De novo assembly of plant body plan: a step ahead of Deadpool

Abstract While in the movie Deadpool it is possible for a human to recreate an arm from scratch, in reality plants can even surpass that. Not only can they regenerate lost parts, but also the whole plant body can be reborn from a few existing cells. Despite the decades old realization that plant cells possess the ability to regenerate a complete shoot and root system, it is only now that the underlying mechanisms are being unraveled. De novo plant regeneration involves the initiation of regenerative mass, acquisition of the pluripotent state, reconstitution of stem cells and assembly of regulatory interactions. Recent studies have furthered our understanding on the making of a complete plant system in the absence of embryonic positional cues. We review the recent studies probing the molecular mechanisms of de novo plant regeneration in response to external inductive cues and our current knowledge of direct reprogramming of root to shoot and vice versa. We further discuss how de novo regeneration can be exploited to meet the demands of green culture industries and to serve as a general model to address the fundamental questions of regeneration across the plant kingdom.


INTRODUCTION
Wear and tear is an inevitable part of normal growth and development in all lifeforms. An organism which cannot repair wounds and regenerate lost body parts will have compromised survival fitness compared to its rivals. Across kingdoms, nature has equipped living cells and tissues of organisms with a remarkable capacity to regenerate in order to conquer biotic and abiotic threats. Both plants and animals share this capacity as a common means of tissue repair despite having independent evolutionary origins (Birnbaum & Sánchez Alvarado, 2008;Sugimoto, Gordon, & Meyerowitz, 2011;Pulianmackal, Kareem, Durgaprasad, Trivedi, & Prasad, 2014). The underlying mechanism of regeneration is likely to have adopted similar strategies in both the kingdoms during evolution in concurrence with other developmental processes (Meyerowitz, 2002). The last decade has witnessed an extensive interrogation into regeneration in plants and animals in an attempt to unearth the molecular mechanism and to draw parallels between the two kingdoms (Takahashi & Yamanaka, 2006;Xu et al., 2006;Gordon et al., 2007;Birnbaum & Sánchez Alvarado, This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. 2008; Sena, Wang, Liu, Hofhuis, & Birnbaum, 2009;Sugimoto, Jiao, & Meyerowitz, 2010;Sanchez Alvarado & Yamanaka, 2014;Kareem et al., 2015;Efroni et al., 2016;Ikeuchi, Ogawa, Iwase, & Sugimoto, 2016). Regeneration has been a focus of attention to study the basic principles of cellular plasticity and other developmental processes. In addition, this knowledge has been exploited in regenerative medicine and horticulture.
Plants have the remarkable ability to regenerate an entire plant from a few existing cells and also to regrow lost parts. The regeneration potential of plants has been recognized for long but it is a more recent discovery in animals (Birnbaum & Sánchez Alvarado, 2008). Regeneration has been observed in plants ranging from lower forms like algae and bryophytes to higher flowering plants (Birnbaum & Sánchez Alvarado, 2008;Duclercq, Sangwan-Norreel, Catterou, & Sangwan, 2011;Pulianmackal et al., 2014;Ikeuchi et al., 2016).

Bud formation upon the decortication of elm tree demonstrated by
Duhamel in 1756 is considered to be the pioneering attempt in plant regeneration (Gautheret, 1985). Experimental approaches to mimic this natural capacity under in vitro culture conditions became feasible (E) Direct shoot regeneration from root explants on suitable inductive medium after realizing the totipotent nature of plant cells (Haberlandt, 1902).
Thanks to the work of Skoog and Miller on the identification of suitable hormonal combinations (Skoog & Miller, 1957), in vitro formation of entire plants or organs is now a simple process which has been widely exploited in various plant species for the past 60 years (Smith, 2013;Vasil & Thorpe, 2013). However, unlike in plants, de novo organogenesis under simple hormonal combinations is yet to be a reality in animals.
De novo organogenesis in plants can be achieved either directly or indirectly (Fig. 1); however, cellular reprogramming is inevitable in both the modes. During direct regeneration, the cells undergo transdifferentiation wherein root cells are reprogrammed to shoot and vice versa (Chatfield et al., 2013;Liu et al., 2014;Kareem et al., 2016). The indirect mode involves the formation of an intermediate regenerative mass of cells called callus from adult stem cells that are distributed throughout the plant body (Sugimoto et al., 2010(Sugimoto et al., , 2011Duclercq, Sangwan-Norreel, Catterou, & Sangwan, 2011). The callus in turn produces root or shoot in response to an appropriate ratio of the key plant hormones auxin and cytokinin (Valvekens, Van Montagu, & Van Lijsebettens, 1988;Gordon et al., 2007;Atta et al., 2009).
Nevertheless, the primary step towards organ regeneration is the acquisition of competence. Once cells have acquired the competence, the dynamic assembly of molecular interactions defines the patterning of new meristem and organ primordia. In this review we discuss de novo regeneration in plants in response to external inductive cues and its plausible comparison with animal counterparts. We focus on celltype specificity, regulatory modules conferring the competence and the assembly of regulatory interactions leading to the completion of regeneration.

