuric
ions.
We conditionally divided all the synthesized pep-
tides into the following groups: peptides with one dis-
ulfide bridge containing no Met (P1–P5, ÄêIIr, APIIIr,
α-ANFr, the (1–14)-somatostatin, oxytocin, octreotide,
and a vasopressin analogue); Met-containing peptides
[ET-1 and u-PAR1-(1–16)]; and peptides with two dis-
ulfide bridges (ET-1 and ET-3).
0
20
min
Fig. 1. Analytical HPLC of the mixture of (a) oxidized
(SS form) and (b) reduced (SH form) of rat atriopeptin II
(Table 1) on an Ultrasphere ODS column (4.6 × 250 mm,
5 µm) eluted with a gradient of 70% acetonitrile in 0.05 M
phosphate buffer (pH 3) from 10 to 60% for 40 min.
Disulfide bridges were closed in aqueous or aque-
ous–organic (methanol–water, DMF–water, and diox-
ane–water) solutions [1] under the condition of high
dilution at the peptide concentrations of 0.5–1 mg/ml at
pH 6.5–8.5. The peptides without Met residues (Table 1)
were cyclized with the use of 5–15-fold excesses of
hydrogen peroxide, whereas 1.5–2-fold excesses were
applied to Met-containing peptides. The processes were
monitored by HPLC. The conditions of analytical
HPLC at which the corresponding SH- and SS-forms of
the peptides were well resolved were found for all of
the studied peptides (Fig. 1). Oxidation was completed
within 15–40 min for all the peptides studied. After the
completion of oxidation, the reaction mixture was acid-
ified with acetic acid to pH 4, and the target peptides
were isolated by the preparative HPLC on Diasorb
130T (C16). The resulting cyclic disulfides of the pep-
tides were characterized by amino acid analysis, mass
spectrometry, and analytical HPLC (Table 3 in the
Experimental section). Their yields are listed in Table 1.
The yields given in Table 1 and in the Results and Dis-
cussion section are not corrected to the content of pep-
tides after lyophilization for the sake of convenience at
comparing with literature data.
RESULTS AND DISCUSSION
This paper generalizes the results of our studies of
the hydrogen peroxide use for the formation of disul-
fide bonds in peptides of various lengths and structures.
Amino acid sequences of the compounds synthesized
are given in Table 1. Potentials and limitations of this
approach to the cyclization were studied in the case of
peptides with one or two disulfide bridges and with var-
ious cycle sizes (from 6 to 17 amino acid residues) that
involved Met, Tyr, and Trp residues sensitive to oxida-
tion. We also studied the reproducibility of cyclization
results at the increased synthesis scale.
The syntheses of antigenic determinants of gp41
(peptides P1–P3 and P5) [6, 10, 11], ET 1 and 3 [7], and
the 1–16 fragment of u-PAR1 [13] were previously
described. The P4 peptide (an antigenic determinant of
gp41 glycoprotein of HIV-2), rat atriopeptins II and III,
the rat 1–28 atrial natriuretic factor, and (1–14)-soma-
tostatin fragment were prepared by the solid phase
method of peptide synthesis on the Wang polymer. The
following protecting groups were used for the side
The closing of disulfide bridges in the peptides with
a relatively small cycle (P1–P5 peptides, oxytocin, and
functions of trifunctional amino acids: But for Ser, Thr, octreotide) proceeded without a considerable formation
Tyr, Asp, and Glu; Boc for Lys, Trt for His, Asn, and of side products. The content of cyclic peptides in the
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 30 No. 2 2004

Table 1. A comparison of yields of the peptide cyclic disulfides prepared by the treatment with iodine or with hydrogen peroxide under a high dilution
Yielda, %
Content in
Designation
Peptide
Amino acid sequence
lyophilizated, %
I2
H2O2
P1
HIV-1 gp41-(598–609) (Z3) fragment
HIV-1 gp41-(598–609) (ELI) fragment
HIV-2 gp41-(593–603) (ROD) fragment
HIV-2 gp-41-(586–607) (ROD) fragment
HIV-1 gp41-(601–617) (BRU) fragment
Oxytocin
H-LGLWGCSGKLIC-NHEt
51[11]
75–85[11]
85–90[11]
80–87[11]
82
82
85
89
90
92
94
87
88
86
82
91
82
84
91
80
P2
P3
ê4
ê5
H-LGIWGCSGKHIC-NHEt
56[11]
H-NSWGCAFRQVC-NHEt
66[11]
H-LQDQARLNSWGCAFRQVCHTTV-OH
H-WGCSGKLICTTAVPWNA-OH
H-CYIQNCPLG-NH2
<5
25[11]
85–90[11]
75
52
Octreotide
