Preliminary in vitro assessment of the potential toxicity and antioxidant activity of Ceiba speciosa ( A . St .-Hill ) Ravenna ( Paineira )

The bark tea of Ceiba speciosa, a tropical tree of the Malvaceae family, is used in the Northwestern Region of Rio Grande do Sul state, Brazil, to reduce blood cholesterol levels. However, there are no scientific data on the efficacy and safety of this plant. The aim of the present study was to evaluate the in vitro antioxidant and toxic potential of bark extracts of C. speciosa. We performed a preliminary phytochemical analysis by high-performance liquid chromatography-diode array detection (HPLCDAD) and evaluated the oxidative damage to proteins and lipids, the radical scavenging effect, and genotoxicity of the lyophilized aqueous extract (LAECs) and the precipitate obtained from the raw ethanol extract (Cs1). The phytochemical profile demonstrated the presence of phenolic and flavonoid compounds. The LAECs and Cs1 prevented damage to lipids and proteins at concentrations of 50 and 10 μg/mL. They also showed a scavenging effect on 2,2-diphenyl-1-pricril-hydrazyl (DPPH) radicals in a concentration-dependent manner. Furthermore, no genotoxic effect was observed at concentrations of 10, 5 and 2 μg/mL in the Comet assay. The present study is the first evaluation regarding the characterization of C. speciosa and its safety, and the results demonstrate its antioxidant potential and suggest that its therapeutic use may be relatively safe.


INTRODUCTION
Given the importance of medicinal plants as sources of new drugs, ethnopharmacology is an important tool for the discovery of species with potential biological activity (Carlini et al., 2006), even in the absence of scientific information on the safety and efficacy of these species.This branch of science studies traditional knowledge and use of plants as alternatives to essential resources for survival, such as food and medicine (Balick, Cox, 1996).Studies of medicinal plants are also of fundamental importance since many of them have a strictly local use and may offer an alternative treatment for various diseases.The popularization of medicinal plants has made them easier to acquire; however, there remains a lack of adequate information about the properties of the plants used and their possible interactions with prescribed drugs, as well as problems in botanical identification, and ignorance in terms of possible side effects (Mashwani et al.¸ 2015;Medeiros, Ladio, Albuquerque, 2015).Technological advances in healthcare as a whole, and, in particular the wide range of treatments and drugs available, have not meant the end of the use of medicinal plants as a therapeutic alternative for a large part of the population.In fact, in recent decades, as well as the everincreasing use of medicinal plants, approximately 30% of all lead compounds evaluated as therapeutic agents have been derived from natural products (Calixto, 2005).Brazil is no exception in this context; therefore, in 2007, the Brazilian Ministry of Health published The National Policy on Medicinal Plants and Herbal Medicines (Brasil, 2007), aimed at the implementation of the therapeutic use of medicinal plants in the Brazilian Unified Health System (SUS).The main aims of this policy were to encourage research focused on ensuring the proper and safe use of these species.In addition, the National List of Medicinal Plants of Interest to Brazilian Public Health System (RENISUS), containing 71 species for therapeutic purposes (Brasil, 2009), was created in order to guide studies and research that support the development of a list of medicinal plants and herbal medicines to be made available for use by the general population.
One of the most significant contributors to toxicity and the development of various diseases are free radicals; however, they are fundamental to biochemical processes and represent an essential part of aerobic life and metabolism.Under normal conditions, reactive oxygen species (ROS) are an essential part of such processes, and the most common ROS include the superoxide anion (O 2

