With respect to the divergence of angiosperms and gymnosperms, many morphological differences of embryogenesis and embryo development have to be considered (Bowe and Coat, 2000; Cairney et al., 2006; Cairney and Pullman, 2007). Many studies using Arabidopsis as a model plant have increased our knowledge of the function of embryogenesis-related genes in angiosperms (Laux et al., 1996; Lotan et al., 1998; Boutilier et al., 2002; Haecker et al., 2004; Breuninger et al., 2008; Jeong et al., 2011; Chung et al., 2016; Horstman et al., 2017). In contrast, little is known about the function of genes that regulate embryogenesis in conifers (Cairney and Pullman, 2007; De Silva et al., 2008; Trontin et al., 2016a). This is because of particular challenges of conifers. These species have long breeding cycles, large physical size, and slow growth. As a result, a number of powerful genetic approaches used in model plants, such as identification of zygotic embryo-defective mutants and T-DNA insertional mutagenesis, are impracticable. Moreover, recalcitrance to vegetative propagation through conventional or tissue culture methods (mostly aging/phase-change effects) often results in lack of an efficient system for direct plant regeneration from selected superior trees. Available methods are, therefore, mostly “retroactive” and based on somatic embryogenesis initiation from immature seeds coupled with cryopreservation to preserve the juvenility of embryogenic tissue. Selected genotypes after field evaluation of somatic seedlings in clonal tests can be propagated from a cryopreserved juvenile stock. This is currently the preferred and powerful strategy in conifers to overcome recalcitrance of mature trees (Klimaszewska and Cyr, 2002; Bonga et al., 2010) to achieve clonal propagation of superior genotypes and for gene functional studies by genetic transformation of embryogenic cultures (Klimaszewska et al., 2007, 2009, 2016).

The first step for assignment of gene functions in conifers is usually based on sequence similarities between conifers and model angiosperms. Most embryogenesis-related genes identified in Arabidopsis (a global network of ca. 300–450 genes) (Tzafrir et al., 2004; De Smet et al., 2010) have homologous candidate sequences in conifers (Cairney and Pullman, 2007; Zhang et al., 2012). The second one is analysis of transgenic plants following overexpression and/or knock-down of the candidate genes. Such reverse genetics approach for functional gene studies remains challenging in most conifers, because significant genomic resources are needed (genome sequence, expressed sequence tags, etc.), as well as efficient genetic transformation and plant regeneration systems (Trontin et al., 2007). Significant progress has been achieved in several conifer species of high economical interest (Pinus, Picea, and Larix) based on Agrobacterium-mediated genetic transformation of embryogenic cultures and plant regeneration by somatic embryogenesis (Trontin et al., 2007; Klimaszewska et al., 2009, 2016; Bonga et al., 2010). Such a system has been developed for reverse genetics in maritime pine (Pinus pinaster Ait.) during the past 15 years (Trontin et al., 2002, 2007, 2016b; Lelu-Walter et al., 2016) and was implemented by multinational EU consortia (e.g., Sustainpine, 2010–2013¹; ProCoGen, 2012–2015²) for functional gene studies (Trontin et al., 2013; El-Azaz et al., 2020).

