Potent Phosphodiesterase Inhibition and Nitric Oxide Release Stimulation of Anti-Impotence Thai Medicinal Plants from “MANOSROI III” Database

Seven plants in the top rank were selected from the “MANOSROI III” database using the two Thai keywords which meant impotence and sexual tonic. Boesenbergia rotunda (L.) Mansf. extract [EDP1-001(1)] gave the highest PDE inhibition activity of 4.36-fold sildenafil, a standard anti-impotence drug. Plumbago indica Linn. extract [EDP2-001(1)] exhibited the highest NO release stimulation activity of 666.85% which was 1.50-fold acetylcholine, a standard drug. Most selected plant extracts were nontoxic to EA.hy926 cells at 1.0 mg/mL. EDP1-001(1) exhibited the LD50 value of acute oral toxicity in male ICR mice of over 5,000 mg/kg body weight. EDP1-001(1) also indicated the improvement of sexual behaviors in the paroxetine-induced sexual dysfunction male mice with the evaluation of number of courtships (NC), mount frequency (MF), intromission frequency (IF), and ejaculatory frequency (EF) at 87.67 ± 6.17, 121.00 ± 23.50, 36.00 ± 3.21, and 13.67 ± 2.96 which were 2.63-, 1.27-, 0.53-, and 0.62-fold sildenafil-treated mice at day 14 of the treatments, respectively. The present study has not only confirmed the traditional use of Thai plants for the treatment of ED but also indicated the potential and application of the “MANOSROI III” database for Thai plant selection to be developed as ED food supplements.


Introduction
Erectile dysfunction (ED) or impotence is a sexual disorder of men who fail to accomplish or maintain an adequate erection of penis to allow the pleasing of sexual intercourse. It is regarded as a major health problem of the ageing population and can cause considerable distress, unhappiness, and relationship problems [1]. Phosphodiesterases (PDEs) are classes of enzymes that are capable of cleaving the phosphodiester both in cyclic adenosine monophosphate (cAMP) and in cyclic guanosine monophosphate (cGMP), which are the important intracellular messengers to stimulate the penile smooth muscle relaxation and generate the penile erection. Sildenafil, the first effective oral drug, belongs to the group called selective phosphodiesterase type 5  inhibitors. This drug preserves the cAMP and cGMP levels by inhibiting the action of phosphodiesterase enzymes, thereby enhancing and prolonging the penile erection. However, it may cause moderate to serious side effects, such as headache, facial flushing, dyspepsia, visual disturbance, heart attack, or stroke and even death [2][3][4]. Nitric oxide (NO) released from the autonomic nerve terminals supplying the penis and the endothelium of penile vascular structures is a primary neurotransmitter that plays a crucial role in penile erection. It activates the soluble guanylyl cyclase enzyme to produce cGMP, resulting in the stimulation of penile smooth muscle 2 Evidence-Based Complementary and Alternative Medicine relaxation and erection of the penis. NO signaling pathway is now understood to be a novel and promising therapy for ED [4][5][6].
The Thai medicinal plant recipe database "MANOSROI III" has been developed since 1993 by Professor Dr. Jiradej Manosroi. The recipes have been collected, interpreted, and recorded in this database. This database contains the Thai medicinal plant recipes for pharmaceuticals, cosmetics, and food supplements from all regions of Thailand of over 89,000 recipes with more than 8,300 medicinal plants, as the folklore wisdom treatments of several diseases and symptoms including ED. Many research outcomes have been produced from this database. Several Thai medicinal plant recipes as well as Thai medicinal plants contained in the recipes have long been traditionally used for enhancing sexual performance. In this present study, the two Thai keywords including "Sueam Sa Mat Tha Phap Thang Phet" and "Bam Rung Kam Lang Thang Phet" which meant impotence and sexual energizer, respectively, have been used to search for the target ED plants from this database. The PDE inhibition and NO release stimulation as well as the phytochemicals and the in vivo acute toxicity of the extracts from the selected plants were investigated. In addition, sexual behavior stimulatory effects of the selected extract were evaluated on paroxetine-induced sexual dysfunction male mice. The results from this study can confirm the traditional use for ED which will be beneficial for the further development of the selected plant as a novel drug for ED treatment.

