Introduction

Somatic embryogenesis (SE), one of the most versatile tools in plant biotechnology, may be used to clonally propagate an individual plant at relatively low cost (Etienne 2005) and under aseptic conditions away from predators or pathogens. Under the right conditions, a single cell may reprogram itself from its original specialization to an embryogenic state (Fehér et al. 2016). This totipotent cell must then reenter the cell cycle and develop into embryos, which must then further develop into a mature embryo (Verdeil et al. 2007). In this way, a large number of plants from hundreds of species have been propagated from all kinds of explants taken from valuable mother plants (Fehér 2019).

Somatic embryos may be used to create synthetic seeds when coupled with encapsulation techniques (Singh et al. 2013), as well as for long-time germplasm conservation through cryopreservation (Von Arnold et al. 2002; Pais 2019). The controlled laboratory conditions required for SE is also beneficial for undertaking studies in developmental biology (Ninković et al. 2010). These practical applications resulted in much investigation toward the development of efficient protocols in many plant species over the last several decades.

Unfortunately, years of research have shown that obtaining somatic embryos is not an effortless task since not all species, and even not all genotypes of one species, are equally able to undergo this extreme biochemical and genetic remodeling (Fehér et al. 2003). Embryogenic competence depends on the cell’s ability to respond to specific signals, such as exogenous plant growth regulators (PGRs), culture conditions, and abiotic stress (Fehér et al. 2016). Afterward, the cell undergoes determination—the process in which cell fate becomes fixed and limited to a particular morphogenetic route (Yeung 1995). Laboratory protocols, therefore, must be tailored to maximize SE induction and minimize conditions that cause the formation of other types of structures, such as callus or roots, or even lead to the death of the initial explant.

The ecological, economic, and social importance of the South American trees Araucaria angustifolia (Bertol.) O. Kuntze (Araucariaceae), Acca sellowiana (O. Berg) Burret (Myrtaceae), and Bactris gasipaes Kunth (Arecaceae) led them to be the focus of decades of research (Stefenon et al. 2009; Ree and Guerra 2015; Guerra et al. 2016). Mainly due to the production of eatable seeds, fruits, and earth-of-palm, these species are an important source of incoming in the Atlantic Forest (A. angustifolia and A. sellowiana) and Amazon Forest (B. gasipaes). Aiming at promoting the conservation of the genetic resources and the genetic improvement, somatic embryogenesis studies in these species began in the 1990s, and general laboratory protocols are now relatively well-established, as will be shown in this review. However, some obstacles remain, and further improvements in those processes are needed. Aiming to outline the state-of-the-art of SE protocols in the conifer A. angustifolia (Fig. 1a–g), the dicot A. sellowiana (Fig. 2a–c), and the monocot B. gasipaes (Fig. 3a, b), this review explores the advances obtained and the remaining constraints in the SE in these three South American forest species. Here we review zygotic and somatic embryogenesis, the role of PGRs and secondary metabolites in developing somatic embryos, and the advances of the “omics” sciences in SE research. We finish this review summarizing the still not solved problems and the studies required to develop tools and strategies towards the clonal mass propagation of these and other forest species.

Fig. 1
figure 1

Araucaria angustifolia. a Adult plant. b Megastrobilus (female cone) approximately 16 months after pollination. c Section of megastrobilus. d Fertilized seed. e Nucellus. f Megagametophyte. g Early zygotic embryo. h Early zygotic embryo stained with double coloration of Evans blue and acetic carmine. i, j, k Calli induced from immature zygotic embryos in the 7th, 14th, and 21st days of culture, respectively. l Late somatic embryo after 60 days in maturation medium (with ABA, PEG, and maltose), individualized from the pro-embryogenic masses (PEMs), by lengthening the suspensor and establishing the polarity. m PEM I composed of a group of embryogenic cells linked to one suspensor cell. n PEM II differs from PEM I by the higher number of aggregates of embryogenic cells (Ec) and two or more suspensor cells. o PEM III with an enlarged clump of embryogenic cells connected to suspensor cells (Sc) and show a disturbed. p Late embryo with established polarity. Bars: a-k 500 µm, m 50 µm, n 100 µm, o, p 200 µm. (Color figure online)

Full size image
Fig. 2
figure 2

Acca sellowiana. a Adult plant of cv. Helena in the germplasm collection of the Research and Extension Agency of Santa Catarina State (EPAGRI), São Joaquim, South Brazil (latitude 28°17′39″, longitude 49°55′56″, altitude 1,415 m). b Flower buds and flowers in clusters with petals and sepal. c Flower bud in the first phenological stages of flowering, in a "balloon" shape (arrow). d Embryogenic callus and e embryogenic cells (red-colored) obtained from petals. Embryogenic cultures were double-stained with Evans blue and acetic carmine. f Embryogenic callus obtained from fillets. g Suspensor cells (blue colored) visualized after double-staining with Evans blue and acetic carmine. h, i Berry type fruits with four locules and seeds. j Detached seed. k, l 21 and 30 days-old embryogenic cultures, respectively, in semi-solid medium with 10 µM Picloram. m, n Somatic embryos at different stages of development after 3–4 months. Bars: 250 µm. (Color figure online)

Full size image
Fig. 3
figure 3

Bactris gasipaes. a Adult plants in a natural population. b Mature (red) and immature (yellow) fruits. c Somatic embryogenesis induced from immature embryos, after successive cycles of subcultures. c Somatic embryogenesis induced from immature embryos. d After successive cycles of subcultures in MS supplemented with 10 Picloram, embryogenic calli are generated and confirmed from the double staining with Evans blue and acetic carmine. e, f Maturation of somatic embryos in culture medium supplemented with 5 µM AIB. g Histological analysis of the somatic embryos, confirming the presence of protoderm (Pt) and procambium (Pc). h Seedlings obtained from the converted somatic embryos. Bars: c, e, f, h: 500 µm. d, g: 200 µm. (Color figure online)

Full size image

Physiology and development of zygotic embryos

The central premise of SE is to mimic zygotic embryogenesis in a controlled environment. Therefore, understanding the steps involved in the production of zygotic embryos by the plants is crucial. Embryo development is divided into stages based on morphology, cellular organization, and cell number.

Although zygotic embryos present similar developing stages in different groups of plants, they have some crucial differences among gymnosperms, dicots, and monocots. Understanding physiological, biochemical, and molecular pathways of zygotic embryogenesis is essential to highlight how somatic cells acquire the competence to form somatic embryos and how competent cells decide when to initiate the formation of somatic embryos (Kadokura et al. 2018).

Araucaria angustifolia

In this species (Fig. 1), after fertilization, the zygotic embryo enters into a free nucleus phase in which nuclear divisions occur without the formation of a cell wall (Burlingame 1915; Johansen 1950). Three cell groups differentiate: the lower cell group forms the cover cells, the upper cell group produces the suspensor cells, and the central cluster constitutes a group of embryogenic cells that remain inactive until the full development of the suspensor (Burlingame 1915; Johansen 1950). In conifers, the formation of multiple embryos is typical during the pro-embryonary stage, either by the occurrence of simple and cleavage polyembryony (Buchholz 1920, 1926; Steeves and Sussex 1989; Williams 2009). Only polyzygotic polyembryony at early stages occurs in Araucaria (Figs. 1g, h) (Gifford and Foster 1989; Goeten et al. 2020). However, only one embryo remains in the mature seed, and the other polyembryos degenerate (Filonova et al. 2002; Bozhkov et al. 2005).

