Embryonal Masses Induced at High Temperatures in Aleppo Pine: Cytokinin Profile and Cytological Characterization

Aleppo pine (Pinus halepensis Mill.), a native species of the Mediterranean region, has been suggested as a species that when introduced in degraded areas could facilitate the long-term colonization and expansion of late-successional species. Due to climate changes, plants need to withstand extreme environmental conditions through adaptation and changings in developmental pathways. Among other paths, plants undergo changes in developmental pathways controlled by phytohormones. At the same time, somatic embryogenesis has been widely used as a model to understand the mechanisms involved in plant response to different stresses. In this study, in order to induce a strong effect of temperature stress on plants regenerated from somatic embryos, higher temperatures (40 ◦C for 4 h, 50 ◦C for 30 min, and 60 ◦C for 5 min) than the control (23 ◦C) were applied during the induction stage of somatic embryogenesis in Pinus halepensis. A morphological characterization of the embryogenic cultures showed small differences in the number of starch grains, lipid bodies, and phenolic compounds between treatments. Results showed that high temperatures (60 ◦C) led to higher rates at the maturation stage of somatic embryogenesis when compared to the control (23 ◦C), strengthening the productivity through the increase in the number of somatic embryos obtained. Finally, analysis of endogenous concentration of cytokinins showed that different conditions applied during the initiation phase of somatic embryogenesis led to different hormonal profiles; isoprenoid cytokinins showed a clear defined pattern with the higher total hormone concentration being found in embryonal masses induced at 50 ◦C for 30 min, while different aromatic cytokinins presented different individual responses to the treatments applied. These differences corroborate the idea that cytokinins could be potential regulators of stress–response processes during initial steps of somatic embryogenesis.

Initiation medium was DCR [33] supplemented with 30 g L −1 sucrose and 3.5 g L −1 gellan gum (Gelrite ® , Duchefa Biochemie, Amsterdam, Netherlands), with a combination of 9.0 µM 2,4-dichlorophenoxyacetic acid and 2.7 µM kinetin, at pH 5.7. After autoclaving, a filter-sterilized solution containing EDM6 amino acid mixture [34] was added to the cooled medium. For temperature treatments, sealed Petri dishes containing initiation medium were preheated for 30 min. For the induction of embryonal masses (EMs) the procedure described in [28] was followed. In brief, immature megagametophytes were cultured at 40, 50, and 60 • C for 4 h, 30 min, and 5 min, respectively. As control conditions, 23 • C were used. After the application of the different treatments, all the megagametophytes were kept at 23 • C for 9 weeks in darkness. Four to eight megagametophytes per 26 Petri dish per treatment were cultured.

Proliferation of EMs
After 9 weeks on the initiation medium, proliferating EMs with an approximate diameter of 16 mm were detached from the megagametophyte. Before the first subculture on the proliferation medium, fresh tissues from all treatments were immersed in liquid nitrogen and immediately stored at −80 • C for further analysis. Proliferation medium had the same composition to that used in the initiation stage, but a higher gellan gum concentration (4.5 g L −1 ). EMs were subcultured every 2 weeks and kept in the dark. Following three subcultures, actively growing EMs were recorded as established embryogenic cell lines (ECLs). Proliferation and the subsequent stages of SE process were carried out at the same temperature (23 ± 2 • C).

Maturation of ECLs
Maturation of SES was carried out following the procedure described in [28]. Thus, after 4 subcultures on the proliferation medium, 75 mg of EMs per dish from different treatments were used. Maturation medium was the DCR medium supplemented with 60 g L −1 sucrose, 75.0 µM abscisic acid, the EDM6 amino acid mixture [34] and 9 g L −1 Gelrite ® . Eight or nine ECLs per treatment and eight Petri dishes per ECL were cultured in the dark for 18 weeks.
To analyze their morphology 160 of the produced SES, 10 from four ECLs per temperature treatment, were selected. For this purpose, two measurements were made using a Leica DMS 1000 microscope and the LAS V4.12 (Leica Application Suite, Wetzlar, Germany) software: the total length of the SES and the width, measured just below the cotyledonary intersection ( Figure 1).

Somatic Embryo Germination
SES were transferred to half-strength macronutrients LP medium [35,36] supplemented with 2 g L −1 of activated charcoal and 9.5 g L −1 Difco ® granulated agar (Becton Dickinson, Franklin Lakes, USA). Twenty embryos per Petri dish and four Petri dishes per 8 ECLs of each treatment were tested. Cultures were placed partially covered for 7 days, and, afterwards were kept under a 16:8 h photoperiod at 100 µ mol m −2 s −1 provided by cool white fluorescent tubes (TFL 58 W/33, Philips, France). The obtained plantlets were subcultured onto fresh medium of the same composition every 4 weeks.

