Synthesis and pharmacological evaluation of prenylated and benzylated pterocarpans against snake venom

Article abstract of DOI:10.1016/j.bmcl.2003.10.044

Edunol (3), a pterocarpan isolated from Harpalyce brasiliana, a plant used in the northeast of Brazil against snakebites, was obtained by synthesis and showed antimyotoxic, antiproteolytic and PLA2 inhibitor properties. These proprieties could be improved

Full text of DOI:10.1016/j.bmcl.2003.10.044

Bioorganic & Medicinal Chemistry Letters 14 (2004) 431–435
Synthesis and pharmacological evaluation of prenylated
and benzylated pterocarpans against snake venom
Alcides J. M. da Silva,a Antonio L. Coelho,a Alessandro B. C. Simas,a
Raphael A. M. Moraes,b Diogo A. Pinheiro,b Fabrıcio F. A. Fernandes,b
Emerson Z. Arruda,b Paulo R. R. Costaa,* and Paulo A. Melob,*
a
´
Laboratorio de Quımica Bioorganica (LQB), Nucleo de Pesquisas de Produtos Naturais, Centro de Ciencias da Saude,
´
ˆ
Bloco H, Universidade Federal do Rio de Janeiro, RJ 21941-590, Brazil
´
ˆ
´
b
´
Departamento de Farmacologia Basica e Clınica, Centro de Ciencias da Saude, Bloco H, Universidade Federal do Rio de Janeiro,
´
ˆ
RJ 21941-590, Brazil
´
Received 9 July 2003; revised 15 October 2003; accepted 25 October 2003
Abstract—Edunol (3), a pterocarpan isolated from Harpalyce brasiliana, a plant used in the northeast of Brazil against snakebites,
was obtained by synthesis and showed antimyotoxic, antiproteolytic and PLA2 inhibitor properties. These proprieties could be
improved through the synthesis of a bioisoster (5), where the prenyl group was replaced by the benzyl group.
# 2003 Elsevier Ltd. All rights reserved.
Viperidae snakebites are an important health problem
in Brazil and other South American countries.1 Among
these snakebites, that of the genus Bothrops induces
local swelling hemorrhage and myonecrosis, which have
been attributed to the venoms proteolytic and phos-
pholipase activity.2 The first aid treatment and man-
agement of this kind of accident are among the most
neglected and poorly studied areas of medicine. The
only specific therapy known so far is the specific anti-
venom, which has low and limited effectiveness.2,3
cabenegrina A-I (1) and cabenegrina A-II (2) (Fig. 1).
More recently a new pterocarpan denominated edunol
(3) was also isolated from the root of the Harpalyce
brasiliana, a plant found in the northeast of Brazil and
used against snakebites.7
While active compounds 1–3 have a C5 side chain on
the A-ring, maackian (4), which is not active against
snake venoms, lacks the A-ring side chain.
During the last decades, several extracts of plants with
activity against snake venoms have been reported in
the literature.4 The naturally occurring coumestan
wedelolactone, isolated from Eclipta prostrata and
some of its synthetic derivatives, such as PLA2 inhibi-
tors, are among the most active products.5 On the other
hand, pterocarpans prenylated in the A-ring are also
described as very active products. For example, the
plant popularly called ‘Cabec¸a-de-Negro’ is main ingre-
6
´
dient in ‘Especıfico Pessoa’, a northeast Brazilian folk
medicine used against snakebites. Two active pter-
ocarpans were isolated from this popular medicine,
* Corresponding authors. Tel.: +55-212-5626791; fax: +55-212-
2702683; e-mail: lqb@nppn.ufrj.br
Figure 1. Prenylated and benzylated pterocarpans.
0960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.10.044

432
A. J. M. da Silva et al. / Bioorg. Med. Chem. Lett. 14 (2004) 431–435
pionaldehyde dimethylacetal, followed by cyclization in
acid medium. Compound 13 was allowed to react with
organomercurial (14) in the presence of PdCl2/LiCl,
leading to the O-benzylated pterocarpan. The protecting
benzyl group was removed by hydrogenolysis leading
the target, compound 5 (Scheme 1).
The activity of edunol derivatives against muscle
damage induced by B. jararacussu crude venom was
tested. This crude venom is myotoxic, inducing tissue
hemorrhage and myonecrosis mainly due to the pre-
sence of proteolytic and phospholipase A2 enzyme
activities.10 Sarcolemmal damage and myotoxicity
induced by B. jararacussu crude venom were assessed in
vitro by the rate of loss of creatine kinase (CK) from
isolated extensor digitorum longus (EDL) muscles
mouse.11 At 0.1–30 mM, many edunol derivatives (com-
pounds 4, 6–9) did not antagonize the increased rate of
CK release induced by the venom in vitro (Table 1).