Selection of Plants from the Thai Medicinal Plant
Recipe Database "MANOSROI III". The aphrodisiac medicinal plants were searched from the Thai medicinal plant recipe database "MANOSROI III" using the two Thai keywords of "Sueam Sa Mat Tha Phap Thang Phet" and "Bam Rung Kam Lang Thang Phet" which meant impotence and sexual tonic, respectively. The searched plants were coded with EDP1 and EDP2 for "Sueam Sa Mat Tha Phap Thang Phet" and "Bam Rung Kam Lang Thang Phet," respectively. To select the plants, the plants were ranked by frequencies of appearing in the recipes and coded with 001, 002, 003, and so forth, respectively. The plants which were in the top three from each of the searched keywords with the total of 6 plants were selected for this study.

Plant Materials.
All selected plants were collected from Chiang Mai Province in Thailand during June to December in 2014. The voucher specimens of the collected plants were kept at MANOSÉ Health and Beauty Research Center (http://www.manose.co), Chiang Mai, Thailand. The plant materials were washed, cut into small pieces, dried at 50 ∘ C, and then ground to powder.

Preparation of the Plant Extracts.
The plant powder was weighed at 20 g. The extracts were prepared from each plant according to the methods indicated in the recipes (containing the selected plants) (methods A to D). The obtained extracts were calculated for yields, kept in glass bottles, and stored at 4 ± 2 ∘ C until use. The details of each preparation method indicated in the recipes were as follows.
Method A: The Recipes Were Orally Administered Together with Honey or Molded as Traditional Pills. Briefly, the plant powder was mixed with 300 mL of distilled water. The mixtures were stirred for 60 min and centrifuged at 9,000 in room temperature (27 ± 2 ∘ C) for 5 min. The supernatants were filtered through Whatman6 no. 1 filter paper (GE Healthcare UK Limited, Buckinghamshire, UK). The filtrates were concentrated by a rotary evaporator (BÜCHI Labortechnik AG, Flawil, Switzerland) at 50 ± 2 ∘ C and lyophilized by a freeze dryer (Christ Alpha 1-2 LD, Martin Christ Gefriertrocknungsanlagen GmbH, Osterode am Harz, Germany).

Method B: The Recipes
Were Boiled with Water. Briefly, the plant powder was mixed with 300 mL of distilled water and boiled for 15 min. After cooling, the mixtures were centrifuged at 9,000 in room temperature (27 ± 2 ∘ C) for 5 min. The supernatants were filtered through Whatman no. 1 filter paper (GE Healthcare UK Limited, Buckinghamshire, UK). The filtrates were concentrated by a rotary evaporator (BÜCHI Labortechnik AG, Flawil, Switzerland) at 50 ± 2 ∘ C and lyophilized by a freeze dryer (Christ Alpha 1-2 LD, Martin Christ Gefriertrocknungsanlagen GmbH, Osterode am Harz, Germany).
Method C: The Recipes Were Used as an Infusion. Briefly, the plant powder was mixed with 300 mL of hot distilled water and stirred for 15 min. After cooling, the mixtures were centrifuged at 9,000 in room temperature (27 ± 2 ∘ C) for 5 min. The supernatants were filtered through Whatman no. 1 filter paper (GE Healthcare UK Limited, Buckinghamshire, UK). The filtrates were concentrated by a rotary evaporator (BÜCHI Labortechnik AG, Flawil, Switzerland) at 50 ± 2 ∘ C and lyophilized by a freeze dryer (Christ Alpha 1-2 LD, Martin Christ Gefriertrocknungsanlagen GmbH, Osterode am Harz, Germany).

