Direct somatic embryogenesis and related gene expression networks in leaf explants of Hippeastrum 'Bangkok Rose'

Abstract

Hippeastrum, a highly diverse genus in the Amaryllidaceae family, is a valuable ornamental bulbous flowering plant. Somatic embryogenesis (SE) is an efficient method for mass production of Hippeastrum plantlets. Previous studies have been devoted to the in vitro propagation of Hippeastrum, but the SE and its regulatory networks are rarely reported. In this study, we established a direct SE method of Hippeastrum 'Bangkok Rose' using leaf bases as expiants. MS supplemented with 1.00 mg * L-1 NAA +1.00 mg * L-1 KT + 0.25 mg * L-1 TDZ was the optimal medium for SE. Histological observations showed that the bipolar somatic embryo originated from the epidermal cell layer and underwent initiation, globular, scutellar and coleoptile stages. During SE, endogenous hormones of IAA, CTK, ABA, and SA were highly accumulated. Transcriptomic analysis revealed the genes encoding auxin biosynthesis/metabolic enzymes and efflux carriers were induced, while the auxin receptor of TIRI and ARE transcriptional repressor of Aux/IAA were down-regulated and up-regulated, respectively, leading to suppression of auxin signaling. In contrast, cytokine signaling was promoted at the early stage of SE, as biosynthesis, transport, and signaling components were up-regulated. Various stress-related genes were up-regulated at the early or late stages of SE. Chromatin remodeling could also be dynamically regulated via distinct expression enzymes that control histone methylation and acetylation during SE. Moreover, key SE regulators, including WOXs and SERKs were highly expressed along with SE. Overall, the present study provides insights into the SE regulatory mechanisms of the Hippeastrum.

Headnote

ABSTRACT

Hippeastrum, a highly diverse genus in the Amaryllidaceae family, is a valuable ornamental bulbous flowering plant. Somatic embryogenesis (SE) is an efficient method for mass production of Hippeastrum plantlets. Previous studies have been devoted to the in vitro propagation of Hippeastrum, but the SE and its regulatory networks are rarely reported. In this study, we established a direct SE method of Hippeastrum 'Bangkok Rose' using leaf bases as expiants. MS supplemented with 1.00 mg * L-1 NAA +1.00 mg * L-1 KT + 0.25 mg * L-1 TDZ was the optimal medium for SE. Histological observations showed that the bipolar somatic embryo originated from the epidermal cell layer and underwent initiation, globular, scutellar and coleoptile stages. During SE, endogenous hormones of IAA, CTK, ABA, and SA were highly accumulated. Transcriptomic analysis revealed the genes encoding auxin biosynthesis/metabolic enzymes and efflux carriers were induced, while the auxin receptor of TIRI and ARE transcriptional repressor of Aux/IAA were down-regulated and up-regulated, respectively, leading to suppression of auxin signaling. In contrast, cytokine signaling was promoted at the early stage of SE, as biosynthesis, transport, and signaling components were up-regulated. Various stress-related genes were up-regulated at the early or late stages of SE. Chromatin remodeling could also be dynamically regulated via distinct expression enzymes that control histone methylation and acetylation during SE. Moreover, key SE regulators, including WOXs and SERKs were highly expressed along with SE. Overall, the present study provides insights into the SE regulatory mechanisms of the Hippeastrum.

Keywords: Hippeastrum; Tissue culture; Somatic embryogenesis; Gene regulation

1. Introduction

After over 200 years of selection and hybridization, Hippeastrum has great diversity in many characteristics like plant shape, flower colors, flowering time, etc. Hippeastrum can be propagated through seeds, offset bulbs, bulbs, basal cuttings, and tissue culture (Datta, 2021). Due to the complex genetic background of Hippeastrum varieties, seed propagation leads to the segregation of traits in progeny, thus unsuitable for commercial production. Propagation through offset bulbs, cuttings, and tissue culture can avoid the problem of segregation, but the efficiency of offset bulblets is low because it needs enough mother bulbs. Bulb cuttings are the primary approach for seedling propagation of Hippeastrum cultivars in commercial production. However, tissue culture is also a cost-effective technique for most plants that could also be considered an ideal method for large-scale production of Hippeastrum seedlings by improving reproductive methods and efficiency, including somatic embryogenesis (SE).

In vitro regeneration of Hippeastrum through tissue culture has been widely studied since 1977 (Seabrook and Cumming, 1977; Fountain, 1979; Kim et al., 2005; Amani et al., 2015; Wang et al., 2020). Expiants from vegetative or reproductive organs were used, including scale, scape, basal plate, leaf, ovary, peduncle, pedicel, petal, anther, and filament. Among them, scale (Amani et al., 2015), scape (Seabrook and Cumming, 1977), peduncle (Seabrook and Cumming, 1977; Fountain, 1979), and pedicel (Kim et al., 2005; Yu et al., 2020) were found to be the optimal expiants for in vitro tissue culture. Plant growth regulators (PGRs) with different combinations were evaluated. The combination of 6benzylaminopurine (6-BA) with naphthaleneacetic acid (NAA) or 2,4-dichlorophenoxyacetic acid (2,4-D) were commonly used combinations to trigger the regeneration of Hippeastrum (Seabrook and Cumming, 1977; Fountain, 1979; Kim et al., 2005; Amani et al., 2015; Wang et al., 2020). Many studies have focused on establishing or improving in vitro propagation of Hippeastrum by adventitious bud regeneration via tissue culture. However, the propagation efficiency is typically low and rarely used in commercial production. Therefore, SE could be an effective method for the large-scale production of Hippeastrum seedlings (Kim et al., 2005; Wang et al., 2020).

SE is an in vitro regeneration process in which the somatic embryos are induced from somatic cells and develop into whole plants (Wang et al., 2023). Somatic embryos are bipolar structures with no vascular connections underlying the plant (von Arnold et al., 2002). Different from in vivo zygotic embryogenesis, which is initiated from fertilization. The SE is triggered by exogenous chemical or physical stimuli produced using PGRs or certain stress conditions such as low or high temperature, osmotic shock, and heavy metals (Nic-Can et al., 2016). SE can be induced either directly from the edge of the explant without an intervening callus phase or indirectly through the proliferation and dedifferentiation of the callus (Mendez-Hernandez et al., 2019). Like zygotic embryogenesis, the development of SE undergoes several developmental stages according to the shape of the embryos. In dicot plants, SE development was classified into globular-shaped, heart-shaped, torpedo-shaped, and cotyledonary stages (Winkelmann, 2016). While in the monocots, they were termed as globular, scuteliar, and coleoptile stages (Zhao et al., 2017). In many cases, the primary somatic embryos induced by expiants did not directly differentiate into a mature plant; instead gave rise to successive cycles of secondary embryos (Bao et al., 2012; Naing et al., 2013). The secondary somatic embryogenesis system enables large-scale production of vegetative propagules in a brief time, which may contribute to rapid clonal propagation and genetic transformation (Bhatia and Bera, 2015; Singh et al., 2019; Liu et al., 2022).

SE is a cellular reprogramming process controlled by the integration of multiple signaling cascades and the integration of numerous genes and transcriptional networks. In many species, WUSCHEL (Zuo et al., 2002), BABY BOOM (BBM) (Boutilier et al., 2002), LEAFY COTYLEDON (LEG) (Braybrook et al., 2006), SOMATIC EMBRYOGENESIS RECEPTOR-LIKE KINASE (SERK) (Schmidt et al., 1997), and AGAMOUS-LIKE15 (AGL15) (Thakare et al., 2008) are conserved SE regulators whose function in SE are coordinated with numerous signals such as hormone, stress, and epigenetic mechanisms. SERKs are leucine-rich repeat receptor-like kinase (LRR-RKK) involved in the acquisition of the embryogénie state of the cells (Kumar and Van Staden, 2019). The coordination and interaction of plant hormones determine the proliferation of meristem cells, the transformation of somatic cells into embryogénie tissues, and the further differentiation and maturation of somatic embryos (Haque and Ghosh, 2017; Mendez-Hernandez et al., 2019). Hormones, especially auxin (AUX) and cytokine (CTK), are central players in the induction of SE (Salaim et al., 2021), and the ratio between auxin and cytokine determined the induction of shoot or root (Elhiti and Stasolla, 2022). Auxin efflux carriers PIN-FORMED1 (PINI) and PIN7 mediated polar transport of auxin generates a signal gradient that specifies the embryonic apical-basal axis in Arabidopsis (Robert et al., 2013; Xiong et al., 2019). Moreover, YUCCAs and AUXIN RESPONSE FACTORS (ARFs), encoding core components of auxin biosynthesis and signaling pathways, interplayed with LEG during SE (Wójcikowska et al., 2013; Wójcikowska and Gaj, 2017). Stresses generated by wounding damages of the expiants, high concentration of PGRs, or external stimuli were thought to be a switch that triggers cellular reprogramming and directs them to the acquisition of cellular totipotency (Nic-Can et al., 2016). Stress-induced genes such as GLUTATIONE-S-TRANSFERASES (GSTs) (Thibaud-Nissen et al., 2003), GERMIN LIKE PROTEINS (GLPs) (Mathieu et al., 2006; Barman and Banerjee, 2015) and HEATSHOCK PROTEINS (Hsps) (Puigderrajols et al., 2002) are highly associated with SE. Epigenetic mechanisms are engineered by DNA methylation, chromatin remodeling, and microRNAmediated regulation (Mahdavi-Darvari et al., 2015; Ikeuchi et al., 2016). In the nucleus, histone proteins serve as a primary material to form the fundamental unit of chromatin, the nucleosome (Pontvianne et al., 2010). Post-translational modifications of histones, including methylation, acetylation, phosphorylation, and ubiquitination are vital for chromatin re-modeling, which plays a crucial role in establishing gene expression patterns and maintaining epigenetic regulations (Jarillo et al., 2009). Recent studies revealed that epigenetic mechanisms are fundamentally involved in regulating SE, possibly by regulating the expression of SE regulators, including WUSCHEL-related homeobox 4 (W0X4), BBM1, and LECI (Li et al., 2011; Kumar and van Staden, 2017; Duarte-Aké et al., 2019). WOX homeobox transcription factors are essential in somatic embryogenesis by promoting vegetative-to-embryonic transition and maintaining cell totipotency (Jha et al., 2020). CLV belongs to the CLE peptide family, which forms a negative-feedback regulation mode with WOX by restricting their activity (Somssich et al., 2016).

