Use of zebrafish (Danio rerio) in experimental models for biological assay with natural products

1 Programa de Pós-graduação em Ciências Farmacêuticas, Departamento de Ciências Biológicas e da Saúde, Universidade Federal do Amapá, Brazil. 2 Laboratório de Pesquisa em Fármacos, Curso de Farmácia, Departamento de Ciências Biológicas e da Saúde, Universidade Federal do Amapá, Brazil. 3 Laboratório de Nanobiotecnologia Fitofarmacêutica, Curso de Farmácia, Departamento de Ciências Biológicas e da Saúde, Universidade Federal do Amapá, Brazil. 4 Facultad de Química, Departamento de Farmacia, Universidad Nacional Autónoma de México, México D.F., México.


INTRODUCTION
The term natural products refers to a living organism, a complete plant or any part of it, or to any chemical compound which had been extracted from these without any changes in its chemical structure (secondary metabolites) (Miranda and Cuéllar, 2001). Natural products are an excellent source of chemical compounds *Corresponding author. E-mail: farmacos@unifap.br.
Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License for new drugs discovery. Until now, bioactive components isolated from plants, fungi and bacteria gave rise to a wide range of therapeutic compounds. Also, a lot of these compounds are useful tools in pharmacology, biochemistry and cell biology and molecular (Crawford et al., 2008).
The improvement of computational chemistry in the 90s and the development of new technologies such as "target validation" and "high-throughput screening", caused the decline in drug research from natural products (Crawford et al., 2008). Many pharmaceutical laboratories eliminated the research program with natural products (Schmid and Smith, 2009). The research with natural products has another inherent challenge: the identification of the mechanism of action of the extracts, the synthesis of complex structures of the active compounds and the difficulty of isolating pure compounds from crude extracts, sometimes in such small quantities, that the "high-throughput screening".
Despite massive investments in new technologies for the development of chemical compounds, the number of new drugs reaching the market has not increased proportionately.
Thus, the computational and combinatorial chemistry cannot achieve its goal of being a primary source of new bioactive candidate compounds for innovative drugs (Yunes and Cechinel, 2001).
In this context, the natural products have recovered space and importance in pharmaceutical industry. They are an inspiring source of new bioactive molecular patterns. At the same time, there is a growing appreciation of natural products, because they have chemical structures, with the ability of interacting effectively with biological macromolecules (Koehn and Carter, 2005;McChesney et al., 2007). Therefore, the research with natural products will continue to expand since these are a source of inspiration for the development of new drugs.
Research on natural products is facing another major difficulty. The analysis "high-throughput screening" has limited effectiveness of their evaluation. These analyses make possible in vitro screening of pharmacological activity of the drugs candidate molecule, by evaluating a single target. The presence of many substances in a natural extract make this process unprecise. These analyses often generate compounds with little or no effectiveness in vivo (Van der Greff et al., 2005). However, the consequence of this deficiency has stimulated the search for new methods of bioassays for a screening of pharmacological activity that is more efficient in the evaluation and target independent (Crawford et al., 2008).
The use of zebrafish in experimental models, has occupied space in the research with natural products, currently being used by several research groups for screening and confirmation of various pharmacological activities. The biological response obtained in this model is similar to that achieved in mammals, helping researchers in refining and reducing the use of rodents. In the future, the study of natural products in zebrafish model will contribute to greater understanding of the biological activity of a wide range of bioactive compounds because this model requires a very small amount for the tests.
The zebrafish (Danio rerio Hamilton-Buchanan, 1822) is well suited to high-throughput applications owing to its high fecundity, rapid extrauterine development and transparency during organogenesis. This fact promotes and increases the relevance of their use in biological assays, which is hard to be matched in vitro (Lin et al., 2013). The objective of this work is to update knowledge on the use of ZF as an experimental model for biological assays with natural products and derived.

