Multi-strategy approach towards optimization of maturation and germination in radiata pine somatic embryogenesis

Somatic embryogenesis is a powerful technique for forestry improvement programs when implementing multi-varietal forestry. However, it still faces some bottlenecks to be efficient in many forestry species. In this work we have studied the effect of some physico-chemical modifications at proliferation, maturation and germination stages of Pinus radiata D. Don somatic embryogenesis, as well as the impact of this modifications on plant conversion, survival, and on the morphology and biochemical profile (carbohydrate and amino acid analysis) of the resulting plantlets. Given the long culture period needed for maturation and germination in radiata pine, an increase of the plant yield could be very beneficial for its commercial application. Following these premises and based on the experiments performed, a preculture of 14 days without plant growth regulators before maturation seems to be beneficial for the development and germination of somatic embryos. Before germination, a cold storage at 4 °C had no detrimental effect and even increased plant conversion in some embryogenic cell lines (85% of germination and 64.6% of planted explants). During germination, supplementation of the culture medium with glutamine and a reduction of the sucrose content had a significant effect on germination (88%) and doubled the percentage of planted explants. Similarly, germinants were demonstrated to be influenced by the light source: fluorescent light enhanced root formation, whereas blue LED light increased the shoot height of somatic plants. Moreover, the highest plants showed lower contents of arginine, asparagine and total carbohydrates. Modification of the physico-chemical environment before or during somatic embryo maturation and germination improves the success of the process.


Introduction
Somatic embryogenesis (SE) is a regeneration pathway where dedifferentiated somatic plant cells become totipotent and develop into embryos that, subsequently, convert to plants (Fehér et al. 2003). The capacity of totipotent cells to proliferate and then to maturate in the right way is very influenced by physico-chemical aspects of in vitro culture, including balance of plant growth regulators (Gao et al. 2021) and medium osmolarity (do Nascimento et al. 2021), among others. Furthermore, stress treatments such as heavy metal ions, ultraviolet radiation, antibiotic applications, osmotic shocks, dehydration, heat or cold shocks, as well as mechanical or chemical stimuli deeply affect the success of the process (Zavattieri et al. 2010;Ibañez et al. 2020).
Since the first reports on Pinus radiata D. Don SE (Minocha 1999), an international effort has been carried out to improve different stages of SE in this species . On this matter, improved SE procedures including the overcome of several bottlenecks to produce somatic embryos (se's) more efficiently (Montalbán et al. 2010) and synchronized (Maruyama and Hosoi 2012; Communicated by Henryk Flachowsky. Ander Castander-Olarieta and Itziar A. Montalbán Li et al. 2022) are essential to implement multi-varietal forestry. Nowadays, there still are problems to be solved along the process (maturation, germination and ex vitro acclimatization) in many forestry species in order to make it a success. Considering the timing from the start of P. radiata maturation to the end of ex vitro acclimatization (at least 25 weeks), it is very important to improve the last phases of the process from a technical and an economical point of view. In this sense, many years ago, different concentrations of abscisic acid (Von Aderkas et al. 2002), activated charcoal (AC) addition (Yao and Wang 2020), modification of carbohydrates and PEG to modify the gel strength (Salaj et al. 2019), the effect of ethylene (Neves et al. 2021), the application of auxin transport inhibitors (Verma et al. 2018) and redox compounds (Kudelko and Gaj 2019) and inhibitors of SE (Nic-Can et al. 2015) were assayed to improve the yield of maturation process. However, the most influential traits on embryo quality identified were the genotype (Kong and Von Aderkas 2007), and the developmental stage of the tissue used as initial explant (Montalbán et al. 2012). In this regard, in Pinus spp., the genotype is determined by the genetics of parental trees with a stronger maternal effect on SE initiation (Niskanen et al. 2004) and se's production (Sun et al. 2022). Having in mind this major drawback, it is necessary to perform experiments with a large number of embryogenic cell lines (ECLs) in order to establish broader conclusions and general protocols that could be applied to several genotypes from different genetic backgrounds.
Based on our experience, the modification of the environmental conditions in one stage of the process has implications not only in that step but also in the subsequent ones (Garcia-Mendiguren et al. 2016;Pereira et al. 2017;. For instance, in previous experiments carried out in our laboratory, temperature was a determining factor, influencing the number and the morphology of embryogenic cell lines and se's, and also the biochemical profile of se's and the resulting somatic plants (Castander-Olarieta et al. 2019Pereira et al. 2021. Similar results were observed when modifying the carbohydrate and amino acid sources during maturation in both P. radiata and P. halepensis (do Nascimento et al. 2021).
Taking all the above-mentioned information into account, the objective of the present study was to analyse the effect of different physico-chemical environments before or during maturation and along germination on the success of plant conversion in order to increase somatic plant production in a wider range of genotypes. Moreover, we have carried out biochemical analyses to determine if some of the treatments applied have implications in the soluble carbohydrate and amino acid profile of the plantlets developed. To achieve these objectives, a multi-strategy approach was followed: during maturation, the type, the concentration and the exposure period of embryonal masses (EMs) to different phytohormones and nutrients was assayed, as well as the inoculum density and agar concentration of the culture medium. Additionally, during the germination process, not only the chemical environment (through amino acid supplementation), but also the physical one (temperature or illumination) were studied; in this regard practical modifications such as cold storage, which could help to adjust planting timing or LED lighting, which could help reducing the cost of the process (Rojas-Vargas et al. 2022) were tested.

