Angio-suppressive triterpenoids from Ardisia cf. elliptica ...€¦ · Further, ardisenone was found to have significant cytotoxic activity in murine cells, and later was identified

128 Chin J Nat Med Mar. 2013 Vol. 11 No. 2 2013 年 3 月第 11 卷第 2 期

Chinese Journal of Natural Medicines 2013, 11(2): 0128−0138doi: 10.3724/SP.J.1009.2013.00128

ChineseJournal of Natural Medicines

Angio-suppressive triterpenoids from Ardisia cf. elliptica (subgenus Tinus) on duck (Anas platyrynchos L.)

chorioallantoic membrane Dennis D. Raga1, 2*, Annabelle A. Herrera2, Consolacion Y. Ragasa3*

1Biology Department, School of Science and Engineering, Ateneo de Manila University, Loyola Heights, Quezon City 1108, Philippines; 2Institute of Biology, University of the Philippines, Diliman, Quezon City, Philippines; 3Chemistry Department and Center for Natural Sciences and Ecological Research, De La Salle University, Manila, 1004, Philippines

Available online Mar. 2013

[ABSTRACT] The dichloromethane extract of the air-dried leaves of Ardisia cf. elliptica (subgenus Tinus) afforded a mixture of bau-erenol (1a), α-amyrin (1b) and β-amyrin (1c). Their structures were identified by NMR spectroscopy. Mixtures of the triterpenes (1a−1c) at ratios of 2 : 2 : 1, 2 : 2 : 3 and 1 : 1 : 1 were tested for their angio-suppressive effects on duck chorioallantoic membrane (CAM). All three ratios were found to be effective in restricting inter-capillary length, while 1a−1c (2 : 2 : 1) was most effective in reducing branch point density with 100% CAM viability and embryo survivability, suggesting a high impact angio-suppressive poten-tial of 1a−1c (2 : 2 : 1). [KEY WORDS] Ardisia cf. elliptica (subgenus Tinus); Myrsinaceae; Bauerenol; α-Amyrin; β-Amyrin; Angio-suppressive

[CLC Number] R965 [Document code] A [Article ID] 1672-3651(2013)02-0128-11

1 Introduction

The genus Ardisia Sw. belongs to the family Myrsina-ceae. Ardisia has 68 recorded species in the Philippines [1], 60 of which are endemic. These are primarily distributed in Mindoro, Polilio, Samar, Panay, Mindanao, Palawan, Leyte, Biliran, Nueva Ecija, Laguna, Bicol, Ilocos Sur, Ilocos Norte, Sambali, Negros Occidental, Negros Oriental, Cebu, Agusan, Pampanga, Batangas, Cagayan and some other areas in Lu-zon. It is commonly known as Tagpo. Ardisia leaves are eaten as a vegetable, used as greens for salads, or cooked with meat or fish [2]. The flowers and fruits may be cooked and used as a flavoring for fish. Young leaves are also eaten by ruminants, while the fruits are eaten by monkeys, wild pigs and birds [2]. Ardisia pyramidalis Pers. fruit approximate analysis [3] re-vealed high fiber content (37.99%), crude protein (13.50%),

[Received on] 21-Feb.-2012 [Research Funding] This project was supported by the grant from the Philippine Council for Health Research and Development (PCHRD) of the Philippines through the Metro Manila Health Re-search and Development Consortium (MMHRDC). [*Corresponding author] Consolacion Y. Ragasa: Tel/Fax: 0632- 5360230, E-mail: [email protected] These authors have no conflict of interest to declare.

crude fat (0.41%), and some minerals such as Ca (0.96%), P (0.21%), K (1.90%), and N (2.16%). Spinasterol, spinasteryl acetate, a mixture of α-amyrin, β-amyrin and bauerenol in a 2 : 1 : 2 ratio, squalene, lutein and triglycerides were isolated from the dichloromethane extract of A. pyramidalis [4].

Recent efforts to elucidate on the bioactitivities of in-digenous Ardisia species have led to discovery of important bioactivities of Philippine indigenous Ardisia species. The presence of a novel compound, ardisenone, isolated from the twigs and leaves of A. iwahigensis Elmer was reported [5]

. Further, ardisenone was found to have significant cytotoxic activity in murine cells, and later was identified as an anti-cancer agent. The angiosuppressive effects of the total methanolic extract of A. pyramidalis leaves on duck chorioallantoic membrane was presented [6]. In a follow-up report [7], the fractionation (hexane and ethyl acetate) and bioassay of the methanolic extract revealed a decrease in potency where the hexane fractions were more effective in suppressing CAM angiogenesis as compared to the ethyl acetate fractions. In addition, recent findings demonstrated the angio-suppressive effects of triterpenes from A. pyrami-dalis [4]. In a related study, the free radical scavenging ability of the ethanolic−water (95%) extract of A. pyramidalis was reported [8]. Similarly, the total methanolic extract of A. squamulosa C. Presl. and its subfractions were found to be

Dennis D. Raga, et al. /Chinese Journal of Natural Medicines 2013, 11(2): 128−138

2013 年 3 月第 11 卷第 2 期 Chin J Nat Med Mar. 2013 Vol. 11 No. 2 129

angiosuppressive as well [9]. Further, CAM treated with the total methanol extract and hexane fractions revealed the sup-pressed expression of factor VIII-related antigen (von Wille-brand factor), CD 32, epithelial membrane antigen, platelet endothelial cell adhesion molecule-1 (CD 31), Cyclin D1 and actin proteins in endothelial cells of CAM vessels. In an-other study [10], the total hexane extract from A. squamulosa was identified to have effects on spermatogenesis with negli-gible toxicity in both systemic and reproductive tissues. De-spite the numerous bioactivities of this genus, there has been very limited information on the chemical composition of indigenous Ardisia species. There is also very limited infor-mation on its biological potentials as sources of phytophar-maceuticals. In the current study, the angio-suppressive ef-fects of mixtures of triterpenes bauerenol, α-amyrin and β-amyrin (1a−1c) from the dichloromethane extract of the leaves of Ardisia cf. elliptica are reported.