Meanwhile, Edunol (3) inhibited the myotoxic activity
in vitro, only at concentrations of 0.1–3 mM (Fig. 2A),
whereas compound 5 showed a potent antivenom activ-
ity (Fig. 2B).
Scheme 1. (i) BuLi (1.2 equiv), THF, 0 ꢀC, 40 min, benzyl bromide; (ii)
PPTS (0.2 equiv), iPrOH, 80 ꢀC; (iii) iodo dimethylacetal, KOH, THF;
(iv) HCl (20%); (v) PdCl2, LiCl, acetone; (vi) H2-Pd/C acetone.
Since the benzyl group is a bioisoster of the prenyl
group, compound 5 is designed as an analogue of 3, as
well as compounds 6 and 7, which have an additional
prenyl or benzyl group on ring D. Compounds 8 and 9
were also designed as analogous of 3. In fact, the com-
pounds 1–4, 8 and 9 get the prenyl or benzyl group
positioned between two alkoxy groups.
The compound 5 was able to fully inhibit the venom’s
myotoxic activity in vitro with an IC50 of 9.97 micro-
molar (Fig. 2B). Interestingly, at 100 mM, this pter-
ocarpan also inhibited 65% the venom phospholipase
activity, as well as more than 80% of its proteolytic
activity.
The natural product Edunol (3) and the analogues (6, 7,
8 and 9) were prepared using ortho-metalation chem-
istry (DMG) on derivatives of 4, as previously descri-
bed.8 Herein, we report the synthesis of the new
analogue 5 using an alternative synthetic pathway.
First, resorcinol derivative 10 was regioselectively
lithiated9 and the resulting intermediate was trapped
with benzyl bromide followed by in situ hydrolysis of
the MOM protecting group, leading to compound 11.
This compound was transformed in to chromen 13 in a
two step sequence involving O-alkylation of 3-iodopro-
Figure 3. Effect of the treatment with edunol (3) and its derivative
compound 5 on the myotoxic effect of B. jararacussu crude venom in
vivo. Panel A shows the effect of intramuscular injection of B. jarar-
acussu crude venom (V; 1.0 mg/g) on plasma creatine kinase (CK)
activity and the effect of the treatment with edunol. Edunol (0.1–
3.0 mg/g) was administered after preincubation in vitro with the
venom. Plasma CK activity of the physiological saline solution (PSS)
control group was 181.16Æ26.44 U/l (n=5). The values reported are
meansÆSEM (n=5). p<0.05 for the difference between the control
(venom alone) and the values for each of the edunol concentration
plus venom (ANOVA). Panel B data shows the inhibition of the
increase of plasma CK activity induced by B. jararacussu crude venom
injection by compound 5, in dose of 1.0 and 10 mg/g, respectively. The
compound was pre-incubated with the venom and administered by
i.m. injection. The mice received the venom injection at a dose of 1 mg/
g and the value of plasma CK activity of PSS control group was
184.34Æ31.10 U/L (n=5). The values were expressed as mean
Æstandard error; p<0.05 (ANOVA) for the difference between the
control (venom alone) and the values from the two different doses of
the compound.
Figure 2. Effect of Edunol (3) or its derivative compound 5 on the
myotoxic effect of B. jararacussu crude venom in vitro. Data in panel
A and B (p<0.05) show the percent of inhibition of the increased rate
of CK release induced by B. jararacussu crude venom (25 mg/mL) on
isolated EDL mouse muscles. Each point represents the mean-
Æstandard error (n=4–6), after 60 min of exposure to the crude
venom alone or associated to 3 or 5, respectively. The basal rate
of CK release in physiologic saline solution was 0.35
Æ0.05 U.gÀ1.hÀ1(n=83). The curve was determined by non-linear fit-
ting to the Hill equation. Each point represents the meanÆstandard
error (n=4–6; IC50=9.97 mM).

A. J. M. da Silva et al. / Bioorg. Med. Chem. Lett. 14 (2004) 431–435
433
Table 1. Effect of Edunol and different pterocarpan derivatives on
the myotoxic effect of B. jararacussu crude venom
Compds
Concentration mM
Effect
% of Inhibitiona
Edunol (3)
0.1–1.0
10.0–30.0
0.1–30.0
0.1–10.0
0.1–30.0
0.1–30.0
0.1–30.0
0.1–30.0
+
À
>40%
0%
0%
>70%
0%
0%
4
5
6
7
8
9
À
++
À
À
À
0%
0%
À
Each compound, Edunol and pterocarpan derivatives, was tested
4–6 times.
a The Data represents the % of the inhibition of the increase of the
rate of CK release induced by 25 mg/mL of B. jararacussu venom on
isolated EDL mouse muscles after 60 min of exposure.