Micromorphological Study
EMs at the end of the initiation phase, right before the first subculture on the proliferation medium, were used for light microscopy analyses following the procedure described in [37]. Two ECLs per treatment, comprising a total of 8 ECLs were analyzed. Samples of 3-5 mm in diameter were collected and fixed in formaldehyde 2.5% (v/v) in 0.1 M sodium phosphate buffer (pH 7.2) overnight at 4 °C. The material was washed twice for 15 min in buffer without fixative and then dehydrated in an increasing series of ethanol aqueous solutions (30-100% v/v), comprising a total of six different solutions, for 30 min. Samples were then infiltrated with Historesin (Leica Historesin, Heidelberg, Germany) and sections of 5 µ m were obtained using a rotatory microtome (Slee Technik, Mainz, Germany). After adhesion to histological slides, sections were stained with 1% (w/v) toluidine blue in an aqueous solution of 1% Borax, pH 9. Samples were analyzed and photographed using an Olympus BX 40 microscope equipped with a computer-controlled Olympus DP 71 digital camera (Olympus, Tokyo, Japan).

Somatic Embryo Germination
SES were transferred to half-strength macronutrients LP medium [35,36] supplemented with 2 g L −1 of activated charcoal and 9.5 g L −1 Difco ® granulated agar (Becton Dickinson, Franklin Lakes, USA). Twenty embryos per Petri dish and four Petri dishes per 8 ECLs of each treatment were tested. Cultures were placed partially covered for 7 days, and, afterwards were kept under a 16:8 h photoperiod at 100 µmol m −2 s −1 provided by cool white fluorescent tubes (TFL 58 W/33, Philips, France). The obtained plantlets were subcultured onto fresh medium of the same composition every 4 weeks.

Micromorphological Study
EMs at the end of the initiation phase, right before the first subculture on the proliferation medium, were used for light microscopy analyses following the procedure described in [37]. Two ECLs per treatment, comprising a total of 8 ECLs were analyzed. Samples of 3-5 mm in diameter were collected and fixed in formaldehyde 2.5% (v/v) in 0.1 M sodium phosphate buffer (pH 7.2) overnight at 4 • C. The material was washed twice for 15 min in buffer without fixative and then dehydrated in an increasing series of ethanol aqueous solutions (30-100% v/v), comprising a total of six different solutions, for 30 min. Samples were then infiltrated with Historesin (Leica Historesin, Heidelberg, Germany) and sections of 5 µm were obtained using a rotatory microtome (Slee Technik, Mainz, Germany). After adhesion to histological slides, sections were stained with 1% (w/v) toluidine blue in an aqueous solution of 1% Borax, pH 9. Samples were analyzed and photographed using an Olympus BX 40 microscope equipped with a computer-controlled Olympus DP 71 digital camera (Olympus, Tokyo, Japan).

Ultrastructural Analysis
Part of the ECLs used for the micromorphological studies were also subjected to transmission electron microscopy (TEM) analysis. Samples of 3-5 mm in diameter were fixed in 2.5% (v/v) glutaraldehyde in 0.1% (w/v) sodium cacodylate buffer and 0.6% (w/v) sucrose overnight. After five washing steps of 20 min, with decreasing concentrations of 0.1% (w/v) sodium cacodylate buffer (1; 0.75; 0.50; 0.25; 0) and increasing concentrations of 0.6% sucrose (w/v) (1; 1.25; 1.50; 1.75; 2) in 2.5% (v/v) glutaraldehyde, the samples were post-fixed using 1% (w/v) osmium tetroxide prepared in 0.1 M sodium cacodylate for 4 h. Samples were washed again three times with the same buffer and dehydrated in an increasing series of acetone aqueous solutions from 30 to 100% (v/v) following the same procedure described above for light microscopy assays. Finally, the material was embedded in Spurr's resin [38] and ultrathin sections (60 nm) were collected and contrasted on grinds using aqueous uranyl acetate followed by lead citrate [39]. The samples were then examined under a TEM JEM 1011 electron microscope (JEOL Ltd., Tokyo, Japan, at 80 kV).
Two technical replicates of 10 mg per ECL were analyzed, using miniaturized purification (pipette tip solid-phase extraction), according to the protocol described by [40]. Samples were extracted in 1 mL of modified Bieleski solvent and homogenized using a MM 301 vibration mill (Retsch GmbH & Co. KG, Haan, Germany) (27 Hz, 5 min, 4 • C) after addition of 3 zirconium oxide beads. Samples were extracted with the addition of stable isotope-labeled internal standards (0.2 pmol for base, ribosides, 9and 7-glucoside CKs and 0.5 for O-glucoside and CK nucleotides). The extracts were ultrasonicated for 3 min and incubated at 4 • C with continuous shaking for 30 min at 20 rpm. After centrifugation (15 min, 20,000 rpm, 4 • C) from the supernatants of each sample, another 3 technical replicates of 300 µL per sample were transferred onto Stage Tips and purified according to the aforementioned protocol, with C18, SDB-RPS, and Cation-SR sorbents.
Previously to the loading of the sample the StageTip sorbents were conditioned with 50 µL acetone (by centrifugation at 2,000 rpm, 10   Mass analysis was carried out following the procedure described by [24], using an Acquity UPLC ® System and a triple-quadrupole mass spectrometer XevoTM TQ-S MS (Waters MS Technologies, Manchester, United Kingdom). All MS data were processed using the MassLynxTM software with TargetLynxTM program (version 4.2., Waters, Milford, USA), and compounds were quantified by standard isotope dilution analysis [42].

