Propagation of IMC67 Plants, Universal Cacao (Theobroma Cacao L.) Rootstock via Somatic Embryogenesis

ABSTRACT One major problem with cacao crops is the variability of plantings obtained from seeds. This has led to considering somatic embryogenesis (SE), a means of obtaining stable clones with minimal somaclonal variation; however, achieving this has not been possible with many genotypes. In this work, we used IMC67 genotype (rootstock in the traditional grafting process) to evaluated various SE protocols; four culture media for induction and expression primary embryogenesis (PSE): induction medium (INDI), expression medium (INDIEXP), primary callus induction medium (PCG), and secondary callus growth medium (SCG), as well as two explants (petal and staminode). Different culture times for PSE (15, 25, 35, 45, 55, 65 days) and SSE (10, 20, 30, 40, 50, 70, 80 days) were evaluated in terms of callus formation, PSE, and SSE. For acclimation, seven substrates were evaluated for plantlet survival. One hundred percent callogenesis was achieved at 50 days with the staminode in INDI, and the most PSE per explant (8) in INDIexp was achieved at 45 days. Secondary somatic embryogenesis (SSE) obtained the most globular and cotyledonary embryos (30) at 30 days in callus multiplication (CM2). The embryos achieved conversion and development into plantlets in maturation medium (MM6). Embryogenic efficiency from explants was 11.7 for PSE, and 15.2 for SSE. Acclimation obtained a 42.1% survival rate in the 70% sand- 30% soil substrate. This work reports a complete regeneration protocol via somatic embryogenesis up to plantlet development in a nursey for the IMC67 rootstock genotype; contributing to standardizing a high-quality plant production system.


Introduction
Theobroma cacao L. grows in humid tropical areas, is part of the Malvacea family, and is an important species (Sánchez, 2018). Cacao beans have nutritional, therapeutic, and cosmetic attributes and are the main raw material for the chocolate industry. Cacao is also a functional product because of its beneficial effects on cardiovascular diseases and high levels of antioxidants (Badrie et al., 2015). Colombia has the right agro-environmental conditions to be a net exporter of cacao. Its cacao beans are cataloged as fine and aromatic (Abbott et al., 2019), with awards at the "Paris Chocolate Show" in 2010Show" in , 2011Show" in , 2015Show" in , and 2019 for samples from the departments of Arauca, Nariño, Antioquia and Santander (Bucaramanga, 2019). Cacao production is increasing in Colombia, moving from 36.7 thousand MT in 2000 to 59.7 thousand MT in 2019. As of November 2019, Colombia recorded 8.33 thousand MT in exports, an 18% increase with respect to the 7.06 thousand MT exported in 2018. Imports (402 MT in November 2019) decreased by 32% with respect to November, 2018 (Fedecacao, 2019). Colombia needs strategies to maintain this production and export rate in order to continue and strengthen this trend, which would further benefit cacao farmers and production areas where cacao is the economies' axis. genotype and evaluate different substrates in the acclimation processes of plantlets developed in a nursery. This is the first report of a complete regeneration protocol via somatic embryogenesis up to plantlet development in a nursery for the IMC67 rootstock genotype.

Plant Material
The plant material consisted of staminodes and petals derived from closed flower buds of the IMC67 rootstock genotype. The immature closed flower buds were collected from field-grown plants of Yariguíes farms owned by Compañía Nacional de Chocolates in Barrancabermeja's municipality in the department of Santander -Colombia. The flowers were transported to Universidad de Antioquia's Laboratory of Plant Physiology in the city of Medellín, in the department of Antioquia -Colombia. The acclimation phase was performed in a nursery at the "La Nacional" farm owned by Compañia Nacional de Chocolates in the municipality of Támesis in the department of Antioquia -Colombia.
The Flower Buds collection and disinfection process was performed in accordance with the methodology described by Henao et al. (2018).

Explant Preparation and Culture
Petals and staminodes were extracted from flower buds using sterile scalpels. Twenty to 25 explants were placed on the culture media (induction medium) per Petri dish. The effect of the detection protocol for the percentage of petal and staminode contamination and necrosis was evaluated in the first experimentstaminode. Contamination by fungi and bacteria (%) was calculated 5 days after in vitro introduction and necrosis at 10 days. Each Petri dish was considered an experimental unit. This experiment had a total of 225 repetitions. The experiment was laid out in a completely randomized design (CRD) with one factor: explant type (staminode and petal).