In the genus Araucaria, embryo development usually occurs accordingly with the model classically established for Conyferophyta by Singh (1978). In this model, three phases can be distinguished: (i) Pro-embryogenic—before the elongation of the suspensor; (ii) early embryogenic—after suspensor’s elongation and before the establishment of meristems; and (iii) late embryogenic—intensive histogenesis, the establishment of the root and shoot meristems (Burlingame 1915; Johansen 1950; Haines and Prakash 1980; Haines 1983).

In A. angustifolia, mature archegonia are observed at the 13th month after pollination, pollen tube germination occurs during the 14th month, and multiple polyembryos are observed at the 15th month (Goeten et al. 2020). During the pro-embryogenic phase, the embryo consists of a small apical region connected to a long tail formed by elongated cells. In this stage, pro-embryos possess three distinct cell types: cap cells, embryonic cells, and suspensor cells. Seventeen months after pollination occurs, cap cells have completely degenerated, giving rise to the early embryogenic stage of the dominant embryo (Steiner et al. 2015; Goeten et al. 2020). The early embryo presents a mass of embryogenic cells structured into an elongate cylindrical body with a rounded and smooth apical region, and a long suspensor system. The late embryogenic stage starts on the 19th month after pollination. The distal and proximal meristematic zones become recognizable in the apical region. In subsequent embryo development, the apical region increases, and procambium differentiation begins in the proximal zone. Upon further development, the shoot apical meristem and the enlargement of emerging cotyledons become apparent, while the remaining suspensor system collapses (Goeten et al. 2020).

Acca sellowiana

The dicotyledon A. sellowiana (Fig. 2) follows the general pattern of embryo development reported for angiosperms. Zygotes form around 21 days after pollination and present a nuclear endosperm (Pescador et al. 2009; Cangahuala-Inocente et al. 2009a). The zygote divides into two cells of unequal size 24 days after pollination: a smaller cell in the apical position and a larger cell in a basal position (Pescador et al. 2009). At 30 days after pollination, the embryo has entered the globular embryo phase with a distinguishable protoderm (Pescador et al. 2009). Endosperm development around the embryo begins at this stage (Pescador et al. 2009), constituted of a cell layer with distinct nuclei (Cangahuala-Inocente et al. 2009a).

At 45 days after pollination, the heart stage embryo has developed procambial strands, while the suspensor is very small or not apparent (Pescador et al. 2009). At this stage, embryos show a well-defined proto-dermal layer (Cangahuala-Inocente et al. 2009a). An elongated torpedo-shaped embryo with short cotyledons is observed after 60 days following pollination, presenting an extended apical-basal axis (Pescador et al. 2009; Cangahuala-Inocente et al. 2009a). By 75 days after pollination, the embryo has developed rudimentary cotyledonary leaves enclosed by the endosperm (Cangahuala-Inocente et al. 2009a). The endosperm is almost wholly consumed 90 days after the pollination. In addition, cells in the meristems are present at the end of the hypocotyls-radicle axis (Cangahuala-Inocente et al. 2009a).

Around 120 days after pollination, the embryo is mature, having developed a long hypocotyl radicular axis and two fleshy and folded cotyledons. The hypocotyl and the cotyledons have a similar width, and the cotyledons are about as long as the hypocotyl (Pescador et al. 2009). At this stage, the embryo fills the seed coat, and the endosperm is absent (Pescador et al. 2009) since embryos show thickened cotyledons with storage compounds (Cangahuala-Inocente et al. 2009a).

Bactris gasipaes

Despite many studies on SE in B. gasipaes, zygotic embryogenesis remains poorly studied. Palm zygotic embryos develop through three distinct phases: (i) proembryo, (ii) globular, and (iii) cotyledon stages (Haccius and Philip 1979).

The development of the proembryo starts with the division of the zygote into a smaller terminal and a larger basal daughter cell. Typically, in the further segmentation steps, the terminal cell divides by a vertical or oblique wall, while the basal one divides by a transverse or vertical wall. Early in development, one of the terminal daughter cells becomes more active, and the upper part of the embryo becomes gradually asymmetrical. This mass of actively dividing cells originates the embryo proper while the remaining terminal cells form the suspensor (Haccius and Philip 1979).

The development of the embryo into the globular stage depends on two main events: the periclinal divisions in the periphery of the quickly enlarging cell complex, forming the protoderm, while the embryo becomes more differentiated from the remaining suspensoral cells (Haccius and Philip 1979).

The early cotyledon stage is characterized by the presence of a well-developed sheathing base. In contrast, the upper part of the cotyledon remains nearly unchanged as a flat disk-shaped meristem. The stem tip is formed precisely at the terminal pole of the globular embryo, showing the typical features of early apical meristem cells (Haccius and Philip 1979). As development progresses, this region bulges out, and a core of procambial cells below it forms the future hypocotyl central cylinder. Subsequently, the whole embryonal body becomes asymmetric due to the increased activity of the cells on one side. Due to this high cell division rate, the meristem quickly surpasses the sheath, forming a massive cylindrical-conical body (Haccius and Philip 1979).

Selecting the explant is a crucial primary step in the SE

It was postulated (Haberlandt 1902) that all plant cells can change from one type to another. The in vitro response to SE induction, however, shows that factors such as genotype, tissue type, or developmental stage lead to vastly differing responses even within the same species. The starting material often dictates success or failure, and it has, therefore, been the focus of many studies.

Araucaria angustifolia

Since the earliest studies, immature zygotic embryos (Fig. 1g) remain the most responsive explants for A. angustifolia, with the developmental stage of the zygotic embryo often dictating its response (Astarita and Guerra 1998; dos Santos et al. 2002, 2008; Silveira et al. 2002). Astarita and Guerra (1998) firstly showed that the developmental stage of the explants influenced cellular mass induction in A. angustifolia. While younger explants (pro-embryos collected in December) were responsive to induction, older explants (collected in February, just about the start of cotyledon formation) lost the ability of induction. A significant reduction in the induction rate for explants collected in February (2.8%) in comparison to explants from December (23.3%) was also reported by dos Santos et al. (2002). Silveira et al. (2002) obtained an induction rate of 52.1% from pre-cotyledonary embryos as explants, while less than 30% of embryogenic induction rates were obtained from pro-embryos and torpedo stage embryos. Immature and mature zygotic embryos were used as source of explants for the induction of somatic embryogenesis in Araucaria angustifolia by dos Santos et al. (2008). Embryogenic cultures were only obtained from immature zygotic embryos, while mature pre-cotyledonary zygotic embryos formed a white translucent mucilaginous cell mass, without embryogenic development (dos Santos et al. 2008).

