Health‐promoting bioactivity and in vivo genotoxicity evaluation of a hemiepiphyte fig, Ficus dubia

Abstract Ficus species have been used as a typical component in food and folk medicine in Asia for centuries. However, little is known regarding the bioactivity and genotoxicity of the recently identified Ficus dubia (FD), an indigenous plant of the tropical evergreen rain forest. FD is unique from other Ficus species because of its highly sought‐after red‐brown latex. Antioxidant properties together with phenolic and flavonoid contents of FD were elucidated. Health‐promoting characteristics were examined by studying the inhibition of enzymes as a drug target for diabetes, hypertension, Alzheimer's disease, and obesity, together with anticancer ability against human colorectal adenocarcinoma, human hepatocellular carcinoma, human ovarian carcinoma, human prostate adenocarcinoma, and human lung carcinoma. Besides, FD genotoxicity was tested using the Drosophila wing spot test. Results showed that both FD root and latex exhibited antioxidant activity due to the presence of phenolics and flavonoids, specifically caffeic acid and cyanidin. The ethanolic fraction of FD root demonstrated a potent antidiabetic mechanism underlying α‐glucosidase inhibitory activity similar to acarbose. This fraction also suppressed lung and ovarian cancer growth, possibly by G1 and G2/M arrest, respectively. All tested fractions lacked mutagenicity in vivo. Results indicated that FD can be developed as novel antidiabetic compounds; however, its bioactive compounds should be further identified.

impairment, reduced ability in daily living, and death. To date, a diverse range of medicines have been used to treat NCDs and NRDs; however, several drugs have been reported to have severe side effects and poor efficacy. Thus, the quest for natural substances having potential to treat NCDs or NRDs with minimal or no side effects is warranted. worldwide. Some 500 Ficus species have been recorded in Asian-Australasian tropical region, with growth habits varying from free-standing woody trees to small shrubs, vines, lithophytes, epiphytes, hemiepiphytes, and rheophytes (Rønsted et al., 2008;Woodland, 2009). Leaves, bark, roots, latex, and fruits of Ficus spp.
have all been utilized as folk medicine in Thai and Chinese cultures and Ayurveda for centuries (Lansky et al., 2008;Nutmakul et al., 2016) and consumed as food. The fruits of Ficus carica triterpenoids, anthocyanins, and flavan-3-ol monomers (Wojdyło et al., 2016). Other Ficus species have also been used as medicine for a wide range of ailments covering the endocrine system, gastrointestinal tract, central nervous system, inflammation, and cancer. An ethanolic extract of Ficus racemosa inner bark, stem milky sap, and frozen fruit, and stem of Ficus tikoua inhibited cyclooxygenase-1 (COX-1) activity, resulting in poor biosynthesis of inflammatory mediators, prostaglandin E2 (PGE 2 ), and prostaglandin D2 (PGD 2 ) (Li et al., 2003). Isolated 6-O-acylβ-D-glucosylβ-s itosterols (6-AGS) from F. carica latex showed anticancer activities against many cancer cells such as Burkitt B-cell lymphoma  and T-cell leukemia (Jurkat cells) (Rubnov et al., 2001). The fruit of F. carica and bark of F. benghalensis have been reported for their potent antidiabetic properties by inhibiting α-glucosidase enzyme in vitro (Ahmed et al., 2011;Wojdyło et al., 2016), leading to reduced glucose absorption in the intestinal tract. Ficus spp. contain numerous health-promoting substances including flavonoids, alkaloids, triterpenoids, tannins, β-sitosterols, and phenolic acids (Lansky et al., 2008). These phytochemicals have been intensively studied for their aptitude to prevent or treat diseases, particularly as anticancer therapy and illness related to metabolic syndromes (Hu et al., 2012;Shahidi & Ambigaipalan, 2015).
Ficus dubia (FD), a hemiepiphytic Ficus, is an indigenous plant that grows in the tropical evergreen rain forests of Southern Thailand, Malaysia, and Brunei (Chantarasuwan & Thong-Aree, 2006). The tree can reach up to 30-35 m with auxiliary aerial roots lengthening to the ground from branches or the trunk. The latex within the wood has been used to prevent infection or healing. Intriguingly, the unique characteristic of FD is its red-brown latex after exposure to the air. Local people believe that this striking latex color may offer health benefits through anticancer and antidiabetic properties.
Although there is no scientific information supporting its ethnopharmacological properties and safety, the price of FD latex is around 100-200 US dollars/kg. This is the first report to investigate the bioactivity, health benefits, and genotoxicity of FD root and latex.
Antioxidant properties were determined, together with phenolic acid and flavonoid profiles of FD root and latex. Health benefits were studied using inhibitory activities against crucial enzymes involved in NCDs and NRDs including α-glucosidase, angiotensinconverting enzyme, acetylcholinesterase, and lipase as drug targets for diabetes, hypertension, Alzheimer's disease, and obesity, respectively. Anticancer activities against five human cancer cell lines as colorectal adenocarcinoma (SW620), hepatocellular carcinoma (HepG2), ovarian carcinoma (SKOV-3), prostate adenocarcinoma (PC3), and lung carcinoma (A549) were also investigated. A genotoxicity assessment of FD was also performed using the in vivo wing spot test (SMART) in Drosophila.