2 Experimental

2.1 General 13C NMR spectra were recorded on a Varian VNMRS

spectrometer in CDCl3 at 150 MHz. Column chromatography was performed with silica gel 60 (70−230 mesh); TLC was performed with plastic backed plates coated with silica gel F254; plates were visualized by spraying with vanillin-sulfuric acid and warming. 2.2 Plant material

Fresh leaves of Ardisia cf. elliptica (Plate 1) were col-lected from a chromafic low altitude area along primary thickets (cogon grassland) at the foot of Mt. Pulido, Barangay Pacalat, Mangatarem, Pangasinan, Luzon Island, Philippines (15°44’16.20 N: 120°16’111.27 E). The plant sample genus was identified at the National Museum, Manila by Dr. Do-mingo S. Madulid of the Phillipine National Herbarium Col-lection of the Philippine National Museum, Philippines in collaboration with Dr. John Pipoly of the Urban Horticul-ture/Climate Change Extension, Broward County Extension, Florida, United State of America and Dr. Tim Utteridge of the South_East Asia Team, Royal Botanic Gardens, Kew, Surrey, United Kingdom. Species identification was not determined yet due to the sheer complexity of the genus Ardisia com-posing of several tribes and subgenera thus the name Ardisia cf. elliptica was used in this paper. For the purpose of identi-fication and verification of future results, a morphologic de-scription and photographic reference of Ardisia cf. elliptica is presented.

Ardisia cf. elliptica is a shrubby tree reaching a height of 2−3 m and a stem diameter from 8−10.5 to 15 cm similar to most Ardisia Sw. species. The oblolanceolate leathery leaves are distichious (alternate whorl), (9.36 ± 0.98) cm long and (3.24 ± 0.41) cm wide (l/w ratio: 2.90 ± 0.23), petioles about (8 ± 0.20) mm long clustered along ends of unbranched branches. Six to eight flowers are found clustered terminally in an umbellate inflorescence originating from the leaf axil of

mature leaves of every branch. The flowers buds placed in panicle are purple, which turns to smooth pink with five pet-als and five sepals. The fruits are crustaceous drupe ranging from green to pink to purple in color with purple as the ma-tured fruit and pink as intermediate. The mature fruits are sweet while the pink fruits are sour. 2.3 Extraction and isolation

Air-dried leaves of Ardisia cf. elliptica (440 g) were ground in a blender and then soaked in dichloromethane for three days and filtered. The filtrate was concentrated in vacuo to afford a crude extract (8.5 g) which was chromatographed using increasing proportions of acetone in dichloromethane at 10 % increments. The 30% acetone in DCM fraction was rechromatographed (3 ×) using 5% ethyl acetate in petro-leum ether to afford a mixture of 1a−1c (150 mg, 2 : 2 : 1). This mixture (80 mg) was rechromatographed using 5% ethyl acetate in petroleum ether (2 ×) to afford a mixture of 1a−1c (52 mg, 1 : 1 : 1). The 40% acetone in DCM fraction was rechromatographed in 5% ethyl acetate in petroleum ether (6 ×) to afford a mixture of 1a−1c (95 mg, 2 : 2 : 3). 2.4 Preparation of test substances

Mixtures of triterpenes (bauerenol, α-amyrin, β-amyrin) with ratios of 2 : 2 : 1, 2 : 2 : 3 and 1 : 1 : 1 (1a−1c) obtained from Ardisia cf. elliptica were tested for their anti-angiogenic potential. Appropriate weights of the mixtures were dissolved in 100 μL dimethyl sulfoxide (DMSO) (Ajax Finechem, Aus-tralia) and reconstituted with 890 μL phosphate buffered saline (1× PBS, Gibco) supplemented with 10 μL PennStrep (Gibco) to obtain 6.0μg·μL−1 of 1a−1c at 10% final DMSO concentration. Lower (10 fold difference) concentrations were also prepared to obtain 0.6 μg·μL−1, at 1% final DMSO concentration. A 1% DMSO and 10% DMSO negative con-trol groups were assigned, along with an environmental con-trol (untouched eggs). 2.5 Chorioallantoic membrane (CAM) vascularity assay