ADVANTAGE OF THE ZEBRAFISH AS EXPERIMENTAL MODEL
The zebrafish (ZF) belongs to the Cyprinidae family. It is a small teleost fish of tropical freshwater. It has a body length of three to four centimetres. It is a species well known for its ornamental use (Poon and Brand, 2013). This small fish is being used every day by the international scientific community. In 1955, it was reported, for the first time, the use of ZF as an experimental model. From that date until now, the use of ZF had an exponential growth ( Figure 1 and Table 1).
The key to the complete success of ZF, as an experimental model, lies in the fact that it has biological characteristics and for maintenance, it is very favourable. The ZF has a short cycle of life and a rapid development, which takes place outside the female's uterus. The ZF produces a large progeny (hundreds of eggs per mating). Another critical feature in its use is the low cost and reduced space required for its maintenance and care. The ZF embryos are small (1 to 5 mm), depending on the stage of development) and are translucent. This fact makes them suitable for handling and use (Lele and Krone, 1996;Bailey et al., 2013). Most of the citations of ZF as experimental model refers to the use of embryos and larvae.
The determining factors for use of ZF as an experimental model are the physiological and pharmacological responses similar to humans and other higher mammals. This makes it suitable for identifying drugs and bioactive natural products with therapeutic potential. The biological characteristics of the ZF combined with the advantages it offers in maintenance and care, make it an ideal model in vivo for "medium throughput screening" (Crawford et al., 2008).
The key advantage of the bioactivity assay of small molecules, by using ZF embryos and larvae is that the compounds can be diluted with non-sterile water. The little weight molecules are rapidly absorbed by the skin and gills of ZF. This allows an increase in the "throughput   screening" of hundreds of molecules per day (Zon and Peterson, 2005;Langheinrich, 2003). Majority of the studies where ZF is used as the experimental model, reported the use of larvae and embryos. One of the big questions in the use of ZF is how far the results could be extrapolated to humans and other higher mammals?
Genetic and pharmacological tests using ZF proved to have representativeness for other organisms, including humans highly. Additionally, the identified genes in the organogenesis of ZF has been consistently validated in mice and humans. In many cases, ZF and human genes have been discovered side by side. The remarkable pharmacological homologies between humans and ZF can be extended to phenotypes modified disease identified through these assays (Peterson and Macrae, 2012).
The ZF has been used as an experimental model for genetic studies. In 1981, the embryonic development of ZF was reported (Streinsinger, 1981). In the 2000s, there was considerable progress in genetics and genomics of ZF. The mitochondrial genome of ZF was fully sequenced in October 2001 (Broughton et al., 2001). In 2013, the Sanger Institute (UK) reported the complete sequence of ZF genome. When comparing the sequence of the ZF genome with the human genome, it was observed that some genes as cell cycle, growth and differentiation, tissue functions, oncogenesis and tumour suppressor, were preserved. Some studies showed that the 70% of the ZF genes are similar to human genes (Stern and Zon, 2003;Howe et al., 2013). These data show ZF as a reliable model for the screening of drugs and natural products (Amatruda et al., 2002;Crawford et al., 2008).
Genetic research on ZF produced a range of mutant strains, which can be used to carry out studies in many different pathologies. Some examples of mutant strains defective of ZF are -Transparent Zebrafish‖ (Casper) (Witte et al., 2008) and rag2E450fs (Qin et al., 2014). Just the Hopkins' insertional mutant collection contains more than 500 recessive mutants with embryonic morphological phenotypes, which include mutations in 335 identified genes (Kishi et al., 2008). These mutants allow studying pathologies related to cartilage development, hematopoiesis, cardiovascular development, among others. The Casper mutant strain (totally transparent) is used for research related to cancer and stem cells, previously limited only to embryos and larvae (Witte et al., 2008). Other genetic testing using ZF are gastrointestinal function tests, vascular development, epilepsy and diabetes (Crawford et al., 2008;Seth et al., 2013).
The ZF mutant strains have utility in toxicological research to clarify the roles of particular genes and their interactions with signalling pathways. Also, they are useful in the pathogenesis caused by damage induced by toxicants. Furthermore, ZF with double or triple mutations may help to elucidate the interaction of the suite of genes. The ZF mutants can be produced more economically and efficiently than in murine (Spitbergem et al., 2003;Hill et al., 2005).
The ZF has proved a versatile model for reverse genetics studies. "Antisense morpholino oligonucleotides" commonly referred as "morpholinos" is the most widely used technique for knockdown in ZF. This method specifically blocks the function of a gene in a ZF embryo. The morpholinos are available as genes tool, and its microinjection is performed in a dose-dependent manner in the early stages of embryo development (single-cellstage embryos). Two types of morpholinos are used to interfere with gene expression of proteins. The ATGmorpholino, that blocks translation of ribosomes and leaves out the embryo without one protein and the splicemorpholino that binds and interferes with RNA splicing, results in a truncated protein form that is used to study a particular area of the protein. However, numerous genes have been functionally analysed in this way, including several identified in the context of large-scale genetic screens reverse (Bill et al., 2009;Heasman, 2002;Biil et al., 2009;Doitsidou et al., 2002;Esguerra et al., 2007).
In recent years, there have been several advances in the ability to generate transgenic lines of ZF. The transgenic methods are well established for ZF. This allows the direct generation of transgenic lines expressing fluorescence under the control of tissuespecific promoters. This radically reduces the time to generate new strains. Traditionally, transgenic reporters were generated by microinjection of linearized plasmid DNA, containing the coding sequence of a reporter protein immediately downstream of a minimal promoter fragment of the gene of interest (Higashijima et al., 1997). However, this approach suffered from some limitations, in particular, the low efficiency of germline integration (Hammond and Moro, 2012).
Advances in the technology have included the introduction of the gateway system and the production of compatible plasmids that can be used in zebrafish (Kawakami et al., 2004;Villefranc et al., 2007), I-SceI cloning, whereby introduction of meganuclease sites increased the efficiency of germline integration (Grabher and Wittbrodt, 2008). More recently, improvements have been achieved by bacterial artificial chromosome (BAC) recombineering, in which fluorophores and Tol2 transposase sites are introduced into a BAC containing the gene or promoter of interest.