NATURE OF THE EXPLANT AND DE NOVO REGENERATION
As per cell theory, new cells arise from pre-existing cells (Virchow, 1859). In plant regeneration, the source of all new tissues is the explant procured from the mother plant which is to be propagated.  (Halperin, 1986;Valvekens et al., 1988;Weigel & Glazebrook, 2002;George, Hall, & De Klerk, 2007;Sugimoto et al., 2010). Factors such as availability, age of explants, response rate, contamination susceptibility, lethality of phenolic exudations, and the objective of tissue culture govern the suitability of the explant (Smith, 2013).
In vitro plant regeneration may be performed for a variety of purposes and each requires the most suitable explant type. For the general purpose of micropropagation, any explant that possesses the maximum regeneration efficiency can be used. However, the generation of haploid plants requires the use of haploid explants such as microspores, egg cells, or tissues bearing the same, such as anther, ovary, or inflorescence (George et al., 2007;Vasil & Thorpe, 2013).
A commonly used explant for hairy root induction by Agrobacterium rhizogenes for secondary metabolite production is root tip (Flores, Hoy, & Pickard, 1987). To generate virus-free plants, shoot apical meristem is the best choice of explant due to the meristematic nature and the lack of connection to differentiated vascular tissue which prevents the spread of viral infection (Slack & Tufford, 1995). The problem of endophytic microbial associations can be overcome by the use of tissues from plants grown in vitro. This can also help in the conservation of the natural population of the donor plant.
Despite the availability of a wide range of explants, the regeneration response relies heavily on the nature of the explant. The response varies between species, genotypes, ecotypes, organs of the same plant, and even between sections of the same organ (Coleman & Ernst, 1989;Akama et al., 1992;Siemens, Torres, Morgner, & Sacristán, 1993;Zhang, Takahata & Xu, 1998;Motte et al., 2014). In Arabidopsis, root and hypocotyl have been shown to be the explants with the highest regeneration potential while cotyledon displays the lowest regeneration potential (Valvekens et al., 1988;Akama et al., 1992). The underlying cause of differences in regeneration potential between various explants may be the presence of more regenerative cells in the explant tissue of higher efficiency. Unlike in Arabidopsis, cotyledon displays the highest shoot regeneration potential in Brassica spp. (Tang et al., 2003;Guo, Zhu, Hu, & Zheng, 2005). This implies that the regeneration potential of the same tissue can vary in different species. In addition, the extrinsic cues such as hormones and culture conditions required for organogenesis may vary for diverse explants (Sugimoto et al., 2010). The endogenous cues from the donor plant to which the explant has been habituated may also have a role in in vitro response. For instance, leaf explants closer to the shoot apex are more responsive in culture (Chaudhuri, Pal, & Jha, 2008). This enhanced response may be due to the relatively young developmental stage of the explants closer to the shoot apex.
Age of the explant is an important factor that influences regeneration capacity (Sugimoto & Meyerowitz, 2013). It has been observed that older leaf explants have reduced root and shoot regeneration efficiency compared to younger leaf explants (Chen et al., 2014;Zhang et al., 2015). The reduced regeneration of root and shoot is partly attributed to the reduced levels of free endogenous auxin and defective cytokinin signaling mediated by micro RNA (miR156), respectively.
In older explants there is a decline in miR156. As a result SQUAMOSA PROMOTER BINDING PROTEIN LIKE (SPL), which is normally under the repression of miR156, interferes with the transcriptional activity of B-type ARABIDOPSIS RESPONSE REGULATORS (ARRs) and impedes the cytokinin signaling pathway. Thus, the shoot regeneration capacity is adversely effected (Zhang et al., 2015). Providing the required hormone extrinsically in the culture medium can recover the regeneration potential of the older explant to some extent under these conditions. Another hormone that is influenced by the aging of the plant is abscisic acid (ABA). ABA in combination with cytokinin enhances shoot regeneration (Paulraj & Yeung, 2012). However, the underlying molecular mechanism is yet to be studied. Age dependence may also be a contributing factor to the loss of plasticity and competence to proliferate and the inability to switch fates in more mature tissues compared to the totipotent embryonic status of younger plant cells.
Regardless of the effort of standardizing the above mentioned factors, the response of a different genotype or ecotype of a plant may still show variation in regeneration efficiency (Motte et al., 2014). The regeneration potential of root explant varies from 0% to 100% in various Arabidopsis accessions. Among the commonly used laboratory ecotypes of Arabidopsis, Wassilewskija (Ws) shows higher regeneration potential than the other popular ecotypes Columbia (Col-0) and Landsberg erecta (Ler) (Chaudhury & Signer, 1989;Akama et al., 1992;Motte et al., 2014). The basic cause of this difference in regeneration capacity is due to the genetic differences that may have arisen as a part of inbreeding and crossbreeding in the case of genotype and as a result of prolonged growth and adaptations to a particular environmental condition in the case of ecotypes. The key quantitative trait loci involved in this variation have been characterized by combined linkage and association mapping (Motte et al., 2014).
As the saying goes, well begun is half done. Therefore, the wise choice of the explant is a prerequisite to obtain desired in vitro response with maximum regeneration capacity. This will require standardization of a combination of culture parameters, keeping in mind the dynamics of the endogenous growth hormones and the physiological status of the explant.

