Constitutional isomers of Reactive Blue 2 - Selective P2Y-receptor antagonists?

Article abstract of DOI:10.1016/S0223-5234(02)01449-6

The anthraquinone derivative Reactive Blue 2 (RB 2) is one of the most widely used P2-receptor antagonists, still claimed to be P2Y-selective. RB 2 is defined as a mixture of two constitutional isomers and commercially available in different identity and purity. A sample of RB 2, offered for sale by RBI, purchased from Biotrend, Koeln, Germany, was chromatographically purified and identified by ¹H- and ¹³C-NMR studies as a 35:65 mixture of the terminal ring F meta and para constitutional isomers. The two constitutional isomers of RB 2 were synthesised and tested alongside with the ortho isomer Cibacron Blue 3GA (CB 3GA) on contractions of the rat vas deferens (RVD) elicited by α,β-methylene ATP (α,β-MeATP), mediated by P2X₁-receptors, and relaxations of the carbachol-precontracted guinea pig taenia coli elicited by adenosine 5′-O-(2-thiophosphate) (ADPβS), mediated by P2Y₁-like-receptors. All compounds inhibited the α,β-MeATP induced contraction of the RVD and the ADPβS induced relaxation of the carbachol precontracted guinea-pig taenia coli. The IC₅₀-values at P2X₁-R were 9.1 μM for CB 3GA, 28.4 μM for RB 2, 19.7 μM for RB 2 meta, and 35.5 μM for RB 2 para. The IC₅₀-values at P2Y₁-like-R were 17.4 μM for CB 3GA, 7.7 μM for RB 2, 12.0 μM for RB 2 meta, and 2.6 μM for RB 2 para. The results clearly show that neither RB 2 as a mixture nor the pure ortho and meta isomer are P2Y₁-like- versus P2X₁-selective antagonists. In contrast the pure para-isomer of RB 2 is a moderately P2Y₁-like- versus P2X₁-selective antagonist.

Full text of DOI:10.1016/S0223-5234(02)01449-6

European Journal of Medicinal Chemistry 38 (2003) 303ꢀ312
/
www.elsevier.com/locate/ejmech
Preliminary communication
Constitutional isomers of Reactive Blue 2 ꢀ
/
selective P2Y-receptor
antagonists?
Markus Glanzel a, Ralph Bultmann b, Klaus Starke b, August W. Frahm a,*
¨
¨
a Chair of Pharmaceutical Chemistry, Albertstr. 25, 79104 Freiburg i.Br., Germany
b Institute of Experimental and Clinical Pharmacology and Toxicology, Albertstr. 25, 79104 Freiburg i.Br., Germany
Received 27 August 2002; received in revised form 2 December 2002; accepted 5 December 2002
Dedicated to Professor Dr. Ernst Mutschler on the occasion of the 70th anniversary of his birthday.
Abstract
The anthraquinone derivative Reactive Blue 2 (RB 2) is one of the most widely used P2-receptor antagonists, still claimed to be
P2Y-selective. RB 2 is defined as a mixture of two constitutional isomers and commercially available in different identity and purity.
A sample of RB 2, offered for sale by RBI, purchased from Biotrend, Koln, Germany, was chromatographically purified and
¨
1
identified by H- and 13C-NMR studies as a 35:65 mixture of the terminal ring F meta and para constitutional isomers. The two
constitutional isomers of RB 2 were synthesised and tested alongside with the ortho isomer Cibacron Blue 3GA (CB 3GA) on
contractions of the rat vas deferens (RVD) elicited by a,b-methylene ATP (a,b-MeATP), mediated by P2X1-receptors, and
relaxations of the carbachol-precontracted guinea pig taenia coli elicited by adenosine 5?-O-(2-thiophosphate) (ADPbS), mediated
by P2Y1-like-receptors. All compounds inhibited the a,b-MeATP induced contraction of the RVD and the ADPbS induced
relaxation of the carbachol precontracted guinea-pig taenia coli. The IC50-values at P2X1-R were 9.1 mM for CB 3GA, 28.4 mM for
RB 2, 19.7 mM for RB 2 meta, and 35.5 mM for RB 2 para. The IC50-values at P2Y1-like-R were 17.4 mM for CB 3GA, 7.7 mM for
RB 2, 12.0 mM for RB 2 meta, and 2.6 mM for RB 2 para. The results clearly show that neither RB 2 as a mixture nor the pure ortho
and meta isomer are P2Y1-like- versus P2X1-selective antagonists. In contrast the pure para-isomer of RB 2 is a moderately P2Y1-
like- versus P2X1-selective antagonist.
´
# 2003 Editions scientifiques et me´dicales Elsevier SAS. All rights reserved.
Keywords: Reactive Blue 2; Cibacron Blue 3GA; P2-receptor antagonist; Rat vas deferens; Guinea-pig taenia coli; P2X1-receptor; P2Y1-like-receptor
1. Introduction
in experimental pharmacology but also as new targets in
drug research [1ꢀ4]. Therefore, there is great demand
/
P2-Receptors (P2-Rs) with the ligand gated ion
channel P2X-receptor- and the G-protein coupled
P2Y-receptor-classes are of growing interest not only
for selective and potent P2-receptor antagonists. Reac-
tive Blue 2 (RB 2) [5,6] is one of the most widely used
P2-receptor antagonists, still claimed to be P2Y-selective
[7ꢀ
indicating that it does not discriminate between P2X-R
and P2Y-R subtypes [10ꢀ18]. Ever since the commer-
cially available dye RB 2 has been used as a tool in
purinoceptor research, there was some doubt about its
/
9], although there is now a number of results
Abbreviations: C.I., colour index; RB 2, Reactive Blue 2; CB 3GA,
Cibacron Blue 3GA; a,b-MeATP, a,b-methylene adenosine 5?-
triphosphate; ADPbS, adenosine 5?-O-(2-thiophosphate); RVD, rat
vas deferens; GPTC, guinea-pig taenia coli; 2D-NMR, two
dimensional nuclear magnetic resonance; COSY, Correlation
Spectroscopy; HETCOR, heteronuclear correlation; NOESY,
Nuclear Overhauser Enhancement Spectroscopy; CC, column
chromatography; PIRPHPLC, paired ion reversed phase high
performance liquid chromatography.
