The vasculature of land plants is critical for growth, as it distributes water and nutrients throughout the plant and provides mechanical support (Ruonala et al., 2017). Each vascular bundle contains three types of tissues: xylem, including water-transporting xylem elements, the mechanically reinforced xylem fibres and parenchyma; phloem, containing the sugar-transporting sieve elements, the companion cells that regulate movement of molecules in and out of the phloem, phloem fibres and parenchyma; and cambium, a layer of undifferentiated cells that produces new xylem cells toward the plant’s main axis and new phloem cells away from it (Nieminen et al., 2015).

During the early stages of vascular development, the position of xylem, phloem and the procambium (precursor of the cambium) is defined by intercellular signals mediated by the auxin and cytokinin hormones, in combination with mobile transcription factors and microRNAs (Carlsbecker et al., 2010; Miyashima et al., 2019). The subsequent production of new vascular cells by the cambium is regulated by the peptide signals CLAVATA3/EMBRYO SURROUNDING REGION (CLE) CLE41/CLE44, which activate the Phloem Intercalated With Xylem (PXY) receptor to regulate the balance between the production of new phloem and xylem cells (Ito et al., 2006; Fisher and Turner, 2007; Hirakawa et al., 2008). The ensuing differentiation of these tissues is guided by cell type-specific transcription factors, such as VASCULAR-RELATED NAC DOMAIN 6 (VND6) and VND7, which specify xylem cell types (Kubo et al., 2005), or Altered Phloem Development (APL) (Bonke et al., 2003) and NAC DOMAIN-CONTAINING PROTEIN 45 (NAC45) (Furuta et al., 2014), which control phloem differentiation.

Among the processes regulated during vascular differentiation, modifications of the cell wall feature prominently, for example in the deposition of lignified hoops that mechanically reinforce xylem elements (Mitsuda et al., 2007; Sugiyama et al., 2017), or in the development of perforated plates that connect adjacent sieve elements (Kalmbach and Helariutta, 2019). Pectic polysaccharides, of which the most abundant form is homogalacturonan, are important components of the plant cell wall that are modified during cell growth and differentiation. Pectins are secreted in a methyl-esterified form and modified by pectin methylesterase (PME) to control the degree of methylesterification. Pectins with low levels of methylesterification bind Ca²⁺ within the cell wall to form a cross-linked gel that surrounds cellulose and is believed to decrease cell wall extensibility (Wolf et al., 2012; Yang et al., 2018). Developmentally controlled pectin methylesterification modulates not only wall extensibility, but also access to further cell wall-modifying enzymes, such as those involved in the deposition of lignin during xylem differentiation (Wolf et al., 2012).

One group of cell wall modifiers that target demethylesterified pectin are the PECTATE LYASE-LIKE (PLL) proteins, which are found throughout the plant kingdom and are encoded by large gene families (Uluisik and Seymour, 2020). PLLs have an N-terminal signal peptide that targets them for secretion into the cell wall, where they cut homogalacturonan at β(1–4) linkages to generate oligogalacturonides (OGs) of different sizes (Sénéchal et al., 2014; Yang et al., 2018). Pectin cleavage occurs during cell expansion or remodelling, and in the middle lamella during cell separation (Yang et al., 2018). Accordingly, PLLs have been implicated in cell expansion processes, such as elongation of cotton fibres (Sun et al., 2020), in secondary wall formation in the xylem, and in cell separation during fruit ripening (Uluisik and Seymour, 2020). Additionally, the OGs produced during pectin cleavage have a signalling role, studied mostly in relation to pathogen attack and as part of the cell wall integrity pathway (Brutus et al., 2010; Ferrari et al., 2013). Externally applied OGs also have developmental consequences, for example by inhibiting the effect of auxin on pea stem elongation (Branca et al., 1988), although a role for OGs as endogenous developmental signals still remains to be established (Ferrari et al., 2013).

Here, we characterise the function of PLL12 in Arabidopsis vascular development. We selected this gene based on ChIP-seq data for the BEL1-like homeodomain transcription factor REPLUMLESS (RPL) (also known as PENNYWISE, PNY and BELLRINGER, BLR) (Byrne et al., 2003; Roeder et al., 2003; Smith and Hake, 2003), which is required for correct vascular patterning (Smith and Hake, 2003; Etchells et al., 2012) and directly interacts with multiple genes that regulate vascular development, such as PXY, CLE41 and NAC45 (Bencivenga et al., 2016). Furthermore, RPL has been implicated in the expression of cell wall enzymes involved in vascular differentiation (Peaucelle et al., 2011; Etchells et al., 2012). We reasoned that less well-characterised RPL target genes, such as PLL12, might reveal novel aspects of vascular development. Our functional analysis showed that PLL12 is expressed in the phloem but has unexpected cell non-autonomous effects on cambium and xylem development.

