Four embryogenic cell lines, 11:12:02, 11:12:04, 09:73:06, and 09:77:03 of Norway spruce (Picea abies (L.) H.Karst.) initiated and established by standard procedures (von Arnold and Clapham 2008) were used in this study. Briefly, the cell lines were maintained through proliferation/multiplication of PEMs in the dark, by subculture every 2 wk on solid, half-strength LP (½ LP) medium (von Arnold and Eriksson 1981), supplemented with 2.21 mg L−1 2,4-dichlorophenoxyacetic acid (2,4-D), 1 mg L−1 N6-benzyladenine (BA), and 10 g L−1 sucrose in 9-cm-diameter Petri plates. For maturation, PEMs were transferred to pre-maturation medium (DKM; Krogstrup 1986, without growth regulators) for 7 d before transfer to DKM maturation medium supplemented with 16 mg L−1 abscisic acid (ABA and 30 mg L−1 sucrose (Dobrowolska et al. 2017)). Proliferation and maturation media were adjusted by HCl and NaOH to pH 5.8 before autoclaving at 121 °C and 103.4 kPa. Maturation lasted 8 wk, during which the cultures were subcultured biweekly to fresh DKM maturation medium. Liquid media of the same composition, but without gelling agent, were autoclaved and used for the bioreactor cultures described in the ‘Experimental set up’ section.
Harvested mature somatic embryos from the experiments were desiccated for 3 wk in darkness and high humidity (100%) in a closed Petri plate together with a smaller Petri plate with water according to standard procedures for desiccation of somatic embryos of Norway spruce (von Arnold and Clapham 2008). Desiccated mature embryos were placed on solidified germination medium for 1 wk in darkness, 2 wk in continuous red light (wavelength 660 nm; TL-D 18W/15, Philips, Stockholm, Sweden) at 5 μmol m−2 s−1 at 20 °C, and then moved under continuous white fluorescent tubes (Fluora L 18W/77, Osram, Johanneshov, Sweden) at 100–150 μmol m−2 s−1 (Dobrowolska et al. 2017). The germination medium had the following composition per liter: 764 mg KNO3, 173 mg NH4NO3, 381 mg KH2PO4, 533 mg MgSO4.7H2O, 83 mg CaCl2·2H2O, 30 g sucrose, 0.5 g casein hydrolysate, 1 mg thiamine, 50 mg inositol, and for solidified medium, 3.5 g gelrite™. Iron and microelements were the same as for DKM medium. The pH was adjusted to 5.8 prior to autoclaving. All steps of the culturing process, from proliferation to germination, took place in culture rooms set at 22 ± 3°C. All chemicals were obtained from Sigma-Aldrich Sweden AB, Stockholm, Sweden.
Experimental set up
Effects of dispersion on PEM proliferation and embryo maturation were studied in cell lines 11:12:02, 11:12:04, 09:73:06, and 09:77:03 on solid medium and in cell lines 11:12:02 and 11:12:04 in temporary immersion bioreactors. For culture on solid medium, 0.5 g of PEMs was dispersed in 40 mL ½ LP liquid medium of the same composition as used for proliferation, using the Mamun (2015) disperser system at a flow rate of 2500 mL min−1 under sterile conditions. Three replicates each of cell lines 11:12:02, 11:12:04, 09:73:06, and 09:77:03 were set up by spreading 40 mL of the dispersed PEMs through pipetting onto a filter paper (Whatman Grade 2 Qualitative Filter Paper Standard Grade, circle, 70 mm, GE Healthcare—Whatman, VWR, Umeå, Sweden) placed on solid proliferation medium in a 9-cm Petri plate. Excess medium fluid was removed by pipetting. As a control, the same amount of non-dispersed aggregates of 3- to 5-mm-diameter tissue clumps of PEMs were transferred by forceps to solid proliferation medium in Petri plates. In total, for the experiments on solid medium, each cell line was represented by three Petri plates with dispersed cultures and three Petri plates with non-dispersed cultures.
For bioreactor culture, PEM aggregates were dispersed using the same dispersion system as used for solid cultures. Six bioreactors of cell lines 11:12:02 and 11:12:04 were set up by dispersion of 2 g of PEM aggregates into 50 mL ½ LP liquid medium spread onto the bioreactor screens. As a control, the same amount of non-dispersed aggregates (3–5 mm in diameter) of PEMs were transferred by forceps from solidified proliferation medium in Petri plates to each of six replicate bioreactors for either cell line. In total, for the experiments in bioreactors, each cell line was represented by six bioreactors with dispersed cultures and six bioreactors with non-dispersed cultures.