Acca sellowiana

Acca sellowiana possesses more tissues able to develop embryogenic cultures compared to A. angustifolia. Petals, styles of floral buds, stamens/filaments (Fig. 2c, d, and f), ovaries, leaf segments, as well as mature and immature zygotic embryos are responsive explants capable of direct and/or indirect SE induction in A. sellowiana (Cruz et al. 1990; Canhoto and Cruz 1994; Stefanello et al. 2005; Cangahuala-Inocente et al. 2007; Canhoto et al. 2009). Cruz et al. (1990) obtained soft-brownish calli originated mainly from cotyledons, apical embryonic meristems, and hypocotyls of zygotic embryos excised from seeds of immature fruits of A. sellowiana. Such calli generated two different kinds of somatic embryos: white embryos lacking chlorophyll and passing through different morphological aspects, and green embryos having a relatively undifferentiated cotyledonary and apical regions and elongated radicles (Cruz et al. 1990). Floral buds A. sellowiana collected before anthesis were used as source of explants (Fig. 2c) (Stefanello et al. 2005). Petals, stamens, and ovaries generated up to 81.7% 86.7%, and 93.3% of embryogenic calli, respectively. Cangahuala-Inocente et al. (2007) succeeded in obtaining embryogenic calli from petals (16.7% of induction rate), styles (100%), and stamens (100%), using different treatments. Leaf segments of an adult tree of A. sellowiana were used by Canhoto et al. (2009), and embryogenic calli were obtained, but they failed to develop further. Despite the positive results obtained with floral tissues concerning embryogenic calli induction and the advantage of allowing the cloning of selected adult genotypes, mature and immature embryos are the most responsive explants and have been preferentially employed.

Bactris gasipaes

Like A. sellowiana, B. gasipaes possesses numerous tissues that are responsive to SE induction. Somatic embryos can be induced from shoot tips (Valverde et al. 1987); adventitious buds (Almeida et al. 2012); leaf primordia (Almeida and Almeida 2006); zygotic embryos (Steinmacher et al. 2007a); inflorescences (Steinmacher et al. 2007b); and thin cell layers from subapical tissues, leaf sheath and apical meristem (Steinmacher et al. 2007c). Shoot tips were used as explants by Valverde et al. (1987), originating 23.8% of embryogenic structures from secondarily formed callus, after several cycles of sub-culturing. These embryos germinated and gave rise to new plantlets. Leaf-primordia were taken from the heart-of palm shoot tips as explants and resulted in an 82% direct embryogenesis induction rate (Almeida and Almeida 2006). Zygotic embryos of B. gasipaes yielded 78% of primary calli (Steinmacher et al. 2007a), while rachillae of immature inflorescences induced only 3% of somatic embryos (Steinmacher et al. 2007b) and thin cell layers (TCL) promoted up to 97% of somatic embryogenesis induction (Steinmacher et al. 2007c). However, the TCL response depended heavily on from which region explants were harvested: 17% of explants taken from apical meristems developed embryogenic callus compared to 3–9% from upper layers or sub-apical tissue (Steinmacher et al. 2007c). Thin cell layers taken from converted somatic embryos of a slow-growing aging culture line gave rise to a young and more vigorous line of embryogenic cultures (Ree et al. 2020).

The pioneer studies on somatic embryogenesis of A. angustifolia, A. sellowiana, and B. gasipaes

The initial studies on SE of A. angustifolia, A. sellowiana, and B. gasipaes were performed in the 1980s–1990s, but the basal components of the culture media and general culture conditions remain the base for all studies aiming at improving the protocols for these species.

Araucaria angustifolia

The first study on the SE in A. angustifolia began with Handro and Ferreira (1980). However, few calli survived after six months of subculture. In the 1990s, the works of Guerra and Kemper (1992) and Guerra et al. (1993), advanced in the protocol and obtained calli after a long period of in vitro propagation. Further, Astarita and Guerra (1998) described the conditions required for the induction and multiplication of embryogenic cultures (Table 1; Fig. 1i–k), as well as the formation of early somatic embryos in this species. These same authors reported that the ability of zygotic embryo explants to induce the formation of embryonal-suspensor masses is directly related to the maturation stage of the seeds (Astarita and Guerra 1998). At the cotyledonary stage, zygotic embryos were unable to produce calli. This study also demonstrated that embryonal-suspensor masses of A. angustifolia could be successfully maintained in LP medium (von Arnold and Eriksson 1981) supplemented with 22 µM of 2,4-D, 11 µM of 6-benzylaminopurine (BAP), and 11 µM of kinetin (Kin). Suspension cultures grew quickly in medium supplemented with sucrose, but other saccharides commonly used as osmotic agents in the culture media (mannitol, sorbitol, inositol, and glucose), are toxic to Araucaria cells (Astarita and Guerra 1998). These aspects have guided the establishment of embryogenic cultures of A. angustifolia with slight modifications, as the use of different culture media and the cultivation in cellular suspension conditions. The supplementation of the culture medium with casein hydrolysate, myo-inositol, and L-glutamine is also typical in the induction and growth of embryogenic cultures of A. angustifolia.

Table 1 Summary of some studies reporting significative advances in the SE protocol for Araucaria angustifolia
Full size table

Acca sellowiana

Somatic embryogenesis of A. sellowiana was first described by Cruz et al. (1990; Table 2). These authors found that SE (Fig. 2k, l) required 2,4-D in the MS culture medium, and the most effective concentrations ranged from 2.26 to 4.52 µM. Embryo development was revealed to be asynchronous, i.e., a callus frequently showed embryos at various developmental stages. Germination of the somatic embryos (Fig. 2m, n) was proposed to occur after transfer to a modified MS medium supplemented with gibberellic acid (GA3) and kinetin (Kin). Both MS and LPm medium are effective basal salts for A. sellowiana SE induction medium. The protocol of SE of A. sellowiana was further improved after several experiments through manipulation of the culture media composition (Canhoto and Cruz 1994; Booz and Pescador 2007), using different basal media (Guerra et al. 1997) and enhancing the auxin induction treatment (Guerra et al. 2001).

Table 2 Summary of some studies reporting significative advances in the SE protocol for Acca sellowiana
Full size table

Bactris gasipaes

B. gasipaes SE (Fig. 3c–e) was first described by Valverde et al. (1987; Table 3). Shoot tips were extracted from young field-grown plants and placed on MS medium supplemented with BAP and Picloram. Callus (Fig. 3c) formed initially, but eventually, both somatic embryos (Fig. 3f) and shoot primordia developed (Valverde et al. 1987). Like the vast majority of palms, SE in B. gasipaes requires both high concentrations of exogenous auxins and activated charcoal (Ree and Guerra 2015). Indeed, explants from multiple B. gasipaes tissues readily undergo SE when cultivated with Picloram and activated charcoal but are generally unresponsive or oxidize rapidly without them (Steinmacher et al. 2007a, b, c; Heringer et al. 2014).

Table 3 Summary of some studies reporting significative advances in the SE protocol for B. gasipaes
Full size table

The role of plant growth regulators in somatic embryogenesis

The correct definition of plant growth regulators types and levels in each phase of the SE plays a central role. In tree species, different plant growth regulators have been used for callus induction, somatic embryo formation, shoot organogenesis, and for the conversion of somatic embryos into plantlets (Giri et al. 2004).

Araucaria angustifolia

Embryogenic cultures of A. angustifolia are multiplied as pro-embryogenic masses (PEM) during the early stages of SE (Steiner et al. 2015; Guerra et al. 2016). According to their development phase, pro-embryogenic masses have three distinct stages: PEM I (Fig. 1m), a small compact cluster of cells with a dense cytoplasm nearby to a single vacuolated and relatively elongated cell; PEM II (Fig. 1n), an aggregate of cells adjacent to more than one vacuolated cell; and PEM III (Fig. 1o), an expanded cluster of cells with a dense cytoplasm (Filonova et al. 2002; Farias-Soares et al. 2014; Steiner et al. 2015; Fraga et al. 2015).