This study established a direct somatic embryogenesis protocol for in vitro propagation of H. 'Bangkok Rose' using leaf bases as expiants. Further characterized the regulatory mechanisms of hormone, stress, and epigenetics during SE by transcriptomic analysis. These findings provided an effective production system and valuable information to understand the molecular regulation of SE in Hippeastrum.

2. Materials and methods

2.1. Plant materials, growth conditions, and the somatic embryo induction

The H. 'Bangkok Rose' plants were grown in the greenhouse in the South China Botanical Garden (Guangzhou, China) under natural light (no more than 800 pmol * m 2 * s 2) with an average temperature changed from 10 °C to 32 °C and relative humidity ranging from 70% to 98%. Sterile seedlings were generated via tissue culture using bulbs planted in greenhouse as expiants. The expiants were inoculated on MS medium supplemented with 5.00 mg * L 1 6-BA and grown in a growth chamber with LED light (27-45 pmol * m 2 * s-1,12h/24 h) at (25 ± 1) °C. The regenerated bulbs around 0.6-0.8 cm in diameter were sub-cultured for 40 days under the same tissue-culture condition as mentioned above. The leave bases of the 40-day-old regenerated seedlings were used as expiants to induce somatic embryos. Around 50 leaf bases were cut into 0.75 cm long segments and tiled on MS medium supplemented with 6-BA (0, 1.00, 2.00, 3.00 mg * L"1), NAA (0, 1.00, 2.00, 3.00 mg * L 2), kinetin (KT) (0, 1.00, 2.00, 3.00 mg * L 2) and thidiazuron (TDZ) (0, 0.25, 0.50, 1.00 mg * L 2) with combinations shown in Table 1. For each combination, five bottles of expiants were cultured (nine inoculated in each bottle) with the same growth condition as mentioned above. The somatic embryogenesis ratio (number of expiants generating somatic embryos/total number of expiants inoculated x 100%) and the number of regenerated buds per explant were calculated 45 days after inoculation.

2.2. Histological observations

Leaf expiants at 0, 10, 20, and 30 days after inoculation (DAI) were fixed in fixation solution [2.5% isovaleraldehyde (Sigma-Aldrich, Germany) + 2.5% paraformaldehyde (Fluka, Buchs, Switzerland) in 0.1 mol * L 1 sodium phosphate buffer solution (pH 7.2)] for 24 h at room temperature, followed by 24 h-vacuum-infiltration at 4 °C. The fixed samples were gradually dehydrated in 30%, 50%, and 70% ethanol for 20 min each, in 80% and 90% ethanol for 15 min each, in 100% ethanol twice for 30 min each, and in 100% propylene oxide (Seebio, Shanghai) twice for 30 min each at room temperature. The dehydrated samples were sequentially immersed in epoxy propane + resin (EPON812) (Wirsam, Cape Town, South Africa) (3:1, v/v) solution for 30 min, epoxy propane + resin (1:1) solution for 1 h, and epoxy propane + resin (1:3, v/v) solution for 2 h. In the end, samples were infiltrated with pure resin at room temperature and baked in an embedding box at 40 °C for 7 days. Before microscopic observation, the embedded samples were sliced into 2 цт sections (Leica Reichert Ultracut S, Austria), and stained with toluene blue, about 20-30 slices were observed under the light microscope (Leica DVM6, Germany).

2.3. Isolation of total RNA, transcriptome sequencing and qRT-PCR

Leaf expiants at 0,10, 20, and 30 DAI (9 expiants for each stage) were collected and freezed in liquid nitrogen. Isolation of the total RNA of leaf expiants was performed using the polysaccharide-polyphenol RNA prep plant total RNA Extraction Kit (TIANGEN, DP441, Beijing, China) following the instructions of the manufacturer. RNA purity was analyzed using an Agilent 2100 Bio-analyzer (Agilent Technologies, Inc., Santa Clara, CA, USA), and the concentration was measured with a Nanodrop ND 2000 (NanoDrop Thermo Scientific, Wilmington, DE, USA). cDNA libraries were constructed, and the no-reference transcriptome sequencing was performed on Illumina HiSeq high-throughput sequencing platform (HiSeq 2000, SanDiego, GA, USA) by Majorbio Bio-pharm Technology Co., Ltd (Shanghai, China).

The analysis of qRT-PCR was conducted to check the RNA-seq data accuracy. The cDNA library was generated using the cDNA synthesis SuperMix kit (Trans, Beijing, China) following the manufacturer's instructions and diluted five times before use. For qRT-PCR, 1 |iL diluted cDNA, 10 gL 2 x Green qPCR SuperMix (Trans, Beijing, China), 0.8 pl primer pair (10 Iimol- L-1), and 8.2 nL ddH2O was correctly mixed and incubated on Light Cycler 480 System (Roche Diagnostics, Germany) for qRT-PCR detection. The amplification program was as follows by Iqbal (Iqbal et al., 2022): preincubation at 94 °C for 30 s for initial denaturation, 40 cycles of denaturation at 94 °C for 5 s, annealing of primers at 57 °C for 15 s, and elongation at 72 °C for 10 s. After the completion of the amplification cycles, a melting curve program was initiated to generate melting curves. The primers used for qRT-PCR (Table SI) were designed by the online tool Primer 3 (http://primer3.ut.ee/). Three technical replicates for each sample were used in the qRT-PCR experiment. The relative gene expression levels were calculated using the 2 AACt method (Livak and Schmittgen, 2001).

2.4. Assembly and annotation of the transcriptome and differential expression analysis of the genes

High-quality clean data were obtained by filtering low-quality reads with over 10% mismatch rate and removing adapter sequences. GC content and the Q20 and Q30 score of clean reads were calculated. The clean reads were subsequently assembled to obtain the H. 'Bangkok Rose' unigenes library. The inserted fragments distribution on unigenes was checked to assess the randomness of mRNA fragmentation and degradation. The length distribution of the inserted fragments was generated to determine the fragment length dispersion. A saturation map was developed using RSeQC-2.3.6 software (https://rseqc.sourceforge, net/) to evaluate the capacity of the libraries and the sufficiency of the mapped reads compared with a unigenes library. The sequencing data (PRJNA862291) was uploaded into the BioProject database on the website of the National Center for Biotechnology Information (NCBI) (https://www.ncbi.nlm.nih.gov/). Functional annotation was carried out by blasting the unigenes sequence or the corresponding amino acid sequence against the databases of Kyoto Encyclopedia of Genes and Genomes (KEGG), Swiss-Prot, NCBI non-redundant protein sequences (NR), Gene Ontology (GO), Clusters of Orthologous Groups (COG), Clusters of Orthologous Groups for Eukaryotic Complete Genomes(KOG), and Protein family (Pfam) using the software of BLAST (http://blast.ncbi.nlm. nih.gov/Blast.cgi). The expression level of each transcript was calculated according to the method of fragments per kilobase of exon per million mapped reads (FPKM). The Differently expressed genes (DEGs) with a P-value <0.05 and a log2-fold change (FC) > 2.0 were identified using the DESeq2 tool (http://www. bioconductor.org/packages/release/bioc/html/DESeq.html). The obtained DEGs were then subjected to KEGG analysis. For timeseries analysis of the transcriptome, genes showing a similar expression pattern were clustered into different profiles using the STEM analysis tool integrated with the Major biological cloud platform (https://www.majorbio.com/). Seven profiles that showed significant change trends were selected and subjected to KEGG enrichment analysis. The top 10 highly enriched pathways were selected for cluster heat map analysis.

2.5. Measurement of the endogenous plant hormones

The hormone contents were determined by Metware Biotechnology Co., Ltd (three replicates per sample) using the briefly explained method by Iqbal (Iqbal et al., 2021). About 1 g fresh leaf base expiants were collected, at 0, 10, 20, and 30 days after inoculation and freezed in liquid nitrogen. After freezing in liquid nitrogen, collected samples were ground into powder and stored at -80 °C until needed. The ground sample (50 mg) was dissolved in 1 mL extraction buffer [methanol/water/formic acid (15:4:1, V/V/V)]. The extract was mixed with 10 pL internal standard solution (100 ng * mL"1 provided by Metware Biotechnology Co., Ltd), then vortexed (ES-VM25, Yesen) for 10 min and centrifuged (5424R, Eppendorf) for 5 min (12 000 r * min at 4 °C). After centrifugation, the supernatant was transferred to clean microcentrifuge tubes (2 mL) and evaporated to dryness. The dried samples were dissolved with 100 pL 80% methanol (V/V), filtered through a membrane filter (0.22 urn, CNW), and subsequently analyzed on UPLC-ESI-MS/MS system [UPLC, ExionLC™ AD (https://sciex.com.cn/); MS, Applied Biosystems 6500 Triple Quadrupole (https://sciex.com.cn/)].