H-dFCFdWKTC-NHCH[CH(OH)CH3]CH2OH
βMpa-YFQNCPRG-NH2
–
80
Deamino-[Arg8]-vasopressin
Rat atriopeptin II
–
81
APIIr
H-SSCFGGRIDRIGAQSGLGCNSFR-OH
H-SSCFGGRIDRIGAQSGLGCNSFRY-OH
H-AGCKNFFWKTFTSC-OH
23[19, 20]
83
APIIIr
Rat atriopeptin III
26[19, 20]
50
Somatostatin-(1-14)
–
51
α-ANFr-(1–28) Rat atrial natriuretic factor-(1–28)
u-PAR1-(1–16) Urokinase receptor-(1–16) fragment
H-SLRRSSCFGGRIDRIGAQSGLGCNSFRY-OH
H-LRC(Acm)MQCKTDGDCRVEE-OH
H-CSCSSLMDKECVYFCHLDIIW-OH
H-CTCFTYLDKECVYYCHLDIIW-OH
20[19, 20]
62
–
–
–
30b[13]
26c[7]
20c[7]
ET-1
ET-3
Endothelin-1
Endothelin-3
a
The yields from the [Cys(Acm)] -derivatives of the corresponding linear peptides are given. The disulfide bridges were formed starting from 100–500 mg of the corresponding linear pre-
cursors of the peptides at their concentrations of 0.5–1.0 mg/ml. No less than three cyclization experiments were carried out for each of the listed compounds; yields varied within 10%.
To simplify the comparison of our results with literature data, the yields were not corrected for the content of peptides in the corresponding lyophilyzates.
2
b
c
d
The yields from the [Cys(SH)] -derivatives of the corresponding peptides are given.
2
The yields from the [Cys(Acm)] -derivatives of the corresponding linear peptides are given.
4
The results for the compounds synthesized using hydrogen peroxide are obtained by the quantitative amino acid analysis.

104
SIDOROVA et al.
Table 2. The results of model experiments on the disulfide bond formation in oxytocin and octreotide using hydrogen perox-
ide in aqueous and organic media at different concentrations of the starting dithiols
Content of target product in reaction
Peptide
Solvent
Concentration, mg/ml
mixture, % according to HPLC
Oxytocin
Water
1
5
82
58
50
87
71
72
58
98
98
96
94
93
10
1
DMF–CH3OH, 1 : 5
5
10
20
1
Octreotide
CH3OH
10
20
50
100
reaction mixtures after the oxidation was sufficiently high a thiol after the treatment with dithiothreitol and, pre-
(85–95%). The yields of cyclic products (see Table 1) sumably, was a dimer of somatostatin; the formation of
from the correspondingAcm-protected precursors were such dimers is characteristic of somatostatin [16].
also sufficiently high (from 69 to 90%).
Note that the yield of all the natural peptides with
one disulfide bridge we prepared by the hydrogen per-
oxide oxidation were significantly higher than those
obtained at the iodine treatment (Table 1).
The cyclization of deamino analogues of some nat-
ural peptides is known to proceed under rather rigorous
conditions. For example, the SS bond in oxytocin is
easily closed even on the exposure to air oxygen,
whereas the closing of SS bridge in deaminooxytocin
requires the treatment with K3[Fe(CN)6] [15]. In the
case of a vasopressin analogue containing a residue of
unnatural β-mercaptopropionic acid, the formation of
SS bond by the action of hydrogen peroxide (see Table 1)
proceeded as quickly and easily as in the natural pep-
tides. The yield of [βMpa1,Arg8]vasopressin was 81%.
The application of hydrogen peroxide to the cycliza-
tion of Met-containing peptides is regarded as problem-
atic due to a very easy oxidation of methionine into the
corresponding sulfoxide [12]. Previously, we published
the results of using hydrogen peroxide for the forma-
tion of SS bridges in Met-containing peptides in detail
[7, 13] and here only briefly describe the cyclization
conditions. We had already found that the process of
formation of SS bridge itself by the action of hydrogen
peroxide is not accompanied by any noticeable oxida-
tion of methionine into the corresponding sulfoxide.
This side reaction proceeds after the completion of
cyclization as a result of acidification of the reaction
mixture before isolation of the target disulfide. There-
fore, an increase in pH of the reaction mixture to 8–9,
the use of a small excess of hydrogen peroxide (no
more than two equiv), and bubbling of an inert gas
through the reaction mixture using a glass capillary for
30 min before the acidification allow a successful use of
hydrogen peroxide for Met-containing peptides ET-1
and u-PAR1-(1–16) too.