Method D: The Recipes Were Orally Administered as a Herbal
Liquor. Briefly, the plant powder was mixed with 300 mL of 40% ethanol in distilled water and stirred for 60 min. The mixtures were centrifuged at 9,000 in room temperature Evidence-Based Complementary and Alternative Medicine 3 (27 ± 2 ∘ C) for 5 min. The supernatants were filtered through the Whatman no. 1 filter paper (GE Healthcare UK Limited, Buckinghamshire, UK). The filtrates were concentrated by a rotary evaporator (BÜCHI Labortechnik AG, Flawil, Switzerland) at 50 ± 2 ∘ C and lyophilized by a freeze dryer (Christ Alpha 1-2 LD, Martin Christ Gefriertrocknungsanlagen GmbH, Osterode am Harz, Germany).

PDEs Activity Inhibition
Assay. The PDEs activity inhibition of the selected plant extracts was examined by the modified method [7]. Briefly, 10 L of the plant extract in distilled water at 1.0 mg/mL was mixed with 20 L of 0.5 mM 3 ,5 -cGMP substrate solution, 5 L of PDE assay buffer, 10 L of 5 kU/ L 5 -nucleotidase solution, and 5 L of 4 U/mL PDE solution in each well of the 96-well half-area microplates (Greiner Bio-One GmbH, Frickenhausen, Germany). The mixtures were incubated at 37 ± 1 ∘ C for 60 min. The reaction was stopped by adding 100 L of the green assay reagent and the mixtures were left to settle at room temperature (27 ± 2 ∘ C) for 20 min. The absorbance of the mixtures was measured at 620 nm by a microplate reader (Bio-Rad Laboratories, California, USA). Since the PDEs activity assay was evaluated by measuring the phosphate amounts from the reaction, the amounts of the phosphate existing in the extracts were determined by using the control system of mixing 10 L of the extract solution with 40 L of distilled water in each well of the 96-well half-area microplates. The mixtures were incubated at 37 ± 1 ∘ C for 60 min and added with 100 L of the green assay reagent. After keeping at room temperature (27 ± 2 ∘ C) for 20 min, the absorbance of the mixtures was measured at 620 nm by a microplate reader. The inhibition of PDE activity was calculated by the following equation: % inhibition of PDEs activity where AR water and AR sample were the absorbance at 620 nm of the mixtures obtained from the reaction of PDEs with distilled water and the plant extracts, respectively; AC sample was the absorbance at 620 nm of the control system to determine the phosphate contents existing in the plant extracts.
Each experiment was carried out in triplicate. To determine the concentrations of the extracts giving 50% inhibition of the PDEs activity (IC 50 ), the concentration-activity curve was prepared by varying the concentrations of the extracts from 2 × 10 −5 to 2 × 10 −1 mg/mL. Distilled water and sildenafil citrate dissolved in distilled water were used instead of the extracts for the negative and positive controls, respectively. Folds of the PDEs activity inhibition in comparison to sildenafil citrate were calculated as the following equation: Folds of the PDEs inhibition activity = IC 50 of sildenafil citrate IC 50 of the samples .
where 540 was the absorbance at 540 nm and the control was the untreated cells.