CELL-TYPE SPECIFICITY, CONCEPT OF ADULT STEM CELLS AND FORMATION OF REGENERATIVE MASS
When explants are exposed to the phytohormone auxin present in callus inducing medium, a proliferating regenerative callus forms.
The regenerative callus can further give rise to either root or shoot depending upon the external inductive stimulus. This is attributed to plant cell totipotency which confers the ability to elaborate an entire plant body. For long, it had been thought that all plant cells are totipotent and can undergo dedifferentiation to generate callus (Birnbaum & Sánchez Alvarado, 2008). But recent developments in the understanding about cell-type specificity and differentiation status during callus formation have redefined the concepts of cellular potency and reprogramming (Sugimoto et al., 2010(Sugimoto et al., , 2011Ikeuchi, Sugimoto, & Iwase, 2013). Unlike previously thought, not all somatic cells can form callus. It is now evident that callus formation is initiated by the activation of a specialized population of partially differentiated cells which span throughout the plant body (Fig. 2). These partially differentiated cells are called adult stem cells (Sugimoto et al., 2011).
In general, adult stem cells are defined as pre-existing stem cells that can proliferate and differentiate into various cell types upon appropriate inductive cues. The adult stem cell populations in plants comprise pericycle and procambium cells (Sugimoto et al., 2010(Sugimoto et al., , 2011Liu et al., 2014) which have been known to generate post-embryonic structures such as lateral root and vascular tissue respectively (Mähönen et al., 2000;De Smet, Vanneste, Inze, & Beeckman, 2006).
However, the concept of adult stem cells has to be better defined in plants as in animals (Sugimoto et al., 2011;Rezza, Sennett, & Rendl, 2014).
Several studies support the highly versatile nature of pericycle cells which empower them as adult stem cells. In a post-embryonic root, pericycle cells remain quiescent all along the root except at the sites opposite to the xylem poles from where lateral root initiates (Malamy & Benfey, 1997;De Smet et al., 2006). Under specific external inductive cues the xylem pole pericycle cells can also give rise to callus or convert into shoot fate (Atta et al., 2009;Sugimoto et al., 2010;Kareem et al., 2015Kareem et al., , 2016. The flexible nature of pericycle cells is explained by their partial differentiation status, but the molecular basis for the partial differentiation status is unknown. However, some studies have shown that non-pericycle cells can also give rise to callus. During the protoplast culture of Arabidopsis leaf after the enzymatic removal of cell wall, mesophyll cells can be reprogrammed into the callus fate (Chupeau et al, 2013). Similar observations have been reported in several Nicotiana spp. (Bourgin, Chupeau, & Missonier, 1979) and also in green algae Bryopsis plumose (Kim, Klotchkova, & Kang, 2001). Moreover, mutants defective in biosynthesis of cell wall components such as pectin and cellulose show hormone-independent callus formation (Frank et al., 2002;Iwai, Masaoka, Ishii, & Satoh, 2002). This opens up the prospect of reprogramming additional cell types for callus induction. The removal of cell wall is likely to impact the state of cells by altering the mechanical properties such as turgor pressure and the stress experienced by cells thereby triggering callus formation.

