Reconstitutive approach for investigating plant vascular development

Genetic experiments have uncovered various key regulators controlling vascular development, such as plant hormones, peptides, receptors, kinases, and transcription factors (reviewed in De Rybel et al. 2016). Vascular development begins during the embryonic stage in Arabidopsis. Vascular cells in roots originate from four vascular initial cells, which are established during the globular stage of embryogenesis (De Rybel et al. 2013; Ohashi-Ito et al. 2013) (Fig. 1a). These vascular initial cells, which are considered as precursors of procambium, undergo several rounds of cell division to ensure a sufficient population of vascular cells. During this stage, the bHLH transcription factor TARGET OF MONOPTEROS 5 (TMO5) and its homologs (TMO5-LIKE 1 and TMO5-LIKE 2) regulate cell division, along with LONESOME HIGHWAY (LHW), under the control of auxin signaling (DyRybel et al. 2014; Ohashi-Ito et al. 2014). As the cell population increases, the xylem axis and two phloem poles gradually become established (Fig. 1b). A mutual regulatory circuit between auxin and cytokinin signaling contributes to the patterning and maintenance of the xylem and phloem domain in roots (Bishopp et al. 2011a, b). In addition, intricate cell–cell communication mediated by mobile miR165/166 and their targets, HD-ZIPIII transcription factor genes, helps determine the identity of metaxylem and protoxylem vessels in the xylem axis (Carlsbecker et al. 2010) (Fig. 1b). Once the vascular cell arrangement is complete, procambial cells acquire strong proliferative activity for secondary radial growth. In particular, proliferating procambial cells that are positioned on a ring-shaped domain are termed cambial cells (Fig. 1c). This transition typically occurs in the upper parts of roots or hypocotyls.

Open image in new window

During this phase of secondary growth, the vascular meristem, which consists of procambial/cambial cells, gives rise to secondary xylem and secondary phloem cells (Fig. 1c). These multipotent cambial cells are considered as vascular stem cells, since they also have a self-renewal activity. The peptide hormone TDIF and its receptor TDR (also known as PXY) play crucial roles in maintaining the vascular meristem (Fisher and Turner 2007; Hirakawa et al. 2008, 2010). Genetic studies have revealed that the TDIF signaling pathway is divided into two distinct cascades, the WUSCHEL-RELATED HOMEOBOX 4 (WOX4) pathway and the GLYCOGEN SYNTHASE KINASE 3 (GSK3s)-BRI1-EMS-SUPPRESSOR 1 (BES1) pathway; these pathways regulate cell proliferation and cell differentiation in cambial cells, respectively (Kondo et al. 2014; Kondo and Fukuda 2015). In addition, the peptide hormones EPIDERMAL PATTERNING FACTORs (EPFs) (Uchida and Tasaka 2013) and signaling by the plant hormone ethylene (Etchells et al. 2012) also help maintain the vascular meristem, together with the TDIF signaling cascades (Kondo and Fukuda 2015). Cytokinin is yet another key regulator of cambial cell proliferation (Matsumoto-Kitano et al. 2008; Nieminen et al. 2008) (Fig. 1c). During the later process of secondary growth, the xylem area dramatically expands, which is accompanied by xylem fiber differentiation. Mobile gibberellin can trigger xylem fiber differentiation in hypocotyls upon the flowering transition (Ragni et al. 2011). Conversely, signaling mediated by the receptors ERECTA (ER) and ER-LIKE 1 (ERL1) negatively regulates xylem fiber differentiation, which helps prevent fiber formation at an immature stage (Ikematsu et al. 2017).

As described above, vascular development is a complex hierarchical process that is regulated by multiple players (Fig. 1), making it difficult to obtain a comprehensive understanding of vascular development. Therefore, we should carefully discuss these regulatory networks by focusing on each developmental process separately.

