The plant growth curve is the basic parameter for a plant physiologist and can be determined using several different methods. The most frequently used methods are based on the counting and/or weighing of cells or tissues. In recent years, image analysis (IA) has been used as a tool to study the growth and/or morphology and also for the study of somatic embryogenesis e.g. Daucus carota [10,11,12], Betula pendula [13], Ipomoea batatas [14], Saccharum spp. [15] and Picea abies [16]. Therefore, we decided to use this method in this study, too. The ESE clusters cultivated on solid culture medium formed masses with whitish or in some cases slightly yellowish colouring. During the 14-day cultivation, the ESE cluster uniformly expanded to the neighbouring area. The typical structures of individual somatic embryos in the ESEs were observed under a light microscope by very carefully disintegrating the ESEs in phosphate buffer. Each embryo consisted of an embryonal group (very intensely dividing cells), embryonal tube cells and embryonal suspensor cells. Embryonal group cells continually divide, and the tubular cells are formed due to their elongation at the distal end. These cells form the embryonal suspensor, whose cells are released into the medium. The ESE cluster is shaped by the production of new tubular and suspensor cells and their segregation in the direction from the embryonal group [17]. The stress reaction increased after the detachment of 50–100 individual somatic embryos into the cultivation medium, which might account for the growth depression observed for 5–7 days. The growth rate increased from the 7^th to 14^th day, when very rapid ESE growth was observed. This very rapid growth lasted until the 35^th–40^th day of sub-cultivation (Fig 1). On the 40^th day after sub-cultivation, the ESE cluster growth slowly decreased (inset in Fig 1). These growth characteristics were observed using two different methods: computer image analysis and weighing of individual ESE clusters of the 2/32 clone (Fig 1). These growth characteristics were repeatedly validated (n = 3). This rapid growth is often detected in ESEs of various species [4,18].

The growth data were interesting and indicated the presence of distinct inner and outer ESE regions (Fig 2A). Significant changes were observed from the 35^th day after ESE sub-cultivation. Therefore, ESEs collected on the 56^th day after sub-cultivation were selected for experiments to evaluate whether the ESE clusters could be divided into two region to analyse their homogeneity. Individual ESEs were stained with 2,3,5-triphenyltetrazolium chloride (TTC, Fig 2B) and acetocarmine (Fig 2C). The outer cluster was coloured red after TTC staining (Fig 2B), while the inner cluster was unstained. TCC staining indicates changes in the metabolic state of the cell respiratory chain oxidative-reduction reactions. In contrast, acetocarmine stained the embryos of the inner cluster more intensely than the embryos of the outer cluster (Fig 2C), which was based on the measurement of the intensity of the signal. The maximal metabolic activity of the ESE cluster was detected on the surface and in the marginal regions near the ESE cluster base. Individual somatic embryos removed from these regions demonstrated the distinct metabolic activity of the embryonal group. These changes can be associated with changes in recently detected transcription factors [19] or various polyamine levels [20].Furthermore, the ESE cluster regions without metabolic activity (after removing individual somatic embryos) did not demonstrate this activity. Isolated regions of the ESE cluster were subsequently cultivated for 45 days (with sub-cultivations after 14 days) to obtain more information about their viability and metabolic activities. After 45 days of cultivation, the ESE clusters were divided into six parts (Fig 3A). Individual somatic embryos were isolated from each of these parts and subsequently tested for viability using the fluorescent probes fluorescein diacetate and propidium iodide. Nearly 75% of the individual somatic embryos from the inner ESE cluster remained unstained, whereas the number of unstained embryos in the outer ESE cluster was limited to 10%. The number of non-viable cells in each cluster was approximately 15%. These results were confirmed by FDA/PI staining, which demonstrated that individual somatic embryos from the outer ESE cluster demonstrated high metabolic activity and were FDA-positive (Fig 3B). FDA positive clusters were determined based on green fluorescence as it is shown in upper inset photo in Fig 3B. Dead cells were stained by red (bottom inset photo in Fig 3B), whereas unstained cells were identified by superimposing figures with those obtained by standard camera [17]. Average viability changes determined by the above-mentioned double FDA/PI staining and esterase activity determination according to cultivation are shown in Fig 3C and Fig 3D, respectively. There was a gradual deceleration of the ESE metabolic activity that was concomitant with the period of cultivation. This effect could have been caused by the gradual growth of FDA/PI unstainable cells. The potential factor responsible for the metabolic changes observed herein is not clear. Nonetheless, polysaccharide compounds could play a significant role in this process [21,22]. Moreover, there have been observed and documented cases, where the decrease in the metabolic activity of the clusters was associated with the external negative influence of metal ions [23]. Karcz and Kurtyka investigated the effect of cadmium(II) ions on growth, proton extrusion and membrane potential in maize coleoptile segments [24]. The cadmium(II) ions induced growth inhibition at almond seedlings [25], green microalga Scenedesmus armatus (Chod.) [26] and others [27,28].