Data Collection and Statistical Analysis
After 7-9 weeks on initiation medium, the number of initiated EMs per Petri dish was registered and the initiation percentages were calculated. Following three subcultures, actively growing EMs were recorded as ECLs and the percentage of proliferation respect to the EMs initiated was calculated. After 18 weeks from the beginning of the maturation stage, the number of ECLs able to form SES and the number of mature somatic embryos produced per ECLs were registered. The percentage of mature SES per gram of ECLs and morphology data were also assessed. In the second subculture on germination medium, the number of germinated SES was evaluated to calculate germination rates.
A one-way analysis of variance was carried out to assess the effect of temperature on SE stages percentages, mature SES produced per gram as well as their morphology data, and different concentration of endogenous cytokinins produced (GraphPad Prism 8.4.1 (676)). Following confirmation of the homogeneity of variances and normality of the samples an ANOVA was made. Whenever the analysis of variance did not fulfill the normality hypothesis, the corresponding non-parametric test, Kruskal-Wallis test, was applied. When significant differences were found (p < 0.05), the Tukey HSD post hoc test or Dunn's multiple comparison test, respectively, were carried out to find out which treatments were statistically different.

Temperature Experiment
Induction at the control temperature (23 • C) presented the lowest initiation rate. The highest initiation ( Figure 2a) and proliferation percentages were achieved when megagametophytes were induced at 40 • C and the lowest proliferation rate was obtained at EMs (Figure 2b) induced at 60 • C for 5 min. Despite the fact that there were no significant differences between treatments at initiation and proliferation rates (Tables 1 and 2), a decreasing pattern could be seen as the temperature rose along the high induction temperatures applied during initiation.     Concerning the effect of different temperatures during initiation on somatic embryo maturation all ECLs transferred to maturation conditions were able to produce somatic embryos ( Table 2).
No statistically significant differences were found for initiation, proliferation, or maturation capacity. Nevertheless, statistically significant differences were found when the number of SES per gram were analyzed, with the treatment of 60 • C for 5 min presenting the highest values ( Table 2), regardless of the fact that this treatment showed lowered rates of initiation when compared to other treatments. Control temperature showed an intermediate value ( Table 2). When the morphological aspects of SES produced were analyzed (Figure 1), the results showed that those produced from the ECLs initiated under control (23 • C) temperatures and at the highest temperature tested (60 • C) were significantly longer compared to somatic embryos resulting from other treatments. Regarding the ratio between length and width, SES from EMs induced at 60 • C for 5 min were significantly more elongated than those from EMs induced at 40 • C for 4 h and 50 • C for 30 min that presented a more barrel-shape form (Tables 3 and 4). Table 3. One-way analysis of variance for length, width, and ratio between length and width of SES produced from P. halepensis megagametophytes induced under different temperature treatments (23 • C, 9 weeks; 40 • C, 4 h; 50 • C, 30 min; 60 • C, 5 min).  No significant differences were found for the percentage of germination ( Figure 2c) ( Table 1). The highest germination rates were achieved in SES from ECLs initiated at 50 • C for 30 min followed by 60 • C for 5 min (Table 2).

Temperature (T)
In contrast with somatic embryos length and the ratio between length and width, no significant differences could be found between SES width coming from different temperature treatments (Table 4).