The formation of SS bonds in somatostatin (10 aa
between two Cys residues) and in atrial peptides (15 aa
between two Cys residues) was accompanied by a for-
mation of side products (possibly disulfide oligomers
[12]), which decreased the content of target cyclic dis-
ulfides in the reaction mixtures to 70–75% (according
to HPLC). In this case, the yields of the corresponding
products dropped to 50–60% (see Table 1). The forma-
tion of SS bridge in oxytocin was the most difficult. Its
yield was practically independent of the concentration
of starting compound (within the range from 0.2 to
1.0 mg/ml), but it significantly changed upon the tran-
sition from aqueous to aqueous–organic solutions in the
reaction medium: 26% in 0.1 M ammonium acetate
These conditions were used for a simultaneous for-
(pH 6.8), 35% in methanol–water (pH 6.8), and 51% in mation of two SS bridges in ET 1 and 3 [7]. The iso-
dioxane–water (pH 6.8). The side product isolated from mers with the natural positions of SS bridges were pre-
the reaction mixture after oxidation was converted into dominantly formed after the treatment with hydrogen
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 30 No. 2 2004

THE USE OF HYDROGEN PEROXIDE
105
Table 3. Characteristics of the cyclic peptides prepared with the use of H2O2
Molecular mass
Purity
Peptide
Empirical formula according
to HPLC
Amino acid analysis*
calcu- found
lated by MS
gp41-(586–607) ë108ç167N35O32S2
(P4)
96.0
2531.9 2531.4a D/N 1.9(2), T 1.7(2), S 0.8(1), E/Q 3.1(3), G 1.0(1), A 1.7(2),
V 1.8(2), L 2.0(2), F 1.1(1), H 1.0(1), W 0.6(1), R 2.2(2)
Oxytocin
97.0
98.2
98.0
1007.2 1007.7b D/N 1.0(1), E/N 1.1(1), G 1.0(1), I 0.9(1), L 0.9(1),Y 0.8(1)
1019.3 1020.3b T 0.9(1), F 2.0(2), W 0.7(1), K 1.0(1)
ë43ç66N12O12S2
Octreotide
ë49ç66N10O10S2
[βå‡1, Arg8]-
vasopressin
1069.2 1068.0b D/N 1.0(1), E/Q 1.1(1), G 1.0(1), Y 0.9(1), F 1.0(1),
R 1.0(1)
ë46ç64N14O12S2
APIIr
98.1
98.3
97.5
98.0
2386.6 2386.0a D/N 2.2(2), S 3.6(4), E/Q 1.2(1), G 5.2(5), A 1.0(1), I 1.9(2),
L 1.0(1), F 2.1(2), R 3.0(3)
ë98ç156N34O32S2
APIIIr
2549.8 2549.5a D/N 2.1(2), S 3.6(4), E/Q 1.2(1), G 5.2(5), A 1.0(1), I 1.9(2),
L 1.0(1),Y 0.9(1), F 2.1(2), R 3.1(3)
ë107ç165N35O34S2
Somatostatin-
(1–14)
1637.9 1640.5b D/N 1.1(1), í 1.8(2), S 0.9(1), G 1.0(1), A 1.0(1), F
2.9(3), W 0.7(1), K 2.1(2)
ë76ç104N18O19S2
α-ANFr
3062.4 3062.4a D/N 2.1(2), S 4.3(5), E/Q 1.2(1), G 5.2(5), A 1.0(1), I 1.8(2),
L 2.0(2), Y 0.9(1), F 2.0(2), R 4.9(5)
ë128ç205N45O39S2
*The contents of proline and cysteine in the hydrolyzates of the corresponding peptides were not determined.
a
Molecular mass was determined by ESI MS.
Molecular mass was determined by MALDI-MS, the [M + H] values are given.
b
+
peroxide as was the case with the air oxygen [17]. How-
ever, the ratio of the natural isomer and two other pos-
sible isomers of ET 1 was 3 : 1 : 0 with the use of air
oxygen and 10 : 1 : 0 with the use of hydrogen peroxide.
In the case of ET 3, this ratio was 4 : 1 : 2 [7]. Moreover,
the oxidation was completed already after 40 min,
whereas the oxidation by the air oxygen required sev-
eral hours.
