Bioactive limonoids from Carapa guianensis seeds oil and the sustainable use of its by-products

Graphical abstract

Highlights

• •Andiroba is used in Brazil as an insect repellent and treatment to various diseases.
• •Limonoids from andiroba oil have shown many in vitro and in vivo activities.
• •Andiroba oil can be used in biodiesel production in an industrial scale.
• •New methodologies of toxicological assays have been applied to substitute animals.
• •Andiroba by-products are able to reduce environmental impacts.

Abstract

Carapa guianensis (Andiroba, Meliaceae) is considered a multipurpose tree. In Brazil, Indigenous people have used it as insect repellent and in the treatment of various diseases. Most biological activities and popular uses are attributed to limonoids, which are highly oxygenated tetranortriterpenoids. More than 300 limonoids have been described in Meliaceae family. Limonoids from Andiroba oil have shown high anti-inflammatory and anti-allergic activities in vivo, by inhibiting platelet activating factors and many inflammatory mediators such as IL-5, IL-1β and TNF-α. It also reduced T lymphocytes, eosinophils and mast cells. In corroboration with the wide popular use of Andiroba oil, no significant cytotoxicity or genotoxicity in vivo was reported. This oil promotes apoptosis in a gastric cancer cell line (ACP02) at high concentrations, without showing mutagenic effects, and is suggested to increase the body's nonspecific resistance and adaptive capacity to stressors, exhibit some antioxidant activity, and protect against oxidative DNA damages. Recently, new methodologies of toxicological assays have been applied. They include in chemico, in vitro, in silico and ex vivo procedures, and take place to substitute the use of laboratory animals. Andiroba by-products have been used in sustainable oil production processes and as fertilizers and soil conditioners, raw material for soap production, biodegradable surfactants and an alternative natural source of biodegradable polymer in order to reduce environmental impacts. This review reinforces the relevance of Andiroba and highlights its ability to add value to its by-products and to minimize possible risks to the health of the Amazonian population.

Introduction

The World Health Organization (WHO) estimates the existence of around 20,000 medicinal plants in 91 different countries (Patra et al., 2018). Currently, medicinal plants play an essential role in many countries, especially those with a predominance of traditional medicine such as least- developed countries (Braga et al., 2020, Porfirio-Dias et al., 2020, Galal et al., 2022, Singh et al., 2022). Plant species are used as home remedies, medicines, and raw materials for the pharmaceutical industry, representing a substantial part of the global drug market (Machado & Mendes, 2021). In this market, most of the drugs introduced in the last 39 years are derived, directly or indirectly, from natural products (Newman & Cragg, 2020). Despite the well-known benefits of Phytotherapy, the traditional use of medicinal plants, which is spread from generation to generation, is based on the belief that plant use is safe and effective. However, it is known that toxicological studies are of vital importance in assessing their safety (Porfirio-Dias et al., 2020). Most of the compounds produced by the secondary metabolism of plants, as well as their effects on the treatment of diseases, are still unknown (Amorim et al., 2020). Therefore, there is an excellent demand for studies on medicinal plants traditionally used by the population (Oyedeji-Amusa et al., 2020, Kaur et al., 2021, Malhotra et al., 2021, Galal et al., 2022).

Within its wide diversity, Brazil contains about 60% of the Amazon rainforest, which composes the richest collection of plant species on the planet (Greenpeace, 2020). In this Region, native people have used several plant species as traditional medicines, and about 300 species native to the Amazon have been cataloged with potential for medicinal, herbal, aromatic, and cosmetics areas (Revilla, 2001). In research with natural products, bioprospecting remains a very important tool in a country with such immense biodiversity, mostly unknown and potentially threatened by anthropic pressures and climate change. However, while the searching alternative bioactive compounds as new natural leading drugs remains the main focus of pharmaceutical companies and research communities, the social environment where these raw materials are obtained, and the waste is disposed of remains a drift of a development that can provide better conditions for those who really depend on this extractive chain to expand (Veiga Junior and Yamaguchi, 2022).

Many products derived from the native Amazonia Region are sold in national and international markets such as acai, guarana, tropical fruits (in the form of pulp, sweets and ice cream), palm hearts (peach palm and acai), various phytotherapics and phytocosmetics (Costa, 2012). In addition, there are other non-wood products with great export value: Brazil nuts, jarina (vegetable ivory), rutila and jaborandi (active ingredients), rosewood (essence of perfume), resins, and oils (Freitas, 2013). Deforestation and desertification caused by human activities and climate change pose major challenges to sustainable development and affect the lives and livelihoods of millions of people (United Nations, 2016). This way, environmental conservation goals are paramount to rural populations and indigenous peoples (Fa et al., 2020). Several Amazon species are oilseeds with great potential to produce notable nutraceuticals, cosmetics, medicines, and energy. Due to their high oil content, the traditional use of some members of the Meliaceae family for many conditions has shown promising pharmacological activities (Penido et al., 2006). This family of angiosperms comprises 52 genera and 3198 listed species. However, only 669 species are accepted (A working list of all plant species). These species are used in traditional medicine for many purposes such as antimicrobials and anti-parasitic (Mendonça and Ferraz, 2007, Oyedeji-Amusa et al., 2020), as repellents and insecticides (Mendonça et al., 2005, Ambrozin et al., 2006, Silva et al., 2009); inflammation and analgesic (Henriques and Penido, 2014, Soares et al., 2021).

Among the natural compounds isolated from different parts of the plants of this family, a variety of mono-, di-, sesqui-, and triterpenoids, coumarins, chromones, lignans, flavonoids, and other phenolic substances have been described (Yadav et al., 2015). There are reports in the literature of a phenolic compound that exhibits protective activity against the neurodegenerative effects of Alzheimer's and Parkinson's diseases (Güzelad et al., 2021, Ogut et al., 2022). However, limonoids are considered the most important natural compounds isolated from Meliaceae species (Yadav et al., 2015, Fan et al., 2022, Luo et al., 2022). Limonoids are highly oxidized tretranortriterpenoids originating from a fused 6/6/6/5 tetracyclic carbon skeleton, with methyl groups at C-4, C-4, C-8, C-10, and C-13 positions and a furan ring at C-17. The structural diversification that leads to different limonoid subtypes is due to cleavages and rearrangements (Higuchi et al., 2017b, Kikuchi et al., 2020). They have several pharmacological activities, mainly as insecticide (Lin et al., 2021), ant-inflammation and analgesic (Henriques and Penido, 2014), antibotulism, and anticancer (Fan et al., 2022). More than 300 limonoids have been isolated and are described as the most abundant and diverse in the Meliaceae family (Paritala et al., 2015).

In this family stands out the genus Carapa, whose seeds are rich in triacylglycerols with high contents of unsaturated fatty acids, such as oleic, palmitic, stearic, and linoleic acids, and an unsaponifiable fraction composed mainly of limonoids, which are chemotaxonomic markers of the Meliaceae family (Ambrozin et al., 2006, Cabral et al., 2013). Its metabolic diversity has resulted in an intensification of chemical studies carried out in several species of this genus, mainly in Carapa guianensis Aublet, known in Northern Brazil as “Andiroba” (Santos et al., 2021).

Andiroba is a neotropical tree, widely distributed throughout the Amazon rainforest, specifically located in the soil-rich regions and swamps of countries such as Brazil, Peru, Ecuador, Colombia, Venezuela, French Guiana, and Suriname, along with several islands in the Caribbean. In the Brazilian Amazon, it is found mainly in seasonal floodplains and wetlands along “igapós” (flooded forests). It is an easily cultivable tree and, therefore, has the potential to recover degraded lands (Ferraz et al., 2002).

