Use of mineral salts to remove recalcitrance to somatic embryogenesis of improved genotypes of cacao (Theobrama cacao L.)

Some improved genotypes of cocoa tree (Theobroma cacao L.), have shown in standard study conditions of absence or a very low response to the induction of somatic embryos. This is the case of the recalcitrant genotypes C8 and C15, partially recalcitrant genotypes C14 and C16. This study aims to improve the production of somatic embryos of recalcitrant genotypes. Staminodes and petals excised from the immature buds of the five genotypes C1, C8, C14, C15 and C16 were used as plant material. These floral explants were cultured on different media containing two type of mineral salts such as potassium sulphate (K2SO4) at concentrations 18, 27, 36 and 45 mM and magnesium sulphate (MgSO4) whose concentrations used are 5, 10, 15 and 20 mM. Calli induction was obtained in the five (05) genotypes at percentages ranging from 50 to 100% with staminodes and from 61 to 100% with petals on all media. The transfer of the callogenic explants on the developmental medium allowed the induction of embryogenic calli and somatic embryos after 84 days only with petal explants. The most important PEC and NSE were obtained with two mineral salts concentrations, 27 mM K2SO4 and 15 mM MgSO4 for all genotypes and varied to 20 to 40%. Potassium sulphate (K2SO4) at 27 and magnesium sulphate (MgSO4) at 15 mM are the concentrations of the most suitable mineral salts to overcome recalcitrance of cocoa genotypes.


INTRODUCTION
The cocoa tree (Theobroma cacao L.), a species of the humid tropics, is mainly grown for its beans, which produce the cocoa powder used to make chocolate (Kone et al., 2019).
Global demand for cocoa continues to increase due to economic growth and public awareness of the benefits for The work done on the cocoa bean has shown that it has virtues on the stimulation of the immune system, the improvement of digestion and cellular detoxification. Cocoa products also have antimicrobial and antioxidant properties beneficial to human health (Sarmadi et al., 2012;Yapo et al., 2013).
Cocoa production is important for many cocoaproducing countries in order to maintain export stability and ensure the continuity of supply of industrial raw materials for the chocolate industries. Cocoa plays an important economic role as a source of foreign exchange for many tropical countries, including Côte d'Ivoire. It is grown by about 6 million farmers globally, and livelihoods of more than 40 million people depend on cocoa (ICCO, 2012;Conseil Café-Cacao, 2014;Beg et al., 2017).
However, demand remains unfulfilled despite the fact that its world production has been increasing considerably for several years (ICCO, 2016). This is due to the fact that in the agricultural sector, cocoa farming is affected by several pathogens, to which is added the aging of orchards, high cost of inputs, use of poor quality plant material, inappropriate agronomic techniques. To overcome these constraints and increase production, elite genotypes, high producers have been created for extension to allow the creation and renewal of plantations (Sonwa, 2002). However, the mode of reproduction of the plant, sexed and allogane, does not make it possible to obtain a homogeneity of the improved material. Since cocoa is naturally pollinated, seed cocoa planting material generally has a very heterogeneous genetic background. As a result, the agronomic performance of seed-derived cocoa plantations is highly variable (Li et al., 1998). To overcome these difficulties, traditional vegetative propagation techniques of plant material, such as cuttings and grafting, have been advised.
These techniques are not very competent (Figueira and Janick, 1993) because of the poor root and air systems of plants associated with slow processes. One of the appropriate solutions to all these problems is somatic embryogenesis. This technique is used to carry out the propagation of desirable cocoa genotypes. However, this method presents some difficulties in the cocoa tree because of the recalcitrance of certain genotypes. This recalcitrance is expressed by the variation of the somatic embryo rate from one genotype to another, often with very low or zero somatic embryo levels. The lifting of this recalcitrance requires the development of a new protocol. Moreover, several studies on the somatic embryogenesis of cacao have revealed a variation of callogenic and embryogenic responses according to genotypes. The work on somatic embryogenesis developed by Li et al. (1998), without being exhaustive, have shown a variation of the responses of 19 cocoa genotypes to somatic embryogenesis. To be used as explants, the response of staminodes to callogenesis ranged from 1 to 100% and the number of somatic embryos per explant ranged from 1 to 46 in this work.
Another somatic embryogenesis protocol was developed by López-Báez et al. (2001) using an induction medium comprising the macro and microelements of Murashige and Skoog (1962) (MS) supplemented with different types and concentrations of phytohormones (2,4-dichlorophenoxyacetic acid or 2,4-acid, 5trichlorophenoxyacetic, and kinetin), out of 12 different genotypes. Once again, a variation of the responses according to the genotypes was found.
The elite genotypes such as C1, C8, C14, C15 and C16 are characterized by good genotypic performance (high productivity) and resistance to diseases. However, low levels of embryogenic calli genotypes C8 (0.03%), C14 (14.32%) (Kouassi et al., 2018) and C16 (8.85%) (Kouassi et al., 2017b) were obtained with somatic embryogenesis. Therefore, it is highly important to develop a protocol allowing maximum high embryo production even in recalcitrant genotypes. The objective of the present study is to improve the production of somatic embryos in the recalcitrant genotypes used for cocoa farming. To achieve this goal, two types of explants (petals and staminodes) will be tested on different media with different concentration of mineral salts.