An embryogenesis-related gene that is crucial for early embryo development in angiosperms is WUSCHEL-related HOMEOBOX 2 (WOX2). In Arabidopsis, WOX2 is required for cell fate decisions and domain delineation in the apical domain during embryo development, including initiation of shoot meristem stem cells (Haecker et al., 2004; Xiao et al., 2006; Breuninger et al., 2008; Zhang et al., 2017). In post-embryonic development, WOX2 seems to be involved in lateral organ formation and separation (Chung et al., 2016). The WOX2 gene encodes a member of the WOX family of homeodomain transcription factors. WOX proteins share a homeobox that encodes a WUS-type homeodomain. In addition to the homeodomain, many WUS/WOX proteins contain a conserved WUS-box (TLPLFPMH) located downstream of the homeodomain in both angiosperms and gymnosperms (Haecker et al., 2004; Nardmann and Werr, 2006; see also Figure 1). Dolzblasz et al. (2016) reported that the canonical WUS-box is essential for stem cell maintenance within shoot apical meristem in Arabidopsis. Moreover, several overexpression studies indicate that WOX family members containing the conserved WUS-box are involved in plant growth and development. For example, induced overexpression of WOX5 in Arabidopsis produces extra layers of stem cells in columella root cap (Zhang et al., 2010). Overexpression of WOX1 in Arabidopsis leads to defects in meristem development (Zhang et al., 2011). Transgenic plants overexpressing WUSCHEL (WUS, founding member of the WOX family) in Arabidopsis led to high frequency of somatic embryo (SE) formation in all tissues without any need for exogenous plant hormone supply (Laux et al., 1996; Zuo et al., 2002). Furthermore, in angiosperms other than Arabidopsis, the overexpression of AtWUS induced somatic embryogenesis and callus formation (Arroyo-Herrera et al., 2008; Solís-Ramos et al., 2009; Bouchabke-Coussa et al., 2013).

In conifers, the transcription factor WOX2 is considered as an important developmental regulator during early somatic embryogenesis (Klimaszewska et al., 2010; Trontin et al., 2016a) and has been proposed as a putative marker for effective initiation of somatic embryogenesis and to predict embryogenic potential (Palovaara and Hakman, 2008; Klimaszewska et al., 2010, 2011; Miguel et al., 2016; Alvarez et al., 2018). Zhu et al. (2016) revealed that the WOX2 from Picea abies (PaWOX2) plays more crucial roles during early embryogenesis (i.e., protoderm formation and suspensor expansion) than during late embryogenesis.

Genomic resources are now accumulating in model conifers, especially for Picea and Pinus species (Birol et al., 2013; Nystedt et al., 2013; Neale et al., 2014; Plomion et al., 2016). For this study, we focused on P. pinaster, as the production of similar genomic sequence data was initiated during the European ProCoGen project (2011–2015²). Moreover, a reference transcriptome has been made available for this species (SustainpineDB database¹) (Canales et al., 2014).

To our knowledge, there has been no report up to now on the overexpression of WOX2 isolated from conifers in both gymnosperms and, more specifically, angiosperms. The genetic transformation of Pinus is a long and difficult process typically achieved in about 12 months. Deregulating WOX2 with same binary vector in both P. pinaster and Arabidopsis would provide additional insights into evolutionary conserved roles of this gene in gymnosperms and angiosperms, particularly during somatic embryogenesis. In this study, we analyzed the constitutive overexpression of WOX2 isolated from P. pinaster (PpWOX2) in both P. pinaster SE and Arabidopsis seedlings. Our results showed that the endogenous expression of PpWOX2 is enhanced in early proliferating embryos and significantly decreased during SE development in P. pinaster. Constitutive overexpression of PpWOX2 in P. pinaster embryogenic cultures affected SE development and promoted callus formation in the root region of regenerated somatic seedling. In Arabidopsis seedlings, PpWOX2 overexpression led to increased propensity to somatic embryogenesis and organogenesis, likely by alteration of expression patterns of embryogenesis-related genes.

In angiosperms, the transcription factor WOX2 is necessary for cell fate and delineation of the apical embryo domain during embryogenesis (Haecker et al., 2004; Xiao et al., 2006). Zhang et al. (2017) demonstrated that WOX2 contributes to the initiation of shoot meristem stem cells in the embryo of Arabidopsis. Several reverse genetic studies have been reported on knockout and overexpression of WOX2 in angiosperms, mainly in Arabidopsis (Haecker et al., 2004; Jeong et al., 2011; Elhiti et al., 2013; Chung et al., 2016; Zhang et al., 2017). For example, loss of AtWOX2 function in zygotic embryos of Arabidopsis has only relatively mild consequences and results in aberrant divisions at the apex of the globular pro-embryo and, occasionally, seedlings with a single cotyledon (Haecker et al., 2004; Jeong et al., 2011). In contrast, overexpression of AtWOX2 in Arabidopsis causes severe growth defects and further morphological phenotypes by impairing plant organ formation (Chung et al., 2016).