The present work aimed to carry out a bibliographical survey regarding the ecological and pharmacological aspects of Carapa guianensis, reviewing the production and the use of andiroba oil and the pharmacological activities of its limonoids, also evaluating its potential as an adaptogenic plant, besides reporting the sustainable use of its by-products.

Methods

This study comprises a non-exhaustive literature review, including works published from January 2002 to August 2022, obtained from different databases (SciFinder, PubMed, Scopus, and Web of Science). The keywords used were: “Carapa guianensis and limonoids”; “Carapa guianensis and oil”; “Carapa guianensis and biomass”; “Carapa guianensis and by-products”; “Carapa guianensis and cancer”; “Carapa guianensis and antioxidant”; “Carapa guianensis and waste”; “Carapa guianensis and toxicology” (Fig. 1).

Fig. 1.

Workflow of the bibliographical research carried out for the conception of this work.

Eligibility criteria were reviews and full publications of in vitro or in vivo pharmacological studies or about the production methods and use of andiroba oil and its by-products and waste, written in the English language. Approximately 389 references came up on the search. After excluding duplicates, 255 references remained and 119 were selected for this review.

Carapa guianensis seed oil: traditional use

Andiroba oil is well known in Brazil and widely employed to heal many skin conditions and is considered a natural insect repellant. The indigenous people from the Amazon rainforest have used different parts of C. guianensis plants, such as barks, flowers, leaves, and seeds, for the treatment of many health conditions. According to Hammer & Johns, 1993, flowers and bark, tea preparations can be used for bacterial and other infections, while hardwood can be used as a fungicide.

The seed oil has a long history of use by natives of the Amazon and in traditional medicine in South America (Silva, 2018). This oil is well-known and widely used in Brazilian folk medicine, especially by the inhabitants of the Amazon rainforest, since they are already familiar with the healing effects of Andiroba oil (Silva et al., 2010). It has been indicated in the literature that all parts of the Andiroba tree are utilized in traditional medicine. However, Andiroba is recognized in traditional medicine and has high anti-inflammatory and analgesic potential including the seed's oil, which is employed to treat inflammation and infections (Soares et al., 2021). Andiroba oil consists mainly of saponifiable fatty acids derivatives, especially oleic, palmitic, stearic, and linoleic acid, as well as many vitamins and minerals, as well as 5% unsaponifiable material, including several tetranortriterpenoids metabolites known as limonoids, such as 17 β-hydroxyazadiradione, xyloccesin k, gedunin, 6-α-acetoxygedunin, 7-desacetoxy-7-oxogedunin, 1,2-di-hydro-3β-hydroxy-7-desacetoxy-7-oxogedunin and methyl angolensate. Chemical analysis of Andiroba oil has attributed the anti-inflammatory, antiparasitic and insect-repellent properties to the presence of limonoids (Henriques and Penido, 2014). The plant species is also used in traditional rituals. The Mundurucus, who inhabit indigenous areas in the Southwest of Pará, east of the state of Amazonas, and west of Mato Grosso, used the andiroba oil to mummify the heads of enemies. The Wayãpi and Palikur people use this oil to treat ticks and lice parasites (Machado and Mendes, 2021). Another topical use has its healing action which helps to reduce the appearance of cellulite, where the massage with andiroba oil in the place where you want to reduce cellulite. It can also be used as a hair toner on the scalp, where it helps both to reduce excess oil and also to reduce dandruff (Milhomem-Paixão et al., 2016). This oil is suggested to exert a wide range of described healing properties related to anti-inflammatory effect (Nayak et al., 2011). These indications have been supported by the literature (Sousa et al., 2021), and may contribute as an agent in pain reduction, acting on the inflammation process of conditions such as rheumatism and arthritis.

Moreover, andiroba oil has several other uses in folk medicine, as described by ethnopharmacology (Table 1).

Table 1.

Ethnopharmacology of Carapa guianensis seed oil.

Traditional uses	References
Insect repellent	Dantas, 2009, Freire et al., 2006, Hammer and Johns, 1993, Mendonça and Ferraz, 2007, Nardi et al., 2016, Oliveira et al., 2003.
Wounds in general, insect bites	Hammer and Johns, 1993, Mendonça and Ferraz, 2007, Nardi et al., 2016.
Antiparasitic (ticks, lice, and other parasites)	Brito et al., 2020, Mendonça and Ferraz, 2007, Miranda-Júnior et al., 2012
Inflammation, contusion, bone pain, catarrh, sore throat, wounds	Van den Berg, 1982, Van den Berg, 1984, Coelho-Ferreira, 2009, Nardi et al., 2016.
Arthritis and Rheumatism	Hammer & Johns, 1993.
Antipyretic and Analgesic	Mendonça and Ferraz, 2007, Miranda-Júnior et al., 2012
Hypoglycemic (diabetes)	Mendonça & Ferraz (2007).
Snake, scorpion, and bee stings	Mendonça & Ferraz (2007).

In Brazil, Andiroba oil is easily found in drugstores, compounding pharmacies, natural products, and cosmetics stores (Nonato et al., 2018). It is sold mainly in open markets in the Northern region of Brazil and markets in the other areas of the country, as well as in the supply of the export trade, being supplied as an input for pharmaceutical and cosmetic industries in Europe and the USA (Mendonça and Ferraz, 2006). Everything from Andiroba is used, and the dough left over from oil extraction is used to make soap, candles, and soap (Moraes et al., 2019). Andiroba oil produces insect repellent candles (Lourenço et al., 2017) and can be used in the cosmetic, food, and textile industries (Amaral & Fierro, 2013). In the treatment of the skin, it has a moisturizing property and thus helps by stimulating its regeneration (Melo et al., 2021). Andiroba oil can be applied topically several times daily to rashes, muscle/joint aches and injuries, wounds insect bites, boils, and ulcers (Nayak et al., 2011). Thus, it can produce shampoos, body oils, perfumes, personal care, and beauty products (Galdino, 2007, Djenontin et al., 2012) and make functional foods or dyes for the textile industry (Sousa et al., 2019).

Adaptogenic and toxic potential effects of Carapa guianensis

The history of using natural adaptogens for human beings is very valuable because they have been used in recovering from physical weakness and illness and enhancing mental functions and memory, among other functions (Todorova et al., 2021). In current medicine, the term adaptogen is related to the compound that promotes nonspecific improving resistance in the live organism and its adaptive ability to stress agents, as a result of interacting with many different natural metabolites, exerting some health benefits to animals and humans (Oliveira and Leitão, 2016, Panossian et al., 2021). Adaptogenic plants have compounds that can promote non-specifically enhancement in resilience and survival in affected cells of different organs and tissues, restoring homeostasis, through the activation of many signaling pathways. Their properties include anti-fatigue, physical endurance enhancement, and immunomodulatory, cardiovascular, anticancer, and antioxidant properties (Liao et al., 2018; Esmaealzadeh et al, 2022).

Andiroba oil and limonoids isolated (6α-acetoxygedunin, 7-deacetoxy-7-oxogedunin, andirobin, gedunin, and methyl-angolensate) have shown a high anti-inflammatory and anti-allergic activity, causing in vivo inhibition on edema in rodents, by depletion of signaling pathways that promote platelet-activating factor bradykinin, and histamine. These compounds also inhibited the inflammatory mediators: eosinophilotactic mediators CCL11/eotaxin and interleukin (IL)-5 and the inflammatory cytokines tumor necrosis factor IL-1β and (TNF)-α. This effect depends on κB (NFκB) activation. Moreover, these limonoids have reduced many populations of cells, such as T lymphocytes, eosinophils, and mast cells, by which they impaired allergy and inflammation (Penido et al., 2006, Henriques and Penido, 2014).

