Evaluating auto-detoxification of Jatropha curcas Linnaeus , 1753 kernel cake with brine shrimp Artemia salina ( Linnaeus , 1758 ) lethality test

Jatropha curcas Linnaeus, 1753 is a plant with several uses for communities in Cameroon and other African countries. However, present methods of detoxifying J. curcas kernel cake (JKC) to add value to its further utilization are too sophisticated for cottage operations in rural areas. This study sought to evaluate two auto-detoxification methods using the shellfish brine shrimp, Artemia salina (Linnaeus, 1758). Two identical J. curcas auto-detoxification apparatuses (JADA) were developed: one simulating diffuse daylight (DJADA) and the other solar irradiation (SJADA) conditions. JKC was pretreated by either soaking or boiling in water or lye. The pretreated samples were then placed in the JADA and either re-moistened to 66% dry matter (DM) or left un-moistened. Among the 14 pretreated samples, water soaked solar without re-moistening (WS1), lye soaked solar without re-moistening (LS1), and water soaked solar with re-moistening (WS2) had LC50 values above 1000 ppm after four weeks. On the other hand, un-moistened diffuse daylight spread (UDC), boiled in water and exposed to diffused daylight without remoistening (WD3), and un-moistened solar spread (USS) were the least detoxified within the same period of four weeks with LC50 values less than 60 ppm. Irrespective of the pretreatment, extracts from the SJADA had higher LC50 values when compared to that of DJADA. This finding suggests that sunlight is an important factor in auto-detoxification. The top three most efficient detoxified treatments: WS1, LS1 and WS2 are recommended for further development and testing with livestock and fish models. This study confirms that JKC is toxic, but under natural conditions is exposed to auto-detoxifying factors which are both endogenous and environmental. When these conditions are optimally manipulated, the detoxification rate is enhanced. Farmers in rural communities can use this strategy to detoxify JKC, transform it into livestock and fish dietary ingredient and consequently enhance its contribution to climate change mitigation.

the critical 2°C rise in global temperature.Easily applied methods such as boiling need further investigation, because phobol esthers (PE), the main toxic component in JKC, are heat stable.They remain in the JKC even when autoclaved at 121°C for 30 min (Aregheore et al., 2003).However, the naturally occurring phobol esters are unstable and are susceptible to oxidation, hydrolysis, transesterificaion and epimerization during isolation procedures (Runumi et al., 2014).By taking into consideration this background, an effort needs to be made to find an easy and cheaper method to detoxify J. curcas for rural communities in countries like Cameroon.The present study was therefore aimed to develop two J. curcas auto-detoxification apparatuses (JADA) and test their efficacy using brine shrimp lethality test.