To simply understand the process of vascular development, many culture systems have been established in numerous laboratories. For instance, callus and suspension cell cultures from various vascular plants can be induced to ectopically form tracheary elements (TEs) at high frequency (reviewed in Devillard and Walter 2014; Iakimova and Woltering 2017). In 1980, a xylogenic culture system was established using isolated mesophyll cells from Zinnia elegans (Fukuda and Komamine 1980). These cultures possess better uniformity in terms of original cell identity compared to callus cultures, making it possible to analyze TE differentiation in a relatively synchronous manner. However, as molecular genetic techniques have become more popular and powerful, researchers have preferred to use the model plant Arabidopsis to study vascular development. Oda et al. (2005) and Kubo et al. (2005) established in vitro xylogenic culture systems using Arabidopsis suspension cells treated with auxin and brassinosteroid. Furthermore, Pesquet et al. (2010) reported that habituated Arabidopsis suspension cells can stably differentiate into TEs in the presence of auxin, cytokinin, and brassinosteroid. Genetic manipulation via conditional induction of VASCULAR-RELATED NAC DOMAIN 6 (VND6) or VND7, which are master regulators of xylem differentiation, led to the establishment of highly efficient xylem culture systems using Arabidopsis suspension cells (Oda et al. 2010; Yamaguchi et al. 2010). The use of these suspension cultures has led to the identification of key components regulating cell wall modifications associated with xylem differentiation, especially from the viewpoint of cell biology. Confocal microscopic analysis revealed that an active ROP domain formed by the competition between GEF and GAP determines the shape and position of wall pits (Oda and Fukuda 2012). Moreover, proteomics analysis using suspension cell cultures led to the identification of many microtubule-associated proteins that control secondary cell wall formation (Derbyshire et al. 2015). Despite their advantages for imaging and proteomics studies, molecular genetic analyses using these culture systems are quite time-consuming, as genetic transformation and subsequent cell culture steps are required, precluding their use for rapid investigations using molecular genetic approaches.

Attempts have been made to develop tissue culture systems that can easily be applied to transgenic Arabidopsis plants. The application of auxin, cytokinin, and brassinosteroid triggers ectopic xylem differentiation in hypocotyl explants (Kubo et al. 2005; Sawa et al. 2005). Recently, the application of bikinin, a specific inhibitor of GSK3s, was found to lead to excess xylem differentiation in the presence of auxin and cytokinin, making it possible to establish a novel culture system for vascular cell differentiation named VISUAL (Kondo et al. 2014, 2015) (Fig. 2a). This effect is not unexpected, since GSK3s play negative roles in xylem differentiation downstream of TDIF-TDR signaling (Kondo et al. 2014). During the differentiation process in VISUAL, mesophyll cells quickly lose their original identity and acquire (pro)cambial cell identity, along with the upregulation of ARABIDOPSIS THALIANA HOMEOBOX GENE-8 (AtHB8) and TDR (Fig. 2b). Consistently, cell cycle- and cell division-related genes are strongly induced at this stage (Kondo et al. 2015). Subsequently, VISUAL-induced (pro)cambial cells differentiate into xylem cells (Fig. 2b). During this process, the TDIF-TDR-GSK3s-BES1 signaling cascade negatively controls xylem differentiation from procambial cells via a process similar to that during in vivo development (Kondo et al. 2014, 2015). Furthermore, genes that are upregulated during the xylem differentiation stage strongly overlap with the inducible gene sets found in VND6-expressing suspension cells (Ohashi-Ito et al. 2010), suggesting that VISUAL mimics vascular development. Another advantage of VISUAL is that the many Arabidopsis mutant and marker lines can easily be used to investigate the molecular functions of genes of interest involved in vascular development.

Open image in new window

In plants, cambial cells give rise to xylem cells as well as phloem cells. Unlike xylem cells, fully differentiated phloem cells do not exhibit marked morphological changes. Recently, Furuta et al. (2014) described the various subcellular events associated with phloem sieve element (SE) differentiation in detail based on observations from Serial Block-Face TEM (SBF-TEM). Phloem SEs contain phloem-specific organelles and subcellular structures, including P-protein bodies, the sieve element reticulum, and sieve plates. Furthermore, callose deposition and enucleation occur during the SE differentiation process (reviewed in Blob et al. 2017; Heo et al. 2017). Nonetheless, it is still difficult to distinguish phloem SEs without proper phloem markers. Therefore, the molecular mechanisms underlying phloem differentiation have remained a mystery. Like xylogenic cultures, several reports describe cultures in which ectopic phloem differentiation is induced from callus (Aloni 1980). Interestingly, VISUAL can stimulate phloem cell differentiation with the appearance of ectopic phloem markers in Arabidopsis cotyledons (Kondo et al. 2016) (Fig. 2b).