Chorioallantoic membrane assay was performed accord-ing to the procedures [4, 7], modified from the earlier proce-dures [11-12], and to determine the angiogenic effects of terpe-noids obtained from Ardisia cf. elliptica. Briefly, fertile mal-lard duck eggs (Anas platyrynchos L.) were obtained from a commercial supplier in Pateros, Metro Manila. Day 0 eggs (n = 6, 2 replicates) were incubated at 37 °C with constant hu-midity at the Institute of Biology, University of the Philip-pines, Diliman Quezon City, Philippines. At the fifth day of incubation, the eggs were candled and inspected for egg vi-ability and position of embryo. On the 7th day, the eggs were wiped with warm 70% ethanol and a small hole using a hand held rotary drill was made at the blunt end (air space) and 50 μL of the proper concentration of each sample was added. The inoculated CAM was sealed using sterile Parafilm and returned to the humidified atmosphere until day 14 after administration of the test samples. Egg viability was monitored every other. On day 14, the eggs were placed on its lateral side to position the CAM and the embryo. The CAM area



Plate 1 Ardisia sp. (A, B) whole plant and arrangement of branches, (C) ventral leaf phylotaxy, (D) flower buds, (E, F) terminal inflorescence and magnified image of the flowers, (G, H) cluster of crustaceous fruit originating from the axil and (I) five sepals

was visually assessed for vascular damage using a Digital Mikroskop Kamera (dnt GMBH, Dietzenbach, Germany) stereomicroscope. Three representative areas or fractal seg-ments were assigned and photo-documented. The CAMs were scored using the CAM scoring guide [12] using an atlas reference as used in a previous report [4]. The frequency of damage was determined through fractal analysis by counting

the number of appearances of the most severe damage ob-served in the three representative areas or fractal segments. Any damage on vasculature and obstruction to normal blood flow was considered a positive anti-angiogenic effect. The CAMs were photographed for the measurement of branching frequency and inter-capillary distance [13]. Branching fre-quency was counted as the number of microvessel branch



points occurring in every capillary segment [13]. Branching point density was determined using the formula:

branch point density=Number of branch point in a blood vessel segment

length of blood vessel segment

Inter-capillary distance was measured in every capillary and microvessel segment using ImageJ 1.40g open access soft-ware [14]. 2.6 Statistical analysis

The results were analyzed using SPSS ver. 13 for Win-dows. One way analysis of variance was performed to deter-mine the significant effects of different ratios of 1a−1c on duck CAM angiogenesis. The results were considered sig-nificant at P ≤ 0.05. Significant differences between group variables were determined by post hoc analysis at 95% Tukey’s test at α = 0.05. Means are reported as mean ± SEM.

3 Structural Identification

Bauerenol (1a): colorless solid. 13C NMR: 36.9 (C-1), 27.7 (C-2), 79.0 (C-3), 38.9 (C-4), 50.4 (C-5), 24.1 (C-6), 116.4 (C-7), 145.2 (C-8), 48.2 (C-9), 35.3 (C-10), 16.9 (C-11), 32.4 (C-12), 37.7 (C-13), 41.5 (C-14), 28.9 (C-15), 37.7 (C-16), 32.0 (C-17), 54.9 (C-18), 35.3 (C-19), 32.0 (C-20), 29.7 (C-21), 31.5 (C-22), 27.5 (C-23), 14.7 (C-24), 13.0 (C-25), 23.7 (C-26), 22.7 (C-27), 40.0 (C-28), 25.6 (C-29), 22.5 (C-30).

α-Amyrin (1b): colorless solid. 13C NMR: 38.8 (C-1),

27.2 (C-2), 79.3 (C-3), 38.8 (C-4), 55.2 (C-5), 18.3 (C-6), 32.9 (C-7), 40.0 (C-8), 47.7 (C-9), 36.9 (C-10), 23.3 (C-11), 124.4 (C-12), 139.6 (C-13), 42.1 (C-14), 28.7 (C-15), 26.6 (C-16), 33.7 (C-17), 59.1 (C-18), 39.6 (C-19), 39.7 (C-20), 31.2 (C-21), 41.5 (C-22), 28.1 (C-23), 15.7 (C-24), 15.6 (C-25), 16.8 (C-26), 23.3 (C-27), 28.1 (C-28), 17.5 (C-29), 21.4 (C-30).

β-Amyrin (1c): colorless solid. 13C NMR: 38.6 (C-1), 27.3 (C-2), 79.0 (C-3), 38.8 (C-4), 54.9 (C-5), 18.4 (C-6), 32.6 (C-7), 38.8 (C-8), 47.7 (C-9), 37.7 (C-10), 23.5 (C-11), 121.7 (C-12), 145.2 (C-13), 41.7 (C-14), 26.1 (C-15), 27.2 (C-16), 32.5 (C-17), 47.6 (C-18), 46.8 (C-19), 31.2 (C-20), 34.7 (C-21), 37.1 (C-22), 28.1 (C-23), 15.6 (C-24), 15.7 (C-25), 16.9 (C-26), 26.1 (C-27), 28.4 (C-28), 33.3 (C-29), 23.7 (C-30).

4 Results

The dichloromethane extracts of the air-dried leaves of Ardisia cf. elliptica afforded the triterpenes 1a-1c by silica gel chromatography. These compounds were identified by comparison of their 13C NMR data with those reported in the literature for bauerenol (1a) [15], α-amyrin (1b) [16], and β-amyrin (1c) [16] (Fig. 1). The ratios of the three triterpenes were determined from the integrations of the olefinic proton resonances at δ 5.39 for bauerenol, δ 5.11 for α-amyrin, and δ 5.16 for β-amyrin.