This study attempted to describe the differences in the cluster structure and to answer the question of whether the inner ESE cluster structure was connected with the transport of different substances from the cultivation medium into the cluster. In particular, the concentration of ¹H nuclei was determined by NMR to assess this method of transport. Concentration of ¹H nuclei determined the concentration of water in ESEs, which was reversely used to find the content of dry matter as it is shown in Fig 4. Scans were taken of the weighted concentration of ¹H nuclei using MR imaging. A 2.0 mm thick incision was made in the selected sample position, and images were taken of a 30 × 30 mm (256 × 256 pixels) area with a resolution of 0.117 mm/pixel. The samples were cultivated in plastic Petri dishes (50 mm in diameter) for 56 days. Measurements were performed on days 0, 7, 14 and 56 (Fig 4). At the same time, the dry weight was also determined. On the basis of the known dry weight in the cultivation medium, we determined the chromatic scale, dry weight quantity and consequently the water level in each cluster region. The obtained data indicated that there was a higher ¹H nuclei concentration in the outer ESE cluster compared with the inner cluster. This difference became more pronounced with the increasing cultivation time. These changes are associated with the changes in water content, which could be of interest for further studies because the fact that partial drying of somatic embryos at high relative humidity or desiccation in the presence of supersaturated solutions of salt is used to improve the efficiency of germination. Hazubska-Przybyl et al. found that ESEs of Norway Spruce used in this study react on the heat stress and their behaviour differed from other ESEs [29]. During the first four weeks of cultivation, the ESE cluster seemed to be compactly aggregated; thereafter, the differences between the ¹H nuclei concentration in the inner and outer clusters were more evident. Additionally, embryos from both the outer and inner clusters appeared to be visibly different. The formation of a cavity was observed in the ESE clusters adjacent to the cultivation medium after more than 40 days of cultivation (Fig 4) and it is label with an asterisk. Although ESEs from the outer cluster were clearly distinguishable and the embryonal group was centrifugal-oriented, the inner parts of the cluster consisted of the mass formed by ESEs and polysaccharide [30,31,32]. The availability of the polysaccharide most likely inhibited the penetration of the fluorescent probes into the ESEs, thereby disabling the determination of ESE viability by double FDA/PI staining [30].

