Mature seeds of blue spruce were collected in 2008 from a tree (denoted as “2” owned by Sheffield’s Seed Company, New York, USA. The seeds were bought from the owner in 2011 and were stored at -40°C since then.

Seeds were rinsed under running water for 18 h, surface-sterilized with 75% medical- grade ethanol for 30 s and rinsed 3–5 times with sterile distilled water (1 min per rinse). Subsequently, seeds were sterilized with 4.2% sodium hypochlorite (NaClO; effective chlorine concentration: 4%; (Enshi Reagent, Hubei, China) for 15 min and rinsed again 3–5 times with sterile distilled water (1 min per rinse). Under sterile conditions, seed coats and endosperm were carefully removed, and the zygotic embryos were excised and cultured on EC induction medium. The medium consisted of full-strength of modified Litvay’s medium (mLV) (Litvay et al., 1985; Klimaszewska and Park, 2001), supplemented with 2.0 mg L^-16-BA (Sigma-Aldrich, Missouri, USA), 4.0 mg L^-1 2,4-D (Sigma-Aldrich, Missouri, USA) and 0.5 g L^-1 L-glutamine (Sigma-Aldrich, Missouri, USA), with addition of, 0.8 g L^-1 casein hydrolysate (Sigma-Aldrich, Missouri, USA) and 10 g L^-1 sucrose (Sigma-Aldrich, Missouri, USA) (Cao and Gao, 2022) as well as 4 g L^-1 gellan gum (Sigma-Aldrich, G1910-1KG, Missouri, USA). One hundred embryos were used in this study. Ten embryos were placed per Petri dish (90 mm diameter x 20 mm depth) (Shanghai Graft Horticultural Products Co., Ltd., Shanghai, China) for inducing EC cell lines. The embryo culture was maintained on the same medium in darkness during the whole EC induction period.

Four stably proliferated EC cell lines (201, 202, 211, 226) were transferred to proliferation medium (PM), which consisted of full-strength mLV supplemented with 4.0 mg L^-1 2,4-D, 1.0 mg L^-1 BA, 0.5 g L^-1 glutamine, 0.8 g L^-1 casein hydrolysate, 10 g L^-1 sucrose and 4 g L^-1 gellan gum. Cultures were maintained in darkness, and the medium was renewed every two weeks. For each EC cell line, 8 Petri dishes were cultured.

EC samples (80 mg per cell line) were collected to determine their initial fresh weights. Subsequently, 4 mL of liquid PM medium without plant growth regulators were added. The cultures were shaken vigorously to disperse the cells and then filtered using a Büchner funnel. The EC cells were transferred onto 8-cm filter papers placed in Petri dishes containing somatic embryo maturation (SEM) medium and incubated in the dark for 8 weeks. The SEM medium consisted of full-strength mLV medium, supplemented with 13.22 mg L^-1 ABA (Sigma-Aldrich, Missouri, USA), 0.5 g L^-1 L-glutamine, 0.8 g L^-1 casein hydrolysate, 30 g L^-1 sucrose, 6.5 g L^-1 gellan gum and 1 g L^-1 activated charcoal.

For cytological, physiological and biochemical analyses, SE cultures were sampled on day 0 (proembryogenic masses stage), 15 (late-stage SE development), 30 (cotyledon-stage) and 45 (mature stage), with 12 g of tissue collected per sample. All samples were immediately frozen in liquid nitrogen and stored at -80°C until further use, following the protocol of Cao and Gao (2022).

At day 60, the SEs were counted after staining, and embryo morphology was examined using a stereomicroscope (OLYMPUS SZX 7, Japan) equipped with a digital camera (Canon DS126271, Japan). For staining, fresh SEs were placed on a glass slide, followed by staining with 0.1% carmine solution for 10 minutes under a coverslip. To ensure uniform cell spreading, the coverslip was gently tapped with the flat end of a pencil. Observations and image acquisition were carried out using an optical microscope (OLYMPUS CX 31, Japan) equipped with a camera (Canon DS126271, Japan). The experiment was conducted in four independent biological replicates.