3. Results

3.1. SE induction and morphological observations

In a previous study, we performed tissue culture using leaf bases and bulb scales of in vitro adventitious buds of over 100 Hippeastrum cultivars as expiants. We found 'Bangkok Rose' presented high SE efficiency using leaf bases as expiants. To optimize the SE efficiency, leaf bases of in vitro adventitious buds of 'Bangkok Rose' were inoculated on MS medium supplemented with different combinations of PGRs, including 6-BA, NAA, KT and TDZ (Table 1). From all combinations, 1.00 mg * L 1 NAA +1.00 mg * L 1 KT + 0.25 mg * L 1 TDZ showed the highest regeneration efficiency with a 53.33% SE ratio, significantly higher than the expiants cultured on other media. On average, each leaf explant produced 3.25 regenerated buds at 45 DAI in the most suitable medium.

We observed the morphological changes during somatic embryogenesis. At around 10 DAI, three to five transparent small protrusions were formed surrounding the wounding edges of the leaf base expiants (Fig. 1, a). These protrusions were supposed to be the initiation sites (is) of regeneration. The protrusions further developed into transparent globular embryos (ge) or scutellar embryos (se) at around 25-35 DAI (Fig. 1, b, c), and these transparent embryos subsequently differentiated into greenish regenerated buds (rb) with coleoptiles at about 40 DAI (Fig. 1, d). These observations suggested that the regeneration using 'Bangkok Rose' leaf bases could be established via direct somatic embryogenesis (DSE). To further confirm the DSE process, we performed histological sections and found that the somatic embryos could originate from the leaf explants' epidermal cell layer (Fig. 1, e). At around 10 DAI, we observed the periclinally divided two-daughter-cell (tdc) pairs and the early-stage somatic proembryo (spe)-like structures in the outmost cell layer of the forefront (Fig. 1, e). The spe-like structure developed into a globular or scuteliar embryo at around 25-35 DAI. The globular, especially the scutellar embryo, showed a typical bipolar structure with many small-sized meristem cells enriched in the apical and basal regions (Fig. 1, f, g). Moreover, a suspensor-like structure (sus) was observed in the base of the globular embryo, which was physiologically isolated from the rest of the maternal tissues (Fig. 1, f). Taken together, the DSE of "Bangkok Rose' originated from the epidermal cell layer of the leaf expiant and experienced distinct development stages before maturation, as in zygotic embryogenesis.

3.2. Changes in endogenous hormones during SE of 'Bangkok Rose'

To study the biosynthesis dynamics of different plant hormones during SE of 'Bangkok Rose', we measured the endogenous concentrations of indole-3-acetic acid (IAA), CTK, abscisic acid (ABA), gibberellic acid (GA), salicylic acid (SA), ethylene (ETH) and jasmonate (JA). The results showed that the concentration of IAA, CTK, ABA, and SA in 10 DAI, 20 DAI and 30 DAI samples were dramatically increased compared to that in 0 DAI samples (Fig. 2, a-c, e), among which IAA, CTK and SA reached peak concentrations at 20 DAI and slightly dropped at 30 DAI. In comparison, ABA concentration showed a continuous increasing pattern during SE (0-30 DAI) (Fig. 2, c). GA concentration presented a sudden increase at 30 DAI but remained low during 0-20 DAI (Fig. 2, d). The change in ETH concentration during SE was mild and did not show any patterns (Fig. 2, f). Interestingly, the concentration of JA in the leaf explant dropped to very low level after inoculation (0 DAI) (Fig. 2, g). In conclusion, the biosynthesis of IAA, CTK, ABA, and SA are induced along with SE, suggesting that these hormones play dominant roles in regulating SE of 'Bangkok Rose'.

3.3. Global mRNA expression profiles during SE of 'Bangkok Rose'

To study the transcrip tomie regulation mechanisms of SE in 'Bangkok Rose', transcriptomic analysis was performed using RNA-seq in the leaf expiants at 0,10, 20, and 30 DAI. After filtering low-quality reads and removing the adapters from the raw RNAseq data of 12 cDNA libraries (three biological replicates for each of the four developmental stages), around 83 GB of clean reads were produced. Samples were clustered separately following the developmental stages in principal component analysis (PGA) (Fig. SI). Overall, the Q30 of all clean reads was more than 94.28%, and the GC content was within 45.51%-47.21%. Detailed information on clean reads can be found in Table S2. Due to the lack of reference genome sequence of Hippeastrum, de novo assembly of the sequencing data was performed, which generated 352 224 transcripts with mean length and N50 length of 893.53 bp and 1 465 bp, respectively. The 211 490 longer transcripts with an average length of 976.71 bp and N50 length of 1 308 bp were selected as unigenes (Table S3). The unigenes were searched against six databases including the KEGG, NR, GO, COG, KOG and Pfam, resulted in a total of 211 490 unigenes annotated with at least one putative function in each database (Table S4). According to the NR homologous species distribution, transcripts of Hippeastrum showed the highest similarity with Asparagus officinalis (39.40%) (Fig. S2).

3.4. Differential gene expression and KEGG enrichment analysis

Comparison of 0 DAI vs. 10 DAI, 10 DAI vs. 20 DAI, and 20 DAI vs. 30 DAI resulted in 26 184 differentially expressed genes (DEGs) (Fig. 3, a). The group of 0 DAI vs. 10 DAI contained the most DEGs (17 280), of which 10 494 were up-regulated, and 6786 were down-regulated. For 10 DAI vs. 20 DAI group, the number of DEGs was 3 484, of which 1 402 and 2 082 were up-regulated and down-regulated, respectively. For group 20 DAI vs. 30 DAI, 5 420 DEGs were identified, with 2 565 being up-regulated and 2 855 being down-regulated. Overlapping analysis of the three comparison groups resulted in 512 overlapped DEGs across the three groups (Fig. 3, b). These results suggested that the initiation stage of SE could require more transcriptional regulations.

Afterward, KEGG enrichment analysis of the three comparison groups was performed. For group 0 DAI vs. 10 DAI (Fig. 3, c), ribosome, phenylpropanoid biosynthesis, glycolysis/gluconeogenesis, and oxidative phosphorylation were the highly enriched KEGG pathways, suggesting that the induction of SE needs a large amount of protein synthesis, secondary metabolites accumulation, and energy supply. For the group of 10 DAI vs. 20 DAI (Fig. 3, d), glutathione metabolism, starch and sucrose metabolism, plant-pathogen interaction, phenylpropanoid biosynthesis, and glycolysis/gluconeogenesis were highly enriched, suggesting that the stress-related signaling, secondary metabolism, and energy homeostasis were actively regulated at this stage. In 20 DAI vs. 30 DAI group (Fig. 3, e), pathways of plant-pathogen interaction, phenylpropanoid biosynthesis, starch and sucrose metabolism and photosynthesis, and plant hormone signal transduction are highly enriched, indicating that in addition to the stress-related signaling and energy homeostasis, plant hormone signaling and the photosynthetic pathway were also involved in the development of this stage.

Time-series expression analysis showed that DEGs formed seven major profiles during SE of 'Bangkok Rose' (Fig. 3, f). Overall, SE's early (0-10 DAI) and late stages (20-30 DAI) presented distinct gene expression patterns. Profile 21 (3286 DEGs), profile 4 (2277 DEGs), profile 5 (2135 DEGs), and profile 20 (1059 DEGs) are the top four DEG-containing modules (Fig. 3, f). Profile 21 was the most powerful module, which showed an up-regulation pattern during 0-10 DAI. Profile 20 showed an up-regulation pattern during 0-10 DAI but a down-regulation pattern during 20-30 DAI. In contrast, profile 4 presented a down-regulation pattern during 0-10 DAI. Profile 5 showed a down-regulation pattern during 0-10 DAI but upregulation during 20-30 DAI (Fig. 3, f). DEGs related to protein processing in the endoplasmic reticulum, and glycerophospholipid metabolism are highly enriched in both profile 21 and profile 20 (Fig. 3, g), indicating that the initiation of SE in the early stage requires a large amount of protein synthesis and activated vesicle trafficking and modification of cell membrane fluidity. In profile 4 and profile 5, DEGs related to ascorbate and aldarate metabolism, oxidative phosphorylation, carbon fixation in photosynthetic organisms, and glycolysis/gluconeogenesis were highly enriched (Fig. 3, g), suggesting that in the early stage of SE, carbon source and energy may not be necessary, as carbon fixation and ATP production might be inhibited (Fig. 3, g). Profile 5 and profile 22 both showed an up-regulation pattern during 20-30 DAI; almost all the pathways showed high enrichment in these two profiles, suggesting that the development of somatic embryos during 20-30 DAI, which was undergoing globular or scuteliar stage (Fig. 1, b, c, f) was regulated via coordination of a large spectrum of pathways.