Lemes et al. (2017) evaluated the in vivo protective effect of this oil on mitomycin C (MMC) and cyclophosphamide (CP), two alkylating chemotherapeutic agents. The results also showed that the andiroba oil had cytotoxicity nor genotoxicity to mouse bone marrow at concentrations of 250, 500, and 1,000 mg/Kg. Further, have shown the ability to modulate the genotoxic activity of MMC and CP, demonstrating its antigenotoxic effect protective in all these concentrations.

Novello et al. (2015) have shown that oil extraction with n-butane (45 min at 25 °C/13 bar), has no residues of this solvent in the final product. This oil has palmitic and oleic acids as the major components and an antioxidant activity of 22.2%, and it was a promising alternative to conventional extraction methods. In another study, an ethanolic extract from this plant also has shown antioxidant activity (IC₅₀ = 7.80 ± 0.002) and a significant statistical correlation with its phenols content quantified by Folin-Ciocalteu (Pereañez et al., 2010). Moreover, an aqueous extract from andiroba barks also had shown some antioxidant activity that was related to their total polyphenol content (TPC), quantified by Fe(bipy)₃²⁺ complexes method (Santana et al., 2015). Moreover, the hydrolyzed extract from the agro-industrial by-product of Andiroba also has shown antioxidant activity similar to Trolox (positive control), evaluated by 2,2-diphenyl-1-picrylhydrazyl (IC₅₀ 16.42 mg/mL) and 2,2′-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (IC₅₀ 6.52 mg/mL). This activity was related to its phenolic acids identified by HPTL and GC–MS (dos Santos et al., 2021).

Plant-originated adaptogens can reduce the risks of cancers as have been reported in studies of this area and they have been got an emphasis on cancer chemotherapy (Arushanyan, 2009, Panossian et al., 2021, Sulaiman and Lakshmanan, 2022). These adaptogens have an important role in cancer prevention and treatment, stabilizing and reversing precancerous conditions or suppressing carcinogenic processes. They can promote selective optimization of gene behavior by protecting and repairing cells or even improving apoptosis in damaged cells, preventing genetic mutations that could conduce to cancer cell development. Moreover, they have antitumor properties, inhibiting invasion and showing anti-metastatic and antiangiogenesis action. When the cancer is active, it can work as an adjunctive therapy to chemotherapy, and radiotherapy, increasing the body’s resistance to radiation effects and other stresses (Yance, 2013). Thus, adaptogens trigger many systemic effects, with complex biological functions that depend on various cancers, revealing many possibilities for personalized therapy and cancer chemoprevention in the future (Sulaiman and Lakshmanan, 2022).

Andiroba oil is suggested to promote apoptosis in a gastric cancer cell line (ACP02). The highest concentration of this oil induced apoptosis at 24 and 48 h, without showing mutagenic effects (no significant enhancement in the frequency of micronuclei) (Porfirio-Dias et al., 2020). Andiroba gel effectively also is showed to decrease the severity of oral mucositis and alleviated pain symptoms in 60 children (0–12 years) submitted to chemotherapy. Compared to laser, this gel has shown a faster response and a more accessible cost, proving to be a promising treatment. Further, it clarifies the cytological action of andiroba on oral mucositis (Cavalcante, 2018).

In vivo studies on andiroba oil (2,000 mg/Kg/14 days) and its nanoemulsion (200 mg/Kg/14 days) have shown the ability to reduce the collateral effects of doxorubicin (40 mg/Kg), an antineoplastic agent used to treat many cancers. Both diminish hematotoxicity, histological alterations in the kidneys and liver, and the frequency of apoptosis. In general, nanoemulsion had a most substantial effect than andiroba oil because nanotechnology enhances the lipid pharmacokinetics in the human body (Melo et al., 2021). According to the data exposed, Andiroba oil is suggested to exert antioxidant and anti-inflammatory properties, besides its apoptotic action in cancer cells and the protective effect in chemotherapy presenting very low or normally no cytotoxicity in vivo.

Andiroba oil and potential productive chain

In Brazil, the last survey bioeconomic socio-study carried out showed production of 115.5 tons, distributed in the states of Amazonas, Manaus, and Pará, with emphasis on the states of Amazonas and Pará, which responded for about 87% of the national production, and the state of Maranhão with the remaining 13% (CONAB, 2022).

In the last two centuries, fossil fuel use has grown up significantly. A relevant alternative to these fossil fuels is biodiesel: a liquid fuel obtained from pure vegetable oil, oils from animal fats, algal cultures, grease, and cooking oil waste (Guo et al., 2015). In the world, 84% of energy is originated from non-renewable sources (Forbes, 2020). In Brazil, about 48% of the energy is renewable, and is a world pioneer in biofuels (ethanol and biodiesel), having reached a position desired by many others that seek renewable energy sources as strategic alternatives to oil (ANP, 2022). Andiroba oil has physical properties, like soybean and cottonseed oils, such as density, viscosity, and calorific value. This makes it a viable alternative for producing biodiesel, as an alternative fuel, for powering electric generators in isolated communities in the Amazon Forest (Carvalho, 2011, Cabral et al., 2013, Iha et al., 2014).

Industrial biodiesel production is performed by transesterification of vegetable oils with methanol by homogeneous catalysis to carry out the cleavage of triglycerides and result in a mixture of many types of fatty acid alkyl esters (Leung et al., 2010). Melo et al., 2014 studied different oil sources, such as nuts and seeds, to generate biodiesel using accelerated oxidation techniques to evaluate their oxidative stability and physical–chemical properties. The results had shown that andiroba oil has more oxidative and thermal stability, presenting a considerable potential for biodiesel production. The enzymatic transesterification by using ethanol instead of methanol is also considered a sustainable and eco-friendly alternative in the biodiesel production process. Thus, Carvalho et al., 2013 have evaluated the potential ethanolysis of andiroba crops industrial waste. These reactions are performed with the microbial lipase from Burkholderia cepacia, held on a silica-polyvinyl alcohol matrix in a solvent-free system. The results showed efficient conversions to ethyl esters in all non-edible raw materials. In another study, Ferreira et al., 2018 evaluated andiroba and babassu (Orbignya sp.) oils, potential triacylglyceride sources of biofuel production using lipase B from Candida antarctica (CALB). In this work, the authors obtained a high yield of ethylic biodiesel, characterized by many methods, with promising results.

Andiroba oil extraction and the potential use of the products and By-Products

Andiroba oil is one of the best-selling natural product derivatives in the Amazon region due to its medicinal properties. It is one of the main non-wood forest products exported to the United States and Europe countries, such as Germany, Spain, and France (CONAB, 2017, CONAB,2021). Despite this, there are no large areas for oil extraction carried out by small groups of families in the region (Amaral and Fierro, 2013, Gomes, 2010, Martins et al., 2012).