In conifers, much attention has been paid to identify genes that are important during embryogenesis [Park et al., 2010; Rupps et al., 2016; reviewed in Cairney and Pullman (2007), Miguel et al. (2016), Trontin et al. (2016a)] but a considerable number of studies are still needed to unravel the precise spatiotemporal function of each gene. Zhu et al. (2016) reported that downregulation of WOX2 of Picea abies (PaWOX2) during early embryogenesis resulted in significant decrease in the yield of mature embryos. In contrast, downregulation of PaWOX2 after late embryo formation had no effect on further embryo development and maturation. In this study, we focused on the functional study on a WOX2 gene isolated from P. pinaster (PpWOX2) by constitutive overexpression of PpWOX2 in both P. pinaster SEs and Arabidopsis seedlings.

The deduced WOX2 protein shares the WUS-type homeodomain and the conserved WUS-box of the WOX family proteins (Figure 1). Based on phylogenetic analyses, the 15 members of the WOX family are divided into three clades: the ancient clade, the intermediate clade, and the WUS clade (modern clade) (van der Graaff et al., 2009; Hedman et al., 2013; Lian et al., 2014; Alvarez et al., 2018). The modern clade contains WUS and WOX1-7 members that are found in seed plants, confirming an evolutionary relationship among WOX genes of this group (Lian et al., 2014; Alvarez et al., 2018). PpWOX2 shares high identity to its homolog in Pinus sylvestris (97%) and Pinus contorta (95%), low global identity to the WOX2 of Picea abies (76%) and Larix decidua (63%), and very low identity to the WOX2 of Arabidopsis (35%) (Hedman et al., 2013; Zhu et al., 2016).

The expression of endogenous PpWOX2 in P. pinaster is high in the SE proliferation step but significantly decreases after SE maturation. No PpWOX2 expression could be detected later in somatic seedlings. These data confirmed (SE development) and extend (somatic plant development) previous results obtained by Alvarez et al. (2018) from a different P. pinaster embryogenic line and following a different protocol for somatic embryogenesis. A similar WOX2 expression pattern has been found in other conifers such as Larix decidua (Rupps et al., 2016), Picea abies (Palovaara et al., 2010a; Hedman et al., 2013), and Picea glauca (Klimaszewska et al., 2011). These differences in WOX2 expression during SE development may be explained by an internal control of PpWOX2 by other genes that are differentially expressed during development, such as polar auxin transport-related genes (PIN genes) (Palovaara and Hakman, 2009). Considering these expression data, it would be interesting to confirm that PpWOX2 can be used in P. pinaster to distinguish in the proliferation stage embryogenic from non-embryogenic tissues. In conifers, the expression of WOX2 is indeed known as a putative marker for effective initiation of somatic embryogenesis and to predict the embryogenic potential (embryogenecity) of a culture (Palovaara and Hakman, 2008; Park et al., 2010; Klimaszewska et al., 2011; Miguel et al., 2016).

Over-expression of PpWOX2 in proliferating SEs of P. pinaster had, apparently, a negative effect on maturation (Figure 3B). Both lines, OE_#11 and OE_#15, were found to have a significant lower maturation ability than the non-transgenic WT control. These results are consistent with those reported by Klimaszewska et al. (2010). In this report, it was shown that induced overexpression of AtWUS in SEs of Picea glauca had a striking effect on SE maturation by disrupting embryo development. According to this result, it may be postulated that WOX genes of (at least) the modern clade influence embryo development in conifers. However, embryos of the empty vector control line (EV-pC05) also showed significant decrease in maturation yield in our experiments compared to non-transgenic WT embryos, suggesting that reduced yield could also result from transgene positional effects and/or technical issues such as differences in physiological aging between transgenic lines and WT controls (discussed in Supplementary Table 1). In maritime pine, physiological aging of embryogenic lines results in reduced maturation yield (Breton et al., 2006; Lelu-Walter et al., 2016; Trontin et al., 2016b). Aging effects could, therefore, result from unperceived differences in management of transgenic and WTs during the long process of genetic transformation. The development of a simplified, rapid, and improved genetic transformation protocol of embryogenic tissues would be profitable to reduce possible protocol-related effects on maturation yield. Furthermore, the choice of a selective agent may influence maturation results. Hygromycin is known to affect negatively maturation yield in P. pinaster when used at 20 mg l^–1 (Trontin et al., 2007). Although hygromycin concentration was reduced to 10–15 mg l^–1 in our experiments, hygromycin could have affected the regeneration of transgenic embryos of the PpWOX2 and control EV-pC05 lines.

