Indirect somatic embryogenesis of Theobroma cacao L. in liquid medium and improvement of embryo-to-plantlet conversion rate

Indirect somatic embryogenesis of Theobroma cacao L. in liquid medium and improvement of... The establishment of cocoa embryogenic cell lines in liquid medium starting from high frequency somatic embryogenesis (HFSE) callus is described. The growth kinetics of the cultures during the multiplication and the expression steps conducted in 250 mL Erlenmeyer flasks were described for three genotypes selected for their agronomical traits (EET95, EET96, and EET103). The glucose and dissolved oxygen concentrations and the absorption of Murashige and Skoog medium macronutrients (nitrate, ammonium, potassium, sulfate, calcium, phosphorus, and magnesium) were monitored. The multiplication of the −1 embryogenic calluses in a medium containing 2,4,5-trichlorophenoxyacetic acid (2,4,5-T) at 1 mg L , initiated with an inoc- −1 ulation density of 20 g L of callus, was achieved. The growth rate was characterized by two phases, with the second being concomitant with a depletion of phosphorus and magnesium, and a decrease in the embryogenic potential of the callus. The expression of the callus embryogenic capacity was conducted in an auxin-free medium. The embryo production starting from 1 −1 and5g L inoculation densities was compared. When placed in the optimal expression conditions in flasks, 1 g of callus produced 1000 to 1500 embryos within 5 to 7 wk. Finally, two paths for improving the plantlet regenerative capacities of cocoa SE produced in liquid medium were identified. Supplementing the expression medium with myo-inositol used as an osmotic −1 agent at a concentration of 50 g L increased the embryo-to-plantlet conversion rate from 13–16% to 40–48%. A 6-wk culture of the embryos on a maturation medium in Petri dishes optimized their subsequent development into plantlets. . . . . Keywords Cocoa Propagation High frequency somatic embryogenesis callus Embryo-to-plantlet conversion rate Myo-inositol Introduction Cocoa production from T. cacao trees is vulnerable in most producing countries due to the low income of farmers, dis- Theobroma cacao L. trees are grown in the humid tropics to eases, pests, and adverse environmental conditions. To face produce cocoa beans. Four main countries produce 75% of the these challenges, integrated propagation approaches based on total cocoa beans: Ghana, Indonesia, Ivory Coast, and Nigeria somatic embryogenesis (SE), combined with traditional prop- (Food and Agriculture Organization, http://www.fao.org/ agation techniques, have been described for the rapid dissem- faostat/en/#data/QC). In producing countries, cocoa is ination of selected clones (Maximova et al. 2005;Miller 2009; considered as a cash crop, but this is a crucial crop for the Guillou et al. 2014). Due to SE, disease-free materials can be confectionary business, as it provides the key raw material for produced with an orthotropic growth habit, which the farmers chocolate-based products. Nestlé utilizes about 12% of the are familiar with in some countries. world cocoa production. Somatic embryos can differentiate either directly from so- matic cells, which undergo a minimum of proliferation (direct SE), or indirectly after extensive proliferation (indirect SE). * Caroline Guillou However, the distinction between the two processes is not caroline.guillou@rdto.nestle.com always clear. In coffee (Coffea arabica L.), Söndahl and Sharp (1977) preferred to use the terms low- and high frequen- Nestlé Research & Development Centre Tours–Plant Science cy somatic embryogenesis (LFSE versus HFSE) to distinguish Research Unit, 101 avenue Gustave Eiffel BP49716, 37097 Tours them. Low frequency somatic embryogenesis refers to the CEDEX 2, France GUILLOU ET AL. sporadic production of embryos from phenotypically indistin- 2011), and supplementing the medium with PEG or sugar guishable embryogenic callus. In contrast, HFSE is character- alcohols, such as mannitol or sorbitol in rice (Oryza sativa ized by the appearance of friable and highly embryogenic L.; Geng et al. 2008). Myo-inositol, as an osmotic agent, has calluses containing pro-embryogenic masses (PEMs) been less often used to improve SE quality, and only in rose (Söndahl et al. 1985). The specific nature of coffee HFSE (Rosa hybrida L.; Castillón and Kamo 2002) and Norway callus allows the use of a liquid medium for embryogenic spruce (Picea abies (L.) H. Karst.; Egertsdotter and tissue proliferation (Zamarripa et al. 1991; Van Boxtel and Clapham 2015). Berthouly 1996; Etienne 2005). Commonly considered as in- While several studies mention a low conversion of cocoa direct SE, HFSE cultures rapidly scale up the number of po- SE into plantlets, very few reports are available on the effects tential plants that can be produced, because it is possible to of individual components on SE quality in this species. multiply the PEMs. Consequently, HFSE is generally pre- Quainoo and Dwomon (2012) evaluated the effect of a range ferred to LFSE (direct SE) for boosting mass propagation of ABA concentrations from 0 to 50 μM applied for a period procedures (Etienne 2005). The multiplication of the embryo- of 0 to 6 wk. They did not observe any significant positive genic calluses (also referred as the Bproliferation^ or effects from these ABA treatments on the subsequent plantlet Bmaintenance^ step) is generally conducted in an auxin- regeneration rates. Niemenak et al. (2015) reported that in −1 based medium. To initiate the regeneration of embryos (also medium containing sucrose at 60 g L compared to −1 referred as Bexpression^ or Bhisto-differentiation^ step), the 30 g L , maturated SEs contained higher levels of proteins calluses are transferred to an auxin-free medium, which allows involved in storage and tolerance to desiccation. them to express their embryogenic potential. An HFSE procedure for establishing cocoa embryogenic In the case of cocoa SE, most of the published literature cell lines in liquid medium is described in this study. The is based on the progress made by Maximova et al. (2002), growth kinetics and the nutrient uptake during the multiplica- which applied to a particular case of SE that is classified as tion and expression steps in liquid medium were characterized secondary embryogenesis (also referred to as for two clones selected for their agronomic value. The study Bcontinuous,^ Brecurrent,^ or Brepetitive^) (Niemenak et also focused on how to improve the conversion of cocoa SE al. 2008; Garcia et al. 2016). This method consists of cul- into viable plantlets by assessing the effects of increasing the turing cotyledon fragments, from 1 to 2 cm long primary osmolality of the medium during the expression step in flasks, SE, for 2 wk on an auxin-based medium. After transfer to and the effects of the duration of the culture on a maturation an auxin-free medium, each fragment produces 5 to 25 medium in Petri dishes. After these different treatments, the secondary SE within 1 y. Because the secondary SE embryo-to-plantlet conversion rates on germination medium emerges from a mass of non-embryogenic cells developing were compared. from the explants, this method corresponds stricto sensu to the HFSE. However, it leads to the regeneration of SE without the multiplication of embryogenic cells. To make Materials and Methods the most of cocoa SE for creating higher mass propagation potential, a protocol was implemented based on obtaining Plant material In this study, four genotypes were chosen: the and subculturing HFSE callus on a solid medium contain- clone Scavina 6 (SCA6) from the genetic cluster Contamana, ing the auxin 2,4,5-trichlorophenoxyacetic acid (2,4,5-T) and the clones EET95, EET96, and EET103 from the genetic (Masseret et al. 2009). cluster Nacional (Motamayor et al. 2008). Contrary to the As reported for many other plant species, converting the SCA6, which is a clone often used for studies due to its high cocoa SE into whole plants is the most inefficient step of the embryogenic capacity (Li et al. 1998; Maximova et al. 2002), process, and it depends on the genotype (Quainoo and EET95, EET96, and EET103 were chosen for their Ariba Dwomon 2012;Maximova et al. 2014). A common approach flavor and agronomical traits. for improving the SE quality is to create a partial dehydration. In vitro tissues were developed from immature floral buds Numerous studies on conifers have shown the positive effect collected in August 2013, from trees grown in fields in the of increasing gelling agent concentrations and/or the duration Nestlé experimental farm in Chollo, Ecuador. of culture on a medium containing abscisic acid (ABA) (Klimaszewska and Smith 1997; Teyssier et al. 2011). To Induction of primary SE and HFSE callus Primary SE were decrease their water content, olive (Olea europaea L.) SE collected from staminodes following the protocol of Li et al. are advantageously placed onto paper or cellulose acetate (1998). The explants were placed on a PCG medium gelled −1 membranes (Cerezo et al. 2011). Restricting water uptake with Gelrite® at 3 g L for 2 wk. The PCG medium contained −1 −1 using high sucrose concentrations is efficient in alfalfa 2mg L 2,4-dichlorophenoxyacetic acid (2,4-D), 5 μgL −1 (Medicago sativa L.; Anandarajah and McKersie 1990)and thidiazuron (TDZ), and 50 mg L isopentenyl adenine (IP). avocado (Persea americana Mill.; Márquez-Martín et al. The complete composition of the medium is listed in Table 1. THEOBROMA CACAO L. INDIRECT SOMATIC EMBRYOGENESIS... 1 2 3 Table 1. Media composition [ Driver and Kuniyuki (1984); McCown and Lloyd (1981); Murashige and Skoog (1962)] Induction Multiplication Expression Maturation Germination Code PCG SCG CC21 CC2 G80 ENR8 (1) (2) (3) (3) (3) (3) Macronutrients DKW WPM MS MS MS /2 MS (1) (2) (1) (1) (3) (1) Micronutrients DKW WPM DKW DKW MS DKW −1 Vitamins (mg L ) Myo-inositol 200.0 100.0 100.0 100.0 100.0 100.0 Nicotinic acid 1.0 1.0 1.0 1.0 0.5 1.0 Thiamine 2.0 10.0 2.0 2.0 0.7 2.0 Pyridoxine-HCl 1.0 0.5 −1 Amino acids (mg L ) Glycine 2.0 2.0 2.0 2.0 2.0 2.0 L-Lysine 0.4 0.4 0.4 0.4 L-Leucine 0.4 0.4 0.4 0.4 L-Arginine 0.4 0.4 0.4 0.4 L-Tryptophane 0.2 0.2 0.2 0.2 Glutamine 250.0 −1 Plant growth regulators (mg L ) 2,4,5-T 1.000 2,4-D 2.000 2.000 Naphthaleneacetic acid 0.010 Kinetin 0.250 Thidiazuron 0.005 Isopentenyl adenine 50.000 0.200 Adenine-H SO 0.250 0.025 2 4 Abscisic acid 1.000 Gibberellic acid 0.020 −1 Coconut milk (mL L)50.0 −1 Other (g L ) Activated charcoal 1 Glucose 20 20 30 30 40 30 Gelrite® 3 3 3 3 pH 5.6 5.6 5.6 5.6 5.8 5.8 The explants were placed in a dark room at 25°C in 55 mm initiated by selecting HFSE calluses (Fig. 1b). From 0.05 to Petri dishes. After 2 wk, the explants were transferred to SCG 0.10 g fresh weight (FW) of the HFSE callus was transferred −1 −1 medium, which incorporated 2 mg L 2,4-D and 0.25 mg L into 10 mL of the multiplication medium (CC21) contained in kinetin. The explants were then placed on the auxin-free CC2 25-mL Erlenmeyer flasks (cycle 1). The medium was renewed medium for 12 wk. every 3 wk by doubling the volume of the suspensions and Three to 4 mo after disinfection, primary somatic embryos transferring them into larger flasks as follows: 20 mL in 50- (PSE) started to develop from primary callus (Fig. 1a). mL flasks at 3 wk (cycle 2), 50 mL in 100-mL flasks at 6 wk Embryos at the torpedo-stage were finely cut with a scalpel (cycle 3), and 100 mL in 250-mL flasks at 9 wk (cycle 4). and the fragments placed on a CC21 medium containing the After 12 wk of culture in liquid medium, the biomass was −1 −1 auxin 1 mg L 2,4,5-T, and 0.25 mg L adenine. Every 6 wk, collected by filtration through a 50-μm filter. The calluses the embryo fragment calluses were subcultured onto fresh were selected and transferred into 100 mL of the CC21 medi- −1 medium for a maximum period of 8 mo. During this time, um with an inoculation density of 20 g L (cycle 5). HFSE callus started to appear. Thereafter, the embryogenic cell lines (Fig. 1c)were subcultured every 3 wk into 250-mL flasks by transferring Multiplication of the HFSE callus in liquid medium The estab- 2 g of callus into 100-mL of fresh CC21 medium. The suspen- lishment of embryogenic cell lines in liquid medium was sions were cultured on an orbital shaker (New Brunswick™ GUILLOU ET AL. Figure 1. The different steps for Theobroma cacao L. propagation by high frequency somatic embryogenesis (HFSE). (a) Induction of primary SE (PSE, primary somatic embryos; LFSE, low frequency somatic embryogenesis calluses; NEC, non-embryogenic calluses); (b) induction of HFSE callus (SSE, secondary somatic embryos; LFSE, low frequency somatic embryogenic calluses); (c) multiplication: embryogenic clumps in liquid medium; (d) expression: an embryo population at the end of the step; (e) germination: plantlets in germination boxes; (f)normal plantlets; (g) abnormal plantlets. Innova® 2300, Nijmegen, The Netherland) at 120 rpm placed 100 mL of expression medium (CC2). The suspensions were in a dark room at 25°C. cultured in a dark room at 25°C at 120 rpm. Growth kinetics of cell lines after eight multiplication cy- Growth kinetics was established using calluses developing cles were established by measuring various parameters after 0, from the same embryogenic cell lines as the ones studied 1, 3, 5, and 7 wk of culture (named respectively in the results for the multiplication kinetics. The expression starting and discussion T0W, T1W, T3W, T5W, and T7W). The pa- from two different inoculation densities, 1 and 5 g −1 rameters measured were biomass concentration (FW), glucose FW L , was compared. The cultures were analyzed after concentration, dissolved oxygen concentration (DO ), pH, os- 1, 3, 5, and 7 wk (named in the results and discussion molality, conductivity, and MS macronutrient concentrations T1W, T3W, T5W, and T7W) for biomass, glucose, DO , (nitrate, potassium, ammonium, sulfate, calcium, phosphorus, and MS ion concentrations. and magnesium). The embryogenic potential of the calluses, To study the effects of medium components on tissue os- collected at different time periods of a multiplication cycle, molality, two experiments were done. In the first experiment, was measured by counting the number of embryos produced the effects on clone EET103 in the expression step of 15, 30, −1 per gram of callus 4 wk after their transfer into the CC2 ex- and 45 g L myo-inositol (corresponding to a tissue osmolal- −1 pression medium. ity level of 337, 441, and 542 mOsm kg H O , respectively) −1 were assessed in CC2 medium containing glucose at 30 g L Expression of the callus embryogenic potential in liquid me- (corresponding to a tissue osmolality level 261 mOsm kg −1 dium To start the regeneration of the embryos, embryogenic H O ). The second experiment compared the expression of calluses were transferred into 250-mL flasks containing clones SCA6 and EET103 in medium supplemented with THEOBROMA CACAO L. INDIRECT SOMATIC EMBRYOGENESIS... −1 myo-inositol at 50 g L ,after 4wk of culture, andin ame- Osmotic potential: The osmolality was measured in −1 dium containing sucrose at 30 or 