Based on the ESE viability determination, the embryos removed from individual ESE cluster regions cultivated for 45 days (6 parts similar to the viability determination) were placed into new cultivation medium and consequently cultivated for the next 45 days, with sub-cultivations after each two week period. This experiment aimed to determine the cluster area from separate regions of a six-week old cluster. The most evident differences between the individual parts of the cluster were noted during the first 14 days of cultivation (Fig 5A). The outer regions were bigger (over 59.4%) than the inner regions. These differences decreased with the next sub-cultivation. Four weeks after the first cultivation, this distinction between the outer and inner clusters was 31.2%, and six weeks after the first cultivation the difference was only 6.5%. These data clearly indicated that there were differences in the growth characteristics between the outer and inner regions of various locations in the primary cluster (Fig 5A). Each region grew at a particular rate. For example, the inner parts where the older embryos predominated had the ability to grow and develop. The time required to reach the primary growth rate depended on the region of the ESE removed from the primary ESE cluster. As mentioned above, the embryos that originated in the outermost region of the primary six-week old clusters grew almost as quickly as the embryos from the two-week old clusters (the common time of sub-cultivation derived from the growth characteristics; please see Fig 5A). Interestingly, there was an increase in all six regions removed from different locations of the primary cluster (three from the outer and three from inner regions of the primary cluster, photos of the representatives of ESEs are as insets in Fig 5A). It was also evident from the obtained data that the innermost embryos (centripetally towards the cluster centre) could grow after sub-cultivation, although they exhibited the slowest rate and required the longest time to reach the common growth rate. The differences between the individual inner and outer regions (according to Fig 3A) were very obvious and observable, especially during the first 14 days of cultivation. The most prominent differences were between the innermost and outermost regions from the primary cluster surface, where the difference in the growth increase was up to 250% after 14 days of the first cultivation, 35% after 28 days and 30.1% after 42 days. There were found differences in some surface structures in seed of Pinus sylvestris L. based on long term experiment with the aim on performing the experiments under various conditions [33]. Based on the obtained data, we could conclude that all embryos from the primary cluster were able to grow and develop despite the fact that vital embryonal cells from the inner regions of the primary cluster seemed to be nonvital due to their lack of double fluorescent FDA/PI staining. Vital (living) cells metabolise FDA to acetate and fluorescein, which is the proper fluorogenic substrate; in contrast, PI penetrates through damaged cytoplasmic membranes and then intercalates into the DNA, causing dead cells to fluoresce red [34]. This is very important because the lack of staining of the inner embryonal cells resulted in our inability to determine their viability despite the fact that these unstained cells were subsequently able to grow and develop.

Next, we attempted to determine the age of the primary clusters. By measuring the oxidative-reduction reaction by TTC, we ascertained that the inner part of the primary ESE cluster remained unstained 65 days after the first cultivation, indicating that the cells were dead. We performed the analogous experiment with ESE clusters after 45 days of cultivation when the individual clusters were separated into six different parts. We found differences in the increase in the growth area between the inner and outer regions of the primary clusters. The time points chosen for this experiment were 4, 8 and 12 days (Fig 5B). After 12 days of cultivation, the increase in the area of the new clusters removed from the outer primary cluster was increased by 86%, which corresponded to the area increase of the two-week clusters. Conversely, the area increase in the regions removed from the inner primary cluster was 7.7% after 12 days of cultivation. Thus, the embryonal cells originating from the inner regions of the ESE primary cluster grew very slowly after their transfer into new cultivation medium. These results indicate that the ESE characteristics change depending on the age (time of cultivation) of the ESEs. The results were further statistically treated and it was found that differences determined after 4 days long cultivation were significant (p < 0.05, labelled with an asterisk, Fig 5B).

The water level in different regions of the clusters was determined by employing the NMR technique and was found to be different. Therefore, we hypothesised that each cluster might have various methods for the transport of compounds from the cultivation medium into the cluster. Additionally, we determined the influence of the cluster orientation on the cultivation medium and the change of the three-dimensional orientation of the cluster on its development. We assumed that the change in the ESE cluster orientation would disrupt transport primarily in the older clusters due to the establishment of pathways. For the current experiment, two and six week old clusters were utilised; one half of these clusters were placed into the Petri dish with the same three-dimensional orientation, whereas the second half of the clusters were placed in an inverse orientation (upside down). The areas of the clusters were determined on days 3, 7, 10 and 14 (Fig 6). A difference in the area increase was detectable between the two- and six-week old clusters with the same three-dimensional orientation as the primary cluster (Fig 6A and Fig 6B). When we compared the cluster areas of 6 weeks old clusters, the differences between normal and upside-down were found significant from 7^th day of the cultivation (p < 0.05, labelled with an asterisk, Fig 6A). On the other hand, no significant difference was found in 2 weeks old clusters, which could show the differences in structure among two and six weeks old clusters. The area increase after two weeks was approximately 40% higher than the increase observed in the six-week old clusters, which was found significant. This difference was probably due to the diverse ontogenetic stages of the investigated clusters [35]. The regions removed from the older ontogenetic embryos demonstrated slower growth than the regions removed from the two-week-old ESE clusters. However, this experiment was aimed at the investigation of different three-dimensional orientations of clusters in new cultivation medium. The difference between two-week-old clusters with the same orientation and the inverse orientation was approximately 10% after 14 days of cultivation and 25% after 42 days of cultivation (Fig 6). Therefore, it may be necessary to keep ESEs in the same position when transferring them into new cultivation medium.