ZEBRAFISH AS EXPERIMENTAL MODEL FOR THE EVALUATION OF NATURAL PRODUCTS
The use of ZF as an in vivo model for the discovery of drug candidate molecules was proposed 58 years ago. This study presents the use of embryos and ZF larvae for the screening of synthetic drugs and natural products (Jones and Huffman, 1957). However, only in the year 2000, was described, the first "screening" using multiwell plates. Since then, over 60 studies have reported the use of ZF for performing whole projects aimed at drug discovery (Rennekamp and Peterseon, 2015).
Many compounds used by humans as flavonoids, alkaloids and some drugs have been tested for their teratogenic and embryotoxic potential in ZF model (Jones et al., 1964;Thomas, 1975;Kim et al., 2009;Stewart and Kalueff, 2014;Lu et al., 2014). Adults ZF were used to confirm the piscicide property of arylnaphthalide, a kind of lignans extracted from Phyllanthus piscatorum, a medicinal plant used by Yanomami (Amazon Indian tribe) as piscicide and antifungal (Gertsch et al., 2003(Gertsch et al., , 2004. It was also used in a recent study to determine the neurotoxic effects of a Chinese Medicinal formulation of Azadirachta indica, by using behavioural models (Bernadi et al., 2013).
Adults ZF can be used for screening anti-inflammatory activity, particularly natural products. A short time ago a model that is based on the intraperitoneal administration of λ-carrageenan was established (Huang et al., 2014). The authors reported a significant increase in abdominal swelling and proinflammatory proteins (iNOS and TNF-α) induced by carrageenan in ZF.

Evaluation of angiogenic activity in Zebrafish model
There are many reports assessing the angiogenic activity of natural products by using ZF as the experimental model. Lam et al. (2008) used transgenic ZF to characterise the pro-angiogenic properties of Angelica sinensis (Dong Quai). This work tested the crude extract, rather report the isolation of any bioactive compound (Lam et al., 2008). Liu et al. (2011) used transgenic ZF to investigate the angiogenic activity of the crude aqueous extract of Rehmannia glutinosa. This work showed the isolation of norviburtinal, a new active compound .
Subsequently, Yu et al. (2013) demonstrated the anti-angiogenic effect of the hydroalcoholic extract of Herba epimedii, a medicinal plant from East Asia. In this in vivo study, transgenic ZF was used . In 2013, one platform for biomonitoring tests (bioassay-guided) was developed. This platform combines the screening of bioactivity on transgenic ZF embryo with the rapid fractionation by thin layer chromatography and initial structural elucidation by mass spectrometry . Using these procedures, the authors identified two compounds angiogenic inhibitors (emodin and coleone) in crude extracts of Oxygonum sinuatum and Plectranthus barbatus .