There are two methods for extraction of andiroba oil: artisanal, carried out by extractors from communities in the North region, which have no electricity, and industrial, which uses mechanical pressing (Sousa et al., 2019). The artisanal process can be divided into three stages: collecting and selecting seeds, preparing the dough, and extracting the oil. Selected seeds are washed and cooked at about 90 °C (Mendonça & Ferraz, 2007). Afterward, placed in the shade to rest for a few days. The oil starts to come off, then the seeds are separated from the husks and pounded in a pestle (Ferraz et al., 2002, Souza et al., 2006, Silva et al., 2012). They are molded into spherical masses and placed in the sun to gradually release the oil by dripping (Souza et al., 2006, Sousa et al., 2019). Due to the lack of standardization in extraction procedures, reports of oil production yield vary widely, and it may be necessary to use 2 to 30 Kg of seeds to obtain one liter of oil (Plowden, 2004, Mendonça and Ferraz, 2007, Silva et al., 2010). The industrial-scale extraction method is carried out without adding organic solvents in a continuous and clean manner (Mendonça and Almeida, 2020). It consists of ding the seeds into smaller pieces, which are placed in an oven at 60–70 °C until they reach 8% humidity. In sequence, it is pressed in the hydraulic press at a temperature of 90 °C (Souza et al., 2006). However, according to Mendonça and Almeida (2020), the pressing is cold. This last process results in an amount of oil rarely exceeding 30% of the seed weight with 8% moisture (Ferraz et al., 2003, Souza et al., 2006).

Nowadays, there is a strong interest in by-products obtained from raw materials from sustainable sources, which can generate a great benefit due to their many applications, such as cosmetics, food supplements, and medicines. In this way, there is the emergence of added value to a by-product that initially had no economic value from disposable raw material natural products. Thus, various technological applications have emerged for these materials, generating agro-industrial derivatives from bark and seeds since they have many bioactive compounds (Castro et al., 2018, Borges et al., 2019). Besides their wastes, Brazilian crops such as andiroba nut and buriti should be explored further for technological applications in the chemicals and materials industries (Alarcon et al., 2020). It is known that the industrial-scale processing of Andiroba seeds to obtain oil generates a significant amount of by-products. About 66% w/w, which is called cake and is a potential source of a number of active metabolites, such as simple phenolic structures and tetranortriterpenoids, the limonoid metabolites (dos Santos et al., 2021). In this process, the oil obtained has many emollient compounds and is considered a high-quality material in the cosmetic industry because this process allows greater efficiency in lipid extraction (Uitterhaegen and Evon, 2017). In this cold pressing are produced 30% w/w of extra virgin oil and 60% w/w remains as a residual cake. Compounds of higher molecular mass, as such polysaccharides, proteins, bound lipids, and even some phenolic compounds, stay retained in the cellular seed wall of this cake. Unfortunately, only a tiny part of this cake is used to produce repellent candles. The rest has been used in composting, as a pig feed supplement, or discarded in the environment (Ferraz et al., 2002, Mendonça and Almeida, 2020). In the artisanal form of extraction, the residents from the communities surrounding the houses can burn the residual cake, as the smoke produced is not irritating to humans and scares away hematophagous insects due to the repellent properties of the andiroba (Brito et al., 2020). Costa-Silva et al. (2008) suggested that andiroba oil can be used as an alternative treatment against ectoparasites with application on infected cow’s horn flies. It was reported that the oil does not appear to have a toxic effect able to inhibit microbial activity. However, they may increase the respiratory activity of soil microorganisms, which may contribute to their rapid degradation. According to the authors, the Andiroba oils showed no risk to springtails and soil microbial population, well-known terrestrial toxicity indicators.

Some studies, e.g., Lira et al. (2021), aim to add value to the production chain of andiroba oil by identifying bioactive compounds in this residual biomass and directing it to food, pharmaceutical, or cosmetic applications of economic benefit. Silva et al. (2009) have isolated six limonoids from the by-products of oil production of andiroba seeds by high-speed countercurrent chromatography (HSCCC). In another study, Pereira et al. (2014) studied the in vitro and in vivo antiplasmodial action of four limonoids isolated from andiroba residual biomass by extraction and chromatography techniques. These limonoids showed promissory inhibitory activity in these. These studies have shown the by-product valorization of andiroba oil extraction, which had not yet been economically used relevant added value to the andiroba seed waste. The section “Pharmacology of limonoids isolated from Andiroba oil” shows this in more detail. Thus, it was possible to isolate limonoids from the by-products originated after oil production from andiroba seeds, in an excellent relative abundance, compared with classical chromatographic methods. Santos et al. (2021) have dehydrated the cake of andiroba seeds obtained by mechanical pressing. This dehydrated cake was extracted by solvent, hydrolyzed, and subsequently characterized by GC–MS. The hydrolyzed extract has 21 compounds, and of these, 13 were identified as phenolic acids. These acids are used in the cosmetic, food, or pharmaceutical industries. The other compounds are substances ordinarily present in andiroba oil, such as steroids, limonoids, and fatty acids. This study has shown which interesting compounds are still present in biomass, even after industrial extraction.

In prospecting for bioactive extracts, the strategies of isolation of major compounds from known species can be extremely effective in generating bioproducts. The residual parts of agricultural production are resources of great value since it still preserves high amounts of bioactive compounds. In Brazil, many bioactive waste raw materials are usually discarded by extractivist, by small producers, and could be used as alternative sources for the development of new active drugs. A literature survey described how these by-products are commonly disposed of inappropriately in nature, causing environmental problems and imbalance, and present substances that may be of interest to the industry (Veiga Junior and Yamaguchi, 2022). In this context, Serafin et al., 2021 proposed the production of activated carbons prepared from waste-derived biomass of C. guianensis and other Amazonian fruit to stock CO₂ at low pressure, reducing global warming. The method used was carbonization in under-flowing N₂ flow atm and activation with KOH. Among the samples studied, the activated carbon derived from andiroba seeds has shown the highest CO₂ adsorption under the pressure of 1 bar.

Limonoids isolated from Andiroba oil

Regarding the compounds characterized by Andiroba, a number of biologically active tetranortriterpenoids metabolites attributed as “limonoids” are presented in the oil. The limonoids already described in the literature, isolated or identified are listed below, separated by type:

-Gedunin-type limonoids: gedunin (1); 6α-acetoxygedunin (2); 1,2-dihydro-3β-hydroxy-7-deacetoxy-7-oxogedunin (3); methyl angolensate (4); 17β-hydroxyazadiradione (5) (Ambrozin et al., 2006); 11β-hydroxygedunin* (6); 6α,11β-diacetoxygedunin*(7); deoxygedunin* (8); 17-glycolyldeoxygedunin* (9) (Oliveira et al., 2018); deacetylgedunin (10); 7-deacetoxy-7-oxogedunin (11) (Silva et al., 2009); 6α-hydroxygedunin (12); andirobin (13); 6α-acetoxyepoxyazadiradione (14) (Pereira et al., 2014); andirolide H (15) (Morikawa et al., 2018); carapanosin C (16); 17-epi-17-hydroxyazadiradione (17) (Higuchi et al., 2017a); carapanolide J (18) (Matsui et al., 2014); 6-hydroxymethyl angolensate (19) (Peron, 2017); epoxyazadiradione (20) (Ninomiya et al., 2016) and carapanin A (21) (Kikuchi et al., 2020) (Fig. 2).

Fig. 2.

Gedunin-type limonoids from Carapa guianensis.

- Mexicanolide-type limonoids: xyloccensin K (22) (Ambrozin et al., 2006); carapanolides A (23), B (24) (Inoue et al., 2012); C (25) (Morikawa et al., 2018); D (26), E (27), F (28), G (29) (Inoue et al., 2014); R (30); S (31); T (32) and U (33) (Miyake et al., 2015); carapanosins E (34) and F (35) (Inoue et al., 2018), and carapanins B (36) and C (37) (Kikuchi et al., 2020) (Fig. 3).

Fig. 3.

Mexicanolide-type limonoids from Carapa guianensis.