Astarita and Guerra (2000) induced embryonal-suspensor mass in A. angustifolia using LP medium supplemented with 1-naphthalenacetic acid (NAA) or 2,4-D, kinetin (9.3 μM), and BAP (12 μM). The higher induction rates (60%) were obtained with 45 μM of 2,4-D supplemented to the culture medium. The induction of somatic embryos from the embryonal-suspensor masses was performed using abscisic acid (ABA) at 38 mM. In association with carbohydrates, the use of ABA did not promote early somatic embryos maturation. However, early somatic embryos were obtained when the embryonal-suspensor masses were transferred to medium supplemented ABA and polyethylene glycol (PEG), suggesting that the use of this PGR is important for this phase of the SE.

On the other hand, dos Santos et al. (2002) and Silveira et al. (2002) reported that SE occurred on explants on medium without PGRs at similar rates (14–15%) to explants with 2,4-D, kinetin, and BAP. However, the absence of PGRs in the multiplication phase led to tissue oxidation and the decline of the embryogenic capacity. Embryogenic cultures maintained in culture media supplemented with PGRs 2,4-D, BA and Kin were white-translucent, while the ones kept in culture medium free of growth regulators showed a progressive browning process (dos Santos et al. 2002). However, even these brownish cultures presented sectors with proliferative capacity, which originated somatic proembryos after selection and subculture on the same culture media composition. The use of BAP and kinetin (1 μM each), combined with maltose and PEG as osmotic agents, showed to be also effective for PEM development (dos Santos et al. 2002; Steiner et al. 2005).

The accumulation of ABA and indoleacetic acid (IAA) in proliferating embryogenic cultures was reported by Steiner et al. (2007a, b), as an effect of the application of exogenous polyamines (PAs) to the culture medium. The authors postulated that the stress caused by the exogenous PAs resulted in the accumulation of ABA and IAA and likely helped to stimulate embryogenic culture conversion into PEMs. However, the direct transfer of embryogenic cultures from multiplication to maturation culture medium containing ABA seems to inhibit the further development into early somatic embryos (Farias-Soares et al. 2014; Fraga et al. 2015).

A two-steps maturation in which embryogenic cultures are first cultivated in a PGR-free medium with high concentrations of osmotic agents (maturation phase I) and then in a culture medium supplemented with ABA with high concentrations of osmotic agents (maturation phase II) was proposed by Steiner et al. 2015; Fraga et al. (2015). This approach showed to be more suitable for early somatic development, while the direct transfer of embryogenic cultures in proliferation to maturation medium supplemented with ABA revealed to be highly harmful. The polarization rate and survival of PEMs after the maturation treatment without ABA was enhanced after maturation treatment with ABA (Steiner et al. 2015; Fraga et al. 2015).

Acca sellowiana

In A. sellowiana, SE can be induced through exogenous 2,4-D and Kin (up to 72% of SE induction), although NAA and 3-indolebutyric acid (IBA) promoted near 60% and 55% of SE induction respectively (Cruz et al. 1990). The use of an auxin shock with high concentrations of 2,4-D (20 μM) for 24 h was proposed as an alternative approach for inducing SE (Guerra et al. 1997, 2001), promoting from 10 to 45% of ES induction, but a lower rate of abnormal embryos, ranging from 16.5 to 45.3% of abnormal embryos, depending on the concentration of 2,4-D and time of exposition (Canhoto et al. 2009).

The use of Picloram (10 μM) and Kin (1 μM) significantly enhanced the induction of callus from A. sellowiana floral explants (81.7, 86.7%, and 93.3% for stamens, petals, and ovaries respectively; Stefanello et al. 2005). Globular somatic embryos grew from the resulting friable calli and developed further in suspension with 2,4-D (1 μM) and 2-iP (1 μM), and then after a final transfer to PGR-free LPm medium.

Cangahuala-Inocente et al. (2007) used different types and concentrations of PGRs to induce SE from A. sellowiana flowers. Picloram, 2-4D, or Dicamba at 2 μM promoted 83.3–100% callogenesis in style and stamen filament. However, the response to the PGRs in this study was genotype-dependent. The combination of 2,4-D (20 μM) with BAP (5 μM) or 2-iP (5 μM) also induced the formation of embryogenetic calli at induction rates ranging from 68.3 to 93.9%; Stefanello et al. 2005).

The conversion of somatic embryos into plantlets is also a PGR-dependent step. Exogenous gibberellic acid (GA3) at 2.0 μM improved plantlet conversion of somatic embryos in the pre-cotyledonary stage (Guerra et al. 1997). However, embryos in the cotyledonary stage failed to convert into plantlets in the same culture medium, and it was suggested the presence of dormancy during the transition of the pre-cotyledonary to the cotyledonary embryo stage (Guerra et al. 1997).

Exogenous GA3, in combination with Kin (Cruz et al. 1990; Canhoto and Cruz 1994) or BAP (Fraga et al. 2012) also improved embryos conversion. Cruz et al. (1990) reported the absence of further development of mature somatic embryos maintained in induction media containing just auxin. Conversion of somatic embryos was just acquired when the culture medium was supplemented with 0.5 mg/L GA3 and 0.1 mg/L Kin. Up to 51% of conversion was obtained by Canhoto and Cruz (1994) using MS medium supplemented with 0.5 mg/L GA3 and 0.1 mg/L Kin. Improved conversion rates were obtained in semi-solid medium with BAP 0.5 μM and GA3 1.0 μM plus activated charcoal, resulting in a conversion rate of 35% (Fraga et al. 2012), as well as in a RITA® temporary immersion system (Fraga et al. 2013). However, embryos may also convert in medium containing BAP alone (0.5 μM; Dal Vesco and Guerra 2001). On the other hand, the conversion of torpedo and cotyledonary-stage somatic embryos into plantlets was obtained upon transfer to PGR-free LPM medium (Stefanello et al. 2005).

Bactris gasipaes

Somatic embryogenesis induction can be obtained with the use of different PGRs as 2,4-D (Stein and Stephens 1991; Steinmacher et al. 2007b), BAP (Almeida and Almeida 2006; Almeida et al. 2012), Picloram (Steimacher et al. 2007b, 2007c, 2011, 2012; Heringer et al. 2013, 2014; Nascimento-Gavioli et al. 2017), Picloram + N6-benzyl-adenine (Valverde et al. 1987), Dicamba (Steinmacher et al. 2007b), and 2,4-D + Picloram + BAP (Maciel et al. 2010).

The use of 7.1 μM BAP for the induction of SE from shoot tips of B. gasipaes promoted active oxidation in the explant (Almeida and Almeida 2006). However, the oxidative process was restricted to the covering tissue, and the explants were duplicated during the first subculture. These cultures were subsequently cultured in MS medium supplemented with a combination of NAA 12.9 μM and BAP 3.55 μM, generating somatic embryos in 82% of explants (Almeida and Almeida 2006).

Comparing the SE induction rate promoted by different PGRs, Steinmacher et al. (2007b) reported induction rates of 1.4%, 10.5%, and 11.4% for 2,4-D (600 μM), Dicamba (300 μM), and Picloram (300 μM) respectively. Despite the positive response of B. gasipaes to Picloram, it was revealed to be genotype dependent (Steinmacher et al. 2007b).