80 g L (tissue osmolalities milliosmoles per kilogram of H Ousing amicro-osmometer −1 differed from 193 to 556 mOsm kg H O ). (Model 6/6M, Löser®, Berlin, Germany). The embryo osmo- lality was determined as following the method of Teyssier et Embryo-to-plantlet conversion on solid medium After collec- al. (2011): Around 10 embryos were stored at − 20°C and tion from the flasks, the embryos at the torpedo-stage were thawed at 50°C for 10 min. After centrifugation at 8000×g individually transferred into 90 × 20-mm Petri dishes contain- for 15 min, the osmolality was measured in 100 μLof the −1 ing G80 maturation medium (Fig. 1d) containing 1 mg L supernatants. −1 ABAand1gL charcoal (Lopez-Baez et al. 1993). Each Macronutrients: Residual concentrations of MS ions in the Petri dish was inoculated with 30–40 embryos. The embryos spent medium were measured by conventional methods were incubated in a dark room at 25°C. (Laboratoire de Touraine, Tours, France). Nitrate and ammo- After 0, 3, 6, and 9 wk of maturation (named in the results nium levels were determined by flow analysis and spectromet- and discussion T0W, T3W, T6W, and T9W), the embryos ric detection according respectively NF-EN ISO 13395:1996 were individually subcultured onto ENR8 germination medi- and NF-EN ISO 11732 (13395:2005). Sulfate level was deter- um in 145 × 100 × 60-mm styrofoam boxes (Dominique mined by ionic liquid chromatography according NF-EN ISO Dutscher S.a, ref. 017002, Issy les moulineaux, France). The 10304-1:2007/2010, and the levels of cations potassium, cal- embryos were germinated with a dark/light cycle of 12/12 h at cium, phosphorus, and magnesium were determined by induc- −2 −1 a light intensity of 65 μmol m s with two LED Tubes tively coupled plasma optical emission spectrometry accord- Green Power Deep Red/White/Medium blue, 18 W, ing NF-EN ISO 11885:2007. All the protocols are available −1 24 μmol S , (Philips Lighting, Zurich, Switzerland) in a cul- on the website https://www.boutique.afnor.org/. ture room at 25°C. Water content (WC): After collection from the flasks, em- The embryo-to-normal plantlet conversion rates (%) were bryos were weighed to determine their FW before being dried measured after 8 wk by counting the embryos that had devel- in an oven (UN55, Memmert®, Schwabach, Germany) at oped into normal plantlets (Fig. 1e). The normal plantlets had 90°C for 24 h. Just after being taken out of the oven, the a rooting apex, and, at least, a 2-cm stem height, and one pair samples were weighed again to determine their dry weight of true leaves with a minimum length of 1 cm (Fig. 1f). These (DW). The WC was calculated using the following equation: fully developed plantlets with standard characteristics were WC (%) = (DW / FW) × 100. able to be successfully acclimated to the greenhouse. In some experiments, the number of embryos was counted, which had Data analysis Statistical analyses, including ANOVA using developed abnormal plantlets characterized by thin shoots, the general linear model (GLM), and tests to compare treat- and dark, thickened, and stunted leaves (Fig. 1g). These types ment averages, were performed using Minitab® 17.3.1 of plantlets are not able to survive to the transfer into the (Minitab® Inc., State College, PA). The relevance and ranking greenhouse. factors were calculated by comparing the averages using the All the chemicals are from Duchefa, Haarlem, The Netherlands Tukey test at the 5% threshold. and Petri dishes from VWR, Fontenay-sous-Bois, France. The media pH was adjusted with either 0.1 or 1 N hydrochloric acid and potassium hydroxide. Media were autoclaved at 121°C for Results and Discussion 20 min and glass vessel at 121°C for 30 min in a Lequeux auto- clave, Paris, France. Induction of HFSE callus Young secondary SE (SSE) became observable after 1 mo of culture on the CC21 solid medium −1 Analytical measurements Glucose: Glucose level in the media containing 1 mg L 2,4,5-T (Fig. 1b). Depending on the was measured using a high-performance liquid chromato- genotype, small embryogenic calluses (HFSE) appeared graph (Dionex, IC5000, Thermo Fisher Scientific®, within 6 to 10 mo (data not shown). They were character- Waltham, MA). The separation was achieved using a 4 × ized by a granular and friable appearance and a yellowish 150-mm ion exchange column (Dionex CarboPac, PA10, color (Fig. 1b). Secondary SE and HFSE calluses did not Thermo Fisher Scientific®). develop directly from the fragments of primary SE, but Dissolved oxygen concentration: Oxygen was non- from brown to dark brown callus. invasively measured by using PSt3 sensor spots placed on The differences observed between the occurrence of cocoa the bottom of the flasks along with an optic meter (Firebox secondary SE and embryogenic callus are very similar to the 3, PreSens®, Regensburg, Germany). differences reported in Arabica coffee, between the LFSE and Conductivity and pH: Both the conductivity and pH values HFSE processes (Söndahl et al. 1985;Etienne 2005). In this were measured directly in the cell culture broth using a digital species, SE takes place in two waves in leaf explant cultures. apparatus (Model 3540, Jenway®, Chelmsford, UK). The first wave (LFSE) corresponds to the sporadic and GUILLOU ET AL. transient production of isolated SE within 2 to 3 mo, and the cell lines have to be initiated in small volumes of medium, second wave (HFSE) corresponds to the production of very generally by transferring 0.05 to 0.10 g FW of selected callus embryogenic callus within 5 to 10 mo. into 10 mL of medium. Next, the volume of the medium was Importantly, the occurrence of cocoa HFSE callus was doubled every 3 wk until 12 wk. At this time (cycle 5), the found only if fragments were isolated from the 2–3-mm tor- biomass consisted of a mixture of secondary SE and calluses pedo-stage, rather than from 1-cm cotyledonary-stage embry- of various types. To initiate the maintenance of the embryo- os, which other authors used as starting material to obtain genic cell lines, the friable and yellowish calluses were select- secondary SE (Maximova et al. 2002;Garcia et al. 2016). ed once again if necessary and transferred into 250-mL flasks −1 In the frame of a core collection establishment, it was ob- with an inoculation density of 20 g FW L to limit the differ- served that almost all of the genotypes that are reactive to entiation of embryos and to maintain HFSE calluses. The cell primary SE also respond positively to secondary SE, which lines were composed of large aggregates, ranging from 0.3 to confirms the results of Maximova et al. (2005), Masseret et al. 1.5 mm (Fig. 1c). (2009), and Guillou et al. (2014). However, there were some Growth kinetics. Typical profiles of the growth and of sub- genotypes that produced primary and secondary SE very well, strate consumption of EET96 and EET103 embryogenic cell but for which obtaining and selecting HFSE calluses remained lines are shown in Figs. 2 and 3. The biomass concentrations −1 a long and laborious process. It is estimated that 50 to 60% of ranged from 20 up to 60–80 g FW L within 5 wk. During the genotypes gave rise to HFSE callus using this protocol this period, the growth was strictly linear and followed the (data not shown). equations y = 11.14x + 20 with an R value of 0.99 for EET96, and y = 8.39x + 20 with an R value of 0.98 for Multiplication of the HFSE callus in liquid medium EET103. Afterwards, the growth rates increased strongly be- Establishment of embryogenic cell lines. The HFSE calluses tween T5Wand T7W, as the biomass concentration was three- must be transferred into a liquid medium with a sufficiently fold higher within only 2 wk. It reached a final value of 224 −1 −1 high inoculation density of 5 to 10 g FW L as a minimum. and166gFWL at T7W for the clones EET96 and EET103, Indeed, with lower inoculum, the suspensions frequently respectively (Figs. 2a, 3a). By T7W, the glucose was tended to differentiate totally into embryos despite the pres- completely consumed from the medium (Figs. 2b, 3b). ence of the 2,4,5-T in the CC21 medium. For this reason, and Because oxygen can be a limiting factor in liquid medium, because of the limited quantity of available HFSE calluses, the the cultures were checked for hypoxia. The DO did not drop −1 Figure 2. Multiplication step: growth kinetics and nutrient uptake of concentrations (mg L ). For a, b,and c, the data represent the averages Theobroma cacao L. cv. EET96 embryogenic cell line. (a)Fresh of four replicates, each one being a 250-mL flask containing 100 mL of −1 −1 −1 weight (g L ); (b) glucose concentration (g L ) and DO (%); (c) CC21 medium initiated with an inoculation density of 20 g L . Bars tissue osmolality, conductivity, and pH; (d) total macronutrients, nitrate, represent standard errors unless too small to visualize. For d, e,and f, −1 and potassium concentrations (mg L ); (e) ammonium, sulfate, and measurements were taken after combining the spent medium of four −1 calcium concentrations (mg L ); (f) phosphorus and magnesium flasks. THEOBROMA CACAO L. INDIRECT SOMATIC EMBRYOGENESIS... −1 Figure 3. Multiplication step: growth kinetics and nutrient uptake of concentrations (mg L ). For a, b,and c, the data represent the averages Theobroma cacao L. cv. EET103 embryogenic cell line. (a)Fresh of four replicates, each one being a 250-mL flask containing 100 mL of −1 −1 −1 weight (g L ); (b) glucose concentration (g L ) and DO (%); (c) CC21 medium initiated with an inoculation density of 20 g L . Bars tissue osmolality, conductivity, and pH; (d) total macronutrients, nitrate, represent standard errors unless too small to visualize. For d, e,and f, −1 and potassium concentrations (mg L ); (e) ammonium, sulfate, and measurements were taken after combining the spent medium of four −1 calcium concentrations (mg L ); (f) phosphorus and magnesium flasks. below 60% throughout the cultures. In another experiment without auxin (Fig. 4). A strongly significant effect (P < (data not shown), after shaking the cultures was stopped, the 0.001) of the clones (F = 12.8), and mostly of the duration DO critical value under which the consumption rate in O (F = 39.5) on the embryogenic potential was detected. It 2 2 decreases was determined and indicated hypoxia regime. For reached its highest value after 3 wk of culture, and then sig- the cocoa calluses cultured in suspensions, this value lies be- nificantly decreased at 5 wk for cultivar EET103. tween 35 and 45%. Consequently, it can be assumed that the The cocoa embryogenic cell lines were characterized by DO was not a limiting factor during the culture period shown biphasic growth, a linear growth up to T5W, followed by faster in the Figs. 2 and 3. The pH started at 4.6 in T0W before reaching values ranging from 6.0 to 7.0 (Figs. 2c, 3c). The tissue osmolality decreased from an initial value of 269, to −1 values close to 50 mOsm kgH O at the end of the culture period, with the medium conductivity dropping from 5.2 to −1 1.9–2.4 mS cm . −1 The sum of the MS macronutrients is equal to 4.0 g L in −1 fresh CC21 medium (T0W). It dropped to 0.8–1.0 g L after 7 wk of culture (Figs. 2d, 3d). The ammonium, potassium, and calcium ions were consumed at 40 to 60% of their initial concentrations at TW7. The nitrate and sulfate were metabo- lized at a higher rate, but 5 to 10% of their initial concentra- tions remained at the same date. Finally, among the sampled macronutrients, only magnesium and phosphorus were completely depleted, which occurred as early as T3W for both Figure 4. Multiplication step: evolution of the embryogenic potential of clones (Figs. 2f, 3f). Theobroma cacao L. embryogenic cell lines EET96 and EET103. After Evolution of the embryogenic potential of the calluses. The various durations of culture in CC21 medium, 0.1 g of callus tissue was transferred into 100 mL CC2 medium. The embryos were counted by the evolution of the tissue qualities within a multiplication cycle naked eye after 4 wk. The data represent the averages of 16 replicates in CC21 medium was monitored by measuring the embryo- −1 (four flasks × four measures flask ). Data followed by different letters genic potential of the calluses (number of embryos that could are significantly different at P <5% (Tukey’s test completed separately −1 for each clone). be produce g FW ) once transferred to expression conditions GUILLOU ET AL. −1 Figure 5. Expression step: growth kinetics of Theobroma cacao L. cv. inoculated with an inoculation density of 1 or 5 g L . Bars represent EET96 embryo regeneration. The data are the averages of four replicates, standard errors unless too small to visualize. For d, e,and f, measurements each one being a 250-mL flask containing 100 mL of CC2 medium were taken after combining the spent medium of four flasks. growth up to T7W. The second phase was concomitant with a Chaturvedi 2012). It should be noted that the growth cycle decrease in the embryogenic capacity of the calluses, which was monitored in terms of fresh biomass. The tissues could may be due to phosphorus and magnesium deficiencies. have a higher WC at the end of the culture period due to the −1 Other authors have described such deficiencies using the MS low osmolality of the medium (50 mOsm kgH O ). It is macronutrients during plant cell cultures in liquid medium possible that the decrease in the embryogenic potential of the (Archambault et al. 1994; Azevedo et al. 2008; Singh and calluses at this time could also be attributed to a higher WC. Figure 6. Expression step: growth kinetics of Theobroma cacao L. cv. represent standard errors unless too small to visualize. For d, e,and f, EET103 embryo regeneration. The data are the averages of four measurements were taken after combining the spent medium of four replicates, each one being a 250-mL flask containing 100 mL of CC2 flasks. −1 medium inoculated with an inoculation density of 1 or 5 g L . Bars THEOBROMA CACAO L. INDIRECT SOMATIC EMBRYOGENESIS... extent on the genetic background (P = 0.021, F = 5.8). The ability of the embryo to develop plantlets, once transferred to maturation followed by a germination medium, tended to increase between T3W and T7W but not significantly (Fig. 7). On the contrary to the embryogenic potential, there was no influence of the inoculation density on the SE quality (P = 0.36). It is not known whether the inhibition of the embryo −1 production in the cultures initiated at 5 g FW L resulted from deficiencies in nutrients due to higher consumption rate of nutrients or to decreased tissue osmolality and me- −1 dium conductivity in the cultures initiated at 5 g FW L (Figs. 5c, 6c). Whatever clone was tested, or the inocula- − + + tion densities used, the DO and the NO ,K ,NH ,and 2 3 4 Ca ions were clearly not limiting factors throughout the Figure 7. Expression step: evolution of the quality of the Theobroma kinetics (Figs. 5d, e, 6d, e). For clone EET96, inhibition was cacao L. cvs. EET96 and EET103 embryos. For the embryo-to-plantlet conversion rates, at least 20 embryos per replicate were transferred onto observed at T5W. At this date, the magnesium and phosphorus G80 medium and after 3 