Antithrombotic activity evaluation using zebrafish model
The antithrombotic activity of compounds isolated from plant extracts has been evaluated using ZF model Shi et al. (2015) glycosides, phenyl and propyl aldehydes and biphenyls from the methanolic extract of Piper wallichii (Mic.). Next, these compounds were tested for their antithrombotic activity by using ZF model. From them, just a lignane (-) syringaresinol, showed excellent antithrombotic activity. Song (2012) reported four new phenolic compounds extracted from Crataegus pinnatifida leaves. These compounds were tested jointly with other known compounds, to evaluate the antithrombotic activity on transgenic ZF. Among the isolated compounds, the eriodictyol is shown to be a potent inhibitor of thrombus formation.

Anticonvulsant activity in zebrafish model
ZF model were used to evaluate the anticonvulsant effect of Curcuma longa extract for the treatment of epilepsy. Seizures were induced with pentylenetetrazol. This extract showed antiepileptic effect (Raoa et al., 2005). Orellana-Paucar (2012) reported the use of ZF model to evaluate the anticonvulsant activity of isolated compounds from Curcuma longa oil. This study used the same pattern of seizures induced by pentylenetetrazol. In conclusion, in this study, the usefulness of a zebrafish seizure model for rapid bioactivity-guided fractionation of natural products and their purified compounds for the identification of novel small molecules with anticonvulsant activity in vivo was demonstrated.
Buenafe (2013) reported the evaluation of the ketonic extract of Salvia miltiorrhiza, Bunge and four bioactive Tanshinones obtained by fractionation of the extract. Pentylenetetrazol model was also used to induce seizure. The extract showed anticonvulsant activity. One of the active tanshinones, tanshinone IIA, also reduced C-fos expression in the brains of PTZ-exposed ZF larvae. Kao et al. (2010) showed that grape seed extract reduced the dihydrofolate reductase activity and thus the growth of S. aureus. Then, ZF was infected with S. aureus pre-incubated with grape seed extract. There was a significant decrease in the inflammatory response and mortality of ZF infected with S. aureus (KAO et al., 2010). Bohni et al. (2013) showed the screening of anti-inflammatory activity of over 80 methanolic extracts of medicinal plants of East Africa, using ZF as an experimental model. The methanolic extract of Rhynchosia viscosa (Roth) DC inhibits leukocyte migration in ZF sectional tails (4 dpf). As a result, five bioactive compounds were isolated (rhynchovisina, genistein, sophoroisoflavona A, licoisoflavona and 3-o-methylorobol). It was proven that dos Santos et al. 887 genistein and sophoroisoflavone possess anti-inflammatory and anti-angiogenic effects. Zebrafish model and cell line RAW 264.7 were used to assess the effects of anti-inflammatory compounds of the ethanol extract of the root of Gentiana dahurica (Gentianaceae). The results showed no cytotoxicity of all the tested compounds and an intense inhibitory activity of the roboric acid and liriodendrin (lignans) (Wang et al., 2013).

Antilipemic activity in zebrafish model
The zebrafish is becoming an increasingly popular model for automated discovery of drug. It is also used for hypercholesterolemic research. The creation of two algorithms was reported for automated analysis of cardio dynamic data acquired by high-speed confocal microscopy, by using data that was obtained using ZF as the experimental model (Littleton et al., 2013).
The hypolipidemic activity of aqueous extracts of turmeric (Curcuma longa) and bay leaves (Lourus nobilis L.) was evaluated in ZF model. The results showed that consumption of bay leaf and turmeric extracts produced a hypolipidemic effect and antioxidant activity . The effect of the aqueous extract of cinnamon (Cinnamomum verum) and clove (Syzygium aromaticum) on the hypercholesterolemic model in ZF was also studied.
Cinnamon and clove aquose extracts showed anti-hypolipidemic activity. They showed strongest anti-glycation and antioxidant activity in this study. Cinnamon and clove extracts (at final 10 µg/mL) had the strongest anti-glycation and antioxidant activity in this study. Cinnamon and clove had the strongest inhibition of activity against copper-mediated low-density lipoprotein (LDL) oxidation and LDL phagocytosis by macrophages (Jin, 2011a).

