Phenotypic plasticity is at the core of the ability of plants to adapt to their changing environments, fundamental to their ability to survive and thrive. In majority of the land plants, the body plan is elaborated post-embryonically in an iterative manner by the shoot and root meristems and its associated procambial tissue (Rensing and Weijers, 2021). The iterative units of the shoot are referred to as phytomers, and the root forms a branching architecture by an iterative patterning of the lateral roots on the primary root, with initiation of the shoot-borne and adventitious roots depending on the species (Perianez-Rodriguez et al., 2014). The other important basis for phenotypic plasticity observed in plants is the high degree of regenerative capacity as witnessed during: (a) formation of asexual propagules coupled with dormancy, (b) vegetative propagation, (c) wound healing and (d) reprogramming of tissue explants in tissue culture to form callus, somatic embryos, shoot and root meristems (Ikeuchi et al., 2019; Mathew and Prasad, 2021). These two hallmarks of plants, i.e., (A) the iterative and adaptive body plan (Hallmark 1 [H1]) and (B) phenotypic plasticity defined by their reprogramming/regenerative potential (RP) (Hallmark 2 [H2]), differentiate them from the animal body plan and development. By contrast, the animal body plan is mostly embryonically determined and shows lesser regenerative capacity and phenotypic plasticity compared to the plants. In this article, we bring forth a “plant-specific” embryonic framework integrating the evolutionary and developmental trajectories of land plant embryogenesis, bringing forth a universal self-organizing logic for plant development and regeneration. This conceptual model also integrates the two hallmarks from a shoot – root perspective, the fates of which are reprogrammable during regeneration. A deeper understanding of how plant phenotypic plasticity is established during embryogenesis is crucial as this knowledge will help develop improved models that explain the modular and adaptable nature of plant growth. Ultimately, this will facilitate designing the crops that will be better equipped to withstand increasingly stressful climates.

To propose a new framework for plant embryogenesis, we first critique the limitations of conventional paradigms derived from animal development. We then trace the evolutionary assembly of the embryophyte genetic toolkit, examining key developmental innovations from bryophytes to angiosperms. Next, we demonstrate the need for a more inclusive model by showing the contrasting and distinct developmental patterns seen in the dicot model Arabidopsis thaliana (hereafter referred to as Arabidopsis) model and monocots. Finally, we propose a “universal” model hypothesis for plant embryogenesis that integrates these developmental, molecular, and evolutionary insights to explain the modular and regenerative nature of plant life.

Both plants (sporophyte phase) and animals begin their lifecycle with the formation of the zygote, which is totipotent, the capacity to give rise to the whole organism (Baroux et al., 2008). The genetic model of plant embryogenesis is based on mutants isolated from Arabidopsis but began with a template that was developed based on an animal model, Drosophila (Mayer et al., 1991). Compared to the stem cell niches located in specific organs in animals, during the process of plant embryo development from the zygote, polarity and patterning of the body axis and differentiation of cell/tissue types result in the segregation of the pluripotent stem cells to specific differentiation programs, i.e., the shoot/root meristems and procambium in plants (Scheres, 2007). Thus, the pluripotent stem cell programs that are assembled during embryo development from a totipotent zygote are wired differently to their respective differentiation programs in plants and animals (Heidstra and Sabatini, 2014). This gives rise to an open indeterminate type of post-embryonic development in plants as opposed to a determinate organismal growth, development, and homeostasis in animals. An illustration comparing the life cycle of a plant, and an animal further shows the fundamental difference in their developmental programs ( Figure 1 ). Plant embryogenesis assembles the meristems that participate in iterative and sequential organ growth during post-embryonic development whereas in animals, embryogenesis forms all the adult organs that undergo further growth during post-embryonic development. (Beyond Animal Blueprints: The Plant Developmental Paradigm; detailed in Box 1 )

The developmental hourglass model emerged out of comparative embryological studies in animals, the narrow part of the hourglass referred to as the phylotypic stage, the least divergent stage during their ontogenies (Irie and Kuratani, 2014). The phylotypic stage represents the highly conserved stage during body plan establishment in the diverse animal lineages. This was further reflected in the gene expression studies of animal embryo development using phylotranscriptomic analysis, which showed a transcriptomic hourglass with the narrow part of hourglass representing the transcripts of least divergent and most conserved genes (Irie and Kuratani, 2011) which function in the establishment of embryonic body plan in diverse animal species (Richardson, 2012). Phylotypic stage represents the empirically viable modern version of the “biogenetic law” – ontogeny recapitulates phylogeny (Niklas et al., 2016). Similar work using the transcriptomes of Arabidopsis and wheat embryo stages representing an ontogenetic sequence showed that the transcriptomic hourglass model is also observed during plant embryogenesis in both dicot and monocot species (Quint et al., 2012; Xiang et al., 2019). However, further studies in Arabidopsis showed that the transcriptomic hourglass is also observed during the post-embryonic stages of development, i.e., seed germination (embryo to vegetative stages) and floral transition (vegetative to reproductive stages) reflective of conserved transcriptomes that defines the organization checkpoints enabling a switch between major developmental programs (Drost et al., 2016). This further distinguishes plant and animal development at the molecular level and highlights the open, iterative post-embryonic development in plants (Hallmark 1 - H1) which begins with the establishment of the embryonic shoot program.