Plant
The plant material consisted of staminodes and petals excised from immature flower buds of the cocoa genotypes coded C1, C8, C14, C15 and C16. The flower buds were taken from the field of cocoa experimentation of the International Center for Agroforestry Research (ICAFR) of Abidjan (Côte d'Ivoire). Five cocoa genotypes were chosen according to their response to somatic embryogenesis. The genotype coded C8 originated from Trinidad and Tobago is recalcitrant to somatic embryogenesis. Genotypes coded C15, recalcitrant, C14 and C16, partially recalcitrant and C1 which is embryogenic, are from Côte d'Ivoire.

Collection and disinfection of flower buds
In the morning before 9 am, 4 to 5 mm long flower buds were collected and placed in jars and stored in a cooler containing ice and sent to the laboratory. Then, buds were disinfected under a laminar flow hood in sterile conditions, first by soaking them in a 1% (w/v) calcium hypochlorite solution, followed by three rinses in sterile distilled water. After that, they were re-dipped in 70% alcohol solution for 30 s and rinsed thoroughly three times with sterile distilled water. Finally, they were immersed a second time in the same solution of calcium hypochlorite 1% (m/v) with three drops of Tween 20 for 10 min and then rinsed thoroughly three times with sterile distilled water.

Isolation and culture of explants
Petals and staminodes were isolated from disinfected flower buds after dissection with a scalpel blade. Petals and staminodes were placed on the calli induction medium with 15 samples per explant (15 petals and 15 staminodes) and per Petri dish under a laminar flow hood for sterile conditions.

Composition of culture media
The media used callus induction medium (I) and their development into somatic embryos (ED medium) are composed of macro and micro elements of DKW (Driver and Kuniyuki, 1984). For testing their effect on calli induction and their ability to induce embryogenic calli, two mineral salts: potassium sulphate (K2SO4) and magnesium sulphate (MgSO4) were added to DKW medium at different concentrations 18, 27, 36 and 45 for potassium sulphate (K2SO4) and 5, 10, 15 and 20 mM for magnesium sulphate (MgSO4). Thus a total of eight induction media varying by the nature and concentration of mineral salts were prepared. Cytokinin used was kinetin at 1.162 μM. The control medium contained 9 mM potassium sulphate (K2SO4) and 3 mM magnesium sulphate (MgSO4). Four (4) weeks after culturing in calli induction medium, explants were transferred to embryo development medium (ED medium) described by Li et al. (1998). The transfer of the explants on this medium was done every 28 days.

Culture conditions
The pH of the media was adjusted to 5.8 for callus induction medium (I) and 5.7 for embryo development medium (EDM) using 1 N NaOH or HCl solutions. Media were solidified with 2 g/L phytagel before being sterilized by autoclaving for 20 min at 121°C and 1 bar. After sterilization, these culture media were dispensed due to 15 ml in sterile Petri dish of 90 mm diameter under a laminar flow hood. Cultures were incubated on continuous darkness conditions in the culture chamber at 24 ± 1°C with a relative humidity of 70%. Petri dishes were arranged on rows according to a completely randomized experimental.

Variables used to assess the success of somatic embryogenesis
Twenty eight (28) days after induction, percentage of callogenic explants (PCE) was evaluated. Eighty four (84) days after induction, percentage of embryogenic calli (PEC) and average number of somatic embryos (NSE) were evaluated on ED medium. These three parameters were calculated according to the following formulas: (1) induction evaluation was made on medium induction of calli (IC), after 28 days of culture; (2) somatic embryos were assessed on ED medium by the percentage of embryogenic calli 84 days after explants induction.
The percentage of callogenic explants (PCE), the percentage of embryogenic calli (PEC) and the mean number of somatic embryos (NSE) per explant were given respectively by the following formulas: Number of explants that induced calli PCE = × 100 Total number of explants cultured Number of calli that induced embryos PEC = × 100 Number of explants that induced callis Koné et al. 35 Number of induced embryos NSE = Number of calli that induced embryos

Statistical analysis of the data
Results were subjected to analysis of variance (ANOVA) with Statistica 7.1 software. For unequal numbers, analysis of variance across the generalized linear model (GLM) was adopted. When a significant difference was observed between averages, the Newman-Keuls multi-range test at a 5% threshold was used to separate the averages. Table 1 shows the percentage of calli induction from staminode and petal explants after 28 days of culture on the various induction media. The analysis of the results of this table after the statistical analyses shows that with the C1 and C14 genotypes, there is no significant difference between the calli induction percentages from the staminode and petal explants on the medium supplemented with different types and concentrations of mineral salts. These calli induction percentages were between 90 and 100% for C1 and between 92 and 100% for C14. Moreover, these percentages are identical statistically to that of the control medium. The contribution of mineral salts did not improve or reduce the responses to the callogenesis of genotypes C1 and C14.