Plant material
Radiata pine ECLs were initiated and proliferated following Montalbán and Moncaleán (2018). Megagametophytes came from ten open pollinated trees from different locations in the Basque Country; information about the origin of each ECL is described in Supplementary material Table S1.

Maturation experiments
Embryo development medium (EDM, Walter et al. 2005) was the basal medium used for initiation, proliferation (EDM-p), and maturation (EDM-m). Media components for different stages of the SE process at control conditions are described in Table 1. Initiation and proliferation were carried out on Petri dishes filled with 18 mL of medium, whereas maturation was carried out in 90 × 20 mm petri dishes filled with 40 mL of medium.
Proliferation was carried out by subculturing EMs every 14 days to fresh medium (5-6 clumps of 12-15 mm diameter each per petri dish). To carry out maturation experiments, Montalbán et al. (2010) methodology was followed. Briefly, embryogenic tissue (ET) was resuspended in basal liquid medium without plant growth regulators and amino acids. In all maturation experiments, aliquots of 5 mL containing 90 mg fresh weight of ET were poured on a filter paper (Whatman, nº2) and the liquid medium was drained in a Büchner funnel; then, the filter paper with the attached ET was laid on semi-solid maturation medium.

Maturation experiment 1
Six ECLs (RE19-10, RE19-23, RE19-37, RE19-55, RE19-68 and RE19-114) were subjected to maturation. The maturation treatments were: M1-Control (filters were laid on EDM-m without subculturing for 16 weeks), M2-The filters with the attached ET were cultured for four weeks on EDM-m without ABA, and then the filters were transferred to EDM-m for 12 weeks, M3-The filters with the ET were cultured on EDM-m for eight weeks, and then they were transferred to EDM-m without ABA for eight weeks.

Maturation experiment 2
Six ECLs (RE19-10, RE19-23, RE19-46, RE19-55, RE19-67 and RE19-114) were precultured on different proliferation media for 14 days before maturation. The maturation pre-treatments were: M4-EMs cultured for 14 days in clumps on EDM-p devoiding plant growth regulators, M5-EMs cultured for 14 days in clumps on EDM-p supplemented with a higher agar concentration (6 g L −1 ) and M1-Control (EMs cultured for 14 days in clumps on EDM-p).  Walter et al. 2005) and half strength macronutrients modified LP medium developed by Quoirin and Lepoivre (1977), and modified by Aitken-Christie (personal communication). BA, 2,4-D and ABA stand for 6-benzyladenine, 2,4-dichlorophenoxyacetic acid and abscisic acid, respectively a Filter -sterilised stock solution was added after autoclaving and cooling to 60 °C After 14 days in different pre-maturation media, all EMs were subjected to maturation at standard conditions as described previously.
After 14 days in different pre-maturation media, all EMs were subjected to maturation at standard conditions as described previously.
After 14 days in different pre-maturation media, all EMs were subjected to maturation at standard conditions as described previously.

Maturation experiment 5
Six ECLs (RE20-10, RE20-62, RE20-94, RE20-99, RE20-105 and RE20-131) were subjected to maturation. The maturation treatments were: M10-45 mg of EMs instead of 90 mg of EMs per filter were cultured on EDM-m, M11-45 mg of EMs per filter were cultured on LGE (Table 1) for 10 days, and then the filters were transferred to EDM-m. Filters with 90 mg of EMs laid on EDM-m without subculturing for 16 weeks were used as control (M1).

Germination experiments
The germination medium used was half strength macronutrients modified LP medium, developed by Quoirin and Lepoivre 1977 and modified by Aitken-Christie (personal communication, LGE, Table 1). Germination was carried out on Petri dishes filled with 18 mL of medium for the first six weeks and then on ecoboxes (Eco2box/green filter: a polypropylene vessel with a ''breathing'' hermetic cover, 125 × 65x80 mm) filled with 90 mL of medium.
Germination under control conditions was carried out by collecting normal se's (white to yellowish, non-germinating se´s, with a distinct hypocotyl region, and at least three cotyledons) from maturation medium after 16 weeks from the start of maturation stage. Fifteen se´s per petri dish (90 × 15 mm) were laid horizontally on each petri dish, with the embryonal root caps pointing downwards. The petri dishes were tilted vertically at an angle of approximately 60º-70º and placed under a 16 h photoperiod provided by cool white fluorescent tubes (TLD 58 W/33; Philips, France) at 80 µmol m −2 s −1 for the first week and then at 120 µmol m −2 s −1 for the rest of germination stage (11 weeks). All in vitro stages of SE were carried out at 21 ± 1 °C.
After six weeks, germinated se's were transferred to ecoboxes. After 12 weeks from the start of germination stage the plantlets were ex vitro acclimatized. Somatic plantlets were transferred to 43 cm 3 individual pots containing blond peat moss (Pindstrup, Ryomgård, Denmark): vermiculite (8:2, v/v) and acclimatized in a greenhouse under controlled conditions at 22 °C, decreasing humidity progressively from 95 to 80%.