−•
), hydrogen peroxide (H 2 O 2 ), peroxyl radicals (ROO • ), and reactive hydroxyl radical (HO • ) (Benhammou, Bekkara, Panovska, 2009;Gaur et al., 2009).The imbalance between free radical production and endogenous detoxification systems (predominantly non-enzymatic and enzymatic antioxidants) can lead to oxidative stress.Damage occurs mainly through the interaction of these radical species and macromolecules such as lipids, proteins, and deoxyribonucleic acid (DNA) (Benhammou, Bekkara, Panovska, 2009).The interaction of free radicals with DNA can lead to mutations and may therefore promote carcinogenesis, and accentuate the processes involved in the development of atherosclerosis, arthritis, and ischemia.ROS detoxification systems include enzymatic and non-enzymatic antioxidants.Enzymatic systems include superoxide dismutase, catalase, glutathione peroxidase, and other enzymes that act specifically against ROS (El-Shenawy, Mohammadden, Al-Fahmie, 2012).Non-enzymatic systems, either endogenous in origin or obtained through the diet, include glutathione, vitamin C, indoles, catechols and polyphenols, bioflavonoids, vitamin E, and carotenoids (Kharrazi et al., 2008).The demand for antioxidants means that plant materials are a significant resource in the fight against diseases associated with oxidative stress.
Despite their popular medicinal use for many different purposes, especially in South America, there have been few pharmacological studies of species of the genus Ceiba, which is not listed by RENISUS (Said, Nahla, Ehsan, 2013).Popularly known as silk tree, kapok and "barriguda", trees of this genus belong to the family Malvaceae and are mainly found in rainforest areas.
Ceiba speciosa (A.St. Hill) is a tropical tree species with a wide geographical distribution that is usually found in the mesophytic semideciduous forests of Paraguay, Argentina and Brazil.In Brazil, they occur abundantly in the states of Rio de Janeiro and Minas Gerais (Beleski-Carneiro, Suqui, Reicher, 2002).In the state of Rio Grande do Sul, this plant is found mainly in the forest region of Alto Uruguai.This species is also a popular ornamental tree in parks and gardens because of its undemanding soil requirements, achieving satisfactory growth even in dry and sandy soils with poor chemical fertility (Veloso et al., 1997;Cappelatti, Schmitt, 2009).The scientific literature describes various popular medicinal uses of several species of the genus Ceiba (Cartaxo, Souza, Albuquerque, 2010;Scarpa, 2004;Ladeji, Omekarah, Solomon, 2003).There are reports of the popular use of this plant in the northwest region of the state of Rio Grande do Sul for the reduction of serum cholesterol, triglyceride and glucose levels.The present study aimed to evaluate the in vitro antioxidant potential and toxicity of C. speciosa, contributing to our knowledge about this poorly studied plant; and to increase ethnopharmacological knowledge by collecting some phytochemical, pharmacological, and toxicological data about the species.

Chemicals
All chemicals used were of analytical grade.Highperformance liquid chromatography (HPLC) solvents were purchased from Merck (Darmstadt, Germany).All other reagents were obtained from Sigma Chemical Co.(St.Louis, MO, USA).

Plant material
The plant material was collected in May 2012 in the municipality of Santo Antônio das Missões, Rio Grande do Sul, Brazil.The species was identified and a voucher specimen (RSPF 12367) was deposited in the Herbarium of the University of Passo Fundo, Passo Fundo, Rio Grande do Sul, Brazil.

Preparation of aqueous extract
The aqueous extract was prepared by decoction of the powder from the dry trunk bark of C. speciosa.The obtained powder was boiled in distilled water for 30 minutes with periodic stirring.This procedure was adopted to simulate the popular use of this plant as a tea.
The extract was filtered and the filtrate was lyophilized resulting in the lyophilized aqueous extract of C. speciosa (LAECs).

Preparation of ethanolic extract
The raw ethanolic extract of C. speciosa was obtained by maceration until exhaustion at room temperature after which the solvent was removed under reduced pressure at 45 ºC using a rotary evaporator.During this process, a large amount of brown precipitate formed (Cs1).Little material remained in the ethanol extract after Cs1 separation, and it was thus discarded.Like the aqueous extract, Cs1 was rich in phenolic compounds and was therefore used for experiments.Further investigations on the complete ethanolic extract are being carried out.

Determination of total polyphenolic content
Measurement of total polyphenols was performed in accordance with the method established by Chandra and Mejia (2004) using a final concentration of 0.150 mg/ mL.The standard curve was prepared using gallic acid.All analyses were performed in triplicate.

Determination of total flavonoid content
The total flavonoid content of LAECs and Cs1 was determined in accordance with Zhishen, Mengcheng and Jianming (1999), with a starting concentration of 1 mg/ mL.The standard curve was prepared using quercetin, and all analyses were performed in triplicate.

Analysis by high-performance liquid chromatography with diode array detection (HPLC / DAD)
Lyophilized aqueous extract of C. speciosa (LAECs) and Cs1 were submitted to preliminary chromatographic analysis to determine their chromatographic profiles and the spectra of their main compounds in the ultraviolet region with emphasis on the quantification of flavonoids and phenolic compounds (quercetin, rutin, caffeic acid, gallic acid, chlorogenic acid, rosmarinic acid and kaempferol).High performance liquid chromatography (HPLC) of the samples was performed with an HPLC system (Shimadzu, Kyoto, Japan) consisting of a prominence auto sampler (SIL-20A), equipped with Shimadzu LC-20 AT reciprocating pumps connected to a DGU 20A5 degasser with a CBM 20A integrator, SPD-M20A UV-VIS-DAD (diode-detector), and Software LC solution 1.22 SP1, according to the method described by Boligon et al. (2009) with minor modifications.The analyses were carried out with a C18 analytical column (5 µm; 4.6 x 250 mm) and a gradient combining solvent A (1% formic acid in water) and B (acetonitrile) with a flow rate of 0.7 ml/min (Table I).The UV absorption spectra of standards as well as the samples were recorded in the range of 230-400 nm.Table II shows the detection wavelengths of phenolic compounds analyzed.

