Plant Polygalacturonases Involved in Cell Elongation and Separation—The Same but Different?

Plant cells are surrounded by the primary cell wall, a rigid framework that needs to be modified in order to allow cell growth. Recent data suggest that in addition to the cellulose-hemicellulose network, the pectin matrix plays a critical role in determining the elasticity of the primary cell wall. Polygalacturonases are key homogalacturonan-hydrolyzing enzymes that function in a wide range of developmental processes. In this review, we present recent progress in understanding the role of polygalacturonases during cell elongation and separation. In discussing the specificities and possible redundancies of polygalacturonases, we focus particularly on newly discovered Arabidopsis mutants that have measurable loss-of-function phenotypes. However, data from other species are included when necessary.


Introduction
In plants, cells are surrounded by a rigid cell wall and are, therefore, fixed in their relative position. As a consequence, the overall shape of the plant body is created by a tightly controlled interplay of cell division and anisotropic cell expansion. The driving force for cell expansion is turgor pressure and it requires controlled relaxation of the cell wall while maintaining cellular integrity [1]. The high osmotic pressure inside the cell, ranging from 0.3 to 1.2 MPa in the case of Arabidopsis thaliana, makes this a

Polygalacturonases
Polygalacturonases belong to the glycosyl hydrolase family 28 and are key HG hydrolyzing enzymes that have been implicated with a wide range of plant developmental processes such as cell elongation, organ abscission, fruit ripening, microspore release, and pollen tube growth [18]. Plant PG genes belong to large gene families and their expansion and diversification can be attributed to whole genome and segmental duplications in association with gene loss, as well as intron gain and (more predominantly) intron loss events [19,20]. Phylogenetic analysis of PG gene structure in Arabidopsis, rice, and other plant species reveals five distinct clades of PG genes which can be further divided into subclades indicating the occurrence of at least four ancestral PG genes before the divergence of monocots and dicots. While sequences from different clades are relatively divergent, they are rather conserved within a clade and tandem-duplicated genes generally fall into the same subclade [19][20][21][22]. Depending on their mode of action, endoand exo-polygalacturonases can be distinguished [23]. Endo-PGs hydrolyze the HG polymer at random sites but require at least four consecutive GalA residues of the HG chain to be de-methylesterified [24,25]. Thus, the methylation pattern of the HG chains directly influences possible endo-PG-mediated HG cleavage. Endo-PG activity might lead to complete hydrolysis of pectin polymers and has therefore the potential to cause rapid cell elongation or even cell separation [12,26].
Exo-PGs on the other hand attack the free ends of de-methylesterified HG polymers and thereby reduce the overall polymer length. It has been speculated that the resulting modification of the pectin matrix might be subtler than the random cleavage by endo-PGs and might, therefore, be used to fine-tune the extensibility of the primary cell wall [27].

Polygalacturonases Involved in Fruit Ripening and Cell Separation
Early on, polygalacturonases were isolated from ripening fruits which implied a role in pectin degradation for tissue softening [28]. In a pioneer work in tomato, down-regulation of polygalacturonase expression by an anti-sense construct lead to decreased de-polymerization of solubilized pectins and increased storage-life of ripe fruits although there was no measureable effect on fruit softening [29,30]. Silencing of PG expression in apple and strawberry on the other hand increased the firmness of the ripe fruit significantly but did not change other ripening parameters. It could be shown that in these transgenic plants, ionically and covalently bound pectin exhibits a lower degree of de-polymerization. Microscopic analysis of transgenic fruits revealed smaller intra-cellular spaces and more cellular adhesion [31][32][33]. This is in agreement with earlier reports demonstrating that pectin degradation by PGs also plays a central role in cell separation in abscission events and dehiscence zones [34][35][36]. In apples, constitutive expression of the fruit-specific MdPG1 gene resulted in a range of novel developmental phenotypes including premature leaf shedding due to reduced cell adhesion in abscission zones, malformed leaves and malfunctioning stomata. As a consequence of the strong constitutive expression of MdPG1 in transgenic apple trees, a decrease of the average molecular weight of pectin chains could be demonstrated [37]. In rice, over-expression of OsBURP16, the non-catalytic PG1β-subunit of the polygalacturonase PG1, reduced cell adhesion in leaves and pectin content [38]. Together these data support the involvement of PGs in cell separation events in plants.
Many polygalacturonase genes are highly expressed in reproductive tissue [21], and for some, their involvement in cell separation events could be demonstrated by loss-of-function phenotypes: The endo-PG QUARTET3 (QRT3) functions in degrading the pollen mother cell wall during microsporogenesis and thus enables the release of unicellular microspores [39]. ARABIDOPSIS DEHISCENCE ZONE POLYGALACTURONASE1 (ADPG1), ADPG2, and QRT2 were reported to act in a redundant manner in anther dehiscence while ADPG2 and QRT2 function partially redundant in floral organ abscission [40].