Germination experiment 1
Somatic embryos obtained from the ECLs of maturation experiments 1 to 4 were subjected to maturation. Two germination media were tested: the standard germination medium (LGE medium) and LGE with the sucrose concentration reduced from 30 to 15 gL −1 and supplemented with 0.5 gL −1 L-glutamine after autoclaving (RG medium). After six weeks, the germinated se´s from LGE and RG were transferred to ecoboxes filled with their respective medium.

Germination experiment 2
Somatic embryos from eight ECLs (RE20-10, RE20-15, RE20-17, RE20-34, RE20-38, RE20-44, RE20-62, and RE20-108) were obtained following the standard maturation protocol. After 16 weeks on maturation medium the se´s were germinated. Also, maturation plates from the same ECLs were stored, after 16 weeks of maturation, further eight weeks in darkness at 4 °C; after this storage period, se´s were subjected to germination. A total of 1260 se´s were cultured, fifteen se´s per Petri dish and six Petri dishes per ECL (630 se´s) for each condition (control or 4 °C) in the abovementioned lighting and temperature conditions. Based on results from germination experiment 1 all se´s were germinated on RG.
At the time of germination, se´s from control and from 4 °C treatment were collected from 3 ECLs (five replicates per ECL and germination treatment) and oven dried at 70 °C for 24 h to calculate the dry weight.
After 12 weeks from the beginning of germination period, when plantlets were ready to be planted ex vitro, the number of roots was recorded and plants were photographed and measured (the shoot and the principal root separately) using ImageJ software (version 1.49).

Germination experiment 4
Somatic embryos from four ECLs (RE20-38, RE20-52, RE20-56 and RE20-62) were tested for the first six weeks in Petri dishes under the same four light conditions mentioned above (45 se´s per ECL and light type): blue, red and white LED lights, all of them at 60 µmol m −2 s −1 . Cool white fluorescent tubes at 80 µmol m −2 s −1 for the first week and at 120 µmol m −2 s −1 for the rest of germination period was used as control. After six weeks, germinated plantlets from different light treatments were transferred to ecoboxes in the same light treatments as in the previous stage.
All cultures were in the same growth chamber at 21 ± 1 °C and with the same 16-h photoperiod.
After 6 weeks in germination medium (when the explants were transferred from Petri dishes to ecoboxes) and after 12 weeks from the beginning of germination period when plants were ready to be planted ex vitro, the plants were photographed and measured (the shoot and the principal root separately) using ImageJ software. After 12 weeks, the number of roots of the plants ready to be planted was also recorded.

Soluble sugar and amino acid extraction and quantification
Fresh entire plantlets from each ECL and light treatment of Germination Experiment 4 were grinded in liquid nitrogen after six weeks in germination medium; 100 mg of the resulting fine powder were transferred to 2 ml microcentrifuge tubes containing 800 µl cold metabolite extraction buffer (methanol:chloroform:water, 2.5:1:0.5, v:v:v). Tubes were centrifuged at 20,000 g for 6 min at 4 °C and the supernatants were transferred to new tubes containing 800 µl of phase separation mix (chloroform:water, 1:1).
After centrifugation at 10,000 g for 5 min at room temperature, the upper aqueous phases containing polar metabolites were saved to new 1.5 ml tubes. Aliquots of 200 ml were then obtained from those tubes and totally dried on a Speedvac to remove the remaining methanol. Then, samples were resuspended in 100 µl ultra-pure water and soluble sugars and amino acids were quantified by HPLC (Agilent 1260 Infinity II). Soluble sugars, including certain sugar alcohols (fructose, glucose, sucrose, mannitol, sorbitol), were separated using an Hi-Plex Ca column (7.7 mm × 300 mm, 8 µm) and detected using a refractive index detector at a flow rate of 0.15 ml min −1 pure water at 80 °C for 30 min. For amino acids (aspartic acid, glutamic acid, asparagine, serine, glutamine, histidine, glycine, threonine, arginine, alanine, tyrosine, cysteine, valine, methionine, tryptophan, phenylalanine, isoleucine, leucine, lysine, hydroxyproline, proline) a Poroshell 120 column (4.6 mm × 100 mm, 2.7 µm) was used coupled to a fluorescence detector. The samples were injected into the column at a flow rate of 1.5 ml min −1 at 40 °C for 18 min with a discontinuous gradient. Solvent A was a mixture of 10 mM Na 2 HPO 4 and 10 mM Na 2 B 4 O 7 (pH 8.2) and solvent B was acetonitrile:methanol:water (45:45:10, v:v:v). The gradient program was the following: min 0-13.40, solvent A 98% and solvent B 2%, min 13.40-13.50, solvent A 43% and solvent B 57%, min 13.50-15.80, solvent B 100%, and min 15.80-18, solvent A 98% and solvent B 2%. Amino acids were estimated after pre-column derivatization by mixing 1 µl sample with 2.5 µl borate buffer, 32 µl diluent (100 ml solvent A and 0.4 ml concentrated H 3 PO 4 ) and 0.5 µl o-phthaldialdehyde. For hydroxyproline and proline analysis, samples were mixed with 0.5 µl fluorenylmethoxycarbonyl protecting group instead of o-phthaldialdehyde. Detection was performed by analysing fluorescence with excitation at 260 nm and emission at 450 nm for o-phthaldialdehyde derivatives and emission at 325 nm for fluorenylmethoxycarbonyl derivatives. Both sugar and amino acid concentrations were determined from internal calibration curves constructed with the corresponding commercial standards. Results were conveniently adjusted considering the concentration step after methanol removal (2 times) and expressed as μmol g FW -1 . In the case of certain amino acids, samples were re-injected after dilution 1/32 to avoid signal saturation and the resulting concentrations were conveniently adjusted.