Thus, we demonstrated a doubtless advantage of
using hydrogen peroxide for the conversion of various
peptide thiols into the corresponding disulfides. The
next stage of our investigation was the study of the
potential use of hydrogen peroxide for the synthesis of
large quantities of oxytocin and octreotide (both pep-
tides exhibit a hormone activity). Octreotide (sandosta-
tin) is a synthetic analogue of somatostatin proposed by
Sandoz [18]. It has the following structure:
H-D-Phe-Cys-Phe-D-Trp-Lys-Thr-Cys-NHCH[CH(OH)CH3]CH2OH.
Both peptides are therapeutic agents certified in cin and octreotide from their linear [Cys(Acm)]2-pre-
Russia, and the necessity of large-scale synthetic
schemes for these peptides (dozens and hundreds
grams), including the cyclization step, is evident. As
follows from Table 1, the use of hydrogen peroxide to
the SS bridge formation in oxytocin and octreotide is
promising under the conditions of high dilution (the
peptide concentration was approximately 1 mg/ml or
cursors were 75 and 80%, respectively. However, tech-
nologically, high dilutions are not good for large-scale
syntheses, because volumes of reaction mixtures dra-
matically increase. For example, even the cyclization of
20 g of the linear precursor of the corresponding pep-
tide requires a 20-l volume of the reaction mixture.
Therefore, we studied the effect of increasing concen-
≈10–3 M). Under these conditions, the yields of oxyto- trations of the linear precursors of the peptides on the
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 30 No. 2 2004

106
SIDOROVA et al.
I, %
100
2012.3
90
80
70
60
50
40
30
20
1006.1
3018.2
2077.4
10
0
4023.5
723.0
1000
2000
3000
4000
5000
6000
m/z
Fig. 2. Mass spectrum of the reaction mixture of the oxytocin preparation at 10-mg/ml concentration of its [Cys(SH) ]-precursor in
2
+
aqueous medium at pH 7.2. The calculated molecular mass of oxytocin is 1007.7. The values of [M + H] are given on the X-axis.
cyclization process. The results of this study are given ing the SS bridge formation facilitates the formation of
in Table 2.
intramolecular disulfides [4]. Our results on the synthe-
sis of the 1–14 fragment of somatostatin also confirm
these publications. Therefore, we tried to cyclize oxy-
tocin in an organic medium the more so, as the solubil-
ity of its linear precursor considerably decreased in
aqueous solution at pH 6.5–8.5. We failed to use meth-
anol as a solvent for this reaction, because the linear
peptide has a low solubility in it under the oxidation
conditions. Satisfactory results were obtained, when
the cyclization was carried out in a 1 : 5 DMF–metha-
nol mixture. One can see from Table 2 that, in this case,
the quality of reaction mixture does not decrease so
substantially with the increase in the concentration of
linear peptide to 10 mg/ml (approximately 10–2 M). The
oxytocin content was 71.6%, and no sharp decrease in
the area of peak corresponding to oxytocin was
observed.
The percent content of oxytocin dramatically
(almost two times) decreased in the reaction mixture
along with increase in the concentration of its linear
precursor (from 81.1% at the peptide concentration of
1 mg/ml to 49.8% at the concentration of 10 mg/ml)
when the SS bonds of oxytocin were formed with the
use of hydrogen peroxide in aqueous medium at pH 7.2
(see Table 2). In this case, the area of the peak corre-
sponding to oxytocin approximately 3.5 times
decreases in the reaction samples brought to equal con-
centrations. The most probable reason for this decrease
is the formation of the corresponding oxytocin oligo-
mers along with its dimeric forms. These oligomers
could be irreversibly sorbed on chromatographic sor-
bents or, under the HPLC conditions, gave no clear
peaks due to exclusion, which did not allow us to regis-
ter the oligomers. A mass-spectrometric analysis of the
reaction mixture of oxytocin preparation at the concen-
tration of 10 mg/ml confirmed the formation of not only
oxytocin and its dimer but also oxytocin trimers and tet-
ramers as well (see Fig. 2). It should be noted that, in
this case, we could not observe any correlation between
the ratios of the target product and by-products in the
HPLC chromatograms (see Table 2) and in mass spec-
tra (Fig. 2).