Prior to the emergence of the embryophytes, gamete fusion led to a zygote that immediately underwent meiosis to start the haploid gametophyte generation. The embryophyte genetic toolkit which became progressively assembled during the evolution of embryogenesis in the five embryophyte groups has its deep origins in the charophycean algal genome (Bowman, 2022). In plants, the reprogramming (Hallmark 2 – H2) and co-option of developmental and signaling pathways paved the way for the assembly of the genetic programs that regulate embryogenesis and seed development in gymnosperms and angiosperms, notably distinct from animal evolution in which lineage based fating during embryogenesis results in organogenesis and assembly of the body plan (Blanpain and Simons, 2013).

With the advent of genome sequencing of algae/early land plants and the understanding of genetic and developmental mechanisms, the possibility of emergence of new models provides a rich resource to understand plant embryogenesis (Szövényi et al., 2021). From an evolutionary perspective, the embryophytes, comprising of bryophytes, lycophytes, monilophytes, gymnosperms and angiosperms showed: decreasing lifespan of the gametophytes, increasing degrees of complexity of their embryonic body plan and the increasingly prolonged lifespan of sporophytes they form (Szövényi et al., 2019).

Characteristic of eukaryotes is a life cycle defined by an alternation between haploid (gametophyte) and diploid (sporophyte) phases, initiated by meiosis and gamete fusion, respectively. In multiple eukaryotic lineages, including embryophytes, a genetic switch based on expression of paralogous homeodomain (HD) proteins in the two gametes directs the haploid to diploid transition in gene expression. In the chlorophyte alga Chlamydomonas, the haploid to diploid transition is mediated by a heterodimer formed between a minus-gamete-expressed KNOX TALE-HD protein and a plus-gamete-expressed TALE-HD BELL protein following gamete fusion (Hisanaga et al., 2021). Some of the early zygotic targets of the KNOX-BELL heterodimer encode enzymes that remodel the cell wall, preparing the zygote to become a dormant dispersal agent (Bowman, 2022). A similar, although modified, system operates in mosses and liverworts, with KNOX genes expressed in the egg cell and BELL genes expressed in both sperm and egg cells (Bowman et al., 2016). Marchantia MpKNOX1 and MpBELL are also expressed in proliferating tissues in developing sporophytes, suggesting similarity with KNOX1 functions in angiosperm meristems that are formed during embryo development. Both KNOX1-BELL and KNOX2-BELL heterodimers have been retained and co-opted into developmental patterning roles throughout the angiosperm diploid sporophyte development (Bowman, 2022).

Several features that give rise to the embryonic body plan from bryophytes to angiosperms, include: (i) early evolution of indeterminacy [e.g., hornwort sporophyte], (ii) evolution of the shoot meristem and its sustained ability to maintain growth, (iii) evolution of the procambium and the vascular initials, (iv) evolution of the root meristem and (v) evolution of the cotyledon and the seed. These developmental milestones were eventually integrated during evolution by divergent and convergent mechanisms to form the bipolar embryo seen in seed plants.

The emergence of the building blocks that paved the way for the formation of the meristems and an embryonic body plan in later land plants began with the establishment of the bryophytes as the first land plants with a sporophyte that has a distinct morphology.

Four directions of research that converge on a more universal model hypothesis for embryogenesis in land plants are as follows: (1) Developmental and genetic models of early land plants (bryophyte, lycophyte and monilophyte models) and the evolution of developmental modules, i.e., shoot apical meristem, root apical meristem, cambium (vascular), embryogenesis, shoot and root patterning (Radoeva et al., 2019b); (2) Further refinement of developmental and molecular mechanisms in Arabidopsis embryogenesis and its correlations with monocot (Poaceae) embryogenesis (Xiang et al., 2019; Armenta-Medina et al., 2021; Hao et al., 2021); (3) Plant developmental plasticity as gleaned from the molecular mechanisms that regulate plant regeneration and wound healing (Ikeuchi et al., 2019; Mathew and Prasad, 2021); and (4) Contrasting plant versus animal embryo development.