Data collection and statistical analyses
For maturation experiments, four maturation replicates (petri dishes) were cultured per ECL and treatment; data on the number of se's per ECL and treatment were recorded after 16 weeks. For those ECLs and treatments that produced se´s, an analysis of variance (ANOVA) was carried out and when necessary, Duncan´s post hoc was performed to determine differences between treatments. In maturation experiment 2, data did not meet homocedasticity; thus, Tamhane post hoc test was performed. In maturation experiment 4, data were transformed to 1/x to meet homocedastivity of variance.
For germination experiments, the percentage of germination was evaluated after six weeks from the start of germination period when germinated se´s were transferred from petri dishes to ecoboxes. The percentage of planted explants was calculated after another six weeks in ecoboxes when the plantlets were planted ex vitro. Regarding the percentage of planted explants in germination experiments 3 and 4, the ECL was considerate as a replicate.
Data of the measurements of shoots and roots in experiment 4 were log (x) transformed to meet homocedasticity, then ANOVA was performed and when necessary, Duncan´s post hoc test was performed.
ANOVA was carried out for germination percentages, for the percentage of planted explants and for the percentage of dry weight of se´s, and when necessary, Duncan´s post hoc was performed. In germination experiment 2, data on germination percentages were transformed to x 2 to meet homocedastivity.
In germination experiments 3 and 4, the number ANOVA was carried out for the number of roots of the somatic plants when planting them ex vitro. These data were 1/x transformed to meet homocedasticity. ANOVA was also carried out for shoot and principal root lenght. In experiment 3, these data were transformed to log (x) to meet homocedasticity of variance. In experiment 4, the data on root length after 6 and 12 weeks were also transformed to log (x) to meet homocedasticity. When necessary, Duncan´s post hoc test was carried out, except for data on shoot length after 6 weeks (germination experiment 4) where Tamhane´s post hoc test was performed as they showed heterocedasticity.
Data for carbohydrate analyses met homocedasticity except those for galactose contents, which were x 2 transformed. In the case of amino acids, data on glutamic acid were x 2 transformed; data on methionine and serine were 1/x transformed and data on aspartic acid were 1/log (x) transformed to meet homocedasticity. All data were subjected to ANOVA, and when they showed significant differences, Duncan´s post hoc test was performed; data on aspartic acid and asparagine showed heterocedasticity for these a Tamhane´s post hoc test was carried out.

Maturation experiment 1
Two ECLs (RE19-10 and RE19-55) out of six did not produced se's in any of the media tested, and RE20-68 produced 3 se´s/g FW in the three-maturation media assayed.
The other three ECLs (RE19-23, RE19-37 and RE19-114) produced se´s in all the maturation media tested, but significant differences were found regarding the genotype and the maturation medium (there were no significant differences for the interaction between these factors, Supplementary material Table S2). Genotype RE20-23 produced the highest number of se´s (353 se´s g −1 FW) and RE19-37 the lowest one (38 se´s g −1 FW).
Regarding maturation media, no significant differences were found between results obtained following the standard procedure (on EDM-m with ABA for 16 weeks) or in those subcultured to EDM-m without ABA for the last eight weeks of maturation; whereas a significantly lower number of se´s was obtained when the ABA was removed from maturation medium for the first four weeks of maturation (Fig. 1A).

Maturation experiment 2
Half of the ECLs tested (RE19-10, RE19-46, and RE19-55) did not produce se´s after preculture in different proliferation media. The ECL, the preculture performed and the interaction between them had a significant effect on the se's production (Supplementary material Table S2). As can be seen in Fig. 1B for RE19-23 and RE19-67, the best results were obtained when the tissue was precultured in EDM-p devoiding growth regulators, but for the less productive line (RE19-114) the best results were obtained after a preculture on EDM-p with a higher gellam gum concentration.
Considering the embryo production for different ECLs and treatments, due to the significant interaction between the factors and the fact that data presented unequal variances, only significant differences between treatments were observed for RE19-67 were a higher number of se´s (161 se´s g −1 FW) was obtained after preculture in medium devoid of plant growth regulators than in control conditions (61 se´s g −1 FW, Fig. 1B).