Figure 4. Effect of compound 5 on the phospholipase and proteolytic
activity of B. jararacussu crude venom. Panel A and B show the inhi-
bition of phospholipase and proteolytic activity of B. jararacussu
crude venom (V; 10.0 mg/mL) by compound 5 (100 mM) respectively.
Panel A shows the phospholipase activity of B. jararacussu crude
venom incubated with hen’s egg-yolk buffered emulsion (pH 7.8),
mixed and carried out at 37 ꢀC during 30 min, alone or in presence of
premixed different concentrations of compound 5. The absorbance
change between each 30 min was measured spectrophotometrically at
925 nm. Data from four identical experiments was measured at 30 min.
p<0.05 for the difference between the control (venom alone) and the
values for the compound 5 plus venom (ANOVA). Panel B shows the
effect of compound 5 on proteolytic activity of the venom. Aliquots of
crude venom alone or venom plus compound 5 (100 mM) were incu-
bated with 0.2% azocasein in 0.2 M Tris–HCl, 20 mM CaCl2, pH 8.8,
for 90 min at 37 ꢀC, in a final volume of 0.8 mL. The reaction was
stopped with 0.4 mL of 10% trichloroacetic acid and fractions were
centrifuged for 10 min. The supernatant (1 mL) was mixed with 2 M
NaOH (0.5 mL) and absorbance was measured at 480 nm. The data
were expressed in % of control enzyme activity.
For in vivo experiments (Fig. 3), the increase of plasma
CK activity was measured after i.m. injection of the
crude venom.11 The compounds 3 and 5 protected the
mice in a dose dependent response. In these experi-
ments, B. jararacussu crude venom (1.0 mg/kg), alone or
preincubated together with 3 or 5 for 15 min at 37 ꢀC (in
vitro), was injected i.m. into each mouse. The final
volume of venom injected, alone or plus 3 or 5, was
0.1 mL. These venom doses were selected based on pre-
vious studies.12 The animals were lightly anesthetized
with diethyl ether immediately before and 2 h after
venom injection for blood collection according to the
guidelines for care and use of laboratory animals.
Plasma was separated by centrifugation and stored at
4 ꢀC for subsequent determination of CK activity. The
procedure for the measurement of CK activity has been
described previously (1, 5–8). Three weeks after venom
injection more than 50% of mice that received i.m.
venom injection did not survive. Meanwhile, the ani-
mals that received treatment with edunol (3) or com-
pound 5 survived the test.4b
and the present work, by using crotalid snake venoms,5
have allowed us not only to confirm the observations
from folk medicine, but also to develop new antisnake
venoms that can neutralize many venom activies.
Compound 5 is a new synthetic drug with potential
therapeutic application, but needs further studies. The
development of a drug against snakebites is a relevant
issue and a challenge, because snake venoms are not
composed of single compounds, being instead
a
These results support our hypothesis that natural com-
pound 3 and synthetic derivative can inhibit snake
venoms and, the new compound 5 also can be improved
and deserves further study as a potential new anti-
venom.
complex mixture causing an even more complex local
reaction.14
In conclusion, the antivenom activity of the natural
product edunol (3) was improved by changing the pre-
nyl group at position 4 to a benzyl group. On the other
hand, the acyivity was completely removed by the
introduction of an additional prenyl or benzyl group in
position 10.
The protective effect of compound 5 is due to inhibition
of phospholipase and proteolytic activity of B. jarar-
acussu, as shown in Figure 4. Compound 3 did not pre-
sent protection against proteolytic or phospholipase
activity.
1.Experimental
1.1. Chemistry
Our data shows that compound 5 antagonized some
activities of B. jararacussu crude venom which contains
peptides and proteins that act on the physiological
mechanisms of the body. To improve snakebite treat-
ment we need to acquire new molecules from plants
as well new synthetic antagonists that can prevent the
toxins from reaching the mammalian targets and
acting as cytotoxic agents. Polyvalent antivenin are
composed of specific antibodies that neutralize
individual venom component and they are limited to
some venom components.13 Our previous investigations
Melting points were determined with a Thomas-Hoover
apparatus and are uncorrected. Column chromato-
graphy was performed on silica gel 230–400 mesh
1
(Aldrich). H NMR spectra were recorded on a Varian
Gemini (200 MHz) instrument using tetramethylsilane
(TMS) as standard and CDCl3 as solvent. J Values are
given in Hz. 13C NMR spectra were obtained at
50 MHz.

434
A. J. M. da Silva et al. / Bioorg. Med. Chem. Lett. 14 (2004) 431–435
1.2. Synthesis of the Pterocarpan (5)
mixture was i.m. injected into the mice. Before and 2 h
after the i.m. injection the animals were lightly anesthe-
tized with diethyl-ether and the blood was collected by
orbital puncture. The plasma was separated by cen-
trifugation and stored at 4 ꢀC for subsequent determi-
nation of creatine kinase (CK) activity.