The overexpression of PpWOX2 in P. pinaster negatively affected SE development with i) significantly increased frequency of deformed SE, especially asymmetrical cotyledon development (lines OE_#11 and OE_#15, Figure 3C) and ii) non-embryogenic callus formation in the root pole of germinating somatic seedlings, with subsequent inhibition of root growth (especially in the highly overexpressing line OE_#15, Figures 3E, 5D,E). Similarly, Klimaszewska et al. (2010) reported the inhibition of root growth in somatic seedlings of Picea glauca after induced overexpression of AtWUS during SE conversion. Following downregulation of WOX2 during early embryogenesis in Picea abies, a unique function of WOX2 was suggested in conifers during protoderm development and suspensor expansion, which are important steps for proper early embryo development (Zhu et al., 2016).

To further investigate the propensity of young P. pinaster somatic seedlings transformed with PpWOX2 by callus formation, needle/hypocotyl and root explants from two transgenic lines were cultured in the mLV medium (Litvay et al., 1985) without PGRs. No formation of embryogenic tissue could be detected from both transgenic lines and controls, suggesting that PpWOX2 constitutive overexpression had no major and direct effect that could stimulate embryogenic potential in young somatic seedlings without PGRs. mLV is the currently preferred basal medium for P. pinaster to initiate somatic embryogenesis from immature zygotic embryos (with or without PGRs, Trontin et al., 2016b) and secondary embryogenesis from cotyledonary SEs and somatic seedlings (with PGRs, Klimaszewska et al., 2009). Furthermore, mLV is useful to induce non-embryogenic callus formation from young to more mature materials with or without PGRs (Trontin et al., 2016b,c). Therefore, we initially considered that the overexpression of PpWOX2 could be sufficient to stimulate the formation of embryogenic tissues from somatic seedling explants. However, as in most conifer species (see Bonga et al., 2010), initiation of somatic embryogenesis in P. pinaster materials older than immature zygotic embryos (Trontin et al., 2016c) or cotyledonary or young germinating SEs (Klimaszewska et al., 2009) is still difficult to achieve and may require the additional use of PGRs such as auxin and/or cytokinins. Recently, it was shown that abscisic acid (ABA) supplementation can also lead to initiation of somatic embryogenesis in the Douglas fir (Pseudotsuga menziesii) (Walther et al., 2021) and has to be taken into consideration as a signal for SE activation. Induction of secondary somatic embryogenesis from germinated SEs could be obtained but at low rate with a combination of 2,4-D (9.5 μM) and BA (4.5 μM) in P. pinaster (Klimaszewska et al., 2009). In Capsicum annuum, ectopic BABY BOOM expression is not sufficient to induce embryogenesis, and exogenous cytokinin is required for SE formation (Heidmann et al., 2011). However, ectopic expression of BABY BOOM in Arabidopsis and Brassica led to spontaneous formation of SEs and cotyledon-like structures in seedlings without the use of exogenous PGRs (Boutilier et al., 2002). The enhanced embryogenic potential of young somatic seedlings in P. pinaster may require a combination of ectopic expression of embryogenesis-related gene(s) and adequate PGR treatment and/or other environmental conditions.