Maturation experiment 3
Only two (RE19-58 and RM19-3) of the six ECLs tested produced se´s in all treatments. In RM19-1, a few se´s were only obtained after control treatment or after preculture in medium with 2,4-D (22 and 19 se´s g −1 FW, respectively).
No significant interaction was observed between the ECLs and the preculture applied (Supplementary material  Table S2). In this regard, RM19-53 produced a significantly higher number of se´s than RE19-58 and in both cases the best pre-treatments were control and preculture in medium with only BA, although the latter did not present significant differences with the preculture with only 2,4D (Fig. 1C).

Maturation experiment 4
Five of the six ECLs tested produced se´s; however, RE20-42 only produced 3 se´s gL −1 after a preculture on medium with kinetin instead BA and in RE19-67 no se´s were obtained in control conditions whereas 19 se´s gL −1 were recorded after preculture on media with kinetin or NAA. For the other three ECLs (RE19-114, RE20-17 and RE20-21) an interaction between the ECL and the preculture applied was observed (Fig. 1D, Supplementary material Table S2). In this regard, significant differences were only observed between the lowest embryo production in control conditions for RE19-114 and the number of somatic embryos obtained after preculture in medium with NAA in this ECL and RE20-17, and the somatic embryo production in RE20-21 independently of the treatment tested (Fig. 1D).

Maturation experiment 5
Somatic embryos were obtained in all ECLs, except for RE20-99, which did not produce se´s on treatment M10 and produced less than 10 se´s per plate in the other treatments. In these ECLs a significant effect of the line, the treatment and the interaction between them was found (Fig. 1E, Supplementary material Table S2). For RE20-131, lowering the initial amount of tissue (M10) had a significant impact when compared to control conditions (M1); a similar effect (although without statistical significance) was observed for RE20-105. In the case of RE20-62 and RE20-94, lowering the initial amount of tissue together with a preculture on germination medium (M11) significantly raised the number of se´s obtained when compared to control conditions (Fig. 1E).

Germination experiment 1
For maturation experiments 1, 3 and 4, the genotype (ECL) had a significant effect on germination percentages. However, no effect was observed for the maturation treatment or for the interaction between these factors (Supplementary material, Table S3).
The germination percentage in se´s from maturation experiment 2 was significantly affected by the previous maturation treatment. In this sense, the se´s obtained after preculture without growth regulators showed a significantly higher germination rate (88%) than those from standard conditions (72%).
Taking all data together from different maturation experiments (1195 germinated se´s from eight genotypes), the germination was not significantly affected by the germination medium. But when considering the percentage of planted explants, the genotype and the germination treatment had a significant effect on the number of plants that were suitable to be planted (Supplementary material Table S4). No interaction between these two factors was observed. A significantly higher percentage of planted explants was recorded in RG medium; 37% of the plants from RG medium were suitable for ex vitro planting, whereas only 16% of the plants from standard in vitro germination were ready to be planted. The acclimatization of planted explants was above 90% for all treatments applied.

Germination experiment 2
No significant differences were found for the percentage of dry weight of fresh se´s and se´s after cold storage. A significant effect of the ECL, the storage method (control vs eight weeks at 4 °C) and an interaction between them was observed for germination percentages (Supplementary  material Table S4). Significantly higher germination percentages were observed after storage at 4 °C (86.0%) than at control conditions (67.9%). When ECLs were compared independently, significantly higher germination percentages were observed when they were cold stored, except for RE20-15 and RE20-108 where no significant differences were observed (Fig. 2A).
The percentage of planted explants was also significantly affected by the ECL, the storage method and the interaction between them. In this case, although the storage at 4 °C led to a significantly higher percentage of planted explants (64.6%) when compared with the results at standard conditions (55.8%), only two ECLs (RE20-15 and RE20-44) showed significantly better percentages of planted explants after cold storage (Fig. 2B). The acclimatization percentage of planted explants was above 90% for all treatments.

Germination experiment 3
The first six weeks of germination were carried out under the same conditions (fluorescent lights); thus, only the number of planted explants (after different light treatments for the last six weeks of germination process) was analysed. These different light treatments did not have a significant effect on the percentage of planted explants, which ranged from 39.6% for plants grown under white LED light to 61.3% for those grown under fluorescent light (Supplementary material Table S5). The acclimatization percentage of planted explants was above 95% for all treatments.
ANOVA revealed that the length of the shoots was only significantly affected by the genotype (Supplementary material Table S6). Regarding this, genotype RE20-44 showed significantly higher shoots than RE20-62 and RE20-10, whereas RE20-21 showed intermediate values.
The principal root length was significantly affected by the genotype and the light treatment, however no significant interaction between these factors was observed (Supplementary material Table S6). In this case, RE20-62 showed significantly higher values than RE20-10 and RE20-21 showing RE20-44 intermediate values.
Plants grown the entire germination period under fluorescent light showed longer roots than those from red and white LED treatments (Figs. 3, 4A, B), whereas those cultured under blue light for the last six weeks of the germination period displayed intermediate values (Fig. 3).
No significant differences were detected for the number of roots in somatic plants from different light treatments before ex vitro planting (Supplementary material  Table S7). The highest number of roots was observed in plants cultured all the germination period under fluorescent lights and the lowest in plants grown under red LED lights the last six weeks of germination period (Fig. 4, B).