In vitro Stage: Callogenesis, primary and secondary somatic embryogenesis
Callus induction starts off cacao's primary somatic embryogenesis process. Callus was induced from petals and staminodesin the second experiment. These explants were cultivated in two semi-solid culture media to evaluate induced embryogenic calli, the first named PCG by Maximova et al. (2002) and the second named INDI by Fontanel et al. (2002) (Appendix 1). Callus percentage (%) was calculated at different cultivation times (10, 20, 30, 40, 50, 60 days) and for two explant types (petals and staminodes) and two culture media (PCG and INDI). A total of 225 repetitions were performed for callogenesis. The experiment was laid out in a CRD with three factors: explant type, culture time and culture media.
In the third experiment, explants were subsequently transferred to evaluate the PSE expression stage in two culture media: SCG (Maximova et al., 2002) and INDIexp (Fontanel et al., 2002) (Appendix 1). Different cultivation times were evaluated for PSE (15, 25, 35, 45, 55, 65, 75 days), as well as two explant types (petals and staminodes). All primary embryo counts were in the globular and cotyledonary development stages. One hundred and ten repetitions were performed for PSE. The experiment was laid out in a CRD with three factors: culture media, culture time, and explant type.
In the fourth experiment, once the primary somatic embryos developed, they were fragmented and moved to the CM2 medium for SSE induction (Fontanel et al., 2002). Different cultivation times were evaluated for SSE (5, 10, 20, 30, 40, 50, 70, 80 days), and all embryo counts were in the globular and cotyledonary development stages. One hundred and ten repetitions were performed for SSE induction. The experiment was laid out in a CRD with one factor: culture time. Embryogenic efficiency (EE) related to average number of embryos per explant was calculated in accordance with the protocol of Garcia et al. (2016). Embryogenic efficiency for PSE and SSE was determined through the relationship between the total number of produced somatic embryos and the total number of initial explants for primary and secondary somatic embryogenesis independently.
Once the SSE developed, they were transferred to the MM6 maturation medium (Henao et al., 2018) (Appendix 1). For the maturation stage, two embryo development experiments were performed from the globular to plantlet stage with 30 repetitions, with an average of 67 explants per Petri dish.
In the fifth experiment, different cultivation times (8, 16, 24, 32 days) were evaluated for abnormal embryo development, and the sixth experiment evaluated the effect of embryo stage on abnormal development. The considered embryo stages were the immature stage (IM) with globular embryos, transition stage (TS) with heart-torpedo/early cotyledonary embryos and mature stage (MA) with mature cotyledonary embryos. The experiments were laid out in a CRD with one factor in each experiment: culture time and developmental stage, respectively.

Culture Conditions
All the cultures were randomly placed in a growth chamber in continuous darkness for callogenesis, PSE and SSE, at an average temperature of 26°C ± 2°C and 70% relative humidity. The plantlets were cultured in 500 ml vessels and placed in a growth chamber under light with a 16-hour photoperiod and a photosynthetic photon flux density (PPFD) of 50 μmol m − 2 per second, at an average temperature of 26°C ± 2°C and 70% relative humidity.

Ex vitro Stage: Acclimation
In the seventh experiment, preliminary tests for ex vitro adaptation were performed in a nursery at the "La Nacional" farm owned by Compañia Nacional de Chocolates. Approximately 130 plants regenerated by somatic embryogenesis were used and the effects of different substrates on plantlets' development and survival were evaluated: (1) 100% soil from the field, (2) 100% sand, (3) 100% coconut coir, (4) 50% soil-50% sand, (5) 70% soil-30% sand, (6) 30% soil-70% sand, (7) 30% soil-70% coconut coir. All substrates were previously sterilized in the laboratory and subsequently taken to the farm. Plantlets with a minimum of 3 cm-long stems and some level of rooting were kept in covered seedbeds for 2-3 days and subsequently transferred to plastic bags (30x14cm). The variable responses were stem length (cm), total number of leaves and survival (%) at 20 days after cultivation. Humidity was controlled every 4 hours by an automatic irrigation system by nebulization.