Instead of embryogenic callus, non-embryogenic callus formation was determined with high frequency in needle/hypocotyl (significant in both lines #11 and #15) and/or root explants (significant in line #15) of PpWOX2 somatic seedlings (Figure 6) cultivated in mLV deprived of PGRs. Non-embryogenic callus formation was similarly significantly enhanced in the transgenic control EV-pC05 line compared to the WT, suggesting some technical issues related with genetic transformation. However, this increase was lower than that observed for the 35S::PpWOX2 lines except in the case of line #11 for the root explants. These results may be explained by alteration in levels of endogenous hormones, particularly auxin, in response to PpWOX2 overexpression in somatic seedlings. Recent studies showed that at the time of stem cell initiation, WOX2 affects the auxin pathway by increase in expression of the auxin transporter PIN1 gene (Zhang et al., 2017). In conifers, Palovaara and Hakman (2009) suggested that polar auxin transport is involved in regulation of the expression of both the auxin efflux carrier (encoded by PIN1) and WOX2. N-1-naphthylphthalamic acid (NPA) treatment of embryos before cotyledon initiation, indeed, disrupted the endogenous auxin pattern and expression of both PIN1-like and WOX2 (Hakman et al., 2009). Interestingly, the expression of PIN homologs was associated with the auxin immunolocalization pattern during cotyledon formation in Picea abies (Palovaara et al., 2010b). It was suggested that correct auxin transport is crucial during transition from early to pre-cotyledonary embryos and that it is involved in the coordinated regulation of WOX2 and PIN1 (Trontin et al., 2016a), which, in turn, could affect both cotyledonary embryo development (especially cotyledon formation) and normal root formation. Further analysis is needed to consider the interplay of PpWOX2 and auxin in P. pinaster and other conifers.

Furthermore, the heterologous overexpression of PpWOX2 in Arabidopsis resulted in transgenic seedlings (T₁ and T₂ generations) with different phenotypes such as normal growth, stunted growth, and embryo-like and leaf-like structures. One-third (35%) of the first generation of PpWOX2-At plants showed embryo-like structures. In addition, a similarly high number of T₁ plants exhibited stunted growth (35%) and leaf-like structures (15%). PpWOX2 expression was comparatively low in T₁ plants exhibiting a normal growth phenotype (similar data to “normal growth”-type of T₂ plants, Figure 9A, right graph). In different phenotypes of T₂ plants, reduced PpWOX2 expression in the normal phenotype compared to stunted- and even more (significant) two-cotyledon phenotypes (Figure 9A) apparently resulted in increase in frequency of the normal phenotype in T₂ plants compared to the other phenotypes (Figure 7B), while a positive link between the expression of PpWOX2 and the frequency of abnormal phenotypes was observed in the T₂ plants. Similarly, the overexpression of PpWOX2 in immature SEs of P. pinaster (Figure 3A) is correlated to decrease in the frequency of normal, well-developed, cotyledonary embryos compared to the WT (Figure 3B), although not significant in our experiments when compared to the EV-PC05 transgenic control. Overall, these results suggest that the high, constitutive expression of PpWOX2 negatively affects the frequency of the normal phenotype in both Arabidopsis and P. pinaster.

PpWOX2 expression is higher in the transgenic T1 and T2 plants than in the WT control material (seedlings, SEs, non-embryogenic callus) and transgenic control EV-pC05 (Figure 9A and Supplementary Table 1). However, the different phenotypes observed in transgenic T1 and T2 plants (Figures 7, 8) are only weakly supported by gene expression data (e.g., lower PpWOX2 expression in the T2 plants with normal and stunted growth than in the “two-cotyledon” plants, Figure 9). Some post-transformation silencing effects related to 35S promoter inactivation (a well-known effect during transgenesis) may be partially responsible for such discrepancy. It can be suspected that transgene inactivation (and associated phenotype reversion) could be high in the T2 generation as the frequency of plants with normal phenotype is significantly higher (60–85%) than in T1 (15–20% only). Different PpWOX2 expression levels revealed by RT-qPCR may, therefore, not reflect the actual levels of PpWOX2 protein production. Our data do not allow for a study on whether transcriptional or post-transcriptional silencing mechanisms could be involved. However, considering the high levels of PpWOX2 expression observed in some plants with abnormal phenotypes, such as the two-cotyledon T2 plants (At#2), and embryo-like and leaf-like T1 structures (Figure 9A), it could largely be an effect of post-transcriptional or even translational regulation of PpWOX2 expression. There is generally low support in our experiments for an effect of PpWOX2 overexpression (Figure 9A) on the germination rate of T1 and T2 plants (Figure 7). Observed data may be more related with various degrees of silencing, as only T2 plants with normal phenotype germinated with excellent rates (similar to those of WT seedlings).