Genes annotated as belonging to the pectin lyase superfamily are enriched among candidate targets of RPL (Bencivenga et al., 2016; Supplementary Table 1); among these, PLL12 (AT5G04310) showed clear binding to RPL in three ChIP-seq biological replicates, but in none of the negative controls (Supplementary Figure 1). Furthermore, PLL12 expression was located specifically in vascular cell types in a root transcriptomic atlas (Brady et al., 2007) and was associated with vascular differentiation in vitro (Kondo et al., 2016). Based on these data, we hypothesised that PLL12 might have a role in cell wall modification during vascular development, in addition to the previously described function in stomatal guard cells (Chen et al., 2021).

To confirm the PLL12 expression during vascular development, we used the GUS reporter fused to the complete PLL12 gene. Previous reporter genes using GUS fused to 2 kb of the putative PLL12 promoter gave weak and variable expression (Sun and Van Nocker, 2010) or suggested widespread expression (Chen et al., 2021). Our ChIP-seq data, however, indicated RPL binding to intron 3 of PLL12 (Supplementary Figure 1), raising the possibility that regulatory sequences have been missed in previous constructs. To include all potential regulatory sequences, we used a genomic fragment that fully complemented the pll12 mutation (see below), with GUS fused in frame after the last exon of PLL12. In nine independent lines, this pPLL12:PLL12-GUS reporter was expressed in stomata and throughout the vasculature of seedlings (Figure 1). In both the root and in the inflorescence, expression initiated close to the apical meristem, indicating that PLL12 begins to function during early stages of vascular development. Cross-sections of the stem showed that within the vascular strands, PLL12 was expressed primarily in the phloem. This localised expression pattern was in agreement with the reported role of PLL12 in stomatal guard cells (Chen et al., 2021) and with reports of phloem-specific PLL12 expression in the root (Brady et al., 2007; Kalmbach et al., 2022) and during in vitro differentiation (Kondo et al., 2016).

We next analysed the function of PLL12, focussing especially on vascular development. For this, we used two independent T-DNA mutant alleles, pll12-1 (SAIL 1207_A07) (Chen et al., 2021) and pll12-2 (SAIL 1149_C06). As reported, pll12-1 had undetectable PLL12 mRNA (Chen et al., 2021), and the same was seen for pll12-2 (Supplementary Figure 2). Both alleles caused similar phenotypes: the mutants were dwarf and late flowering, with slow stem growth (Figure 2). Complementation with full genomic fragments (from 1,884 bp upstream of the start codon to 702 bp downstream of the stop codon) confirmed that the pll12 mutations caused these phenotypes, and that the selected genomic region contained all regulatory and coding regions required for function (Figure 2). Because the mutants had comparable and fully recessive phenotypes, from this point on we used the previously characterised pll12-1 mutant.

The vascular pattern of pll12-1 appeared normal and was uninterrupted in leaves and in the inflorescence apex (Supplementary Figure 3). The vascular bundles contained all the expected cell types in their correct positions, but appeared smaller, especially in the mature region of the inflorescence stem (Figure 3). To detect growth defects, we measured the size and cell numbers in different regions of the stem vasculature. To compensate for differences in the growth rate between the mutants and the wild type, stems were collected when the fifth flower had matured (i.e., after a longer period of growth in pll12-1). Within stems at this stage, we collected sections at two positions: at the base of the fifth flower, corresponding to the immature region of the stem, which is still elongating (less than 1 cm of the apex, see Figure 2), and near the first silique, where stem elongation is complete and the vascular bundles have started to deposit secondary cell walls (more than 2 cm from the apex, Figure 2). In pll12-1, the size of immature vascular strands was not different from the wild type (number of cells per cross-section, mean ± SD: 101.4 ± 19.9 in wt, 98.7 ± 15.9 in pll12-1, Supplementary Table 2). However, as the stem matured, differences became clearer (213.9 ± 27.7 in wt, 142.4 ± 33.5 in pll12-1, Supplementary Table 2) and, contrary to the expectation based on the expression pattern, were not specific to the phloem. Instead, all three regions of the vascular strand (phloem, cambium and xylem) had reduced growth, measured either by cross-sectional area or cell numbers (Figures 3D,F–H). The region most affected was the cambium, which already had significantly fewer cells in the mutant before differences were detectable in the phloem and xylem (Figures 3G,H). Furthermore, the cambium was the only region with no detectable growth (measured as either cross-sectional area or by number of cells in cross-section), in contrast to the phloem and xylem, which grew at about half the wild-type rate. These results showed that PLL12 is not essential for the initial vascular patterning or differentiation but is required for subsequent growth of all regions of the vascular strand, particularly the cambium.