The solubility of the linear dithiol precursor of oct-
reotide is limited in aqueous media with pH 6.5–8.0
even at the concentration of 1 mg/ml, and no compari-
son of the results of its cyclization in aqueous and
organic solvents is possible. Bauer and Pless [18]
formed SS bridge in octreotide in highly dilute solu-
tions (0.25 mg/ml) using a dioxane–water mixture. Sur-
prisingly, the octreotide linear precursor was easily
cyclized in methanol without noticeable formation of
There are some reports in literature that the intro- oligomeric by-products in the wide range of concentra-
duction of organic solvents into reaction mixtures dur- tions (from 1 to 100 mg/ml) (Fig. 3). The content of tar-
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 30 No. 2 2004

THE USE OF HYDROGEN PEROXIDE
107
get product in the reaction mixture insignificantly
changed from 98 to 93% (Table 2) and no considerable
decrease in the area of peak corresponding to octreotide
(no more than 5–10%) was also observed in the sam-
ples of the oxidation reaction mixtures brought to the
equal concentrations (HPLC). We cannot now explain
this interesting phenomenon and can only presume that
the absence of intermolecular interactions at such high
concentrations of the peptide is associated with forma-
tion of a compact conformation of the octreotide mole-
cule in methanol that favors the predominant formation
of the intermolecular SS bridge. A mass-spectrometric
analysis of the reaction mixture (Fig. 4) demonstrated
that, under these conditions, the main oxidation product
was the target cyclic peptide, and the corresponding oli-
gomers were present in insignificant amounts.
A220
We chose 10 mg/ml as the working concentration
and used it for the preparation of oxytocin and oct-
reotide in organic solvent media starting from 20 g of
the linear dithiols (see the Experimental section). Note
that, in the model experiments, the corresponding linear
precursors were purified to >95% HPLC-purity (the
results are given in Table 2). However, when carrying
out the large-scale experiments, we used the corre-
sponding Acm- and SH-derivatives without any prelim-
inary purification, and the content of the main sub-
stance in the starting SH-peptides was 74% for oxyto-
cin and 82% for octreotide (according to HPLC). The
yields of oxytocin and octreotide in the preparative
experiments were 44 and 64%, respectively.
0.5
f
e
d
c
b
a
Thus, we demonstrated the advantages of hydrogen
peroxide for the formation of disulfide bonds in natural
peptides of various structures. A well-reproducible
technology of the oxidation of peptide thiols at high
concentrations into the corresponding disulfides by the
treatment with hydrogen peroxide in organic solvents
was developed by the examples of large-scale experi-
ments with oxytocin and octreotide.
0
20
40 min
Fig. 3.Analytical HPLC of (a) [Cys(SH) ]-precursor of oct-
2
reotide and the reaction mixtures of its oxidation by the
treatment with hydrogen peroxide in methanol at the pep-
tide concentrations of (b) 1, (c) 10, (d) 20, (e) 50, and (f)
100 mg/ml on an Ultrasphere ODS column (4.6 × 250 mm,
5 µm) eluted with a gradient of 70% acetonitrile in 0.05 M
phosphate buffer (pH 3) from 10 to 70% for 30 min.
EXPERIMENTAL
KH2PO4 (pH 3, buffer A) and the second, with 80%
acetonitrile in 0.1% trifluoroacetic acid (buffer B); the
flow rate was 1 ml/min and the detection was at 220 nm.
The preparative HPLC was carried out on a Beckman
Prep 350 chromatograph (United States) on a Diasorb-
130-C16T column (10 µm, ZAO Biokhimmak CT,
Russia) with detection at 226 nm. The cyclic peptide
disulfides were eluted with a gradient of 80% acetoni-
trile in 0.01 M ammonium acetate (pH 4.5). Acetoni-
trile for HPLC was from Technopharm (Russia).
Derivatives of L-amino acids (Bachem, Switzer-
land), phenol (Merck, Germany), mercury acetate, ani-
sole, p-cresol, thiophenol, EDT, DIC, βMpa, D-phenyl-
alanine, D-tryptophan (Fluka, Switzerland), and 50%
hydrogen peroxide (Solvay Interox, Belgium) were
used in this study. Before the use, hydrogen peroxide
was diluted with deionized water to the desired concen-
tration and titrated with 1 N solution of KMnO4.
Dichloromethane, N-methylpyrrolidone, piperidine,
methanol, and trifluoroacetic acid (Applied Biosystems
GmbH, Germany) were used in the solid phase peptide
syntheses. For analytical HPLC, a Gilson chromato-
graph (France) equipped with Ultrasphere ODS (4.6 ×
250 mm, 5 µm) column (Beckman, United States) and
Vydac C18 (4.6 × 250 mm, 5 µm, pore size of 300 Å)
column (Sigma, United States) was used. The first col-
Peptides were hydrolyzed with 6 N HCl containing
2% of thioglycolic acid at 110°ë for 24 and 48 h and
subjected to amino acid analysis after the hydrolysis on
a Biotronik 5001 automatic analyzer (Germany). Mass
spectra were registered on a Finnigan MAT TSQ 700
quadrupole mass spectrometer (Germany) with electro-
umn was eluted with 70% acetonitrile in 0.05 M spray ion source (API) (Finnigan MAT, United States)
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 30 No. 2 2004

108
SIDOROVA et al.