wk on ENR8 medium. The data (averages of were the nutrients with the lowest concentrations, but their four replicates) followed by different letters are significantly different at −1 levels remained higher than 1.8 mg L in both cultures. For P < 5% (Tukey test completed for each clone separately). clone EET103, inhibition was quantifiable at T7W, and the phosphorus was totally exhausted at T5W in the cultures ini- −1 Considering these results, a frequency of 3 wk to subcul- tiated at 5 g L . For these cultures, very low concentrations in −1 +2 −1 −2 ture, the cell lines was selected for the standard procedure. glucose(0.7gL ), in Mg (0.9 mg L ), and in SO −1 Under this 3-wk regime, the multiplication rate of the biomass (7 mg L ) at T7W were noted. ranged from 2 to 3 at each subculture. However, as reported Effect of medium renewals. For EET95 cell line, the ex- −1 early in the history of plant SE (Smith and Street 1974), a pression step initiation wascompared with 1 and5gFWL , decline of the capacity of the cocoa calluses to regenerate with and without renewing the medium twice a week. Once embryos was observed with the number of subculture cycles again, a strong inhibition of the expression was linked to the (data not shown). Therefore, this number was restricted to higher inoculation density (Fig. 8a). After 33 d, less than 50% 10 cycles, a total of 30 wk of culture, in a liquid multiplication of the biomass was composed of embryos in cultures initiated −1 medium. In T. cacao, the long-term decrease in the embryo- with 5 g L , instead of complete differentiation in those ini- −1 genic potential of secondary SE has been associated with an tiated at 1 g L (Fig. 8b). As a consequence of the medium increase of global DNA methylation levels (Rodrίguez Lόpez renewals, the inhibition was totally suppressed as the biomass −1 et al. 2010;Adu-gyamfi et al. 2016;Quinga et al. 2017). increased to 150 g L , with 95% of the biomass consisting of −1 embryos. The inhibition of the differentiation at 5 g FW L Expression of the callus embryogenic capacity in liquid medi- did not result from a limitation in oxygen because the DO um Growth kinetics. Using calluses produced from the cell level stayed above 70% (Fig. 8c). When the medium was lines described above, the growth kinetics was compared renewed, the DO dropped to a lower level than without re- during the expression step starting with two inoculation newal, as early as the 14th day (Fig. 8c)(P =0.011, t test), −1 densities, 1 and 5 g FW L for the clones EET96 and which indicated a higher demand for oxygen in the suspen- EET103 (Figs. 5 and 6, respectively). The highest concen- sions. This fact suggests that the growth and the differentiation trations in embryos were reached after 5 wk for the cultures process were hampered before the 21st day, when the expres- −1 −1 initiated at 5 g FW L and ranged between 2500 and 3500 sion was started at 5 g FW L . If this observation does not −1 embryos L (Figs. 5a, 6a). permit exclusion of nutrient deficiencies as the cause of inhi- If the embryo concentrations were proportional to the size bition, it does provide an argument in favor of the presence of of the inoculum up to T3W, this was no longer the case after inhibitory compounds released by the tissues. this date. In fact, the highest values of embryogenic potential, Based on these results, the standard procedure selected −1 expressed as the number of embryos per inoculated gram, consisted of starting the expression with 1 to 2 g FW L −1 were obtained in the cultures inoculated at 1 g FW L at inoculum and renewing the medium after 3 to 4 wk of culture. T5W for clone EET96 and T7W for clone EET103, and The embryos were collected after three to four additional −1 ranged between 1000 and 1500 embryo g . The embryogenic weeksofgrowthinliquidmediumbeforedirecttransfer onto potential highly depends on the duration (P <0.001,F =40.9), the maturation medium, or after a 3 to 5 wk of development in on the inoculation density (P < 0.001, F = 18.2), and to a lesser a temporary immersion bioreactor (Guillou et al. 2014). GUILLOU ET AL. Figure 8. Expression step: effect of the inoculation density and of a twice-weekly medium renewal of Theobroma cacao L. cv. EET95. (a) Embryo suspensions after 5 wk. From left to right: inoculation density of 1 and −1 5gL with medium renewal. (b) Total (whole biomass) and embryo FW at 5 wk. The data (averages of three replicates) followed by different letters are significantly different at P <5% (Tukey test). (c) Evolution of the DO .DO was measured just 2 2 before the medium was renewed. The data are the averages of three replicates. It was observed that inhibition of the cocoa SE expression depletion on the inhibition of cocoa SE production. Generally, was linked to high initial densities, similar to results reported to the naked eye, embryos are smaller in the cultures initiated −1 in other plant species like coffee (Zamarripa et al. 1991), car- at 5 g FW L as early as the third week at which time the rot (Daucus carota L.; Kobayashi et al. 2000), or larch (Larix concentrations of phosphorus and the magnesium were clearly leptolepis Gordon; Umehara et al. 2005). Among the elements not yet limited. This point was confirmed by the experiment monitored, deficiencies in phosphorus and Mg are most that renewed the medium twice weekly. The demand in oxy- likely to occur during the expression step, similarly to the gen in the high-density cultures was hampered before the end multiplication step. However, these ions were consumed more of the 3rd week. Taken together, these observations suggest slowly than during the multiplication phase, because they that this inhibition was most likely associated with the diffu- were only completely consumed at T5W or T7W, instead of sion of SE self-inhibitors from the tissues into the medium, at T3W. Interestingly, Minyaka et al. (2008) reported that similar to those identified in other species, such as 4- supplementing a medium with MgSO improved the induc- hydroxybenzyl alcohol in carrot (Kobayashi et al. 2000), or tion and the development of secondary SE on solid medium. vanillyl benzyl ether in larch (Umehara et al. 2005). Optimal cocoa SE probably requires these elements in higher Effect of an increase in the tissue osmolality. Myo-inositol concentrations than the elements provided by the MS basal is known to be involved in seed development, seed desicca- medium commonly used in plant cell cultures. Nevertheless, tion, signaling pathways, cell wall biosynthesis, and various results are insufficient to determine a definitive role of nutrient stress-related responses (Loewus and Murthy 2000). For this THEOBROMA CACAO L. INDIRECT SOMATIC EMBRYOGENESIS... Table 2. Effect of myo-inositol supplementation during the expression plantlet conversion rates were measured after 3wkonmaturation step on the embryo quality of Theobroma cacao L. cv. EET103. The medium and 8 wk on germination medium (data are the averages of −1 expression step was initiated with an inoculation density of 5 g L in five replicates of at least 20 embryos). (± are the standard errors, data 250-mL flasks containing 100 mL of CC2 medium. Fresh weight (FW), followed by different letters are significantly different according to the embryo number, and water content (WC) were measured after 30 d of Tukey test (P <5%)) expression (data are the averages of five replicates). The embryo-to- Glucose Myo-inositol Molarity Osmolality fresh medium FW Embryo number WC Embryo-to-plantlet conversion rate −1 −1 −1 −1 −1 gL gL MmOsmkgH O gL Number L % % Normal 30 0 0.17 261 65 ± 27 a 1114 ± 131 a 90.2 ± 2.7 a 16.4 ± 2.1 a 30 15 0.25 337 71 ± 13 a 876 ± 191 ab 88.9 ± 0.5 ab 25.4 ± 1.9 ab 30 30 0.33 441 45 ± 21 ab 584 ± 242 bc 82.7 ± 3.5 bc 27.7 ± 2.3 b 30 45 0.42 542 32 ± 7 b 272 ± 118 c 82.5 ± 1.1 c 33.0 ± 3.2 b reason, myo-inositol was tested as an osmotic agent rather 48.3% instead of 13.3% for SCA6, and 40.8% instead of than other sugar alcohols, such as mannitol or sorbitol, in 15.8% for EET103. Concomitantly, the rates of conversion order to produce higher frequencies of cocoa SE conversion into abnormal plantlets were reduced. The incidence of this into plantlets. supplementation in medium in which the glucose was re- In the first experiment, the growth and the embryo forma- placed by sucrose was also checked. In this case, this sucrose tion were partially inhibited with myo-inositol at 30 and treatment led to a significant increase of the embryo quality for −1 45 g L (Table 2). After 30 d, the embryo WC decreased SCA6, but not for EET103. When the sucrose concentration −1 from 90% in the control to 82%, with the highest myo- was increased from 30 to 80 g L , the embryo quality tended inositol concentration. When the tissue osmolality was in- to improve though not significantly for both clones, despite creased, the morphology of the embryos changed. They the fact that this treatment reduced their WC. showed a higher degree of opacity, and more elongated hypo- As observed by Niemenak et al. (2015), a treatment with a cotyls with anthocyanin pigmentation (Fig. 9). After culture high sucrose concentration modifies the appearance of the on a maturation medium followed by a germination medium, embryos, which indicates a higher degree of maturation. the aptitude of the embryos to develop normal plantlets was However, myo-inositol was more efficient than sucrose for improved proportionally to the myo-inositol concentration, increasing the SE quality. Increasing the tissue osmolality from 16 to 33% with myo-inositol at a concentration of generally inhibited the formation of new embryos from −1 45 g L (Table 2). micro-calluses attached to the embryos. Consequently, the Therefore, in the second experiment, the medium was sup- embryo populations appeared more uniform at the end of the −1 plemented with myo-inositol at 50 g L once the embryos expression step. The comparison of the effect of myo-inositol reached the early torpedo-stages (after 4 wk of culture). After with other osmotic agents (mannitol, sorbitol, and maltose), an additional 4 wk, the increase in the fresh-medium myo- with identical molarity, and the continuation of these treat- −1 inositol concentration from 10 to 45 g L led to an increase ments beyond the expression stage such as during maturation −1 in the embryo osmolality, from 252 to 600 mOsm kgH O on solid medium, is currently underway. for SCA6, and from 304 to 502 for EET103 (Table 3). This Embryo maturation on solid medium. The embryo-to- coincided with slight but statistically significant tissue WC plantlet conversion rates were compared for clones SCA6, decreases of 6%. This treatment was followed by significantly EET96, and EET103 after 0, 3, 6, and 9 wk of maturation higher conversion rates into normal plantlets, with a rate of on G80 medium (Table 1). When the embryos were directly −1 Figure 9. Expression step: effect of supplementing CC2 medium with without myo-inositol; (b) embryos in CC2 medium with 15 g L of −1 different myo-inositol concentrations on Theobroma cacao L. cv. myo-inositol; (c) embryos in CC2 medium with 30 g L of myo- −1 EET103 embryos after 4 wk. (a) Embryos in control CC20 medium inositol; (d) embryos in CC2 medium with 45 g L of myo-inositol. GUILLOU ET AL. Table 3. Effect of an increase of the osmolality during the last week of content (WC) and osmolality were measured at the end of the the expression step on the quality of the embryos cvs. SCA6 and EET103. expression (data are the averages of four replicates of 30 embryos for −1 After 4 wk of expression on CC2 medium containing glucose at 30 g L , WC, of four replicates of 10 embryos for the osmolality). For the −1 embryos were cultured (50 embryos 100 mL )fouradditional wk conversion rate, data are the averages of four replicates of 30 embryos. in CC2 supplemented or not with myo-inositol or sucrose at different (± are the standard errors, data followed by different letters are concentrations. The embryos were subcultured on maturation medium significantly different according to the Tukey test (P <5%)) during 3 wk then 8 wk on germination medium. The embryo water Clone Glucose Sucrose Myo-inositol Molarity Osmolality fresh Osmolality WC Embryo-to-plantlet conversion rate medium embryos −1 −1 −1 −1 −1 gL gL gL MmOsmkgH O mOsm kg H O % % Normal % Abnormal 2 2 SCA6 30 0.17 274 252 ± 15 d 92.4 ± 0.4 a 13.3 ± 7.7 cd 67.5 ± 12.9 a 30 50 0.44 556 600 ± 29 a 86.3 ± 0.5 c 48.3 ± 7.9 ab 45.8 ± 9.2 ab 30 0.09 193 314 ± 15 c 91.7 ± 0.2 b 9.2 ± 9.9 b 63.3 ± 7.2 a 30 50 0.37 468 600 ± 17 a 86.0 ± 0.6 d 61.7 ± 13.5 a 32.5 ± 10.3 b 80 0.23 373 557 ± 7 b 87.4 ± 0.2 c 33.3 ± 8.2 bc 55.5 ± 10.7 a EET103 30 0.17 274 304 ± 5 c 92.6 ± 0.1 a 15.8 ± 5.7 b 71.6 ± 5.8 a 30 50 0.44 556 502 ± 15 ab 86.7 ± 0.4 c 40.8 ± 6.3 a 50.8 ± 5.7 b 30 0.09 193 252 ± 65 d 92.1 ± 0.3 b 15.8 ± 3.2 b 57.5 ± 11.1 ab 30 50 0.37 468 533 ± 26 a 85.5 ± 0.6 d 26.7 ± 6.1 ab 61.6 ± 11.7 ab 80 0.23 373 472 ± 13 b 86.7 ± 0.8 c 25.0 ± 16.4 ab 62.5 ± 7.9 ab transferred from the CC2 expression medium onto the ENR8 51, and 47%, respectively) were observed after 6 wk of mat- germination medium (T0W), their conversion into normal uration. On average for the three clones, the frequency of plantlets was practically nil (Fig. 10). For the three clones regeneration into normal plantlets increased from 13 to 47% SCA6, EET96, and EET103, the best conversion rates (43, when the maturation duration was extended from 3 to 6 wk. Meanwhile, the regeneration into abnormal plantlets dropped from 31 to 6%. After 9 wk of maturation, the embryo quality decreased. During this kinetic period on maturation medium, neither the osmotic potential of the embryos nor their WC changed (data not shown). The WC never decreased below 90%, particularly between T3W and T6W (Fig. 11 shows the evolution of the appearance of the embryos). The higher qual- ityafter 6wk of maturation could notbe correlated with a greater opacity, or any other signs of maturation such as a partially dehydrated appearance. However, the embryos started to present signs of oxidation after 9 wk, which could be responsible for a lower quality at this period. The positive effect of the maturation step could come from certain components of the G80 medium, and most probably from active charcoal and/or from ABA, which are both pres- −1 −1 ent at 1 g L and1mg L , respectively. Active charcoal is known to trap toxic compounds such as residual or endoge- nous auxins, which affects the SE quality in soybean (Buchheim et al. 1989). The combination of charcoal and −1 ABA at 5 to 25 mg L improved the production of well- shaped SE in coniferous species, despite the fact that ABA was trapped by charcoal (Pullman et al. 2005). Free ABA was indeed undetectable in the G80 medium in the present exper- Figure 10. Effect of the duration of culture on G80 maturation medium iment (data not shown). Finally, the hypothesis that the posi- on the conversion into normal or abnormal plantlets Theobroma cacao L. tive impact of the maturation on the SE quality does not result cvs. SCA6, EET96, and EET103. Selected embryos were subcultured on G80 maturation