H

H

HHO

1b

H

H

HHO

1c

H

HHO

1a

H1

47

10

11

14

18 22

23

25

26

27 28

29

30

Fig. 1 Chemical structures of bauerenol (1a), α-amyrin (1b), and β-amyrin (1c) isolated from the dichloromethane extract of Ardisia sp. subgenus Tinus leaves

Table 1 Survivability of embryos exposed to triterpenes isolated from Ardisia sp. subgenus Tinus

Lethality ( f )(Hours post treatment) Treatment

48 h 96 h 144 h 192 h Total mortality

at 216 h (n = 10) Survivability (%)

Environmental control 0 0 0 0 0 100

1% DMSO + PBS 0 0 0 0 0 100

10% DMSO + PBS 0 0 0 0 0 100

0.23 μg·μL−1 1a−1c (2 : 2 : 1) 0 0 0 0 0 100

2.3μg·μL−1 1a−1c (2 : 2 : 1) 0 0 0 0 0 100

0.28μg·μL−1 1a−1c (1 : 1 : 1) 0 0 0 0 0 100

2.8μg·μL−1 1a−1c (1 : 1 : 1) 0 0 0 0 0 100

0.2μg·μL−1 1a−1c (2 : 2 : 3) 0 0 1 2 3 70

2.0μg·μL−1 1a−1c (2 : 2 : 3) 0 0 0 0 0 100



Chorioallantoic membrane treated with either concentra-tions of 1a−1c (2 : 2 : 1), 1a−1c (1 : 1 : 1), and 2.0 μg·μL−1 1a−1c (2 : 2 : 3) had no incidence of mortality, similar with those CAM administered with 1% or 10% DMSO + PBS and environmental control (Table 1). On the other hand, CAMs administered with 0.2 μg·μL−1 1a−1c (2 : 2 : 3) had incurred 30% (3/10) mortality in between 144 h to 192 h. It is possible

that the observed mortality at the lower concentration of 1a−1c (2 : 2 : 3) may not be solely due to treatment administration. The independent and variable response may have been one of the reasons for such, considering that there was no incidence of mortality in those eggs treated with the higher concentra-tion (2.0 μg·μL−1) of 1a−1c (2 : 2 : 3).

Plate 2 Vascular area of CAM exposed to (A) environmental control (untouched eggs), negative controls administered with (B) 1% DMSO + PBS, and (C) 10% DMSO + PBS, and (D) 0.23 μg·μL−1, and (E) 2.3 μg·μL−11a-1c (2 : 2 : 1), (F) 0.28 μg·μL−1, and (G) 2.8 μg·μL−1 1a-1c (1:1:1), (H) 0.2 μg·μL−1, and (I) 2.0 μg·μL−11a−1c (2 : 2 : 3)



The vascular integrity of CAMS administered with 1% (Plate 2B) and 10% (Plate 2C) DMSO + PBS appears to be normal, with no noticeable differences in terms of signs of vascular trauma, such as petechial hemorrhages, blood vessel occlusion, capillary injection (hyperemia) and ghost vessels. Such observations were similar to those from the CAM ob-tained from the environmental control (Plate 2A) setup. The absence of trauma indicates high vascular integrity among those CAMs in the negative and environmental control groups.

Chorioallantoic membranes administered with 1a−1c (2 : 2 : 1) demonstrated minimal signs of vascular assault in the form of petechial hemorrhage (blood droplets covering 25%

of CAM area), hyperemia (signs of increased blood flow or capillary injection), and ghost vessels where capillaries ap-pear devoid of blood flow. The vascular area of those CAMs administered with 0.23 μg·μL−1 1a−1c (2 : 2 : 1) (Plate 3A) had signs of hyperemia (green arrow) and petechial hemor-rhages at its very early (yellow arrow) and intermediate (Plate 3B) stages in those areas where capillary injection seems to be at a higher pressure and velocity. Hyperemia, petechial hemorrhage and intermediate ghost vessels (black arrow) were noted on CAMs administered with 2.3 μg·μL−1 1a−1c (2 : 2 : 1) (Plate 3C-D). Increased blood flow was seen along the main arterial trunk, while small, but numerous, petechial hemorrhages were seen on capillary sprouts.

Plate 3 Vascular area of CAM administered with (A−B) 0.23 μg·μL−1 1a−1c (2 : 2 : 1) and (C-D) 2.3 μg·μL−1 1a−1c (2 : 2 : 1) with indications of blood occlusions forming ghost vessels (black arrow), hyperemia (green arrow), and petechial hemorrhages (yellow arrow)

CAMs administered with 1a−1c (1 : 1 : 1) revealed characteristic signs of vascular damage, such as the formation of ghost vessels and petechial hemorrhage. Preliminary ghost vessels (Plate 4A) were noted on the vascular area of CAM administered with 0.23 μg·μL−1 1a−1c (1 : 1 : 1). Small petechial hemorrhages (Plate 4B) were also seen at the apical region of capillary sprouts. Similar observations were noted on the vascular area of CAM administered with 2.3 μg·μL−1

1a−1c (1 : 1 : 1) (Plate 4C-D). On the average, minimal damage was noted on the vascular area of CAMs adminis-tered with both concentrations of 1a−1c (1 : 1 : 1), which

correlates with the overall CAM viability and embryo sur-vivability in the 1a−1c (1 : 1 : 1) treatment.