F I G U R E 2 Schematic representation of the two stages involved in callus mediated de novo regeneration
It is interesting to examine if cellular reprogramming during pluripotent callus formation will lead to embryonic ground state. Callus displays a gene expression pattern resembling that of the basal half of the embryo. From this, one may infer that callus has basal embryolike features. But several lines of evidence suggest that the formation of pluripotent callus largely follows the molecular developmental program of lateral root initiation (Sugimoto et al., 2010). The strongest evidence is that aberrant lateral root formation 4 (alf4) mutant which fails to form lateral root (Celenza, Grisafi, & Fink, 1995) also does not form callus (Sugimoto et al., 2010). Similarly plethora3 (plt3);plt5;plt7 triple mutant makes lateral root primordia (LRP) but these cells are not pluripotent as they fail to develop all cell types of lateral root (Prasad et al., 2011;Hofhuis et al., 2013). This mutant is able to make callus but it is not pluripotent as it fails to regenerate organs (Kareem et al., 2015). But neither alf4 nor plt3;plt5;plt7 mutants show any defects in embryogenesis (Celenza et al., 1995;Prasad et al., 2011). Therefore, callus formation from various parts of plants follows a lateral root development pathway and callus predominantly displays a gene expression pattern similar to lateral root meristem (Sugimoto et al., 2010;Kareem et al., 2015). These studies suggest that callus may not be an undifferentiated tissue; rather it possesses differentiated rootlike features. Further, callus does not go back to the zero point of embryonic state to achieve the pluripotent state. An intriguing question arises why callus derived from various organs should retain rootspecific genes. Insights into the functional significance of root-specific genes in the callus emerge from recent studies which show that calluses which fail to express root stem cell maintenance regulators are not pluripotent (Kareem et al., 2015).

Pericycle competence
The formation of callus is under the tight control of the molecular machinery which determines cellular competence to re-enter the cell cycle. One of the important factors that provides competence to pericycle cells is ALF4 whose loss of function mutant is defective in callus formation from multiple organs such as root, cotyledon, and petal (Sugimoto et al., 2010). ALF4 is a transcription factor expressed in multiple organs to maintain the cells in a mitotically competent state.
Initially alf4 mutant has been described to be deficient in LRP formation due to the absence of cell division in pericycle cells (Celenza et al., 1995). These studies provide strong evidence for the direct link between the necessity of pericycle competence for the formation of LRP and callus (Sugimoto et al., 2010). The necessity of ALF4 for the division of leaf protoplast suggests its key role in making non-pericycle cells also competent to enter the regeneration program (Chupeau et al., 2013). In addition, ALF4 is required for the wound healing process during grafting (Melnyk, Schuster, Leyser, & Meyerowitz, 2015). A recent study demonstrates that the cell layer signal of very-long-chain fatty acids (VLCFAs) restricts the callus forming capacity of pericycle cells in Arabidopsis partly by inhibiting ALF4 transcription (Shang et al., 2016 (Fukaki, Tameda, Masuda, & Tasaka, 2002;Shang et al., 2016). These studies imply that the developmental programs leading to callus formation might not be sufficient to trigger lateral root formation.

Hormonal signaling cascade and induction of key regulators
While VLCFA regulated ALF4 is the initial trigger, auxin-rich culture medium induces a cascade of molecular players. For instance, auxin activates four LATERAL ORGAN BOUNDARIES DOMAIN (LBD) transcription factors that act downstream to AUXIN RESPONSE FACTOR 7 (ARF7) and ARF19 to initiate callus formation (Fan, Xu, Xu, & Hu, 2012).
The ectopic overexpression of these transcription factors, namely LBD16, LBD17, LBD18, and LBD29, can trigger callus formation without the supply of exogenous auxin. In contrast, suppression of these transcription factors inhibits callus formation even in the presence of auxin, suggesting the key role of these genes in callus formation (Fan et al., 2012). The functional significance of both ARF and LBD genes in lateral root formation has been identified (Okushima, Fukaki, Onoda, Theologis, & Tasaka, 2007) which emphasizes the notion that callus formation largely follows a lateral root development program.
The downstream pathway of LBD mediated callus formation is vague.
However, based on studies on LBD mediated lateral root development the downstream mechanism of callus formation can be inferred. During lateral root development, LBD18 activates cell wall loosening factor EXPANSIN .
Together with LBD33, LBD18 induces the cell cycle regulator E2 PRO-MOTER BINDING FACTOR a (E2Fa) which dimerizes with DIMERIZA-TION PARTNER (DP) protein for cell cycle entry (Berckmans et al., 2011). However, these cell cycle regulators alone are insufficient to trigger callus compared to LBD genes (Ikeuchi et al., 2013). Additional cell cycle regulators induced by LBD and other transcription factors need to be identified to further our understanding of the key transcription factor mediated mechanism of callus formation. Nevertheless, cell cycle activation is a key regulatory checkpoint during callus formation.
A recent study has shown that callus formation is also regulated by miRNA mediated interplay between auxin and cytokinin signaling . miR160 acts as a key repressor of callus formation by cleaving the mRNA of auxin signaling gene ARF10 and subsequently activating cytokinin signaling gene ARR15. Thus endogenous hormonal signaling interactions play a critical role in callus formation despite the external hormone supplement.