These converge on the two hallmarks (e.g., shoot development as iterative units, root branching elaboration and the reprogramming potential of plants can be used to test the following hypotheses:

In the context of our model, a “module” refers to a discrete, functionally integrated unit within the developing embryo. These modules are characterized by a specific set of genes, regulatory interactions, and resulting cellular behaviors that contribute to a distinct developmental process. For example, a module might encompass the genes and pathways responsible for establishing apical-basal polarity, or those governing cotyledon formation.

To test these hypotheses, the Arabidopsis ontogenetic sequence of embryogenesis will be described in 4 steps, each step defined by the two hallmarks [ Box 3-(I) ]. Some of the underlying mechanisms and cell fates are provided in Box 3-(II) . The linkages between these steps and their evolvability can be used to develop an integrative “universal” model hypothesis for plant embryogenesis from a developmental ( Figure 3 ) and evolutionary perspective ( Figure 4 ). Arabidopsis embryogenesis is well-defined and provides a detailed exemplar to overlay the model’s components while testing the evolutionary assembly of these modules necessitates further studies of comparative functional genomics and developmental studies across embryophyte phylogeny.

The ability of plant cells to form somatic embryos has been reported in several plant species from various tissues and the genetic mechanisms underlying this reprogramming event have been explored (Horstman et al., 2017a). Specifically, ectopic expression of the AINTEGUMENTA-LIKE/PLETHORA (AIL/PLT) clade gene BABYBOOM/PLT4 can induce the formation of somatic embryos (Boutilier et al., 2002). The BBM genes (BBM1, BBM2 and BBM3) in rice have been shown to be critical to male transmitted pluripotency factors to initiate embryo development after fertilization; the triple bbm1 bbm2 bbm3 mutant causes embryo arrest and abortion of the seeds (Khanday et al., 2019). A recent study in Arabidopsis shows that members of the AIL/PLT family, PLT1, PLT2, PLT3 and PLT4/BBM play an important role in the early initiation of embryogenesis and patterning based on: (1) the lethality of plt2/bbm double mutant; (2) the expression of PLT2 and BBM genes in the zygote and apical/basal cells after the first division during early embryogenesis, and; (3) the resemblance between the PLT regulome during early embryogenesis and that of meristematic cells (Kerstens et al., 2024). The redundant roles of the PLT genes in early embryogenesis and in the later specification of the root meristem activity suggests that the PLT genes were evolutionarily recruited for embryonic divisions and meristematic potential and later adapted for meristematic functions in the embryo of bipolar seed plants (Kerstens et al., 2024).

Another example of dual adaptation during seed evolution and reprogramming is observed in the functions of LEAFY COTYLEDON (LEC) gene family members. LEC transcription factors are master regulators of the seed maturation process during late embryogenesis which, when overexpressed during the vegetative phase, reprogram cells to induce somatic embryogenesis (Braybrook and Harada, 2008). Interestingly, BBM-induced somatic embryo formation involves the activation of the LEC TF- containing gene regulatory network (LEC1-ABI3-FUS3-LEC2 [LAFL]) (Horstman et al., 2017b). Recent studies suggest that the LAFL regulatory network is likely involved in phase transition during development via modulation of epigenetic pathways (Niu and He, 2019; Chen and Du, 2022). This highlights the central role of epigenetic mechanisms through which the master regulators like LEC and BBM reprogram cells towards totipotency (Peng et al., 2023).

The embryophyte genetic toolkit has repeatedly evolved to give rise to apomixis in both monocots and dicots (Bowman, 2022). Apomixis, a naturally occurring process of asexual reproduction through seeds whereby the embryo is a genetic clone of the mother plant, enables the preservation of a hybrid genotype over multiple generations and is found throughout the plant phylogeny. In the moss Physcomitrium, ectopic overexpression of the homeobox gene BELL1 induces embryo formation and diploid sporophytes from specific gametophytic cells without fertilization (Horst et al., 2016). Naturally apomictic mechanisms in some angiosperms also appear to involve BLH1 misexpression in the female gametophyte, reminiscent of the eostre mutant (Bezodis et al., 2022). In Hieracium praealtum, the BLH1 orthologue is expressed in apomictic aposporous initial cells and early embryo sac cells (Okada et al., 2013). In Arabidopsis, ectopic expression under a germline-specific promoter of KNOX and BELL genes not normally expressed in the gametophytes both disrupts germ cell specification and causes defects in cell identity throughout gametophyte development – some mirroring events seen in natural apomicts (Bezodis et al., 2022). In Pennisetum squamulatum, multiple copies of BABY BOOM-like (BBM-like) genes that encode transcription factors previously implicated in somatic embryogenesis are found in the genetic locus associated with apomixis (Conner et al., 2015). The transfer of one of the BBM-like genes into pearl millet, rice and maize, is sufficient to trigger embryo formation without fertilization. Alternatively, the dandelion (Taraxacum officinale) PARTHENOGENESIS (PAR) gene encodes a protein with zinc finger and EAR domains (DNA binding and transcriptional repression) to activate embryo formation without fertilization (Wang and Underwood, 2023). Inducing apomixis in crops offers significant breeding potential by enabling the fixation of superior hybrid genotypes across generations through seeds. Transferring genes like BBM-like into crops such as rice and maize have successfully triggered embryo formation without fertilization, demonstrating a viable pathway to engineer this trait (Conner et al., 2017).