VISUAL involves the normal steps of SE development, including callose deposition, enucleation, and the formation of the SE reticulum (Kondo et al. 2016). However, sieve plates are not found in VISUAL-induced phloem cells. Cell-sorting of phloem cells produced via VISUAL using the fluorescent marker SIEVE ELEMENT OCCLUSION-RELATED 1 (SEOR1) revealed their similarity to root SEs (Brady et al. 2007) in terms of gene expression profiles (Fig. 3). This finding, together with their morphological similarity, suggests that root SEs can almost be equated with VISUAL phloem cells. Combining three different transcriptome data sets using VISUAL, I generated a coexpression network for phloem differentiation in which phloem-related genes could be divided into two major gene clusters: early phloem genes and late phloem genes (Kondo et al. 2016) (Fig. 3). One of early phloem genes encodes an leucine-rich repeat receptor BARELY ANY MERISTEM 3 (BAM3), which regulates root protophloem development by perceiving the small peptide CLAVATA3/ESR-RELATED 45 (CLE45) (Depuydt et al. 2013; Hazak et al. 2017). Moreover, the expression profiles of genes encoding OCTOPUS (OPS) and BREVIS RADIX (BRX), which are polarly localized at the plasma membranes of protophloem precursor cells (Truernit et al. 2012; Depuydt et al. 2013), are similar to those of early phloem genes (Kondo et al. 2016). By contrast, CALLOSE SYNTHASE 7 (CALS7) (Barratt et al. 2011) and NAC45/86-DEPENDENT EXONUCLEASE-DOMAIN PROTEINS (NENs) (Furuta et al. 2014), which are involved in callose deposition and enucleation, are categorized as late phloem genes. ALTERED PHLOEM DEVELOPMENT (APL), encoding a master regulator of phloem differentiation (Bonke et al. 2003), is also classified in the late phloem cluster, suggesting that network analysis can distinguish cell specification-related genes from differentiation-related genes (Kondo et al. 2016). Further analysis of combinations of various transcriptome datasets would enable us to dissect the sequential differentiation process in more detail.

Open image in new window

Several additional studies of phloem differentiation using VISUAL have recently been reported. SUPPRESSOR OF MAX2-LIKE 4 (SMXL4) and SMXL5 are expressed in root phloem tissues in planta (Wallner et al. 2017; Wu et al. 2017). Wallner et al. subjected the smxl4 smxl5 double mutant to the VISUAL system to investigate the contribution of SMXLs to phloem differentiation, finding that SMXLs redundantly function in phloem differentiation but not in xylem differentiation (Wallner et al. 2017) (Fig. 3). In addition, Breda et al. (2017) utilized VISUAL for proteomics analysis to identify the phosphorylation site of the phloem-specific protein OPS (Fig. 3), since roots do not contain enough vascular cells for biochemical analysis. Thus, VISUAL is useful not only for investigating vascular development without taking other developmental processes into account, but also for amplifying cells during the differentiation process that can be used for omics analyses.

Since VISUAL can be used to induce xylem cells as well as phloem cells, the VISUAL process appears to correspond to secondary vascular development, in which vascular cambial cells give rise to xylem or phloem cells (Fig. 2b). Although the xylem and phloem cell differentiation processes are gradually being uncovered, the mechanism underlying the multipotency of vascular stem cells remains to be elucidated. Interestingly, bes1 mutants have defects in xylem differentiation in the VISUAL system (Kondo et al. 2014, 2015). Further genetic investigations exploring the bes1 mutants during phloem differentiation via VISUAL will help address this issue. Furthermore, elucidating the reprogramming process from mesophyll cells into (pro)cambial cells represents a major challenge for understanding plant totipotency. Bikinin treatment induces the formation of (pro)cambial cells via an unknown mechanism (Kondo et al. 2015); uncovering this mechanism may provide clues about the process underlying entry into the vascular cell lineage.

In summary, VISUAL is a reconstitutive approach to in vitro plant vascular cell production that can simplify investigations of complex vascular developmental processes. The availability of this tool raises the possibility of investigating unsolved issues concerning the regulation of developmental processes, including the multipotency of vascular stem cells and stem cell establishment. In the future, novel concepts proposed using such approaches should be investigated in vivo, which would greatly increase our understanding of vascular development in plants.