/
identity and purity [18ꢀ23], both of which are prerequi-
/
sites for reliable receptor characterisation and subdiffer-
entiation.
RB 2 belongs to the anthraquinoneꢀchlorotriazinyl
/
reactive dyes, which formerly were used in textile
industry and later became important as ligands in
affinity chromatography for purification of enzymes
* Corresponding author.
E-mail address: august.wilhelm.frahm@pharmazie.uni-freiburg.de
(A.W. Frahm).
´
0223-5234/03/$ - see front matter # 2003 Editions scientifiques et me´dicales Elsevier SAS. All rights reserved.
doi:10.1016/S0223-5234(02)01449-6

304
M. Glanzel et al. / European Journal of Medicinal Chemistry 38 (2003) 303ꢀ
/312
¨
and biopolymers such as interferons and albumins
[24,25]. These dyes contain a unique combination of
ionic and hydrophobic sites. Yet, as commercial pre-
parations they are of highly heterogeneous nature and
varied in their actual chemical composition of the major
components not only from one manufacturer to the
conclusions made in the past concerning the effect of
ligand structures on P2-receptors, since small changes in
the structure may cause large differences in the degree of
bioselective interaction.
In 1988 Burton et al. synthesised all three terminal
ring F isomers of the anthraquinone dye RB 2, again
characterised just by TLC and PIRPHPLC. They also
analysed a number of commercial dye preparations and
in contrast to Hanggi and Carr found them to be
composed primarily of the ortho sulphoanilino isomer
[27].
Burton et al. claimed that Hanggi and Carr seemingly
made a false identification of the dichlorotriazinyl
analogue of Cibacron blue as the major constituent of
those investigated commercial dye preparations [26,27]
since they did not investigate the possibility of the third
constitutional isomer with the ortho sulphonate group.
Additional confusion stems from unfortunate errors
in the assignment of generic names, which virtually
destroyed the utility of the Colour Index (C.I.) system
with respect to these dyes.
other but also from batch to batch [26ꢀ29]. From the
/
1970s these dyes have also been applied as P2-receptor
research tools, since they display a certain degree of
antagonistic properties. But reliable receptor research is
only warranted with ligands of structural homogeneity
and unequivocal designation.
The structure of RB 2 (see Fig. 1) was claimed to be a
ring F meta/para sulphonate mixture [30]. This structure
has been assigned the dye constitution Index number
61 211 [31] and is found under the CAS Registry No.:
12236-82-7. For a time the so defined mixture has also
been offered for sale under the name of Cibacron Blue
F3GA. But Hanggi and Carr found out that of three
commercial preparations of Cibacron Blue F3GA only
one (Pierce) contained the two constitutional isomers
and the content did not exceed 10% of the total mixture.
Instead they identified the dichlorotriazinyl analogue of
Cibacron blue as the major constituent of all prepara-
tions [26]. Their identification was done by stepwise
synthesis of the two pure meta and para isomers and
comparative thin layer chromatography (TLC) and
paired-ion RP-HPLC. The authors did not provide a
complete NMR-spectroscopic characterisation, because
of the difficulty in exactly interpreting details of NMR
data and of variations from run to run, since strong
concentration and pH dependent self association phe-
nomena were observed [32].
The extent of the confusion may be illustrated by
generic dye names of the C.I. (3rd Ed. C.I.) [31] for ‘RB
2’, classified as isomeric meta- and para-mixture.
Manufacturers dye
names
Generic dye names (3rd Ed.
C.I.)
Procion Blue HB (ICI) $
Cibacron Blue 3G
(CibaꢀGeigy)
/
Procion Blue H-B
Cibacron Blue F-3GA
C.I. RB 2
$
/
/
$
/
(C.I. Constitution number
61211)
Nevertheless, Cibacron dyes were predominantly
composed of the ortho-isomer. Moreover, RB 2 was
also sold under the names of Cibacron Blue F3GA and
Procion Blue HBS, respectively. All these dyes were
withdrawn with the C.I. 3rd Ed. 2nd Revision.
They discovered also significant errors in the struc-
tural assignment for currently available preparations of
some of these dyes. These errors call into question some
In follow up, a new Cibacron Blue 3GA (CB 3GA)
was marketed by BDH, Pierce, and Polysciences, said to
contain the ortho-isomer different from C.I. RB 2.
Sigma re-evaluated their so-called RB 2 material as the
F-ring ortho isomer, and renamed it as CB 3GA. The
ring F meta and para isomeric mixture was then sold as
Basilen Blue E-3G, which was finally replaced by the
RBI product RB 2 after the Sigma merger with RBI.
Burton et al. advocate either the analysis of commercial
preparations of sulphonated dyes prior to biological use
or the structure selective synthesis, since unpurified and
ill-defined dye samples are hazards of reliable research
work including receptor-ligand-binding work [27]. Care-
ful analysis of the purity and structural composition of
RB 2 preparations is mandatory but is not guaranteed
for most of the published P2-receptor work. Therefore,
we have investigated thoroughly the story of the RB 2
chemistry from the beginning.
Fig. 1. RB 2, mixture of the terminal ring F meta and para
constitutional isomers.

M. Glanzel et al. / European Journal of Medicinal Chemistry 38 (2003) 303ꢀ
/312
305
¨
2. Chemistry
comparison. Up to now a complete NMR-spectroscopic
characterisation was lacking in the literature, due to
severe difficulties with the interpretation of the NMR
data [32]. Therefore, we purchased and synthesised
numerous compounds with comparable structures to
RB 2 fragments. Based on the analysed NMR shift
values for these compounds we have developed a system
of increments which helped us to identify unequivocally
the isomers of RB 2 and CB 3GA. 2D-NMR-techniques
like 1H,1H-COSY, NOESY, 1H-13C-HETCOR, long-
The commercially available dye RB 2 which was used
in this study was offered by RBI and distributed by
Biotrend (Koln, Germany) in a purity of 97% with
¨
respect to the mixture of two constitutional isomers. In
comparison, RB 2 products of other suppliers provide
merely a dye content of ca. 60%, but have also been used
in literature work so far. The same applies to CB 3GA
which we purchased from Sigma with a dye content of
ca. 65%. In this study CB 3GA was purified by column
chromatography from organic impurities as well as from
inorganic salts such as NaCl and sodium phosphate.