Statistical Analysis
Results were subjected to an analysis of variance (ANOVA), and a comparison test was performed based on residuals to test normality. If they were not normal, the Kruskal-Wallis and Mann Whitney test was used. For unequal numbers, an analysis of variance across the generalized linear model (GLM) was performed for some stages with a Poisson variable response distribution. Dispersion, box and whisker diagrams were used. A significance of 95% was determined for all comparisons using the R project v3.6.1 software.

Results and Discussion
Explants' responses to somatic embryogenesis in cacao are genotype-dependent. In order to develop a reliable and applicable protocol for the somatic embryogenesis of cacao genotypes, IMC67 was tested in differing culture conditions during this study, from in vitro establishment to acclimation. In the first experiment of the study, microbial contamination by fungi (Supplement, Table 1) and bacteria (Supplement, Table 2) was found not to be related to the type of explant. However, it should be noted that different climate conditions during flower collection lead to changes in the abundance of microorganisms on the crops, and fungi emergence is higher during rainy seasons than in dry seasons (data not shown). The percentage of tissue necrosis was found to be significantly higher for petals (Supplement , Table 3), possibly because petals are more external on the floral structure and have a larger surface area exposed to disinfectants, which can cause more oxidative damage. Flowers were collected during the afternoon and floral tissues were kept between 10-15°C during transportation to reduce the percentage of necrosis during experiments (data not shown).