Chorioallantoic membrane exposed to 1a−1c (2 : 2 : 3) was observed for signs of vascular trauma. CAMs adminis-tered with 0.20 μg·μL−1 of 1a−1c (2 : 2 : 3) had signs of petechial hemorrhage along the arterial length and apical region of capillary sprouts (Plate 5A&B). Vascular occlusion was also noted, resulting in an early stage ghost vessel (Plate 5B). Numerous blood droplets from apical capillary petechial hemorrhage were also observed on the vascular area of CAMs administered with 2.0 μg·μL−1 (Plate 5C&D).



Plate 4 Vascular area of CAM administered with (A−B) 0.28 μg·μL−1 1a−1c (1 : 1 : 1) and (C−D) 2.8 μg·μL−1 1a−1c (1 : 1 : 1) with indications of blood occlusions forming ghost vessels (black arrow), and petechial hemorrhages (yellow arrow)

Plate 5 Vascular area of CAM administered with (A−B) 0.20 μg·μL−1 1a−1c (2 : 2 : 3) and (C−D) 2.0 μg·μL−1 1a−1c (2 : 2 : 3) with indications of blood occlusions forming ghost vessels (black arrow), and petechial hemorrhages (yellow arrow)



The number of branch points was counted and the inter- capillary length was measured in CAM administered with samples isolated from Ardisia cf. elliptica. The mean in-ter-capillary length in between two branch points of CAM in the environmental control (EC) group (untouched eggs) were observed to be the longest compared to all other treatments [(7.53 ± 0.34) mm]. This also corresponds to a higher branch point density (3.21 ± 0.63), indicating that there was a higher number of branching segments in between two branch points (Table 2). Longer inter-capillary segment and higher branch point density also indicates a more extensive network of blood vessels along the vascular area of CAMs. The branch point density (P = 0.000 1) and inter-capillary length (P = 0.000 1) of blood vessels along the vascular area of CAM administered with the test samples were significantly affected compared to the environmental and negative controls. In-ter-capillary length of CAMs administered with 1% DMSO + PBS [(4.70 ± 0.44) mm] was found significantly (P = 0.000 1) different from that of the EC eggs, but with similar (P = 0.732) effects with that of the 10% DMSO + PBS [(3.85 ± 0.612) mm] treatment. The observed difference in the in-ter-capillary length of vessel segments between the EC and DMSO + PBS groups indicate that either concentration of DMSO + PBS (1% and 10%) had significant effects (P = 0.0001) on inter-capillary distance. Treatment with either 1% DMSO + PBS (P = 0.208) or 10% DMSO + PBS (P = 0.976) on the other hand did not affect the branch point density rela-tive to the EC treatment. Branch point density at either con-centration of DMSO + PBS was also found not to be signifi-cantly different (P = 0.977).

The inter-capillary length of CAMs administered with 0.23 μg·μL−1 1a−1c (2 : 2 : 1) was significantly reduced (P = 0.000 1) to (1.01 ± 0.06) mm, which was found to be similar (P = 1.00) with the mean inter-capillary length of those CAM administered with 2.3μg·μL−1 1a−1c (2 : 2 : 1) [(0.92 ± 0.04) mm) (Table 2). Branch point density of CAM treated with 0.23 μg·μL−1 1a−1c (2 : 2 : 1) was significantly (P = 0.015) reduced to 0.10 ± 0.014 indicating that there was a low inci-dence of branch points occurring in a given blood vessel segment as compared to the 1% DMSO + PBS (1.98 ± 0.24) treated CAM. When the CAM was administered with 2.3μg·μL−1 1a−1c (2 : 2 : 1), the higher concentration sig-nificantly reduced the branch point density to a much greater extent (0.11 ± 0.016) as compared to 10% DMSO + PBS (2.60 ± 0.22). The effects obtained from the lower concentra-tion of 1a−1c (2 : 2 : 1), however, were not significantly dif-ferent (P = 0.100) compared with the effects on branch point density at the higher concentration.

Treatment with 0.28 μg·μL−1 1a−1c (1 : 1 : 1) resulted in a shorter mean inter-capillary length [(0.55 ± 0.04) mm] among the other treatments. Administration with a higher concentration (2.8 μg·μL−1 1a−1c (1 : 1 : 1) resulted in a mean inter-capillary length [(0.76 ± 0.03) mm] comparable (P = 1.00) to that of the lower concentration. The branch

point density of the blood vessels was not significantly (P = 0.811) affected with the administration of 0.28 μg·μL−11a−1c (1 : 1 : 1) obtaining a mean density of 3.210 ± 0.88 as com-pared to that of the mean density of the 1% DMSO + PBS treated CAM. The same effect was observed when the CAM was exposed to a higher concentration (2.8 μg·μL−1) of 1a−1c (1 : 1 : 1). There was no significant (P = 0.969) difference on the vascular branch point density of CAM exposed to 2.8 μg·μL−1 1a−1c (1 : 1 : 1) compared to that of the branch point density of the 10% DMSO + PBS treatment. Administration with 1a−1c (2 : 2 : 1) presents a less pronounced an-gio-suppression by restriction of the inter-capillary length. Such an effect is complimented with the 100% CAM viabil-ity and embryo survivability.