SE germination tests were initiated on day 45. SE samples were first dried in a 6-well cell culture plate (Corning-Costar 3516; Corning Incorporated, New York, USA). Two layers of dry, sterile filter paper were placed in three wells, on which SEs were positioned, while the remaining three wells contained sterile water to maintain humidity. Plates were sealed with Parafilm (Pechiney Plastic Packaging, Chicago, USA) and incubated in darkness at 4°C for 14 days. Following desiccation, SEs were transferred to germination medium consisting of full-strength mLV, supplemented with 2 g L^-1 activated charcoal, 10 g L^-1 sucrose and 4 g L^-1 Gelrite. Cultures were maintained in darkness at 23°C for 7 days, followed by 14 days under light conditions of 8 h photoperiod with a light intensity of 100 μmol m^-2 s^-1 under cool white, fluorescent tubes. Germinated SEs were then transferred to culture flasks (Greiner Bio – One, Kremsmünster, Germany) for continued growth. After 3 months, plantlets were transplanted into a greenhouse and acclimatized in a substrate composed of soil, vermiculite and perlite at a ratio of 3:1:1.

MDA content was determined using the MDA A003–1 kit; Nanjing Jiancheng Bioengineering Institute (Nanjing, China). Embryo tissues of 0.2 g along with 0.6 mL of PBS solution (0.1 mol L^-1) were homogenized in mortar on ice. After centrifugation at 1800 g for 10 min, the supernatant of 80 μL and 80 μL of Reagent I were mixed thoroughly in a glass tube, followed by addition of 1200 μL of Reagent II and 400 μL of Reagent III according to the manufacturer’s instructions. The tubes were incubated in a boiling water bath for 40 minutes, followed by cooling down under running water and centrifuged at 1800 g for 10 minutes. Supernatants of 200 μL were loaded in a 96-well cell plate and OD values at 532 nm were obtained using an enzyme counter (BIOTEK, Vermont, USA) according to the manufacturer’s instructions. Each treatment was conducted with 3 biological replicates.

Endogenous IAA, ABA, GA and ZR contents were determined by Enzyme-linked Immunosorbent Assays (ELISA) using the Mlbio ELISA kits; Shanghai Enzyme-Link Bio-technology Co., Ltd. (Shanghai, China). Embryogenic tissues of 0.5 g along with 4.5 mL of 0.01 mol L^-1 phosphate buffer (pH=7.2-7.4) were homogenized in mortar on ice. After centrifugation at 1800 g for 20 min at room temperature, the supernatant was taken and stored on ice or fridge. The extracts and reagents were added according to the manufacturer’s instructions to determine the contents of endogenous IAA, ABA, GA and ZR. Each treatment was conducted with 3 biological replicates.

As stated in material section, all experiments or biochemical analyses were repeated and the number of biological replicates varied among the different analyses, which are specified in each section as stated above. SPSS 23 software was used for ANOVA and multiple comparison tests using the LSD method. SigmaPlot 12.0 software was used for graphing.

Out of 100 seed explants, 53 seeds resulted in EC formation, giving the induction rate of 53%. From the EC cell lines, the best four were used in further studies.

All four cell lines maintained stable proliferation on the PM medium; however, their SE maturation capabilities differed significantly. After 15 days of maturation culture, cell line 202 yielded the highest number of late-stage SEs, reaching 1,625 SEs g^-1 FW (Figure 1B), followed by Cell line 201 with 1,350 SEs g^-1 FW (Figure 1F). In contrast, cell line 211 showed slower SE development, yielding only 325 SEsg^-1 FW (Figure 1J), while cell line 226 exhibited only callus proliferation and failed to produce any mature SE (Figure 1N).

Cytological analysis revealed distinct differences in internal callus cell structures of the four cell lines from day 0 on PM medium (Figures 1A, E, I, M). Cell line 202, which exhibited the highest embryogenic potential, contained an abundant number of early SEs characterized by large embryonic heads and well-defined structures (Figure 1A). Cell line 201 (Figure 1E) showed a similar but less pronounced pattern, producing fewer SEs and smaller embryonic heads than 202 (Figure 1A). Cell line 211 formed only a limited number of poorly structured SEs (Figure 1I), whereas cell line 226 (Figure 1M) failed to produce early SEs displaying numerous scattered cell clusters instead. After 15 days of culture on the PM medium, morphological differences became more pronounced: cell lines 202 and 201 had already produced many late-stage SEs (Figures 1B, F). By day 30, the SE-competent lines (202, 201, and 211) formed cotyledonary SEs (Figures 1C, G, K). At day 45, mature SEs formed were as follows: 1,625 ± 82 SEs g^-1 FW for 202 (Figure 1D); 1,350 ± 178 Ses g ^-1 FW for 201 (Figure 1H); and 325 ± 22 SEs g^-1 FW for 211 (Figure 1I). Cell line 226 (Figure 1P) continued to show only proliferation with no SE development.

To obtain healthy plantlets from SEs, desiccation is needed before the SEs can germinate (Figure 2A). The germination test showed SEs germinated normally on the germination medium (Figure 2B) and the germinated SEs grew vigorously on a large culture container (Figure 2C) after 3 months. The plantlets were well established ex vitro on the substrate (Figure 2D).