3.5. Expression of the plant hormone-related genes during SE of 'Bangkok Rose'

To understand the regulatory mechanisms of different hormones during SE of 'Bangkok Rose', expression patterns of varying hormone biosynthesis, transport, and signaling-related genes were analyzed (Fig. 4; Table S5). Auxin/IAA was the most effective hormone for SE induction (Nic-Can and Loyola-Vargas, 2016). IAA biosynthetic gene tryptophan aminotransferaserelated proteins (TAR, two genes) presented an increasing pattern along with SE (0-30 DAI). YUCCA (one gene) slightly increased at 10 DAI but decreased later (Fig. 4, a). Genes of gretchen hagen 3 (GH3, three genes) and dioxygenase for auxin oxidation (DAO, one gene) control the deactivation of IAA; both were up-regulated during SE (10-30 DAI), especially in the early stages (0-10 DAI) (Fig. 4, a). The auxin efflux carrier of PINFORMED (PIN) and influx carrier of AUXIN1/LIKE-AUX1(AUX/ LAX) control the polar distribution of auxin (Leyser, 2018). PIN (PINI and PIN5) genes in 10-30 DAI samples were up-regulated compared to that in 0 DAI samples; however, LAXs (LAX2 and LAX11) were down-regulated, suggesting that PIN-mediated polar auxin distribution plays a vital role during SE. The auxin response factors (ARFs) are the key transcription factors controlling auxin-responsive gene expression (Leyser, 2018). A total of eight ARF genes were identified, and except for ARF5 and ARF16 genes, all were up-regulated along with SE, while other six ARF genes were down-regulated in the early stage of SE (10 and 20 DAI) (Fig. 4, a). Interestingly, the auxin receptor of transport inhibitor response protein 1 (TIRI, three genes) and ARF transcriptional repressor of auxin/indole-3-acetic acid (Aux/IAA, three genes) in 10-30 DAI samples were downregulatedand up-regulated respectively compared to that in 0 DAI samples (Fig. 4, a), however, both regulations led to restriction of auxin signaling during SE.

Cytokinin signaling is thought to play a role in the early globular stage of the somatic embryo in Arabidopsis (Mullerand Sheen, 2008). In this study, a gene of isopentenyl transferase (IPT) encoding a key enzyme that catalyzes cytokinin biosynthesis was found to be increasingly up-regulated during 0-20 DAI, and its expression level peaked at 20 DAI and subsequently dropped at 30 DAI (Fig. 4, b), which is consistent with the accumulation pattem of endogenous CTK in the leaf expiants during SE (Fig. 2, b). CTK influx carriers, including purine permeases (PUPs, nine genes) and equilibrative nucleoside transporter (ENT, one gene) were all up-regulated at the early stages of SE (0-10 DAI) (Fig. 4, b). Similarly, CTK signaling components, including histidine kinase (HK, nine genes) and histidine-containing phosphor transfer proteins (HPts, two genes), both were significantly up-regulated at the initiation stage of SE (10 DAI), suggesting that CTK might play an essential role in the initiation of SE in 'Bangkok Rose'.

ABA was also reported as necessary for initiating SE in Arabidopsis (Su et al., 2013). In this study, we found two ABA biosynthetic genes, including 9-cis-epoxycarotenoid dioxygenase (NCED) and ABA2 were slightly up-regulated at the early stage of SE (10-20 DAI) (Fig. 4, c). UGT71C5 and CYP707A control the deactivation of ABA via glycosylation and catabolism, respectively, UGT71C5 was down-regulated in 10, 20 and 30 DAI samples, and CYP707A was down regulated in 10 and 30 DAI samples copared to that in 0 DAI samples. In contrast, BGs (two genes) encoding ß-glucosidases that transform ABA-GE to active ABA were up-regulated in 10-20 DAI samples. These results suggested that the accumulation of endogenous ABA in leaf expiants during SE (Fig. 2, c) could result from the increased biosynthesis and decreased inactivation/degradation of ABA. Similar to CTK, ABA influx carriers of NRT1/PTR (NPF) (seven genes) were up-regulated in 10 and 20 DAI samples. ABA signaling component phosphatase 2C (PP2C, ten genes) were detected mainly with varied expression patterns. While SNFl-related protein kinase 2 (SnRK2, four genes), which drives the ABA response via phosphorylating the downstream substrates, were up-regulated in 10-30 DAI samples compared to that in 0 DAI samples (Fig. 4, c), suggesting an enhanced ABA signaling during SE.

GA3 is also involved in the induction of SE, even though its function could be varied in distinct species (Elhiti et al., 2013). In this study, we identified three GA biosynthetic genes, including ent-kaurene oxidase (KO) ent-kaurenoic acid oxidase (KAO), and GA20-oxidase (GA2Oox) (Fig. 4, d). KO was down-regulated in 10-20 DAI samples, KAO was down-regulated in 10-30 DAI samples compared to that in 0 DAI samples, while GA20ox was highly expressed at 30 DAI, which is consistent with the accumulation of endogenous GA during SE (Fig. 2, d), suggesting that the GA20ox could function as a rate-limiting enzyme in the biosynthesis of GA in Hippeastrum. GA3 signaling components, including GA insensitive dwarf 1 (GIDI), GID2, DELLA, and SLY1SPINDLY (SPY), were also detected in the transcriptome (Fig. 4, d). GIDI (one out of three genes) and GID2 (two genes) were slightly up-regulated in the early stage of SE (10 DAI); this could lead to the enhancement of GA signaling via inhibition of DELLA, which is the key negative regulator of GA signaling. Genes of DELLA and another GA negative regulator SPY were also detected in this study, two out of three DELLA genes were down-regulated, and one was up-regulated in 10-30 DAI samples compared to that in 0 DAI samples. SPY showed varied expression patterns among different transcripts during SE (0-30 DAI).

The endogenous level of SA was increased during SE of 'Bangkok Rose' (Fig. 2, e). This could result from the up-regulation of the SA biosynthetic gene phenylalanine ammonia-lyase (PAL) (6 genes were up-regulated in 10 DAI samples, out of which 5 genes were up-regulated in both 10 and 20 DAI samples) (Fig. 4, e). SA signaling-related genes were not detected in the transcriptome, which indicates that SA signaling was not involved in SE of 'Bangkok Rose', at least not engaged via transcription regulation.

Even though many genes related to ETH biosynthesis and signaling were detected, we cannot conclude whether they have a role in the SE of 'Bangkok Rose' as their expression did not show particular patterns along with SE and sometimes even contradicted each other. For example, the biosynthetic gene S-adenosyl methionine synthetase (SAMS) was up-regulated in 10-30 DAI samples compared to that in 0 DAI samples.; However, the downstream enzyme ACC oxidase (ACCO) was down-regulated in 10-30 DAI samples compared to that in 0 DAI samples. Fourteen ethylene response factors (ERFs) were detected; among them, four were up-regulated and ten were down-regulated in 10-30 DAI samples compared to that in 0 DAI samples (Fig. S3, a).

JA was supposed to be not essential for SE of 'Bangkok Rose' as endogenous JA concentration in leaf expiants was dramatically reduced after inoculation (Fig. 2, g). The transcriptomic data supported this hypothesis as almost all the genes related to JA biosynthesis, metabolism, and signaling were down-regulated in 10-30 DAI samples compared to that in 0 DAI samples (Fig. S3, b).

3.6. Expression of the stress-related genes during SE of 'Bangkok Rose'

It has been known that the onset of SE is influenced by the expression of hundreds of stress-related genes in response to the effects of in vitro culture conditions (Nic-Can et al., 2016). This study found that various stress-related genes were induced in 'Bangkok Rose' leaf expiants during SE (Fig. 5; Table S6). Glutatione-S-transferase (GST), thaumatin and peroxidase are associated with oxidative stress control in responding to biotic and abiotic stresses (Doroshow, 1995; Vaish et al., 2020; Park and Kim, 2021). In this study, we identified 12 GST, 5 thaumatin genes, and 18 peroxidases, and most of them were up-regulated in 10-30 DAI samples compared to that in 0 DAI samples, especially in the early stages of SE (10 DAI or 20 DAI) (Fig. 5, a-c). Wound-induced cell wall proteins of expansin were involved in cell wall extension during plant growth and were also found to be essential regulators in response to abiotic stresses (Zhou et al., 2015; Jadamba et al., 2020). This study identified nine expansin genes, all of them wereup-regulated in the 10 DAI samples, six of them were up-regulated in 10-20 DAI samples, and five of them were up-regulated in 10-30 DAI samples compared to that in 0 DAI samples (Fig. 5, d). Chitinases play a critical part against biotic and abiotic stresses (Vaghela et al., 2022), and are also supposed to be associated with SE (Fráterová et al., 2013). This study identified eight chitinase genes, and seven of them were up-regulated in 10-30 DAI samples compared to that in 0 DAI samples Fig. 5, e). Like the oxidative-related genes and expansin, chitinases were also more remarkably induced in the early stages (10-20 DAI ) than in the late stage (30 DAI). Heat-shock proteins (Hsps) are well-known stress proteins that were involved in response to almost every biotic/abiotic stress (Al-Whaibi, 2011). In this study, we identified 30 Hsps, among which 13 of them including HSP7M-2, HSP7M-4, HSP70-2, HSP70-4, HSP80-3, HSP81, HSP81-2, HSP81-3, HSP81-4, HSP81-5, HSP90-4 and HSP90-5 were remarkably up-regulated in 10-30 DAI samples compared to that in 0 DAI samples. Interestingly, most up-regulated Hsps showed peak expression levels at 30 DAI (Fig. 5, f), which is different from the other stress-related genes mentioned above. Germin-like protein (GLP) was another stress-related protein family that was found to be expressed during SE (Mathieu et al., 2006; Barman and Banerjee, 2015). Consistently, we also identified nine GLP genes in the transcriptome, and sixof them were dramatically induced in 10-30 DAI samples compared to that in 0 DAI samples (Fig. 5, g). In conclusion, our analysis revealed that the SE of 'Bangkok Rose' was correlated with various stress responses at the transcriptional level.