medium in 90 × 20-mm Petri dishes (40 embryos per from the G80 medium itself, but from the microenvironment dishes). For their conversion into plantlets, they were transplanted onto provided by Petri dishes cannot be excluded. This hypothesis ENR8 germination medium. The conversion rates were measured after is currently being investigated. In spruce, a confined environ- 8 wk. Data (averages of 10 replicates of 40 embryos) followed by ment induced a larger number of well-formed mature embryos different letters are significantly different according to Tukey test completed separately for each clone (P <5%). and a lower rate of precocious germination than a ventilated THEOBROMA CACAO L. INDIRECT SOMATIC EMBRYOGENESIS... Figure 11. Evolution of the embryos during the maturation step of Theobroma cacao L. cv. EET96. T0W, embryos after the harvest without maturation phase; T3W, embryos after 3 wk on maturation medium; T6W, embryos after 6 wk on maturation medium; T9W, embryos after 9wk on maturation medium. environment, and the addition of 5% CO or 10% O in a treatments increased the rate of conversion into normal plantlets 2 2 vented environment promoted SE (El Meskaoui and while they decreased the conversion into abnormal plantlets, Tremblay 1999;El Meskaoui et al. 2006). which indicated that the morphological issue was not the con- sequence of somaclonal variation, but most likely due to phys- iological disorders resulting from incomplete maturation. Conclusions One of the important strategies for future improvements of the quality of the embryogenic cocoa tissues, including the embryogenic potential of the callus at the end of multiplica- This is the first detailed report describing the establishment of tion, and the capacity of the embryos to regenerate plants at cocoa embryogenic cell lines and their growth kinetics in liq- the end of the expression, consists of implementing these steps uid medium. This protocol is based on obtaining and rigor- in a bioreactor system that continuously monitors and controls ously selecting HFSE callus. The main points necessary to the environmental conditions. There are some studies that in- succeed in establishing the cell lines are as follows: (1) at dicated SE optimization in a stirred bioreactor by controlling the beginning of the process, the primary embryos used as the DO (Archambault et al. 1994;DeFeria et al. 2003; explants must be taken at the torpedo-stage to obtain HFSE callus, and (2) particular attention must be focused on the Shimazu and Kurata 2003). In the framework of the Nestlé Cocoa Plan, the method inoculation densities of the suspensions, as in the case of described in this study was used to regenerate 260,000 plants, Robusta coffee (Ducos et al. 2007), because high densities which, after their acclimation, have been sent to producing hamper the precocious embryo differentiation, which tends countries, mainly Ecuador and Ivory Coast (Guillou et al. to occur despite the presence of the 2,4,5-T in the medium. 2014). The Indonesian Coffee and Cocoa Research Institute Once placed in the optimal expression conditions in flasks, (ICCRI) scaled up its production to deliver 74 M cocoa so- 1 g of callus produced 1000 to 1500 embryos. Compared to matic seedlings during the period from 2009 to 2011 and certain other species, the resulting embryogenic potential is rel- 5 −1 modified certain steps for particular genotypes (Sena Gomes atively low. Productivity levels of over 5 × 10 embryos g of et al. 2015). To address the issues of cocoa production by callus have been frequently reported in carrot, up to 2 × 10 −1 diffusing more productive varieties with resistance to biotic embryos g of callus in Robusta coffee (Zamarripa et al. 1991), and 0.4 to 0.8 × 10 in Arabica coffee (Van Boxtel and and abiotic stresses, the application of HFSE will be of interest for future breeding programs, and for mass propagation of Berthouly 1996;DeFeria et al. 2003, Etienne 2005). The fre- validated planting materials. quent occurrence of abnormal cocoa plantlets with shoots con- tinuously producing cotyledon-like leaves and long internodes Acknowledgements We would like to thank the Master 2 students who was noted, which is a phenomenon that has been reported by contributed enthusiastically to this research: Quentin Claverie (University other authors (Maximova et al. 2005; Quainoo and Dwomon of Bordeaux) for the studies on the growth kinetics, Gaetan Augier 2012). Two paths were identified for improving the quality of (University of Angers) for myo-inositol experiments, Mathilde Piquet (University of Rennes) for the study on the duration of maturation. We cocoa SE regenerated in liquid medium. One was to increase the would also like to thank Dr. James Mc Carthy for the critical review of the osmolality during the expression step and the second was to manuscript. extend the maturation step in Petri dishes. Interestingly, both GUILLOU ET AL. Open Access This article is distributed under the terms of the Creative Guillou C, Fillodeau A, Brulard E, Verdier D, Simon M, Landmann A, Commons Attribution 4.0 International License (http:// Lausanne F, Fontanel A, Ducos JP, Buchwalder A, Broun P (2014) creativecommons.org/licenses/by/4.0/), which permits unrestricted use, Nestlé Cocoa Plan: cocoa propagation by somatic embryogenesis. distribution, and reproduction in any medium, provided you give appro- In: Park YS, Bonga JM (eds) Proceedings of the 3rd International priate credit to the original author(s) and the source, provide a link to the Conference of the IUFRO Unit 2.09.02 on BWoody plant production Creative Commons license, and indicate if changes were made. integrating genetic and vegetative propagation technologies^,8–12 September, vol 2014. Vitoria-Gasteiz, Spain, pp 75–80 Klimaszewska K, Smith DR (1997) Maturation of somatic embryos of Pinus strobus is promoted by a high concentration of gellan gum. Physiol Plant 100:949–957 References Kobayashi T, Higashi K, Sasaki K, Asami T, Yoshida S, Kamada H (2000) Purification from conditioned medium and chemical identi- Adu-Gyamfi R, Wetten A, Lopez CMR (2016) Effect of cryopreservation fication of a factor that inhibits somatic embryogenesis in carrot. and post-cryopreservation somatic embryogenesis on the epigenetic Plant Cell Physiol 41:268–273 fidelity of cocoa (Theobroma cacao L.). PLoS One 11(7):e0158857. Li Z, Traore A, Maximova S, Guiltinan MJ (1998) Somatic embryogen- https://doi.org/10.1371/journal.pone.0158857 esis and plant regeneration from floral explants of cacao Anandarajah K, McKersie BD (1990) Manipulating the desiccation tol- (Theobroma cacao L.) using thidiazuron. In Vitro Cell Dev Biol- erance and vigor of dry somatic embryos of Medicago sativa L. with Plant 34:293–299 sucrose, heat shock and abscisic acid. Plant Cell Rep 9:451–455 Loewus FA, Murthy PPN (2000) myo-Inositol metabolism in plants. Plant Archambault J, Williams RD, Lavoie L, Pepin MF, Chavarie C (1994) Sci 150:1–19 Production of somatic embryos in a helical ribbon impeller bioreac- Lopez-Baez O, Bollon H, Eskes A, Petiard V (1993) Somatic embryo- tor. Biotechnol Bioeng 44:930–943 genesis and plant regeneration from flower parts of cocoa Azevedo H, Dias A, Tavares RM (2008) Establishment and characteriza- Theobroma cacao L. C R Acad Sci III 316:579–584 tion of Pinus pinaster suspension cell cultures. Plant Cell Tissue Márquez-Martín B, Sesmero R, Quesada MA, Pliego-Alfaro F, Sánchez- Organ Cult 93:115–121 Romero C (2011) Water relations in culture media influence matu- Buchheim JA, Colburn SM, Ranch JP (1989) Maturation of soybean ration of avocado somatic embryos. J Plant Physiol 168:2028–2034 somatic embryos and the transition to plantlet growth. Plant Masseret B, Gianforcaro M, Bouquet JF, Brulard E, Florin B (2009) Physiol 89:768–775 Somatic embryogenesis applied to the creation of a cocoa collection. Malays Cocoa J 5:1–10 Castillón J, Kamo K (2002) Maturation and conversion of somatic em- Maximova SN, Alemanno L, Young A, Ferriere N, Traore A, Guiltinan bryos of three genetically diverse rose cultivars. Hortic Sci 37:973– MJ (2002) Efficiency, genotypic variability, and cellular origin of primary and secondary somatic embryogenesis of Theobroma cacao Cerezo S, Mercado JA, Pliego-Alfaro F (2011) An efficient regeneration L. In Vitro Cell Dev Biol-Plant 38:252–259 system via somatic embryogenesis in olive. Plant Cell Tissus Organ Maximova SN, Florez S, Shen X, Niemenak N, Zhang Y, Curtis W, Cult 106:337–344 Guiltinan MJ (2014) Genome-wide analysis reveals divergent pat- De Feria M, Jiménez E, Barbón R, Capote A, Chávez M, Quiala E (2003) terns of gene expression during zygotic and somatic embryo matu- Effect of dissolved oxygen concentration on differentiation of so- ration of Theobroma cacao L., the chocolate tree. BMC Plant Biol matic embryos of Coffea arabica cv. Catimor 9722. Plant Cell 14:185 Tissue Organ Cult 72:1–6 Maximova SN, Young A, Pishak S, Miller CR, Traore A, Guiltinan MJ Driver JA, Kuniyuki AH (1984) In vitro propagation of Paradox walnut (2005) Integrated system for propagation of Theobroma cacao L. In: rootstock. Hortic Sci 19:507–550 Jain SM, Gupta PK (eds) Protocol for somatic embryogenesis in Ducos JP, Labbe G, Lambot C, Pétiard V (2007) Pilot scale process for woody plants. Springer, Dordrecht, the Netherlands, pp 209–227 the production of pre-germinated somatic embryos of selected ro- McCown BH, Lloyd G (1981) Woody plant medium (WPM)—amineral busta (Coffea canephora) clones. In Vitro Cell Dev Biol-Plant 43: nutrient formulation for microculture of woody plant species. Hortic 652–659 Sci 16:453 Egertsdotter U, Clapham D (2015) Method for maturing and synchroniz- Miller CR (2009) An integrated in vitro and greenhouse orthotropic clon- ing conifer somatic embryos. Patent Number WO 2011123038 al propagation system for Theobroma cacao L. Ph.D. The El Meskaoui A, Desjardins Y, Tremblay FM (2006) The effects of in vitro Pennsylvania State University, USA gaseous environment on the maturation of somatic embryo of black Minyaka E, Niemenak N, Sangare A, Omokolo DN (2008) Effect of spruce (Picea mariana [mill.] BSP) and white spruce (Picea glauca MgSO and K SO on somatic embryo differentiation in 4 2 4 [Moench] Voss). Moroccan. J Biol 2:12–20 Theobroma cacao L. Plant Cell Tissue Organ Cult 94:149–160 El Meskaoui A, Tremblay FM (1999) Effects of sealed and vented gas- Motamayor JC, Lachenaud P, Da Silva E Mota JW, Loor R, Kuhn DN, eous microenvironments on the maturation of somatic embryos of Brown JS (2008) Geographic and genetic population differentiation black spruce with as special emphasis on ethylene. Plant Cell Tissue of the Amazonian chocolate tree (Theobroma cacao L.). PLoS ONE Organ Cult 56:201–209 3(10):e3311. https://doi.org/10.1371/journal.pone.0003311 Etienne H (2005) Somatic embryogenesis protocol: coffee (Coffea Murashige T, Skoog F (1962) A revised medium for rapid growth and arabica L. and C. canephora P.) In: Jain S. Gupta M, Pramod K bioassays with tobacco cultures. Physiol Plant 15:473–497 (eds) Protocol for somatic embryogenesis in woody plants. Springer, Niemenak N, Kaiser E, Maximova SN, Laremore T, Guiltinan MJ (2015) Dordrecht, Netherlands, pp 167–179 Proteome analysis during pod, zygotic and somatic embryo matura- Garcia C, Corrêa F, Findley S, Almeida AA, Costa M, Motamayor JC, tion of Theobroma cacao.JPlant Physiol 180:49–60 Schnell R, Marelli J-P (2016) Optimization of somatic embryogen- Niemenak N, Saare-Surminski K, Rohsius C, Ndoumou DO, Lieberei R esis procedure for commercial clones of Theobroma cacao L. Afr J (2008) Regeneration of somatic embryos in Theobroma cacao L. in Biotechnol 15:1936–1951 temporary immersion bioreactor and analyses of free amino acids in different tissues. Plant Cell Rep 27:667–676 Geng P, La H, Wang H, Stevens EJC (2008) Effect of sorbitol concentra- tion on regeneration of embryogenic calli in upland rice varieties Pullman G, Gupta P, Timmis R, Carpenter C, Kreitinger M, Welty E (Oryza sativa L.). Plant Cell Tissue Organ Cult 92:303–313 (2005) Improved Norway spruce somatic embryo development THEOBROMA CACAO L. INDIRECT SOMATIC EMBRYOGENESIS... through the use of abscisic acid combined with activated carbon. Söndahl MR, Nakamura T, Sharp WR (1985) Propagation of coffee. In: Henke R, Hughes KW, Constantin MJ, Hollaender A, Wilson CM Plant Cell Rep 24:271–279 Quainoo AK, Dwomon IB (2012) The effect of abscisic acid in the conver- (eds) Tissue culture in forestry and agriculture. Basic Life Sciences, sion of cocoa somatic embryos into plantlets. Front Plant Sci 2:6–10 vol 32. Springer, Boston, pp 215–232 Quinga LAP, De Freitas Fraga HP, Do Nascimento Vieira L, Guerra MP Söndahl MR, Sharp WR (1977) High frequency induction of somatic (2017) Epigenetics of long-term somatic embryogenesis in embryos in cultured leaf expiants of Coffea arabica L. Z Theobroma cacao L.: DNA methylation and recovery of embryo- Pflanzenphysiol 81:395–408 genic potential. Plant Cell Tissue Organ Cult 131:295–305 Teyssier C, Grondin C, Bonhomme L, Lomenech AM, Vallance M, Rodríguez López CM, Wetten AC, Wilkinson MJ (2010) Progressive Morabito D, Label P, Lelu-Walter MA (2011) Increased gelling erosion of genetic and epigenetic variation in callus-derived cocoa agent concentration promotes somatic embryo maturation in hybrid (Theobroma cacao) plants. New Phytol 186:856–868 larch (Larix×eurolepsis): a 2-DE proteomic analysis. Physiol Plant Sena Gomes AR, Andrade Sodré G, Guiltinan M, Lockwood R, 141:152–165 Maximova SN (2015) Chap 4: Tissue culture. In: Laliberté B, End Umehara M, Ogita S, Sasamoto H, Koshino H, Asami T, Fujioka S, M (eds) Supplying new cocoa planting material to farmers: a review Yoshida S, Kamada H (2005) Identification of a novel factor, of propagation methodologies. Bioversity International, Rome, Italy, vanillyl benzyl ether, which inhibits somatic embryogenesis of pp 67–87 Japanese larch (Larix leptolepis Gordon). Plant Cell Physiol 46: Shimazu T, Kurata K (2003) Dynamic dissolved oxygen concentration 445–453 control for enhancing the formation rate of torpedo-stage embryos in Van Boxtel J, Berthouly M (1996) High frequency somatic embryogen- carrot somatic embryo culture. J Biosci Bioeng 95:384–390 esis from coffee leaves. Plant Cell Tissue Organ Cult 44:7–17 Singh M, Chaturvedi R (2012) Evaluation of nutrient uptake and physical Zamarripa A, Ducos JP, Bollon H, Dufour M, Pétiard V (1991) parameters on cell biomass growth and production of spilanthol in Production of somatic embryos of coffee in liquid medium: effects suspension cultures of Spilanthes acmella Murr. Bioprocess Biosyst of inoculation density and renewal of the medium. Cafe Cacao The Eng 35:943–951 35:233–244 Smith SM, Street H (1974) The decline of embryogenic potential as callus and suspension cultures of carrot (Daucus carota L.) are serially subcultured. Ann Bot 38:223–241 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png In Vitro Cellular & Developmental Biology - Plant Springer Journals

Indirect somatic embryogenesis of Theobroma cacao L. in liquid medium and improvement of embryo-to-plantlet conversion rate

Free
15 pages
Loading next page...