At day 0, MDA content was lowest in cell lines 202 and 201 and highest in cell line 226 (Figure 4E). By day 15, MDA levels increased in cell lines 202 and 201, while remaining unchanged in cell lines 211 and 226. At days 30 and 45, MDA content rose significantly in cell lines 202, 201, and 211, but decreased in cell line 226.

Progression through the proembryogenic mass (PEM) stages (Filonova et al., 2000) is a key determinant of somatic embryogenesis (SE) capacity. Highly embryogenic lines (202, 201) advanced to PEM III and formed early bipolar embryos, whereas line 226 remained arrested at PEM II, exhibiting signs of oxidative stress and cellular damage. This oxidative stress likely resulted from elevated ROS levels originating from culture conditions: 2,4-D activates NADPH oxidase to generate superoxide anions, BA enhances mitochondrial respiration and ROS production, and sucrose-induced osmotic stress further amplifies oxidative imbalance. Therefore, morphogenetic differentiation, rather than callus proliferation, determines SE success, consistent with previous findings in larch and spruce (Bozhkov et al., 2010; Savane et al., 2023).

Embryogenic lines (202 and 201) also exhibited early and coordinated accumulation of soluble sugars and starch, providing both metabolic energy and structural carbon for embryo development. Concurrent increases in total protein (TP) reflected the synthesis of storage proteins and antioxidant enzymes such as SOD and CAT (Businge et al., 2012; Hazubska-Przybył et al., 2022). In contrast, the non-embryogenic line 226 showed limited reserve buildup and weak stress adaptation, underscoring the link between metabolic readiness and embryogenic potential.

ROS act as essential developmental signals during SE, but their levels must remain tightly controlled. In this study, SOD activity in embryogenic lines peaked around day 45, followed by increases in CAT and POD activity, forming a sequential “conversion–clearance” mechanism that maintained optimal ROS balance. This regulated redox environment promoted differentiation while preventing oxidative injury. In contrast, non-embryogenic lines exhibited insufficient antioxidant activity and excessive lipid peroxidation, indicating ROS imbalance. These findings are consistent with reports in Pinus and Lycium barbarum (Cui et al., 1999; Rahal et al., 2014). Moderate MDA accumulation in embryogenic lines supports the “ROS metabolic switch” hypothesis, wherein controlled ROS flux facilitates cellular reprogramming, while excessive or insufficient ROS suppresses SE (Cui et al., 1999; Fatima et al., 2011; Wu et al., 2020).

Endogenous hormones closely interact with ROS metabolism to regulate SE progression (Méndez-Hernández et al., 2019; Chen et al., 2021). Auxin (IAA) increased during induction across all lines, but similar IAA levels in embryogenic and non-embryogenic lines suggest that responsiveness—rather than concentration—determines SE competence (Ayil-Gutiérrez et al., 2013; Kępczyńska and Kępczyńsk, 2023). Embryogenic lines also exhibited elevated gibberellin (GA) levels during maturation, supporting embryo elongation (Kim et al., 2019), and a moderate rise in zeatin riboside (ZR), facilitating cell proliferation and differentiation.

Abscisic acid (ABA) peaked at day 45 in embryogenic lines, coinciding with starch and protein accumulation. This suggests that ABA promotes reserve mobilization and desiccation tolerance (Yan & Chen, 2017). Elevated POD activity and moderate MDA levels during the ABA peak indicate functional ABA–ROS crosstalk, aligning antioxidant activity with developmental transitions. The low ABA and high GA/ABA ratios in line 226 disrupted this coordination, impairing maturation.

SE competence in blue spruce depends on the synchronization of morphological differentiation, energy reserve metabolism, ROS homeostasis, and hormonal signaling. Among these, ABA–ROS interplay appears central to integrating stress adaptation with developmental programming. Highly embryogenic lines maintain this network effectively, whereas non-embryogenic lines exhibit disrupted integration, poor metabolic balance, and limited morphogenetic progression.

Biochemical markers such as high TP content, elevated SOD and POD activity, and moderate MDA accumulation consistently indicated embryogenic potential, suggesting their utility for early screening of responsive cell lines. Manipulating hormonal ratios—particularly optimizing ABA relative to GA and IAA—and enhancing antioxidant capacity may further improve SE outcomes. Future work should explore the transcriptional and metabolic regulators mediating ABA–ROS interactions and reserve mobilization to inform rational optimization of culture systems for conifer SE.