Cytological Characterization of Embryogenic Cultures
Three cell types were revealed in the embryogenic cultures analyzed by light microscopy: (1) Small round-shaped embryogenic cells (ECs) strongly stained and displaying a dense cytoplasm, a high ratio nucleus/cytoplasm, a conspicuous nucleoli and where mitotic figures were often seen; (2) highly vacuolated elongated suspensor cell (SCs), which have a reduced cytoplasm between the cell membrane and the tonoplast and (3) tube-like cells (TLCs) that seemed to be in transition between ECs and SCs sharing characteristics of both cell types (Figure 3a-e). All three types of cells could be found in all samples analyzed. These different types of cells usually appeared in clusters forming proembryogenic masses (PEM).
high ratio nucleus/cytoplasm, a conspicuous nucleoli and where mitotic figures were often seen; (2) highly vacuolated elongated suspensor cell (SCs), which have a reduced cytoplasm between the cell membrane and the tonoplast and (3) tube-like cells (TLCs) that seemed to be in transition between ECs and SCs sharing characteristics of both cell types (Figure 3a-e). All three types of cells could be found in all samples analyzed. These different types of cells usually appeared in clusters forming proembryogenic masses (PEM).  Two different stages of development were found: PEM I (Figure 3a) with smalls groups of ECs associated with one or two SCs; and PEM II, with a higher number of ECs and SCs in the cluster (Figure 3b). No PEM III, which consists of larger clusters of ECs and SCs, were found.
Despite the organized and clear cell polarization of PEM II found in some sections ( Figure 3b) the majority of PEMs found in the analyzed samples showed poor cellular organization. However, although all three types of cells were present (ECs, SCs, and TLCs), all of them tangled together without defined polarization and an unbalanced proportion between embryonal areas and suspensors, with a big number of TLCs and SCs, was often observed (Figure 3c,e). Both polarized and non-polarized areas were detected for all treatments, therefore, it seems that there are not differences respect to cellular organization between treatments.
Finally, a large number of starch grains (S) on ECs on samples from all treatments (Figure 3c,d) and some phenolic compounds (PCs) were found only in one sample from an embryogenic culture induced at 60 • C for 5 min (Figure 3e   ECs were characterized by a dense cytoplasm and a large central nucleus containing one or more nucleolus. Many cytoplasmic organelles, indicatory of intense metabolic activity such as mitochondria, Golgi complex, and PVs were found around the nucleus (Figure 4a-c). Quite common was the presence of vacuoles containing phenolic compounds in ECs (Figure 4d,e) in samples from different temperatures. These polyphenolic-containing vacuoles were particularly abundant in samples induced at 50 • C for 30 min. On the contrary, PCs could not be found in explants cultured under control temperatures. These observations confirm those obtained with light microscopy where PCs where only found in cultures kept at 60 • C for 5 min.
Concerning the starch found by light microscopy in samples from all treatments, TEM analysis displayed the same results. A great amount of starch was found in all treatments in ECs (Figure 4a,f) and some could also be found in TLCs (Figure 4f). It can also be noted that, contrary to light microscopy where no evident differences were clear between treatments, TEM allowed the identification of a higher number of starch in samples from cultures initiated at the control temperature (23 • C). A great amount of lipid bodies (LB) were also detected on samples from all treatments (Figure 4e) particularly in samples cultured under the lowest temperature (control).
Regarding TLCs, that seem to be in transition between ECs and SCs, cells were more elongated and had a higher number of PVs when comparing to ECs (Figure 4g). It appeared that by progressive destruction of the cytosol and organelles, PVs were formed on the cytoplasm and together started to form a large central vacuole (V). SCs appeared as PVs started rising in number toward the cell periphery and converged to the formation of a large central V that occupied the majority of cell volume, restraining the cytoplasm to a narrow layer confined between tonoplast and plasma membrane (Figure 4i). No clear differences among treatments were found for TLCs or PVs.