| Plant material and collection
Roots and latex of F. dubia were collected from Narathiwat Province, the Southern province of Thailand. The plant was collected and identified by Dr. Bhanumas Chanatarasuwan (Taxonomist), and a voucher specimen (Chantarasuwan 040117-1) was deposited at Thailand Natural History Museum, National Science Museum, Thailand.

| Root preparation and extraction
Ficus dubia roots were prepared according to (Jin et al., 2015) with slight modification. In brief, roots were washed with tap water twice, followed by deionized water. They were then cut into small pieces with a thickness of 1.5-2 cm and dried at 40°C for ten days. The dried samples were finely powdered with an electric mill and kept in a desiccator until required for analysis.
Normal extraction: Ten grams of finely powdered roots were mixed with 100 ml of deionized water (RW30) or 80% ethanol (REt30). Then, mixtures were shaken at 30°C for 12 hr. The supernatants were collected by centrifugation at 939 g for 10 min, and the residues were subsequently re-extracted using the same solvent and temperature. A rotatory evaporator was employed to remove the solvent.
Hot extraction: The sample was extracted following the same method as normal extraction, except that only deionized water was used, and the extraction temperature was set at 80°C (RW80).

| Latex preparation and extraction
Ficus dubia latex was freeze-dried. Then, dried latex powder was redissolved in deionized water at 1 mg/ml. The solution was filtered through Whatman ® No.1 paper several times to remove any latex residuals and then freeze-dried again. The dried latex sample was stored at −20°C until required for analysis.

| Cytotoxicity analysis
The MTT assay was used to determine cellular toxicity against normal cells, or human cancer cell lines as formerly described (Pitchakarn et al., 2017). Colorectal adenocarcinoma (SW620), hepatocellular carcinoma (HepG2), ovarian carcinoma (SKOV-3), prostate adenocarcinoma (PC3), and lung carcinoma (A549) were used and all cell lines were purchased from the American Type Culture Collection (ATCC). All cancer cells were exposed to various concentrations of RW30, REt30, RW80, and latex for 48 hr at 37°C in a 5% CO 2 humidified atmosphere before analysis. At the indicated time, MTT dye was added. Formazan production was determined by spectrophotometry at 540/630 nm. To determine the selectivity index (SI) of REt30, mouse embryonic fibroblast (3T3-L1) was used to represent normal cells. The SI index was calculated as the IC 50 ratio of REt30 on 3T3-L1 to each cancer cell line at the same duration time of treatment. An SI value ≥2 indicated that the cytotoxic REt30 was selective against cancer cells (Badisa et al., 2006).

| Cell cycle analysis by flow cytometry
A549 and SKOV3 cells were treated with increasing concentrations of REt30 for 48 hr, and then, cell suspensions were prepared and stained with propidium iodide (Guava ® cell cycle reagent; Guava Technologies) according to the Guava ® Cell Cycle Assay protocol.
Cell cycle phase distributions were determined on a Guava ® PCA Instrument using CytoSoft Software. Cisplatin at IC 50 (14 and 10 µM for A549 and SKOV3, respectively) was used as a positive control.