 
/lp/springer_journal/indirect-somatic-embryogenesis-of-theobroma-cacao-l-in-liquid-medium-wDys1LNgfC
Publisher
Springer US
Copyright
Copyright © 2018 by The Author(s)
Subject
Life Sciences; Plant Sciences; Cell Biology; Developmental Biology; Plant Breeding/Biotechnology; Plant Genetics and Genomics
ISSN
1054-5476
eISSN
1475-2689
D.O.I.
10.1007/s11627-018-9909-y
Publisher site
See Article on Publisher Site

Abstract

The establishment of cocoa embryogenic cell lines in liquid medium starting from high frequency somatic embryogenesis (HFSE) callus is described. The growth kinetics of the cultures during the multiplication and the expression steps conducted in 250 mL Erlenmeyer flasks were described for three genotypes selected for their agronomical traits (EET95, EET96, and EET103). The glucose and dissolved oxygen concentrations and the absorption of Murashige and Skoog medium macronutrients (nitrate, ammonium, potassium, sulfate, calcium, phosphorus, and magnesium) were monitored. The multiplication of the −1 embryogenic calluses in a medium containing 2,4,5-trichlorophenoxyacetic acid (2,4,5-T) at 1 mg L , initiated with an inoc- −1 ulation density of 20 g L of callus, was achieved. The growth rate was characterized by two phases, with the second being concomitant with a depletion of phosphorus and magnesium, and a decrease in the embryogenic potential of the callus. The expression of the callus embryogenic capacity was conducted in an auxin-free medium. The embryo production starting from 1 −1 and5g L inoculation densities was compared. When placed in the optimal expression conditions in flasks, 1 g of callus produced 1000 to 1500 embryos within 5 to 7 wk. Finally, two paths for improving the plantlet regenerative capacities of cocoa SE produced in liquid medium were identified. Supplementing the expression medium with myo-inositol used as an osmotic −1 agent at a concentration of 50 g L increased the embryo-to-plantlet conversion rate from 13–16% to 40–48%. A 6-wk culture of the embryos on a maturation medium in Petri dishes optimized their subsequent development into plantlets. . . . . Keywords Cocoa Propagation High frequency somatic embryogenesis callus Embryo-to-plantlet conversion rate Myo-inositol Introduction Cocoa production from T. cacao trees is vulnerable in most producing countries due to the low income of farmers, dis- Theobroma cacao L. trees are grown in the humid tropics to eases, pests, and adverse environmental conditions. To face produce cocoa beans. Four main countries produce 75% of the these challenges, integrated propagation approaches based on total cocoa beans: Ghana, Indonesia, Ivory Coast, and Nigeria somatic embryogenesis (SE), combined with traditional prop- (Food and Agriculture Organization, http://www.fao.org/ agation techniques, have been described for the rapid dissem- faostat/en/#data/QC). In producing countries, cocoa is ination of selected clones (Maximova et al. 2005;Miller 2009; considered as a cash crop, but this is a crucial crop for the Guillou et al. 2014). Due to SE, disease-free materials can be confectionary business, as it provides the key raw material for produced with an orthotropic growth habit, which the farmers chocolate-based products. Nestlé utilizes about 12% of the are familiar with in some countries. world cocoa production. Somatic embryos can differentiate either directly from so- matic cells, which undergo a minimum of proliferation (direct SE), or indirectly after extensive proliferation (indirect SE). * Caroline Guillou However, the distinction between the two processes is not caroline.guillou@rdto.nestle.com always clear. In coffee (Coffea arabica L.), Söndahl and Sharp (1977) preferred to use the terms low- and high frequen- Nestlé Research & Development Centre Tours–Plant Science cy somatic embryogenesis (LFSE versus HFSE) to distinguish Research Unit, 101 avenue Gustave Eiffel BP49716, 37097 Tours them. Low frequency somatic embryogenesis refers to the CEDEX 2, France GUILLOU ET AL. sporadic production of embryos from phenotypically indistin- 2011), and supplementing the medium with PEG or sugar guishable embryogenic callus. In contrast, HFSE is character- alcohols, such as mannitol or sorbitol in rice (Oryza sativa ized by the appearance of friable and highly embryogenic L.; Geng et al. 2008). Myo-inositol, as an osmotic agent, has calluses containing pro-embryogenic masses (PEMs) been less often used to improve SE quality, and only in rose (Söndahl et al. 1985). The specific nature of coffee HFSE (Rosa hybrida L.; Castillón and Kamo 2002) and Norway callus allows the use of a liquid medium for embryogenic spruce (Picea abies (L.) H. Karst.; Egertsdotter and tissue proliferation (Zamarripa et al. 1991; Van Boxtel and Clapham 2015). Berthouly 1996; Etienne 2005). Commonly considered as in- While several studies mention a low conversion of cocoa direct SE, HFSE cultures rapidly scale up the number of po- SE into plantlets, very few reports are available on the effects tential plants that can be produced, because it is possible to of individual components on SE quality in this species. multiply the PEMs. Consequently, HFSE is generally pre- Quainoo and Dwomon (2012) evaluated the effect of a range ferred to LFSE (direct SE) for boosting mass propagation of ABA concentrations from 0 to 50 μM applied for a period procedures (Etienne 2005). The multiplication of the embryo- of 0 to 6 wk. They did not observe any significant positive genic calluses (also referred as the Bproliferation^ or effects from these ABA treatments on the subsequent plantlet Bmaintenance^ step) is generally conducted in an auxin- regeneration rates. Niemenak et al. (2015) reported that in −1 based medium. To initiate the regeneration of embryos (also medium containing sucrose at 60 g L compared to −1 referred as Bexpression^ or Bhisto-differentiation^ step), the 30 g L , maturated SEs contained higher levels of proteins calluses are transferred to an auxin-free medium, which allows involved in storage and tolerance to desiccation. them to express their embryogenic potential. An HFSE procedure for establishing cocoa embryogenic In the case of cocoa SE, most of the published literature cell lines in liquid medium is described in this study. The is based on the progress made by Maximova et al. (2002), growth kinetics and the nutrient uptake during the multiplica- which applied to a particular case of SE that is classified as tion and expression steps in liquid medium were characterized secondary embryogenesis (also referred to as for two clones selected for their agronomic value. The study Bcontinuous,^ Brecurrent,^ or Brepetitive^) (Niemenak et also focused on how to improve the conversion of cocoa SE al. 2008; Garcia et al. 2016). This method consists of cul- into viable plantlets by assessing the effects of increasing the turing cotyledon fragments, from 1 to 2 cm long primary osmolality of the medium during the expression step in flasks, SE, for 2 wk on an auxin-based medium. After transfer to and the effects of the duration of the culture on a maturation an auxin-free medium, each fragment produces 5 to 25 medium in Petri dishes. After these different treatments, the secondary SE within 1 y. Because the secondary SE embryo-to-plantlet conversion rates on germination medium emerges from a mass of non-embryogenic cells developing were compared. from the explants, this method corresponds stricto sensu to the HFSE. However, it leads to the regeneration of SE without the multiplication of embryogenic cells. To make Materials and Methods the most of cocoa SE for creating higher mass propagation potential, a protocol was implemented based on obtaining Plant material In this study, four genotypes were chosen: the and subculturing HFSE callus on a solid medium contain- clone Scavina 6 (SCA6) from the genetic cluster Contamana, ing the auxin 2,4,5-trichlorophenoxyacetic acid (2,4,5-T) and the clones EET95, EET96, and EET103 from the genetic (Masseret et al. 2009). cluster Nacional (Motamayor et al. 2008). Contrary to the As reported for many other plant species, converting the SCA6, which is a clone often used for studies due to its high cocoa SE into whole plants is the most inefficient step of the embryogenic capacity (Li et al. 1998; Maximova et al. 2002), process, and it depends on the genotype (Quainoo and EET95, EET96, and EET103 were chosen for their Ariba Dwomon 2012;Maximova et al. 2014). A common approach flavor and agronomical traits. for improving the SE quality is to create a partial dehydration. In vitro tissues were developed from immature floral buds Numerous studies on conifers have shown the positive effect collected in August 2013, from trees grown in fields in the of increasing gelling agent concentrations and/or the duration Nestlé experimental farm in Chollo, Ecuador. of culture on a medium containing abscisic acid (ABA) (Klimaszewska and Smith 1997; Teyssier et al. 2011). To Induction of primary SE and HFSE callus Primary SE were decrease their water content, olive (Olea europaea L.) SE collected from staminodes following the protocol of Li et al. are advantageously placed onto paper or cellulose acetate (1998). The explants were placed on a PCG medium gelled −1 membranes (Cerezo et al. 2011). Restricting water uptake with Gelrite® at 3 g L for 2 wk. The PCG medium contained −1 −1 using high sucrose concentrations is efficient in alfalfa 2mg L 2,4-dichlorophenoxyacetic acid (2,4-D), 5 μgL −1 (Medicago sativa L.; Anandarajah and McKersie 1990)and thidiazuron (TDZ), and 50 mg L isopentenyl adenine (IP). avocado (Persea americana Mill.; Márquez-Martín et al. The complete composition of the medium is listed in Table 1. THEOBROMA CACAO L. INDIRECT SOMATIC EMBRYOGENESIS... 1 2 3 Table 1. Media composition [ Driver and Kuniyuki (1984); McCown and Lloyd (1981); Murashige and Skoog (1962)] Induction Multiplication Expression Maturation Germination Code PCG SCG CC21 CC2 G80 ENR8 (1) (2) (3) (3) (3) (3) Macronutrients DKW WPM MS MS MS /2 MS (1) (2) (1) (1) (3) (1) Micronutrients DKW WPM DKW DKW MS DKW −1 Vitamins (mg L ) Myo-inositol 200.0 100.0 100.0 100.0 100.0 100.0 Nicotinic acid 1.0 1.0 1.0 1.0 0.5 1.0 Thiamine 2.0 10.0 2.0 2.0 0.7 2.0 Pyridoxine-HCl 1.0 0.5 −1 Amino acids (mg L ) Glycine 2.0 2.0 2.0 2.0 2.0 2.0 L-Lysine 0.4 0.4 0.4 0.4 L-Leucine 0.4 0.4 0.4 0.4 L-Arginine 0.4 0.4 0.4 0.4 L-Tryptophane 0.2 0.2 0.2 0.2 Glutamine 250.0 −1 Plant growth regulators (mg L ) 2,4,5-T 1.000 2,4-D 2.000 2.000 Naphthaleneacetic acid 0.010 Kinetin 0.250 Thidiazuron 0.005 Isopentenyl adenine 50.000 0.200 Adenine-H SO 0.250 0.025 2 4 Abscisic acid 1.000 Gibberellic acid 0.020 −1 Coconut milk (mL L)50.0 −1 Other (g L ) Activated charcoal 1 Glucose 20 20 30 30 40 30 Gelrite® 3 3 3 3 pH 5.6 5.6 5.6 5.6 5.8 5.8 The explants were placed in a dark room at 25°C in 55 mm initiated by selecting HFSE calluses (Fig. 1b). From 0.05 to Petri dishes. After 2 wk, the explants were transferred to SCG 0.10 g fresh weight (FW) of the HFSE callus was transferred −1 −1 medium, which incorporated 2 mg L 2,4-D and 0.25 mg L into 10 mL of the multiplication medium (CC21) contained in kinetin. The explants were then placed on the auxin-free CC2 25-mL Erlenmeyer flasks (cycle 1). The medium was renewed medium for 12 wk. every 3 wk by doubling the volume of the suspensions and Three to 4 mo after disinfection, primary somatic embryos transferring them into larger flasks as follows: 20 mL in 50- (PSE) started to develop from primary callus (Fig. 1a). mL flasks at 3 wk (cycle 2), 50 mL in 100-mL flasks at 6 wk Embryos at the torpedo-stage were finely cut with a scalpel (cycle 3), and 100 mL in 250-mL flasks at 9 wk (cycle 4). and the fragments placed on a CC21 medium containing the After 12 wk of culture in liquid medium, the biomass was −1 −1 auxin 1 mg L 2,4,5-T, and 0.25 mg L adenine. Every 6 wk, collected by filtration through a 50-μm filter. The calluses the embryo fragment calluses were subcultured onto fresh were selected and transferred into 100 mL of the CC21 medi- −1 medium for a maximum period of 8 mo. During this time, um with an inoculation density of 20 g L (cycle 5). HFSE callus started to appear. Thereafter, the embryogenic cell lines (Fig. 1c)were subcultured every 3 wk into 250-mL flasks by transferring Multiplication of the HFSE callus in liquid medium The estab- 2 g of callus into 100-mL of fresh CC21 medium. The suspen- lishment of embryogenic cell lines in liquid medium was sions were cultured on an orbital shaker (New Brunswick™ GUILLOU ET AL. Figure 1. The different steps for Theobroma cacao L. propagation by high frequency somatic embryogenesis (HFSE). (a) Induction of primary SE (PSE, primary somatic embryos; LFSE, low frequency somatic embryogenesis calluses; NEC, non-embryogenic calluses); (b) induction of HFSE callus (SSE, secondary somatic embryos; LFSE, low frequency somatic embryogenic calluses); (c) multiplication: embryogenic clumps in liquid medium; (d) expression: an embryo population at the end of the step; (e) germination: plantlets in germination boxes; (f)normal plantlets; (g) abnormal plantlets. Innova® 2300, Nijmegen, The Netherland) at 120 rpm placed 100 mL of expression medium (CC2). The suspensions were in a dark room at 25°C. cultured in a dark room at 25°C at 120 rpm. Growth kinetics of cell lines after eight multiplication cy- Growth kinetics was established using calluses developing cles were established by measuring various parameters after 0, from the same embryogenic cell lines as the ones studied 1, 3, 5, and 7 wk of culture (named respectively in the results for the multiplication kinetics. The expression starting and discussion T0W, T1W, T3W, T5W, and T7W). The pa- from two different inoculation densities, 1 and 5 g −1 rameters measured were biomass concentration (FW), glucose FW L , was compared. The cultures were analyzed after concentration, dissolved oxygen concentration (DO ), pH, os- 1, 3, 5, and 7 wk (named in the results and discussion molality, conductivity, and MS macronutrient concentrations T1W, T3W, T5W, and T7W) for biomass, glucose, DO , (nitrate, potassium, ammonium, sulfate, calcium, phosphorus, and MS ion concentrations. and magnesium). The embryogenic potential of the calluses, To study the effects of medium components on tissue os- collected at different time periods of a multiplication cycle, molality, two experiments were done. In the first experiment, was measured by counting the number of embryos produced the effects on clone EET103 in the expression step of 15, 30, −1 per gram of callus 4 wk after their transfer into the CC2 ex- and 45 g L myo-inositol (corresponding to a tissue osmolal- −1 pression medium. ity level of 337, 441, and 542 mOsm kg H O , respectively) −1 were assessed in CC2 medium containing glucose at 30 g L Expression of the callus embryogenic potential in liquid me- (corresponding to a tissue osmolality level 261 mOsm kg −1 dium To start the regeneration of the embryos, embryogenic H O ). The second experiment compared the expression of calluses were transferred into 250-mL flasks containing clones SCA6 and EET103 in medium supplemented with THEOBROMA