Endogenous Cytokinins Quantification
Statistically significant differences with a very defined pattern were found concerning the concentrations of total isoprenoids CKs, CKs bases, CKs ribosides, and CKs nucleotides (Table 5), Figure 5). Samples coming from control conditions showed the lowest values in all the CKs groups analyzed. On the contrary, tissues coming from EMs induced at 50 • C for 30 min contained a significantly higher concentration compared to the control. The other treatments presented intermediate values (Figure 5a). The same results were obtained for CKs ribosides (Figure 5b). Samples coming from EMs initiated at 50 • C for 30 min showed significantly higher concentration of CKs bases when compared to all treatments (Figure 5c). In the same way, CKs nucleotides concentrations were significantly higher in samples coming from 50 • C treatment (Figure 5d). It should also be noted that this was the functional group that presented the highest concentration of isoprenoid endogenous CKs.  When the total contents of different isoprenoids CKs types were analyzed, the results showed the same pattern observed for isoprenoids functional groups (Figure 6a-d).
cZ types showed significant differences between samples from all treatments and the control (Figure 6a). EMs induced at 50 °C for 30 min contained amounts significantly higher of tZ, DHZ, and iP types than the other treatments assayed (Figure 6b-d). When the total contents of different isoprenoids CKs types were analyzed, the results showed the same pattern observed for isoprenoids functional groups (Figure 6a-d).  cZ types showed significant differences between samples from all treatments and the control (Figure 6a). EMs induced at 50 • C for 30 min contained amounts significantly higher of tZ, DHZ, and iP types than the other treatments assayed (Figure 6b-d).
Examining the content of each isoprenoid CKs individually, cZ and iP concentrations in samples from 50 • C showed statistically significant differences from other treatments, apart from 40 • C that presented intermediate values (Figure 7a,b). In tZ and DHZRMP mean values of samples from 50 • C treatment resulted in the highest values of cytokinins (Figure 7c,d). cZRMP and tZRMP presented higher concentrations in samples cultured at highest temperatures (50 and 60 • C) and DHZR showed the same pattern as cZ and iP (Figure 7e-g). The ratios between CK bases and CK nucleotides (Figure 8a), between total DHZ and cZ types ( Figure 8b) and between [bases + ribosides] and nucleotides ( Figure 8c) presented a decreasing pattern as the temperature increased, with samples from the control presenting a statistically significant higher ratio than the ones from 60 °C treatment. Contrary, the ratio between cZ and tZ types showed to be statistically significant higher in samples from higher temperatures comparing to the control (Figure 8d). The ratios between CK bases and CK nucleotides (Figure 8a), between total DHZ and cZ types ( Figure 8b) and between [bases + ribosides] and nucleotides ( Figure 8c) presented a decreasing pattern as the temperature increased, with samples from the control presenting a statistically significant higher ratio than the ones from 60 • C treatment. Contrary, the ratio between cZ and tZ types showed to be statistically significant higher in samples from higher temperatures comparing to the control (Figure 8d).
The majority of CKs N-glucosides (7-G/9-G) and O-glucosides were under the limit of detection in samples from all treatments applied. No statistically significant differences were found neither for cZR, tZR, DHZ, DHZR, iPR, and iPMP, nor for the ratios between CK bases and ribosides, iP, and Z types and tZ and DHZ (Table 6). The majority of CKs N-glucosides (7-G/9-G) and O-glucosides were under the limit of detection in samples from all treatments applied. No statistically significant differences were found neither for cZR, tZR, DHZ, DHZR, iPR, and iPMP, nor for the ratios between CK bases and ribosides, iP, and Z types and tZ and DHZ (Table 6). When the effect of different induction temperatures in the endogenous concentrations of aromatic CKs was analyzed, statistically significant differences were found concerning the concentrations of BA and K types, as well as BA, Kn, K9G, and oT ( Table 7). Concentrations of BA types in samples coming from 40 °C treatment were significantly lower than samples from the control (Figure 9a). BA concentrations showed the same pattern, and samples coming from 50 and 60 °C treatment presented intermediate values (Figure 9b). K types and Kn mean values obtained in  When the effect of different induction temperatures in the endogenous concentrations of aromatic CKs was analyzed, statistically significant differences were found concerning the concentrations of BA and K types, as well as BA, Kn, K9G, and oT ( Table 7). Concentrations of BA types in samples coming from 40 • C treatment were significantly lower than samples from the control (Figure 9a). BA concentrations showed the same pattern, and samples coming from 50 and 60 • C treatment presented intermediate values (Figure 9b). K types and Kn mean values obtained in samples cultured under 50 • C were significantly higher when compared to the other treatments. It should be noted that the Kn presented, by far, the higher concentration for all treatments (Figure 9c,d). K9G concentration was the highest in samples coming from 60 • C initiation treatment (Figure 9e). oT concentrations in samples cultured at the control treatment were significantly lower when compared to higher temperatures (Figure 9f). samples cultured under 50 °C were significantly higher when compared to the other treatments. It should be noted that the Kn presented, by far, the higher concentration for all treatments (Figure  9c,d). K9G concentration was the highest in samples coming from 60 °C initiation treatment ( Figure  9e). oT concentrations in samples cultured at the control treatment were significantly lower when compared to higher temperatures (Figure 9f).  The majority of aromatic CKs N-glucosides (7-G/9-G) (apart from K9G), BARMP, oTR and all of pT types were under limit of detection in all samples from different treatments. Finally, it is worth to mention that statistically significant differences were not found for BAR, total mT types, mT, mTR or KR (Table 8). The majority of aromatic CKs N-glucosides (7-G/9-G) (apart from K9G), BARMP, oTR and all of pT types were under limit of detection in all samples from different treatments. Finally, it is worth to mention that statistically significant differences were not found for BAR, total mT types, mT, mTR or KR (Table 8).

