Journal of Ethnopharmacology 92 (2004) 53–56

Evaluation of the bioactivity of triterpene mixture isolated

from Carmona retusa (Vahl.) Masam leaves
Irene M. Villaseñor∗ , Arlyn P. Canlas, Karen M. Faustino, Katherine G. Plana
Institute of Chemistry, University of the Philippines, Diliman, 1101 Quezon City, Philippines
Received 2 February 2003; received in revised form 13 January 2004; accepted 26 January 2004

Abstract

The major constituent of Carmona retusa (Vahl.) Masam. leaves is an intractable mixture of triterpenes, namely ␣-amyrin (43.7%), ␤-amyrin
(24.9%), and baurenol (31.4%). At a dosage of 100 mg/kg mouse, the triterpene mixture exhibited 51% analgesic activity but only showed 20%
anti-inflammatory activity. Kruskal–Wallis one-way analysis of variance by ranks showed that the triterpene mixture is as active as mefenamic
acid, a commercially available analgesic, at α = 0.01. The charcoal tracing test showed a 29% anti-diarrheal activity for the triterpene mixture,
which increased to 55% at a dosage of 250 mg/kg mouse. At the higher dosage, the triterpene mixture differed significantly from its solvent
control at α = 0.01. Results of the micronucleus test showed that the triterpene mixture did not exhibit mutagenic nor anti-mutagenic activity
at α = 0.001. There was no significant decrease in blood glucose levels (bgl) in alloxan-induced diabetic mice after administration of the
triterpene mixture. The triterpene mixture was inactive against Escherichia coli and possessed moderate activities against Staphylococcus
aureus, Candida albicans, and Trichophyton mentagrophytes.
© 2004 Elsevier Ireland Ltd. All rights reserved.

Keywords: Analgesic; Anti-diarrheal; ␣-Amyrin; ␤-Amyrin; Baurenol; Carmona retusa (Vahl.); Masam

1. Introduction Our present research is on the determination of the

The Department of Health of the Philippines included in
its primary health care the use of medicinal plant prepara-
tions, targeted symptomatics rather than curatives, and came
up with a list of 10 priority plants for dosage formulation
and clinical trials. Carmona retusa (Vahl.) Masam. (Boragi-
naceae) is one of these priority plants under the Department
of Science and Technology–Philippine Council for Health
Research and Development–National Integrated Program on
Medicinal Plants. The dried leaves are available as 250 mg
tablets and are recommended as anti-colic and anti-diarrheal.
The tablets have undergone clinical trial phases 1 and 2 as
well (Cortes-Maramba et al., 1991). The leaves are used as
a stomachic, for cough, fever, and secondary and constitu-
tional syphilis (de Padua et al., 1982; Quisumbing, 1978).
During the processing for the isolation of an anti-mutagenic
constituent (Villaseñor and Edu, 1993; Villaseñor et al.,
1993), white solids crystallized out in huge quantities.
GC–MS analysis showed an intractable mixture of triter-
penes consisting of structural isomers: ␣-amyrin, ␤-amyrin,
and baurenol (Villaseñor et al., 1992).
∗ Corresponding author. Fax: +63-2-9205427.
E-mail address: ivillase@chem.upd.edu.ph (I.M. Villaseñor).
anti-diarrheal constituent from the leaves of Carmona re-
tusa. The triterpene mixture was again isolated as the major
constituent and they turned out to be anti-diarrheal. Appre-
hensive that we might again miss their other bioactivities,
we investigated this triterpene mixture for its possible
analgesic, anti-inflammatory, mutagenic, anti-mutagenic,
anti-diabetic, and anti-microbial activities.
2. Materials and methods
2.1. Plant material
Carmona retusa leaves were bought from a herbal store.
A voucher specimen was deposited at the Dr. Jose Vera
Santos Herbarium at the Institute of Biology, University of
the Philippines, Diliman with accession number 08903 and
authenticated by L.L. Co.
2.2. Isolation and identification of active compounds
Air-dried Carmona retusa leaves were homogenized
in methanol, filtered and concentrated in vacuo. The
MeOH extract was then partitioned between water and