Moreover, in Arabidopsis, the embryo-like and leaf-like phenotypes in T₁ plants, persistence of the stunted growth phenotype in T₁ (35%) and T₂ plants (up to 15%), and production of T₂ plants with two cotyledons (5–20%) or formation of embryo-like structures on the root tip (1.7%) directed us to analyze the endogenous expression of several major embryogenesis-related genes (AtWOX2, AtLEC1, and AtWUS) to investigate a putative association with observed phenotypes. As hundreds of genes are probably involved in Arabidopsis embryogenesis, an extension of this expression study and additional transcriptomic and/or proteomic approaches are required to gain further insight into intrinsic processes (Tzafrir et al., 2004; De Smet et al., 2010; Elhiti et al., 2013).

PpWOX2 constitutive over-expression in Arabidopsis had a globally low effect on endogenous AtWOX2 and AtWUS expressions, which are roughly similar in transgenic lines and controls from the T₁ and T₂ generations.

The expression of AtLEC1 was much more affected, especially in the T₁ generation (significant overexpression). Indeed, transgenic PpWOX2 plants from the T₁ generation with embryo-like and leaf-like structures showed very high expression of AtLEC1 compared to the WT non-transgenic (seedlings, SEs, and non-embryogenic callus) and transgenic (EV-pC05_At) controls. These strong results suggest that overexpression of PpWOX2 in Arabidopsis may stimulate embryogenesis and organogenesis through coordinated overexpression of AtLEC1. Consistent with this outcome, previous reports support that the over-expression of AtLEC1 in Arabidopsis induced somatic embryogenesis in vegetative cells (Lotan et al., 1998).

In addition, the expression of PpWOX2 in transgenic plants with embryo-like and leaf-like structures was much higher (10 − 35-fold) than in transgenic plants with normal growth. Therefore, a plausible explanation is that ectopic expression of PpWOX2 in Arabidopsis might act redundant with AtWOX2 in inducing embryogenesis and/or organogenesis. Another possible explanation is that the overexpression of PpWOX2 in Arabidopsis might result in the over-expression of AtWUS, which activates the expression of LEC1 and promotes somatic embryogenesis. It was reported that AtWUS promoted somatic embryogenesis and activated the expression of LEC1, LEC2, and FUS3 in cotton (Zheng et al., 2014). The authors suggested that AtWUS may alter PIN expression, which could lead to establishment of new auxin gradients. Subsequently, a new auxin response was formed and stimulated the expression of LEC1, LEC2, and FUS3. In addition, the overexpression of GgWUS from Gnetum gnemon (Coniferopsida) in Arabidopsis was observed to induce somatic embryogenesis and organogenesis (Nardmann et al., 2009). In fact, WUS is involved in regulation of both meristematic stem cells (pluripotent) and embryogenic stem cells (totipotent) (Elhiti et al., 2013). Moreover, WUS is known as an embryogenesis marker in embryonic cells, and its overexpression in Arabidopsis resulted in somatic embryogenesis and repeated formation of adventitious shoots in the absence of auxin (Zuo et al., 2002; Elhiti, 2010). We did not obtain clear results of AtWUS expression during our experiments. The data showed some high expression of AtWUS compared to the non-transgenic WT seedling control, but the differences were not significant compared with the other non-transgenic (WT-SE, WT-callus) and transgenic (EV-pC05_At) controls. The heterologous expression of PpWOX2 in Arabidopsis (which only has 35% similarity with AtWOX2, Figure 1) cannot apparently direct increased expression of either AtWUS (Figure 9C) or AtWOX2 (Figure 9B), which, in turn, could affect the expression of AtWUS and/or AtLEC1. Specific experiments are, therefore, needed to confirm this hypothesis.