Establishment of pluripotent state
The sheer formation of callus may not determine the ability to regenerate. Callus needs to be pluripotent for the subsequent regeneration of different organs such as shoot and root. The pluripotent state of callus is established by root stem cell maintenance regulators (Kareem et al., 2015) such as AP2/ERF transcription factors PLT1 and PLT2 (Aida et al., 2004;Mähönen et al., 2014 (Kareem et al., 2015). PLT3, PLT5, and PLT7 are rapidly induced upon external auxin application. The cumulative loss of function of these genes (plt3;plt5;plt7) leads to the formation of callus lacking the expression of root stem cell maintenance regulators despite the auxin-rich culture medium. The mutant callus completely loses the potential to regenerate and the reconstitution of root stem cell maintenance regulators re-establishes the pluripotent state.
In addition to PLT1 and PLT2, other root-specific stem cell maintenance regulators are also likely to function in a similar fashion under the control of PLT3, PLT5, and PLT7. Root stem cell maintenance regulators are predominantly expressed in callus during its induction phase and reinstate the pluripotent state. These are downregulated when the callus is shifted to cytokinin medium for shoot induction. When the high level is sustained even upon shoot induction by forced expression in a heterologous fashion, a default pathway of root formation is opted for despite the abundance of cytokinin. This signifies the tight temporal and spatial control of root stem cell maintenance regulators during regeneration. The necessity of root stem cell maintenance regulators to establish the pluripotent state in callus emphasizes the significance of the root-like trait of callus derived from any part of the plant (Sugimoto et al., 2010;Kareem et al., 2015). Intriguingly, some of the mutant combination of root stem cell maintenance regulators such as PLT2 and PLT4 are embryonic lethal (the mutant cannot make all the embryo cell types) (Galinha et al., 2007). This suggests that, in all the scenarios such as embryogenesis, lateral root formation, and callus formation, the common function of root stem cell maintenance regulators is to establish the pluripotent state. Future studies are required to understand how root stem cell regulators establish the pluripotency.

Wound induced callus formation
In addition to auxin induced callus formation, pluripotent callus can also be formed through other pathways. Wound induced callus formation is one such pathway (Iwase et al., 2011

PLAUSIBLE COUNTERPART OF CALLUS IN OTHER ORGANISMS
The formation of blastema in animals can be considered analogous to callus formation. Blastema formation is one of the first steps in epimorphic regeneration predominantly seen in certain lower ani-

DE NOVO ASSEMBLY OF REGULATORY INTERACTIONS, RECONSTITUTION OF SHOOT STEM CELLS, AND DE NOVO ORGANOGENESIS
Following the formation of pluripotent callus, a subsequent hormonal treatment is required to accomplish de novo organ regeneration ( Fig. 2). Culturing the callus on cytokinin-rich shoot inducing medium (SIM) favors shoot regeneration, whereas auxin-rich root inducing medium (RIM) promotes root regeneration (Che et al., 2006;Gordon et al., 2007). Root regeneration from the callus is a relatively simple process as callus displays a root-like trait. However, de novo shoot regeneration from callus is a more complex process.
During de novo shoot formation, culturing on SIM instigates the callus to create spatial domains of hormone perception that lead to partitioning of cell identity and re-specification of cell fate from root-like fate to shoot fate. In concurrence with the assembly of regulatory molecular interactions, these cells acquire the competence to regenerate shoot progenitors. Subsequently, formation of shoot promeristem and functional shoot apical meristem (SAM) with lateral organ patterning ensues (Gordon et al., 2007; 2015) (Fig. 3). Morphologically, this process involves the greening of callus (also denoted as regenerating green foci) followed by the initiation of shoot. These sequential developmental events during de novo organogenesis are discussed in more detail in the following sections.

Uncoupling intermediate developmental phases of shoot regeneration: competence to regenerate shoot progenitors and completion of shoot formation
De novo shoot organogenesis progresses through several intermediate developmental phases. Formation of any of these phases such as green callus, shoot progenitor, or shoot promeristem is not an assur- of their origin. In parallel, de novo organogenesis from iPSCs in animals also follows multiple steps. Regeneration of organs such as kidney (Yokote, Yamanaka, & Yokoo, 2012) and intestine (Spence et al., 2011) from iPSCs progresses through intermediate developmental phases when it is exposed to the stepwise application of inductive cues.