Inter-capillary length of the blood vessels in CAM ad-ministered with 0.20 μg·μL−1 1a−1c (2:2:3) was reduced to (0.92 ± 0.04) mm, which was significantly (P = 0.000 1) different from that of the inter-capillary length of the 1% DMSO+PBS control. The shortening effect of the in-ter-capillary distance observed at the lower concentration was found to have no significant (P = 1.00) difference with that of the shortening effects observed in the 2.0 μg·μL−1 1a−1c (2 : 2 : 3) treatment [(0.93 ± 0.05) mm]. Branch point density of CAMs was not affected by the administration of 0.20 μg·μL−1 1a−1c (2 : 2 : 3). Blood vessel segments obtained a mean inter-capillary length of 1.27 ± 0.66 which was comparable (P = 0.973) to the effects of those CAM treated with 1% DMSO + PBS. When CAMs were exposed to a higher con-centration (2.0 μg·μL−1) of 1a−1c (2 : 2 : 3), a lower branch point density (0.21 ± 0.30) was observed which was signifi-cantly (P = 0.002) different from that of the 10% DMSO + PBS control. However, the angio-suppressive effect observed using the higher concentration of 1a−1c (2 : 2 : 3) is coupled with 30% (3/10) mortality. The results indicate that the effec-tive concentration of 1a−1c (2 : 2 : 3) capable of eliciting angio-suppresive effects is at 2.0 μg·μL−1.

In terms of inter-capillary length, the blood vessel in-ter-capillary length of CAMs administered with all concen-trations of 1a−1c (2 : 2 : 1), 1a−1c (1 : 1 : 1), and 1a−1c (2 : 2 : 3) was shorter compared to that of the environmental or negative controls (1% and 10% DMSO + PBS). Compounds 1a−1c (1 : 1 : 1) had obtained the shortest inter-capillary length compared among other treatments. Post hoc analysis however, reveals no significant (P > 0.05) differences within the effects obtained in other treatments (Fig. 2). This indicates that all samples tested are as effective as each other in terms of shortening the capillary distance between two branch points. The number of branch points originating from a given capillary length was greatly restricted with no significant (P > 0.05) differences from the effects of the vascular area of CAMs administered with both concentra-tions of 1a−1c (2 : 2 : 1) and 2.0 μg·μL−1 1a−1c (2 : 2 : 3). Both concentrations of 1a−1c (1 : 1 : 1) and the lower con-centration of 1a−1c (2 : 2 : 3) were observed to have similar



effects with that of the environmental control and the two negative control groups. The data suggests that 1a−1c (2 : 2 : 1) is angio-suppressive by restricting blood vessel in-ter-capillary elongation and branch point formation similar

to that of the higher concentration of 1a−1c (2 : 2 : 3). The remainder of the substances tested were also an-gio-suppressive, but to a lesser extent by only restricting inter-capillary elongation.

Table 2 Branch point density and inter-capillary length of blood vessels in CAM administered with test samples from Ardisia sp. subgenus Tinus

Treatment Branch point ( f ) Branch point density Inter-capillary length (mm)

Environmental control 6.29 ± 1.09 3.21 ± 0.63 7.53 ± 0.34

1% DMSO + PBS 18.61 ± 2.78 1.98 ± 0.24 4.70 ± 0.44

0.23 μg·μL−1 1a−1c (2 : 2 : 1) 8.00 ± 1.52 0.10 ± 0.014* 1.01 ± 0.06**

0.28 μg·μL−1 1a−1c (1 : 1 : 1) 8.75 ± 2.17 3.10 ± 0.88 0.55 ± 0.04**

0.20 μg·μL−11a−1c (2 : 2 : 3) 9.00 ± 1.52 1.27 ± 0.66 0.92 ± 0.04**

10% DMSO + PBS 15.43 ± 1.34 2.60 ± 0.22 3.85 ± 0.612

2.3 μg·μL−1 1a−1c (2 : 2 : 1) 9.29 ± 0.94 0.11 ± 0.016** 0.92 ± 0.04**

2.8 μg·μL−11a−1c (1 : 1 : 1) 7.38 ± 1.19 1.88 ± 0.43 0.76 ± 0.03**

2.0 μg·μL−1 1a−1c (2 : 2 : 3) 8.36 ± 1.38 0.21 ± 0.30** 0.93 ± 0.05**

*P < 0.05, **P < 0.001 vs respective control

Fig. 2 Inter-capillary distance and branching point density of CAM administered with samples from Ardisia sp. subgenus Tinus. Means followed by the same letter are not significantly different at 5% Tukey’s test

5 Discussion

The different ratios of bauerenol, α-amyrin, and β-amyrin in each of the mixtures of triterpenes isolated from Ardisia cf. elliptica demonstrated potent angio-suppressive activity at varying efficacy. In a previous report from this laboratory, a mixture of bauerenol, α-amyrin and β-amyrin (1a−1c) from the dichloromethane extract of Ardisia pyra-midalis leaves, a congener of Ardisia cf. elliptica were ob-served to have angio-suppressive activity at a ratio of 2:1:2[4]. In the current study, the three ratios of 1a−1c were tested for their angio-suppressive activity on duck chorioallantoic membrane as well. The vascular area was observed with oc-clusion resulting in ghost vessels, hyperemia, and petechial to severe hemorrhages. Such observations were parallel to those