Considering the emergence of various phenotypes in the T₂ generation following PpWOX2 overexpression, our data showed that the normal/abnormal plant phenotype is associated with altered AtWOX2, AtWUS, and AtLEC1 expression (PpWOX2_At#2). It is suggested that variable levels of ectopic overexpression of PpWOX2 could affect the expression of these three embryogenesis-related genes.

One possible further explanation on how overexpression of PpWOX2 in Arabidopsis can stimulate somatic embryogenesis and organogenesis is the high similarity of the WUS family-specific homeodomain and the WUS-box from PpWOX2 and AtWOX2 proteins. In plants, the WOX homeobox is assigned to a subfamily of homeobox transcription factors that are involved in plant embryonic patterning. However, the function of the WOX homeobox has not yet been analyzed in plants. In animals, it is confirmed that a homeobox gene family (HOX), which is expressed in specific embryo domains, has a major regulatory role during early pattern formation, similar to that of the WOX homeobox family (Haecker et al., 2004). In this study, the complete protein sequences of PpWOX2, AtWOX2, and AtWUS show low similarity (35%), but they share high similarity (78%) in the WUS-type homeodomain and WUS-box. Wu et al. (2007) suggested that functional redundancy in the WOX family was not solely determined by overall protein sequence similarity but by high similarities of the homeodomains.

Thus, it can be concluded that high similarities of the WUS-type homeodomain and WUS-box among PpWOX2, AtWOX2, and AtWUS might be sufficient for functional redundancy in these WOX proteins. This speculation is supported by several lines of evidence:

First, the homology of WOX homeodomains among the WOX genes supports three major clades: the ancient, intermediate, and modern clades. According to phylogenetic analysis, the WOX1-7 and WUS genes are grouped in the modern clade. Some subgroups from the modern clade have been lost in several species (e.g., lack of WOX1/6 subgroup in rice). These studies imply that WOX members in the modern clade may have a conserved and redundant function (Hedman et al., 2013; Lian et al., 2014; Alvarez et al., 2018).

Second, in angiosperms, the expression of WUS and WOX5 is specific to the shoot and root regions, respectively. Furthermore, Sarkar et al. (2007) demonstrated that WOX5 and WUS are interchangeable for stem cell control in Arabidopsis. In contrast, Nardmann et al. (2009) reported that Ginkgo biloba WUS (GbWUS) and Pinus sylvestris WUS (PsWUS) are expressed in both the shoots and the roots, and suggested that the WUS and WOX5 genes are the result of an angiosperm-specific gene duplication.

Our results suggest that constitutive overexpression of PpWOX2 in Arabidopsis seedlings alters the expression of embryogenesis-related genes (AtLEC1 and/or AtWUS), which, in turn, could promote the formation of SEs and organs. Further reverse genetics studies through inducible overexpression of combinations of embryogenesis-related genes and possibly an additional (hormonal) stimulus might be helpful to overcome recalcitrance to somatic embryogenesis of already differentiated tissues in Pinus pinaster. These findings might be helpful to gain insights into conifer embryogenesis and, in best case, to develop strategies for induction of somatic embryogenesis from adult conifer trees (Lelu-Walter et al., 2016; Trontin et al., 2016c).

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: https://www.ncbi.nlm.nih.gov/, KY773924.1.

AR, KZ, JR, and SH conceived and designed the research project with expert guidance from J-FT for the maritime pine work. J-FT provided PN519 as starting material for maritime pine. J-FT and AR provided the plasmids and methods for the plant transformation experiments. SH and AR conducted the experiments, analysed the data and wrote the manuscript. KZ and AR supervised the lab work. All authors reviewed and approved the final version of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.