In vitro Stage: Callogenesis, PSE and SSE
In the second experiment, callogenesis induction in the PCG medium with both explants (petals and staminodes) was a median 25%, while it surpassed 87% for both explants in the INDI medium, with 94.7% callus formation for the staminode explant and 88% for the petal explant (Supplement, Figure S1). In addition, it was observed that callus formation occurred at a median of 85% at 10 days and up to 100% at 50 days. Therefore, the best response to callus formation was with the INDI medium with a staminode explant 10 days after cultivation (Figure 1). The results are in accordance with that which Monsalve et al. (2005) reported for IMC67, where the staminode explant demonstrated a greater callogenic response. The response can be explained because staminodes are adjacent to the meristematic area on a flower's structure. In addition, in a culture medium, the entire staminode's surface is in contact with it, which is not the petal's case due to its curved structure (Gallego et al., 2016).
In accordance with Gallego et al. (2016), three types of calli were also recognized: a vitreous granular-white callus (TC), waxy granular-brown callus (WBC) and filamentous white callus (FC) (Figure 2). Embryos always originated from brown granular calli (WBC), which is considered an embryogenic callus for cacao. The presence of polyphenols in cacao is always associated with embryogenic development and, depending on their position inside or outside of the proembryogenic tissue, they can act as process inductors or repressors (Chanatásig and Aguilar, 2004;Kouassi et al., 2017). Furthermore, Kouassi et al. (2018) found that embryogenic differentiation can be promoted through phenols, which, as antioxidants, are better substrates for oxidative enzymes, allowing auxin release. In addition, the results are consistent with Daouda et al. (2019), in which the combination of 2,4-D and KIN is suitable for inducing callogenesis and, subsequently, primary embryo formation. PSE formation came about from the WBC embryogenic callus (Figure 2(a)), although two other clearly recognizable types of calli were present. The first consisted of round, translucent glassy-white cells (TC) (Figure 2(c)) and the second type consisted of waxy-yellow cells (WC) (Figure 2(b)), which did not complete the full embryogenic process. These results were comparable to those obtained by Gallego et al. (2016) for genotype ICS95 and Henao et al. (2018) for genotypes EET8, ICS1, ICS39, ICS60, ICS95.
In the third experiment, a relationship between the type of medium, type of explant and different culture times was found in PSE induction in a globular state. PSE showed a significant increase at 45 days of cultivation, with a median of 8 embryos. In terms of cotyledonary embryos, there was a maximum production of 38 embryos at 55 days (Figure 3). A significant difference with respect to the explant and culture media was observed for both types of embryos, obtaining more embryos with the staminode explant ( Figure 4). The same was true for the INDIexp medium ( Figure 5).
Various authors such as Minyaka et al. (2017), Ajijah and Hartati (2019) and Iracheta et al. (2019) affirm that the callogenic and embryogenic response in cacao is a dependent genotype, and culture conditions and type of explant influence it. Monsalve et al. (2005) performed a study with the IMC67 genotype. Despite achieving a high frequency of callus formation, they only obtained primary embryos up to the heart stage. Minyaka et al. (2008), using IMC67, also obtained a larger percentage of embryo formation with the staminode explant. However, PCG and SCG mediums with 6.0 mM MgSO4 concentrations were used for this work. The differences in embryogenic responses between these studies can be due to the different growth regulators, culture medium components and type and  physiological state of the used explant (Boutchouang et al., 2016). Various authors have confirmed that DKW basal salts are more appropriate for the somatic embryogenesis of cacao than MS basal salts (Ajijah et al., 2016;Li et al., 1998). The positive response to the combination of nutrients from DKW basal salts can be the result of a higher concentration of calcium, sulfur and magnesium. These macronutrients can be crucial during the first induction stages and expression of proembryogenic structures and embryos (Ajijah et al., 2016).
Somatic embryogenesis is begun in the INDI induction medium, which contains auxin and cytokinin growth regulators (with auxin (2,4 D) and cytokinin (KIN)). Cells respond to different stress factors, changing their development program to a specific physiological condition by reprogramming gene expression toward acquiring an embryonic competence (Fehér et al., 2016;Méndez et al., 2019). For cacao, it has been identified that stress factors, such as lesions, salt content and growth regulators like TDZ, KIN and NAA can trigger floral tissue somatic cell conversion into competent embryonic callus and embryo formation (Ajijah and Hartati, 2019). This embryogenic response in tissues occurs, not only because of the external addition of growth regulators, but also due to the result of interactions between exogenous regulators and endogenous contents from other regulators in the plant tissue. The presence of KIN in the culture medium can specifically favor embryogenic tissue induction and regeneration. This entire process is a coordinated action with auxins and amino acids, which offer a hormonal balance that can lead to regenerating various organs (Guillermo et al., 2003).
Releasing the hormonal stimulus of the explants from the INDI medium, which are moved to the INDexp medium, enables their embryogenic potential's expression and development (Guillou et al., 2018;Iracheta et al., 2019). The INDexp culture medium, which does not have growth regulators, responds well. This may be because of the presence of various amino acids, such as arginine, glycine, leucine, lysine and tryptophan. These amino acids are an effective and immediate source of nitrogen for the cells and are considered essential for embryogenesis (Méndez et al., 2019). The negative responses obtained from the SCG medium may have occurred by using high concentrations of auxin, which can suppress organized growth and nullify the embryogenic mass' polarity, inhibiting the embryogenic process begun by endogenous auxins. Although it has theoretically been said that somatic embryogenesis begins with a culture medium with high auxin levels, especially 2,4-D (Smertenko and Bozhkov, 2014), it is important to note that the correct balance between internal and external stimuli is required. The obtained results could suggest that the concentration of 2,4-D affects embryogenic development in this genotype. The SCG culture medium also contains TDZ, which normally increases purine supply, assisting cell development and metabolite accumulation (Victor et al., 1999). However, the presence of TDZ in the SCG induction medium may have limited embryogenic responses. According to Guillermo et al. (2003), this regulator is highly resistant to cytokinin oxidase's actions, which suggests it can remain active for prolonged periods of time, inhibiting the expected response.
In the fourth experiment, SSE (Figure 6(d-f)) was achieved in the CM2 culture medium drawing from PSE fragments (Figure 6(a-c)) originated from the INDexp medium. The most SSE was obtained with the staminode explant (Supplement, Figure S2), as well as a significant difference at 30 days, with a median of 30 globular and cotyledonary embryos (Figure 7). According to Daouda et al. (2019), using auxins, in this case 2,4,5-T, is common for inducing SSE in cacao. It has been used for numerous genotypes with positive results. Embryogenic efficiency by explant was 11.7 for PSE and 15.2 for SSE. A 29.8% increase was observed for each explant between PSE and SSE. This result was similar to the one reported by Garcia et al. (2016) for the TSH1188 genotype, with 12% EE.   In the fifth experiment, 52.1% of embryos reached the cotyledonary stage and developed into plantlets during SSE development and conversion in the MM6 medium. The remaining percentage includes necrotic embryos and malformations such as embryonic fusion, embryos with a single cotyledon, multicotyledonary embryos and embryos without apical and radicular development or atrophied development (Figure 6(g-i)). The lowest number of malformed embryos in the MM6 medium was obtained at 24 and 32 days (Figure 8), with a median of 2 embryos and no significant difference between them. The most malformations were 37 embryos at 8 days. In the sixth experiment, no significant difference was observed between the IM stage and TS stage, in which the most malformations were obtained. This coincides with the first 8 days of the experiment. Upon comparing this with the MA stage, it had significant differences with respect to the IM and TS stages. The MA stage had the least malformed embryos and, therefore, the embryos must be transferred from the CM2 medium to MM6 in the MA (Figure 9).
Malformations are typical in cacao during the embryogenic process and have been reported by various authors (Garate et al., 2017;Garcia et al., 2019). They have tried solving this issue using liquid media (Guillou et al., 2018). Having tested the IMC67 genotype and with researchers' experience with other genotypes such as CCN51 and TSH565, it has been determined that malformations increase upon prolonging the SSE's individuation process. Therefore, they must be separated from the globular stage. In addition, prolonged time under the effect of growth regulators can trigger these malformations (using 2,4,5-T has been reported to be a trigger) (Kahia et al., 2017;Lázaro et al., 2015). Embryo development and conversion in the MM6 medium are the result of the effect of ANA and GA 3 growth regulators, activated carbon and other culture medium components, such as amino acids, salts, vitamins, high concentrations of macroelements, especially nitrogen as NH 4 NO 3 , and high carbon concentrations originating from glucose. ANA and GA 3 have also been used by Iracheta et al. (2019) to favor cotyledonary embryo development. However, they add 3.7 µM of abscisic acid. Moreover, various authors have been able to develop somatic embryos from Juglans regia (Ali et al., 2010) and Cucurbita pepo (Urbanek et al., 2004) in the culture medium with MS salts, supplemented with NAA and GA 3 . In addition, GA 3 has also been used to elongate and develop embryos and plantlets in other species (Kordestani and Karami, 2008). NH 4 NO 3 also favors embryo conversion to plantlets as a source of inorganic nitrogen (Menke and Zimny, 2001). Activated carbon can have a significant effect on embryo development because it can absorb substances, such as toxic metabolites and phenolic exudates, which inhibit the process, preventing correct maturation. Additionally, when the culture medium becomes darker, it allows accumulating auxins and/or cofactors due to the generated environment (López et al., 2005).