CACAO L. INDIRECT SOMATIC EMBRYOGENESIS... −1 myo-inositol at 50 g L ,after 4wk of culture, andin ame- Osmotic potential: The osmolality was measured in −1 dium containing sucrose at 30 or 80 g L (tissue osmolalities milliosmoles per kilogram of H Ousing amicro-osmometer −1 differed from 193 to 556 mOsm kg H O ). (Model 6/6M, Löser®, Berlin, Germany). The embryo osmo- lality was determined as following the method of Teyssier et Embryo-to-plantlet conversion on solid medium After collec- al. (2011): Around 10 embryos were stored at − 20°C and tion from the flasks, the embryos at the torpedo-stage were thawed at 50°C for 10 min. After centrifugation at 8000×g individually transferred into 90 × 20-mm Petri dishes contain- for 15 min, the osmolality was measured in 100 μLof the −1 ing G80 maturation medium (Fig. 1d) containing 1 mg L supernatants. −1 ABAand1gL charcoal (Lopez-Baez et al. 1993). Each Macronutrients: Residual concentrations of MS ions in the Petri dish was inoculated with 30–40 embryos. The embryos spent medium were measured by conventional methods were incubated in a dark room at 25°C. (Laboratoire de Touraine, Tours, France). Nitrate and ammo- After 0, 3, 6, and 9 wk of maturation (named in the results nium levels were determined by flow analysis and spectromet- and discussion T0W, T3W, T6W, and T9W), the embryos ric detection according respectively NF-EN ISO 13395:1996 were individually subcultured onto ENR8 germination medi- and NF-EN ISO 11732 (13395:2005). Sulfate level was deter- um in 145 × 100 × 60-mm styrofoam boxes (Dominique mined by ionic liquid chromatography according NF-EN ISO Dutscher S.a, ref. 017002, Issy les moulineaux, France). The 10304-1:2007/2010, and the levels of cations potassium, cal- embryos were germinated with a dark/light cycle of 12/12 h at cium, phosphorus, and magnesium were determined by induc- −2 −1 a light intensity of 65 μmol m s with two LED Tubes tively coupled plasma optical emission spectrometry accord- Green Power Deep Red/White/Medium blue, 18 W, ing NF-EN ISO 11885:2007. All the protocols are available −1 24 μmol S , (Philips Lighting, Zurich, Switzerland) in a cul- on the website https://www.boutique.afnor.org/. ture room at 25°C. Water content (WC): After collection from the flasks, em- The embryo-to-normal plantlet conversion rates (%) were bryos were weighed to determine their FW before being dried measured after 8 wk by counting the embryos that had devel- in an oven (UN55, Memmert®, Schwabach, Germany) at oped into normal plantlets (Fig. 1e). The normal plantlets had 90°C for 24 h. Just after being taken out of the oven, the a rooting apex, and, at least, a 2-cm stem height, and one pair samples were weighed again to determine their dry weight of true leaves with a minimum length of 1 cm (Fig. 1f). These (DW). The WC was calculated using the following equation: fully developed plantlets with standard characteristics were WC (%) = (DW / FW) × 100. able to be successfully acclimated to the greenhouse. In some experiments, the number of embryos was counted, which had Data analysis Statistical analyses, including ANOVA using developed abnormal plantlets characterized by thin shoots, the general linear model (GLM), and tests to compare treat- and dark, thickened, and stunted leaves (Fig. 1g). These types ment averages, were performed using Minitab® 17.3.1 of plantlets are not able to survive to the transfer into the (Minitab® Inc., State College, PA). The relevance and ranking greenhouse. factors were calculated by comparing the averages using the All the chemicals are from Duchefa, Haarlem, The Netherlands Tukey test at the 5% threshold. and Petri dishes from VWR, Fontenay-sous-Bois, France. The media pH was adjusted with either 0.1 or 1 N hydrochloric acid and potassium hydroxide. Media were autoclaved at 121°C for Results and Discussion 20 min and glass vessel at 121°C for 30 min in a Lequeux auto- clave, Paris, France. Induction of HFSE callus Young secondary SE (SSE) became observable after 1 mo of culture on the CC21 solid medium −1 Analytical measurements Glucose: Glucose level in the media containing 1 mg L 2,4,5-T (Fig. 1b). Depending on the was measured using a high-performance liquid chromato- genotype, small embryogenic calluses (HFSE) appeared graph (Dionex, IC5000, Thermo Fisher Scientific®, within 6 to 10 mo (data not shown). They were character- Waltham, MA). The separation was achieved using a 4 × ized by a granular and friable appearance and a yellowish 150-mm ion exchange column (Dionex CarboPac, PA10, color (Fig. 1b). Secondary SE and HFSE calluses did not Thermo Fisher Scientific®). develop directly from the fragments of primary SE, but Dissolved oxygen concentration: Oxygen was non- from brown to dark brown callus. invasively measured by using PSt3 sensor spots placed on The differences observed between the occurrence of cocoa the bottom of the flasks along with an optic meter (Firebox secondary SE and embryogenic callus are very similar to the 3, PreSens®, Regensburg, Germany). differences reported in Arabica coffee, between the LFSE and Conductivity and pH: Both the conductivity and pH values HFSE processes (Söndahl et al. 1985;Etienne 2005). In this were measured directly in the cell culture broth using a digital species, SE takes place in two waves in leaf explant cultures. apparatus (Model 3540, Jenway®, Chelmsford, UK). The first wave (LFSE) corresponds to the sporadic and GUILLOU ET AL. transient production of isolated SE within 2 to 3 mo, and the cell lines have to be initiated in small volumes of medium, second wave (HFSE) corresponds to the production of very generally by transferring 0.05 to 0.10 g FW of selected callus embryogenic callus within 5 to 10 mo. into 10 mL of medium. Next, the volume of the medium was Importantly, the occurrence of cocoa HFSE callus was doubled every 3 wk until 12 wk. At this time (cycle 5), the found only if fragments were isolated from the 2–3-mm tor- biomass consisted of a mixture of secondary SE and calluses pedo-stage, rather than from 1-cm cotyledonary-stage embry- of various types. To initiate the maintenance of the embryo- os, which other authors used as starting material to obtain genic cell lines, the friable and yellowish calluses were select- secondary SE (Maximova et al. 2002;Garcia et al. 2016). ed once again if necessary and transferred into 250-mL flasks −1 In the frame of a core collection establishment, it was ob- with an inoculation density of 20 g FW L to limit the differ- served that almost all of the genotypes that are reactive to entiation of embryos and to maintain HFSE calluses. The cell primary SE also respond positively to secondary SE, which lines were composed of large aggregates, ranging from 0.3 to confirms the results of Maximova et al. (2005), Masseret et al. 1.5 mm (Fig. 1c). (2009), and Guillou et al. (2014). However, there were some Growth kinetics. Typical profiles of the growth and of sub- genotypes that produced primary and secondary SE very well, strate consumption of EET96 and EET103 embryogenic cell but for which obtaining and selecting HFSE calluses remained lines are shown in Figs. 2 and 3. The biomass concentrations −1 a long and laborious process. It is estimated that 50 to 60% of ranged from 20 up to 60–80 g FW L within 5 wk. During the genotypes gave rise to HFSE callus using this protocol this period, the growth was strictly linear and followed the (data not shown). equations y = 11.14x + 20 with an R value of 0.99 for EET96, and y = 8.39x + 20 with an R value of 0.98 for Multiplication of the HFSE callus in liquid medium EET103. Afterwards, the growth rates increased strongly be- Establishment of embryogenic cell lines. The HFSE calluses tween T5Wand T7W, as the biomass concentration was three- must be transferred into a liquid medium with a sufficiently fold higher within only 2 wk. It reached a final value of 224 −1 −1 high inoculation density of 5 to 10 g FW L as a minimum. and166gFWL at T7W for the clones EET96 and EET103, Indeed, with lower inoculum, the suspensions frequently respectively (Figs. 2a, 3a). By T7W, the glucose was tended to differentiate totally into embryos despite the pres- completely consumed from the medium (Figs. 2b, 3b). ence of the 2,4,5-T in the CC21 medium. For this reason, and Because oxygen can be a limiting factor in liquid medium, because of the limited quantity of available HFSE calluses, the the cultures were checked for hypoxia. The DO did not drop −1 Figure 2. Multiplication step: growth kinetics and nutrient uptake of concentrations (mg L ). For a, b,and c, the data represent the averages Theobroma cacao L. cv. EET96 embryogenic cell line. (a)Fresh of four replicates, each one being a 250-mL flask containing 100 mL of −1 −1 −1 weight (g L ); (b) glucose concentration (g L ) and DO (%); (c) CC21 medium initiated with an inoculation density of 20 g L . Bars tissue osmolality, conductivity, and pH; (d) total macronutrients, nitrate, represent standard errors unless too small to visualize. For d, e,and f, −1 and potassium concentrations (mg L ); (e) ammonium, sulfate, and measurements were taken after combining the spent medium of four −1 calcium concentrations (mg L ); (f) phosphorus and magnesium flasks. THEOBROMA CACAO L. INDIRECT SOMATIC EMBRYOGENESIS... −1 Figure 3. Multiplication step: growth kinetics and nutrient uptake of concentrations (mg L ). For a, b,and c, the data represent the averages Theobroma cacao L. cv. EET103 embryogenic cell line. (a)Fresh of four replicates, each one being a 250-mL flask containing 100 mL of −1 −1 −1 weight (g L ); (b) glucose concentration (g L ) and DO (%); (c) CC21 medium initiated with an inoculation density of 20 g L . Bars tissue osmolality, conductivity, and pH; (d) total macronutrients, nitrate, represent standard errors unless too small to visualize. For d, e,and f, −1 and potassium concentrations (mg L ); (e) ammonium, sulfate, and measurements were taken after combining the spent medium of four −1 calcium concentrations (mg L ); (f) phosphorus and magnesium flasks. below 60% throughout the cultures. In another experiment without auxin (Fig. 4). A strongly significant effect (P < (data not shown), after shaking the cultures was stopped, the 0.001) of the clones (F = 12.8), and mostly of the duration DO critical value under which the consumption rate in O (F = 39.5) on the embryogenic potential was detected. It 2 2 decreases was determined and indicated hypoxia regime. For reached its highest value after 3 wk of culture, and then sig- the cocoa calluses cultured in suspensions, this value lies be- nificantly decreased at 5 wk for cultivar EET103. tween 35 and 45%. Consequently, it can be assumed that the The cocoa embryogenic cell lines were characterized by DO was not a limiting factor during the culture period shown biphasic growth, a linear growth up to T5W, followed by faster in the Figs. 2 and 3. The pH started at 4.6 in T0W before reaching values ranging from 6.0 to 7.0 (Figs. 2c, 3c). The tissue osmolality decreased from an initial value of 269, to −1 values close to 50 mOsm kgH O at the end of the culture period, with the medium conductivity dropping from 5.2 to −1 1.9–2.4 mS cm . −1 The sum of the MS macronutrients is equal to 4.0 g L in −1 fresh CC21 medium (T0W). It dropped to 0.8–1.0 g L after 7 wk of culture (Figs. 2d, 3d). The ammonium, potassium, and calcium ions were consumed at 40 to 60% of their initial concentrations at TW7. The nitrate and sulfate were metabo- lized at a higher rate, but 5 to 10% of their initial concentra- tions remained at the same date. Finally, among the sampled macronutrients, only magnesium and phosphorus were completely depleted, which occurred as early as T3W for both Figure 4. Multiplication step: evolution of the embryogenic potential of clones (Figs. 2f, 3f). Theobroma cacao L. embryogenic cell lines EET96 and EET103. After Evolution of the embryogenic potential of the calluses. The various durations of culture in CC21 medium, 0.1 g of callus tissue was transferred into 100 mL CC2 medium. The embryos were counted by the evolution of the tissue qualities within a multiplication cycle naked eye after 4 wk. The data represent the averages of 16 replicates in CC21 medium was monitored by measuring the embryo- −1 (four flasks × four measures flask ). Data followed by different letters genic potential of the calluses (number of embryos that could are significantly different at P <5% (Tukey’s test completed separately −1 for each clone). be produce g FW ) once transferred to expression conditions GUILLOU ET AL. −1 Figure 5. Expression step: growth kinetics of Theobroma cacao L. cv. inoculated with an inoculation density of 1 or 5 g L . Bars represent EET96 embryo regeneration. The data are the averages of four replicates, standard errors unless too small to visualize. For d, e,and f, measurements each one being a 250-mL flask containing 100 mL of CC2 medium were taken after combining the spent medium of four flasks. growth up to T7W. The second phase was concomitant with a Chaturvedi 2012). It should be noted that the growth cycle decrease in the embryogenic capacity of the calluses, which was monitored in terms of fresh biomass. The tissues could may be due to phosphorus and magnesium deficiencies. have a higher WC at the end of the culture period due to the −1 Other authors have described such deficiencies using the MS low osmolality of the medium (50 mOsm kgH O ). It is macronutrients during plant cell cultures in liquid medium possible that the decrease in the embryogenic potential of the (Archambault et al. 1994; Azevedo et al. 2008; Singh and calluses at this time could also be attributed to a higher WC. Figure 6. Expression step: growth kinetics of Theobroma cacao L. cv. represent standard errors unless too small to visualize. For d, e,and f, EET103 embryo regeneration. The data are the averages of four measurements were taken after combining the spent medium of four replicates, each one being a 250-mL flask containing 100 mL of CC2 flasks. −1 medium inoculated with an inoculation density of 1 or 5 g L . Bars THEOBROMA CACAO L. INDIRECT SOMATIC EMBRYOGENESIS... extent on the genetic background (P = 0.021, F = 5.8). The ability of the embryo to develop plantlets, once transferred to maturation followed by a germination medium, tended to increase between T3W and T7W but not significantly (Fig. 7). On the contrary to the embryogenic potential, there was no influence of the inoculation density on the SE quality (P = 0.36). It is not known whether the inhibition of the embryo −1 production in the cultures initiated at 5 g FW L resulted from deficiencies in nutrients due to higher consumption rate of nutrients or to decreased tissue osmolality and me- −1 dium conductivity in the cultures initiated at 5 g FW L (Figs. 5c, 6c). Whatever clone was tested, or the inocula- − + + tion densities used, the DO and the NO ,K ,NH ,and 2 3 4 Ca ions were clearly not limiting factors throughout the Figure 7. Expression step: evolution of the quality of the Theobroma kinetics (Figs. 5d, e, 6d, e). For clone EET96, inhibition was cacao L. cvs. EET96 and EET103 embryos. For the