Sample Treatment and Nitric Oxide Measurement.
The nitrite concentration was measured using the Griess reaction as described by Green et al. [8]. The EA.hy926 cells (1 × 10 5 cells/well) were plated in each well of 96-well microplate and incubated at 37 ∘ C for 24 h in a humidified 5% CO 2 incubator (Contherm Scientific Ltd., Petone, New Zealand). The cultured cells were treated with 100 L of the plant extract, acetylcholine chloride (a positive control), and L-NAME (a negative control) at 1.0 mg/mL in DMEM-no phenol red medium. After incubation at 37 ∘ C for 24 h, 100 L of the cultured cell supernatants was mixed with an equal volume of the Griess reagent and incubated at room temperature (25 ∘ C) for 10 min. The absorbance was measured at 540 nm using a microplate reader (Bio-Rad Laboratories, California, USA). The NO production of the EA.hy926 cells was calculated from the obtained absorbance by comparing with the standard concentration curve of sodium nitrite (NaNO 2 ) at 1-200 M in the DMEM-no phenol red medium. Each experiment was carried out in triplicate. To determine the concentrations of the extracts giving 50% stimulation of the NO production activity (SC 50 ), the concentration-activity curve was prepared by varying the extract concentrations from 1 × 10 −4 to 1.0 mg/mL. Folds of the NO release stimulation of the extract in comparing to acetylcholine chloride were calculated as the following equation: Folds of the NO release stimulation activity = SC 50 of acetylcholine chloride SC 50 of the samples .  [11]. The male mice were weighed and randomly divided into two groups (5 mice per group). The animals were orally given using an oral feeding tube the vehicle (distilled water) as a control group in Group I and in Group II a single dose of the selected plant extract at 5,000 mg/kg body weight. Mortality and toxicity apparent signs or behavioral alterations of mice were observed daily for 14 days. The body weight changes of the mice were checked at day 14. At the end of the experiment, each mouse was euthanized and the organs (heart, lung, liver, kidney, spleen, stomach, intestine, and testis) were collected for histopathology. The organs were isolated and preserved in 10% formalin in phosphate buffered saline (PBS) at pH 7.4. Each organ was weighed and the abnormal-weight organs were embedded in paraffin, cut into 6 and 12 m thickness, stained with hematoxylin and eosin, and examined under a light microscope (Olympus, Tokyo, Japan).

Evaluation of Sexual Behavior Stimulatory Effects
2.9.1. Animals. Healthy eight-week-old ICR male and female mice weighing 35-40 g and 30-35 g, respectively, were obtained from the National Laboratory Animal Center, Nakhon Pathom, Thailand. The mice were maintained under the standard condition for laboratory animal housing controlled temperature at 24 ± 1 ∘ C and humidity at 50 ± 10% with the cycle of 12 h light and 12 h dark. The mice were fed with commercially available animal foods and allowed water ad libitum.

Induction of Estrous Period to the Female Mice.
The female mice were induced to be in the estrous period by the procedure modified from the method of Aslam and Sial [12]. Briefly, the female mice were orally given 17ethinylestradiol at 2 mg/kg body weight and subcutaneously injected with progesterone at 20 mg/kg body weight, 48 h and 6 h, respectively, before exposure to the male mice.
The estrous period of the female mice was confirmed by vaginal smear examinations under microscopic observation [13].

Induction of Sexual Dysfunction or ED (Erectile Dysfunction) of the Male Mice.
The male mice were induced to be in the state of sexual dysfunction by daily oral administration for 21 days of paroxetine suspension at 20 mg/kg body weight using a metal feeding tube [14]. The paroxetine suspension was prepared daily in Tween-80 suspended in normal saline solution. After 21 days, an estrous female mouse was introduced into each paroxetine-treated male mouse in a separate cage, and the sexual behaviors including numbers of courtship (NC), mount frequency (MF), intromission frequency (IF), and ejaculatory frequency (EF) were observed via a video record for 12 h between 6.00 p.m. and 6.00 a.m. The male mice which showed the minimum of 25% reduction in NC, MF, IF, and EF when compared with the normal male mice were considered as sexually dysfunctional and were used for the further study. The observing staffs were free from treatment information and have evaluated the male sexual behaviors according to the criteria of the previous studies [15,16] as follows: (i) Courtship: it is the amount of frequency it takes for a male to mount the female for the first time.
(ii) Mounting the female: the male normally mounts from the rear, sometimes posing his forelegs over the female's back, and rapidly makes.
(iii) Intromission vaginal penetration: this behavior starts with a mount, but suddenly the male makes a deep thrust forward and stops pelvic thrusting.
(iv) Ejaculation: this behavior starts with an intromission, but, after vaginal penetration (the deep forward thrust), the male remains on the female for 1-3 s.
Evidence-Based Complementary and Alternative Medicine 5 2.9.4. Animal Grouping and Extract Administration. The paroxetine-induced sexual dysfunction male mice were randomly divided into three groups with 3 animals in each group. 300 l of distilled water, 0.02 g/kg body weight of sildenafil (a standard PDE-5 inhibitor drug), and 1 g/kg body weight of the plant extract were orally administered to each animal group using a metal oral feeding tube, once daily for 14 days. After 30 mins of administration, an estrous female mouse was introduced into each sample-treated male mouse in a separate cage and the sexual behaviors of the male mice were monitored at days 1, 7, and 14. The sexual behaviors including NC, MF, IF, and EF were observed via a video record for 12 h between 6.00 p.m. and 6.00 a.m.