| In vivo genotoxicity analysis in Drosophila
To determine the mutagenic potential of RW30, REt30, RW80, and latex, the somatic mutation and recombination test (SMART) were Wings from flies expressing round wings were cut and mounted with Faure's solution. The phenotypes and statistical analysis were performed as previously explained (Frei & Wurgler, 1988;Graf et al., 1984).

| Antioxidant activity
Oxidants contribute to degenerative diseases including diabetes, cancer, and premature aging (Gupta et al., 2014). Enhancing the antioxidant capacity in the body can prevent these clinical disorders.
Ficus dubia was first determined for its antioxidant activity through the two mechanisms of hydrogen atom transfer (HAT) and single electron transfer (SET). The HAT mechanism can be measured by ABTS and ORAC assays, while the SET reaction can be measured by DPPH and FRAP assays (Liang & Kitts, 2014). As shown in Table 1, among all extracts, REt30 exhibited the highest antioxidant activities determined by DPPH and FRAP assays, indicating oxidant reduction via SET reaction. Ethanol reduced the polarity index and enhanced extraction conditions of FD root by enriching bioactive compounds.
A previous study found that the SC 50 value determined by DPPH radical scavenging assay of methanolic extracted F. benjamina root grown in Pakistan was 58.81 ± 4.50 μg/ml or 4.3-fold greater than REt30 (Imran et al., 2014), indicating that species, growth location, and extraction methods all contributed to the result. A contradiction was shown between ABTS and ORAC of latex (Table 1); however, FD latex quenched oxygen radicals better than the root (3-to 5-fold).
The HAT reaction mechanism may be responsible for reducing oxygen radicals. The ORAC assay was proven to be more relevant to the body as it uses a biological radical source (Prior et al., 2003).

| Phytochemical analysis
The presence of phytochemicals such as phenolic and flavonoid compounds may contribute to antioxidant capacity (Pandey & Rizvi, 2009). Therefore, we further investigated total phenolic (TPCs) and total flavonoid contents (TFCs) of FD. Results in Table 2 show that FD root was rich in both phenolic and flavonoid compounds, especially in REt30. These results were consistent, suggesting that high antioxidant activity of REt30 following DPPH and FRAP may associate with high amounts of TPCs and TFCs (Tables 1 and 2) (Imran et al., 2014). Remarkably, the latex fraction showed the highest amount of TPCs at 1-to 2-fold higher than the root extract, whereas TFC level was the lowest at 3-and 4-fold lower than root extracts, implying that oxygen radical absorbance capacity (ORAC) in latex may primarily result from phenolic compounds. Furthermore, TPCs and TFCs of F. religiosa latex were reported as 2.76 ± 0.84 mg GAE/g extract and 1.84 mg QE/g extract, respectively (Yadav, 2015). We then hypothesized that the red-brown pigments presented in FD latex might belong to groups of phenolic and flavonoid compounds.
Moreover, considering the TFCs in RW80, some flavonoids presented in the root may be heat-labile types.
HPLC analysis was applied for further investigation of phenolic acids and flavonoids in FD using twenty-one phenolic and flavonoid standards. First, undigested FD extracts were subjected for analysis; however, no phenolic acids or flavonoids were identified (data not shown). Thereafter, prior to HPLC analysis, all FD extracts were subjected to acid hydrolysis to remove sugar groups. The chromatograms are shown in Figure S1. From these chromatograms, only two compounds were identified and quantified. FD root contained both caffeic acid and cyanidin as phenolic acid and flavonoid, respectively (Table 3). The REt30 fraction had the highest amount of caffeic acid and cyanidin that correlated with these results (Table 2).
Intriguingly, F. benjamina root contained chlorogenic acid, syringic acid, p-coumaric acid, and ferulic acid but not caffeic acid (Imran et al., 2014). group, contribute to red and purple colors in fruits. Our data indicated that the red-brown color of FD latex may not originate from cyanidin.