Polygalacturonases Involved in Cell Expansion
The previously mentioned examples of polygalacturonases involved in fruit ripening and cell separation might lead to the conclusion that the primary role of endo-PGs is the more or less complete breakdown of the pectin matrix in terminal developmental situations like tissue softening, abscission, or dehiscence. This is clearly not the whole story, since, recently, the endo-PG POLYGALACTURONASE INVOLVED IN EXPANSION1 (PGX1) was shown to be involved in hypocotyl elongation and floral patterning [41]. Furthermore, in vivo assays for pectin-degrading enzymes suggested that during cotyledon expansion in cotton endo-PG and exo-PG activity could both be detected. Interestingly, their appearance differed temporally during cell elongation with high endo-PG activity at an early phase, followed by an increase of exo-PG activity during a later phase when endo-PG activity decreased [26]. This implies a scenario where different pectin hydrolyzing enzymes might act on the same substrate but in a consecutive manner.
The pectinase LeXPG1 was isolated from tomato seed protein extracts and gene expression was detected in the embryonic root, the developing vasculature as well as the embryo-surrounding endosperm. Based on this expression pattern, it was suggested that LeXPG1 might play a role in cell elongation as well as tissue softening in the embryo and the endosperm, respectively. Interestingly, LeXPG1 displays calcium-dependent exo-PG activity [42]. Furthermore, exo-polygalacturonase activity had been observed during abscission events in citrus explants before [43]. Endo-PGs and exo-PGs seem to be both involved in cell elongation as well as cell separation events. Therefore, their enzymatic activity does not directly correlate with one or the other process but rather a combination of both activities seems to be necessary in both events. Hence, it is tempting to speculate if cell elongation and cell separation in principal rely on identical pectin modifications.
The Arabidopsis genome contains 69 PG genes [35,44]. Several of these arose from tandem duplications and belong to the same phylogenetic subclade [21]. Intuitively, one would speculate that these tandem-duplicated genes might act in a redundant manner if the expression pattern of these genes did not diverge. Only for a handful of PG genes is there direct evidence for their function during development, based on loss-of-function phenotypes. For many other Arabidopsis PG genes, the lack of an obvious loss-of-function phenotype might indeed be a result of genetic redundancy [18].
Recent findings might shed some further light on this situation: Cell elongation defects were reported in embryos of the Arabidopsis nimna (nma) mutant [45]. NMA codes for a putative exo-polygalacturonase and is preferentially expressed in reproductive tissue. Cells of nma mutant embryos fail to elongate as early as the zygote stage and severe cell elongation defects can be further observed in the suspensor while cells of the embryo-proper seem to recover from their defects at later stages of embryo development (Figure 1) [45]. This might indicate that other polygalacturonases can take over NMA function in the embryo-proper. Peptides from five PGs including NMA were found to be present in cell wall fractions of 5 day-old etiolated hypocotyls of Arabidopsis [46]. While NMA is obviously present in elongating hypocotyls, the nma mutation does not seem to have any measurable effect on hypocotyl length [45]. It appears that other PGs are able to compensate for the loss of NMA activity in this case. For the embryonic suspensor, the situation seems to be different: While there are several closely related PG genes expressed in the suspensor of globular stage embryos according to published microarray data ( Figure 2) [47], the strong cell-elongation defects observed in nma mutants indicate that none of these can fulfill NMA function [45]. The reason for this might be different temporal expression, sub-cellular localization, enzyme activity, or substrate specificity.
A similar situation was observed for the closely related ADPG1, ADPG2, and QRT2 genes ( Figure 2) [40]. Loss of all three genes causes an impaired pod shatter phenotype and compromises anther dehiscence. While the pADPG1::ADPG1 transgene was able to fully complement the pod shatter defects of the triple mutant, QRT2 and the closely related PG gene At1g48100 failed to do so when expressed under the ADPG1 promoter [40]. Again, this would argue for a distinct function of these proteins in the cell separation process possibly caused by different enzymatic activity or substrate preference.
Expression analysis of three closely related tomato PG genes (TAPGl, TAPG2, and TAPG4) indicates temporal regulation during leaf and flower abscission. The temporal expression pattern of these genes suggests that they might act consecutively to fulfill a stepwise modification of the pectin matrix [48]. Phylogeny was created with protein sequences of two sub-clades A1a and A14 [21] using neighbour-joining with bootstrap values of 100 [49]. Some of the PGs mentioned in this review (Medicago sativa PG3, Solanum lycopersicum XPG1 and a bacterial exo-PG from Yersinia enterocolitica) were also included as outgroups. Scale bar represents amino acid substitutions per site; (B) Mean expression values of selected Arabidopsis polygalacturonase genes closely related to NMA in the embryo and suspensor based on publicly available microarray data [47].

Perspectives
Pectin plays a central role in determining the physicochemical properties of the primary cell wall. Modifications of the pectin matrix are thus elementary for cell elongation by determining the extensibility of the cell wall. The degree of methyl-esterification is one important aspect that seems to be tightly regulated, but recent data emphasizes the importance of pectin-hydrolyzing polygalacturonases in cell elongation processes [41,45]. The temporally regulated activity of endoas well as exo-PGs seems essential for both cell elongation and cell separation processes. To what degree the vast number of PG genes in the Arabidopsis genome reflects genetic redundancy or displays the need for a high number of specific enzymes in these cell wall processes is still unclear. With recently described Arabidopsis mutants like nma or pgx1, which show obvious and quantifiable loss-of-function phenotypes, this question can now be addressed. Promoter-swap and complementation experiments are powerful tools to support the biochemical analysis of these proteins. Complementation experiments with endoand exo-PGs will allow a better understanding to what degree these enzymes are functionally redundant or are involved in separate non-exchangeable steps in pectin modification during cell elongation processes.
The active site in PG proteins is well conserved but substrate recognition motifs are not well understood [23]. In vivo complementation assays along with biochemical studies guided by protein-structure data might help unraveling specific modes of substrate recognition.
Technical advances have greatly helped our understanding of the primary cell wall composition [5,[50][51][52]. Studying the effect of well-characterized PGs might indirectly give further insight in cell wall composition and the structure and modifications of the pectin matrix.
Furthermore, with the recent advances in genome-editing tools, like the CRISPR/Cas9 system and their application in plant biology, the study of many tandem-duplicated PG genes is now technically possible [53,54].
Understanding the substrate-specificity and the nature of the pectin modification carried out by specific polygalacturonases will not only increase our understanding of plant cell wall biology during cell elongation but will also be valuable for their use in commercial products and technical processes like biofuel production.