Temperature Experiment
In the present study temperatures higher than used in standard routine were applied (40, 50, and 60 • C) during the initiation stage of P. halepensis SE for different periods (4 h, 30 min, and 5 min, respectively). No statistically significant differences could be found between initiation, proliferation, maturation or germination rates obtained in the different treatments assayed during the induction of EMs. These results are in agreement with the ones obtained when the same temperatures and induction times were studied in P. radiata [43] where no significant differences were also found. Nonetheless, the control presented a lower initiation rate when compared to higher temperatures, in contrast with the results obtained in previous studies carried out in our laboratory [29].
Also, the ECLs initiated under different temperatures did not show significant differences regarding their capacity to produce SES, corroborating the results found in Pinus spp. previous studies [29,43]. However, samples induced at 60 • C for 5 min showed significantly higher number of somatic embryos, doubling the number of embryos produced in samples cultured at 40 or 50 • C. Similar results have been obtained in P. radiata, where although not statistically significant differences were found, the highest embryo production was achieved at 60 • C [43]. Similar experiments carried out in P. pinaster [44] showed that an increase of temperature up to 28 • C resulted on initiation rates similar to those observed in standard conditions but significantly higher number of SES were produced. This fact reinforce the idea that temperature exerts a selective pressure in the SE initial stages, that can result in an initiation decrease but higher rates for the forthcoming steps [45,46].
It seems that the temperature treatment that triggered a deeper effect on the efficiency of P. halepensis SE, was 60 • C for 5 min. In spite of the association of stress with adaptation mechanisms and the induction of SES formation, cell responses to the applied stresses depends on its intensity [47]. The different induction times for each temperature could in a way affect samples response to the different high temperatures applied.
Regarding the morphological analysis of the SES obtained, more elongated embryos were obtained from samples induced at 60 • C for 5 min and from the control treatment, while the ones obtained from 40 • C for 4 h or 50 • C for 30 min were smaller and more barrel-shaped. Barrel-shaped form has been described as a symptom of poor quality embryos, which is usually accompanied by low germination rates in P. taeda [48], P. Pinaster [49], and P. radiata [43]. This is contrary to our results since no significant differences were found in the percentage of germinating somatic embryos, and, in fact, the highest percentage of germination was achieved in SES from ECLs initiated at 50 • C for 30 min. These results indicate that the shape of P. halepensis SES may not be related with their ability to germinate as long as they are well formed.

Cytological Characterization of Embryogenic Cultures
Previous work on Picea abies [50] showed that embryogenic cultures appeared in three distinct levels or organization: PEM I, PEM II, and PEM III, formed by three different cell types (ECs, TLCs, and SCs) attached together in clusters in different proportions.
In Aleppo pine three different types of cells (ECs, TLCs, and SCs) in embryogenic cultures from all different temperature treatments (23,40,50, and 60 • C) were also found. Nonetheless, only two of the organizational structures described by [50] could be found, since no PEM III were present. In Picea abies [50] PEM III were described as enlarged clumps of densely cytoplasmic cells loosely attached to each other that do not present polarity while in the case of P. radiata these cells formed compact clusters which showed a well-organized structure with a clear polarization in samples from the control and a higher disorganization at samples from higher temperatures treatments [43]. Also, despite the fact that well organized and polarized PEMs could be observed from different treatments, the common scenario featured high cellular disorganization with a disproportionate number of TLCs and SCs and no apparent discrepancies in the analyzed samples could be found between treatments when it comes to these features. In P. sylvestris [51] a continuous loop of embryo degeneration and differentiation of new embryos during the initiation of embryogenic tissue was suggested. However, they could not find if the differentiation pattern during the initiation stage would give rise to differences in the embryos obtained. This can explain the higher number of TLCs and SCs found in our samples and the fact that all of the analyzed embryogenic cultures produced somatic embryos despite their disorganization A large number of starch was found both in light microscopy and TEM analysis from all treatments. In fact, TEM analysis showed that starch grains were present both in ECs and TLCs and in higher number in samples from the control. Some authors support the idea that, in ECs, the accumulation of starch around the nucleus is a signal of loss of embryogenic competence and it is also linked with the appearance of dead cells in meristematic centers, leading to low initiation and proliferation rates [43,52]. This was not the case in P. halepensis, since no differences were found for initiation or proliferation rates between treatments. Other authors pointed out that starch, as a primary source of energy for cell proliferation, is abundantly found at competent embryogenic cells that will later differentiate into PEMs. High metabolic activity of these cells will deplete this reserve in such a way that it will no longer be present when embryos start to organize [53,54]. In Picea glauca a precise pattern of storage compound accumulation during SE was identified with the first accumulation of starch further followed by the formation of lipid bodies and protein vacuoles [55]. In P. halepensis lipid bodies were also found in all treatments, although more abundantly in the control. Numerous protein bodies of different sizes or lipid bodies were also observed in ECs of other species [54,56]. In the case of P. halepensis, the presence of the above mentioned storage compounds could not be directly correlated with the progress of SE and their function could probably be related with house-keeping functions of the cells more than with a control of the embryogenic process.
Finally, PCs were found in cultures induced at the higher temperatures. These compounds started to accumulate in the internal face of membrane vacuoles and were particularly notorious when cultures were initiated at 50 • C. The accumulation of this type of secondary metabolites has been described as playing an essential role preventing oxidative damage caused by different types of stresses [27,57], explaining their appearance in samples coming from high temperature treatments and not in control ones. In P. radiata differences were also seen among different treatments respect to PCs with larger accumulations at higher temperatures [43]. Also, opposite to our results, a positive association between accumulation of phenolic compounds and somatic embryo formation has been indicated in Acca sellowiana [58].