Design and construction of J. curcas auto-detoxification apparatuses (JADA)
A natural convection greenhouse solar drier with chimney (Ekechukwu, 1999) was modified to a J. curcas auto detoxification apparatus (JADA) by compartmentalizing the access door air inlet in such a way that various levels of air access could be achieved.This modification allowed the adjustment of the apparatus from full detoxification (when air access is completely closed) to full drying mode (the access is fully open).Between these two extreme modes, partial drying and detoxification was also possible.The JADA was constructed to have a triangular drying chamber with a height of 2.7 m, fitted on a rectangular base measuring 2.3 m wide and 5 m long.This chamber had an attached 2.3 m high vertical cylindrical chimney at one end, while at the other end a door existed for air inlet and access to the chamber.Thus, the total height of the chimney from the base of the apparatus was 5 m and had a north-south orientation.In Buea, located at 4.1667° N, and 9.2333° E, the chimney was located in the south and the door in the north.The floor was insulated with 30 cm thick compacted wood shaving covered with black plastic to keep it in place and absorb light.On the top of the floor were two rock piles gathered to retain any trapped heat during the day.The rock piles were covered externally by black corrugated sheets which trapped heat from the sun.Two identical J. curcas auto-detoxification apparatuses were constructed at the Faculty of Agriculture and Veterinary Medicine Farm in the University of Buea.In the first, the north was covered with transparent plastic, and the rest of the apparatus covered with black plastic to simulate diffuse daylight conditions and was named diffuse daylight J. curcas auto-detoxification apparatus (DJADA).In the second, the northern end with the door was covered with black plastic sheets, while the rest of the apparatus was covered with transparent plastics to maximize the effects of solar irradiation and was called solar J. curcas auto-detoxification apparatus (SJADA).
Author(s) agree that this article remains permanently open access under the terms of the Creative Commons Attribution License 4.0 International License (South West, North West, Littoral, Adamawa, North and Extreme North).The seeds were transported to the University of Buea where they were manually de-shelled to produce kernels.The kernels were de-oiled using a hydraulic press and J. curcas kernel cake (JKC) was produced.The JKC was ground into powder using a plate mill, and the powder was homogenized by hand mixing.The JKC samples were either physically or chemically pre-treated before being placed in the two types of JADA for detoxification.
Physical pretreatment was performed by soaking or boiling the JKC samples in water.In the physical pretreatment by soaking, homogenized JKC was soaked overnight in water at the ratio of 1:2 w/v by adding fixed weight of JKC to fixed volume of water and stirring the mixture to ensure complete dispersion in water.This procedure resulted in a pretreatment additional moisture content of 66.67%.This material was thereafter divided into 4 treatments (WS1, WS2, WD1 and WD2) with 4 replicates per treatment.In the boiling method, homogenised JKC was poured into boiling water in an aluminium pot at the ratio of 1:2 w/v and stirred continuously while still on fire for 30 min to produce dough.The resultant dough was sub-divided into two treatments (WS3 and WD3) each with 4 replicates.
Chemical pretreatment was achieved by soaking or boiling the JKC samples in lye.The lye used was produced locally by modifying the procedure described by Kent et al. (2014).About 50 kg of wood ash from a local palm oil processing plant was leached for three days in a 120 L plastic tank fitted with a tap.In the chemical pretreatment by soaking, homogenized JKC was over soaked overnight in lye at the ratio of 1:2 w/v by adding fixed weight of JKC to fixed volume of lye and stirring the mixture to ensure complete dispersion in lye.This procedure offered a pretreatment additional moisture content of 66.67%.This material was thereafter divided into 4 treatments (LS1, LS2, LD1 and LD2) with 4 replicates per treatment.In the boiling method, homogenised JKC was poured into boiling lye at the ratio of 1:2 w/v and stirred continuously while still on fire for 30 min to produce dough.The resultant dough was sub-divided into two treatments (LS3 and LD3) each having 4 replicates.

Auto-detoxification of the JKC
The 12 pretreated JKC samples (WS1, WS2, WS3, WD1, WD2, WD3, LS1, LS2, LS3, LD1, LD2 and LD3) plus two untreated or unmoistened samples (USS and UDC) were subjected to autodetoxification using the JADA.The two JADA modifications were operated at full detoxification mode.The samples were placed in steel plates that were arranged on two tables within the JADA; each treatment had four replicates.The WS1, WS2, WS3, LS1, LS2, LS3 and USS were placed in the SJADA while WD1, WD2, WD3, LD1 LD2, LD3 and UDC were placed in DJADA (Table 1).During the detoxification process, WS2 and WD2 and LS2 and LD2 were remoistened daily with water and lye, respectively, to obtain a moisture level of 66% every morning.However, there was no remoistening on the day proceeding the day of sample collection for evaluation (Table 1).

Preparation of crude methanol auto-detoxified J. curcas kernel cake extracts (CMJKCE)
Forty grams of each treatment sample (10 g per replicate) were collected weekly to assess the level of auto-detoxification.The four replicates per treatment were pooled and further dried within the JADA for another week.Each sample was homogenised, powdered and 20 g of it were mixed with 200 ml of methanol for 72 h with regular agitation.Filtration using Whatman No. 541 filter papers followed and the CMJKCE were obtained by complete evaporation of the solvent using a rotary evaporator.