embryo-to-plantlet conversion rates, at least 20 embryos per replicate were transferred onto observed at T5W. At this date, the magnesium and phosphorus G80 medium and after 3 wk on ENR8 medium. The data (averages of were the nutrients with the lowest concentrations, but their four replicates) followed by different letters are significantly different at −1 levels remained higher than 1.8 mg L in both cultures. For P < 5% (Tukey test completed for each clone separately). clone EET103, inhibition was quantifiable at T7W, and the phosphorus was totally exhausted at T5W in the cultures ini- −1 Considering these results, a frequency of 3 wk to subcul- tiated at 5 g L . For these cultures, very low concentrations in −1 +2 −1 −2 ture, the cell lines was selected for the standard procedure. glucose(0.7gL ), in Mg (0.9 mg L ), and in SO −1 Under this 3-wk regime, the multiplication rate of the biomass (7 mg L ) at T7W were noted. ranged from 2 to 3 at each subculture. However, as reported Effect of medium renewals. For EET95 cell line, the ex- −1 early in the history of plant SE (Smith and Street 1974), a pression step initiation wascompared with 1 and5gFWL , decline of the capacity of the cocoa calluses to regenerate with and without renewing the medium twice a week. Once embryos was observed with the number of subculture cycles again, a strong inhibition of the expression was linked to the (data not shown). Therefore, this number was restricted to higher inoculation density (Fig. 8a). After 33 d, less than 50% 10 cycles, a total of 30 wk of culture, in a liquid multiplication of the biomass was composed of embryos in cultures initiated −1 medium. In T. cacao, the long-term decrease in the embryo- with 5 g L , instead of complete differentiation in those ini- −1 genic potential of secondary SE has been associated with an tiated at 1 g L (Fig. 8b). As a consequence of the medium increase of global DNA methylation levels (Rodrίguez Lόpez renewals, the inhibition was totally suppressed as the biomass −1 et al. 2010;Adu-gyamfi et al. 2016;Quinga et al. 2017). increased to 150 g L , with 95% of the biomass consisting of −1 embryos. The inhibition of the differentiation at 5 g FW L Expression of the callus embryogenic capacity in liquid medi- did not result from a limitation in oxygen because the DO um Growth kinetics. Using calluses produced from the cell level stayed above 70% (Fig. 8c). When the medium was lines described above, the growth kinetics was compared renewed, the DO dropped to a lower level than without re- during the expression step starting with two inoculation newal, as early as the 14th day (Fig. 8c)(P =0.011, t test), −1 densities, 1 and 5 g FW L for the clones EET96 and which indicated a higher demand for oxygen in the suspen- EET103 (Figs. 5 and 6, respectively). The highest concen- sions. This fact suggests that the growth and the differentiation trations in embryos were reached after 5 wk for the cultures process were hampered before the 21st day, when the expres- −1 −1 initiated at 5 g FW L and ranged between 2500 and 3500 sion was started at 5 g FW L . If this observation does not −1 embryos L (Figs. 5a, 6a). permit exclusion of nutrient deficiencies as the cause of inhi- If the embryo concentrations were proportional to the size bition, it does provide an argument in favor of the presence of of the inoculum up to T3W, this was no longer the case after inhibitory compounds released by the tissues. this date. In fact, the highest values of embryogenic potential, Based on these results, the standard procedure selected −1 expressed as the number of embryos per inoculated gram, consisted of starting the expression with 1 to 2 g FW L −1 were obtained in the cultures inoculated at 1 g FW L at inoculum and renewing the medium after 3 to 4 wk of culture. T5W for clone EET96 and T7W for clone EET103, and The embryos were collected after three to four additional −1 ranged between 1000 and 1500 embryo g . The embryogenic weeksofgrowthinliquidmediumbeforedirecttransfer onto potential highly depends on the duration (P <0.001,F =40.9), the maturation medium, or after a 3 to 5 wk of development in on the inoculation density (P < 0.001, F = 18.2), and to a lesser a temporary immersion bioreactor (Guillou et al. 2014). GUILLOU ET AL. Figure 8. Expression step: effect of the inoculation density and of a twice-weekly medium renewal of Theobroma cacao L. cv. EET95. (a) Embryo suspensions after 5 wk. From left to right: inoculation density of 1 and −1 5gL with medium renewal. (b) Total (whole biomass) and embryo FW at 5 wk. The data (averages of three replicates) followed by different letters are significantly different at P <5% (Tukey test). (c) Evolution of the DO .DO was measured just 2 2 before the medium was renewed. The data are the averages of three replicates. It was observed that inhibition of the cocoa SE expression depletion on the inhibition of cocoa SE production. Generally, was linked to high initial densities, similar to results reported to the naked eye, embryos are smaller in the cultures initiated −1 in other plant species like coffee (Zamarripa et al. 1991), car- at 5 g FW L as early as the third week at which time the rot (Daucus carota L.; Kobayashi et al. 2000), or larch (Larix concentrations of phosphorus and the magnesium were clearly leptolepis Gordon; Umehara et al. 2005). Among the elements not yet limited. This point was confirmed by the experiment monitored, deficiencies in phosphorus and Mg are most that renewed the medium twice weekly. The demand in oxy- likely to occur during the expression step, similarly to the gen in the high-density cultures was hampered before the end multiplication step. However, these ions were consumed more of the 3rd week. Taken together, these observations suggest slowly than during the multiplication phase, because they that this inhibition was most likely associated with the diffu- were only completely consumed at T5W or T7W, instead of sion of SE self-inhibitors from the tissues into the medium, at T3W. Interestingly, Minyaka et al. (2008) reported that similar to those identified in other species, such as 4- supplementing a medium with MgSO improved the induc- hydroxybenzyl alcohol in carrot (Kobayashi et al. 2000), or tion and the development of secondary SE on solid medium. vanillyl benzyl ether in larch (Umehara et al. 2005). Optimal cocoa SE probably requires these elements in higher Effect of an increase in the tissue osmolality. Myo-inositol concentrations than the elements provided by the MS basal is known to be involved in seed development, seed desicca- medium commonly used in plant cell cultures. Nevertheless, tion, signaling pathways, cell wall biosynthesis, and various results are insufficient to determine a definitive role of nutrient stress-related responses (Loewus and Murthy 2000). For this THEOBROMA CACAO L. INDIRECT SOMATIC EMBRYOGENESIS... Table 2. Effect of myo-inositol supplementation during the expression plantlet conversion rates were measured after 3wkonmaturation step on the embryo quality of Theobroma cacao L. cv. EET103. The medium and 8 wk on germination medium (data are the averages of −1 expression step was initiated with an inoculation density of 5 g L in five replicates of at least 20 embryos). (± are the standard errors, data 250-mL flasks containing 100 mL of CC2 medium. Fresh weight (FW), followed by different letters are significantly different according to the embryo number, and water content (WC) were measured after 30 d of Tukey test (P <5%)) expression (data are the averages of five replicates). The embryo-to- Glucose Myo-inositol Molarity Osmolality fresh medium FW Embryo number WC Embryo-to-plantlet conversion rate −1 −1 −1 −1 −1 gL gL MmOsmkgH O gL Number L % % Normal 30 0 0.17 261 65 ± 27 a 1114 ± 131 a 90.2 ± 2.7 a 16.4 ± 2.1 a 30 15 0.25 337 71 ± 13 a 876 ± 191 ab 88.9 ± 0.5 ab 25.4 ± 1.9 ab 30 30 0.33 441 45 ± 21 ab 584 ± 242 bc 82.7 ± 3.5 bc 27.7 ± 2.3 b 30 45 0.42 542 32 ± 7 b 272 ± 118 c 82.5 ± 1.1 c 33.0 ± 3.2 b reason, myo-inositol was tested as an osmotic agent rather 48.3% instead of 13.3% for SCA6, and 40.8% instead of than other sugar alcohols, such as mannitol or sorbitol, in 15.8% for EET103. Concomitantly, the rates of conversion order to produce higher frequencies of cocoa SE conversion into abnormal plantlets were reduced. The incidence of this into plantlets. supplementation in medium in which the glucose was re- In the first experiment, the growth and the embryo forma- placed by sucrose was also checked. In this case, this sucrose tion were partially inhibited with myo-inositol at 30 and treatment led to a significant increase of the embryo quality for −1 45 g L (Table 2). After 30 d, the embryo WC decreased SCA6, but not for EET103. When the sucrose concentration −1 from 90% in the control to 82%, with the highest myo- was increased from 30 to 80 g L , the embryo quality tended inositol concentration. When the tissue osmolality was in- to improve though not significantly for both clones, despite creased, the morphology of the embryos changed. They the fact that this treatment reduced their WC. showed a higher degree of opacity, and more elongated hypo- As observed by Niemenak et al. (2015), a treatment with a cotyls with anthocyanin pigmentation (Fig. 9). After culture high sucrose concentration modifies the appearance of the on a maturation medium followed by a germination medium, embryos, which indicates a higher degree of maturation. the aptitude of the embryos to develop normal plantlets was However, myo-inositol was more efficient than sucrose for improved proportionally to the myo-inositol concentration, increasing the SE quality. Increasing the tissue osmolality from 16 to 33% with myo-inositol at a concentration of generally inhibited the formation of new embryos from −1 45 g L (Table 2). micro-calluses attached to the embryos. Consequently, the Therefore, in the second experiment, the medium was sup- embryo populations appeared more uniform at the end of the −1 plemented with myo-inositol at 50 g L once the embryos expression step. The comparison of the effect of myo-inositol reached the early torpedo-stages (after 4 wk of culture). After with other osmotic agents (mannitol, sorbitol, and maltose), an additional 4 wk, the increase in the fresh-medium myo- with identical molarity, and the continuation of these treat- −1 inositol concentration from 10 to 45 g L led to an increase ments beyond the expression stage such as during maturation −1 in the embryo osmolality, from 252 to 600 mOsm kgH O on solid medium, is currently underway. for SCA6, and from 304 to 502 for EET103 (Table 3). This Embryo maturation on solid medium. The embryo-to- coincided with slight but statistically significant tissue WC plantlet conversion rates were compared for clones SCA6, decreases of 6%. This treatment was followed by significantly EET96, and EET103 after 0, 3, 6, and 9 wk of maturation higher conversion rates into normal plantlets, with a rate of on G80 medium (Table 1). When the embryos were directly −1 Figure 9. Expression step: effect of supplementing CC2 medium with without myo-inositol; (b) embryos in CC2 medium with 15 g L of −1 different myo-inositol concentrations on Theobroma cacao L. cv. myo-inositol; (c) embryos in CC2 medium with 30 g L of myo- −1 EET103 embryos after 4 wk. (a) Embryos in control CC20 medium inositol; (d) embryos in CC2 medium with 45 g L of myo-inositol. GUILLOU ET AL. Table 3. Effect of an increase of the osmolality during the last week of content (WC) and osmolality were measured at the end of the the expression step on the quality of the embryos cvs. SCA6 and EET103. expression (data are the averages of four replicates of 30 embryos for −1 After 4 wk of expression on CC2 medium containing glucose at 30 g L , WC, of four replicates of 10 embryos for the osmolality). For the −1 embryos were cultured (50 embryos 100 mL )fouradditional wk conversion rate, data are the averages of four replicates of 30 embryos. in CC2 supplemented or not with myo-inositol or sucrose at different (± are the standard errors, data followed by different letters are concentrations. The embryos were subcultured on maturation medium significantly different according to the Tukey test (P <5%)) during 3 wk then 8 wk on germination medium. The embryo water Clone Glucose Sucrose Myo-inositol Molarity Osmolality fresh Osmolality WC Embryo-to-plantlet conversion rate medium embryos −1 −1 −1 −1 −1 gL gL gL MmOsmkgH O mOsm kg H O % % Normal % Abnormal 2 2 SCA6 30 0.17 274 252 ± 15 d 92.4 ± 0.4 a 13.3 ± 7.7 cd 67.5 ± 12.9 a 30 50 0.44 556 600 ± 29 a 86.3 ± 0.5 c 48.3 ± 7.9 ab 45.8 ± 9.2 ab 30 0.09 193 314 ± 15 c 91.7 ± 0.2 b 9.2 ± 9.9 b 63.3 ± 7.2 a 30 50 0.37 468 600 ± 17 a 86.0 ± 0.6 d 61.7 ± 13.5 a 32.5 ± 10.3 b 80 0.23 373 557 ± 7 b 87.4 ± 0.2 c 33.3 ± 8.2 bc 55.5 ± 10.7 a EET103 30 0.17 274 304 ± 5 c 92.6 ± 0.1 a 15.8 ± 5.7 b 71.6 ± 5.8 a 30 50 0.44 556 502 ± 15 ab 86.7 ± 0.4 c 40.8 ± 6.3 a 50.8 ± 5.7 b 30 0.09 193 252 ± 65 d 92.1 ± 0.3 b 15.8 ± 3.2 b 57.5 ± 11.1 ab 30 50 0.37 468 533 ± 26 a 85.5 ± 0.6 d 26.7 ± 6.1 ab 61.6 ± 11.7 ab 80 0.23 373 472 ± 13 b 86.7 ± 0.8 c 25.0 ± 16.4 ab 62.5 ± 7.9 ab transferred from the CC2 expression medium onto the ENR8 51, and 47%, respectively) were observed after 6 wk of mat- germination medium (T0W), their conversion into normal uration. On average for the three clones, the frequency of plantlets was practically nil (Fig. 10). For the three clones regeneration into normal plantlets increased from 13 to 47% SCA6, EET96, and EET103, the best conversion rates (43, when the maturation duration was extended from 3 to 6 wk. Meanwhile, the regeneration into abnormal plantlets dropped from 31 to 6%. After 9 wk of maturation, the embryo quality decreased. During this kinetic period on maturation medium, neither the osmotic potential of the embryos nor their WC changed (data not shown). The WC never decreased below 90%, particularly between T3W and T6W (Fig. 11 shows the evolution of the appearance of the embryos). The higher qual- ityafter 6wk of maturation could notbe correlated with a greater opacity, or any other signs of maturation such as a partially dehydrated appearance. However, the embryos started to present signs of oxidation after 9 wk, which could be responsible for a lower quality at this period. The positive effect of the maturation step could come from certain components of the G80 medium, and most probably from active charcoal and/or from ABA, which are both pres- −1 −1 ent at 1 g L and1mg L , respectively. Active charcoal is known to trap toxic compounds such as residual or endoge- nous auxins, which affects the SE quality in soybean (Buchheim et al. 1989). The combination of charcoal and −1 ABA at 5 to 25 mg L improved the production of well- shaped SE in coniferous species, despite the fact that ABA was trapped by charcoal (Pullman et al. 2005). Free ABA was indeed undetectable in the G80 medium in the present exper- Figure 10. Effect of the duration of culture on G80 maturation medium iment (data not shown). Finally, the hypothesis that the posi- on the conversion into normal or abnormal plantlets Theobroma