De novo assembly of multiple regulatory interactions
The most critical part during de novo shoot regeneration is the positioning of shoot progenitors on the callus surface. Unlike embryonic SAM development, the positional information required for shoot progenitor formation followed by functional SAM development is not predetermined in de novo shoot regeneration. The positional information is absolutely governed by hormonal signaling mediated regulatory interactions during regeneration (Fig. 3). These hormonal signals activate certain key regulators of shoot regeneration. Various studies have shown that the key shoot stem cell regulator WUSCHEL (WUS) is one of the earliest shoot determinants induced in cytokininrich medium (Gordon et al., 2007;Chatfield et al., 2013;Kareem et al., 2015Kareem et al., , 2016.  (Leibfried et al., 2005;Buechel et al., 2009). Thus cytokinin and WUS act in a positive feedback loop. In addition, ectopic overexpression of WUS is sufficient to make de novo shoots from callus on cytokinin-free medium (Kareem et al., 2015). In association with WUS activation on cytokinin-rich SIM, CUC2 gets restricted into a small population of rapidly dividing cells (Gordon et al., 2007). These rapidly dividing cells are the future shoot progenitor cells. Initially, WUS is observed to be expressed in the periphery of CUC2-expressing cells and thus both of these regulatory molecules act in mutually exclusive functional domains.
These non-overlapping functional domains of WUS and CUC2 create a partition in callus for hormonal response and cellular identity. Similar to the CUC2 expression pattern, PLT3, PLT5, and PLT7 genes are also restricted to rapidly dividing shoot forming cells (Kareem et al., 2015).
It has been shown that WUS expression is epigenetically regulated by both DNA methylation and histone modifications during shoot regeneration (Li et al., 2011). It is possible that epigenetic regulators It gets confined to the center of dome shaped shoot progenitors and further leads to the formation of shoot promeristem (Fig. 4B). The CUC2 expression pattern is also changed and it is visualized in the peripheral zone of shoot promeristem in a radial pattern (Gordon et al., 2007).

Organization of functional shoot meristem and completion of shoot formation
Subsequent to the formation of shoot promeristem, polar localization of PIN1 marks the future anlagen of organ initiation. Concomitant with polarized PIN1 upregulation, the key regulator of shoot meristem maintenance SHOOT MERISTEMLESS (STM) appears in the promeristem (Gordon et al., 2007). STM prevents the meristematic cells from differentiating into lateral organs. Simultaneously several other regulators including auxin signaling molecule MONOPTEROS (MP) are upregulated in shoot promeristem (Ckurshumova, Smirnova, Marcos, Zayed, & Berleth, 2014). Also, the stem cells residing at the center of shoot meristem express CLAVATA3 (CLV3) (Gordon et al., 2007). A functional SAM is thus established (Fig. 4C). The subsequent developmental events for the patterning of lateral organs are a recapitulation of in planta shoot development.
Several other factors have also been implicated in shoot regeneration such as ENHANCER OF SHOOT REGENERATION 1 (ESR1) and ESR2 which trigger shoot regeneration in a hormone-independent medium (Banno, Ikeda, Niu, & Chua, 2001;Ikeda, Banno, Niu, Howell, & Chua, 2006). ESR2 activates CUC1 at the transcript level to promote shoot regeneration (Ikeda et al., 2006). The roles of CUC1 and CUC2 to promote shoot regeneration have been demonstrated (Daimon, Takabe, & Tasaka, 2003;Hibara, Takada, & Tasaka, 2003) and CUC genes are activated by PLT3, PLT5, and PLT7 (Kareem et al., 2015). Various regulators of de novo shoot regeneration such as homeobox containing proteins, KNOX proteins, double AP2 containing proteins, and ARFs are conserved across the plant species. Thus the regulatory actions of these genes are likely to be conserved. Despite the growing knowledge about the key regulators, regulatory interactions among several of these are yet to be established to generate the larger regulatory network involved in de novo shoot regeneration.

Whether or not to remember embryonic developmental phases
A beautiful work by Efroni et al. (2016) has very recently shown that, upon severe damage, tissue repair follows an embryo-like sequence to reconstitute stem cells in regenerating root. Similarly, during de novo shoot regeneration, the shoot stem cell niche is reconstituted in the pluripotent callus. A careful follow-up of the temporal sequence of gene expression in live imaging of a regenerating callus demonstrates the progressive assembly of the spatial expression of a battery of genes that eventually culminate in the formation of a confined shoot stem cell niche (Gordon et al., 2007). During very early embryogenesis, key shoot stem cell regulators confine their expression to a few cells at the tip of the apical pole of the embryo (Radoeva & Weijers, 2014). CUC gets localized to the periphery of these cells. In strong wus mutant, the embryonic SAM cannot be traced. But, post-embryonically, shoot meristem displays "on" and "off" activity in wus mutant (Laux, Mayer, Berger, & Jurgens, 1996;Mayer et al., 1998). Mutants like wus and stm make de novo shoots albeit at a lower frequency (Barton & Poethig, 1993;Gordon et al., 2007). Based on existing studies on the spatio-temporal expression pattern of shoot-specific genes and their mutant phenotype during embryogenesis and de novo shoot regeneration, it is difficult to infer that reconstitution of a shoot stem cell niche during de novo shoot stem cell formation follows an embryo-like sequence. These differences could be attributed to a different mode of regeneration, that is, tissue repair upon disruption of multicellularity versus assembly of the entire plant system from a regenerative mass.
The future discovery of additional shoot-specific regulators and the real-time live imaging of the dynamic cellular events during embryogenesis and de novo shoot regeneration might reveal if there is any cellular process during de novo shoot regeneration that follows an embryo-like sequence.