| Enzyme inhibitory properties
Ficus species have been recognized for their value as rich sources of phytochemicals with therapeutic potential. To confirm this, FD extracts were tested for their beneficial health properties against key enzymes involved in NCDs and NRDs including α-glucosidase, ACE, AChE, and lipase as drug targets for diabetes, hypertension, Alzheimer's disease, and obesity, respectively. tively, and approximately 2.5-4.5-fold higher than REt30 (Table 4).
Bioactive compound contents suggested that high inhibitory activities, as observed in the ethanolic fraction, might be due to the biological functions of caffeic acid and cyanidin since the contents of both phenolic acids were found to be significantly higher in ethanolic than water fractions (Table 3). Likewise, contents of these phenolic acids were higher in water extracted at higher temperature than lower temperature. Caffeic acid exhibited an IC 50 value of 149.17 μg/ ml against α-glucosidase and 77.46 μg/ml against ACE (Bhullar et al., 2014;Ishikawa et al., 2007), while cyanidin displayed an IC 50 of 1.15 μg/ml for α-glucosidase (Tadera et al., 2006). Interestingly, even though latex exhibited high content of caffeic acid, its activity against α-glucosidase was not observed. Thus, cyanidin, which was absent in latex, might be the main phenolic acid responsible for antiα-glucosidase activity. This hypothesis was confirmed by higher α-glucosidase inhibitory activity detected in the fraction containing cyanidin than caffeic acid (Ishikawa et al., 2007;Tadera et al., 2006).

The AChE and lipase inhibitory activities of all FD extracts were
low compared to α-glucosidase and ACE. Even though ethanol could extract more phenolics from FD material than water, these phenolics might not be suitable inhibitors for AChE. Considering the active site of AChE (Bajda et al., 2013), only small and less bulky compounds can fit in, leading to particular types of AChE inhibitors. Interestingly, RW30 extracted using water at low temperature (30°C) exhibited higher AChE inhibitory effectiveness than extraction with water at high temperature (80°C). Thus, data suggested that bioactive compounds extracted from FD root with water might be small compounds that acted as more effective AChE inhibitors than larger compounds from ethanolic extract.
Surprisingly, no inhibitory activity was detected in REt30 against lipase, even though this fraction exhibited higher TPCs and TFCs than water extracts (Table 2). Therefore, phenolics might not be the main compounds responsible for antilipase activity in FD root extracts. Previous literature suggested that particular peptides/proteins and polysaccharides play an important role in inhibiting lipase reaction (Gargouri et al., 1986;Hu et al., 2013). Concurring with our data, ethanolic extract of F. carica showed comparable low antilipase activity (Mopuri et al., 2018). showed no toxicity to all tested cancer cells ( Figure S2). REt30 and latex may be toxic to cancer cells.

TA B L E 4
The IC 50 values and percent inhibition of RW30, RW80, REt30, and latex against α-glucosidase, AChE, ACE and lipase exhibited slight anticancer activity, and the MTT assay was repeated with maximum dose set at 500 µg/ml. Results in Figure 1 and Table 5 show that REt30 caused significantly decreased cancer cell survival, especially for A549, SKOV3, and HepG2 cells with IC 50 values of 268 ± 46, 323 ± 15, and 337 ± 39 µg/ml, respectively, whereas latex showed minor toxic effects only in A549 cells with an IC 50 value ≥500 µg/ml. Our data indicated that growth inhibition of cancer cells by REt30 was more effective than latex. Likewise, the selectivity index (SI) of REt30 was assayed in parallel using mouse embryonic fibroblast (3T3-L1) as a normal cell representative. An SI value ≥2 reflects high selectivity of the extract (Badisa et al., 2006). REt30 exhibited anticancer activity (Table 5), with SI values ranging from 2.3 to 3.0, implying a high degree of selectivity against cancer cells, particularly A549 cells.
HPLC analysis (Table 3) showed caffeic acid as the main phenolic compound existing in REt30 at 0.0136 µg/ml in 500 µg/ml extract.
Caffeic acid was shown to inhibit A549 growth with an IC 50 value of 27 µg/ml (Lin et al., 2012), while cyanidin presented in REt30 of 500 µg/ml extract at 0.0023 µg/ml, and was not toxic to A549 cells (Chen et al., 2006). Many studies reported that REt30 affected the growth of some cancer cells, possibly by disturbing cell cycle progression, while p53, a frequently mutated tumor suppressor gene found in cancer, contributed to the sensitivity of cancer cells to chemotherapeutic agents. Thus, A549 (wildtype p53) and SKOV3 (p53 mutated) were selected for cell cycle study (Mukhopadhyay & Roth, 1997;Ryu et al., 2009). Progression of A549 and SKOV3 after treatment with REt30 at 300-500 µg/ ml was compared with untreated and cisplatin-treated control. The F I G U R E 1 Effect of REt30 and latex on cell viability of human colorectal adenocarcinoma (SW620), human hepatocellular carcinoma (HepG2), human ovarian carcinoma (SKOV-3), human prostate adenocarcinoma (PC3), and human lung carcinoma (A549). Data are expressed as mean ± standard deviation (SD) of three experiments. The percentage of cell viability of each cell line (% cell viability) was calculated, and statistical significance of each cell line was analyzed by Student's unpaired t test against its control group. *p <.05; **p <.01; ***p <.001; and ****p <.0001 While by contrast to cisplatin, REt30 at 300 and 400 µg/ml significantly arrested A549 cells in the G1 phase (approximately 70%-80% and 1.3-fold greater than the untreated control) (Figure 3). Cell distribution of A549 was not changed by REt30 at 500 µg/ml, possibly due to cell death induction as observed under the microscope (data not shown). An aqueous extract of F. religiosa was shown to induce G1 phase arrest of human cervical cancer cells (SiHa, wild-type p53 (Lee et al., 2006)) by disturbing G1 phase proteins (Choudhari et al., 2013). Consistent with our data, damaged cancer cells with wild-type p53 status were arrested at G1 before apoptosis induction (Shaw et al., 1992); however, p53 mutant cancer cells underwent G2/M arrest (Kastan et al., 1991). A high dose of REt30 was required to inhibit A549 growth and this should be explored for p53mediated apoptosis in lung cancer.