The use of Picloram (0, 10, 20, or 40 μM) in interaction with 2-isopentyladenine (2-iP; 0 or 5 μM) for the SE of B. gasipaes using mature embryos as explants revealed a higher induction rate (78%) for Picloram (20 μM) in the absence of 2-iP (Steinmacher et al. 2007c). In this study, 1 μM of silver nitrate (an inhibitor of the hormone ethylene) in the SE induction medium enhanced the embryogenic competence of subepidermal cells adjacent to vascular bundles (Steinmacher et al. 2007c). Picloram at concentrations of 300 μM and 600 μM (Steinmacher et al. 2007c) were used to induce SE in TCL explants, generating from 8 to 43% of embryogenic calli. In addition to its role in inducing SE, Picloram can also be used in the multiplication medium to promote secondary SE (Steinmacher et al. 2007b; Heringer et al. 2014; Ree et al. 2020).

After SE induction, culture multiplication and development require a medium with either different concentrations of growth regulators or complete removal. Steinmacher et al. (2007b, c) and Heringer et al. (2013) reported the embryogenic cultures multiplication using a culture medium enriched with 40 μM 2,4-D + 10 μM 2-iP, while Stein and Stephens (1991) and Almeida et al. (2012) used free-PGR medium for this step. ABA (5 μM) improved somatic embryo development when added to either a semi-solid medium or when included in a RITA® temporary immersion system (Heringer et al. 2014). In this study, embryos multiplied in the temporary immersion system showed greater conversion into plantlets.

PGR-free medium has been sufficient for plantlet conversion (Valverde et al. 1987; Stein and Stephens 1991; Almeida et al. 2012). However, the inclusion of NAA (12.9 μM) and BAP (3.55 μM Almeida and Almeida 2006;) or NAA (0.44 μM) and 2-iP (24.5 μM) Steinmacher et al. 2007b, c; Steinmacher et al. 2011; Heringer et al. 2013) in culture medium improved somatic embryo conversion into plantlets (Fig. 3h).

Several non-PGR compounds are key factors in the embryogenetic development

In addition to the choice of the explant and of PGRs, several biochemical compounds are essential in the regulatory mechanisms of SE. For instance, carbohydrates provide both a carbon and energy source and influence the control of somatic embryos induction, growth, and maturation. The polyamines (PAs) are known to play a vital role into the regulation of somatic embryogenesis (Kong et al. 1998; Thorpe and Stasolla 2001), while nitric oxide plays a key role as intra- and inter-cellular messenger, inducing biochemical processes as the formation of embryogenic-like cells (Silveira et al. 2006).

Araucaria angustifolia

The inclusion of either sucrose or maltose as the carbon source influences A. angustifolia embryogenic culture behavior: 3% sucrose promotes higher induction rates (56.6%) but also the formation of abnormal pro-embryos, while 3% maltose (35.4% of induction rate) reduced cell proliferation to favor further development of somatic pro-embryos with bipolar morphology (Steiner et al. 2005).

Maltose showed to be beneficial because the transition from pro-embryonic cultures to somatic embryos is a critical point in A. angustifolia (Steiner et al. 2015). The combined use of maltose (60 g L−1) and sucrose (30 g L−1) positively influenced somatic embryo maturation, promoting the development of apical buds from the embryonal shoot meristem in more than 50% of explants on a PGR-free culture medium (dos Santos et al. 2008). However, embryos did not form roots, and so conversion was not completed. It was shown that the use of pre-maturation treatments supplemented with maltose (90 g L−1) or lactose (90 g L−1) and polyethylene glycol (PEG) enhanced the transition of PEM III to early somatic embryos (Farias-Soares et al. 2014).

Silveira et al. (2006) demonstrated that the PAs caused a reduction in the growth of A. angustifolia suspension cultures. However, exogenous polyamines into PGR-free demi-solid BM medium improved A. angustifolia embryogenic culture proliferation: 1 mM putrescine (Put) promoted more than twofold increase in cultures fresh-mass, while 0.01 mM Spermine (Spm) and 1 mM spermidine (Spd) fostered almost twofold proliferation (Steiner et al. 2007a, b). Spermine may be associated with a reduction in the activities of proton pumps, such as H + -ATPase and H + -PPases, reducing the cellular growth and promoting the development of PEMs in embryogenic cultures (Dutra et al. 2013). Furthermore, the addition of exogenous Spm increased IAA content (3–3.5 μg g−1 of fresh mass) in embryogenic cultures, and exogenous Put or Spd increased ABA accumulation (0.2–0.3 μg g−1 of fresh mass Steiner et al. 2007a, b;). The addition of Put increased nitric oxide release from the embryogenic culture, while Spd and Spm have a contrary effect. Interestingly, embryogenic cells accumulated more nitric oxide than the suspensor cells, suggesting that nitric oxide biosynthesis might be related to the maintenance of the polarity present at the PEM II stage (Silveira et al. 2006).

Also, the quality and quantity of early somatic embryos in A. angustifolia are related to nitric oxide emission, which shows a clear relationship with the ratio of reduced glutathione/glutathione disulfide (Vieira et al. 2012). Low reduced glutathione concentrations (0.01 and 0.1 mM) displayed potential nitric oxide scavenger activity in the pre-maturation culture medium. Furthermore, low reduced glutathione concentrations increased the number of early somatic embryos formed in cell suspension culture medium within a few days after inoculation. In semisolid culture medium, high levels of reduced glutathione (5 mM) promoted the development of globular embryos. These embryos emitted nitric oxide from their apexes, suggesting that nitric oxide plays a role in embryo development (Vieira et al. 2012).

Aiming to improve PEM III-to-early somatic embryos transition, Farias-Soares et al. (2014) proposed a pre-culture phase where embryogenic cultures are firstly cultivated in a medium with fluridone (FLD) and subsequently, in a pre-maturation medium with high osmotic potential (carbon source and polyethylene glycol). The use of the preculture with FLD significantly increased the frequency of PEM III (60%) in comparison to the control treatment (40%).

Acca sellowiana

Amino acids greatly influence A. sellowiana SE. Comparisons between somatic and zygotic embryogenesis (Pescador et al. 2013) revealed differences in the endogenous abundance of different amino acids, corroborating this dependence. Glutamine was the most abundant amino acids in zygotic embryogenesis at the globular stage (9.0 μmol g−1 fresh mass). However, in cotyledonary-staged somatic embryos, the highest amount (3.0 μmol g−1 fresh mass) was observed from direct SE. The same was observed for the levels of glutamic acid at the globular (3.8 μmol g−1 fresh mass in zygotic embryos) and cotyledonary (2.5 μmol g−1 fresh mass in embryos obtained from indirect SE) stages (Pescador et al. 2013). Asparagine and GABA, in turn, revealed the highest amount of zygotic embryos at the globular stage (48 μmol g−1 and 10 μmol g−1 fresh mass respectively), without any peak in somatic embryos in any stage of development (Pescador et al. 2013).

The addition of exogenous glycine (4 mM) to the culture medium significantly improved somatic embryo induction and development in A. sellowiana, with an estimation of 43.2% of SE induction and 25.5 viable somatic embryos/explant obtained after 18.8 and 14.5 weeks in culture, respectively (Dal Vesco and Guerra 2001). Similarly, the supplementation of the induction medium with 10 mM of γ-aminobutyric acid (GABA) promoted higher induction of globular embryos, whereas 8 mM of GABA induced the development of embryos at the torpedo stage (Booz and Pescador 2007). Exogenous GABA also reduced the occurrence of abnormal embryo development, suggesting that it plays a pivotal role in embryo development (Booz et al. 2009).