I, %
100
1020.2
90
80
70
60
50
40
30
20
10
0
2041.2
3063.0
1000
1500
2000
2500
3000
3500
m/z
Fig. 4. Mass spectrum of the reaction mixture of the octreotide preparation at the concentration of 10 mg/ml of its [Cys(SH) ]-pre-
2
+
cursor in methanol. The calculated molecular mass of octreotide is 1019.3. The values of [M + H] are given on the X-axis.
(ESI-MS) and on an Analytical Compact MALDI 4 2 h at 25°ë. The polymer was filtered off and washed
mass spectrometer (Kratos, UK) (MALDI-MS). Nitro-
gen in cylinders of o.s.ch. (special purity) grade (Rusia)
was used in this study.
with TFA (3 × 1 ml). The combined filtrate was evapo-
rated to the volume of about 1 ml and mixed with anhy-
drous ether. The precipitated product was filtered,
washed with ether and ethyl acetate, and dried. The
resulting [Cys(Acm)]2 derivatives of the P4, ÄêIIr,
ÄêIIIr, α-ANFr peptides, and the somatostatin-(1–14)
were dissolved in 10% acetic acid and fractionated on a
column with Diasorb (25 × 250 mm) in portions of
0.15–0.20 g. The peptides were eluted with a gradient
of 80% acetonitrile in 0.1% TFA (0.5%/min) from 10 to
50% at a flow rate of 12 ml/min. The fractions contain-
ing the target products were combined, evaporated, dis-
solved in water, and lyophilized.
Solid phase peptide synthesis. We synthesized
peptides on a 431A automatic peptide synthesizer
(Applied Biosystems). The linear Acm-precursors of
P4, ÄêIIr, ÄêIIIr, α-ANFr, and somatostatin-(1–14)
were prepared according to the Fmoc-strategy starting
every time from 0.25 mmol of the corresponding Fmoc-
aminoacyl polymer (Bachem, Switzerland). The con-
tent of starting amino acid in the Fmoc-aminoacyl poly-
mers was 0.5–0.7 mmol/g. A copolymer of styrene and
1% divinylbenzene with 4-hydroxymethylphenoxyme-
thyl anchoring group (the Wang polymer) was used as
a support. The amino acid chains were elongated
according to the standard programs for the one-step
coupling of Fmoc-amino acids. The synthetic cycle
involved a 20-min activation of the attached amino acid
(1 mmol) in the presence of equimolar amounts of DIC
and HOBt in NMP, deprotection of the α-amino groups
by the treatment with 20% solution of piperidine in
NMP for 20 min, coupling with the fourfold excess of
Our procedures of direct conversion of the
[Cys(Acm)]2-precursors into the cyclic disulfides by
iodine [6, 10, 11, 19, 20] and deprotection of Cys resi-
dues and the subsequent formation of the disulfide
bonds by hydrogen peroxide [6, 10, 11] were previ-
ously described and used for the preparation of P4 pep-
tides, atrial peptides, and somatostatin-(1–14) (Table 1)
acylating agent (1 mmol) in NMP for 90 min, and all and in the procedure of cyclization of a vasopressin
the necessary intermediate washings of the peptidyl
polymer.
analogue (see below). The yields and characteristics of
the cyclic peptides we prepared using hydrogen perox-
ide for the first time are given in Tables 1 and 3. The
content of peptide material in the lyophilized cyclic
products was determined by the quantitative amino acid
When the solid phase synthesis was completed, the
Nα-deprotected peptidyl polymers were suspended in a
mixture (10 ml) containing 82.5% TFA, 5% phenol, 5%
water, 5% thioanisole, and 2.5% EDT and stirred for analysis; it was from 82 to 90%.