- Phragmalin-type limonoids: guianolides A (38) and B (39) (Inoue et al., 2013); carapanolides I (40) (Inoue et al., 2014); K (41); L (42) (Matsui et al., 2014); M (43); N (44); O (45); P (46); Q (47) (Inoue et al., 2015); V (48); W (49) and X (50) (Miyake et al., 2015); carapanosins A (51) and B (52); carapanolide H (53); swietephragmins D (54) and G (55) (Higuchi et al., 2017a); carapanosin D (56) (Inoue et al., 2018) and guianofruits E (57); F (58); G (59); H (60) and I (61) (Tsukamoto et al., 2019) (Fig. 4).

Fig. 4.

Phragmalin-type limonoids from Carapa guianensis.

- Chukrasone-type limonoids: guianofruits A (62); B (63) (Sasayama et al., 2018); C (64) and D (65) (Tsukamoto et al., 2019). There are also two compounds, guianolactones A (66) and B (67), which are classified as “rearranged pentacyclic limonoids” (Higuchi et al., 2017b), due to their skeletons containing non-standard cleavages and rearrangements, (Fig. 5).

Fig. 5.

Chukrasone-type and rearranged pentacyclic limonoids from Carapa guianensis.

Pharmacology of limonoids isolated from Andiroba oil

Many in vitro and in vivo pharmacological studies were carried out with limonoids isolated from Carapa guianensis seed oil. As previously described, the oil is widely traded worldwide and has been evaluated for the anti-inflammatory and cytotoxic effects of tetraterpenoids and limonoids isolated from C. guianensis. All the results in vitro are shown in Table 2, and the in vivo results are shown in Table 3.

Table 2.

In vitro pharmacological studies of limonoids isolated from andiroba seeds/oil.

Activity	Compounds	Results	References
Murine L1210 leukemia cell line	Carapanolide A	IC50: 8.7 µM	Inoue et al. (2012)
	Carapanolide F	IC50: 15.9 µM	Inoue et al. (2014)
Murine P388 leukemia cell line	Gedunin	IC50: 16.0 µM	Sakamoto et al. (2013)
	6α-acetoxygedunin	IC50: 11.4 µM
	Deacetylgedunin	IC50: 11.2 µM
Human HL-60 leukemia cell line	Gedunin	IC50: 15.2 µM	Sakamoto et al. (2013)
	6α-acetoxygedunin	IC50: 10.4 µM
	Deacetylgedunin	IC50: 10.4 µM
	Carapanolide D	IC50: 11.0 µM	Inoue et al. (2014)
Antiplasmodial against P. falciparum FCR-3 strain (ATCC 30932, chloroquine-sensitive)	Gedunin	EC50: 2.5 µM	Tanaka et al. (2012)
	6α-acetoxygedunin	EC50: 2.8 µM
	Methyl angolensate	EC50: 15 µM
	7-deacetoxy-7-oxogedunin	EC50: 2.5 µM
	6α-hydroxygedunin	EC50: 9.0 µM
	Andirobin	EC50: 12 µM
	6α-acetoxyepoxyazadiradione	EC50: 4.5 µM
	Andirolide H	EC50: 4.0 µM
Antiplasmodial against P. falciparum K1 strain	6α-acetoxyazadiradione	IC50 = 15.4 µM	Pereira et al. (2014)
	andirobin (2)	IC50 = 15.3 µM
	6α-acetoxygedunin (3)	IC50 = 7.0 µM
	7-deacetyl-7-oxogedunin (4)	IC50 = 20.7 µM
Inhibitory effects on LPS-induced NO production in murine peritoneal macrophages	Carapanolide J	IC50: 37.4 µM	Matsui et al. (2014)
	Epoxyazadiradione	IC50: 5.9 µM
	Carapanolide T	IC50: 22.0 µM	Miyake et al. (2015)
	Carapanolide U	IC50: 23.3 µM
	Gedunin	IC50: 4.6 µM	Sakamoto et al. (2015)
	6α-acetoxygedunin	IC50: 7.9 µM
	Deacetylgedunin	IC50: 8.4 µM
	7-deacetoxy-7-oxogedunin	IC50: 12.8 µM
	6α-hydroxygedunin	IC50: 19.1 µM
	Gedunin	IC50: 4.6 µM	Ninomiya et al. (2016)
	6α-acetoxygedunin	IC50: 7.9 µM
	Epoxyazadiradione	IC50: 8.2 µM
	7-deacetoxy-7-oxogedunin	IC50: 12.8 µM
	17β-hydroxyazadiradione	IC50: 10.8 µM	Higuchi et al. (2017a)
	Carapanosin C	IC50: 13.7 µM
	Swietephragmins D	IC50: 4.9 µM
	Carapanosin E	IC50: 23.9 µM	Inoue et al. (2018)
	Carapanosin F	IC50: 11.8 µM
	Carapanin B	IC50: 12.6 µM	Kikuchi et al. (2020)
	Carapanin C	IC50: 29.5 µM
Effects on collagen synthesis-promoting in normal human dermal fibroblasts	Gedunin	Inhibition: 133.3%±3.6 (3 μM)	Morikawa et al. (2018)
	6α-acetoxygedunin	Inhibition: 152.8%±6.8 (100 μM)
	Methyl angolensate	Inhibition: 114.5 ± 2.9 (30 μM)
	Deacetylgedunin	Inhibition: 119.4 ± 3.1 (3 μM)
	7-deacetoxy-7-oxogedunin	Inhibition: 129.5%±3.0 (30 μM)
	6α-hydroxygedunin	Inhibition: 121.4 ± 2.7 (30 μM)
	Carapanolide C	Inhibition: 123.0 ± 3.5 (10 μM)
Inhibitory effect on hepatitis C virus	Deacetylgedunin	Entry: EC50 = 41.03 ± 3.55 μM Replication: EC50 = 27.88 ± 1.97 μM Assembly: EC50 = 44.22 ± 3.72 μM	Kouan et al. (2021)
	7-deacetoxy-7-oxogedunin	Entry: EC50 = 50.54 ± 4.24 μM Replication: EC50 = 24.81 ± 2.66 μM Assembly: EC50 = 34.30 ± 4.24 μM

EC₅₀ = half efficient concentration; IC₅₀ = half inhibitory concentration.

Table 3.

Pharmacological investigations on in vivo experiments using limonoids isolated from andiroba seeds/oil.

Activity	Compounds	Result	References
Effects on the armyworm, Spodoptera frugiperda	7-deacetoxy-7-oxogedunin	Larval phase duration: 16.3 (±1.18) days Pupal weight: 244.7 mg Larval mortality: 33.3%	Sarria et al. (2011)
Suppression of Plasmodium berghei Suppression in infected mice	6α-acetoxygedunin	% Parasite inhibition (5 days of treatment) Oral: 65.7 Subcutaneous: 44.2 IC50 = 7.0	Pereira et al. (2014)
Phytotoxicity	Gedunin	IC50 = 2.64 μM	Nebo et al. (2015)
Insecticidal, antifeedant on Reticulitermes speratus	Gedunin	PC95 = 218.4 μg/disc (30 d)	Lin et al. (2021)
	17β-hydroxyazadiradione	PC95 = 235.6 μg/disc (30 d)
	Deacetylgedunin	PC95 = 113.7 μg/disc (30 d)
Insecticidal, antifeedant on Plutella xylostella	Epoxyazadiradione	AR = 37.2% at 2000 μg/mL (48 h)
Insecticidal, antifeedant on Spodoptera littoralis	7-deacetoxy-7-oxogedunin	AFD at 1000 μg/mL (3–10 h)
	Xyloccensin K	AFD at 1000 μg/mL
Insecticidal, antifeedant on Spodoptera litura	Methyl angolensate	PFI = 65.3 at 1 μg/cm2 (24 h)
Insecticidal, poisonous on Atta sexdens rubropilosa	6α-acetoxygedunin	S50 = 8 d at 100 μg/mL
	1,2-dihydro-3β-hydroxy-7-deacetoxy-7-oxogedunin	S50 = 9 d at 100 μg/mL
	Deacetylgedunin	S50 = 9 d at 100 μg/mL
	7-deacetoxy-7-oxogedunin	S50 = 11 d at 100 μg/mL
Insecticidal, poisonous on Spodoptera frugiperda	Gedunin	LC50 = 39 μg/mL (7 d)
	Methyl angolensate	MR: 40% at 50 mg/Kg (7 d)
Insecticidal, growth regulatory on Ostrinia nubilalis	6α-acetoxygedunin	reduced growth at 50 μg/mL

PC₉₅ = 95% protective concentrations (μg/disc); AFD = antifeedant activity; PFI = percentage feeding index; S₅₀ = survival average 50%; LC₅₀ = lethal concentration 50%; MR = mortality rate.