| Genotoxicity analysis in Drosophila
Scant information regarding genotoxicity has been documented; however, the historical use of Ficus species in folk medicine with no reports of any adverse side effects suggests safety in humans.
Before investigating FD extract in both animal and human studies, genotoxicity testing is mandatory. An in vivo short-term assay involving the Drosophila wing spot test (SMART) was used to determine the mutagenic properties of FD extract. The SMART assay is a more sensitive mutagenic assay compared to rodents (Delgado-Rodriguez et al., 1999) and is capable of detecting several types of DNA damages, including point mutations and DNA breaks (Wurgler & Graf, 1985). First, the survival rate of trans-heterozygous third instar larvae was studied after exposure to FD extracts. The number of surviving flies obtained from larvae raised on each experimental medium containing three different concentrations of RW30, RW80, REt30, and latex is illustrated in Figure 4. Compared to the negative control, results implied that not all FD extracts induced toxicity; therefore, 2 mg/ml was further used for wing spot analysis. Table 6 show that, in agreement with a previous study (Srichamnong et al., 2018), a nontoxic dose of urethane strongly displayed total spots/wing at 24.63 compared to 0.53 for the negative control. In addition, compared to urethane, RW30, RW80, REt30, and latex revealed no induction of either small, large, or twin spots even at a high dose of 2 mg/ml ( Table 6), suggesting that RW30, RW80, REt30, and latex could be considered as genome safe. We also studied the antigenotoxicity of FD extracts against urethane using the same assay. Before wing analysis, the larvae were cotreated with 20 mM urethane and three concentrations of FD extracts. Again, urethane exhibited total spots/wing at about 30 (Table 7); however, all FD extracts revealed negligible or weak inhibitory effects toward urethane mutagenicity (5%-40% of inhibition).

Findings in
Further investigations using other reference mutagens might elucidate the antimutagenicity potential of FD extract.

| CON CLUS ION
This is the first comprehensive bioactivity and genotoxicity assess-

F I G U R E 4
Survival percentage of adult flies obtained from trans-heterozygous mwh+/+flr 3 larvae with improved high bioactivation cross-fed on standard medium (NDM) and experimental medium containing 20 mM urethane or RW30, RW80, REt30, and latex (0.25, 1, and 2 mg/ml)

CO N FLI C T O F I NTE R E S T
All authors declare that there are no conflicts of interest.

E TH I C A L A PPROVA L
The

DATA AVA I L A B I L I T Y S TAT E M E N T
Data available on request from the authors.  (12)