All reagents were of ACS purity and were purchased from Sigma Aldrich Chemical Corp. (USA) unless otherwise noted. Celulysin was purchased from Calbiochem (USA). All solutions were prepared using deionised water (18.2 MΩ, Iwa 20, Watek, Czech Republic). Culture media were prepared using plant cell culture chemicals purchased from Duchefa Biochemie BV (Haarlem, The Netherlands) unless otherwise indicated. The pH value was measured using the WTW inoLab Level 3 with a level 3 terminal (Weilheim, Germany) controlled by a personal computer program (MultiLab Pilot; Weilheim, Germany). The pH-electrode (SenTix-H, pH 0–14/3 M KCl) was regularly calibrated with a set of WTW buffers (Weilheim, Germany). Deionised water underwent demineralisation by reverse osmosis using the Aqua Osmotic 02 instrument (Aqua Osmotic, Tisnov, Czech Republic) and was subsequently purified using the Millipore RG (Millipore Corp., USA, 18 MΏ) MiliQ water system.

A culture of early somatic embryos (ESEs) of the Norway spruce (Picea abies (L.) Karst.) clone 2/32 was used in the current experiments. The clone was originally established at the Department of Plant Biology of Mendel University in Brno, Czech Republic, according to a procedure described by Durzan et al. [36]. The ESEs were maintained on a semisolid (Gelrite Gellan Gum, Merck, Germany) half-strength LP medium [37] modified by Havel [38,39]. The concentrations of 2,4-dichlorofenoxyacetic acid and N6 benzyladenine were 4.4 and 9 μM, respectively. The pH was adjusted to 5.7/5.8 prior to autoclaving (121°C, 100 kPa, 20 min). The organic part of the medium (except saccharose) was sterilised by filtration through a 0.2 μm polyethylensulfone membrane (Puradisc 25 AS, Whatman, USA). Briefly, ten ESE clusters were cultivated in one plastic Petri dish (100 mm in diameter) containing 30 ml of medium. Sub-cultivation of the stock cultures was performed at 2 week intervals. The stock and experimental cultures were maintained in a cultivation box in the dark at 23±2°C.

Tomography based on magnetic resonance was used to study changes in the ¹H nucleus spin density in ESEs. The experiments were performed on a 4.7 T/120 mm (i.e., 200 MHz for ¹H nuclei) MRI scanner (Magnex magnet, MR Solution electronics and software) at the Institute of Scientific Instruments, Brno. Actively shielded Magnex gradient coils yielded a maximum gradient field magnitude of 180 mT/m. The samples to be measured were cultivated in a plastic Petri dish (50 mm in diameter). MR images were acquired using a classic Spin Echo pulse sequence with the following parameters: echo time (TE) = 12 ms, repetition time (TR) = 3.8 s, matrix size = 256 × 256 pixels (30 × 30 mm with resolution 0.117 mm per pixel), slice thickness = 2 mm and number of averages of the MR signal was (NA) = 5. The obtained data were processed with the MAREVISI and MATLAB programs.

Data were processed using MICROSOFT EXCEL® (USA) and STATISTICA.CZ Version 8.0 (Czech Republic) unless noted otherwise. Results are expressed as mean ± standard deviation (S.D.) unless noted otherwise (EXCEL®). Statistical significances of the differences between cluster area and time of cultivation were determined using STATISTICA.CZ. Differences with p < 0.05 were considered significant and were determined by using of one way ANOVA test (particularly Scheffe test), which was applied for means comparison.