In vitro studies

The following mechanisms of hepatoprotective action of 17 limonoids were evaluated by the in vitro studies: i) D-GalN-induced cytotoxicity in primary cultured mouse hepatocytes; ii) LPS-induced NO production in mouse peritoneal macrophages and iii) tumor necrosis factor (TNF-α)-induced cytotoxicity in L929 cells. This study showed that none studied limonoids decreased the cytotoxicity caused by D-GalN in mouse hepatocytes. However, 1, 2, 11, and 20 showed a higher inhibitory effect in NO production without remarkable cytotoxic effects at the effective concentration. According to the work, this effect was more substantial in gedunin-type limonoids. For this class of limonoids, it was observed that compounds with a 7 α-acetoxy have more potent activity than those with 7 α-hydroxy or 7- α-keto group groups. Compounds with an α,β-epoxy-δ-lactone in the D-ring displayed more activity than those with an α,β-epoxy, or α,β -unsaturated cyclopropane group. The limonoids 5, 11, 20, 23, and 25 presented an antiproliferative effect in L929 cells, a TNF-α-sensitive cell line (Ninomiya et al., 2016).

In the study performed by Ferraris et al. (2011), in vitro incubation of eosinophils with limonoids 1, 2, 4, 11, and 13 resulted in the inhibition of eosinophil adhesion to tumor necrosis factor-α (TNF-α)-primed tEND.1 endothelial cell. Other experiments conducted proved that these limonoids also blocked T-cell activation. Besides that, all the limonoids tested, except 2, also promoted downregulation of nuclear factor-κB (NFκB) nuclear translocation in OVA-challenged splenocytes. The analysis of these results allows concluding that the antiallergic effects of these limonoids depend on their potential to suppress T lymphocyte activation and eosinophil migration (results not shown in tables).

The limonoid compound 23 (Inoue et al., 2012) and 28 (Inoue et al., 2014) showed in vitro anticancer activity on murine leukemia cells (L1210). The first compound presented high cytotoxicity with IC₅₀ = 8.7 µM, and the latter showed moderate cytotoxicity. In the study carried out by Sakamoto et al. (2013), 1, 2, and 10 were evaluated and showed a high cytotoxic activity in leukemia cell lines, murine (P388), and human cells (HL-60). For this latter cell line, 26 exhibited a strong inhibitory effect (Inoue et al., 2014).

Tanaka et al. (2012) have reported the antimalarial activity of compounds 1, 2, 4, 11, 12, 13, 14, and 15 against the FCR-3 strain of Plasmodium falciparum (ATCC 30932, chloroquine-sensitive). In the assays, 15 showed the best in vitro antimalarial activity due to its better selectivity index.

Compounds 18, 20, 41, and 42 were evaluated for their inhibitory effect on NO production in LPS-stimulated RAW264.7 cells (mouse peritoneal macrophages), and the cytotoxicity of these limonoids was evaluated to determine safe concentrations. The limonoid 18 showed an inhibitory effect on the NO production in concentrations (3–100 μM) and no cytotoxicity. Compound 20 showed higher inhibitory activity on NO production at 3 and 10 μM, considered non-toxic concentrations, having an IC₅₀ = 5.9 μM. Compound 41 had shown inhibitory activities on NO production but also presented cytotoxicity, while compound 42 showed no cytotoxic or significant effect inhibitory on this assay. These results concluded that 18 could be a promising inhibitor of NO production (Matsui et al., 2014).

In other similar works, using the same model, compounds 1, 2, 10, 11, 12 (Sakamoto et al., 2015), 5, 16, 54, (Higuchi et al., 2017a), 32, 33 (Miyake et al., 2015), 34, 35 (Inoue et al., 2018), 36 and 37 (Kikuchi et al., 2020) were considered non-toxic and exhibited similar or superior inhibitory activities compared to NG-monomethyl-L-arginine (L-NMMA), a nitric oxide synthase inhibitor. These results suggest that the compounds have the potential to act as anti-inflammatory agents.

Inoue et al. (2015) investigated triglyceride (TG) metabolism-stimulating activities in human hepatocellular carcinoma cells (HepG2) pretreated with high concentrations of glucose. According to the authors, among the twelve limonoids tested, only compounds 1, 10, and 11 significantly reduced TG levels in hepatocytes and demonstrated potential as preventive agents for hepatic steatosis (results not shown in tables).

In the study carried out, by Pereira et al. (2014), in residual seed biomass, the limonoids 2, 11, 13, and 14 were evaluated in vitro on the multi-drug-resistant Plasmodium falciparum K1 strain, with IC₅₀ = 7.0–20.7 μM and the compound 2 has the most antiplasmodial activity (Table 2). They showed a moderate inhibitory activity and were not cytotoxic to MRC-5 human fibroblasts. The limonoids 2 and 11, isolated in more amounts, were also studied in vivo, and the results and details are shown in the section on the subject and Table 3.

In another study, Morikawa et al. (2018) examined ten-limonoidś collagen synthesis-promoting in normal human dermal fibroblasts (NHDFs). Among these, 1, 2, 10, 4, 11, 12, and 25 have significantly promoted collagen synthesis without cytotoxicity at effective concentrations. Compounds 1 and 10 have the most potent activity in this group. Besides that, for gedunin-type limonoids, regarding the collagen synthesis-promoting activity in NHDFs, the results showed that 6α-acetoxy and 6α-hydroxy groups decrease the activity. Moreover, the limonoids with the 7α-acetoxy group have more potent activity than 7α-hydroxy or 7-keto substituents, and compounds with an α,β-epoxy-γ-lactone moiety in the D-ring have shown to be more potent activity than α, β-unsaturated cyclopentanone moiety.

The study developed by Kouam et al. (2021) evaluated the inhibitory effect of limonoids 10 and 11 on the hepatitis C virus (HCV). The compounds inhibited HCV infection, mainly on entry (via CD81 receptor blockade) and replication (by decreasing NS5B expression) stages. In addition, they also increased the expression of Class-III phosphatidylinositol 4-kinase alpha (the newly identified major host cell factor that regulates HCV replication) and 2′,5′-oligoadenylate synthase-3. The results indicate the possibility of using 10 and 11 as alternative therapy against HCV infection.

In vivo studies

Limonoids isolated from C. guianensis were evaluated against the armyworm of Spodoptera frugiperda (J. E. Smith), one of the major pests of many crops in the Americas and tropical maize. The limonoids 11 and β-photogedunin have promoted a reduction in the pupal weight of 17.8 and 31.5 mg at 50 mg/Kg, respectively. Compound 11 has prolonged the time of the larval stage to approximately 1.2 days at 50.0 mg/Kg. It was shown that the limonoid β-photogedunin has a stronger activity (Sarria et al., 2011).