3.7. Expression of the epigenetics and SE-related genes during SE of 'Bangkok Rose'

This study identified numerous genes regulating DNA methylation and chromatin re-modeling (Fig. 6; Table S7). Domains-rearranged methyltransferases (DRMs) belong to a subfamily of DNA (cytosine-5)-methyl transferases (DNMTs) and are responsible for the de nouo methylation of DNA by miRNAdirected pathway (Pavlopoulou and Kossida, 2007). Five DRM genes were identified in the transcriptome of H. 'Bangkok Rose', and they were dramatically down-regulated in 10-20 DAI, but start to be up-regulated at 30 DAI (Fig. 6, a). A substantial number of genes encoding histone modification enzymes, including histone-lysine N-methyltransferase (HKMTases), Lysine-specific demethylase (LSD), histone acetyltransferase (HAT) and histone deacetylase (HDAC), were identified in this study. ASH and TRX proteins are SET domain-containing HKMTase (Yu et al., 2009) that catalyze the methylation of histone. In contrast, Jumonji (JMJ) domain-containing histone lysine demethylases and their homologs in rice, SE14s catalyze the demethylation of histone (Klose et al., 2006; Yokoo et al., 2014). These two antagonistic protein groups showed similar gene expression patterns during SE. Most of the genes in these two groups were down-regulated in the early stages(0-10 DAI) and later up-regulated in the late stage(20-30 DAI) (Fig. 6, b, c), which is similar to the gene expression pattern of DRMs (Fig. 6, a). HAT and HDAC antagonistically catalyze the acetylation and deacetylation of histone, respectively (Nitsch et al., 2021). Intriguingly, these two gene groups also showed consistent expression patterns that were either up-regulated or down-regulated in 10-20 DAI samples, and all up-regulated at the late stage of SE (30 DAI) (Fig. 6, d, e). These results suggested that histone methylation and acetylation were dynamically regulated in various stages of SE.

It has been reported that chromatin modifications could lead to expression level alteration of several SE marker genes, including WUSCHEL-related homeobox (WOX), LEG, and BBM (Li et al., 2011; Duarte-Aké etai., 2019). This study detected many WOX (five genes), CLAVATA (CLV, six genes), and somatic embryogenesis receptor-like kinase (SERK, seven genes) genes. In this study, we found five WOX genes were induced in 10-20 DAI samples compared to that in 0 DAI samples, among which W0X6 and W0X9 showed peak expression levels at 30 DAI, while the expression level of W0X8 peaked at 10 DAI (Fig. 6, f). In contrast, CLVs (six genes) were down-regulated in 10-20 DAI, but their expression levels slightly increased at 30 DAI (Fig. 6, g). We identified seven SERK genes, of which four (SERKI-2, SERKI-3, SERK4-1, SERK4-2) were significantly induced at 30 DAI, and two (SERKI-1, SERK2) showed peak expression level at 10 DAI, suggesting that the function of different SERK genes could be spatiotemporally varied.

3.8. Validation of gene expression using qRT-PCR

Twelve DEGs were selected for quantitative real-time polymerase chain reaction (qRT-PCR) to confirm the accuracy and reproducibility of RNA-Seq results (see Fig. 7). Seven DEGs including ORR9 (TRINITY_DN775_cO_gl), ASK9 (TRINITY_DN30453_c0_gl), IAA1 (TRINITY_DN6982_cO_gl), IAA17 (TRINITY_DN901_c0_gl), PINI (TRINITY_DN11056_c0_gl), GID2 (TRINITY_DN1738_cO_g2) and TIRI (TRINITY_DN16577_cO_g2) are related to plant hormones. Four DEGs, including SERKI (TRINIT Y_DN566l_c0_g2), W0X6 (TRINITY_DN18256_cO_g2), W0X8 (TRINITY_DN10376_c0_gl), and CLE13 (TRINITY_DN17689_cO_g2) are key regulators in somatic embryogenesis. One DEG CYP92A6 (TRINITY_DN2625_cO_gl) belongs to the P450 cytochrome. In 10 out of 12 genes, the gene expression patterns determined by qRT-PCR were consistent with RNA-Seq results. Two genes, ASK9 and W0X8 showed minor variation at 10 DAI and 20 DAI, respectively. These results confirmed that the transcriptome data was dependable and accurate.

4. Discussion

In vitro tissue culture propagation of Hippeastrum using scape (Seabrook and Cumming, 1977), peduncle, pedicel (Kim et al., 2005), and scale (Amani et al., 2015) as expiants have been reported in previous studies. While the leaf from the field or greenhouse as explant has yet to be established. Compared to these floral organs and bulbs, leaves are low-cost and readily obtainable materials considered ideal for tissue culture propagation. In a previous study, we found that the regeneration efficiency from leaf expiants depends highly on varieties. Only two cultivars, including 'Bangkok Rose' and 'Dancing Queen' out of over 100 cultivars, were efficiently reproduced from leaf expiants via in vitro tissue culture.

Moreover, somatic embryogenesis was successfully induced only when the leaf bases were used as expiants, and we failed to induce SE using the middle and tip segments of the leaf. This is probably because a meristem-like region was maintained in the base region of plant leaves to provide new cells for differentiation via constant cell proliferation (Gonzalez et al., 2012). Plants can be regenerated through either DSE or ISE or a mixture of the two pathways according to the tissue culture condition and explant type (von Arnold et al., 2002)., two types of SE pathways have been successfully established in many species include Curcuma amada, Digitalis lanata, and Camellia oleífera (Shajahan et al., 2016; Bhusare et al., 2020; Zhang et al., 2021). The induction of ISE is highly dependent on 2,4-D, which functions as a potent enhancer for callus induction and proliferation. In this study, we established a DSE method for 'Bangkok Rose' on MS medium supplemented with 1.00 mg * L 1 NAA, 1.00 mg * L 1 KT, and 0.25 mg * L 1TDZ. Even though this combination of PGRs was the best optimal in terms of SE and the adventitious bud regeneration ratio, we cannot rule out the possibility that 'Bangkok Rose' could also be reproduced via ISE pathway as 2,4-D has not been evaluated in our study. Furthermore, ISE from plantlet leaves of 'Blossom Peacock' and 'Royal Velvet' has been reported in a previous study (Wang et al., 2020).

AUX, CTK, ABA, and JA are the most commonly used PGRs for SE induction in most plant species (Gulzar et al., 2020). Among all PGRs, AUX and CTK are the key factors triggering the embryogenic response due to their strong regulatory effect on cell division and differentiation (Nic-Can and Loyola-Vargas, 2016; Shi et al., 2020). It is well recognized that the balance between AUX and CTK determines the state of cell di- or dedifferentiation (Elhiti and Stasolla, 2022). In this study, we found IAA (the most abundant AUX in higher plants) and CTK were significantly accumulated along with SE, and their concentrations peaked at 20 DAI and then dropped slightly at 30 DAI. AUX and CTK have been reported to act synergistically in the shoot and root development regulation in Arabidopsis (Jones et al., 2010). It is plausible that a similar mechanism occurs in the regulation of SE. In our transcriptomic analysis, the IAA biosynthetic enzyme TAR showed an increasing expression pattern and peaked at 30 DAI; however, the downstream enzyme YUCCA was significantly down-regulated at 30 DAI. This could explain why the biosynthesis of IAA was slightly decreased at 30 DAI compared to 20 DAI. It has been reported that the gradual accumulation of IAA in the somatic embryos is crucial for SE. The auxin gradients were supposed to be established by PINl-mediated polar auxin transport (Su et al., 2009). In 'Bangkok Rose', the polar auxin transport in somatic embryos could be the responsibility of PINI and PIN5, as both these genes were significantly up-regulated along with SE. In carrot (Michalczuk et al., 1992), conifer (Vondráková et al., 2011), white oak (Corredoira et al., 2017) and Pepper (PérezPastrana et al., 2021), auxin was supposed to function in the SE. However, this seems like not the case in Hippeastrum, as we found the auxin transcription factor of ARFs (six out of eight genes) were significantly down-regulated in the early stages of SE.

Moreover, auxin receptors of TIRls were also down-regulated, and ARF transcriptional repressors of Aux/IAA were up-regulated during SE (10-30 DAI), especially in the early stages of SE (0-10 DAI); all these transcriptional regulations led to the restriction of auxin signaling in the early stages of SE (10-20 DAI). Interestingly, the biosynthesis of auxin was induced in the early stages (10-20 DAI), this could be due to the up-regulation of auxin biosynthesis genes like TARs and YUCCAs in the early stages of SE. Still, the downstream signaling components were transcriptionally repressed, implying a possible feedback regulation mode of auxin during SE of'Bangkok Rose'. In contrast to auxin, the downstream CTK signaling components, including HK, HPts, and the influx carriers of PUPs and ENT were significantly up-regulated in the early stage of SE, suggesting a strong CTK signaling output in the early stage of SE. It is well documented in many studies that a high CTK: auxin ratio induces shoots, while a low ratio generates roots (Elhiti and Stasolla, 2022). Consistent with this notion, our study revealed that a high CTK: auxin signaling output ratio is required for the early development of somatic embryos from leaf expiants of 'Bangkok Rose'.