Cangahuala-Inocente et al. (2014) measured free amino acids in A. sellowiana embryogenic cultures for 30 days. Free amino acid content was highest on the 6th day of culture (3676.7 μg g−1 fresh weight), decreased during epidermal cell proliferation, and then peaked again on the 15th day (2884.9 μg g−1 fresh weight). Following this, total amino acid levels decreased up to the 24th day (335.5 μg g−1 fresh weight), remaining constant until the 30th day of culture (Cangahuala-Inocente et al., 2014). In addition, early-stage somatic embryos (globular and heart) showed significant variation in the total content of free amino acids in comparison with more advanced somatic embryos, at the torpedo and cotyledonary stages (Cangahuala-Inocente et al. 2014). These authors suggested that this pattern implied that the many developmental stages in somatic embryo formation created different demands for amino acids.

Zygotic embryos contained higher concentrations of sucrose, fructose, raffinose, and myo-inositol compared to somatic embryos (Pescador et al. 2008; Cangahuala-Inocente et al. 2009b). During SE, soluble sugars levels increased from 0.7 to 2.7 mg g−1 fresh weight during the first 60 days of induction. Afterward, sugar levels remained stable until day 120, increasing in heart and cotyledonary stages, and decreasing in the torpedo and pre-cotyledonary stages (Cangahuala-Inocente et al. 2009a). These findings indicate that reserves were being mobilized to fuel embryo development (Cangahuala-Inocente et al. 2009a). The addition of maltose in the culture medium improved embryo conversion to plantlets in comparison to sucrose; however, it increased the formation of achlorophyllous somatic plantlets (19.47% of normal and 48.93% of abnormal achlorophyllous plants; Pavei et al. 2018).

Moreover, the supplementation of the culture medium with 1 mM glutathione (GSH) also demonstrated positive effects, accelerating the induction rate (80% versus 56.67% without GSH). The process of somatic embryo formation, in turn, presented the best results with a concentration of 0.1 mM of GSH (56.17, 34.83, and 57.17 embryos formed at lobular, heart, and torpedo stages respectively) in comparison to the absence of GSH (38.83, 21.83, and 20.83 embryos) and increasing the embryonic synchrony (Pavei et al. 2018).

Bactris gasipaes

Steinmacher et al. (2012) demonstrated that arabinogalactan proteins seem to play an important role in SE of B. gasipaes. Secondary somatic embryos were observed on 71% of B. gasipaes explants in the control treatment, whereas only 33% developed in the presence of 30 μM ßGlcY (an inhibitor of arabinogalactan proteins). Furthermore, arabinogalactan proteins were prominent within the extracellular matrix of somatic embryos, tending to accumulate in areas from which new secondary somatic embryos emerge. Higher content of arabinose in embryogenic cultures in comparison to fibrous non-embryogenic was reported Ree et al. (2020). No arabinose was observed in eight-year-old cultures, which had developed into amorphous masses of hyperhydritic, malformed somatic embryos. However, these authors point out that there is no clear link between the higher concentration of arabinose found in the embryogenic cultures and the arabinogalactan proteins observed by Steinmacher et al. (2012) in the cell walls of embryogenic cells and closely associated with the extracellular matrix of globular embryos and regions where new secondary somatic embryogenesis occurs (Ree et al. 2020).

Significant differences between embryogenic and non-embryogenic B. gasipaes cultures were identified by Nascimento-Gavioli et al. (2017) concerning the content of phenols, PAs, and amino acids. Non-embryogenic cultures contained higher total phenolic compounds (24.5 μg g−1 fresh mass vs 7.9 μg g−1 in embryogenic cultures) and PAs (2068.6 μg g−1 fresh mass vs 359.9 μg g−1 in embryogenic cultures), and lower concentrations of endogenous total amino acids (57,072.7 μg g−1 fresh mass vs 79,378 μg g−1 in embryogenic cultures), except GABA. Non-embryogenic cultures presented a three-fold higher concentration of GABA (867.0 μg g−1 fresh mass) in comparison to embryogenic cultures (256.7 μg g−1 fresh mass; Nascimento-Gavioli et al. 2017). This is especially interesting because exogenous GABA increases the induction of globular embryos in A. sellowiana (Bozz and Pescador 2007) and also seems to be a key factor for B. gasipaes.

Recently, nitric oxide synthesis and storage proteins were found to be differentially accumulated in a vigorous young B. gasipaes culture line compared to an eight-year-old culture line that had lost its ability to form somatic embryos (Ree et al. 2020).

The highlights from the omics sciences

Advances in biotechnological tools allow high throughput analyses of entire groups of genes (genomics), transcripts (transcriptomics), proteins (proteomics), and metabolites (metabolomics). In turn, these omics sciences have been used trying to elucidate critical processes involved in SE of recalcitrant species.

Araucaria angustifolia

The pioneer omics studies in A. angustifolia aimed at elucidating key factors in zygotic embryogenesis, which might help to understand the control points in SE. Proteomics analyses revealed a gradual accumulation of soluble protein in the megagametophyte was observed during the development of A. angustifolia seed, with the highest amount detected in mature seeds, reaching a tenfold increase from the globular zygotic embryo phase until the mature zygotic embryo phase (Santos et al. 2006). Storage and late embryogenesis abundant (LEA) proteins were the mainly accumulated proteins (Santos et al. 2006; Silveira et al. 2008; Balbuena et al. 2009). Proteins with chitinolytic activity—believed to play a critical role in embryogenesis—were observed in all phases of seed development, but the expression of class IV chitinases was observed only in the late stages (Santos et al. 2006). However, chitinase accumulation may be correlated with the developmental program of seed maturation but can also be a response to abiotic factors or pathogenic attack (Santos et al. 2006). Further, Silveira et al. (2008) suggested that the increased synthesis of storage and LEA proteins in pre-cotyledonary stage zygotic embryos is correlated with high levels of ABA and the differential expression of the mitogen-activated protein kinases (implicated in ABA signaling) and the abscisic acid receptor PYL8 in mature zygotic embryos of A. angustifolia was reported by Balbuena et al. (2011). These results were related to SE as the successful conversion rates of embryogenic cultures when using torpedo/early-cotyledonary zygotic embryos may be mostly effect of a boom in the gene expression levels and activation of different biochemical pathways (Balbuena et al. 2009).

The proteomics comparative analyses between cultures with contrasting embryogenic potential shed light on A. angustifolia SE Jo et al. (2014). found a significant difference in the abundance of proteins comparing responsive (3.22 mg g−1 fresh weight) and blocked (2.43 mg g−1 fresh weight) cell lines. Among the detected proteins, enzyme S-adenosylmethionine synthase and the mitochondrial ATPase beta subunit were observed with a high score in the responsive cell line but not in the blocked one. The authors suggested that the mitochondrial ATPase is a potential marker for the selection of embryogenic culture lines responsive to maturation treatments in A. angustifolia. Using a quantitative proteomic approach, Santos et al. (2016) identified 35 proteins that were more abundant in a responsive genotype of A. angustifolia embryogenic cultures, while 71 proteins were more abundant in the non-responsive culture (Santos et al. 2016). This study also revealed an increase in the number of proteins associated with cell defense, anti-oxidative stress responses, and storage reserve deposition in the responsive embryogenic cultures. The detection of vicilin and GST proteins exclusively in the responsive cell line suggests that they have potential as markers in the early selection of cell lines with responsiveness to maturation conditions (Santos et al. 2016).