Plant cells are developmentally plastic with reference to their cell fate, as displayed by their regenerative capacity to form shoot/root meristems, somatic embryos and callus (Birnbaum and Alvarado, 2008). These regenerative potentials are predominantly activated in the pericycle-like cells in the shoot and the pericycle cells of the root in Arabidopsis, which led to the hypothesis that regeneration occurs via a root development pathway (Sugimoto et al., 2010). Subsequent studies demonstrate that genetic pathways regulating root meristem formation (PLT1/PLT2) during embryonic and post-embryonic stages of development are activated by PLT3/PLT5/PLT7 during the formation of de novo root and shoot meristems from callus. However, in the case of shoot meristem specification, PLT 3/5/7 activation is followed by activation of the shoot meristem pathway gene CUC2 in a two-step process (Kareem et al., 2015). A recent study shows that the mechanical conflict created by differential cell wall loosening in callus progenitor cells which form shoot meristems versus surrounding cells is regulated by CUC2 (Varapparambath et al., 2022). CUC2 activates the expression of XTH9, that encodes a cell wall loosening enzyme in cells surrounding the progenitor cells. This results in the localization of PIN1 auxin transport protein and the polarity protein SOSEKI2 in the progenitor cells, activating de novo shoot meristem formation (Varapparambath et al., 2022). Other pathways triggered by auxin and cytokinin involved in the specification of the shoot/root meristem specification (e.g., WOX and KNOX genes), and the LBD genes involved in lateral root/adventitious/de novo root formation have overlapping roles in the assembly of embryonic body plan in dicot and monocot species (Palovaara et al., 2016; Ikeuchi et al., 2019; Mathew and Prasad, 2021; Rashotte, 2021).

This “universal” model hypothesis of embryo development elucidates four steps that integrates the two hallmarks: (A) an iterative pattern of the phytomer (embryonic phytomer with apical meristem and iterative formation of post-embryonic phytomers by the SAM) and (B) the suppressed regenerative states that are reprogrammable ( Figure 3 ). This model takes into account (1) the post-embryonic activity of the shoot apical meristem that iteratively form phytomers with axillary meristems; (2) the diverse outputs of the post-embryonic root types from root founder cells and collet zone (radicle/primary root; anchor root; shoot-borne root; adventitious root; lateral root; nodal root/crown root) (Dolan et al., 1993; Chandler, 2011; Gonin et al., 2019; Zhang et al., 2023) [ Box 5 ]; and (3) the regenerative potential of plant tissues (e.g.,form shoot/root meristems, somatic embryos, callusing and wound healing). A diagrammatic representation of a GRN that functions as an activation – suppression switch is provided in Box 4 . A detailed exploration of these GRNs that represent the activation - suppression switches will be the focus of an ensuing article.

The emergent ‘proto-meristems,’ which gave rise to shoot and root programs during the early evolution of land plants (as discussed in the Evolutionary origins section), represent foundational elements. These elements convergently and progressively integrated to form the bipolar embryonic body plan observed in extant lycophytes, monilophytes, gymnosperms, and angiosperms. The evolutionary origin and formation of meristems in different plant groups ( Figure 4 ) provides a foundation for understanding the developmental and genetic pathways that gave rise to the “universal embryonic body plan” defined by dicot and monocot embryos ( Figures 2 , 3 ). With the sequencing of genomes from across the land plant phylogeny (including early-diverging groups like bryophytes, lycophytes, and monilophytes, as well as later-diverging gymnosperms and angiosperms), comes a better understanding of the genetic and molecular mechanisms that underly embryonic and post-embryonic development in these different plant species. Looking within the angiosperms (dicots and monocots), an evolutionary perspective of the plant embryonic body plan provides a rich template to integrate the modular assembly of the shoot units during post-embryonic development (Hallmark 1) with the reprogramming potential as revealed by the regenerative pathways (Hallmark 2) into one framework. The proposed universal plant embryonic body plan that is distinct from the animal embryo models developed by this revised approach facilitates probing deeper into the genome of the land plants with reference to their genetics and adaptation. This model also offers a framework for interpretation and guides future research into the precise nature of the GRNs, further enabling the identification of both conserved and diversifying gene regulatory networks. The genetic mechanisms uncovered will, in turn, provide an enhanced ‘toolkit’ for understanding plant adaptation offering alternate blueprints for developing stress resilient crops that will be better adapted to future food production challenges in a changing global climate.