Statistical Analysis.
All assays were performed in triplicate of independent experiments. The data were presented as means ± standard deviation (SD). Statistical evaluation was determined by the one-way analysis of variance (oneway ANOVA) with post hoc test of multiple comparisons. Statistical significance was considered at the values of less than 0.01 and 0.05.

Selection of Plants from the Thai Medicinal Plant Recipe
Database "MANOSROI III". Seven instead of six medicinal plants in the top three ranks searching by the two Thai keywords ("Sueam Sa Mat Tha Phap Thang Phet" and "Bam Rung Kam Lang Thang Phet") were obtained and used to prepare the extracts according to the details shown in Table 1, since there were two medicinal plants that gave the same frequency in the selected aphrodisiac recipes.

Preparation of the Extracts.
The extracts of the seven selected medicinal plants were prepared according to the traditional methods (methods A-D) from the recipes containing these plants. Fifteen extracts were obtained because some selected plants were from the recipes which indicated more than one traditional preparation method. The percentage yields and physical appearance of the extracts were demonstrated in

PDEs Activity Inhibition of the Selected Plant Extracts.
The dose of 15 extracts at 200 g/mL which was calculated from the traditional dose indicated in the recipes was tested for PDEs inhibition compared with sildenafil, a positive control PDE-5 inhibitor drug. The dose of sildenafil at 0.45 g/mL was determined from the maximum plasma concentration of a single oral dose of tablet containing 100 mg sildenafil in healthy male volunteers [17]. The IC 50 values and PDEs inhibition percentages of the extracts and sildenafil were shown in Table 2. All extracts [EDP1-001(1), EDP1-001(2), EDP1-002(2), EDP1-003, EDP2-001(1), EDP2-002(2), and EDP2-003(2)] exhibited the PDEs inhibition percentages in the range of 2.72-65.28% of 0.18-4.36-fold sildenafil. Sildenafil provided quite low inhibitory activity to PDEs because the dose of sildenafil (0.45 g/mL) used in the test was lower than the extracts (200 g/mL). The IC 50 value of sildenafil was higher than EDP1-001(1). Also, sildenafil is a selective inhibitor of phosphodiesterase (PDE) type 5, but the nonselective phosphodiesterases (PDEs) assay was performed in this study. So, the different potency of PDEs inhibition activity between EDP1-001(1) and the standard sildenafil may be due to the different inhibition mechanism on PDEs. The mechanism of sildenafil is the potential endogenous increases in cGMP by inhibiting its breakdown at the catalytic site [18]. Orallo et al. [19] showed that (±) naringenin as a natural flavonoid characterized as an agent with clear vasorelaxant effects on rat aortic smooth muscle is probably mediated by an increase in systolic cAMP and cGMP concentrations. However, some plant extracts showed PDEs stimulation effect (with the minus PDEs stimulation value). This may be due to affinity to noncatalytic CGMP blinding sites on the PDEs [20]. The highest PDEs inhibition percentage was observed in EDP1-001(1) extract at 65.28 ± 2.09% of 4.36-fold sildenafil. This extract gave the IC 50 value of PDEs inhibition activity at 0.0026 ± 0.0025 mg/mL of 2.23-fold sildenafil (the IC 50 value of sildenafil was 0.0058 ± 0.0032 mg/mL). This extract was prepared by method A using aqueous cold extraction of Boesenbergia rotunda (L.) Mansf. rhizome ( Table 1). As previously reported, the ethanolic extract of this plant at 0.1 mg/mL showed complete inhibitory effect (100% inhibition) against PDEs [21]. This ethanolic extract at 50 g/mL also exhibited the selective inhibition against PDE-5 activity of 40.86 ± 3.94% [22]. In addition, this present study has also demonstrated high PDEs inhibition activity of EDP1-002(2), EDP2-003(2), EDP1-003, and EDP0-001 (2)    have long been traditionally used in humans. In fact, this long traditional use in humans is similar to the clinical trial step in the research and development of the modern drugs [6].