Besides IAA and CTK, ABA and SA were also remarkably accumulated along with SE of H.'Bangkok Rose'. ABA was recently reported to regulate SE via the ABSCISIC ACID INSENSITIVE 3 (ABI3) and ABI4 transcription factors (Chen et al., 2021). Even though ABI genes were not detected in our transcriptomic analysis, they could be activated via phosphorylation by snRK2 kinases, as several genes encoding snRK2 were significantly induced along with the SE of 'Bangkok Rose'. The role of SA in SE varied among distinct species and could act as dose-dependent (Roustan et al., 1990; Hong et al., 2008; Grzyb et al., 2018). In 'Bangkok Rose', except for the SA biosynthetic genes of PALs, we did not detect any SA-related genes in the transcriptomic analysis, suggesting that the involvement of SA in SE may not be mediated via transcriptional regulations.

Induction of genes related to oxidative stress and redox homeostasis during the onset of SE was commonly observed in soybean (Thibaud-Nissen et al., 2003), conifer (de Vega-Bartol et al., 2013) and maize (Salvo et al., 2014). Similarly, in 'Bangkok Rose', we also detected many oxidative-related genes, including GSTs, thaumatin, and peroxidases, which were notably upregulated during the early stages of SE (0-10 DAI). GLPs are extracellular matrix-associated proteins that were found to be involved in redox homeostasis (Barman and Banerjee, 2015). We also found many GLP genes that were up-regulated during SE of 'Bangkok Rose', especially in the early stage (0-10 DAI). It has been well-studied in conifer that GLPs are recognized as early SE markers due to their expressions being highly associated with the early growth stages of SE (Mathieu et al., 2006). Unlike these early responsive genes mentioned above, the Hsps are likely late responsive, as most of the Hsps detected in this study are significantly up-regulated at 30 DAI. This is consistent with the previous studies in cork oak and white spruce that Hsps including Hspl7 in cork oak (Puigderrajols et al., 2002) and PgEMB22, PgEMB27, and PgEMB29 in white spruce (Dong and Dunstan, 1996) were highly accumulated at the late stages of SE.

Epigenetic mechanisms were recently found to be critical factors during plant SE (Kumar and van Staden, 2017). Several transcription factor genes, W0X4, BBM1, and LECI, crucial for the induction and development of SE were found to be epigenetically regulated by H3K27me3 and H3K9me2 (Li et al., 2011; Duarte-Aké et al., 2019). In this study, we found that genes control histone methylation (ASHHs and TRXs) showed similar expression patterns to those that control histone demethylation (SEs and JMJs), and genes control acetylation (HACs and HATs) showed similar expression pattem to those that control deacetylation (HDTs and HDAs) during SE. This result demonstrated that histone methylation and acetylation were under dynamic regulation via the co-expression of the antagonistic enzymes during SE of'Bangkok Rose'. Moreover, the SE of 'Bangkok Rose' might require a low H3K27me3 and H3K9me2 level in the early stages and high acetylation modifications at the late stage of SE. Consistent with our result, low levels of H3K27me3 and H3K9me2 were necessary for the induction and development of SE in many studies (Grafi et al., 2007; Lafos et al., 2011; Nic-Can et al., 2013).

Epidermal patterning factor (EPF) and EPF-like (EPFL) are plant-specific secretory peptides that control stomata development by regulating the division and differentiation of the stomatai stem cell in plant epidermis (Hunt and Gray, 2009; Hunt et al., 2010). In 'Bangkok Rose', we have identified several EPF/ EPFL genes that are significantly up-regulated (EPFL2, EPFL3-1, EPFL3-2, EPFL5, EPFL6, EPFL8) or down-regulated (EPFL1, EPF1, EPFL9) during SE (Fig. S3, c; Table S8). Since we have found that the SE of 'Bangkok Rose' is initiated in the epidermal cell layer, EPF/EPFL peptides could contribute to the amplifying and further differentiation of the embryogénie stem cells in the epidermis. However, the involvement of EPF/EPFL peptides in SE has been rarely reported, and need to be further validated in the future.

In conclusion, our study revealed a transcriptional regulatory network governing SE of 'Bangkok Rose' (Fig. 8). In the early developmental stages of SE (10-20 DAI), auxin signaling was suppressed. In contrast, CTK signaling and biosynthesis were both promoted to allow a high CTK: auxin signaling ratio, which is required for the early development of somatic embryos (Elhiti and Stasolla, 2022). Meanwhile, stress-responsive genes, including oxidative-related genes, expansin, and chitinases were up-regulated to trigger SE initiation. In addition, the dynamic histone methylation was low, which is also crucial for somatic embryo development (Grafi et al., 2007; Lafos et al., 2011; Nic-Can et al., 2013). In the late stage of SE (30 DAI), heat-shock proteins were up-regulated, and the dynamic histone acetylation also increased to promote the maturation of the somatic embryos via yet unknown mechanisms. Along with SE, the key regulators, including WOXs (W0X6, W0X8, and W0X9), and SERKs (SERKI and SERK4) were highly expressed to regulate the induction and maturation of the Hippeastrum somatic embryos.

Declaration of interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work was funded by Guangdong Basic and Applied Basic Research Foundation (Grant No. 2023A1515010237), the 2021 Dongguan Provincial Rural Revitalization Program (Grant No. 20211800400022), the Guangdong Key Technology Research and Development Program (Grant Nos. 2020B020220005; 2022B1111040003), the Guangdong Modern Agricultural Industry Technology System Program (Grant No. 2023KJ121), the South China Botanical Garden, the Chinese Academy of Sciences (Grant No. QNXM-02).

Supplementary materials

Supplementary data to this article can be found online at https://doi.Org/10.1016/j.hpj.2023.02.013.

Sidebar

Received 13 May 2023; Received in revised form 3 July 2023; Accepted 25 September 2023

Available online 17 November 2023

Footnote

* Corresponding authors.

E-mail addresses: [email protected]; [email protected]

Peer review under responsibility of Chinese Society of Horticultural Science (CSHS) and Institute of Vegetables and Flowers (IVF), Chinese Academy of Agricultural Sciences (CAAS)

References

REFERENCES

Al-Whaibi, M.H., 2011. Plant heat-shock proteins: a mini review. J King Saud Univ Sci, 23: 139-150.

Amani, S., Zarei, H., Azar, A.M., Mashayekhi, K., 2015. Micropropagation of Hippeastrum hybridům. Cumhuriyet Sei J, 36: 594-605.

Bao, Y., Liu, G., Shi, X., Xing, W., Ning, G., Liu, J., Bao, M., 2012. Primary and repetitive secondary somatic embryogenesis in Rosa hybrida 'Samantha'. Plant Cell Tissue Organ Cult, 109: 411-418.

Barman, A.R., Banerjee, J., 2015. Versatility of germin-like proteins in their sequences, expressions, and functions. Punct Integr Genom, 15: 533-548.

Bhatia, S., Bera, T., 2015. Somatic embryogenesis and organogenesis. In: Bhatia, S., Sharma, K., Dahiya, R., Bera, T. (Eds.), Modern Applications of Plant Biotechnology in Pharmaceutical Sciences. Academic Press, Boston, pp. 209-230.

Bhusare, B.P., John, C.K., Bhatt, V.P., Nikam, T.D., 2020. Induction of somatic embryogenesis in leaf and root expiants of Digitalis lanata Ehrh.: direct and indirect method. South Afr J Bot, 130: 356-365.

Boutilier, K., Offringa, R., Sharma, V.K., Kieft, H., Ouellet, T., Zhang, L., Hattori, J., Liu, C.M., van Lammeren, A.A.M., Miki, B.L.A., Custers, J.B.M., van Lookeren Campagne, M.M., 2002. Ectopic expression of baby boom triggers a conversion from vegetative to embryonic growth. Plant Cell, 14: 1737-1749.

Braybrook, S.A., Stone, S.L., Park, S., Bui, A.Q., Le, B.H., Fischer, R.L., Goldberg, R.B., Harada, J.J., 2006. Genes directly regulated by LEAFY COTYLEDON2 provide insight into the control of embryo maturation and somatic embryogenesis. Proc Natl Acad Sei USA, 103: 3468-3473.

Chen, B., Fiers, M., Dekkers, B.J.W., Maas, L., van Esse, G.W., Angenent, G.C., Zhao, Y., Boutilier, K., 2021. ABA signalling promotes cell totipotency in the shoot apex of germinating embryos. J Exp Bot, 72: 6418-6436.

Corredoira, E., Cano, V., Bárány, L, Solis, M.T., Rodriguez, H., Vieitez, A.M., Risueño, M.C., Testillano, P.S., 2017. Initiation of leaf somatic embryogenesis involves high pectin esterification, auxin accumulation and DNA demethylation in Quercus alba. J Plant Physiol, 213: 42-54.

Datta, S.K., 2021. Amaryllis/hippeastrum. In: Datta, S.K., Gupta, Y.C. (Eds.), Floriculture and Ornamental Plants. Springer, Berlin, pp. 1-27.

de Vega-Bartol, J.J., Santos, R.R., Sirnoes, M., Miguel, C.M., 2013. Normalizing gene expression by quantitative PCR during somatic embryogenesis in two representative conifer species: Pinus pinaster and Picea abies. Plant Cell Rep, 32: 715-729.