From residual seed biomass, Pereira et al. (2014) isolated a large amount of limonoids 2 and 11 and evaluated them in mice infected with Plasmodium berghei NK65 strain-chloroquine sensitive by oral and subcutaneous treatment for 5 and 7 days. Of these two compounds, limonoid 2 showed a significant antimalarial property in vivo, with suppression of parasitemia of 65.7%, at oral doses of 100 mg/Kg/day. The oral doses of this limonoid have conducted stronger parasite suppression than subcutaneous injection. Oral and subcutaneous doses of 11 have an effect on dose response.

In another study, Nebo et al. (2015) demonstrated the phytotoxic activity of compound 1 which was tested in a wheat coleoptile assay and further tested on Lycopersicon esculentum (tomato), Lepidium sativum (cress), Allium strain (onion) and Lactuca sativa (lettuce). The phytotoxic activity was, for some parameters evaluated, comparable to that of Logan, a commercial herbicide. The results obtained have indicated that 1 and 5α,6β,8α,12α-tetrahydro-28-norisotoonafolin has a high phytotoxic activity and may be involved in ecological interactions between these plant species.

Matsumoto et al. (2019a) elucidated the molecular mechanism of adipogenesis suppression in adipocytes by the limonoid 11. The compound reduced intracellular triglyceride accumulation in a dose-dependent manner. Its action occurred in the early phase of adipocyte differentiation through repression of insulin receptor-1 (IRS-1)/Akt substrate-mediated glucose uptake in adipocytes. In another study of the same group, this compound also had its anti-obesity effect examined in high-fat diet (HFD)-fed mice. The results have shown the reduction of body weight gain and an increase in serum triglyceride level and insulin sensitivity (Matsumoto et al., 2019b). Both results are not shown in tables.

Braga et al. (2020) listed many biological properties associated with compound 1 and their mechanisms of action. The activities described were anticancer (ovarian, breast, cervical, colon, oral, stomach, pancreatic, prostate, lung, brain, stem cells, and skin cancers and leukemia), neuroprotective, antidiabetic, antiallergic, insecticidal, herbicidal, antifeedant, nematicidal, antibacterial antifungal, antiparasitic and others. The work also highlights the importance of the discovery that compound 1 is able to inhibit the heat shock protein 90 (Hsp90). This is one of the most abundant proteins in cells and is involved in the activities of several oncogenic proteins associated with neurodegenerative, vascular, and metabolic diseases. The results are not shown in the tables.

Lin et al. (2021) evaluated 102 triterpenoids from 19 Meliaceae plants regarding their insecticidal potential against 29 insect species. Among these, 14 limonoids found in andiroba seeds/oil have shown insecticidal activity against 7 of the insect species tested. The mechanisms of action studied were based on anti-feeding, poisonous, and growth regulators.

Costa-Silva et al. (2007), in a study approaching the toxic effects of andiroba oil on fertility and pregnancy in Wistar rats and offspring development, concluded that the administration of the oil did not induce maternal toxicity, did not produce an abortifacient effect, and also did not alter the normal development of the offspring nor the behavioral parameters. However, the motor activity, in treated rats has increased, suggesting a possible central action, which should be studied in more detail. In 2007, the same group evaluated liver toxicity and other parameters in pregnancy mice. It was shown that andiroba oil orally in a dose of 2.0 g/Kg/14 days did not produce any sign of toxic effect or death in mice, nor cause alterations in fetus characteristics such as weight. In 2018, they evaluated acute and subacute administration of C. guianensis seed oil and concluded that the andiroba oil did not produce toxic effects in male Wistar rats. However, they noticed an increase in the alanine transaminase (ALT) serum level and absolute and relative liver weights. These last results may indicate possible hepatic toxicity. Miranda-Júnior et al. (2012) also studied the acute toxicity of andiroba oil evaluated by gavage of andiroba oil in mice (2.0 g/Kg), for 14 days. The animals were assessed by a measure of body weight and hematological and biochemical parameters. The results confirmed the traditional use of andiroba oil as antiplasmodial, which also had no toxicity in mice bioassays. Milhomem-Paixão et al. (2016) demonstrated that exposure to andiroba oil did not result in hematotoxic, genotoxic, or mutagenic effects in female Swiss mice. No clinical or behavioral changes were observed during the treatment period.

New approach methodologies in toxicology

Natural products have shown many advantages over synthetic drugs in recent years. For this reason, researchers have turned their attention to exploring their therapeutic potential and plant-derived natural products play an important role as an alternative source for drug discovery (Atanasov et al., 2021, Singh et al., 2023). Great interest from scientific communities and the pharmaceutical industry has emerged in recent years as it has provided or inspired the development of various pharmaceuticals for global health challenges across the world. Consequently, the study of toxicology and the mechanisms involved in these drug candidates has gained prominence in predicting their clinical safety. In the past, preclinical toxicology relied on the relationship between treatment and reported effect to establish the margin of safety for test drugs. More recently, many advances in this subject have empowered researchers in the field of toxicology with more information, including the reduction of side effects, translatability to humans and evolution to safer biomarkers (Pognan et al., 2023).

Currently, these scientific advances have been conducted in the improvement, application and validation of new approach methodologies (NAMs). These methodologies can be defined by any technique or procedure, or combination of them, that can obtain information about risk assessment or chemical hazards. They include in chemico, in vitro, in silico and ex vivo procedures, and are progressively used by regulatory agencies around the world, due to their ability to produce the necessary information, reducing the use of animals (Fig. 6) (Stucki et al., 2022).

Fig. 6.

New approach methodologies (NAM) to avoid animal testing (adapted from Gądarowska et al., 2022).

Some toxic effects occur due to covalent interaction between a drug and biological macromolecules. Chemical assays are available to identify these toxic compounds, which differ widely according to the macromolecule (a peptide) and the analytical technique applied. Thus, the covalent reaction of a molecule with DNA can lead to mutagenesis, with an immunoprotein can cause sensitization, and with an epithelial cell can result in irritation. By understanding these interactions, it will be possible to predict the possibility of toxicity without the use of animal testing. These in chemico methods are not a new in vitro approach because they use physicochemical measurements and do not use biological materials. There are interesting possibilities to obtain information in chemico using in silico methods and providing computational tools for screening purposes. To use these two strategies, it is necessary to integrate their testing strategies to assess individual toxicity parameters (Cronin et al., 2009, Price et al., 2022). In the context of the NMA, Kojima et al. (2023) established a new systematic in vitro methodology using BALB/c3T3 cells, to evaluate compounds with acute oral toxicity (LD₅₀ greater than 2,000 mg/Kg), avoiding tests on animals. This methodology uses Neutral Red Uptake (NRU) cytotoxicity and has been previously validated by the National Toxicology Program (NTP) Interagency Center for the Evaluation of Alternative Toxicological Methods (NTP, 2006). Andiroba oil has healing properties, related to anti-inflammatory activity, and is used in skin treatment, to stimulate its regeneration, as reported in the section “Carapa guianensis seed oil: traditional use” of this work. The acute oral toxicity of this oil has only been performed in animal models. Costa-Silva et al. (2008) administered andiroba oil to Wistar rats orally and concluded that there was no sign of toxicity or death in rats at doses up to 5.0 g/Kg (acute test). In the subacute treatment (1.5 g/Kg), this oil did not promote alterations in the biochemical parameters evaluated, except for an increase in the serum levels of alanine aminotransferase (ALT) in the group, indicative of possible hepatotoxic effects. However, the biochemical toxicity is considered almost null in oral administration. Henriques and Penido (2014) also evaluated the acute and subacute oral toxicity of this oil in Swiss mice. The LD₅₀ was 22.3 g/Kg, higher than fexofenadine (LD₅₀ = 4.5 g/Kg) and dexamethasone (LD₅₀ = 5.8 mg/Kg), both already used in the treatment of acute and chronic allergic reactions. In the context of NMA, it may be a future proposal to develop a study of the acute oral toxicity for this oil, in this new experimental model, using NRU in BALB/c3T3 cells to enhance the mechanisms involved.