Leukocyte and plasma samples
The experimental protocols used for the cytotoxicity and genotoxicity analyses, including venous blood collection, were approved by the Research Ethics Committee of the Federal University of Pampa (Universidade Federal do Pampa [UNIPAMPA]), under registration number 27045614.0.0000.5323.
The venous blood was collected by venipuncture after 12 hours' overnight fasting using top purple Vacutainer® (BD diagnostics, Plymouth, UK) tubes with heparin from a single healthy non-medication-using donor aged over 18 years old.The blood sample was centrifuged at 3000 rpm for 10 minutes and the white  ).The number of leukocytes was standardized (8×10³ cells/mL) for measurements of cell viability and oxidative damage of DNA (Pereira et al., 2015).

Cell viability test
Cell viability in human leukocytes was assessed by the loss of membrane integrity using the Trypan Blue method (Burow et al., 1998) to determine the concentrations of LAECs and Cs1 to be used in in vitro tests.Stock solutions of LAECs and Cs1 (2 µg/mL) were prepared in PBS.From these, aliquots were transferred to Eppendorf tubes and diluted in a suspension of leukocytes to achieve concentrations of 1000, 750, 500, 250, 100, 50, 10, 5 and 2 µg/mL.Phosphate-buffered saline with H 2 O 2 was used as the positive control.The negative control consisted of a PBS and leukocyte suspension.The concentrations chosen for this study showed cell viability above 85%.The analysis was performed in triplicate.

Radical-scavenging capacity -DPPH assay
Evaluation of in vitro antioxidant capacity was performed by measuring the abilities of LAECs and Cs1 to quench the 2,2-diphenyl-1-pricril-hydrazyl (DPPH) radical, according to the method of Sharma and Bhat (2009).The analysis was performed in triplicate.

Lipid peroxidation test
Assessment of lipid damage was performed by quantification of thiobarbituric acid reactive species (TBARS) (Ohkawa, Ohishi, Yagi, 1979).Initially, LAECs and Cs1 solutions were prepared at the previously chosen concentrations.Two hundred microliters of plasma were mixed with 100 µL H 2 O 2 100 µmol/L (1:10 dilution) and 200 µL LAECs or Cs1 solutions.For the positive control 200 µL H 2 O 2 was used in place of the LAECs and Cs1 solutions.In the negative control, they were replaced by saline solution.The analysis was performed in triplicate.

Determination of protein carbonyl groups
Determination of protein carbonyl groups was performed according to the method of Morabito et al. (2004).Initially, LAECs and Cs1 solutions were prepared at the previously chosen concentrations.To perform the test, 200 µL plasma was added to an Eppendorf tube and mixed with 100 µL H 2 O 2 solution 100 µmol/L (1:10 dilution), and 200 µL LAECs or Cs1 solutions.As a positive control, 200 µL H 2 O 2 was used in place of the LAECs and Cs1 solutions.In the negative control, they were replaced by 200 µL PBS.The analysis was performed in triplicate.

Comet assay
The Comet assay was used to measure the single and double strand breaks in DNA (Singh et al., 1995).An Eppendorf tube containing 200 µL extract and 200 µL leukocyte suspension was prepared.All steps after blood collection were performed in the dark or under dimmed red light to prevent additional DNA damage.
The Comet assay was performed in triplicate, analyzing 100 nuclei per sample of each concentration s t u d i e d .T h e s e 1 0 0 c e l l s w e r e t h e n e x a m i n e d microscopically at 40x magnification and classified according to tail damage: 0 being no damage to 4 being maximum damage.The sum of these values was used to ascertain a damage index (0-400) for each treatment.The analysis was performed in triplicate.

Statistical analysis
Data were expressed as mean ± standard deviation.Concentrations of flavonoid and phenolic compounds obtained by HPLC were analyzed using one-way analysis of variance (ANOVA) followed by the Tukey test.Other results were analyzed using one-way ANOVA followed by the Bonferroni test.Results were considered statistically significant when p<0.05.Statistical analysis was performed using the software GraphPadPrism 5.0.