The remaining organic matter was then investigated
NMR-spectroscopically as was the purchased RB 2
mixture. RB 2 was found to consist of two constituents
in a ratio of 35:65 by intensive 1H- and 13C-NMR
investigations. The isomerism was exclusively found in
ring F with a meta- and para-sulfonate group, respec-
tively. These constitutional isomers contained in RB 2
are different from the dye CB 3GA, which is the F-ring-
ortho-sulfonate isomer of the two RB 2 constituents.
The NMR-data of all three anthraquinone derived
constitutional isomers are summarised in Table 1 for
1
range-1H-13C-, H-15N-HETCOR and J-resolved spec-
tra were necessary to resolve all the problems in signal
identification. Later on we were able to verify the
correlation between our NMR data and the correspond-
ing, by now isomerically pure isomers. NMR data
obtained from the pure isomers (Table 1) show that
chemical shifts for the anthraquinone core and the side
chain rings D and E of the two RB 2 isomers differ by
less than 9
/
0.2 ppm in 13C- and 90.04 ppm in 1H-
/
NMR-spectra. The same applies for differences between
the CB 3GA ortho isomer and the other two isomers,
although the differences come up to 9
0.7 ppm in 13C-
/
1
0.1 ppm in H-NMR-spectra. These significant
chemical shift variations are restricted to terminal ring F
and 9
/
Table 1
13C- and 1H-NMR chemical shifts d (ppm) of CB 3GA and the constitutional isomers RB 2 meta and RB 2 para
NMR-signal number
CB 3GA
13C
RB 2 meta
13C
RB 2 para
13C
1H
1H
1H
1
144.2
142.5
122.7
140.7
111.6
125.7
132.5
132.5
125.7
134.0
181.8
109.3
182.4
133.4
134.0
121.7
137.4
130.7
122.1
123.5
163.4
168.3
163.2
134.1
136.2
126.8
122.6
129.4
121.1
ꢀ
ꢀ/
8.06
11.99
/
144.2
142.8
122.5
141.1
111.1
125.8
132.4
132.8
125.8
134.0
181.6
109.0
182.2
133.5
133.3
121.9
137.1
131.2
121.4
124.1
163.9
168.2
162.9
137.4
118.4
148.7
121.2
127.7
120.9
ꢀ
ꢀ/
7.98
12.05
/
144.2
142.7
122.5
141.0
111.2
125.8
132.5
132.9
125.8
134.0
181.7
109.1
182.3
133.5
133.5
122.0
137.2
131.2
121.5
123.9
163.7
168.3
162.9
138.2
119.7
125.8
143.4
125.8
119.7
ꢀ
ꢀ/
7.97
12.04
/
2
3
4
4a
5
ꢀ/
ꢀ/
ꢀ/
8.26
7.83
7.83
8.26
8.29
7.83
7.83
8.29
8.29
7.85
7.85
8.29
6
7
8
8a
9
ꢀ/
ꢀ/
ꢀ/
ꢀ/
ꢀ/
ꢀ/
ꢀ/
ꢀ/
ꢀ/
ꢀ/
ꢀ/
ꢀ/
ꢀ/
ꢀ/
ꢀ/
ꢀ/
ꢀ/
ꢀ/
9a
10
10a
1?
2?
3?
4?
5?
6?
2ƒ
4ƒ
6ƒ
1§
2§
3§
4§
5§
6§
7.66
7.57
7.60
ꢀ/
ꢀ/
ꢀ/
10.29
8.38
7.31
10.32
8.5
7.33
10.35
8.5
7.30
ꢀ/
ꢀ/
ꢀ/
ꢀ/
ꢀ/
ꢀ/
10.34
10.26
10.30
ꢀ/
ꢀ/
7.77
7.09
7.50
8.30
ꢀ/
7.83
ꢀ/
7.60
7.60
7.60
7.60
ꢀ/
7.75
7.36
7.36
ꢀ/

306
M. Glanzel et al. / European Journal of Medicinal Chemistry 38 (2003) 303ꢀ
/312
¨
nuclei. This made it difficult to quantify the ratio of the
two isomers RB 2 meta and RB 2 para contained in the
RB 2 mixture. In the 1H-NMR spectra, the NH-4?-signal
is the only one that features a sufficient shift difference
between the two isomers and does not overlap with
other signals. In the 13C-spectra, the C-1§ signals of the
two isomers are well separated. In addition, the signals
of C-2§ can be used, considering that the 13C-signal in
the para-isomer represents the two carbons C-2§ and C-
6§. Integration of the NH-4? signal resulted in a ratio of
62:38, integration of the C-1§ and the C-2§ signal in a
ratio of 66:34 in each case for the benefit of the para
isomer.
RB 2 as the purified isomeric mixture unfortunately
could not be separated by any type of CC into the F-ring
meta and para isomers. Therefore, we synthesised
structure-selectively the two constituents contained in
RB 2 as shown in Fig. 2.
Fig. 2. Synthesis of RB 2 isomers.

M. Glanzel et al. / European Journal of Medicinal Chemistry 38 (2003) 303ꢀ
/312
307
¨
The constitutional isomers of RB 2 were synthesised
according to procedures of Wojtkiewicz and Kraska [34]
and of Hanggi and Carr [26]. In the first step, 1-amino-
4-bromo-9,10-dioxo-9,10-dihydroanthracene-2-sulfonic
acid (bromaminic acid) (764 mg, 2 mmol) and 2,5-
diaminobenzenesulfonic acid (760 mg, 3 mmol) were
added to a stirred solution of sodium carbonate (500
mg) and sodium sulphite (400 mg) in water (50 mL) at
room temperature (r.t.). Copper(I) chloride (50 mg) was
added and the mixture was stirred for ca. 8 h, monitor-
ing the formation of 1-amino-4-(4-amino-3-sulfopheny-
lamino)-9,10-dioxo-9,10-dihydroanthracene-2-sulfonic
acid (1) by reversed phase TLC using a mobile phase of
40ꢀ
tain a pH of 5ꢀ
analytical reversed phase TLC using again a mobile
phase of methanolꢀwater (2:3). Finally the solvent was
evaporated and the blue residue was purified by flash
column chromatography on RP-18 silica gel using a
/
60 8C. Sodium carbonate was used again to main-
/
7. Reaction progress was monitored by
/
methanolꢀwater (1:3) eluent to give the constitutional
/
isomers RB 2 meta (3) and RB 2 para (4), respectively.