Phytochemical Determination of the Selected Plant
Extracts. The phytochemical contents in the selected plant extracts were presented in Table 3. High contents of flavonoids and saponins were found in EDP1-001(1), whereas alkaloids and xanthones (phenolic compounds) were observed in EDP1-003. Bioactive compounds from rhizome extracts of B. rotunda have been classified mainly into two major groups including flavanones (alpinetin, pinostrobin, and pinocembrin) and chalcones (boesenbergin, cardamonin, panduratin A, and 4-hydroxypanduratin A) [30]. These compounds have estrogenic or androgenic activities and show phosphodiesterase inhibition activity [31]. Some plants have demonstrated PDEI activity related to their saponin contents. Mimaki et al. [32] reported that Lilium regale and Lilium henryi have inhibitory effects on cAMP PDE. The main active substances presented in these plants are steroidal saponins. Kuroda et al. [33] showed that ethanol extract of Allium chinense has inhibitory activity on cAMP PDE, probably because of its saponin contents. These major bioactive compounds can be used as markers for phytochemical profile standardization. In this study, the HPLC fingerprint of the EDP1-001(1) extract using flavone as a marker was performed for quality control, with the flavone content of 58 g/mg extract ( Figure 2). Both flavonoids and phenolic compounds have been reported to be related to NO production stimulation [27][28][29], with the possible mechanism of increasing mRNA and protein expression of endothelial NO synthase (eNOS) in rats [34]. NO might be stimulated and released directly in the endothelial cells by activating the eNOS. Upadhyaya et al. have reported that flavonoids in Shixiao San increased the expression of eNOS [35]. In addition, an increase in the eNOS mRNA expression of the endothelial cells could be observed after being incubated with red wine polyphenol extract [29]. Thus, flavonoid contents in EDP1-001(1) appeared to be responsible for the NO release stimulation activity.
3.6. Acute Oral Toxicity. The EDP1-001(1) extract was further investigated for acute oral toxicity in male mice because it gave the highest activity of PDEs inhibition and high NO release stimulation activity with no cytotoxicity in the EA.hy926 cells (Table 4). Mortality, abnormal animal behaviors, toxic signs, and body weight changes compared to the control group of the male mice fed with a single dose of 5,000 mg/kg body weight were not observed. Hence, the median lethal dose (LD 50 ) of the EDP1-001(1) extract was more than 5,000 mg/kg body weight. In addition, no macroscopic change of the internal organs of all treated mice was indicated. Most internal organ weights of the extract-treated mice were not different from the control group, except the liver and intestine weights which increased significantly ( < 0.05) when compared with the control group. However, the histopathological examination of these two organs was not different from the control group. Therefore, the B. rotunda extract [EDP1-001(1)] was safe and appeared to have potential for the further development as a product.