Brine shrimp bioassay
The brine shrimp lethality test (BSLT) as described by Meyer et al. (1982) was used to test toxicity of CMJKCE after some modifications.The toxicity is reported as LC50 values in mg/ml (ppm).Brine shrimps [Artemia salina (Linnaeus, 1758)] were hatched from eggs in rectangular dishes divided by a perforated barrier, under constant aeration using natural seawater.The perforated barrier divided the dish into two chambers.The bigger chamber receiving the eggs was in the dark while the smaller chamber receiving the hatched nauplii in anticipation was under constant light.Forty eight hours were allowed for the eggs to hatch and the photropic nauplii to mature.Ten nauplii were collected using a pipette and placed in marked vials, each containing 4 ml of natural seawater.Two hundred miligrams of each examined CMJKCE were dissolved in 2 ml of pure dimethyl sulfoxide (DMSO) to get a solution of 100,000 ppm (100,000 mg/l).A unit volume of this solution was diluted in one mililitre of natural seawater to give a concentration of 50,000 ppm.Ten-fold serial dilutions were further performed in sea water to produce concentrations of 5000, 500, 50 and 5 ppm.One milliliter of each of these dilutions was delivered into pre-marked vials containing 4 ml of natural sea water and 10 nauplii.As a result, the final concentrations of CMJKCE were 10,000, 1000, 100, 10 and 1 ppm.Three replicates of each concentration were prepared for CMJKCE samples.The negative controls were dilutions of DMSO in seawater without CMJKCE.Undetoxified whole J. curcas kernel cake served as the positive control.The vials were kept incubated under light for 24 h after which the surviving nauplii were counted.The nauplii were counted against a lighted background using a 3× magnifying hand lens.They were considered dead if they lay immobile at the bottom of the vial.The mortality percentage was then calculated.The mortality data was corrected using Abbott's (1925) formula (Apu et al., 2013), since a mortality percentage more than 10% was recorded in the controls.Subsequently there were corrections for 0 and 100% as proposed by Ghosh (1984).The formulae used in calculating and correcting brine shrimp mortality are presented in Table 2: The surviving nauplii were killed by addition of 5 ml of 5% v/v phenol to each vial.The vials were crosschecked to ensure that all the nauplii were dead before discarding.

Determination of 50% lethal concentration (LC50)
The lethal concentration of CMJSCE resulting in 50% mortality of brine shrimp (LC50) was determined from the 24 h counts by a plot of percentage of the shrimps killed against the logarithm of the CMJKCE concentration.The best-fit line was obtained from the curve data by means of regression analysis (MS Excel version 7) and the LC50 was derived from the slope of the best-fit line obtained.

Statistical analysis
Levene's test for equality of variances and t-test for equality of means were performed on LC50 using the IBM SPSS Statistics version 22 (IBM Corp. Released 2013).The mean (±SEM) were quarreled.

Selection of the most promising auto-detoxified JKC (ADJKC) treatments for further development
Selection of the most promising ADJKC treatments was achieved

RESULTS
The level of auto detoxification of the different experimental groups has been ranked using the mean of the LC 50 values for all the treatments from weeks 1 to 4 (Table 3).From the results presented in Table 3, WS1, LS1 and WS2 treatments all had LC 50 values above 1000 ppm after four weeks of exposure in the JADA, indicating that they were the most detoxified.On the other hand UDC, WD3 and USS were the least detoxified within the same period of 4 weeks with LC 50 of less than 60 ppm.The extent of lethality was found to be proportional to the concentration of the CMJKCE.High mortality rates were recorded at 1000 and 10,000 ppm, while lower values were recorded at 1 and 10 ppm.Mortality rates of brine shrimp exposed to whole J. curcas kernel cake just after oil extraction are presented in Figure 1.Thereafter, corrected brine shrimp mortality rates in the 4 th week for UDC, WS1, LS1 and WS2 are presented in Figures 2 to 5, respectively.The means (and standard error of the means) of the LC 50 of the extracts of solar and diffused daylight detoxified JKC extracts over the four week detoxification period are presented in Table 4. Irrespective of the pretreatment, SJADA treated extracts had higher LC 50 values compared to the DJADA ones, with the WS1, LS1 and WS3 having significantly higher values (P<0.05)compared to WD1, LD1 and WD3 groups, respectively (Table 4).