Total flavonoid and total polyphenolic contents
Total flavonoid contents, expressed as microgram equivalents of quercetin per gram of extract sample, were 240 µg and 229 µg of quercetin/g for LAECs and Cs1, respectively.Total polyphenolic contents, expressed as microgram equivalents of gallic acid per gram of extract sample, were 425 µg and 470 µg of gallic acid/g for LAECs and Cs1, respectively.

Cell viability test
The aim of cell viability testing (Figure 2) was to choose concentrations for the in vitro studies.

Radical-scavenging capacity -DPPH assay
LAECs and Cs1 demonstrated significant radical scavenging activity in the DPPH assay (Figure 3).They exhibited a scavenging effect of 85.13% and 88.95% respectively at a concentration of 50 µg/mL.The inhibition rate was concentration-dependent, being 49.12%, 27.52%, and 13.32% for LAECs and 73.12%, 45.35%, and 21.36% for Cs1, both at concentrations of 10, 5 and 2 µg/mL, respectively.Dimethyl sulfoxide (DMSO) was only used to help dissolve Cs1, and it was included in the analysis  and figure to demonstrate that it did not have a statistically significant effect, and therefore no effect on the results.

Lipid peroxidation test
LAECs and Cs1 had an almost constant effect on lipid peroxidation at all concentrations tested, with similar effects to those of the negative control (Figure 4).

Determination of protein carbonyl groups
At concentrations of 50 and 10 µg/mL, LAECs demonstrated a carbonylation index statistically similar to that of the negative control (Figure 5); however, LAECs at 5 and 2 µg/mL and Cs1 at all concentrations showed results statistically similar to that of the positive control.

Comet assay
Lyophilized aqueous extract of C. speciosa at concentrations of 10, 5, and 2 µg/mL and Cs1 at concentrations of 5 and 2 µg/mL produced statistically similar results to the negative control, showing no significant DNA damage; however, LAECs at 50 µg/mL and Cs1 at 50 and 10 µg/mL presented a score of 1 for DNA damage (Figure 6).

DISCUSSION
Studies of the toxicity and the protective effects of species used in traditional medicine are essential.In this      et al. (2013), who found a high concentration of flavonoids (7.7 mg quercetin/g of extract) and phenolic compounds (3.3 mg gallic acid/g of extract) in C. speciosa leaves.LAECs showed higher concentrations of phenolic compounds than Cs1, with the exception of quercetin.
In the present study, a cell viability assay was performed to determine the concentrations to be used in the analysis (Koko et al., 2008).The cell viability remained above 85% at relatively high concentrations, indicating that LAECs and Cs1 were non-toxic to human leukocytes.Antioxidant tests were performed using a DPPH assay; this is a widely accepted method for evaluation of in vitro antioxidant activity of extracts and is based on the capacity of antioxidant compounds present in the sample to capture DPPH radicals (Benhammou, Bekkara, Panovska, 2009;Ayala-Zavala et al., 2012;Mohsin, Mahadevan, Kurup, 2014;Wan-Ibrahim, Sidik, Kuppusamya, 2010).The results showed that LAECs and Cs1 inhibited ≥ 50% of radicals at concentrations of 10 µg/mL and 50 µg/ mL, showing that both samples appear to protect the studied biological matrix, since the concentrations tested were able to neutralize the radical DPPH.In a study by Ayala-Zavala et al. (2012), the authors examined several parameters to assess the antioxidant activity of Phellinus gilvus, Phellinus rimosus, and Phellinus badius, and found a correlation between the total polyphenolic content and DPPH radical scavenging activity; therefore, the inhibitory potential found in this work indicates an elevated concentration of these phytochemicals and their high antioxidant power.
Following determination of the antioxidant potential of C. speciosa, the in vitro effects of different concentrations of LAECs and Cs1 on some important parameters of oxidative stress were evaluated.An imbalance between production and inactivation of ROS can lead to oxidative stress and damage to various biological macromolecules, and proteins exposed to ROS can suffer oxidative modifications that can lead to changes in their functions, and, consequently, affect cellular metabolism (Gupta, Ballal, 2015).In this context, measurement of protein carbonyl groups in biological samples is a reliable parameter with which to assess ROS-mediated protein oxidation (Levine et al., 1990).In this study, LAECs at concentrations of 10 and 50 µg/ mL showed a protein carbonyl content similar to the negative control, demonstrating no evidence of oxidative damage, and suggesting that these samples prevented H 2 O 2 -induced protein carbonylation.However, the results for LAECs at concentrations of 2 and 5 µg/mL were similar to those observed for the positive control, suggesting the occurrence of oxidative damage.At all concentrations tested, Cs1 demonstrated a carbonyl group content similar to the positive control, which indicates the occurrence of protein damage.These results may reflect the high concentrations of quercetin in Cs1 because, despite the beneficial actions of flavonoids, studies have demonstrated that quercetin has an in vitro pro-oxidant action (Choi, Chee, Lee, 2003;Heim, Tagliaferro, Bobilya, 2002;Yang et al., 2012).Studies have reported that, in general, phenolic compounds exhibit pro-oxidant activity in vitro but not in vivo, and the conditions for this arise mainly in the presence of transition metals and acidic pH (Eghbaliferiz, Iranshahi, 2016).This suggests that further studies are necessary to determine whether these factors are present in our samples.In other species, ROS can target lipids and initiate the lipid peroxidation process, which may lead to molecular cell damage (Gupta & Ballal, 2015).In the present study, LAECs and Cs1 showed no differences compared to the negative control at all concentrations tested in the H 2 O 2 -induced lipid peroxidation assay, which indicates the absence of lipid oxidative damage and hence the potential antioxidant action of C. speciosa, consistent with the results of the DPPH test.These results are an important indication of the safety of this species, because the occurrence of lipid peroxidation is directly related to various disease processes such as carcinogenesis and atherosclerosis.
Besides protein and lipids, DNA is also a significant target for oxidative damage mediated by ROS.Samples were subjected to the Comet assay, which is considered a sensitive method for the detection of single and double DNA strand breaks (Collins et al., 2008).LAECs at concentrations of 2, 5 and 10 µg/mL, and Cs1 at concentrations of 2 and 5 µg/mL, showed no significant DNA damage.Both samples at higher concentrations (50 µg/mL of LAECs and 10 and 50 µg/mL of Cs1) showed low levels of DNA damage (score 1).These results indicate that further genotoxic studies are required in order to deepen knowledge about the safety of C. speciosa.