To reveal physiological consequences of the histological defects described above, we investigated whether pll12-1 affected phloem and xylem function. Using carboxyfluorescein diacetate (CFDA) as a tracer, reduced phloem transport from hypocotyls to roots was shown in pll12 seedlings (Kalmbach et al., 2022). Although we were not able to adapt this method to measure phloem transport in the inflorescence stem, it would be reasonable to assume that a similar defect would occur in the stem vasculature. As an indirect method to detect defects in phloem function in the stem, we assayed for high starch levels after a dark period, which are typically seen in plants with compromised sugar transport (Barratt et al., 2010; Bezrutczyk et al., 2018). Consistent with a defect in sugar export, staining with Lugol’s Iodine reagent revealed a strong accumulation of starch in the cortex of the inflorescence stem in pll12-1 compared to the wild type (Figure 4). To assess xylem function, we measured transport of rhodamine B dye from roots to the inflorescence apex. Both visually (Figures 4E,F) and based on the dye concentration in the apex (normalised to protein concentration, Figure 4G), the mutant showed a clear reduction in transport through the xylem. We conclude that the reduced growth of vascular bundles in pll12 was accompanied by a decrease in both sugar and water transport, which might be the cause of the general inhibition of growth in the mutant. Similarly, slow growth of the cambium and xylem in the mutant might be an indirect consequence of a reduced sugar supply, caused by a primary defect in the phloem. Alternatively, PLL12 might be required to produce a phloem-derived signal that promotes development of the cambium and xylem.

To test whether increased or ectopic PLL12 expression might be sufficient to promote cambium and xylem development, we generated plants in which PLL12 transcription could be induced ubiquitously. To do this, we used an pOp:PLL12 construct expressed in an RPS5A:LhGR background; when tissues are exposed to dexamethasone, this system induces PLL12 expression under the widely expressed RIBOSOMAL SUBUNIT 5A promoter (Supplementary Figure 4). After inflorescence tips were treated five times at 2-day intervals, dexamethasone-treated plants showed variable inhibition of growth, ranging from short internodes to death of the shoot apex (Figures 5A–C). These phenotypes were associated with different levels of lignin deposition in the cortex and vascular bundles within 1–2 mm below the shoot meristem (Figures 5D–F). In apices with a mild phenotype, it was possible to trace the earliest deposition of ectopic lignin to strands of cells leading basally away from the shoot apical meristem—this position is typically occupied by procambial strands in the wild type (Supplementary Figure 5), suggesting that cells at early stages of vascular differentiation were particularly sensitive to ectopic PLL12 expression.

Ectopic lignification is one of the responses elicited by cell wall fragments during pathogen attack (Ferrari et al., 2013), so lignification could have been a defence response activated by pectin fragments produced when PLL12 was overexpressed. However, closer examination of the dexamethasone-treated apices revealed cortex cells with lignification in a pitted pattern and openings between adjacent cells (Figure 5I), similar to root metaxylem and to ectopic metaxylem cells induced by the regulators of xylem differentiation, VND6 and SND5 (Figures 5G–I; Kubo et al., 2005; Zhong et al., 2021). Pitted cells were also seen within the vascular bundles of plants with a mild phenotype after PLL12 induction, but not in uninduced controls (Supplementary Figure 5). Further supporting the idea that ectopic PLL12 expression re-directed cortex cells to vascular development, ectopic expression of the regulator of cambium development, PXY, was detected by in situ hybridisation after PLL12 induction (Figure 5N). Conversely, the normal PXY expression in the cambium of the wild type was disrupted in the disorganised vascular bundles seen after dexamethasone treatment (Figures 5J–O). Together with the inhibition of cambium and xylem growth in the pll12-1 mutant (Figure 3), the results above supported a role for PLL12 in the regulation of cambium and xylem development, although the experiments could not distinguish whether PLL12 promoted cambium and xylem identity sequentially, or separately in different cells.

The stronger effects of PLL12 overexpression close to the meristem and in cortex cells suggested that the response to a putative PLL12-produced signal depended on developmental stage and cell type. To study the effect of ectopic PLL12 in a different context, we also looked at seedlings. RPS5A:LhGR pOp:PLL12 plants were grown on medium without dexamethasone until the first pair of leaves emerged, then the seedlings were moved on to dexamethasone or mock treatment plates. Over the subsequent 7 days, a gradient of phenotypes became evident, ranging from small rosettes to stunted seedlings with severely deformed leaves (Figure 6), although root growth was initially unaffected (Supplementary Figure 6). In the most extreme cases, leaf blades failed to expand and became chlorotic (Figures 6D,E). Confocal imaging of these leaves stained with mPS-PI showed that the veins had irregular thickness, were discontinuous and showed cambium-like cell divisions along the leaf vasculature (Figures 6F–L). Thus, although ectopic PLL12 only appeared to induce extensive lignification in the inflorescence apex, it interfered with the differentiation of vascular cells both in the inflorescence and in seedlings.