DISCUSSION
The JADAs were easy to construct onsite because of their simple design and low cost which is characteristic of natural convection greenhouse solar driers with chimney.These characteristics attributed to natural convection greenhouse driers (Weiss and Buchinger, http://www.aee-intec.at/0uploads/dateien553.pdf) make development of the JADA a suitable tool for J. curcas detoxification in remote rural communities.Other solar dryer designs may not be appropriate for re-design into a JADA.This is because in the JADA, there is need for the crop to have optimised access to environmental conditions such as light, temperature, moisture, wind, oxygen and microbes that contribute to its autodetoxification.Besides, J. curcas is a tropical and subtropical crop, grown in areas with ubiquitous solar radiation.
The proportional response of brine shrimp to the increased concentration of the CMJKCE confirms that auto-detoxification was effective in reducing JKC deleterious effects.The present study also supports the earlier findings of Chikpah and Demuyakor (2012), who found that approximately 60% reduction in crude phorbol ester levels can be achieved within 21 days of  spontaneous fermentation of J. curcas kernel meal.According to Meyer et al. (1982), several extracts derived from natural products which had LC 50 ≤ 1000 μg/ml using BSLT were known to contain physiological active principles while those with LC 50 values > 1000 ppm are considered inactive.Thus, exposure of the JKC in the JADA for four weeks rendered them inactive and therefore the detoxified cake extract was less toxic to , further increase in concentration was not effective in increasing mortality rates of A. salina.100% mortality was only achieved at a high concentration of 6000 mg L -1 . This observation prompted the testing of CMJKCE at a concentration of 10,000 ppm in the present study.
The J. curcas plant contains many biologically active phytochemicals including proteins, peptides and diterpenes which exhibit a wide range of biological activities (Devappa et al., 2010(Devappa et al., , 2011a)).However, the seeds contain anti-nutritional and toxic factors such as phytate, trypsin inhibitor, lectin, curcin and phorbol esters (PEs) (Makkar et al., 1997).In the majority of the cases, al. (2011b) observed that after a mortality "plateau" of toxicity of seeds and seed derivatives is attributed to the presence of PEs (Devappa et al., 2011b).In a previous study, Kinghorn et al. (1977) had demonstrated that there are differences in the toxicity of specific purified phorbol esters.While purified PEs such as phorbol 12tetradconoate 13-acetate, phorbol 12, 13-didecanoate, and phorbol 12, 13-dibenzoate induced toxicity with an effective mortality dose (ED 50 ) of 3.8, 6.8 and 11.8 mg L -1 respectively, the phorbol and 4a-phorbol 12, 13didecanoate were relatively nontoxic (ED 50 > 1000 mg L -1 ).Referring to the above study, Devappa et al. (2011b) concluded that bioactivity of PEs depends on their chemical structure/configuration and purity.
The BSLT only indicates the level of toxicity of an extract or a substance and does not identify the toxic component.However, it can be used anywhere in the world as a general toxicity test because the eggs are readily available.It is a rapid, inexpensive and simple bioassay for testing plant extract lethality which in most cases correlates reasonably well with cytotoxic and antitumour properties (McLaughlin et al., 1991).This assay is considered as a useful tool for preliminary assessment of toxicity and it has been used for the detection of fungal toxins, plant extract toxicity, heavy metals, pesticides and the cytotoxicity testing of dental materials (Harwing and Scott, 1971;McLaughlin et al., 1991;Martinez et al., 1998;Barahona and Sanchez-Fortun, 1999;Pelka et al., 2000).Comparing the BSLT with the respective snail bioassay, Devappa et al. (2011b) concluded that the BSLT is less sensitive towards J. curcas PEs with respect to toxicity screening.Nonetheless, the use of A. salina and Daphnia magna Straus, 1820 is preferred for assaying a large number of PE samples because the test can be performed in 96-well plates (Ruebhart et al., 2008).At the same time, BSLT is the only feasible choice in countries like Cameroon, where the common bladder snail [Physa fontinalis (Linnaeus, 1758)], un-adapted to tropical environments, is not readily available and the response of local snails to PEs has not yet been studied.In addition, the preliminary testing of ADJKC for toxicity should not only be focused on PEs but should detect additional toxins developed during the process.Makkar et al. (2012) recognized that no information is available on the nature of the phorbol ester degraded products and  their possible toxicity.The naturally occurring PEs are unstable and are susceptible to oxidation, hydrolysis, trans-esterification and epimerization during isolation procedures (Runumi et al., 2014).Auto detoxification could exploit the opportunities presented by these PEs properties to render JKC innocuous.Irrespective of the pretreatment, SJADA extracts had higher LC 50 values compared to the DJADA ones.This observation indicates that exposure of JKC to sun-light has a significant effect on the auto-detoxification