Functional tests were carried out with the RB 2
mixture, the two pure constitutional isomers of RB 2
and with the pure ortho-isomer CB 3GA in order to
determine the P2Y1-like- versus P2X1-selectivity.
methanolꢀ
the solution was filtered and the residue was washed
with water (3ꢁ20 mL) and methanol (3ꢁ20 mL). The
/
water (2:3). After completion of the reaction
3. Pharmacology
/
/
combined filtrates were evaporated. The residue was
purified by flash column chromatography on RP-18
We have previously reported a structureꢀactivity-
/
study [33] in which the purified dyes RB 2, CB 3GA and
other structurally related compounds were studied on
contractions of the rat vas deferens (RVD) elicited by
a,b-methylene ATP (a,b-MeATP; mediated by native
P2X1-receptors which are presumably identical to the
recombinant P2X1-receptor subtype) and relaxations of
the carbachol-precontracted guinea-pig taenia coli
(GPTC) elicited by adenosine 5?-O-(2-thiodiphosphate)
(ADPbS; mediated by native P2Y1-like-receptors which
are so far not cloned but pharmacologically similar to
the recombinant P2Y1-receptor subtype). In this former
silica gel using a methanolꢀwater (1:3) eluent. Accord-
/
ing to our observations a nitrogen atmosphere did not
improve the reaction. Higher reaction temperatures as
given in the procedure of Pearson et al. [35] did reduce
the reaction time, but also the yield of the product on
the expenses of a higher amount of side products.
In the next step a stoichiometric amount of cyanuric
chloride in an ice-cooled aqueous solution of acetone
was added to an aqueous solution of 1 at 0ꢀ
resulting solution was stirred at 0ꢀ5 8C for 1 h while 2
M sodium carbonate solution was added to keep the pH
in the range 5ꢀ7. Then, in the final reaction step an
/
5 8C. The
/
study, CB 3GA (Kdꢂ
potent than RB 2 (Kdꢂ
receptor, whereas CB 3GA (Kdꢂ
(Kdꢂ
/
1.6 mM) was considerably more
9.9ꢀ11.4 mM) at the P2X1-
2.9 mM) and RB 2
/
/
/
excess of metanilic and sulphanilic acid, respectively, in
an aqueous solution was added to the solution of 1-
amino-4-[4-(4,6-dichloro-[1,3,5]triazin-2-ylamino)-3-sul-
fophenylamino]-9,10-dioxo-9,10-dihydroanthracene-2-
sulfonic acid (2) and the temperature was increased to
/
/
3.4 mM) were almost equipotent at the P2Y1-like-
receptor. It also became clear that RB 2 and CB 3GA
display little P2Y1-like- versus P2X1-selectivity (selectiv-
ity ratios were 2.9ꢀ3.4 and 0.6, respectively). Never-
/
Fig. 3. Effects of antagonists on (a) contraction of RVD elicited by 3.2 mM a,b-MeATP (CB 3GA m; RB 2 mixture j; RB 2 meta '; RB 2 para %;
means9S.E.M. from nꢂ5ꢀ11 experiments) and on (b) relaxation of carbachol-precontracted GPTC elicited by 0.1 mM ADPbS (CB 3GA k; RB 2
mixture I; RB 2 meta ^; RB 2 para \; means9S.E.M. from nꢂ4ꢀ7 experiments).
/
/
/
/
/
/

308
M. Glanzel et al. / European Journal of Medicinal Chemistry 38 (2003) 303ꢀ312
/
¨
Fig. 4. Selectivity profiles: effects of (a) CB 3GA, (b) RB 2 mixture, (c) RB 2 meta and (d) RB 2 para on P2X1-receptors of RVD (filled symbols) vs.
effects on P2Y1-like-receptors of GPTC (blank symbols). Means9S.E.M. from nꢂ4ꢀ11 experiments.
/
/
/
theless unequivocal and reliable results are only gained
from structurally pure entities.
The RB 2 mixture and the pure constitutional isomers
inhibited the a,b-MeATP induced contraction in RVD,
mediated by P2X1-receptors, as well as the ADPbS
induced relaxation in GPTC, mediated by P2Y1-like-
receptors.
In the present study, purified CB 3GA, the RB 2
mixture and the constitutional isomers of RB 2, RB 2
meta and para, were investigated to answer the question
whether they differ in their inhibitory potencies and
selectivities. So we decided to determine the IC50 values
of all pure constitutional isomers in addition to the RB 2
mixture. Contractions of the RVD were elicited every 60
min by a 3.2 mM concentration of a,b-MeATP. Relaxa-
tions of the carbachol-precontracted GPTC were elicited
every 60 min by a 0.1 mM concentration of ADPbS.
Both agonist concentrations are ca. the EC50 values
4. Results and discussion
Commercially available CB 3GA and the RB 2
mixture were chromatographically purified and identi-
1
fied by H- and 13C-NMR studies. The two constitu-
tional isomers of RB 2 were synthesised. All compounds
were investigated in functional tests.
The results are summarised in Table 2. Comparing the
under the present conditions [33,36ꢀ39]. The antago-
/
nists were added in vivo in increasing concentrations
ranging from 1 to 100 mM. IC50 values were calculated
for each experiment by interpolation from the two
nearest data points. Means were tested for a significant
pure constitutional isomers, CB 3GA (IC50ꢂ
the most potent compound at the P2X1-receptor. RB 2
meta (IC50ꢂ19.7 mM) is more potent than RB 2 para
(IC50ꢂ35.5 mM). At the P2Y1-like-receptor, RB 2 meta
(IC50ꢂ12.0 mM) and CB 3GA (IC50ꢂ17.4 mM) are
nearly equipotent, whereas RB 2 para (IC50ꢂ2.6 mM)
/
9.1 mM) is
/
difference by the MannꢀWhitney test. Concentration
/
/
response curves of the compounds are shown in Figs. 3
/
/
and 4.
/

M. Glanzel et al. / European Journal of Medicinal Chemistry 38 (2003) 303ꢀ
/312
309
¨
Table 2
Antagonist IC50 values at P2X1-receptors (RVD), at P2Y1-like-
receptors (GPTC), and selectivity ratios P2Y1-like versus P2X1
Germany), 2,5-diaminobenzenesulphonic acid, cyanuric
chloride (2,4,6-trichloro-s-triazine), m- and p-amino-
benzenesulphonic acids were purchased from Aldrich
(Taufkirchen, Germany), CB 3GA from Sigma and RB
Compound IC50 (mM)
Selectivity, P2Y1-like vs. P2X1
2 from RBI (Biotrend, Koln, Germany).