Temporary immersion bioreactors (Mamun 2015; Businge et al. 2017) were used for production of mature somatic embryos through a first step of proliferation of embryogenic masses (both dispersed and non-dispersed), followed by maturation in the same bioreactor. Immersion periods were set to a period of 1 min in every 12 h for all developmental stages. After 12–14 d, the proliferation medium was replaced by pre-maturation medium for 7 d, followed by maturation medium for 8 wk. During maturation, the bioreactors were subcultured biweekly, by replacing the maturation medium in the flasks associated with the bioreactor with fresh culture medium.
Methods for evaluation of embryo maturation and synchronization
To study the effect of dispersion on embryo maturation, PEMs were multiplied in proliferation medium for one subculture cycle (14 d), and then taken through the pre-maturation and maturation steps, as described above. After 8 wk of maturation, samples of tissue composed of PEMs and mature embryos were randomly collected from each bioreactor and transferred to Petri plates. Three samples were collected from each dispersed bioreactor, as were several tissue clumps from non-dispersed bioreactors. All embryos were then harvested from the collected tissue samples under a microscope (Nikon SMZ-10, Nikon Corp., Japan) using ×10 magnification. The total number of mature embryos in each replicate and the number of mature embryos per gram of fresh weight (gFW-1) at the start of the proliferation treatment were determined. The proportion of embryos collected out of the total number of embryos present in the bioreactor was estimated, based on the area of tissue sampled compared to the total area covered by tissue in the bioreactor. The sampled area was measured by image analysis of an overview photo of the bioreactor, after the samples had been removed. The length of each harvested embryo was measured using image analysis and analyzed as described in (Fig. 1).
Image analysis of PEMs, bioreactor cultures, and mature embryos
Images of dispersed PEMs, cultures in bioreactors at the end of maturation, and mature embryos after harvest were acquired by a digital single-lens reflex (DSLR) camera (Nikon D5100, Nikon Corp., Tokyo, Japan). At the start of the experiments, three samples of dispersed PEMs submerged in liquid medium were collected from each bioreactor, poured into Petri plates, and photographed for estimating the size distributions. Each sample contained 651 ± 155 dispersed PEMs. The area occupied by each dispersed PEMs was obtained from image analysis using ImageJ 1.48v (National Institute of Health, Bethesda, MD) and was considered as the dimension of a dispersed PEMs. Results from the replicates were combined to get the estimated mean and standard deviation of the size of dispersed aggregates in each cell line (Fig. 1).
At the end of maturation, bioreactors were opened, and two images of each bioreactor culture were analyzed in ImageJ to estimate the area occupied by the harvested mature embryos. Since the embryos had developed rather homogeneously from the dispersed PEMs, the number of embryos gFW−1 of PEMs in dispersed bioreactor was estimated from the number of harvested embryos and the area occupied by them.
A total of 1037 and 1176 images of mature embryos were captured from dispersed and non-dispersed PEMs, respectively, of cell line 11:12:02, and 727 and 573 images of mature embryos from dispersed and non-dispersed PEMs, respectively, of cell line 11:12:04. On solid culture medium, images of somatic embryos from 24 plates of dispersed and non-dispersed cultures, containing an average of 129 and 78 mature embryos per plate, respectively, were analyzed. The images were trimmed using GIMP 2.8.10 (GNU Image Manipulation Program, a free and open-source graphics editor, https://www.gimp.org/) and converted into a binary image using ImageJ. For each embryo, the length of the major axis of the ellipse drawn around the embryo was acquired in ImageJ. The axis length was considered as the length of an embryo. The number of embryos was then obtained for each of the embryo length intervals of 0.0–1.0, 1.0–2.0, 2.0–3.0, 3.0–4.0, and 4.0–5.0 mm, giving δ as 1 mm.
Differences between dispersed and control cultures were measured as the: (a) number of mature embryos gFW−1 of PEMs; (b) coefficient of variation (CV) for number of mature embryos per bioreactor; (c) average length of embryos, (q̄), in a length interval; (d) width of the normalized length distribution function, φ(q̄), at 50, 66, and 75% of maximum; (e) CV of lengths of embryos; and (f) confidence interval of CV. The number and length of embryos, and the width of the curve, φ(q̄), were normally distributed (see Supplementary document published online for the details and φ(q̄) plots).