Germination experiment 4
Plants germinated under different lights showed significantly different germination percentages (Supplementary material  Table S5). Although plants germinated in red light showed significantly lower germination percentages than those from other light treatments (between 88.9 and 92.2%), this percentage was also high (82.8%). The percentage of planted explants was not significantly affected by the light treatment, ranging from 46.2% for plants developed for 12 weeks under red light to 68.5% for those under fluorescent light. The acclimatization of planted explants was above 95% for all treatments.
The shoot length of the somatic plantlets after 6 weeks and 12 weeks of germination, was significantly affected by Fig. 2 A Germination percentage for different embryogenic cell lines at different storage conditions in germination experiment 2. B Percentage of planted explants for different embryogenic cell lines at different storage conditions in germination experiment 2. Different letters indicate significant differences the genotype and the light treatment, but there wasn´t a significant interaction between them. In this sense, genotype RE20-56 showed significantly shorter shoots than the other genotypes tested. Regarding light treatments, the shoots of the plantlets grown under blue light were significantly longer than the rest after 6 weeks (Figs. 4C, D, 5A); after 12 weeks (Fig. 5B) the same tendency was observed.
The length of the principal root in somatic explants after 6 weeks and 12 weeks was significantly affected by the genotype, the light treatment and the interaction between these variables (Supplementary material, Tables S8 and S9). After 6 weeks, RE20-52 showed significantly higher values than RE20-38 and RE20-62, and these presented significantly longer roots than RE20-56. Plantlets grown under blue LED or fluorescent lights showed significantly longer roots than those grown under white or red LED lights (Fig. 6A). After 12 weeks, only RE20-56 showed significantly lower root lengths than the other genotypes tested. At the end of germination period, fluorescent light led to significantly higher root lengths than blue LED lights, and plantlets growing under white and red LED light had significantly shorter roots than those from blue LED or fluorescent lights. The effect of light treatment can also be seen when considering the interaction between variables, particularly in the genotypes with longer roots (RE20-38, RE20-52, and RE20-62) under fluorescent light (Fig. 6B).
The number of roots of the somatic plants when they were transferred to ex vitro conditions was significantly higher in plants grown under fluorescent light (1.80) than in those cultured under different LED light treatments (lower than 1.04).

Soluble sugar and amino acid extraction and quantification
The concentration of the carbohydrates analysed separately in the explants grown under different lights did not differ statistically (Supplementary material Table S10). The total carbohydrate content was significantly higher in explants grown under red LED light (10.16 µ/g FW) than in explants under blue LED light (8.05 µ/g FW); this content showed intermediate values in somatic plants after white LED and fluorescent lights (Supplementary  material Table S11). The glucose and fructose contents in the somatic plantlets were significantly higher than the Amino acid contents (µmol /g fresh weight) in somatic plantlets grown for 6 weeks under different light treatments (Germination experiment 4). Different letters or numbers indicate significant differences mannitol and sucrose contents, and these were significantly higher than the galactose amounts.
Only the contents of arginine and asparagine differed significantly in plants grown under different lights. Somatic plantlets grown under blue LED light showed significantly lower contents of these amino acids than those grown under the other lights tested (Fig. 7, Supplementary material Table S13).