¨
P2X1
P2Y1-like
CB 3GA a
RB 2 b
9.19
28.492.6
1.1 12.09
RB 2 para 35.595.4 2.69
/
0.8 17.49
/
4.7
2.9
4.3
0.4
0.5
3.7
6.1.2. Analytical studies
1H- and 13C-1D- and -2D-NMR spectra were re-
corded on a Varian Unity 300 spectrometer operating at
299.5 and 75.4 MHz, respectively. Chemical shift values
d (in ppm) are referenced to the signal of DMSO in
/
7.79
/
RB 2 meta 19.79
/
/
1.6
13.7
/
/
IC50 values are from experiments of Figs. 3 and 4 (nꢂ
Differences between IC50 values at P2X1- and P2Y1-like-receptors are
/
4ꢀ/11).
DMSO-d6, with dTMS
dTMS dDMSO
39.7 for 13C-NMR. The chemical shifts
of proton multiplets were determined via the corre-
ꢂ
/
dDMSOꢃ
2.49 for 1H-NMR and
/
significant for RB 2 and RB 2 para (P B
/0.05).
ꢂ
/
ꢃ
/
a
Ortho isomer.
Mixture (35:65) of meta and para isomers.
b
1
1
sponding H-13C-HETCOR spectra. H-NMR data are
listed in the following order: multiplicity (s, singlet; d,
dublett; t, triplet; m, multiplet), number of nuclei,
assignment. Flash chromatography was carried out on
reveals stronger inhibition. As a result, the para-isomer
of RB 2 displays moderate P2Y1-like- versus P2X1-
selectivity of 13.7, whereas the other constitutional
isomers are non-selective. The commercially available
RB 2 mixture with IC50s of 28.4 mM (P2X1) and 7.7 mM
(P2Y1-like), respectively, shows a P2Y1-like- versus
P2X1-selectivity ratio of 3.7 only.
Silica gel RP C18 (32ꢀ63 mm) (ICN Biomedicals,
/
Eschwege, Germany). TLC was performed on Silica
gel aluminium sheets RP-18 F254s (E. Merck, Darm-
stadt, Germany).
The results demonstrate that the meta RB 2 isomer
like the ortho isomer CB 3GA is a non selective P2-
6.1.2.1. 1-Amino-4-(4-amino-3-sulfophenylamino)-9,10-
dioxo-9,10-dihydroanthracene-2-sulfonic acid (1). To a
stirred solution of 500 mg sodium carbonate and 400 mg
sodium sulphite in 50 mL water were added 764 mg (2
receptor antagonist whereas the para-isomer *
/
con-
tained in RB 2 with 65% only * is a moderate P2Y1-
/
like-selective antagonist.
mmol)
1-amino-4-bromo-9,10-dioxo-9,10-dihydroan-
thracene-2-sulfonic acid (bromaminic acid) and 760 mg
(4 mmol) 2,5-diaminobenzenesulfonic acid. Then 50 mg
copper(I) chloride were added and the mixture was
stirred for ca. 8 h, monitoring the formation of product
by reversed phase TLC using a mobile phase of
5. Conclusions
It is clearly demonstrated that RB 2 as a meta/para
mixture does not discriminate the P2X1-receptor of
RVD and the P2Y1-like-receptor of GPTC [33] *
contrast to most of published work and to commercial
advertisement in receptor ligand catalogues [40] *
whereas the pure para isomer of RB 2 is a comparatively
potent and selective antagonist at the P2Y1-like-receptor
in comparison with the native P2X1-receptor.
The present results prompted us to further investigate
other structural modifications of RB 2 and CB 3GA
with respect to potency, selectivity and reversibility at
the P2Y1-like-receptor. Structure-selectivity-relation-
ships are under exploration and will be published in
the near future.
methanolꢀ
in about 8 h the mixture was filtered and the residue was
washed with water (3ꢁ20 mL) and methanol (3ꢁ20
/water (2:3). After completion of the reaction
/
in
/
/
/,
mL). The combined filtrates were evaporated. The blue
residue was separated from impurities by flash column
chromatography on reversed phase silica gel using a
methanolꢀwater (1:3) eluent. The combined fractions
/
were filtered and evaporated. The blue residue was dried
under vacuum.
1
C20H13N3O8S2Na2 [533.44] H-NMR (DMSO-d6) d
5.65 (s, 2H, NH2-4?), 6.70 (d, 1H, H-5?), 6.95 (dd, 1H, H-
6?), 7.34 (d, 1H, H-2?), 7.80 (s, 1H, H-3), 7.81 (m, 2H, H-
6, H-7), 8.28 (m, 2H, H-5, H-8), 9-11 (br, 2H, NH2-1),
12.05 (s, 1H, NH-4); 13C-NMR (DMSO-d6) d 108.69
(C-9a), 109.47 (C-4a), 116.13 (C-5?), 122.56 (C-3), 124.10
(C-2?), 125.51 (C-1?), 125.67 (C-5), 125.76 (C-8), 126.57
(C-6?), 130.97 (C-3?), 132.29 (C-6), 132.42 (C-7), 133.73
(C-10a), 134.02 (C-8a), 143.04 (C-4), 143.19 (C-4?),
143.58 (C-2), 143.92 (C-1), 181.11 (C-10), 181.34 (C-9).