Quantification of Endogenous Cytokinins
A significantly higher concentration of isoprenoid CKs is present at EMs induced at 50 • C when compared to the control. Samples from 40 and 60 • C treatments presented concentrations that could be similar to one or another, but were generally intermediate. This pattern was detected through the analysis of the endogenous content of total CKs, CK ribosides, CK nucleotides, cZ, tZ and DHZ types, and finally for cZ, tZ, iP, cZRMP, tZRMP, DHZRMP, and DHZR. The results obtained in P. halepensis are in line with the concept that short-term or mild stresses stimulate CKs accumulation, while prolonged or more severe stresses are generally associated with downregulation of active CKs levels [59][60][61]. Furthermore, the ribosylated form of the initial precursor of CKs biosynthesis, iPA, across the time of heat stress exposure in P. radiata showed an increase as other author pointed out [62]. Contrary, when the same temperatures and induction times were analyzed in P. radiata, the results showed a decreasing tendency when applying high temperatures for isoprenoid CKs, especially for those applied for the longest periods of time (40 and 50 • C) [63].
Although the physiological function of each individual CKs is not yet completely understood they can be classified into different functional groups: (1) Active forms (bases), (2) translocation forms (ribosides), (3) precursors (nucleotides), and (4) storage and inactivated forms usually bound to glycosides (N-glucosides (7-G/9-G) and O-glucosides) [22,64]. Regarding the analysis of different isoprenoid functional groups, the results showed that the highest concentration was obtained for CK nucleotides. As reviewed by [19], it has been demonstrated that exogenously applied free cytokinin bases are rapidly metabolized into the corresponding ribosides and nucleotides and that, in general, high concentrations of these are found in developing organs. In this sense, when the same induction conditions were studied in P. radiata EMs a higher concentration of CKs bases was found [63], while at mature SES of the same species CKs ribosides showed the highest concentration in general [24].
The majority of N-glucosides (7-G/9-G) and O-glucosides, both for isoprenoids and aromatic CKs, were below the limit of detection. In terms of physiological activity, they exhibit little or no activity in CK bioassays [65]. These data are in agreement with those reported for P. radiata when high temperatures were analyzed [63], and opposite to those obtained when different water availability and temperatures were tested at maturation stage of P. radiata SE, despite the low values obtained [24], or when different CKs were examined on the induction of organogenesis [66]. The latter authors hypothesized that this type of forms are normally used by the plant as a form of detoxification in the face of excess hormone. Following this hypothesis, we can say that, in our case, the endogenous content of the hormone is not such that it needs this detoxification mechanism. Data obtained with Norway spruce showed that CKs O-glucosides are the most abundant group during all stages of the SE process, including at callus stage, whereas CKs N-glucosides were practically absent [67]. This can be explained by the difference in the tissues and species analyzed, and the fact that these metabolites have distinctive importance at different stages of SE in these species.
When endogenous concentrations of aromatic CKs were analyzed, a different pattern was obtained when compared to isoprenoid CKs. In K types, despite the fact that samples from 50 • C showed the higher concentrations, 40 • C showed a significantly lower concentration and, in BA types, samples from 40 • C showed significantly lower concentrations when compared with samples from the control. It seems that, when it comes to BA and K types, 40 • C for 4 h acted more as a severe stress, contrary to its effect on isoprenoids CKs [59][60][61]. Also, our results show that the different temperature treatments did not lead to a clear accumulation of these hormones. Reports in apple trees [68], where concentration of Kn tended to reduce with the intensity of water deficit, and in soybean [69], where its growth under salt stress was significantly promoted by elevated Kn levels that effectively improved the quantities of isoflavones presented a more clear function of Kn in response to stress.
Kinetin showed to be the most abundant cytokinin in our samples which might be related to its exogenous application in the culture media. Despite the fact that the majority of naturally occurring CKs in plants are the N 6 -isopentenyl conjugated adenine derivatives, in addition to a small amount of N 6 -aromatic CK species [64,70], exogenous addition of CKs to culture media is known to have an effect by altering and increasing the endogenous CKs pool [66,71]. Similarly, reports in Solanum tuberosum [72], Pinus pinea [73], and Pinus radiata [66] found that aromatic CKs accounted for more than 90% of the total endogenous CK pool although isoprenoid forms are generally the most abundant CKs in plant materials that have not been hormonally treated.