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 30 No. 2 2004

THE USE OF HYDROGEN PEROXIDE
109
H-D-Phe-Cys-Phe-D-Trp-Lys-Thr-Cys -NH-CH-
båp‡-Tyr-Phe-Gln-Asn-Cys-Pro-Arg-Gly-NH2
([bMpa1,Arg8]vasopressin). βMpa(Acm)-Tyr-Phe-
[CH(OH)CH3]-CH2OH (octreotide). Crude trifluoro-
acetate of H-D-Phe-Cys-Phe-D-Trp-Lys-Thr-Cys-thre-
oninol (20 g of a raw synthetic product containing 82%
of the target product according to HPLC) was dissolved
in methanol (2 l), adjusted with 25% ammonia to pH
7.5–8 (6 ml), and mixed with a solution of hydrogen
peroxide (6 ml, 53.0 mmol). The reaction mixture was
Gln-Asn-Cys(Acm)-Pro-Arg-Gly-NH2
(0.2
g,
0.16 mmol) was dissolved in a mixture of acetic acid
(15 ml) and methanol (5 ml), mercuric acetate (0.2 g,
0.64 mmol) was added to the solution, the reaction mix-
ture was stirred for 1.5 h at 20°ë, and hydrogen sulfide
was bubbled through it for 30 min. The precipitate was
filtered off and washed with a 2 : 1 mixture of acetic
acid and methanol (15 ml). The filtrate was evaporated
to the volume of approximately 2 ml, and ethyl acetate
(20 ml) was added. The precipitated solid was filtered,
washed with ethyl acetate and hexane, and dried. The
resulting dithiol was dissolved in water (150 ml),
adjusted with 25% ammonia to pH 8.0, and treated with
a dilute solution of hydrogen peroxide with exact con-
centration (3 ml, 1.6 mmol). The reaction mixture was
stirred for 30 min with the HPLC monitoring of the SS
bridge formation. Acetic acid (2 ml) was added, and the
reaction mixture was evaporated to the volume of 50 ml
and fractionated on a column (25 × 250 mm) with Dia-
sorb. Its elution with a gradient of 80% acetonitrile in
0.01 M ammonium acetate (pH 4.5) (12 ml/min,
1%/min, from 20 to 60%), and the fractions containing
the target product were combined and evaporated. The
residue was dissolved in water and lyophilized, to give
the cyclic peptide; yield 0.13 g (81.0%); mass spectrum
(MALDI-MS): found m/z 1068.0 [M + H+], calculated
1069.24 for ë46ç64N14O12S2. Amino acid analysis: Asx
1.01(1), Glx 1.10 (1), Gly 1.00 (1), Tyr 0.88 (1), Phe
1.01 (1). The content of peptide in the lyophilized mate-
rial was 88%. The purity of the peptide was 98.0%
according to HPLC.
3
kept for 1 h at 25°ë with the HPLC monitoring. The
reaction mixture was acidified with acetic acid to pH 4,
evaporated to the volume of 1 l at 30°ë, and diluted
with water (1 l). Methanol was evaporated, and the
remaining solution was fractionated on a Diasorb col-
umn (50 × 250 mm). The target octreotide was eluted
with a gradient of 80% acetonitrile in 0.01 M ammo-
nium acetate (pH 4.5) (0.3%/min, from 20 to 60%) at a
flow rate of 50 ml/min; yield 8.6 g (64%); mass spec-
trum (MALDI-MS): found m/z 1020.3 [M + H]+, calcu-
lated 1019.3 for ë49ç66N10O10S2. Amino acid analysis:
Thr 0.87 (1), Phe 2.01 (2), Trp 0.69 (1), Lys 1.03 (1).
The content of peptide in the lyophilized material was
87%, and the purity of the peptide was 98.2% according
to HPLC.
REFERENCES
1. Andreu, D., Albericio, F., Sole, N.A., Munson, M.C.,
Ferrer, M., and Barany, G., Peptide Synthesis Protocols,
Pennington, W. and Dunn, M.B., Eds., Totowa, NJ:
Humana, 1994, pp. 91–169.
2. Sidorova, M.V., Kudryavtseva, E.V., and Evstigne-
eva, R.P., Usp. Khim., 1998, vol. 67, pp. 611–630.
3. Hardy, P.M., Ridge, B., Rydon, H.N., Dos, S.P., and Ser-
räo, F.O., J. Chem. Soc. (C), 1971, pp. 1722–1731.
H-Cys-Tyr-Ile-Gln-Asn-Cys -Pro-Leu-Gly-NH2
4. Moroder, L., Besse, D., Musiol, H.-J., Rudolph-
Bohner, S., and Siedler, F., Biopolymers, 1996, vol. 40,
pp. 207–234.