Dong, J.Z., Dunstan, D.I., 1996. Characterization of three heat-shock-protein genes and their developmental regulation during somatic embryogenesis in white spruce [Picea glauca (Moench) Voss]. Planta, 200: 85-91.

Doroshow, J.H., 1995. Glutathione peroxidase and oxidative stress. Toxicol Lett, 82-83: 395-398.

Duarte-Aké, F., Nic-Can, G., De-la-Peña, C., 2019. Somatic Embryogenesis: Polycomb Complexes Control Cell-To-Embryo Transition. In: Alvarez-Venegas, R., De-la-Peña, C., Casas-Mollano, J.A. (Eds.), Epigenetics in Plants of Agronomic Importance: Fundamentals and Applications. Springer, Berlin, pp. 339-354.

Elhiti, M., Stasolla, C., 2022. Transduction of signals during somatic embryogenesis. Plants, 11: 178.

Elhiti, M., Stasolla, C., Wang, A., 2013. Molecular regulation of plant somatic embryogenesis. In Vitro Cell Dev Biol Plant, 49: 631-642.

Fountain, W., 1979. Investigations of in vitro culture of Hippeastrum as a method of rapid propagation [Ph. D. Dissertation]. Louisiana: Louisiana State University and Agricultural and Mechanical College.

Fráterová, L., Salaj, T., Matusikova, L, Salaj, J., 2013. The role of chitinases and glucanases in somatic embryogenesis of black pine and hybrid firs. Cent Eur J Biol, 8: 1172-1182.

Gonzalez, N., Vanhaeren, H., Inzé, D., 2012. Leaf size control: complex coordination of cell division and expansion. Trends Plant Sei, 17: 332-340.

Grafi, G., Ben-Meir, H., Avivi, Y., Moshe, M., Dahan, Y., Zemach, A., 2007. Histone methylation controls telomerase-independent telomere lengthening in cells undergoing dedifferentiation. Dev Bio, 306: 838-846.

Grzyb, M., Kalandyk, A., Mikuła, A., 2018. Effect of tiba, fluridone and salicylic acid on somatic embryogenesis and endogenous hormone and sugar contents in the tree fern Cyathea delgadii Sternb. Acta Physiol Plant, 40: 1-11.

Gulzar, B., Mujib, A., Malik, M.Q., Sayeed, R., Mamgain, J., Ejaz, B., 2020. Genes, proteins and other networks regulating somatic embryogenesis in plants. J Genet Eng Biotechnol, 18: 31.

Haque, S.M., Ghosh, B., 2017. Regeneration of cytologically stable plants through dedifferentiation, redifferentiation, and artificial seeds in Spathoglottis plicata Blume. (Orchidaceae). Hortic Plant J, 3: 199-208.

Hong, P.I., Chen, J.T., Chang, W.C., 2008. Effects of salicylic and acetylsalicylic acid on direct somatic embryogenesis in Oncidium. J Plant Biochem Biotechnol, 17: 149-153.

Hunt, L., Bailey, KJ., Gray, J.E., 2010. The signalling peptide EPFL9 is a positive regulator of stomatai development. New Phytol, 186: 609-614.

Hunt, L., Gray, J.E., 2009. The signaling peptide EPF2 controls asymmetric cell divisions during stomatai development. Curr Bio, 19: 864-869.

Ikeuchi, M., Ogawa, Y., Iwase, A., Sugimoto, K., 2016. Plant regeneration: cellular origins and molecular mechanisms. Dev, 143: 1442-1451.

Iqbal, S., Pan, Z., Hayat, F., Bai, Y., Coulibaly, D., Ali, S., Ni, X., Shi, T., Gao, Z., 2021. Comprehensive transcriptome profiling to identify genes involved in pistil abortion of Japanese apricot. Physiol Mol Biol Plants, 27: 1191-1204.

Iqbal, S., Bai, Y., Hayat, F., Coulibaly, D., Khalil-ur-Rehman, M., Shi, T., Gao, Z., 2022. Comprehensive transcriptomic and metabolomic analysis revealed distinct flavonoid biosynthesis regulation during abnormal pistil development in Japanese apricot. Genomics, 114: 110451.

Jadamba, C., Kang, K., Paek, N.C., Lee, S.L, Yoo, S.C., 2020. Overexpression of rice Expansin? (Osexpa?) confers enhanced tolerance to salt stress in rice. Int J Mol Sei, 21: E454.

Jarillo, J.A., Pineiro, M., Cubas, P., Martinez-Zapater, J.M., Jarillo, J.A., Pineiro, M., Cubas, P., Martinez-Zapater, J.M., 2009. Chromatin remodeling in plant development. Int J Dev Biol, 53: 1581-1596.

Jha, P., Ochatt, S.J., Kumar, V., 2020. Wuschel: a master regulator in plant growth signaling. Plant Cell Rep, 39: 431-444.

Jones, B., Gunnerås, S.A., Petersson, S.V., Tarkowski, P., Graham, N., May, S., Doležal, К., Sandberg, G., Ljung, К., 2010. Cytokinin regulation of auxin synthesis in Arabidopsis involves a homeostatic feedback loop regulated via auxin and cytokinin signal transduction. Plant Cell, 22: 2956-2969.

Kim, MJ., Kim, Y.J., Kim, H.S., Ko, J.A., 2005. Effective in vitro propagation from pedicel culture of Hippeastrum hybridům Hort. 'Dazzler'. Korean J Plant Resources, 18: 382-389.

Klose, R.J., Kailin, E.M., Zhang, Y., 2006. JmjC-domain-containing proteins and histone demethylation. Nat Rev Genet, 7: 715-727.

Kumar, V., van Staden, J., 2017. New insights into plant somatic embryogenesis: an epigenetic view. Acta Physiol Plant, 39: 194.

Kumar, V., van Staden, J., 2019. Multi-tasking of serk-like kinases in plant embryogenesis, growth, and development: current advances and biotechnological applications. Acta Physiol Plant, 41: 31.

Lafos, M., Kroll, P., Hohenstatt, M.L., Thorpe, F.L., Clarenz, O., Schubert, D., 2011. Dynamic regulation of H3K27 trimethylation during Arabidopsis differentiation. PLoS Genet, 7: 61002040.

Leyser, O., 2018. Auxin signaling. Plant Physiol, 176: 465-479.

Li, W., Liu, H., Cheng, Z.J., Su, Y.H., Han, H.N., Zhang, Y., Zhang, X.S., 2011. DNA methylation and histone modifications regulate de novo shoot regeneration in Arabidopsis by modulating WUSCHEL expression and auxin signaling. PLoS Genet, 7: e!002243.

Livak, KJ., Schmittgen, T.D., 2001. Analysis of relative gene expression data using real-time quantitative PGR and the 2-ллст method. Meth, 25: 402-408.

Liu, R., Xue, Y., Ci, H., Gao, J., Wang, S., Zhang, X., 2022. Establishment of highly efficient plant regeneration of Paeonia ostii 'Fengdan' through optimization of callus, adventitious shoot, and rooting induction. Hortic Plant J, 8: 777-786.

Mahdavi-Darvari, F., Noor, N.M., Ismanizan, L, 2015. Epigenetic regulation and gene markers as signals of early somatic embryogenesis. Plant Cell Tiss Org, 120: 407-422.

Mathieu, M., Lelu-Walter, M.A., Blervacq, A.S., David, H., Hawkins, S., Neutelings, G., 2006. Germin-like genes are expressed during somatic embryogenesis and early development of conifers. Plant Mol Biol, 61: 615-627.

Mendez-Hernandez, H.A., Ledezma-Rodriguez, M., Avilez-Montalvo, R.N., Juarez-Gomez, Y.L., Skeete, A., Avilez-Montalvo, J., De-la-Pena, C., Loyola-Vargas, V.M., 2019. Signaling overview of plant somatic embryogenesis. Front Plant Sci, 10: 77.

Michalczuk, L., Cooke, T.J., Cohen, J.D., 1992. Auxin levels at different stages of carrot somatic embryogenesis. Phytochemistry, 31: 1097-1103.

Muller, B., Sheen, J., 2008. Cytokinin and auxin interaction in root stem-cell specification during early embryogenesis. Nat, 453: 1094-1097.

Naing, A.H., Kim, C.K., Yun, B.J., Jin, J.Y., Lim, K.B., 2013. Primary and secondary somatic embryogenesis in Chrysanthemum cv. Euro. Plant Cell Tiss Org, 112: 361-368.

Nic-Can, G.I., Avilez-Montalvo, J.R., Aviles-Montalvo, R.N., Márquez-López, R.E., Mellado-Mojica, E., Galaz-Ávalos, R.M., Loyola-Vargas, V.M., 2016. The Relationship between Stress and Somatic Embryogenesis. Somatic Embryogenesis: Fundamental Aspects and Applications. Springer, Berlin, pp. 151-170.

Nic-Can, G.I., López-Torres, A., Barredo-Pool, F., Wrobel, K., Loyola-Vargas, V.M., Rojas-Herrera, R., De-la-Peha, C., 2013. New insights into somatic embryogenesis: LEAFY COTYLEDONI, BABY BOOM1 and WUSCHEL-RELATED HOMEOBOX4 are epigenetically regulated in Coffea canephora. PLoS One, 8: e72160.