cacao L. tive impact of the maturation on the SE quality does not result cvs. SCA6, EET96, and EET103. Selected embryos were subcultured on G80 maturation medium in 90 × 20-mm Petri dishes (40 embryos per from the G80 medium itself, but from the microenvironment dishes). For their conversion into plantlets, they were transplanted onto provided by Petri dishes cannot be excluded. This hypothesis ENR8 germination medium. The conversion rates were measured after is currently being investigated. In spruce, a confined environ- 8 wk. Data (averages of 10 replicates of 40 embryos) followed by ment induced a larger number of well-formed mature embryos different letters are significantly different according to Tukey test completed separately for each clone (P <5%). and a lower rate of precocious germination than a ventilated THEOBROMA CACAO L. INDIRECT SOMATIC EMBRYOGENESIS... Figure 11. Evolution of the embryos during the maturation step of Theobroma cacao L. cv. EET96. T0W, embryos after the harvest without maturation phase; T3W, embryos after 3 wk on maturation medium; T6W, embryos after 6 wk on maturation medium; T9W, embryos after 9wk on maturation medium. environment, and the addition of 5% CO or 10% O in a treatments increased the rate of conversion into normal plantlets 2 2 vented environment promoted SE (El Meskaoui and while they decreased the conversion into abnormal plantlets, Tremblay 1999;El Meskaoui et al. 2006). which indicated that the morphological issue was not the con- sequence of somaclonal variation, but most likely due to phys- iological disorders resulting from incomplete maturation. Conclusions One of the important strategies for future improvements of the quality of the embryogenic cocoa tissues, including the embryogenic potential of the callus at the end of multiplica- This is the first detailed report describing the establishment of tion, and the capacity of the embryos to regenerate plants at cocoa embryogenic cell lines and their growth kinetics in liq- the end of the expression, consists of implementing these steps uid medium. This protocol is based on obtaining and rigor- in a bioreactor system that continuously monitors and controls ously selecting HFSE callus. The main points necessary to the environmental conditions. There are some studies that in- succeed in establishing the cell lines are as follows: (1) at dicated SE optimization in a stirred bioreactor by controlling the beginning of the process, the primary embryos used as the DO (Archambault et al. 1994;DeFeria et al. 2003; explants must be taken at the torpedo-stage to obtain HFSE callus, and (2) particular attention must be focused on the Shimazu and Kurata 2003). In the framework of the Nestlé Cocoa Plan, the method inoculation densities of the suspensions, as in the case of described in this study was used to regenerate 260,000 plants, Robusta coffee (Ducos et al. 2007), because high densities which, after their acclimation, have been sent to producing hamper the precocious embryo differentiation, which tends countries, mainly Ecuador and Ivory Coast (Guillou et al. to occur despite the presence of the 2,4,5-T in the medium. 2014). The Indonesian Coffee and Cocoa Research Institute Once placed in the optimal expression conditions in flasks, (ICCRI) scaled up its production to deliver 74 M cocoa so- 1 g of callus produced 1000 to 1500 embryos. Compared to matic seedlings during the period from 2009 to 2011 and certain other species, the resulting embryogenic potential is rel- 5 −1 modified certain steps for particular genotypes (Sena Gomes atively low. Productivity levels of over 5 × 10 embryos g of et al. 2015). To address the issues of cocoa production by callus have been frequently reported in carrot, up to 2 × 10 −1 diffusing more productive varieties with resistance to biotic embryos g of callus in Robusta coffee (Zamarripa et al. 1991), and 0.4 to 0.8 × 10 in Arabica coffee (Van Boxtel and and abiotic stresses, the application of HFSE will be of interest for future breeding programs, and for mass propagation of Berthouly 1996;DeFeria et al. 2003, Etienne 2005). The fre- validated planting materials. quent occurrence of abnormal cocoa plantlets with shoots con- tinuously producing cotyledon-like leaves and long internodes Acknowledgements We would like to thank the Master 2 students who was noted, which is a phenomenon that has been reported by contributed enthusiastically to this research: Quentin Claverie (University other authors (Maximova et al. 2005; Quainoo and Dwomon of Bordeaux) for the studies on the growth kinetics, Gaetan Augier 2012). Two paths were identified for improving the quality of (University of Angers) for myo-inositol experiments, Mathilde Piquet (University of Rennes) for the study on the duration of maturation. We cocoa SE regenerated in liquid medium. One was to increase the would also like to thank Dr. James Mc Carthy for the critical review of the osmolality during the expression step and the second was to manuscript. extend the maturation step in Petri dishes. Interestingly, both GUILLOU ET AL. Open Access This article is distributed under the terms of the Creative Guillou C, Fillodeau A, Brulard E, Verdier D, Simon M, Landmann A, Commons Attribution 4.0 International License (http:// Lausanne F, Fontanel A, Ducos JP, Buchwalder A, Broun P (2014) creativecommons.org/licenses/by/4.0/), which permits unrestricted use, Nestlé Cocoa Plan: cocoa propagation by somatic embryogenesis. distribution, and reproduction in any medium, provided you give appro- In: Park YS, Bonga JM (eds) Proceedings of the 3rd International priate credit to the original author(s) and the source, provide a link to the Conference of the IUFRO Unit 2.09.02 on BWoody plant production Creative Commons license, and indicate if changes were made. integrating genetic and vegetative propagation technologies^,8–12 September, vol 2014. Vitoria-Gasteiz, Spain, pp 75–80 Klimaszewska K, Smith DR (1997) Maturation of somatic embryos of Pinus strobus is promoted by a high concentration of gellan gum. Physiol Plant 100:949–957 References Kobayashi T, Higashi K, Sasaki K, Asami T, Yoshida S, Kamada H (2000) Purification from conditioned medium and chemical identi- Adu-Gyamfi R, Wetten A, Lopez CMR (2016) Effect of cryopreservation fication of a factor that inhibits somatic embryogenesis in carrot. and post-cryopreservation somatic embryogenesis on the epigenetic Plant Cell Physiol 41:268–273 fidelity of cocoa (Theobroma cacao L.). PLoS One 11(7):e0158857. Li Z, Traore A, Maximova S, Guiltinan MJ (1998) Somatic embryogen- https://doi.org/10.1371/journal.pone.0158857 esis and plant regeneration from floral explants of cacao Anandarajah K, McKersie BD (1990) Manipulating the desiccation tol- (Theobroma cacao L.) using thidiazuron. In Vitro Cell Dev Biol- erance and vigor of dry somatic embryos of Medicago sativa L. with Plant 34:293–299 sucrose, heat shock and abscisic acid. Plant Cell Rep 9:451–455 Loewus FA, Murthy PPN (2000) myo-Inositol metabolism in plants. Plant Archambault J, Williams RD, Lavoie L, Pepin MF, Chavarie C (1994) Sci 150:1–19 Production of somatic embryos in a helical ribbon impeller bioreac- Lopez-Baez O, Bollon H, Eskes A, Petiard V (1993) Somatic embryo- tor. Biotechnol Bioeng 44:930–943 genesis and plant regeneration from flower parts of cocoa Azevedo H, Dias A, Tavares RM (2008) Establishment and characteriza- Theobroma cacao L. C R Acad Sci III 316:579–584 tion of Pinus pinaster suspension cell cultures. Plant Cell Tissue Márquez-Martín B, Sesmero R, Quesada MA, Pliego-Alfaro F, Sánchez- Organ Cult 93:115–121 Romero C (2011) Water relations in culture media influence matu- Buchheim JA, Colburn SM, Ranch JP (1989) Maturation of soybean ration of avocado somatic embryos. J Plant Physiol 168:2028–2034 somatic embryos and the transition to plantlet growth. Plant Masseret B, Gianforcaro M, Bouquet JF, Brulard E, Florin B (2009) Physiol 89:768–775 Somatic embryogenesis applied to the creation of a cocoa collection. Malays Cocoa J 5:1–10 Castillón J, Kamo K (2002) Maturation and conversion of somatic em- Maximova SN, Alemanno L, Young A, Ferriere N, Traore A, Guiltinan bryos of three genetically diverse rose cultivars. Hortic Sci 37:973– MJ (2002) Efficiency, genotypic variability, and cellular origin of primary and secondary somatic embryogenesis of Theobroma cacao Cerezo S, Mercado JA, Pliego-Alfaro F (2011) An efficient regeneration L. In Vitro Cell Dev Biol-Plant 38:252–259 system via somatic embryogenesis in olive. Plant Cell Tissus Organ Maximova SN, Florez S, Shen X, Niemenak N, Zhang Y, Curtis W, Cult 106:337–344 Guiltinan MJ (2014) Genome-wide analysis reveals divergent pat- De Feria M, Jiménez E, Barbón R, Capote A, Chávez M, Quiala E (2003) terns of gene expression during zygotic and somatic embryo matu- Effect of dissolved oxygen concentration on differentiation of so- ration of Theobroma cacao L., the chocolate tree. BMC Plant Biol matic embryos of Coffea arabica cv. Catimor 9722. Plant Cell 14:185 Tissue Organ Cult 72:1–6 Maximova SN, Young A, Pishak S, Miller CR, Traore A, Guiltinan MJ Driver JA, Kuniyuki AH (1984) In vitro propagation of Paradox walnut (2005) Integrated system for propagation of Theobroma cacao L. In: rootstock. Hortic Sci 19:507–550 Jain SM, Gupta PK (eds) Protocol for somatic embryogenesis in Ducos JP, Labbe G, Lambot C, Pétiard V (2007) Pilot scale process for woody plants. Springer, Dordrecht, the Netherlands, pp 209–227 the production of pre-germinated somatic embryos of selected ro- McCown BH, Lloyd G (1981) Woody plant medium (WPM)—amineral busta (Coffea canephora) clones. In Vitro Cell Dev Biol-Plant 43: nutrient formulation for microculture of woody plant species. Hortic 652–659 Sci 16:453 Egertsdotter U, Clapham D (2015) Method for maturing and synchroniz- Miller CR (2009) An integrated in vitro and greenhouse orthotropic clon- ing conifer somatic embryos. Patent Number WO 2011123038 al propagation system for Theobroma cacao L. Ph.D. The El Meskaoui A, Desjardins Y, Tremblay FM (2006) The effects of in vitro Pennsylvania State University, USA gaseous environment on the maturation of somatic embryo of black Minyaka E, Niemenak N, Sangare A, Omokolo DN (2008) Effect of spruce (Picea mariana [mill.] BSP) and white spruce (Picea glauca MgSO and K SO on somatic embryo differentiation in 4 2 4 [Moench] Voss). Moroccan. J Biol 2:12–20 Theobroma cacao L. Plant Cell Tissue Organ Cult 94:149–160 El Meskaoui A, Tremblay FM (1999) Effects of sealed and vented gas- Motamayor JC, Lachenaud P, Da Silva E Mota JW, Loor R, Kuhn DN, eous microenvironments on the maturation of somatic embryos of Brown JS (2008) Geographic and genetic population differentiation black spruce with as special emphasis on ethylene. Plant Cell Tissue of the Amazonian chocolate tree (Theobroma cacao L.). PLoS ONE Organ Cult 56:201–209 3(10):e3311. https://doi.org/10.1371/journal.pone.0003311 Etienne H (2005) Somatic embryogenesis protocol: coffee (Coffea Murashige T, Skoog F (1962) A revised medium for rapid growth and arabica L. and C. canephora P.) In: Jain S. Gupta M, Pramod K bioassays with tobacco cultures. Physiol Plant 15:473–497 (eds) Protocol for somatic embryogenesis in woody plants. Springer, Niemenak N, Kaiser E, Maximova SN, Laremore T, Guiltinan MJ (2015) Dordrecht, Netherlands, pp 167–179 Proteome analysis during pod, zygotic and somatic embryo matura- Garcia C, Corrêa F, Findley S, Almeida AA, Costa M, Motamayor JC, tion of Theobroma cacao.JPlant Physiol 180:49–60 Schnell R, Marelli J-P (2016) Optimization of somatic embryogen- Niemenak N, Saare-Surminski K, Rohsius C, Ndoumou DO, Lieberei R esis procedure for commercial clones of Theobroma cacao L. Afr J (2008) Regeneration of somatic embryos in Theobroma cacao L. in Biotechnol 15:1936–1951 temporary immersion bioreactor and analyses of free amino acids in different tissues. Plant Cell Rep 27:667–676 Geng P, La H, Wang H, Stevens EJC (2008) Effect of sorbitol concentra- tion on regeneration of embryogenic calli in upland rice varieties Pullman G, Gupta P, Timmis R, Carpenter C, Kreitinger M, Welty E (Oryza sativa L.). Plant Cell Tissue Organ Cult 92:303–313 (2005) Improved Norway spruce somatic embryo development THEOBROMA CACAO L. INDIRECT SOMATIC EMBRYOGENESIS... through the use of abscisic acid combined with activated carbon. Söndahl MR, Nakamura T, Sharp WR (1985) Propagation of coffee. In: Henke R, Hughes KW, Constantin MJ, Hollaender A, Wilson CM Plant Cell Rep 24:271–279 Quainoo AK, Dwomon IB (2012) The effect of abscisic acid in the conver- (eds) Tissue culture in forestry and agriculture. Basic Life Sciences, sion of cocoa somatic embryos into plantlets. Front Plant Sci 2:6–10 vol 32. Springer, Boston, pp 215–232 Quinga LAP, De Freitas Fraga HP, Do Nascimento Vieira L, Guerra MP Söndahl MR, Sharp WR (1977) High frequency induction of somatic (2017) Epigenetics of long-term somatic embryogenesis in embryos in cultured leaf expiants of Coffea arabica L. Z Theobroma cacao L.: DNA methylation and recovery of embryo- Pflanzenphysiol 81:395–408 genic potential. Plant Cell Tissue Organ Cult 131:295–305 Teyssier C, Grondin C, Bonhomme L, Lomenech AM, Vallance M, Rodríguez López CM, Wetten AC, Wilkinson MJ (2010) Progressive Morabito D, Label P, Lelu-Walter MA (2011) Increased gelling erosion of genetic and epigenetic variation in callus-derived cocoa agent concentration promotes somatic embryo maturation in hybrid (Theobroma cacao) plants. New Phytol 186:856–868 larch (Larix×eurolepsis): a 2-DE proteomic analysis. Physiol Plant Sena Gomes AR, Andrade Sodré G, Guiltinan M, Lockwood R, 141:152–165 Maximova SN (2015) Chap 4: Tissue culture. In: Laliberté B, End Umehara M, Ogita S, Sasamoto H, Koshino H, Asami T, Fujioka S, M (eds) Supplying new cocoa planting material to farmers: a review Yoshida S, Kamada H (2005) Identification of a novel factor, of propagation methodologies. Bioversity International, Rome, Italy, vanillyl benzyl ether, which inhibits somatic embryogenesis of pp 67–87 Japanese larch (Larix leptolepis Gordon). Plant Cell Physiol 46: Shimazu T, Kurata K (2003) Dynamic dissolved oxygen concentration 445–453 control for enhancing the formation rate of torpedo-stage embryos in Van Boxtel J, Berthouly M (1996) High frequency somatic embryogen- carrot somatic embryo culture. J Biosci Bioeng 95:384–390 esis from coffee leaves. Plant Cell Tissue Organ Cult 44:7–17 Singh M, Chaturvedi R (2012) Evaluation of nutrient uptake and physical Zamarripa A, Ducos JP, Bollon H, Dufour M, Pétiard V (1991) parameters on cell biomass growth and production of spilanthol in Production of somatic embryos of coffee in liquid medium: effects suspension cultures of Spilanthes acmella Murr. Bioprocess Biosyst of inoculation density and renewal of the medium. Cafe Cacao The Eng 35:943–951 35:233–244 Smith SM, Street H (1974) The decline of embryogenic potential as callus and suspension cultures of carrot (Daucus carota L.) are serially subcultured. Ann Bot 38:223–241

Journal

In Vitro Cellular & Developmental Biology - PlantSpringer Journals

Published: Jun 4, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off