Discussion
Pine SE has experienced several improvements since the first reports in 1980s (Gupta and Durzan 1986;Mohan Jain et al. 1989). Some of them have highlighted the importance of capturing many genotypes by improving ET initiation; however, many of these ECLs have not been tested for their embryo production, and when so, this production varies greatly among genotypes. So, optimization of SE process is critical in conifer biotechnology towards multi-varietal forestry that uses elite varieties to cope with environmental and socio-economic issues (Llebrés et al. 2018).
In line with this, it is important to develop maturation and germination protocols that have been tested in several ECLs from different parental trees. In the present study, considering the existing knowledge (Lelu-Walter et al. 2016 and references therein) and other untested environmental conditions in Pinus radiata SE, we have tried to improve the last steps of in vitro SE to be able to obtain plants in a wider range of genotypes.
Although it is current practice in other Pinus (P. pinaster, Arrillaga et al. 2019) or P. sylvestris Abrahamsson et al. 2018), when ABA was eliminated from the maturation medium, during the first weeks of culture, a lower se's was obtained. On the other hand, elimination of ABA after eight weeks did not lead to a reduction in the number of se's obtained. In almost all pine SE protocols, maturation is accomplished by the addition of ABA to the culture medium and by increasing sugar and/or gelling agent concentrations (Maruyama and Hosoi 2019). Transferring ET to maturation medium typically provokes a sharp increase in the endogenous profile of ABA, but the dynamics of this phytohormone in following developmental stages of embryo development contrast among species and ECLs (Von Aderkas 2001). In this sense, prolonged levels of ABA were correlated with high levels of embryo production in P. pinaster (Arrillaga et al. 2019), similarly to what observed in Vales et al. (2007) for loblolly pine. However, in Norway spruce, endogenous ABA reaches a peak during the first weeks of maturation but progressively decreases at later stages until germination (Vondrakova et al. 2018). Based on our results in radiata pine, ABA and low water availability should be concomitant during early stages of maturation, while exogenous ABA seems not to be essential during further stages, suggesting that from one point onwards endogenous ABA levels are independently regulated to meet the ABA-mediated requirements for germination, such as storage protein accumulation (Tereso et al. 2007) or precocious germination avoidance (Rodriguez-Gacio et al. 2009).
Overall, preculture with just one of the phytohormones currently used for proliferation (BA and 2, led to lower number of se's, especially when removing BA from the culture medium. To this day, all published studies in pine SE require the combination of both auxins and cytokinins during proliferation , and in several angiosperms, authors highlight the importance of a proper auxin-cytokinin balance to directing SE (Martín et al. 2000;Avilez-Montalvo et al. 2022). Nonetheless, removal of both auxin and cytokinin equalled or improved the results during maturation, depending on the ECL. Avoiding long culture of ET in media containing 2,4-D or reducing the content of this hormone before maturation could help to increase embryo yield as 2,4-D has been proved tochange DNA pattern and induces DNA mutations, altering the expression of genes required during embryo formation and compromising polar auxin transport (García et al. 2019). In this regard, the somatic embryos generated after preconditioning without phytohormones also presented higher germination rates than the ones generated under standard conditions.
Although BA and 2,4-D are the most popular phytohormones used at early stages of SE for their affordability and successful performance, in some species kinetin instead or BA (Pinus halepensis, Pereira et al. 2020) or NAA instead of 2,4-D (Picea abies, Ramarosandratana and van Staden 2004) have led to better results. We have previously described that the physico-chemical environment of cultures has consequences on the subsequent stages of the process (García-Mendiguren et al. 2016;Moncaleán et al. 2018;Castander-Olarieta et al. 2019). In these experiments, the somatic embryo production after NAA preculture was similar or better when compared with standard proliferation protocol. NAA is commonly used in many angiosperm SE protocols, for example in Digitalis ferruginea (Verma et al. 2014). In this regard, although a higher callogenesis was obtained using 2,4-D, the maximum number of somatic embryos was obtained when NAA was added to the medium.
Different authors have stated that excessive tissue proliferation could hinder the development of somatic embryos (Lelu-Walter et al. 2006). Following this hypothesis, some authors have sought a reduction on proliferation rates preconditioning the ET in different media (reducing nutrient or hormone concentrations) to improve maturation (Carneros et al. 2009). Thereby, in maturation experiment 4, a reduction of the initial amount of ET with or without a preculture on a medium with AC was assayed. Despite the interaction between the ECLs and the treatments, it could be observed that within each line lowering the initial amount of inoculum did not have a detrimental effect on somatic embryo production. Moreover, the preculture on LGE promoted a higher number of se´s per gram of fresh weight of ET when compared with control conditions in two of the genotypes tested. AC is added in some SE maturation protocols in the suspension medium (Lelu-Walter et al 2008), and it is known for its adsorption of inhibitory compounds such as toxic metabolites or phenolic exudations (Thomas 2008) that could be adverse for somatic embryo maturation. In previous experiments, we observed a negative effect of AC in the suspension medium; this may be due to an excessive interaction of AC with the embryogenic cells which produced abnormal se's (Montalbán et al. 2010). In the current experiment, this effect could have been attenuated by the presence of a physical barrier (the filter paper) between the AC and the ET.
Regarding germination, a higher percentage of somatic plantlets was acclimatized successfully when the medium was supplemented with glutamine; in the same way, Carlsson et al. (2019) observed that glutamine was the preferred source of nitrogen in P. abies somatic germinants, accounting for half of the assimilated nitrogen. In the same species, Tikkinen et al. (2018) reported that lowering inorganic nitrogen concentration (ammonium and nitrate) from 1980 to 583 mg of nitrate had a positive effect on root length. However, it must be noticed that the total inorganic nitrogen concentration was much superior in their control medium than in ours (950 mg). In line with this study, Llebrés et al. (2018) found that Pinus strobus x P. wallichiana se´s developed longer roots in medium devoiding inorganic nitrogen. Similarly, increasing organic nitrogen concentration during maturation had a long-term positive effect during both germination and acclimatization in our previous study with radiata pine (do Nascimento et al. 2021).