CONCLUSION
This study revealed that C. speciosa extracts have promising antioxidant potential, which may be related to their high polyphenolic content.Phenolic compounds are the constituents responsible for the antioxidant potential of natural products (Krishnaveni et al., 2013;Loganayaki, Siddhuraju, Manian, 2013) from species of the genus Passiflora and Citrus, both known for being rich in flavonoids and with well-established antioxidant activity.Furthermore, LAECs and Cs1 showed low toxicity in vitro, especially at a concentration of 10 µg/mL.C. speciosa extracts showed either no damage or at the most a low degree of DNA damage (grade 1) in the Comet assay and they also prevented lipid peroxidation at all concentrations tested.Aqueous extract (LAECs), which is used in folk medicine, showed no damage to proteins at concentrations of 10 and 50 µg/mL.
The use of plants is widely accepted by the public, and this fact is closely related to the popular belief that natural products are free of adverse effects.However, the popularization of the use of medicinal plants, their ease of access, and lack of professional guidance on their correct use, represent a health risk (Al-Arifi, 2013;Jeong et al., 2012).This study describes a preliminary evaluation of the biological activity and safety of C. speciosa and shows that this species exhibits in vitro antioxidant potential and low toxicity, which may be related at least in part to its high concentration of polyphenols.The results provide an important foundation for knowledge about this species, which is popularly used for the reduction of serum cholesterol, triglyceride and glucose levels.Further phytochemical analyses are necessary to determine the chemical compounds responsible for the antioxidant activity of C. speciosa and to correlate them with other possible pharmacological or even toxicological effects of this very promising plant.

TABLE I -
Gradient elution used for the evaluation of lyophilized aqueous extract (LAECs) and precipitate obtained from the raw ethanol extract (Cs1).Solvent A (1% formic acid in water).

TABLE II -
Wavelengths used for the detection of phenolic compounds by High-Performance Liquid Chromatography with Diode Array Detection (HPLC/DAD)

TABLE III -
Composition of lyophilized aqueous extract of C. speciosa (LAECs) and raw ethanol extract (Cs1) Results are expressed as mean ± standard deviation of three determinations.Means followed by different letters differ by Tukey test (p <0.05).LOD -Limit of Detection; LOQ -Limit of Quantification.

TABLE IV -
Composition of lyophilized aqueous extract of C. speciosa (LAECs) and raw ethanol extract (Cs1), at the concentrations used for the in vitro tests