6. Experimental protocols
6.1. Chemistry
6.1.1. Chemicals
Bromaminic acid (1-amino-4-bromoanthraquinone-2-
sulphonic acid) was supplied by Bayer AG (Leverkusen,
Germany) and purchased from Sigma (Taufkirchen,
6.1.2.2. 1-Amino-4-[4-(4,6-dichloro-[1,3,5]triazin-2-
ylamino)-3-sulfophenylamino]-9,10-dioxo-9,10-
dihydroanthracene-2-sulfonic acid (2). An ice-cooled

310
M. Glanzel et al. / European Journal of Medicinal Chemistry 38 (2003) 303ꢀ
/312
¨
solution of 35 mg (0.19 mmol) cyanuric chloride in
water (10 mL) and acetone (10 mL) was added to a
148.74 (C-3§), 162.85 (C-2ƒ), 163.88 (C-6ƒ), 168.20 (C-
4ƒ), 181,62 (C-9), 182.15 (C-10).
solution of 100 mg (0.19 mmol) 1 in water (10 mL) at 0ꢀ
5 8C. The resulting solution was stirred at 0ꢀ5 8C for 1 h
while 2 M sodium carbonate solution was added
dropwise to keep the pH in the range 5ꢀ7. The reaction
was monitored for completeness by reversed phase TLC
using a mobile phase of methanolꢀwater (2:3). After
/
/
6.1.2.4. 1-Amino-4-{4-[4-chloro-6-(4-
sulfophenylamino)-[1,3,5]triazin-2-ylamino]-3-sulfo-
phenylamino}-9,10-dioxo-9,10-dihydroanthracene-2-
sulfonic acid (4) (RB 2 para). A solution of 165 mg
(0.95 mmol) of sulphanilic acid in water (10 mL) was
added to the stirred solution of 130 mg (0.19 mmol) 2 in
/
/
completion of the reaction the solvent was evaporated
and the blue residue was purified by flash column
chromatography on reversed phase silica gel using a
water (20 mL) and acetone (10 mL) at 0ꢀ
/
5 8C. Then the
temperature was increased to 40ꢀ60 8C and stirring was
/
methanolꢀwater (1:3) eluent. The combined fractions
/
continued for ca. 2 h. A 2 M sodium carbonate solution
was added dropwise to keep the pH in the range of pH
were filtered and evaporated. The blue residue was dried
under vacuum.
1
C23H12N6O8Cl2S2Na2 [681.39] H-NMR (DMSO-d6)
5ꢀ/7. Reaction progress was monitored by reversed
phase TLC using a mobile phase of methanolꢀ
/
water
(2:3). Finally the solvent was evaporated and the blue
d 7.27 (dd, 1H, H-6?), 7.56 (d, 1H, H-2?), 7.83 (m, 2H,
H-6, H-7), 7.98 (s, 1H, H-3), 8.00 (d, 1H, H-5?), 8.27 (m,
residue was purified by flash column chromatography
2H, H-5, H-8), 10.7 (br, 1H, NH-4?), 9ꢀ11 (br, 2H,
/
on reversed phase silica gel using a methanolꢀwater
/
NH2-1), 12.0 (br, 1H, NH-4); 13C-NMR (DMSO-d6) d
109.19 (C-9a), 111,54 (C-4a), 121.16 (C-2?), 122.61 (C-3),
123.25 (C-6?), 125.63 (C-5?), 125.85 (C-5, C-8), 129.44
(C-4?), 132.58 (C-6), 132.99 (C-7), 133.46 (C-10a), 134.03
(C-8a), 135.00 (C-1?), 139.50 (C-3?), 140.62 (C-4), 142.69
(C-2), 144.27 (C-1), 152.91 (C-2ƒ), 154.27 (C-4ƒ, C-6ƒ),
181.77 (C-9), 182.46 (C-10).
(1:3) eluent. The combined fractions were filtered and
evaporated. The blue residue was dried under vacuum.
1
C29H17N7O11S3Na3Cl [840.10] H-NMR (DMSO-d6)
d 7.30 (dd, 1H, H-6?), 7.60 (m, 5H, H-2?, H-2§, H-3§, H-
5§, H-6§), 7.85 (m, 2H, H-6, H-7), 7.97 (s, 1H, H-3),
8.29 (m, 2H, H-5, H-8), 7-11 (br, 2H, NH2-1), 10.30 (s,
1H, NH-6ƒ), 10.35 (s, 1H, NH-4?), 12.04 (s, 1H, NH-4);
13C-NMR (DMSO-d6) d 109.11 (C-9a), 111.21 (C-4a),
119.68 (C-2§, C-6§), 121.50 (C-5?), 121.98 (C-2?), 122.47
(C-3), 123.92 (C-6?), 125.80 (C-3§, C-5§), 125.83 (C-5,
C-8), 131.19 (C-4?), 132.49 (C-6), 132.86 (C-7), 133.47
(C-1?,C-10a), 134.00 (C-8a), 137.22 (C-3?), 138.24 (C-
1§), 140.99 (C-4), 142.69 (C-2), 143.39 (C-4§), 144.17
(C-1), 162.89 (C-2ƒ), 163.73 (C-6ƒ), 168.29 (C-4ƒ), 181.69
(C-9), 182.25 (C-10).
6.1.2.3. 1-Amino-4-{4-[4-chloro-6-(3-
sulfophenylamino)-[1,3,5]triazin-2-ylamino]-3-sulfo-
phenylamino}-9,10-dioxo-9,10-dihydroanthracene-2-
sulfonic acid (3) (RB 2 meta). A solution of 165 mg
(0.95 mmol) of metanilic acid in water (10 mL) was
added to the stirred solution of 130 mg (0.19 mmol) 2 in
water (20 mL) and acetone (10 mL) at 0ꢀ
/
5 8C. Then the
temperature was increased to 40ꢀ60 8C and stirring was
/
continued for ca. 2 h. A 2 M sodium carbonate solution
was added dropwise to keep the pH in the range of pH
6.1.2.5. 1-Amino-4-{4-[4-chloro-6-(2-
sulfophenylamino)-[1,3,5]triazin-2-ylamino]-3-sulfo-
phenylamino}-9,10-dioxo-9,10-dihydroanthracene-2-
sulfonic acid (CB 3GA). The purchased CB 3GA was
purified by flash column chromatography on reversed
5ꢀ/7. Reaction progress was monitored by reversed
phase TLC using a mobile phase of methanolꢀ
/water
(2:3). Finally the solvent was evaporated and the blue
residue was purified by flash column chromatography
phase silica gel using a methanolꢀwater (1:3) eluent. The
/
on reversed phase silica gel using a methanolꢀ
/
water
combined fractions were filtered and evaporated. The
blue residue was dried under vacuum.
(1:3) eluent. The combined fractions were filtered and
evaporated. The blue residue was dried under vacuum.