0378-8741/$ – see front matter © 2004 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.jep.2004.01.017
54 I.M. Villaseñor et al. / Journal of Ethnopharmacology 92 (2004) 53–56
hexane. The hexane layer was collected and concentrated in
vacuo.
The hexane extract (21.7 g) was subjected to vacuum liq-
uid chromatography using a 9.5 cm column filled to 7 cm
height of silica gel 60G (Merck) as stationary phase. The
column was eluted with 500 ml hexane, 10% gradient ratios
of EtOAc in hexane, and EtOAc. The column fractions were
pooled according to analytical thin layer chromatography
results.
The fraction eluted out with 20% EtOAc in hexane was
rechromatographed using silica gel and eluted with hexane
and 2% gradient ratios of EtOAc in hexane. Yellowish crys-
tals, precipitated out in 4% EtOAc in hexane, were collected
and washed with acetone until white crystals were obtained.
The white crystals were analyzed by GC–MS at the
National Pesticide Analytical Laboratory of the Bureau
of Plant Industry using an HP 6890 spectrophotometer:
J&W 125-0212 column (30.0 m length, 320.0 ␮m diam-
eter, 0.25 ␮m film thickness), helium was used as carrier
gas, 1.0 psi, 1 ␮l sample splitless injection at an isothermal
temperature of 240 ◦ C, constant flow of 1.3 ml/min.
2.3. Animals
Swiss Webster albino mice were purchased from the Bu-
reau of Animal Industry and were acclimatized for at least
a week before the start of the experiments. Five mice were
used for each test sample.
2.4. Anti-diarrheal assay: charcoal tracing method
(PCHRD, 1989)
The charcoal meal, which was used as an indicator, was
prepared as follows: 2 g of activated charcoal was mixed
with 20 ml of castor oil to form a uniform suspension. This
suspension was then mixed with 10 ml of coconut oil.
After 12 h fasting, the triterpene mixture was adminis-
tered orally to mice at dosages of 100 and 250 mg/kg BW
of mouse. After 1 h, the mice were fed with the charcoal
meal. Twenty minutes later, they were sacrificed by cer-
vical dislocations, their abdomens were opened, and their
small intestines were excised from the pylorus to the ce-
cum. The total length of the intestine and the length that
the charcoal meal has traveled were measured. Loperamide
(Immodium® ) was used as reference.
2.5. Analgesic bioassay: acetic acid-induced writhing test
(Villaseñor et al., 2002)
Approximately 30 min after administration of the test
samples, 0.7% acetic acid was injected intraperitoneally
(0.2 ml/20 g mouse). The number of writhings for each
mouse was then counted for 15 min starting from 5 min
after acetic acid injection. The triterpene mixture was dis-
solved in corn oil and was injected intraperitoneally at a
dosage of 100 mg/kg mouse. Mefenamic acid, the positive
control, was dissolved in water (3.5 mg/kg mouse) and was
administered orally.
2.6. Mutagenicity/antimutagenicity bioassay: micronucleus
test
Swiss Webster albino mice, 7–12 weeks old, were used.
The required weight of the triterpene mixture (100 mg/kg
BW of mouse) was dissolved in corn oil while the required
weight (55 mg/kg BW of mouse) of tetracycline (Upjohn)
was dissolved in distilled water. Tetracycline, the positive
control, was administered orally using a gavage while the
test sample was injected intraperitoneally to the test animals.
The protocol used was described by Schmid (1975).
2.7. Anti-diabetic assay: alloxan-induced (Esmerino et al.,
1998)
Male mice were fasted for 24 h with water ad libitum and
their blood glucose levels (bgl) were then determined using
one touch basic blood glucose meter and test strips. Dia-
betes was induced by intraperitoneal injection of alloxan at
a dosage of 150 mg/kg mouse. Their bgl were again deter-
mined after 2 days. Animals with bgl over 13.8 mmol/l were
considered diabetic. The test samples were then adminis-
tered orally at t = 0. The bgl were monitored at 30-min inter-
vals. The triterpene mixture (100 mg/kg BW of mouse) was
dissolved in dimethylsulfoxide while glipizide (0.77 mg/kg
BW of mouse), the positive control, was dissolved in
water.
2.8. Anti-inflammatory bioassay: carrageenan-induced
mouse paw edema test (Winter et al., 1962)
The initial volume of the mouse hind paw was measured
using a plethysmometer. One hour after intraperitoneal in-
jection of the triterpene mixture at a dosage of 100 mg/kg
BW of mouse, 0.01 ml carrageenan was injected intrader-
mally into the left hind paw. After 3 h, the volume of the
mouse hind paw was measured. The difference in the initial
and final volumes was then computed and compared with
those of the solvent control. Indomethacin, the positive con-
trol, was administered orally at a dosage of 1.4 mg/kg BW
of mouse.
2.9. Antimicrobial assay: agar cup method
The anti-microbial assay was done by the Microbiolog-
ical Services Laboratory of the Natural Sciences Research
Institute using the agar cup method. The test organisms used
were Escherichia coli UPCC 1195, Staphylococcus aureus
UPCC 1143, Candida albicans UPCC 2168, and Trichophy-
ton mentagrophytes UPCC 4193. Aliquots of the cell sus-
pensions of the bacteria, yeast and molds were transferred
into pre-poured nutrient agar, glucose yeast peptone agar and
potato dextrose agar plates, respectively. The plates were
 I.M. Villaseñor et al. / Journal of Ethnopharmacology 92 (2004) 53–56 55