Ex vitro Stage: Acclimation
In the last experiment, no significant difference was found between treatments 20 days after the ex vitro adaptation in number of leaves (Supplemment , Table 4) and stem length (Supplement, Table 5). A higher survival rate was obtained in 70% sand -30% soil from the area (42.1%). The lack of nutrients in sand can be compensated by the cacao crop's soil, which generally favors plantlet growth. (Supplemment, Table 6). The treatment with the lowest survival rate was in the 70% sand -30% coconut coir substrate (22.2%). This is contrary to other reports, which have shown that this mixture has a significant effect on survival, growth expressed as stem length and the number of formed leaves. Coconut coir offers high porosity, aeration capacity and water retention capacity, and sand offers good drainage, consistency and porosity, which could reduce the risk of flooding and favor good rooting (Vasquez, 2005) (Figure 10).

Conclusion
This article offers a study for propagating genotype IMC67 cacao (T. cacao L.) by somatic embryogenesis, achieving 100% callogenesis in 50 days with the INDI staminode, the most PSE's per explant in the globular stage (8) in INDIexp at 45 days to be fragmented and work as the basis for SSE induction Figure 9. Malformed secondary cotyledonary somatic embryos with respect to the development stage in which they were transferred to the MM6 maturation medium. IM (immature), TS (transition) and MA (mature). and the most globular and cotyledonary SSE's (30) at 30 days in CM2. An embryogenic efficiency of 11.7 embryo per explant was obtained for PSE and 15.2 for SSE. Cacao regeneration, maturation and somatic embryo conversion into plants were finally achieved.
In perspective, using a hormone-free medium between secondary embryogenesis and the embryo maturation stage has been proposed for future tests, which would favor the embryogenic potential's expression, prematuration and a reduction of malformations. Evaluating a germination medium that focuses on developing roots to obtain a more robust plant has also been proposed. We recommend evaluating the treatment matrix again with more experimental units and repetitions for ex vitro adaptation.