1
C29H17N7O11S3Na3Cl [840.10] H-NMR (DMSO-d6)
1
C29H17N7O11S3Na3Cl [840.10] H-NMR (DMSO-d6)
d 7.09 (t, 1H, H-4§), 7.31 (dd, 1H, H-6?), 7.50 (t, 1H, H-
5§), 7.66 (d, 1H, H-2?), 7.77 (d, 1H, H-3§), 7.83 (m, 2H,
H-6, H-7), 8.06 (s, 1H, H-3), 8.26 (m, 2H, H-5, H-8),
8.30 (d, 1H, H-6§), 8.38 (d, 1H, H-5?), 7-11 (br, 2H,
NH2-1), 10.29 (s, 1H, NH-4?), 10.34 (s, 1H, NH-6ƒ),
11.99 (s, 1H, NH-4); 13C-NMR (DMSO-d6) d 109.30
(C-9a), 111.57 (C-4a), 121.08 (C-6§), 121.67 (C-2?),
122.05 (C-5?), 122.56 (C-4§), 122.71 (C-3), 123.52 (C-
6?), 125.71 (C-5, C-8), 126.80 (C-3§), 129.44 (C-5§),
130.73 (C-4?), 132.48 (C-6, C-7), 132.89 (C-6), 133.39 (C-
10a), 133.96 (C-1?, C-8a), 134.09 (C-1§), 136.17 (C-2§),
137.38 (C-3?), 140.65 (C-4), 142.45 (C-2), 144.17 (C-1),
d 7.33 (dd, 1H, H-6?), 7.36 (m, 1H, H-5§), 7.57 (d, 1H,
H-2?), 7.75 (br, 1H, H-4§), 7.83 (m, 3H, H-2§, H-6, H-
7), 7.98 (s, 1H, H-3), 8.29 (m, 2H, H-5, H-8), 8.5 (br, 1H,
H-5?), 7-11 (br, 2H, NH2-1), 10.26 (s, 1H, NH-6ƒ), 10.32
(s, 1H, NH-4?), 12.05 (s, 1H, NH-4); 13C-NMR (DMSO-
d6) d 109.04 (C-9a), 111.10 (C-4a), 118.37 (C-2§), 120.94
(C-6§), 121.15 (C-4§), 121.40 (C-5?), 121.85 (C-2?),
122.48 (C-3), 124.07 (C-6?), 125.77 (C-5, C-8), 127.68
(C-5§), 131.15 (C-4?), 132.43 (C-6), 132.79 (C-7), 133.30
(C-1?), 133.47 (C-10a), 134.00 (C-8a), 137.11 (C-3?),
137.35 (C-1§), 141.05 (C-4), 142.80 (C-2), 144.20 (C-1),

M. Glanzel et al. / European Journal of Medicinal Chemistry 38 (2003) 303ꢀ
/312
311
¨
163.16 (C-2ƒ), 163.35 (C-6ƒ), 168.32 (C-4ƒ), 181,79 (C-9),
182.42 (C-10).
for a significant difference by the Mannꢀ
/
Whitney test.
P B0.05 was taken as the limit of statistical significance.
/
6.2. Pharmacology
6.2.1. Methods
6.2.5. Materials
Adenosine 5?-O-(2-thiodiphosphate) trilithium (AD-
PbS), carbachol chloride and a,b-methylene adenosine
5?-triphosphate dilithium (a,b-MeATP) were purchased
from Sigma. All drugs were dissolved in distilled water.
Solutions of drugs were added to the organ bath in
aliquots not exceeding 100 mL.
Male Wistar rats (250ꢀ
/
350 g) or guinea pigs of either
sex (400ꢀ800 g) were decapitated, the vasa deferentia
/
(rat) or the ventral taenia coli (guinea-pig) removed and
cleaned of adherent tissue. The incubation medium
contained (mM): NaCl 118, KCl 4.8, CaCl2 2.5,
KH2PO4 0.9, NaHCO3 25, glucose 11, ascorbic acid
0.3, and disodium EDTA 0.03. It was gassed with 95%
O2/5% CO2 and kept at 37 8C.
Acknowledgements
We are grateful to the Deutsche Forschungsge-
meinschaft, the European Commission, the Fonds der
Chemischen Industrie and the grant ‘Interdisciplinary
Projects’ of the University of Freiburg for their support.
6.2.2. Contraction of rat vas deferens
Prostatic thirds of the vas deferens were suspended
vertically in a 5.9 mL organ bath. The lower end was
fixed and the upper end attached to an isometric force
transducer (K30, Hugo Sachs Elektronik, Hugstetten,
Germany). The initial tension applied was 9.8 mN
(Graphtec thermal pen recorder, Ettlingen, Germany),
but tissues subsequently relaxed to ca. 3 mN during a 60
min equilibration period. The medium was replaced
every 15 min. Contractions were elicited by a,b-MeATP.
a,b-MeATP was washed out immediately after contrac-
tions had peaked. a,b-MeATP was added every 60 min
at a concentration of 3.2 mM.
References
[1] V. Ralevic, G. Burnstock, Pharmacol. Rev. 60 (1998) 413ꢀ
/492.
[2] M. Williams, M.F. Jarvis, Biochem. Pharmacol. 59 (2000) 1173ꢀ
/
1185.
[3] G. Burnstock, M. Williams, J. Pharmacol. Exp. Ther. 295 (2000)
862ꢀ
[4] G. Burnstock, Clin. Med. 2 (2002) 45ꢀ
[5] D.I.B. Kerr, A. Krantis, Proc. Austr. Physiol. Pharmacol. Soc. 10
(1979) 156ꢀ157.
[6] L.K. Choo, J. Pharm. Pharmacol. 33 (1981) 248ꢀ
[7] Y. Fischer, C. Becker, C. Loken, J. Biol. Chem. 274 (1999) 755ꢀ
/869.
/
53.
/
6.2.3. Relaxation of guinea-pig taenia coli
/
250.
Strips of about 10 mm of the taenia coli were
suspended vertically in a 5.9 mL organ bath. The lower
end was fixed and the upper end attached to a K30
transducer. The resting tension was repeatedly adjusted
to 9.8 mN during an initial 60 min equilibration period.
The medium was replaced every 15 min. After equili-
bration, carbachol was added to the medium at 15 min
intervals for 2 min each. Initially, the maximal contrac-
tion of each strip was determined by a single addition of
carbachol (300 nM). During the following two to four
applications, the concentration of carbachol giving
about 80% of the maximum was determined. This
/
761.