PLL12 has been shown to function in stomatal guard cells, where it affects cell wall properties and the speed of stomatal opening and closure (Chen et al., 2021). Here, we show that PLL12 is also required for the development and function of the stem vasculature. Loss of PLL12 function did not have obvious effects on patterning or cell differentiation, but inhibited growth of vascular bundles, with the earliest effects visible in the fascicular cambium, followed by reduced cell numbers in both phloem and xylem. The reduced cell numbers throughout the stem vascular bundles were accompanied by reduced transport of both water and sugar. The simplest interpretation of these phenotypes is that PLL12 function is limiting for the activity of the fascicular cambium, and consequently for the enlargement of vascular bundles during stem growth.

Despite the reduced growth across the vasculature (Figures 3G,H and Supplementary Table 2), PLL12 expression was specific to the phloem, in line with previous reports (Brady et al., 2007; Kalmbach et al., 2022). Thus, the effects on cambium and xylem growth were likely indirect consequences of a primary role of PLL12 in the phloem. In support of a phloem function for PLL12, a recent pre-print reports that pll12-1 has a defect in long-distance phloem transport, attributed to subtle changes in the formation of sieve plates between phloem elements (Kalmbach et al., 2022). However, defects in long-distance sugar transport would not readily account for the disproportionate effect of pll12-1 on the growth of established vascular bundles, in comparison to earlier stages. Moreover, a role exclusively in sieve plate formation would not easily explain why PLL12 overexpression induced ectopic cambium and xylem features and disrupted vascular differentiation in seedlings. To explain both loss- and gain-of-function phenotypes, and taking into account the phloem-specific expression, we propose that in addition to the reported role in sieve element development (Kalmbach et al., 2022), PLL12 activity influences an intercellular signal that coordinates development across vascular tissues.

PLL12 contains a canonical pectate lyase C domain, with conserved residues implicated in catalysis, Ca ^{+ +} binding and exocytosis (Chen et al., 2021), so this protein is likely to cleave demethylesterified pectin within the cell wall. Accordingly, immunolocalisation showed increased levels of calcium-crosslinked pectin in pll12-1 stomatal walls (Chen et al., 2021) and mutation of the predicted pectate lyase catalytic site abolished the ability of PLL12 to complement the mutant (Kalmbach et al., 2022). Furthermore, PLL12 is homologous to a Zinnia PLL protein for which pectate lyase activity has been demonstrated in vitro (Domingo et al., 1998). Pectin cleavage would be expected to release oligogalacturonides (OGs), which bind to the WAK1 receptor and activate the cell wall integrity and pathogen defence signalling pathways, whose downstream responses include lignification (Caño-Delgado et al., 2003; Brutus et al., 2010; Ferrari et al., 2013). However, the induction of metaxylem-like pitted cell walls, combined with ectopic PXY expression, suggested that lignification after ectopic PLL12 expression was an aspect of vascular differentiation, rather than a generic defence or cell wall stress response.

Thus, the expected biochemical function of PLL12, together with its non-cell-autonomous effects and its ability to ectopically induce cambium and xylem identity, raise the possibility that a cell wall-derived signal participates in the regulation of vascular differentiation. The most straightforward candidate signal would be OGs, although their developmental roles remain speculative (Seifert and Blaukopf, 2010). Supporting a link to xylem differentiation, genes that respond early to OG treatment (Moscatiello et al., 2006) included SND5, which is part of the transcriptional network controlled by PXY (Smit et al., 2020) and encodes a NAC domain protein that regulates secondary wall deposition in the xylem (Zhong et al., 2021). However, indirect effects on other signalling pathways are also possible. External application of OGs has been shown to interfere with auxin signalling (Branca et al., 1988; Ferrari et al., 2013). However, the phenotypes caused by ectopic PLL12 activation were different from those seen in plants with an overall inhibition of auxin signalling; for example, there was little effect on roots, where auxin plays a central role (Israeli et al., 2020). The interrupted veins seen after ectopic PLL12 induction in cotyledons and leaves are also reminiscent of phenotypes seen after application or overexpression of the PXY ligand CLE41/44 (Hirakawa et al., 2008), or in mutants that affect phosphoinositide-regulated vesicle traffic, which may regulate auxin transport (Sieburth et al., 2006; Carland and Nelson, 2009). In the future, it will be interesting to explore how PLL12 function interacts with these signalling pathways.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

RS: conceptualisation, software, funding acquisition, and supervision. MB and VS: investigation. MB and RS: writing—original draft. All authors: formal analysis, data curation, and writing—review and editing.

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.