Border cell release: Cell separation without cell wall degradation?

ABSTRACT

Plant border cells are specialized cells derived from the root cap with roles in the biomechanics of root growth and in forming a barrier against pathogens. The mechanism of highly localized cell separation which is essential for their release to the environment is little understood. Here I present in situ analysis of Brachypodium distachyon, a model organism for grasses which possess type II primary cell walls poor in pectin content. Results suggest similarity in spatial dynamics of pectic homogalacturonan during dicot and monocot border cell release. Integration of observations from different species leads to the hypothesis that this process most likely does not involve degradation of cell wall material but rather uses unique cell wall structural and compositional means enabling both the rigidity of the root cap as well as detachability of given cells on its surface.

Abbreviations

CW

Calcofluor White

EGCG

(–)-epigallocatechin gallate

HG

homogalacturonan

MS

Murashige-Skoog medium

PME

pectin methylesterase

Cell separation occurs throughout the life of a plant as part of genetically and physiologically highly regulated developmental processes which include, for instance: organ abscission, dehiscence or separation of pollen microspores.¹ Cell walls form the interface between cells and therefore play an essential role in both cell adhesion and separation. The latter is usually coupled to cell wall degradation that is accomplished by various carbohydrate acting enzymes (CAZymes) especially polysaccharide hydrolases or lyases.² In many cases this is well supported by genetic evidence; for instance mutants in genes encoding polygalacturonases exhibit phenotypes in floral organ abscission³ or microspore separation.⁴ However, among all the cell separation processes, the border cell release represents an interesting biologic paradox. The root cap has to be rigid to allow for the penetration through the substrate and to protect the meristem and on the other hand it continuously grows and its surface is a site for a recurring cell separation.^5,6 In our recent publication,⁷ we showed that the pea root cap cell walls become structurally different and enriched in various cell wall epitopes once they reach the surface of the root cap. Both in situ analyses and tests on cell-wall-component-degrading enzyme activities suggested that no extensive cell wall degradation is involved in this process.⁷

Brachypodium distachyon is already a well-established model organism for grasses.⁸ It is generally accepted that grasses produce primary cell walls distinct from dicots.^9-11 One of the striking differences between type I and type II cell walls is the content of pectins, notably homogalacturonan (HG), the principal determinant of cell adhesion.¹² The question is how this difference would relate to the mechanism of border cell separation. Here I present analyses which were performed on the 2-days-old Brachypodium seedlings (Bd21–3) grown on vertical half MS plates in a similar manner as Arabidopsis (Fig. 1A). Brachypodium border cells are more homogeneous than those from pea and are less curved. The measurements determined their length to be 24.5–58.4 μm (average 38.0 μm) and their diameter was between 9.9–16.6 μm (average 13.2 μm). Brachypodium whole seedlings can be treated by their transfer to liquid MS medium in 12-well micro plate supplemented with biologically active chemicals. The Fig. 1 (B,C) presents the outcome of such an experiment using the pectin methylesterase (PME)-inhibitor (–)-epigallocatechin gallate¹³ (EGCG) which, in similar manner to pea,⁷ inhibited the border cell release to the media. Sectioning of resin embedded and toluidine stained Brachypodium root apex shows that the Brachypodium root cap is a triangular structure which does not fully cover the root meristem (Fig. 2A). Like in pea, the cell walls become gradually thicker as they approach the surface (Fig. 2B,C). Distinction of 2 parts of extracellular material from the detaching last layer of cells could be observed: one part is diffused between cells and one remains as an intact cell wall of released cells. In situ analysis of 2 major cell wall components: HG and β-(1,4)-glucans (represented mainly by cellulose) using the oligosaccharide probe COS⁴⁸⁸ and Calcofluor White (CW), respectively, are presented in Fig. 2 (D–F). COS⁴⁸⁸ is specific for HG with a low degree of esterification¹⁴ – a substrate for pectolytic enzymes. However, fluorescence scans and the signal plot profile (Fig. 2G) indicated that there is no loss of either Calcofluor White or COS⁴⁸⁸ signal during the event of detachment. Quite the opposite, the general upregulation of COS⁴⁸⁸ binding sites could be observed. COS⁴⁸⁸ also stained the dissolving cell wall material present in intercellular spaces between detaching cells in a form of a diffuse signal. This shows the dynamics of HG within a monocot root cap is similar to that of pea and could be a part of some common mechanism of separation regardless of cell wall types.