To a mixture of PdCl2 (0.194 g, 1.1 mmol) and LiCl
(0.047 g, 1.1 mmol) in acetone (10 mL) was added chro-
men 13 (0.360 g, 1.1 mmol) in acetone (10 mL). This
mixture was stirred for 15 min at 0 ꢀC and then 2-chloro-
mercurio-4,5-methylenedioxyphenol 13 (0.414 g, 1.1
mmol) in acetone (10 mL) was added to it. The suspen-
sion thus obtained was stirred for 12 h at 25 ꢀC. After
this time, brine (30 mL) was added to it and the mixture
was extracted with acetyl acetate (100 mL), the organic
extract dried (Na2SO4), and subjected to column chro-
matography to give the O-benzyl pterocarpan as a white
solid (0.280 g, 55% yield), mp 78–80 ꢀC. 1H NMR
(CDCl3) d 3.50 (m, 1H), 3.60 (t, J=11.1 Hz, 1H), 4.06
(s, 2H), 4.27 (dd, J=11.1, 4.5 Hz, 1H), 5.08 (s, 2H), 5.50
(d, J=6.3 Hz, 1H), 5.90 (d, J=5.5, 1H), 5.91 (d,
J=5.5 Hz, 1H), 6.42 (s, 1H), 6.68 (d, J=8.5 Hz, 1H),
6.71 (s, 1H), 7.26 (m, 11H). Anal. calcd for C30H24O5:
C, 77.57; H, 5.21. Found: C, 77.50; H, 5.19. The solu-
tion of the O-benzyl pterocarpan (0.208 g, 0.40 mmol) in
acetone was hydrogenated (3 atm) in the presence of Pd-
C (10%). After 3 h the catalyst was filtered and the
product purified by chromatography on silica to give (5)
in 80% yield, mp 60–62 ꢀC. 1H NMR (CDCl3) d 3.50 (t,
J=11.0, 1H), 3.64 (m, 1H), 4.03 (s, 2H), 4.27 (dd,
J=11.0, 4.4 Hz, 1H), 5.49 (d, J=6.5 Hz), 5.90 (d,
J=5.5 Hz, 1H), 5.91 (d, J=5.5 Hz, 1H), 6.43 (s, 1H),
6.55 (d, J=8.4 Hz, 1H), 6.71 (s, 1H), 7.25 (m, 6H).
Anal. calcd for C23H18O5: C, 73.79; H, 4.85. Found: C,
73.82; H, 4.78.
3.Enzymatic assays
Proteolytic and Phospholipase activity were performed
according to refs 12a and b respectively. Briefly, Phos-
pholipase activity of B. jararacussu crude venom was
evaluated by incubating 10 mg of the venom with hen’s
egg-yolk buffered emulsion (pH 7.8), mixed and carried
out at 37 ꢀC during 30 min., alone or in presence of
premixed different concentrations of compounds. The
absorbance change between each 30 min. was measured
spectrophotometrically at 925 nm. Proteolytic activity
of the venom was evaluated by added aliquots of crude
venom alone or venom plus compound 5 (100 mM) and
incubated with 0.2% azocasein in 0.2 M Tris–HCl,
20 mM CaCl2, pH 8.8, for 90 min at 37 ꢀC, in a final
volume of 0.8 mL. The reaction was stopped with
0.4 mL of 10% trichloroacetic acid and fractions were
centrifuged for 10 min. The supernatant (1 mL) was
mixed with 2 M NaOH (0.5 mL) and absorbance was
measured at 480 nm. The data were expressed in% of
control enzyme activity. Each point represents the data
from four identical experiments. p<0.05 for the differ-
ence between the control (venom alone) and the values
for the compound 5 plus venom (ANOVA).
1.3. Data analysis
The concentration–response curves were fitted by the
Hill equation, using nonlinear regression. Data from
assays of venom-induced CK release was well fitted by a
single-site model.