De novo regeneration and somatic embryogenesis: two routes to the same destination
In addition to de novo organogenesis, in vitro regeneration can also be accomplished by somatic embryogenesis (Zimmerman, 1993;Pulianmackal et al., 2014). induces somatic embryogenesis while its low level is associated with de novo shoot organogenesis (Nowak & Gaj, 2016). Since during somatic embryogenesis root and shoot pole is pre-specified, it is not dependent on the assembly of regulatory interactions for positional information. In contrast, reconstitution of stem cells and assembly of spatio-temporally controlled regulatory interactions are central to positional information during de novo organogenesis.

DIRECT REPROGRAMMING
Besides callus mediated regeneration, direct reprogramming of root to shoot and shoot to root occurs in nature as well as under in vitro conditions. Although the root and shoot of plants arise from distinct embryonic regions, they can still regenerate organs different from their own origin, such as root from shoot and vice versa. This can largely be attributed to common evolutionary origins of root and shoot. It is believed that roots of modern day plants evolved from the shoots of primitive plants (Gifford & Foster, 1989;Friedman, Moore, & Purugganan, 2004). Therefore, these two structures are likely to retain the potential to get reprogrammed to the other fate. Not surprisingly, plants such as ferns, many monocots, dicots such as curry leaves, guava, and some tree species possess the natural capacity to generate shoot from root as a part of vegetative propagation (Wittrock, 1884;Holm, 1925). Such reprogramming potential has been widely exploited in horticulture. For instance, a variety of plant parts such as shoot cuttings, root cuttings, and even leaf cuttings have been used for plant propagation. Stem cuttings have been used for propagating common garden plants like chrysanthemum, carnation, and rose (Stangler, 1956) while plants such as raspberry and horseradish can be multiplied by root cuttings (Dore, 1953). Leaves of Bryophyllum, Saintpaulia, and Sansevieria are capable of forming entire plantlets bearing root and shoot (Goethe, 1820;Broertjes, Haccius, & Weidlich, 1968;Henson & Wareing, 1977).  (Ieda et al., 2010;Vierbuchen et al., 2010;Sekiya & Suzuki, 2011).

Root to shoot
An ideal example of direct reprogramming is the induction of shoot from root. Culturing of root explants on cytokinin-rich medium promotes the direct conversion of LRP into shoot (Atta et al., 2009;Chatfield et al., 2013;Kareem et al., 2015Kareem et al., , 2016. Real-time monitoring of dynamic cellular events using live imaging enables us to understand the early molecular events during LRP to shoot conversion ( and shoot-specific genes are expressed (Kareem et al., 2016). Subsequently, there is an elevation in the expression of shoot-specific genes and disappearance of root-specific genes. The interactions of multiple shoot-specific regulatory molecules establish the shoot meristem identity in regenerating structure followed by organ initiation (Chatfield et al., 2013;Kareem et al., 2016). However, this conversion process is highly influenced by a number of factors such as stage of the LRP, ecotype of the plant, cytokinin concentration, and external conditions such as light and temperature (Kareem et al., 2016).

Shoot to root
Similar to the conversion of root to shoot, direct conversion of shoot to root is also possible under in vitro culture conditions. This approach is of significance when not all propagules used in vegetative propa-  . Upon wounding, YUCCA (YUC) gene mediated auxin biosynthesis occurs at the site of injury which is essential for the de novo root regeneration (Chen, Tong et al., 2016). The WUSCHEL-RELATED HOMEOBOX 11 (WOX11) and WOX12 transcription factors respond to this auxin maximum by upregulating LBD16 and LBD29 resulting in a switch from procambium and nearby parenchyma cell fate to root founder cell fate . LBDs may also be involved in the consequent transition to root primordia. So the wound induced adventitious root formation in the absence of external hormone application highlights that the default pathway adopted is of root fate. Wounding induces the NAC pathway that is independent of the auxin maxima mediated cell fate switching (Chen, Cheng et al., 2016). Therefore, similar to wound induced callus formation and subsequent regeneration, the wound response pathway can also initiate direct root organogenesis.
In addition to the external inductive cues, cell fate transition can also occur by the ectopic activation of intrinsic genetic factors. For instance, inducible ectopic overexpression WUS can trigger shoot from root on hormone-free medium (Gallois, Nora, Mizukami, & Sablowski, 2004;Kareem et al., 2015). Similarly ectopic induction of PLT2 can activate root formation from shoot apex (Galinha et al., 2007). Taken together, it can be comprehended that direct shoot and root organogenesis are under the control of extrinsic as well as intrinsic factors.
The molecular dissection of massive cellular reprogramming during cell fate transition can be instrumental in understanding the mechanism underlying transdifferentiation. Furthermore, this knowledge can be used to enhance the direct regeneration capacity of recalcitrant species by adopting a combinatorial approach of standardization.