Melanogenic activity in zebrafish model
ZF larvae are an ideal model for studies related to melanogenic activity, as they allow an easy observation of phenotypic pigmentation process. Kim (2008) reported on anti-melanogenic activity of hanginine A, an isolated isoflavone from the branches of Lespedeza cyrtobotrya. The authors concluded that the hanginine A promotes a similar effect on 1-phenyl-2-thiourea (PTU), an inhibitor of the enzyme tyrosinase in a dose-dependent way. Park et al. (2013) investigated the melanogenic inhibitory activity of Arctigenin, an isolated compound from the aqueous extract from Fructus arctii on zebrafish cell line B16BL6 embryos and Melan-A.
The results obtained proved that the treatment arctigenin (10 uM) produces a moderate decrease of the pigment deposition in ZF, 15 dpf.
In other work, ZF model was used to evaluate the antimelanogenic effect of rengiolona, a compound obtained from the Eurya emarginata extract. The results showed an inhibition of body pigment of ZF, besides the reduction of melanin levels and activity of the tyrosinase enzyme .

Studies on neurodegenerative diseases in zebrafish model
The ZF has been used in the study of neurodegenerative diseases such as Parkinson's disease, Huntington's and Alzheimer's diseases. Zhang (2012) reported the use of ZF model and PC-12 cell line for evaluating the neuroprotective effect of ethanol extract of the fruit of Alpinia oxyphylla. This is a plant used in traditional Chinese medicine. The results showed that the extract prevented and restored dopaminergic neurodegeneration induced by 6-OHDA (6-hydroxydopamine) and attenuated deficits in locomotor activity in Parkinson's disease model in ZF. Chong (Chong et al., 2013).

Toxicological studies on studies on zebrafish model
The ZF has been widely used for environmental toxicity studies to characterise the damage caused by various pollutants. The ZF is considered the gold standard for evaluating environmental toxicity (Scholz et al., 2008). With the expansion of nanotechnology, the ZF has been gaining ground as an experimental model for environmental health and safety tests of various nanomaterial (Lin et al., 2013;Scholz et al., 2008). The ZF model has been used for assessing the toxicity of silver nanoparticles (Muth-Köhne et al., 2013), silica (Duan et al., 2013) and other nanomaterials (Jevgenij et al., 2013). Furthermore, ZF has been very usefully in evaluating plant extracts and isolated compounds. Thus, the toxic effect of celastrol, one terpenoid isolated from Tripterygium wilfordii Hook F., was assessed. The results showed that celastrol, at micromolar concentrations, affects the healthy development of ZF embryos .
In 2012, the nephrotoxicity caused by aristolochic acid compound isolated from Aristolochia asarum extract was evaluated. The result showed that aristolochic acid-induced nephropathy, defects in blood circulation and cardiac malformations, were mediated by inflammation process (Ding and Chen, 2012). Zhang et al. (2014) evaluated the toxicity of 10 compounds isolated from the ketonic extract of Kadsura oblongifolia, over cardiac function and embryonic development of ZF. The results showed that kadsulignane, meso-dihydroguaiaretic acid and kadsufolina produced edema in ZF embryos, a diminution of the heart rate and also interfered with the development of the zebrafish heart .

Conclusion
The use of zebrafish is already established as a powerful research platform for the discovery of new drugs. Also, the zebrafish is a useful model in other areas of science. It has become an experimental model for screening extracts and components derived from it, due to the ease in which small molecules and natural products can be studied in this species. A large number of tests that can be performed by using zebrafish are one of the attractions of this model since it is not possible to carry out such a large number of tests in others experimental models. Assays based on zebrafish can aid in the isolation of bioactive molecules from plant extracts, which were identified in a large-scale screening; this increases the biological relevance of such findings. Another decisive factor in using the zebrafish as an experimental model for large-scale screening are a few extracts or the isolated compounds required for tests. Therefore, the growing number of publications and innovative models for research on a lot of diseases with this species, reveal the importance of zebrafish as an experimental model for the screening of natural products.