Sexual Behavior Stimulatory
Effects. Sildenafil at 0.02 g/ kg body weight was calculated from the normal dose of sildenafil in human (100 mg/60 kg body weight) and adjusted by the dose conversion factor to mice [36]. The selected plant extract, EDP1-001(1), at 1 g/kg body weight was obtained from the traditional indication in the recipe and also adjusted by the conversion factor from human to mice. The sexual behaviors including number of courtships (NC), mount frequency (MF), intromission frequency (IF), and ejaculatory frequency (EF) of the paroxetine-induced sexual dysfunction male mice treated with the EDP1-001(1) extract compared to the control groups were shown in Tables 5 and 6. The paroxetine-induced sexual dysfunction male mice that were orally given daily the EDP1-001(1) extract for 7 and 14 days exhibited sexual behavior improvement when compared with the mice treated with distilled water (the control group). At day 7 of the treatment, the NC and MF of the EDP1-001(1)treated mice were at 26.33 ± 3.48 and 17.00 ± 12.34 which were 1.44-and 17.00-fold the distilled water-treated mice (the control group), respectively. At day 14 of the treatment, the NC, MF, IF, and EF of the EDP1-001(1)-treated mice were at 87.67±6.17, 121.00±23.50, 36.00±3.21, and 13.67±2.96 which were 2.63-, 1.27-, 0.53-, and 0.62-fold the sildenafil-treated mice (the positive control group), respectively. The EDP1-001(1) appeared to be a potential extract for the treatment of sexual dysfunction. However, the extract needed to be orally administered continuously for 7-14 days to obtain an improvement of sexual dysfunction symptoms, whereas sildenafil exhibited the improvement particularly at the initial treatment of day 1. The higher sexual behavior stimulatory effects of the EDP1-001(1) extract at 1 g/kg body weight than sildenafil at 0.02 g/kg body weight might be due to the regulatory mechanism of male reproductive system that are not related only to the inhibition of PDEs activity, but also to the influence on androgens such as testosterone and the release of NO. This in vivo sexual dysfunction improvement of the EDP1-001(1) extract has confirmed the in vitro results of both the PDEs inhibition and the NO release stimulation with the IC 50 value of 0.0026 ± 0.0025 mg/mL (2.23-fold sildenafil) and the SC 50 value of 3.96 ± 5.57 mg/mL (0.002fold acetylcholine), respectively.

Conclusions
Several Thai medicinal plants appeared in the Thai medicinal plant recipe database "MANOSROI III" have long been traditionally used to treat various diseases and symptoms including ED. Seven aphrodisiac medicinal plants selected from the top three ranks searched from the "MANOSRO III" database by the 2 Thai keywords ("Sueam Sa Mat Tha Phap Thang Phet" and "Bam Rung Kam Lang Thang Phet") were prepared as 15 extracts. Only 7 extracts at 200 g/mL exhibited the PDEs inhibition in the range of 2.72-65.28% which were 0.18-4.36-fold sildenafil at 0.45 g/mL, while 10 out of 15 extracts at 1.0 mg/mL showed the NO release stimulation in the range of 33.53-666.85% in the EA.hy926 cells which were 0.08-1.50-fold acetylcholine at 1 mg/mL. The EDP1-001(1) extract prepared from B. rotunda by method A gave the highest PDEs inhibition activity with the IC 50 value of 0.0026 ± 0.0025 mg/mL and the high NO release stimulation with the SC 50 value of 3.96 ± 5.57 mg/mL. The EDP1-001(1) extract contained the high contents of flavonoids and saponins in which flavonoids have been reported to be related to NO production stimulation. The EDP1-001(1) was nontoxic to EA.hy926 cells at 1.0 mg/mL. The EDP1-001(1) extract showed no acute toxicity with no mortality, no abnormal animal behaviors, no toxic signs, and no body weight changes after oral administration in male ICR mice at a single dose of 5,000 mg/kg body weight. This extract also exhibited the sexual behavior improvement in the paroxetine-induced sexual dysfunction male mice after oral daily administration for 14 days when compared with the control group. At day 14 of the treatment, the NC, MF, IF, and EF of the EDP1-001(1)-treated mice were at 87.67 ± 6.17, 121.00±23.50, 36.00±3.21, and 13.67±2.96 which were 2.63-, 1.27-, 0.53-, and 0.62-fold the sildenafil-treated mice (positive control group), respectively. The results from this study have indicated not only the potential of the selected Thai plant extract for the further development as food supplements for ED but also the benefit of the Thai medicinal plant recipe database "MANOSROI III" as a convenient source for an efficient searching and ranking of the target plants for the treatment of various symptoms including ED.

Conflicts of Interest
The authors report no conflicts of interest.