process.Auto-detoxification is a self-detoxification process induced by endogenous and environmental factors including enzymes, microbes, sunlight, temperature, humidity and wind.It is the natural way to transform the toxic J. curcas seeds into an innocuous material.These processes take considerable time under natural conditions but their duration can be shortened by human manipulation.Materials are pre-treated through physical, chemical or both ways, in order to induce the necessary changes that contribute in their detoxification.One of the known changes is a process called autoxidation.Autoxidation is any oxidation that occurs in open air or in presence of oxygen (and sometimes UV radiation) and forms peroxides and hydroperoxides.In a wider sense, any oxidation process which produces peroxy radicals, irrespective of their origin, is also considered as an autoxidation process (Michael, 1981).The main toxin of J. curcas seeds is represented by the phorbol esters that are highly susceptible to autoxidation.Schmidt and Hecker (1975)  autoxidation under different conditions of storage.It was found that in addition to the traces of polar products that were not isolated, the aldehyde, 0-epoxide, hydroperoxide, hydroxylated A5-6 isomer, ketone and 6, 7-seco hemiacetal are formed from TPA.If TPA is stored as a powder at 25°C in diffuse daylight for 3 months, it is converted preferentially into the hydroperoxide, with smaller amounts of the other products.If spread on a large surface as a thin film (on glass plates, beads, or silica gel thin layers) and kept in diffuse daylight at 25°C, TPA is oxidized very rapidly.In this case, the 6, 7-seco hemiacetal is formed preferentially, in addition to smaller amounts of the other products.This rapid conversion is dependent on light and the reaction is slowed down in the dark.Moreover, among the various autoxidation products that could be obtained from TPA, a sample of 4a-phorbol-12, 13-diacetate yielded (~30%) the corresponding 20aldehyde as the only conversion product after storage for 1 year at room temperature in the dark.Storage in the dark is mandatory to avoid the known "lumi reaction" of 4a-phorbol, which is independent of the presence of oxygen (Schmidt and Hecker, 1975).It can be concluded from the above study that all phobol esters do not follow the same path of autoxidation in the presence of light.This may explain the differences in auto-detoxification observed between the SJADA and DJADA.The challenge for auto-detoxification of J. curcas seed materials is to maintain an adequate post-detoxification value for their utilization as feed ingredients for livestock and fish.
The top three most detoxified treatments: WS1, LS1 and WS2 are recommended for further development and testing as livestock and fish dietary ingredients by using the respective farm animal models.WS3 could also be considered for further development.It could be possible that pre-treatment heating influenced some factors that subsequently affected the level of auto-detoxification of WS3.Aregheore et al. (2003), found that autoclaving at 121°C for 30 min did not change the PE content in J. curcas kernel meal.However, the concentration of lectins was significantly (P<0.05)reduced by the heat.Makkar et al. (2012) recognize that the main category of toxic compounds, the phorbol esters, is to a large extent heat stable.Therefore, other strategies must be applied for their removal.Alatise et al. (2014), Fakunle et al. (2013) and Annongu et al. (2010) concluded that Clarias gariepinus (Burchell, 1822) could tolerate up to 30% of boiled J. curcas kernel meal in the diet in place of soybean meal.One of the challenges concerning the future development of WS3 treatment is that it uses heat as a pre-treatment prior to auto-detoxification.This may increase its carbon footprint particularly in countries like Cameroon where 98.7% of rural households rely on wood as their only fuel source (Ego, 2009).Moreover, Cameroon's wood biomass demand grows nearly at the same pace with total energy demand; approximately 2.4% per year (AEEP, 2013).

Conclusion
This study has confirmed that JKC is quite toxic.However, under natural conditions, it is exposed to autodetoxifying factors which are both endogenous and environmental.When these conditions are optimally manipulated as in the SJADA, the detoxification rate is enhanced.Solar radiation plays an important role in auto detoxification process of JKC.Farmers in rural communities can use this strategy to detoxify JKC, transform it into animal feed and consequently enhance its contribution to climate change mitigation.The role of microbial succession in auto-detoxification needs to be further investigated.Although the BSLT is rather inadequate in identifying the toxic ingredients in ADJKC, it is useful to determine the level of detoxification.The top three most detoxified treatments: WS1, LS1 and WS2 are recommended for further development and testing as livestock and fish dietary ingredient.

Table 1 .
Summary of JKC auto-detoxification treatment groups.

Table 2 .
Formulae used in calculating and correcting brine shrimp mortality.

Table 3 .
LC50 of crude methanol auto detoxified J. curcas kernel cake extracts (CMJKCE) ranked from highest to least detoxified (mean of 4 weeks).