Nic-Can, G.I., Loyola-Vargas, V.M., 2016. The role of the auxins during somatic embryogenesis. In: Loyola-Vargas, V.M., Ochoa-Alejo, N. (Eds.), Somatic Embryogenesis: Fundamental Aspects and Applications. Springer, Berlin, pp. 171-182.

Nitsch, S., Zorro Shahidian, L., Schneider, R., 2021. Histone acylations and chromatin dynamics: concepts, challenges, and links to metabolism. EMBO Rep, 22: e52774.

Park, EJ., Kim, Т.Н., 2021. Thaumatin-like genes function in the control of both biotic stress signaling and ABA signaling pathways. Biochem Bioph Res Co, 567: 17-21.

Pavlopoulou, A., Kossida, S., 2007. Plant cytosine-5 DNA methyltransferases: structure, function, and molecular evolution. Genomics, 90: 530-541.

Pérez-Pastrana, J., Testillano, P.S., Barany, L, Canto-Flick, A., Álvarez-López, D., Pijeira-Fernández, G., Avilés-Viñas, S.A., Peña-Yam, L., Muñoz-Ramírez, L., Nahuat-Dzib, S., Islas-Flores, L, Santana-Buzzy, N., 2021. Endogenous auxin accumulation/localization during zygotic and somatic embryogenesis of Capsicum chínense Jacq. J Plant Physiol, 258: 153333.

Pontvianne, F., Blevins, T., Pikaard, C.S., 2010. Arabidopsis histone lysine methyltransferases. Adv Bot Res, 53: 1-22.

Puigderrajols, P., Jofré, A., Mir, G., Pla, M., Verdaguer, D., Huguet, G., Molinas, M., 2002. Developmentally and stress-induced small heat shock proteins in cork oak somatic embryos. J Exp Bot, 53: 1445-1452.

Robert, H.S., Grones, P., Stepanova, A.N., Robles, L.M., Lokerse, A.S., Alonso, J.M., Weijers, D., Friml, J., 2013. Local auxin sources orient the apical-basal axis in Arabidopsis embryos. Gurr Biol, 23: 2506-2512.

Roustan, J.P., Latche, A., Fallot, J., 1990. Inhibition of ethylene production and stimulation of carrot somatic embryogenesis by salicylic acid. Biol Plant (Prague), 32: 273-276.

Salaün, G., Lepiniec, L., Dubreucq, B., 2021. Genetic and molecular control of somatic embryogenesis. Plants, 10: 1467.

Salvo, S.A.G.D., Hirsch, C.N., Buell, C.R., Kaeppler, S.M., Kaeppler, H.F., 2014. Whole transcriptome profiling of maize during early somatic embryogenesis reveals altered expression of stress factors and embryogenesis-related genes. PLoS One, 9: elll407.

Schmidt, E.D., Guzzo, F., Toonen, M.A., de Vries, S.C., 1997. A leucine-rich repeat containing receptor-like kinase marks somatic plant cells competent to form embryos. Dev, 124: 2049-2062.

Seabrook, J.E.A., Cumming, B.G., 1977. The in vitro propagation of amaryllis (Hippeastrum spp. Hybrids). In Vitro, 13: 831-836.

Shajahan, A., Raju, C.S., Thilip, G., Varutharaju, K., Faizal, K., Mehaboob, V.M., Aslam, A., 2016. Direct and Indirect Somatic Embryogenesis in Mango Ginger (Curcuma Amada Roxb.). In: Loyola-Vargas, V.M., Ochoa-Alejo, N. (Eds.), Somatic Embryogenesis: Fundamental Aspects and Applications. Springer, Berlin, pp. 367-379.

Shi, T., Iqbal, S., Ayaz, A., Bai, Y., Pan, Z., Ni, X., Hayat, F., Saqib Bilal, M., Khuram Razzaq, M., Gao, Z., 2020. Analyzing differentially expressed genes and pathways associated with pistil abortion in Japanese apricot via RNA-Seq. Genes, 11: 1079.

Singh, R., Dubey, A.K., Sanyal, L, 2019. Optimisation of adventitious shoot regeneration and Agrobacterium-mediated transformation in Canna x generalis (Canna Lily). Hortic Plant J, 5: 39-46.

Somssich, M., Je, B.I., Simon, R., Jackson, D., 2016. CLAVATA-WUSCHEL signaling in the shoot meristem. Dev, 143: 3238-3248.

Su, Y.H., Su, Y.X., Liu, Y.G., Zhang, X.S., 2013. Abscisic acid is required for somatic embryo initiation through mediating spatial auxin response in Arabidopsis. Plant Growth Regül, 69: 167-176.

Su, Y.H., Zhao, X.Y., Liu, Y.B., Zhang, C.L., O'Neill, S.D., Zhang, X.S., 2009. Auxin-induced WUS expression is essential for embryonic stem cell renewal during somatic embryogenesis in Arabidopsis. Plant J, 59: 448-460.

Thakare, D., Tang, W., Hill, K., Perry, S.E., 2008. The MADS-domain transcriptional regulator AGAMOUS-LIKE15 promotes somatic embryo development in Arabidopsis and soybean. Plant Physiol, 146: 1663-1672.

Thibaud-Nissen, F., Shealy, R.T., Khanna, A., Vodkin, L.O., 2003. Clustering of microarray data reveals transcript patterns associated with somatic embryogenesis in soybean. Plant Physiol, 132: 118-136.

Vaghela, B., Vashi, R., Rajput, К., Joshi, R., 2022. Plant chitinases and their role in plant defense: a comprehensive review. Enzym Microb Technoi, 159: 110055.

Vaish, S., Gupta, D., Mehrotra, R., Mehrotra, S., Basantani, M.K., 2020. Glutathione S-transferase: a versatile protein family. 3 Biotech, 10: 321.

von Arnold, S., Sabala, L, Bozhkov, P., Dyachok, J., Filonova, L., 2002. Developmental pathways of somatic embryogenesis. Plant Cell Tiss Org, 69: 233-249.

Vondráková, Z., Eliášova, К., Fischerova, L., Vágner, M., 2011. The role of auxins in somatic embryogenesis of Abies alba. Cent Eur J Biol, 6: 587-596.

Wang, S., Wang, G., Li, H., Li, F., Wang, J., 2023. Agrobacterium tumefaciens-mediated transformation of embryogénie callus and CRISPR/Cas9-mediated genome editing in 'Feizixiao' litchi. Hortic Plant J, 9: 947-957.

Wang, C., Zhang, M., Wang, J., Wang, Z., Sun, H., 2020. Establishment of an efficient regeneration system in Hippeastrum vittatum with plantlet leaves. Acta Hortic Sin, 47: 301-309. (in Chinese)

Winkelmann, T., 2016. Somatic versus Zygotic Embryogenesis: Learning from Seeds. In vitro Embryogenesis in Higher Plants. Humana Press, New Jersey, pp. 25-46.

Wójcikowska, В., Gaj, M.D., 2017. Expression profiling of auxin response factor genes during somatic embryogenesis induction in Arabidopsis. Plant Cell Rep, 36: 843-858.

Wójcikowska, В., Jaskóła, K., Gąsiorek, P., Meus, M., Nowak, К., Gaj, M.D., 2013. LEAFY ICOTYLEDON2 (LEC2) promotes embryogénie induction in somatic tissues of Arabidopsis, via YUCCA-mediated auxin biosynthesis. Planta, 238: 425-440.

Xiong, F., Liu, H.H., Duan, C.Y., Zhang, B.K., Wei, G., Zhang, Y., Li, S., 2019. Arabidopsis JANUS regulates embryonic pattern formation through pol Il-mediated transcription of W0X2 and PIN7. iScience, 19: 1179-1188.

Yokoo, T., Saito, H., Yoshitake, Y., Xu, Q., Asami, T., Tsukiyama, T., Teraishi, M., Okumoto, Y., Tanisaka, T., 2014. Sel4, encoding a JmjC domain-containing protein, plays key roles in long-day suppression of rice flowering through the demethylation of H3K4me3 of RFT1. PLoS One, 9: e96064.

Yu, B., Huang, L.L., Zhu, Y., Zhu, G.F., Sun, Y.B., 2020. Induction of embryogénie calli from immature pedicels and efficient plant regeneration of Hippeastrum. Acta Hortic Sin, 47:907-915. (in Chinese)

Yu, Y., Bu, Z., Shen, W.H., Dong, A., 2009. An update on histone lysine methylation in plants. Prog Nat Sci, 19: 407-413.

Zhang, M., Wang, A., Qin, M., Qin, X., Yang, S., Su, S., Sun, Y., Zhang, L., 2021. Direct and indirect somatic embryogenesis induction in Camellia oleífera Abel. Front Plant Sci, 12: 644389.

Zhao, P., Begcy, K., Dresselhaus, T., Sun, M.X., 2017. Does early embryogenesis in eudicots and monocots involve the same mechanism and molecular players? Plant Physiol, 173: 130-142.

Zhou, S., Han, Y.Y., Chen, Y., Kong, X., Wang, W., 2015. The involvement of expansins in response to water stress during leaf development in wheat. J Plant Physiol, 183: 64-74.

Zuo, J., Niu, Q.W., Frugis, G., Chua, N.H., 2002. The WUSCHEL gene promotes vegetative-to-embryonic transition in Arabidopsis. Plant J, 30: 349-359.