incubated at room temperature for 24, 48, and 72 h, respec- Table 2

tively. Anti-inflammatory assay of the triterpene mixture using the
carrageenan-induced mouse paw edema assay
Test samples Dosage Average change in Percentage
2.10. Statistical analysis
(mg/kg) volume (ml) ±S.D. inhibition
Triterpene mixture 100 0.016 ± 0.016 20
Data were analyzed using the Kruskal–Wallis one-way 250 0.030 ± 0.024 −50
analysis of variance by ranks. This is an extremely useful test
Corn oil 10 ml 0.020 ± 0.012
for deciding whether k independent samples are from dif-
ferent populations. For the anti-diabetic assay, t-test: paired Indomethacin 1.4 0.011 ± 0.012 69
two sample for means was used. Water 10 ml 0.036 ± 0.022

Table 3
3. Results and discussion Anti-diarrheal assay of the triterpene mixture using the charcoal tracing
method
The gas chromatogram of the triterpene mixture (1.33% Test samples Dosage Average percentage Percentage
yield) showed the presence of three compounds with re- (mg/kg) distance (mm) inhibition
tention times of 19.84, 21.00, and 23.23 min with relative traveleda ± S.D.
amounts of 24.9, 43.7, and 31.4%, respectively. Their mass Triterpene mixture 100 43.8 ± 5.7 29
spectra showed that they are structural isomers with a molec- 250 28.1 ± 26.3 55
ular ion peak at m/e 426 and the following prominent peaks Corn oil 10 ml 61.9 ± 13.8
with their relative abundances: Rt = 19.84 min—426 (8%),
Loperamide 10 36.5 ± 7.7 45
218 (100%), 205 (42%), 189 (10%); Rt = 21.00 min—426
(12%), 218 (100%), 205 (20%), 189 (16%); Rt = 426 Water 10 ml 65.9 ± 7.2
(32%), 411 (28%), 393 (8%), 247 (100%), 229 (62%), 205 a Percentage of distance traveled = (distance traveled by the charcoal/
(14%), 187 (8%). Reverse fit data and comparison with lit- length of the small intestine) × 100.
erature (Villaseñor et al., 1992) showed that Rt = 19.84 min
was ␤-amyrin, Rt = 21.00 min was ␣-amyrin, and Rt = The triterpene mixture exhibited a 29 and 55% anti-
23.23 min was baurenol. diarrheal activity at dosages of 100 and 250 mg/kg BW of
At a dosage of 100 mg/kg BW of mouse, the triterpene mouse, respectively (Table 3). KW: 13.79 denoted differ-
mixture exhibited analgesic activity, as shown by a 51% re- ences in responses at α = 0.01. At this level of significance,
duction in the number of writhes induced by acetic acid the variances of loperamide and the triterpene mixture at
(Table 1). The computed Kruskal Wallis statistic (KW = 250 mg/kg BW of mouse differed significantly from their
14.46) showed that there were differences in the responses solvent controls, while that of the triterpene mixture at
of the mice to the different treatments at 0.01 level of signif- 100 mg/kg BW of mouse was not significantly different
icance. Comparisons of the treatments versus solvent con- from that of corn oil.
trols showed that mefenamic acid differed from water at Results of the micronucleus test (Table 4) showed that the
α = 0.01 while the triterpene mixture differed from corn triterpene mixture is not mutagenic as it approximates the
oil at α = 0.10. Comparison between treatments showed number of MN-PCE induced by corn oil, the solvent control.
that the triterpene mixture is as active as mefenamic acid, a It is not anti-mutagenic because it was not able to decrease