(oxytocin). A crude trifluoroacetate of H-Cys-Tyr-Ile-
Gln-Asn-Cys-Pro-Leu-Gly-NH2 (20 g, contains 74% of
the main product according to HPLC) was dissolved in
a 1 : 5 DMF–methanol mixture (2 ml), adjusted with
25% ammonia to pH 7.5–8 (6 ml), and a solution of
hydrogen peroxide (6 ml, 53.0 mmol) was added to the
peptide solution. The reaction mixture was kept for 1 h
at 25°C with the HPLC monitoring, acidified to pH 4
with acetic acid, and evaporated to the volume of
200 ml at 30°ë. The residue was mixed with ether. The
precipitate was filtered, washed with ether, dried, dis-
solved in 3% acetic acid (1 l) and fractionated on a Dia-
sorb column (50 × 250 mm). The target peptide was
eluted with a gradient of 80% acetonitrile in 0.01 M
ammonium acetate (pH 4.5) (0.3%/min, from 10 to
50%) at a flow rate of 50 ml/min. The yield of oxytocin
was 6.2 g (47%); mass spectrum (MALDI-MS): found
m/z 1007.7 [M + H]+, calculated 1007.2 for
ë43ç66N12O12S2. Amino acid analysis: Asx 0.98 (1),
Glx 1.08 (1), Gly 1.00 (1), Ile 0.90 (1), Leu 0.80 (1), Tyr
0.84 (1). Content of the peptide in the lyophilized mate-
rial was 94%, and purity of the peptide was 97.0%
according to HPLC.
5. Misicka, A. and Hruby, V.J., Pol. J. Chem., 1994, vol. 68,
pp. 893–899.
6. Kudryavtseva, E.V., Sidorova, M.V., Ovchinnikov, M.V.,
Bespalova, Zh.D., and Bushuev, V.N., J. Peptide Res.,
1997, vol. 49, pp. 52–58.
7. Kudryavtseva, E.V., Sidorova, M.V., Molokoedov, A.S.,
Ovchinnikov, M.V., and Bespalova, Zh.D., Bioorg.
Khim., 1999, vol. 25, pp. 107–116.
8. Kudryavtseva, E.V., Sidorova, M.V., Ovchinnikov, M.V.,
Bespalova, Zh.D., Bushuev, V.N., and Evstigneeva, R.P.,
Bioorg. Khim., 1996, vol. 22, pp. 370–375.
9. Papsuevich, O.S., Chipens, G.I., and Mikhailova, S.V.,
Neirogipofizarnye gormony (Neurohypophysial Hor-
mones), Riga: Zinatne, 1986.
10. Sidorova, M.V., Kudryavtseva, E.V., Molokoedov, A.S.,
Ovchinnikov, M.V., and Bespalova, Zh.D., Bioorg.
Khim., 1995, vol. 21, pp. 675–683.
3
The octreotide content in various reaction mixtures was from 70
to 75%.
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 30 No. 2 2004

110
SIDOROVA et al.
11. Sidorova, M.V., Kudryavtseva, E.V., Antopol’skii, M.L., 17. Kumagaye, S.-I., Kuroda, H., Nakajima, K.,
Pal’keeva, M.E., Ovchinnikov, M.V., and Bespa-
lova, Zh.D., Bioorg. Khim., 1996, vol. 22, pp. 273–279.
Watanabe, T.X., Kimura, T., Masaki, T., and Sakakibara, S.,
Int. J. Peptide Protein Res., 1988, vol. 32, pp. 519–526.
12. Eichler, J. and Hougthen, R.A., Protein Peptide Lett.,
18. Bauer, W. and Pless, J., Eur. Pat. Appl. 29 579 (1981);
US Patent 4395403 (1983).
1997, vol. 4, pp. 157–164.
13. Kudryavtseva, E.V., Sidorova, M.V., Ovchinnikov, M.V.,
and Bespalova, Zh.D., J. Peptide Sci., 2000, vol. 6,
pp. 208–216.
14. King, D.S., Fields, C.G., and Fields, G.B., Int. J. Peptide
Protein Res., 1990, vol. 36, pp. 255–266.
19. Ovchinnikov, M.V., Bespalova, Zh.D., Molokoedov, A.S.,
Titov, M.V., Revenko, I.V., Vinogradov, V.A.,
Korobov, N.V., and Zhukovskii, S.V., Bioorg. Khim.,
1988, vol. 14, pp. 777–782.
15. Hope, D.V., Murti, V.V.S., and du Vigneaud, V., J. Biol.
Chem., 1962, vol. 237, pp. 1563–1566.
16. McCurdy, S.N., User Bulletin Peptide Synthesizer
Applied Biosystems, 1988, no. 25.
20. Ovchinnikov, M.V., Bespalova, Z.D., Molokoedov, A.S.,
Titov, M.I., Revenko, I.V., Vinogradov, V.A.,
Korobov, N.V., and Zhukovskii, S.V., Collect. Czech.
Chem. Commun., 1989, vol. 54, pp. 796–802.
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 30 No. 2 2004