Evaluation of calli induction in the various media after 28 days
With the C8, C15 and C16 genotypes, mineral salts improved the response to callogenesis. A beneficial effect of using potassium sulfate mineral salts and magnesium sulfate was clearly observed on the induction of calli for genotype C15.
However, the addition of MgSO 4 mineral salt at a concentration of 5 mM resulted in a reduction in calli induction percentages of genotypes C8 (staminodes 75.08% ± 0.17 and petals 84%) and C16 (the staminodes 74.33% ± 1.01). For all genotypes, calli percentages ranged from 74 to 100% with staminodes and from 92 to 100% with petals. The petals yielded calli at slightly higher percentages than staminodes in all genotypes (Figure 1).

Evaluation of calli and somatic embryos in the various media after 84 days
The calli obtained, transferred to the embryonic development medium (EDM medium), gave embryos 84 days after the initiation of the cultures. The percentage of embryogenic calli (PEC) and the average number of embryos (NSE) produced per explant are shown in In the same column, the averages followed by the same letter are statistically equal (test of Newman-Keuls to the threshold of 5%).  In the same column, the averages followed by the same letter are statistically equal (test of Newman-Keuls to the threshold of 5%).

Effect of mineral salts on the development of somatic embryos
After transfer of the callies to the embryo development medium (EDM), some continued to proliferate while others died. From the 84th day after the initial culture, embryogenic calli and embryos were observed for all genotypes. Figures 2 and 3 illustrate an example of embryogenic calli and embryos developed from petal explants of genotypes C1, C14 and C15.

DISCUSSION
The response of explants of T. cacao to somatic embryogenesis is dependent on genotype.
In order to develop a protocol applicable to certain genotypes deemed recalcitrant to somatic embryogenesis, two types of explants, staminodes and petals of five (05) genotypes, from the most embryogenic (C1) to at least embryogenic, C14, C16, C15 and C8 were used as plant material. A total of eight (08) induction media with four (04) concentrations of two (02) inorganic salts that are potassium sulfate (K 2 SO 4 ) at concentrations 18, 27, 36 and 45 mM and magnesium sulfate (MgSO 4 ) at concentrations 5, 10, 15 and 20 mM were carried out for the avoidance of recalcitrance in the studied cocoa genotypes.
The use of mineral salts in the induction medium did not have a negative impact on the callogenesis responses of staminode and petal explants in the five (05) genotypes C1, C14, C15, C8 and C16 in the most cases. The various mineral salts used, improved the responses to callogenesis, particularly in the most embryogenic C1, the partially recalcitrant C14 and the recalcitrant C15 genotypes. A reduction of the rate of calli induction was however observed when the concentration of 5 mM MgSO 4 was added to the culture medium with the two types of explant (staminodes and petals) for the most recalcitrant genotype C8 and the partially recalcitrant genotype C16. These results show that the responses to callogenesis varied according to the genotype of cacao, the type and concentration of the mineral salt used.
The petal and staminode explants of the genotypes tested allowed, however, the induction of calli with levels higher than 74% on all media. These results are similar to those obtained by Minyaka et al. (2008) who showed that petal and staminode explants respond to calli production with the mineral salts K 2 SO 4 and MgSO 4 .
In addition, the petals had higher callogenesis percentages than staminodes for the five (05) genotypes evaluated. This response shows that petals respond better to callogenesis than staminodes with mineral salts. Similar results have been obtained by Da Silva et al. (2008) who found that petals are better adapted to calli  formation than staminodes in T. cacao. Staminodes and petals could therefore indistinctly of genotype provide high percentages of callogenic explants with a maximum for the petals. These results, however, contradict those of Bahoya (2012). Indeed, these authors have shown that staminodes of genotypes named genotype 1, genotype 2 and genotype 3 used during their work were more conducive to callogenesis than petals. This shows that the response to callogene in addition to being dependent genotype is also a function of the explant. These authors have worked on genotypes that are different from those used in the present study. In the study, the percentage of embryogenic explants and the average number of somatic embryos were obtained with the calli derived from the petal explants in the five (05) genotypes C1, C8, C14, C15 and C16 of cacao and not with staminode explants. This result shows that the protocol developed in this study is more adapted to petal explants and the response to somatic embryogenesis also depends on the genotype. These results are consistent with those of Issali et al (2008), Kouassi et al. (2017a) and Eliane et al. (2019) which revealed that petals are better adapted to somatic embryo production than staminodes.
With mineral salts, inhibition of embryo production was observed with the concentration of 45 mM K 2 SO 4 for the genotypes C8, C14, C15 and C16 and with 5 mM MgSO 4 for the genotypes C8, C15 and C16. This shows that too low concentrations are sometimes insufficient to lift the recalcitrant however when they are too high they cause toxicity. These concentrations of mineral salts are too high for K 2 SO 4 and too low for MgSO 4 to allow the lifting of recalcitrance. This shows that the lowest concentrations are sometimes insufficient to overcome the recalcitrant ; however, when they are too high they cause toxicity regardless of the compound. Inhibition of somatic embryo production of T. cacao would be due, in addition to the insufficiency or toxicity of certain compounds used, to a strong secretion of ethylene and polyphenols by the explants of certain genotypes according to Fang et al. (2014) and Minyaka et al. (2017).