commercially available analgesic, at α = 0.01. the number of MN-PCE induced by tetracycline, the positive
The triterpene mixture did not exhibit anti-inflammatory control. KW: 24.66 implied that there were differences in
activity (Table 2). KW: 4.89 indicated differences in re- responses at α = 0.001. At this level of significance, the
sponses only at α = 0.20. At this level of significance, in- triterpene mixture is neither mutagenic nor anti-mutagenic
domethacin differed from water while the triterpene mixture
is statistically similar to corn oil. Table 4
Mutagenicity and anti-mutagenicity assay of the triterpene mixture using
the micronucleus test
Table 1
Test samples Dosage (Number of MN-PCEa /
Analgesic assay of the triterpene mixture using the acetic acid-induced
(mg/kg) 1000 PCEb ) ± S.D.
writhing test
Triterpene mixture 100 1.44 ± 0.88
Test samples Dosage Average number Percentage Tetracycline 55 5.66 ± 1.58
(mg/kg) of writhes ± S.D. inhibition Corn oil 10 ml 1.11 ± 0.93
Triterpene mixture 100 26 ± 12 51 Triterpenes + tetracycline 100/55 5.22 ± 1.39
Corn oil 10 ml 53 ± 12 Corn oil + tetracycline 10 ml/55 5.22 ± 1.39
Mefenamic acid 3.5 13 ± 9 76 a Micronucleated polychromatic erythrocytes.
Water 10 ml 54 ± 6 b Polychromatic erythrocytes.
56 I.M. Villaseñor et al. / Journal of Ethnopharmacology 92 (2004) 53–56

Table 5
Anti-diabetic assay of the triterpene mixture using alloxan-induced diabetic mice
Test samples Normal bgl Alloxan-induced bgl Blood glucose levels (bgl) at 30-min intervals (percentage decrease in bgla )
(mmol/l) t = 0 (mmol/l)
t = 30 min t = 60 min t = 90 min t = 120 min t = 150 min

Triterpene mixture 2.9 20.0 15.5 (22.5) 16.1 (19.5) 16.3 (18.5) 15.9 (20.5) 15.8 (21.0)
DMSO 3.0 18.7 18.0 (3.7) 17.9 (4.3) 17.9 (4.3) 18.0 (3.7) 18.0 (3.7)
Glipizide 2.7 23.5 19.5 (17.0) 19.0 (19.1) 19.8 (15.7) 19.5 (17.0) 18.5 (21.2)
Water 3.3 25.9 25.4 (1.9) 25.6 (1.2) 24.6 (5.0) 24.1 (7.7) 23.9 (7.7)
a Percentage of decrease in bgl = (alloxan-induced bgl − sample bgl)/alloxan-induced bgl × 100.

Table 6 also available as 250 mg tablets. They are used in the symp-
Anti-microbial assay of the triterpene mixture using the agar cup method tomatic relief of diarrhea. Our results showed that the major
Organism Test samplesa Average Antimicrobial constituent of Carmona retusa, an intractable triterpene mix-
clearing indexb ture, may relieve diarrhea and the stomach ache associated
zone (mm) with it. The triterpene mixture exhibited potent analgesic and
Escherichia coli Triterpene mixture 0 0 anti-diarrheal activities. It did not exhibit mutagenic activity
Tetracycline 25 1.5
and hence, it is safe to use as a herbal medicine. Hence, re-
Staphylococcus Triterpene mixture 13 0.3 sults of this study validated the popular and safe use of the
aureus Chloramphenicol 25 1.5 plant as an anti-diarrheal.
Candida albicans Triterpene mixture 18 0.8
Clotrimazole 25 1.5
Trichophyton Triterpene mixture 16 0.6 Acknowledgements
mentagrophytes Clotrimazole 40 3.0
a 30 ␮l of 1 ␮g/␮l solution for the positive controls; 200 ␮l of The authors would like to thank the National Research
10,000 ppm for the triterpene mixture. Council of the Philippines (NRCP) for the funding support.
b Antimicrobial index = (diameter of clearing zone − diameter of

well)/diameter of well.