A label-free proteomic analysis comparing EC proliferation of cell cultures in response to presence/absence of PGR supplementation (Fraga et al. 2016) found that PIN-like proteins (transmembrane proteins that transport auxin as their substrate) were synthesized exclusively in cultures grown in PGR-supplemented medium (4 μM 2,4-D, 2 μM BAP, 2 μM Kin), indicating the possibility of polar auxin transport in this embryogenic culture cell line. Additionally, embryogenic cultures grown on PGR-supplemented medium showed up-accumulation of stress-related proteins compared to those grown on PGR-free medium, suggesting that PGRs created a stressful environment. In the same way, enhanced expression of proteins involved with protein folding and stabilization processes in PGR-free treatment could play a protective function in response to the stress conditions caused by in vitro culture (Fraga et al. 2016).

The global DNA methylation (GDM) comparison between embryogenic cultures of A. angustifolia induced on PGR-free culture medium and embryogenic cultures induced on PGR-supplemented medium revealed an increased rate of GDM in the former ones (Fraga et al. 2016). In the first cycle of the subculture, PGR-free cultures presented 20.04% of GDM, while supplemented cultures revealed 12.28% of GDM. The levels of GDM decreased in both cultures, equaling about 10% in cycle 5. Unstable behavior of increase/decrease of the GDM level persisted until the cycle 17 multiplication in the PGR-free culture, while a gradual and steady increase in GDM was observed in the PGR-supplemented culture (Fraga et al. 2016). The behavior of GDM of the cultures was observed during 17 cycles and compared to the levels of the initial zygotic embryo explant. Throughout all cycles, PCR-supplemented cultures presented GDM levels inferior to the zygotic embryo, while the PGR-free cultures presented higher GDM in the first and the ninth cycles, and lower in all other cycles of subculture. The authors reported the existence of promising results with embryogenic cultures induced and maintained in PGR-free culture medium (Fraga et al. 2015), while low responsiveness to maturation treatment was observed for derived from PGR-supplemented cultures.

Analysis of gene expression in zygotic embryos of A. angustifolia at different developmental stages was used to study embryonic gene expression and establish markers for monitoring normal development and detect abnormalities in the somatic embryogenesis process of A. angustifolia (Schlögl et al. 2012a, b). The genes ARGONAUTE (AGO), CUP-SHAPED COTYLEDON1 (CUC) WUSCHEL-related WOX (WOX), S-LOCUS LECTIN PROTEIN KINASE (LecK), LEAFY COTYLEDON 1 (LEC), and VICILIN 7S (VIC) revealed up-regulated expression until the cotyledonary stage, decreasing expression in mature zygotic embryos. The SCARECROW-like (SCR) gene revealed high expression at the start of embryo development, and in post-embryogenic structures. SCR is up-regulated in embryogenic cultures maintained in absence of growth regulators, and down-regulated when the embryogenic culture is transferred to a medium supplemented with growth regulators or in the maturation medium (Schlögl et al. 2012a). Subsequently studies on the gene expression during somatic embryogenesis if A. angustifolia were implemented, evaluating some of these genes.

Up-regulation of AGO, CUC, WOX, LecK, and VIC genes was observed during the transition of somatic embryos from stage I to stage II (Schlögl et al. 2012b). This study also revealed the influence of the presence/absence of both auxin and cytokinin in the expression of AGO and SCR genes during the multiplication phase of embryogenic cultures. On the other hand, this influence was not observed for the RPG and LEC genes (Schlögl et al. 2012b).

In another study, Steiner et al. (2012a, b) showed that a putative homolog of the SOMATIC EMBRYOGENESIS RECEPTOR-like KINASE (SERK) gene family member (SERK1) preferentially expressed in embryogenic cell aggregates of A. angustifolia, but not in non-embryogenic cell cultures. Additionally, in situ hybridization results showed that SERK1 transcripts initially accumulated in groups of cells at the periphery of the embryogenic culture and then are restricted to the developing embryo proper (Steiner et al. 2012a, b).

Acca sellowiana

According to proteomic analysis, A. sellowiana zygotic and somatic embryos had similar protein profiles at defined developmental stages, suggesting that key proteins play conserved roles during development (Cangahuala-Inocente et al. 2009b). Eleven differential proteins (spots) were detected in 2-D electrophoresis analysis of different somatic embryogenesis stages, with one exclusively observed in the torpedo, four in the pre-cotyledonary, and none in the cotyledonary stage. This pattern was interpreted as evidence that few genes are involved in the control of hystodifferentiation and morphogenesis of A. sellowiana somatic embryos (Cangahuala-Inocente et al. 2009b).

Fraga et al. (2013) evaluated the proteomics pattern of normal and off-type phenotype somatic plantlets obtained at 10- and 20-days conversion in RITA® temporary immersion system. The average number of detected proteins for the 10-days conversion plantlets was 772 and 642 for normal and off-type, respectively. For the 20-days conversion plantlets, the average number of proteins was 1138 and 1177 for normal and off-type, respectively. A total of 10 exclusive spots were identified in the off-type and two in the normal plantlets at 10 days conversion. At 20 days conversion, the off-type plantlets revealed five exclusive proteins, and the normal plantlets had 12 exclusive proteins (Fraga et al. 2013). These authors suggested that this high similarity between protein profiles of normal and off-type plantlets suggests that changes in a few proteins’ expression may be responsible for abnormal somatic plantlets formation. Moreover, a vicilin-like storage protein was only found in off-types of 20-days conversion plantlets, suggesting that these plantlets may present an abnormality in the mobilization of storage compounds, a fact that causes reduced vigor (Fraga et al. 2013).

Global DNA methylation in A. sellowiana was studied by Fraga et al. (2012) and Cristofolini et al. (2014). Small differences were observed in the pattern of fragments PCR amplification of DNA from normal and off-type somatic plantlets after enzymatic digestion with the methylation-sensitive restriction endonuclease HpaII. Besides, higher methylation was observed in indirect somatic embryogenesis, in comparison to direct somatic embryogenesis (Cristofolini et al. 2014). These authors concluded that somatic embryos cultured continuously on 2,4-D-containing medium gradually accumulated increased DNA methylation over time. Such an increase in GDM as an effect of 2,4-D cultivation overtime was also reported by Fraga et al. (2012). On the other hand, the reduction of DNA methylation using the hypomethylating compound 5-Azacytidine (5-AzaC) enhanced SE in A. sellowiana. Somatic embryos of A. sellowiana treated with 10 µM 5-AzaC and 200 µM 2,4-D stimulated SE induction, although it decreased embryo-to-plantlet conversion (Fraga et al. 2012).

Bactris gasipaes

Analysis of GDM was used to evaluate the genetic stability of regenerated cultures after cryopreservation of B. gasipaes embryogenic cultures. Exposure of the cultures to PVS3 vitrification solution (50%w/v sucrose, 50%w/v glycerol) increased global DNA methylation in comparison to non-treated cultures (24.56% for non-treated cultures vs 25.28–25.83% of GDM depending on incubation time in PVS3 solution; Heringer et al. 2013).

Similarly, culture multiplication in a temporary immersion systems RITA® and modified twin flasks showed lower GDM rate (27.5% and 28.0% respectively), when compared to Petri dishes (28.7%), Erlenmeyer flasks (29.2%), and twin flasks (28.8%) systems (Heringer et al. 2014). This lower GDM rate was ascribed to the high protein expression and embryogenic capacity, characteristics of totipotent cells.