High temperatures application overlap with higher endogenous concentrations of oT, cZ types and higher ratios between cZ types and tZ types. Within the CKs identified in response to mistletoe infestation in P. sylvestris, the levels of biologically highly active (tZ, iP) and less active (cZ, DHZ) CKs were also found to be generally enhanced [74]. It should be noted that, despite of the fact that there were no significant differences, the ratio between iP types and Z types also decreased as the temperatures applied rose. Z-type cytokinins are derived from iP-type compounds and not vice versa [22,60,64] and both cell fate and organ formation have been associated with local concentration gradients of these hormones [75]. Despite the fact that oT is present in inferior concentrations comparing to mT (Figure 8; Table 7) it seems to have a stronger effect in stress response as well as in cellular activity since mT did not presented significant variations. Contrary, the results found in Pinus pinea [73] suggest that a higher cellular activity that lead to a stronger callogenesis response was related to higher concentrations of mT as well as BAR and KR that also did not presented significant differences in our case.
Apart from their influence on stress response per se, differences obtained in CKs profiles can explain some of the behaviors seen throughout different SE stages between the different treatments applied. After all, CKs are known to be responsible for the regulation of a number of aspects of plant growth and development, such as cell differentiation and growth [25].
Although there were also no statistically significant differences at proliferation and germination stages, high concentrations of tZ types, tZRMP and iPR of samples from 50 and 60 • C treatments overlays with the highest germination rate achieved during SE, at the same time that the treatments with lower concentrations of these hormones presented lower proliferation rates. In coffee it was reported that the transition to the callus stage was accompanied by significant decrease of Z-type (cis-and trans-) [76]. Experiments carried out in P. pinea [77] indicated that tZR, tZ, and DHZ had a relevant role in the callogenic process. Also, in P. radiata [62] ZR seems to be a key element to further acclimation and the recovery mechanism in plants while iPR and DHZR seems to be highly implied, across time, at the heat stress response.
It must be stressed out that samples from 60 • C and control treatment (23 • C) had a similar endogenous low hormonal concentration of cZ, iP and DHZR, and these are the treatments that were able to induce a significantly higher number of somatic embryos. It was observed in [78] that high concentrations of Z and active CKs in general are important during the initial cell division phase of somatic embryogenesis, but not for the later stages of embryo development and maturation. However, as reported in P. radiata, low levels of iP types leads to higher success in the SE process [24,63]. Additionally, the only CK N-glucoside found in our samples, K9G, presented a significantly higher concentration at 60 • C, the treatment that produced more somatic embryos. In Cocos nucifera [79] the isoprenoid cytokinin profiles showed a predominant pattern of 9-conjugation as a major metabolism route and the detection of high levels of inactive metabolic products has been cited as evidence of high CK turnover.

Conclusions
As far as we know, this is the first study developed in Aleppo pine that focus in the initiation of embryonal masses under high temperatures analyzing both the phytohormones involved in the success of the SE process as well as the cytological characterization of embryogenic cultures. We found that it is possible to improve the efficiency of the somatic embryogenesis process, under high temperatures, strengthening its productivity through the significant increase of somatic embryos obtained at samples induced at 60 • C for 5 min. Different hormonal profiles overlap with the application of temperature stress as lower endogenous concentrations of oT and cZ types cytokinins were found at control. Also, a correlation between high rates at specific steps of the process and higher concentrations of tZ types, tZRMP, and iPR led to higher germination rates.
Further investigations on the effect of temperature stress during induction phase of P. halepensis SE should focus on complementary molecular pathways, as well as the epigenetics involved along the process. Also, the analysis of physiological parameters in somatic embryogenesis-derived plants