of the effects noted on CAMs administered with different ratios of 1a−1c. The ratio of 2 : 1 : 2 used in the previous study however was coupled with a very high mortality rate at 50 and 100 μg·mL−1, while those ratios in the current study are at 0−30% mortality using a very low sample concentra-tions. It is possible that the ratios 1 : 1 : 1, 2 : 2 : 1, and 2 : 2 : 3 of 1a−1c might also have high mortality if compared to the same concentration used in the previous report. The triter-penes bauerenol (1a), α-amyrin (1b) and β-amyrin (1c) in the 1a−1c mixture were previously noted to have related bioac-tivities, such as anti-nociceptive, anti-inflammatory, 5-lipoxygenase and α-glycosidase inhibitory effects. Another study reported that α-amyrin and β-amyrin mixtures isolated from the resin of Protium kleinii Cuatrec. have anti-nociceptive properties [17-18]. Bauerenol isolated from the



dichloromethane/ methanol extract from the roots of Antho-cleista schweinfurthii Gilg was reported to have α-glucosidase inhibitory activity [19]. Similarly, a mixture of the triterpenes α-amyrin palmitate, α-amyrin palmitoleate, β-amyrin palmitate and β-amyrin palmitoleate in a 13 : 4 : 3: 1 ratio was reported to have antinociceptive and anti-inflammatory effects in mice [20]. Treatment with α- and β-amyrin regulated pain and inflammatory responses in mammalian subjects [21-23]. Amyrins were earlier reported to have inhibitory effects on 5-lipoxygenase activity [24]. Pain or injury induces white blood cells to produce prostaglandins, some endogenous mediators and other lipoxygenase products which stimulate nociceptive neurons that could result in the abdominal responses [25-26]. Non-steroidal anti-inflammatory drugs (NSAIDs) work peripherally in reducing the produc-tion of prostaglandins through the cycloxygenase (COX-1 & 2) enzyme inhibition [27]. NSAIDs also block COX-2 en-zymes that produce prostaglandins specifically for pain and inflammation [27]. Recent findings indicate that cyclooxy-genase-2 (COX-2) acts as a potent inducer of angiogenesis [28- 29], and NSAIDs have been identified to have substantial preclinical evidence demonstrating that NSAIDs and COXIBs have anti-angiogenic properties [30]. In addition to this, the inhibition of VEGF and proinflammatory cytokines is also related to NSAID activity [31-32] suggesting the effective an-giogenic suppression observed on the CAM.

6 Conclusions and Recommendation

Mixtures of triterpenes 1a−1c at different ratios were found to have angio-suppressive effects on duck chorioallan-toic membrane by restricting inter-capillary length and branch point formation. The mixture of 1a−1c at a 2 : 2 : 1 was found to be the most effective in regulating capillary length and branch point density with no incidence of mortal-ity. Species identification of Ardisia cf. elliptica, and an in-vestigation of the specific angiogenesis modulating factors occurring in the CAM vessels, will be useful to confirm the reported angiogenic regulation.

References

[1] Merril ED. An Enumeration of Philippine Flowering Plants [M]. vol 3. Manila Bureau of Printing, 1967: 256-266.

[2] FAO Monograph #32, 1984, Food and fruit-bearing forest species 2: Examples from Southeast Asia. Forest References Development, Forest Resources Division, Forestry Depart-ment.

[3] Catibog C, 1978. In FAO Monograph #32. 1984. Food and fruit-bearing forest species 2: Examples from Southeast Asia. Forest References Development, Forest Resources Division, Forestry Department. 21.

[4] Raga DD, Alimboyouguen AB, Shen CC, et al. Triterpenoids and an anti-angiogenic sterol from Ardisia pyramidalis Cav. Pers [J]. Philipp Agric Scient, 2011a, 94(2): 103-110.

[5] Horgen FD, Guinaudeau H, Pezzuto JM, et al. Isolation and structure elucidation of ardisenone: a new cytotoxic alkylphe-

nol from Ardisia iwahigensis [J]. J Nat Prod, 1997, 60(5): 533-535.

[6] Herrera AA, King REC, Ipulan LADG. Effects of oral admini-stration of crude leaf extracts of Aglaia loheri Blanco and Ardisia pyramidalis (Cav.) Pers on mouse embryo morphology and maternal reproductive performance [J]. J Med Plants Res, 2011, 5(16): 3904-3916.

[7] Herrera AA, Amor EC. Antiangiogenic activity of extracts and fractions from an endemic plant Ardisia pyramidalis (Cav.) Pers. from Bataan, Philippines using duck in ovo chorioallan-toic membrane assay [J]. J Med Plant Res, 2011, 5(6): 2637-2646.

[8] Jacinto SD, Ramos EF, Siguan APT, et al. Determining the antioxidant property of plant extracts: A laboratory exercise [J]. Asian J Biol Ed, 2011, 5: 22-25.

[9] Herrera AA, Ipulan LA, Tameta ADC. Transmission electron micrsocopy and immunohistochemical characterization of the angiosuppressive activity of the crude methanol extract and hexane fraction of Ardisia squamulosa Presl (Myrsinaceae from Bataan, Philippines [J]. Asian Life Sci, 2012, 21(1): 95-105.

[10] Raga DD, Pocsidio GN, Herrera AA. Effects of the oral ad-ministration of nonpolar extract from Ardisia squamulosa Presl (Myrsinaceae) leaves on spermatogenesis in rats [J]. Pharmacog Res, 2011b, 3(4): 255-260.