Figure 1.

Brachypodium border cells. (A) Two-days-old Brachypodium seedling germinated and grown on vertical half MS plate. (B) Close-up to the root cap of Lugol-stained root tip with released single border cells. BC, border cell. (C,D) The effect of the PME inhibitor (–)-epigallocatechin gallate (EGCG) on the release of border cells. Brachypodium seedlings were incubated in 1 ml liquid half MS in a 12-well plate (C) without and (D) with 100 μM EGCG for 24 hours and analyzed with Lugol staining and chloral hydrate clearing. Note the border cells in EGCG treated sample are formed but not fully released and are trapped around the root tip. Scale bars: 0.5 mm in (A) and 30 μm in (B-D).

Figure 2.

In situ cell wall analysis on resin sections of Brachypodium root cap. (A-C) LR White resin sections (1 μm thick) of Brachypodium root tip stained with toluidine blue. (A) Longitudinal section through the root tip with indicated root cap: RC, root meristem: RM and border cell: BC. (B) Cross-section through the root cap. Note the thickness of the cell wall increases toward the surface. (C) Close-up to the region marked by dashed square in B. The arrow depicts the diffuse intercellular material and the arrowhead the intact cell wall of a detaching cell. (D-F) Staining of the region of the root cap with a detaching layer of cells. (D) Staining with β-(1,4)-glucan-specific dye Calcofluor White (CW) and (E) COS⁴⁸⁸ probe specific for HG with low degree of esterification. (F) Overlay of the 2 channels with indicated line of signal plot profile in G and 3 regions of interest. Note the COS⁴⁸⁸ staining of the intercellular material (arrows). (G) The signal plot profile of CW and COS⁴⁸⁸ signal intensities over the dashed line in F. The numbers indicate the cell wall regions marked in F. Note there is no loss of the total amount of the signals when the cells separate (compare the total area of the peaks marked 1 and 2). Scale bars: 30 μm.

Compelling observations from different models^7,14 form the base for my hypothesis about possible mechanism of detachment. I propose that the elevated PME activity on the secreted HG act as a calcium chelating agent disassembling the already formed egg boxes (gluing complexes of HG with calcium) within the middle lamella. In other words, the altered ratio between the available extracellular calcium and de-methylesterified HG might be an important factor in disassembly of the middle lamella. The secretion of pectins with high levels of site decorations (especially arabinan) can further facilitate inhibition of the de novo formation of egg boxes at cell interface.⁷ Such a mechanism could ensure that the border cells will possess intact cell walls required for their post-release function, as well as the integrity and rigidity of the root cap. Separation by means of degrading enzymes cannot be hitherto completely ruled out. In theory, this would require their precise expression and activity to be meticulously localized. These enzymes would need to be active exclusively at the cell-to-cell interface during detachment otherwise the degradation of the whole root cap is at stake. It is important to note that it has been reported that pectolytic activity in the pea root cap correlates with developmental stage at which the border cells are being released.¹⁵ However, I speculate that this measured activity could be actually attributed to other cell activities like elongation or cell division.

All species analyzed in our laboratory so far have some advantages and disadvantages as models for border cells. Arabidopsis is the most established plant genetic model; however, as a member of Brassicaceae family it produces a special type of border-like cells. Moreover, cell death occurs during the root cap developmental cycle.¹⁶ Pea provides enough border cells to accomplish -omics analyses but it is less amenable to in vivo microscopy and genetic experiments due to the lack of the genome sequence. Increasing availability of cell wall mutants, more genetic and molecular biology resources and moderate size might make Brachypodium a feasible future model for border cell wall physiology. Brachypodium produces relatively homogenous populations of border cells and although their isolation for biochemical analyses would still be challenging, it does not seem to be impossible. In conclusion, border cell release represents an interesting but relatively neglected cell wall-related phenomenon with many exciting open questions.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Acknowledgments

I would like to thank European Commission Marie Skłodowska-Curie Actions for providing me the fellowship (project CeWalDyn, agreement number 329830) and Dr. Stjepan Krešimir Kračun for a helpful discussion and critical comments on the manuscript.