The genotoxicity of natural compounds and their derivatives is crucial in determining an assessment for humans. In the current scenario of strong pressure to avoid animal testing, alternative in silico methodologies can promote faster and more cost-effective detection (Kumar et al., 2021). For this purpose, refined techniques of simulations and artificial intelligence (AI) are used along with impressive computational machinery and data analysis tools to model pharmacokinetics, pharmacodynamics and development and testing of toxicological hypotheses (Cronin et al., 2022, Stucki et al., 2022). Many AI-based methods have been created to study chemical-biological interactions associated with oxidative stress and DNA damage using networks, logistic regression models, random forests, deep learning tools and support vector machines (Davidovic et al., 2021). In toxicology, guidelines for systemic bioavailability, physiologically based kinetics (PBK) and multiscale models can be applied to assess internal exposure of organs and/or tissues. To assess the safety of the compound at low concentrations, toxicological concern limits are commonly used. In hazard prediction, the most used in silico modeling methods are the Quantitative Structure-Activity Relationships ((Q)SARs) and the comparative method (Cronin et al., 2022, Stucki et al., 2022).

To predict mutagenicity, new models and machine learning (ML) methods, such as structure–activity relationships and linear statistical methods, have shown remarkable performance and reliability. Thus, the computational prediction of mutagenicity allows evaluating the toxicity profile of the compound, minimizing the time consumed in the last stage of drug development. These prediction methodologies were developed using deep neural network (DNN), support vector machine, k-nearest neighbor and random forest. The DNN model has the highest prediction accuracy (Kumar et al., 2021).

Costa-Silva et al. (2007) studied the toxic effects of Andiroba oil on fertility and pregnancy in female Wistar rats (2.0 g/Kg/14 days) and on offspring development. The results showed that the administration of the oil did not induce maternal toxicity, had no abortive activity, and also did not change the normal development of the offspring or the behavioral parameters. Araujo-Lima et al. (2018) evaluated three Andiroba oil samples obtained by different procedures. Oil 1 was obtained by pressure at room temperature, oil 2 by autoclaving, drying and pressing, and oil 3 by soxhlet extraction with petroleum ether. The results showed that these samples are a composition of different oil fatty acids, such as palmitic, oleic, linoleic and stearic acids, without much variation between the studied oils. They also observed that oil 1, obtained without the use of high temperatures, was the safest to use, not showing mutagenicity or micronucleus induction. In a study carried out by Melo et al. (2018), the in vivo genotoxic and antigenotoxic potential of Andiroba oil and its nanoemulsion was evaluated at a concentration of 2,000 mg/Kg for 14 days of treatment. No hematotoxicity, cytotoxicity or genotoxicity was observed in this in vivo assay, suggesting that the oil can be used safely by the population regarding genetic damage. This study confirmed the results reported by Arencibia-Arrebola et al., 2013, Milhomem-Paixão et al., 2016. In this last work, the same exposures and doses were used, in addition to Andiroba oils with the same major compounds (oleic, palmitic and linoleic acids) for in vivo studies. No clinical or behavioral changes were observed during this treatment period. These findings suggest that Andiroba oil is not mutagenic, genotoxic or cytotoxic. Therefore, we consider that these compounds at the doses and forms of administration tested in this study could not induce genotoxicity. However, Milhomem-Paixão et al. (2017) demonstrated that Andiroba oil and the nanoemulsion can have an in vitro cytotoxic effect only at high concentrations, which was not observed in in vivo tests. In that work, the results of the cytotoxic difference between the tests may be related to the forms of absorption. There is no barrier to the absorption of compounds by cells in in vitro studies, while in vivo absorption involves the permeability of metabolites with the gastrointestinal tract, their transport to organs and hepatic metabolism. Despite many reported biological activities of Andiroba, there is still little knowledge reported about the toxicity and ecotoxicity of natural products derived from this species. The indiscriminate use of a wide variety of products derived from seed oil may pose a risk to the health of the Amazonian population, since some of the main constituents of the oil are suggested to promote apoptosis by various mechanisms, such as damage to cellular DNA, depending on the concentration tested (Milhomem-Paixão et al., 2016). In this context, new in-depth toxicological and genotoxic investigations are needed to assess the efficiency and safety of natural products commonly used to treat diseases among Amazonian and other populations. Assessing potential genotoxicity is particularly important, as such toxicity is considered vital in the development of diseases such as cancer (Ren et al., 2017).

Conclusion

Andiroba is one of the main non-timber forest products exported to the United States and Europe, and has several applications in the cosmetic, food, textile and repellent industries. The growing interest in the traditional uses of products derived from Brazilian biodiversity, especially from the Amazon, encourages the adoption of management practices and improvement of productive systems for collecting and cultivating species, which is of great value to integrate many families and communities of small local producers. However, there is little incentive on the part of institutions to convert research results into patents and to use all information on new technologies, which can lead to the loss of opportunities for commercial and economic exploitation of Brazilian biodiversity.

Many studies report in vitro and in vivo pharmacological activities of Andiroba oil and its isolated limonoids. Studies have even been carried out with the use of this oil in nanoemulsions for various applications, from protection against side effects caused by antineoplastic agents to larvicidal action. In addition, the possibility of extracting active compounds from the residual biomass of Andiroba was evaluated. On the other hand, the by-products of the oil production process can be sustainable, giving rise to soil conditioners and biodegradable polymers, in addition to fertilizers, reducing environmental impacts. However, there is a potential that has not yet been fully explored for oil and its derivatives and, in this way, it is possible to add value to its production chain. Far beyond producing repellent candles, the oil can be used to generate biodiesel in an extraction method on an industrial scale, without adding organic solvents in a continuous and clean way.

Although genotoxicity and cytotoxicity have not been reported for Andiroba oil, it is known that some of the main compounds found (fatty acids and limonoids) are suggested to lead to cellular apoptosis via mitochondrial pathway in high concentrations. Thus, being exposed to these natural agents can also result in toxic effects, considering the indiscriminate and prolonged use of Andiroba oil derivatives in Amazon Region. In this sense, therefore, it is desirable to find new effective test methods to identify and measure toxic effects, such as genotoxicity, of traditionally used natural products. Unfortunately, to date, no single test is capable of detecting all relevant genotoxic features and therefore a combination of different testing techniques is needed to achieve this goal. It is known that many limitations are restricting the role of new discoveries and approaches in this field. New toxicological approaches are needed to circumvent current limitations, such as cross-species and quantitative comparisons in model animals. Thus, the improvements in this domain could also guarantee the safe use of its derivative products by detecting some possible adverse effects in patients who traditionally use natural products.

CRediT authorship contribution statement

Vagner Pereira da Silva: Conceptualization, Methodology. Lavínia de Carvalho Brito: Writing – review & editing. André Mesquita Marques: Writing – review & editing. Flávia da Cunha Camillo: Writing – review & editing. Maria Raquel Figueiredo: Supervision.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

No data was used for the research described in the article.