[11] Leung T, Miller JM, Bilbao KV, et al. The chick chorioallan-toic membrane as a model tissue for surgical retinal research and stimulation[J]. Retina, 2004, 24(3): 427-434.

[12] Burdick JD, Gao Y, Kanengiser B, et al. Comparative assess-ment of two eye area cosmetic formulations through evalua-tion of alternative eye irritation methods relative to endpoints measured in a human clinical subacute study design [J]. The Toxicologist, 2003, 72(S1): 220.

[13] Norby K. Microvascular density in terms of number, and length of microvessel segments, per unit tissue volume in mammalian angiogenesis [J]. Microvasc Res, 1998, 55(1): 43-53.

[14] Perreira T, Rasband W. 2011, Image J. National Institutes of Health, USA).

[15] Cerda-Garcia-Roxas CM, Hernandez-Vidal HH, Joseph- Nathan P. 13C NMR assignments of D: C-friedours-7-ene de-rivatives. Evidence of an abnormal methyl group chemical shift [J]. Magn Res Chem, 1966, 34: 777-781.

[16] Mahato SB, Kundo AP. 13C NMR spectra of pentacyclic triterpenoids – A compilation of some salient features [J]. Phytochemistry, 1994, 37: 1517-1575.

[17] Otuki MF, Ferreira J, Lima FV, et al. Antinociceptive proper-ties of mixture of alpha-amyrin and beta-amyrin triterpenes: evidence for participation of protein kinase C and protein kinase A pathways [J]. J Pharmacol Exp Ther, 2005a, 313(1): 310-318.

[18] Otuki MF, Vieira-Lima F, Malheiros A, et al. Topical anti-inflammatory effects of the ether extract from Protium kleinii and α-amyrin pentacyclic triterpene [J]. Eur J Pharma-col, 2005b, 507(1-3): 253-259.

[19] Mbouangouere RN, Tane P, Ngamga D. A new steroid and α-glucosidase inhibitors from Anthocleista schweinfurthii [J]. Res J Med Plant, 2007, 1(3): 106-111.

[20] Ragasa CY, Puno MRA, Sengson JMAP, et al. Bioactive triterpenes from Diospyros blancoi [J]. Nat Prod Res, 2009, 23(13): 1252-1258.



[21] Oliveira FA, Lima Jr RC, Cordeiro WM, et al. Pentacyclic triterpenoids, α, β-amyrins, suppress the scratching behavior in a mouse model of pruritus [J]. Pharmacol Biochem Behavior, 2004, 78(4): 719-725.

[22] Oliveira FA, Chaves MH, Almeida FRC, et al. Protective effect of α-amyrin, and β-amyrin, a triterpenes mixture from Protium heptaphyllum (Aubl,) March. trunk wood resin against acetaminophen induced liver injury in mice [J]. J Ethnopharmacol, 2005a, 98(1-2): 103-108.

[23] Oliveira FA, Costa CLS, Chaves MH, et al. Attenuation of capsaicin-induced acute and visceral nociceptive pain by α- and β-amyrin, a triterpene mixture isolated from Protium heptaphyllum resin in mice [J]. Life Sci, 2005b, 77: 2942-2952.

[24] Sailer ER, Subramanian LR, Rall B, et al. Acetyl-11-keto- boswellic acid (AKBA): structure requirements for binding and 5-lipoyxgenase inhibitory activity [J]. Br J Pharmacol, 1998, 117: 615-618.

[25] De Souza E, de Lira D, de Quieroz A, et al. The anti-nociceptive and anti-inflammatory activities of caulerpin, a bisin-dole alkaloid isolated from seaweeds of the genus Caulerpa

[J]. Mar Drugs, 2009, 7(4): [26] Smith HS. Arachidonic acid pathways in nociception [J]. J

Supp Oncol, 2006, 4(6): 277-287. [27] Sawynok J. Topical and peripherally acting analgesics [J].

Pharmacol Rev, 2003, 55(1): 1-20. [28] Szekanecz Z, Koch AE. Angiogenesis and its targeting in

rheumatoid arthritis [J]. Vasc Pharmacol, 2009, 51(1): 1-7. [29] Bian XW, Chen JH, Jiang XF, et al. Angiogenesis as an im-

munopharmacologic target in inflammation and cancer [J]. Int Immunopharmacol, 2004, 4(12): 1537-1547.

[30] Monnier Y, Zaric J, Rüegg C. Inhibition of angiogenesis by non-steroidal anti-inflammatory drugs: From the bench to the bedside and back [J]. Current Drug Targets Inflamm Allergy, 2005, 4(1): 31-38.

[31] Sunila ES, Kuttan G. Piper longum inhibits VEGF and proin-flammatory cytokines and tumor-induced angiogenesis in C57BL/6 mice [J]. Int Immunopharmacol, 2006, 6(5): 733-741.

[32] Su SJ, Yeh TM, Chuang WJ, et al. The novel targets for anti-angiogenesis of genistein on human cancer cells [J]. Bio-chem Pharmacol, 2005, 69(2): 307-318.

Documents

Angio-suppressive triterpenoids from Ardisia cf. elliptica ...€¦ · Further, ardisenone was found to have significant cytotoxic activity in murine cells, and later was identified