[8] J.D. Troadec, S. Thirion, D. Petturiti, P. Poujeol, Eur. J. Physiol.
440 (2000) 770ꢀ
[9] T. Zenteno-Savin, I. Sada-Ovalle, G. Ceballos, R. Rubio, Eur. J.
Pharmacol. 410 (2000) 15ꢀ23.
[10] E.M. Silinsky, V. Gerzanich, J. Physiol. (London) 464 (1993)
197ꢀ212.
[11] G.P. Connolly, P.J. Harrison, Eur. J. Pharmacol. 259 (1994) 95ꢀ
/
777.
/
/
/
99.
[12] B.S. Khakh, P.P. Humphrey, A. Surprenant, J. Physiol. (London)
484 (1995) 385ꢀ
[13] D.J. Trezise, I. Kennedy, P.P.A. Humphrey, Br. J. Pharmacol.
112 (1994) 282ꢀ288.
[14] H. Suzuki, S. Kokubun, Br. J. Pharmacol. 112 (1994) 117ꢀ
[15] J.S. Wiley, R. Chen, G.P. Jamieson, Arch. Biochem. Biophys. 305
(1993) 54ꢀ60.
[16] A.J. Brake, M.J. Wagenbach, D. Julius, Nature (Washington,
DC) 371 (1994) 519ꢀ523.
[17] X. Bo, Y. Zhang, M. Nassar, G. Burnstock, R. Schoepfer, FEBS
/
395.
/
/122.
concentration (50ꢀ90 nM) was used for the remainder
/
of the experiment, the 15 min rhythm being kept
throughout. ADPbS was added during the plateau of
the carbachol response and washed out together with
the latter after the ensuing relaxation was maximal.
Relaxations were expressed as percentage of the respec-
tive carbachol contraction. ADPbS was added every 60
min at a concentration of 0.1 mM.
/
/
¨
Lett. 375 (1995) 129ꢀ
/
133.
[18] R. Bultmann, K. Starke, Naunyn-Schmiedeberg’s Arch. Pharma-
¨
col. 349 (1994) 74ꢀ
[19] M.K. McMillan, S.P. Soltoff, L.C. Cantley, R.A. Rudel, B.R.
Talamo, Br. J. Pharmacol. 108 (1993) 453ꢀ461.
[20] X. Bo, B. Fischer, M. Maillard, K.A. Jacobson, G. Burnstock, Br.
J. Pharmacol. 112 (1994) 1151ꢀ1159.
[21] G. Hansmann, R. Bultmann, F. Tuluc, K. Starke, Naunyn-
/80.
/
6.2.4. Statistics
IC50 values were calculated for each experiment by
interpolation from the two nearest data points. Antago-
/
¨
Schmiedeberg’s Arch. Pharmacol. 356 (1997) 641ꢀ
[22] W.W. Lin, D.M. Chuang, J. Pharmacol. Exp. Ther. 269 (1994)
926ꢀ931.
/
652.
nist IC50 values are expressed as the arithmetic mean9
/
S.E.M. from nꢂ4ꢀ11 experiments. Means were tested
/
/
/

312
M. Glanzel et al. / European Journal of Medicinal Chemistry 38 (2003) 303ꢀ312
/
¨
[23] J. Pfeilschifter, Biochem. J. 272 (1990) 469ꢀ
[24] Y.D. Clonis, N.E. Labrou, V.Ph. Kotsira, C. Mazitsos, S.
Melissis, G. Gogolas, J. Chromatogr. 891 (2000) 33ꢀ44.
[25] A. Denizli, E. Piskin, J. Biochem. Biophys. Methods 49 (2001)
391ꢀ416.
[26] D. Hanggi, P. Carr, Anal. Biochem. 149 (1985) 91ꢀ
[27] S.J. Burton, S.B. McLoughlin, C.V. Stead, C.R. Lowe, J.
Chromatogr. 435 (1988) 127ꢀ137.
[28] C.R. Lowe, Top Enzyme Ferment. Biotechnol. 9 (1984) 78ꢀ
[29] B.H. Weber, K. Willeford, J.G. Moe, D. Piszkiewicz, Biochem.
Biophys. Res. Commun. 86 (1979) 252ꢀ258.
[30] J. Panchartek, Z.J. Allen, F. Muzik, Col. Czech. Chem. Commun.
25 (1960) 2783ꢀ2799.
/
472.
[33] F. Tuluc, R. Bultmann, M. Glanzel, A.W. Frahm, K. Starke,
¨
¨
Naunyn Schmiedeberg’s Arch. Pharmacol. 357 (1998) 111ꢀ
/
/
120.
[34] W. Wojtkiewicz, J. Kraska, Chem. Abstr. 61 (1964) 16203f.
[35] J.C. Pearson, S.J. Burton, C.R. Lowe, Anal. Biochem. 158 (1986)
/
/
104.
382ꢀ/389.
[36] R. Bultmann, K. Starke, Naunyn Schmiedeberg’s Arch. Pharma-
¨
/
col. 352 (1995) 477ꢀ
/482.
/161.
[37] R. Bultmann, B. Pause, H. Wittenburg, G. Kurz, K. Starke,
¨
Naunyn Schmiedeberg’s Arch. Pharmacol. 354 (1996) 481ꢀ
/
/
490.
[38] R. Bultmann, H. Wittenburg, B. Pause, G. Kurz, P. Nickel, K.
¨
Starke, Naunyn Schmiedeberg’s Arch. Pharmacol. 354 (1996)
/
[31] Society of Dyers and Colourists, Colour Index, Lund, Humphries,
Bradford and London, vol. 3 and 4, third ed., Society of Dyers
and Colourists, London, 1975.
498ꢀ504.
/
[39] H. Wittenburg, R. Bultmann, B. Pause, C. Ganter, G. Kurz, K.
¨
Starke, Naunyn Schmiedeberg’s Arch. Pharmacol. 354 (1996)
[32] A. Foris, in: K. Venkataraman (Ed.), Analytical Chemistry of
491ꢀ
[40] Cell Signalling & Neuroscience Product Catalogue, Sigma-Al-
drich, Taufkirchen, Germany, 2002ꢀ2003, p. 738.
/
497.
Synthetic Dyes, Wiley-Interscience, New York, 1977, pp. 217ꢀ
/
254.
/