Ree et al. (2020) analyzed a two-year-old vigorous embryogenic cell line and an eight-year-old slow-growing cell line of the same genotype that had lost its ability to form somatic embryos. Shotgun proteomic analysis highlighted several key differences. Eight-year-old formerly embryogenic cultures contained an up-accumulation of proteins involved in cell wall formation and reformation, peroxidases, and phospholipase D (Ree et al. 2020). Embryogenic cell cultures accumulated significantly more proteins involved in cellular redox control, fermentation, protein degradation and recycling, NO-related synthesis and storage, and proteins involved in epigenetic regulation. Treatments with 5-AzaC (4, 16, or 64 µM) led to a small proportion of treated eight-year-old cultures regaining the ability to form new somatic embryos, and these embryos showed similar protein profiles to two-year-old embryogenic cultures (Ree et al. 2020). This resumption of somatic embryo growth by the old cultures after the treatment with the hypomethylation compound suggests that hypermethylation may have occurred over time. The reversion of this age-dependent hypermethylation may have activated some essential genes that may show reduced expression over time, allowing them to return to their formerly embryogenic state.

Challenges in getting somatic embryos

Several advances were reached until now towards the development of protocols for SE of A. angustifolia, A. sellowiana, and B. gasipaes (Tables 1, 2 and 3). However, converting somatic cells into somatic embryos is not an easy task, and among the advances, some problems remain. Considering the several aspects involved in the initiation and permanence of SE process, several variables may be investigated towards improving SE workflows. In this sense, we point out some strategic perspectives (Table 4) towards the effective generation of plants of these species through the route of SE.

Table 4 Summary of investigations necessary for overpassing the main remaining constraints in the SE workflows of A. angustifolia, A. sellowiana, and B. gasipaes
Full size table

Araucaria angustifolia

A suggested line of research for SE in A. angustifolia is to test the supplementation of the cultures with different exogenous modulators that could enhance the embryogenic potential for the further development of somatic embryos. Also, studies on embryogenic cultures maturation under low temperatures are lacking and could be a start point for key physiological chains. This species grows under below 10 °C during the winter and from pollination (end of the first year’s winter) to embryo (seed) maturation (from autumn to winter of the 2nd year) takes about 22–23 months (Mantovani et al. 2014; Goeten et al. 2020). Thus, temperature stress may be an important factor in the physiological development and maturation of the A. angustifolia somatic embryos and should be investigated.

Finally, studies on microRNAs present in zygotic embryos and embryogenic cultures may help to characterize potential control points in the conversion of A. angustifolia embryos into plantlets, complementing the few studies on gene expression control during SE of this species.

Acca sellowiana

It was demonstrated that somatic embryo conversion in A. sellowiana was significantly improved when pre-cotyledonary somatic embryos were placed in a 4.5 °C cold treatment for 48 to 80 h (Guerra et al. 1997). On the other hand, embryogenic cultures maintained at 4.0 °C for 20 days did not improve the plantlets' conversion from somatic embryos (Mengarda et al. 2009). Thus, further investigations on this issue are needed to understand the effect of low temperature in the SE of this species.

In addition, the development of cryopreservation protocols is an important step towards promoting the germplasm conservation for A. sellowiana. Guerra et al. (2001) demonstrated that the mechanical resistance of the alginate capsule allied with the storage at low temperatures prevents the germination of the encapsulated somatic embryos of A. sellowiana. Additionally, encapsulated somatic embryos significantly increase the survival of plantlets in comparison to direct conversion (Cangahuala-Inocente et al. 2007). Thus, the encapsulation of somatic embryos into synthetic seeds and their cryopreservation rise as a potential strategy for the conservation of cloned genotypes of this species. However, encapsulation resulted in lower frequencies of conversion in comparison to naked somatic embryos (Cangahuala-Inocente et al. 2007), and no efficient cryopreservation protocol was established for this species until now. So, improving protocols for encapsulation and cryopreservation of A. sellowiana is a current demand in the SE workflow of the species.

Finally, the rates of abnormal somatic embryos in A. sellowiana as high as 76.3% reported by Pescador et al. (2008) and 78% by (Cristofollini et al. 2014) are a big concern for the SE in this species. Adjusting PGR concentrations, time-length of the induction phase, and conditions for somatic embryos conversion into temporary immersion systems are also key issues for future studies towards an efficient protocol of SE of A. sellowiana.

Bactris gasipaes

The main bottleneck to B. gasipaes micropropagation through SE is acclimatization since the converted plantlets lack protection against severe dehydration outside the humid in vitro vessels, leading to a high mortality rate. In such a case, the use of gas-porous membranes can not only improve acclimatization but also benefit the conversion of somatic embryos. The use of these membranes allows the efficiency of gas exchanges, reducing the accumulation of ethylene, and favoring the maintenance of CO2 concentration, stimulating photosynthesis. In addition, gas-porous membranes reduce the humidity inside the culture flasks, benefiting the ex vitro acclimatization of the seedlings and reducing mortality.

Since the main difficulty in acclimatizing in vitro generated plantlets concerns the reduced photosynthetic capacity and malfunctioning of stomata, the use of priming and/or biopriming approaches may be another alternative to help the acclimatization of B. gasipaes plantlets. The integrated use of priming and biopriming was employed in micropropagation of hybrid sugarcane plants (Saccharum sp.) in temporary immersion bioreactors, improving vigor, multiplication, and adaptability rates (Bernal et al. 2008).

Concluding remarks

Understanding how SE processes occur in A. angustifolia, A. sellowiana, and B. gasipaes is not only important for researchers working solely with these species. It can also be a significant instrument supporting the knowledge on this issue for other conifers, dicots, and monocots from tropical and subtropical regions of the world. For instance, exploring different techniques such as droplets vitrification over different support materials, adjusting rewarming bath temperature and rewarming duration, and the supplementing the cryoprotectant medium with salts, Ree and Guerra (2020) reached culture regrowth rates above 90%. Such advances may now be employed towards expanding these factors for use in other species to increase the regrowth potential of vitrified in vitro cultures of plants with recalcitrant seeds.

Tropical and subtropical forests are quite important ecosystems for the equilibrium of the climate and maintenance of the world’s biodiversity. Guaranteeing forests’ subsistence under a scenario of global climatic changes and unsustainable anthropic use of the natural resources is, thus, essential for the future of mankind. In this context, the use of biotechnological tools is a key strategy towards promoting the forests safeguard in a relatively prompt and efficient way.

So far, numerous advances were reached in SE of tree species, encouraging the use of this biotechnological tool into new perspectives, in addition to genetic improvement and production of clonal forests. Our growing insight into SE and other in vitro micropropagation procedures brings the hope that the beauties will overpass the problems, allowing the progress of the conservation and the sustainable use of the plant genetic resources of the world’s tropical and subtropical forests.

Lastly, global environmental change has and will have various effects on the reproductive biology of plant populations (e.g. Aguilar et al. 2006; Aizen and Vázquez 2006). Climate change, habitat fragmentation, and pollinators decline will all directly impact the capacity of plant populations to reproduce successfully, with consequences for their demography, evolution, and long-term persistence. Further understanding of the biology of plant reproduction will therefore be of crucial importance for dealing with these environmental challenges and for maintaining biodiversity, genetic resources, and human well-being. When considering the agriculture of Europe as an example, the impacts of pollination failure are likely to be great since 84% of the crop species depend on animal pollination, and the estimates of the global service of pollinators for crops were valued as high as US$ 200 billion in 1993 (Wilcock and Neiland 2002). Taking together all these findings led us to consider that in the worst scenario zygotic embryonic and seed development would be limiting. In this case, the development of reliable in vitro regenerative systems based on SE should be a feasible alternative to be considered.