PERSPECTIVE
Most organisms are born with the gift of regeneration that has been tested time and again, over the course of evolution. However, the extent of this capability varies, perhaps to suit the adversities posed by their natural habitats. While some organisms like axolotl can regenerate a lost body part like the Deadpool superhero, plants are much ahead of the animal kingdom in their regeneration capability.
F I G U R E 5 Larger regulatory network involved in two-step mechanism of shoot regeneration. A detailed description of the regulatory interactions is given in the text. Auxin requires the activity of ALF4 to induce the callus (Sugimoto et al., 2010). WUS activates the expression of TOPLESS (TPL) (Busch et al., 2010) and TPL represses PLT1 and PLT2 (Smith & Long, 2010 during embryogenesis). MP (Ckurshumova et al., 2014) and CUC (Daimon et al., 2003) induce STM expression during regeneration. STM in turn activate CUC during in-planta shoot development (Spinelli, Martin, Viola, Gonzalez & Palatnik, 2011).
Provided that the ideal inductive cues are given, plants have the unique ability to be reborn from a handful of cells or even a single somatic cell (Vasil & Vasil, 1980) which houses the blueprint and the formula necessary to recreate a complete plant. In a very recent study, an externally imposed electric field influences plant regeneration in addition to other external inductive cues like hormones and injury (Kral, Ougolnikova, & Sena, 2016). The underlying molecular basis is yet to be revealed.
Despite the distinctions, it is interesting to note the convergence in the mechanisms underlying regeneration in both plant and animal kingdoms. A common challenge that both the systems face is the maintenance of the spatio-temporal balance between self-renewal and differentiation. While the finer details are different, a common theme emerges that master transcription factors (genetic control) are necessary and sufficient to recreate this process in both plants and animals (Gallois et al., 2004;Takahashi & Yamanaka, 2006;Smith et al., 2013;Kareem et al., 2015). However, epigenetic regulation is critical for finetuning this process and has a major role to play in vivo as well ( (Farthing et al., 2008;Luo et al., 2012). However, the reverse is poorly understood in the plant system compared to animals (Liang & Zhang, 2013). Studies may uncover similar homologous mechanisms in the plant system (Weiste & Droge-Laser, 2014;De Lucas et al., 2016).
This crosstalk between genetic and epigenetic regulators occurs in a bi-directional manner. The signaling pathways acting in concert with transcription factors can control cell fate decisions via epigenetic modifications such as DNA methylation, and histone acetylation and methylation (Anzola et al., 2010;Fagnocchi, Mazzoleni, & Zippo, 2016). The interplay between genetic and epigenetic factors may be responsible for making cells containing the same genetic information differentiate into diverse kinds of cells with varying levels of potency.
However, among the two modes of regulation, epigenetics may be more easily influenced by a mere modification of extrinsic culture conditions (Moussaieff et al., 2015) compared to modulating the genetic regulators.
These studies, however, have only scratched the tip of the iceberg.
There is a long way to go before we acquire a better understanding of the galaxy of genetic and epigenetic players involved in regeneration. The next step then would be to unravel the mechanism by which the crosstalk occurs between these regulators. Although plants have been hailed for their higher regeneration potential, we are yet to nail down all the factors responsible for this cellular plasticity. In addition, although the signature factors of embryonic, root, and shoot development and de novo regeneration have been elucidated, the larger regulatory network involving these players is still being described (Fig. 5).
It is interesting to note that similar or the same set of factors in different contexts can regulate distinct modes of regeneration. Ectopic overexpression of WUS or PLT5 can induce de novo shoot regeneration (Gallois et al., 2004;Kareem et al., 2015) and somatic embryogenesis (Zuo, Niu, Frugis, & Chua, 2002;Tsuwamoto, Yokoi, & Takahata, 2010;Siligato et al., 2016). Similarly, ALF4 plays significant role in callus formation (Sugimoto et al., 2010) as well as wound healing during grafting (Melnyk et al. 2015). Furthermore, a careful analysis at high cellular resolution and an interpretation of regulatory interactions during de novo regeneration are required to identify whether these reflect embryonic or post-embryonic developmental programs in a scenario where organisms utilize the very same set of regulators in two different developmental contexts.
In the long run, the lessons learnt in regeneration from the dynamic array of species can be applied to enhance the regeneration capacity