Acknowledgements
Our research was supported by grants from PRONEX
(No. 41.96.0888.00), FAPERJ, FUJB-UFRJ, and
CAPES, in Brazil. A.J.M.S is a postdoctoral fellow
of FAPERJ (No. 26/151.081/97). P.R.R.C is CNPq
fellows. D.A.P. is recipient of CNPq fellowship.
R.A.M.M. is recipient of CNPq (PIBIC) fellowships.
F.F.A.F. is recipient of FAPERJ fellowship. E.Z.A. is
recipient of CAPES fellowship.
2.Myotoxicity assays
In vitro and in vivo myotoxicity was evaluated at room
temperature as previously described in references 4a,b,c.
Briefly, Extensor digitorum longus muscles were blot-
ted, weighed rapidly and then transferred to sample
collecting units of 2.5 mL capacity, where they were
superfused continuously at a rate of 3 mL/min with
physiological saline solution (PSS) equilibrated with
95% O2/5% CO2. At 30 min intervals, the solutions
perfusing the muscles were collected and replaced with
fresh media. The CK activity was determined using a
diagnostic kit purchased from Sigma Chemical Co. The
rate of CK release from the isolated muscles is expressed
as international enzyme units released into the medium
per gram of muscle per h of collection (U.gÀ1.h.À1).
Bothrops jararacussu venom alone or associated to the
edunol or its derivatives was added to the nutrient
solution which superfused the isolated muscles. In vivo
the myotoxicity was evaluated by injecting intramuscu-
larly (i.m.) the venom dissolved in physiological saline
solution (PSS, 0.1 mL) alone or preincubated with edu-
nol or compound 5 (15 min, 37 ꢀC), after which the
References and notes
1. Rosenfeld, G. In Venomous Animals and their Venoms;
Academic Press Ed.: New York, 1971; vol. 2, p 345.
2. Watt, G. In The Venomous Reptiles of Latin America.
Campbell, J. A., Lamar, W. W. Eds., Cornell University
Press: Ithaca, London, New York, 1989; p 6.
3. Chippaux, J. P.; Goyffon, M. Toxicon 1998, 36, 823.
4. (a) Mors, W. B.; Nascimento, M. C.; Parente, J. P.; Silva,
M. H.; Melo, P. A.; Suarez-Kurtz, G. Toxicon 1989, 27,
1003. (b) Melo, P. A.; do Nascimento, M. C.; Mors, W. B.;
Suarez-Kurtz, G. Toxicon 1994, 32, 595. (c) Melo, P. A.;
Ownby, C. L. Toxicon 1999, 37, 199.
5. da Silva, A. J. M.; Melo, P. A.; Silva, N. M.; Brito, F. V.;
Buarque, C. D.; de Souza, D. V.; Rodrigues, V. P.; Poc¸ as,
E. S.; Noel, F.; Albuquerque, E. X.; Costa, P. R. R.
¨
Bioorg. Med. Chem. Lett. 2001, 11, 283.
6. Nakagawa, M.; Nakanishi, K.; Darko, L. L.; Vick, J. A.
Tetrahedron Lett. 1982, 23, 3855.

A. J. M. da Silva et al. / Bioorg. Med. Chem. Lett. 14 (2004) 431–435
435
7. Silva, G. L.; Matos, F. J. A.; Silveira, E. R. Phytochem-
istry 1997, 46, 1059.
8. Costa, P. R. R.; da Silva, A. J. M.; Coelho, A. L.; Simas,
A. B. C. Tetrahedron Lett. 2001, 42, 4111.
9. Costa, P. R. R.; Coelho, A. L.; Simas, A. B. C. Synthesis
2001, 42, 4111.
10. (a) Ferreira, M. L.; Moura-da-Silva, A. M.; Franca, F. O.;
Cardoso, J. L.; Mota, I. Toxicon 1992, 30, 1603. (b) Mebs,
D.; Ownby, C. L. Pharmacol. Ther. 1990, 48, 223.
11. (a) Melo, P. A.; Suarez-Kurtz, G. Toxicon 1988, 26, 87.
(b) Melo, P. A.; Ownby, C. L. Toxicon 1996, 34, 653.
˜
12. (a) Garcia, E. S.; Guimaas, J. A.; Prado, J. L. Arch. Bio-
chem. Biophys. 1978, 315. (b) Marinetti, G. V. Biochim.
Biophys. Acta 1965, 98, 554.
13. Chippaux, J. P.; Goyffon, M. Toxicon 1998, 36, 823.
14. Sanchez, E. F.; Fretas, T. V.; Ferreira-Alves, D. L.;
Velarde, M. R.; Diniz, M. N.; Agostinicotta, G.;
Cordeiro, M. N.; Diniz, C. R. Toxicon 1992, 30, 95.