Furthermore, sucrose was 50% reduced in the medium supplemented with organic nitrogen and this photomixotrophic conditions (considering the ventilation provided by ecoboxes) could have helped later acclimatization of the somatic plantlets in agreement with other authors (Pinheiro et al. 2021;Arencibia et al. 2018). However, the percentage of acclimatization was very high for all planted explants, regardless of the sucrose concentration in the germination medium. This survival (over 90%) doubled previous reports using LGE medium and ecoboxes in our lab . This could be due to a change in the peat employed and in the substitution of perlite for vermiculite. Different authors have reported that sucrose concentrations could influence the composition of exodermis in in vitro propagated plants (Martins et al. 2020), so testing an in vitro transition to photoautotrophic conditions by progressively removing sucrose from germination medium will be one of our future goals.
Although P. radiata se´s obtained following our standard protocol do not need any post-maturation treatment to germinate successfully, it has been hypothesized that a cold treatment before germination leads to a partial desiccation that benefits germination (Lipavská et al. 2004). Contrary to this, in our experiments the se´s did not present significant differences in their dry weight percentage before and after cold storage; they were not moved from the maturation plates and those were not unsealed for the entire maturationstorage period; this together with the fact that they were stored in a high-capacity refrigerator with a constant temperature kept the se´s free from condensation or other visible changes, contrary of what was reported for some cold storage experiments by Tikkinen et al. 2018. To this respect, previous experiments during longer periods (data not shown) suggest that it is possible to store the se´s for periods up to 9 months, in accordance with Varis et al. 2017.
The light spectra employed during the different germination experiments had a significant impact on several aspects of the plantlets produced; overall red and white LED lights gave the worst results for most of the parameters analysed: germination rates, number of plants ready to be planted ex vitro, shoot and root lengths and number of roots. Fluorescent turned out to be the most appropriate source of light in terms of root formation and elongation, while blue LED light had a positive effect during early germination as an inducer of shoot elongation. Bibliography on this topic is usually contradictory and is typically influenced by the species, the genotype and the developmental stage of the plantlets. In Norway spruce, Varis et al. (2021) reported shorter roots and shoots in se's germinated under blue light, whereas Riikonen et al. (2016) showed a positive effect of high blue proportions during the first steps of germination. In Pinus sylvestris, red light seems to be the preferred one throughout the germination process (Merkle et al. 2006;Alakärppä et al. 2019;Pashkovskiy et al. 2021), contrasting with the results obtained in our study with radiata pine. On the other hand, Ranade et al. (2016) did not observe any detrimental effect of blue light on Scots pine hypocotyl growth when compared with other monochromatic light sources, and even showed enhanced fibre length and width in subsequent growing seasons of outplanted seedlings. Interestingly, Montagnoli et al. (2018) demonstrated that roots show more plasticity to light spectra than shoots. In this sense, as seedlings grow, shoots tend to be less influenced by the light source than roots, which could explain the lack of shoot morphological differences in Germination Experiment 3.
The light source, directly or indirectly, also had a significant impact on the metabolite profile of the emblings. Plantlets germinated under blue LED light had significantly lower levels of both arginine and asparagine, and a lower total carbohydrate pool, being the plantlets germinated under red and white LED lights the ones with the highest levels of all these metabolites. Following the same pattern observed in other pine seedlings (Tershkikh et al. 2005;Cánovas et al. 2007), asparagine, arginine and glutamine were the most abundant amino acids in our work. Asparagine and arginine are important amino acids during pine germination. Storage proteins accumulated during maturing embryos are rich in both amino acids, and when germination starts, especially after radicle emergence, these amino acids are released and asparagine becomes the most abundant free amino acid (Cañas et al. 2006). These molecules are used for the synthesis of other metabolites, building biomass and provide energy during germination (Carlsson et al. 2019). Then, arginine is converted to ornithine and urea and the subsequent hydrolysis of urea leads to the formation of ammonium to be used in the synthesis of glutamine (Cañas et al. 2016). However, in our study glutamine was added into the germination medium, so we cannot confirm if the high glutamine levels have an endogenous or exogenous origin. In any case, lower levels of both arginine and asparagine under blue LED light did not have a detrimental effect in the germination capacity or morphological quality of the plantlets. As reported by several authors in Arabidopsis thaliana (Lam et al. 1998;Wong et al. 2004), the asparagine synthetase genes modulate the levels of asparagine depending on the light conditions. Nonetheless, it is worth highlighting that the light sources leading to the poorest growth characteristics presented the highest metabolite pools. Considering this, together with the idea that asparagine is a nitrogen reserve used to preserve the excess of mobilized nitrogen that is not immediately used (Cañas et al. 2016), we could hypothesize that the low levels of this amino acids under blue LED are more likely linked to the enhanced growth parameters observed after 6 weeks of germination under these light conditions, rather than to a direct effect of the light.
Summarizing, a preculture of 14 days without plant growth regulators seem to be beneficial for the development and germination of se´s. Further, storage at 4 °C has no detrimental effect, and can even increase plant conversion. Likewise, supplementation with glutamine and reduction of sucrose during germination improves the acclimatization success of the plantlets. Based in our experiments it would be also interesting in the future to test germination under blue LED light for the first six weeks to favour shoot growth and then transferring the germinated se´s to fluorescent lights to enhance root development.

Acknowledgements
The authors are grateful to Iratxe Urreta, Sonia Suárez-Álvarez and Carolina Da Rocha Carvalho for their technical assistance.

Data availability
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.