because its variance is statistically similar to that of the
solvent control.
Diabetes was induced by intraperitoneal injection of al-
loxan using a dosage of 150 mg/kg BW of mouse. The
triterpene mixture was tested for its anti-diabetic activity
at a dosage of 100 mg/kg BW of mouse. It proved to be a
more effective anti-diabetic than the known and commer-
cially available glipizide (Table 5) in terms of percentage
decrease in bgl. However, statistical analysis using t-test:
paired two sample for means showed that there was a sig-
nificant decrease for glipizide during the first 30 min at α =
0.10 (tstat : 2.14 > tcrit : 1.64) but there was none for the
triterpene mixture (tstat : 0.56 < tcrit : 1.64).
The triterpene mixture showed a weak to moderate
anti-candidal activity against Candida albicans and anti-
dermatophytic against Trichophyton mentagrophytes, which
causes 90% of chronic dermatophyte infections, but is in-
active against Escherichia coli (Table 6). KW: 11.0 showed
differences in treatments at α = 0.10. At this level of signif-
icance, the triterpene mixture did not show the same activity
as the positive controls, chloramphenicol and clotrimazole.
4. Conclusion
Carmona retusa is used as tea in the Philippines. A de-
coction is usually prepared from the dried leaves. They are
References
Cortes-Maramba, N.P., Dayrit, F.M., de Castro, N.F., Estrada, H.R.,
Lim-Sylianco, C.Y., Lingao, A.L., Quijano, R.F., Quintana, E.G., 1991.
Selection and scientific validation of medicinal plants for primary health
care. Technical Report Series No. 12. Philippine Council for Health
Research and Development, Department of Science and Technology.
de Padua, L., Lugod, G., Pancho, J., 1982. Handbook on Philippine
Medicinal Plants, vol. 1. Documentation and Information Section,
Office of the Director of Research, UPLB, Philippines, p. 21.
Esmerino, L.A., Ranali, J., Rodrigues Jr., A.L., 1998. Blood glucose
determination in normal and alloxan-diabetic rats after administration of
local anesthetics containing vasoconstrictors. Brazilian Dental Journal
9, 33–37.
PCHRD (Philippine Council for Health Research and Development), 1989.
Recommended Bioassay Procedures for Medicinal Plants Products.
Department of Science and Technology, p. 4.
Quisumbing, E., 1978. Medicinal Plants of the Philippines. Katha Pub-
lishing Co., Inc. Philippines, p. 773.
Schmid, W., 1975. The micronucleus test. Mutation Research 31, 9–15.
Villaseñor, I.M., Edu, D.A., Bremner, J.B., 1993. Structure of an antimu-
tagen from Carmona retusa leaves. Carcinogenesis 14, 123–125.
Villaseñor, I.M., Edu, D.A., 1993. Antimutagen from the leaves of Car-
mona retusa (Vahl.) Masam. Mutation Research 298, 215–218.
Villaseñor, I.M., Edu, D.A., Bremner, J.B., 1992. Triterpenoids of Car-
mona retusa (Vahl.) Masam. Philippine Journal of Science 121, 413–
416.
Villaseñor, I.M., Canlas, A.P., Pascua, M.P.I., Sabando, M.N., Soliven,
L.A.P., 2002. Bioactivity studies on Cassia alata Linn. leaf extracts.
Phytotherapy Research 16, 93–96.
Winter, C., Risley, E., Nuss, G., 1962. Carrageenan-induced edema in hind
paw of the rat as an assay for anti-inflammatory drugs. In: Proceedings
of the Society for Experimental Biology and Medicine III, p. 544.