Cooperative lignification of xylem tracheary elements

Abstract

The development of xylem tracheary elements (TEs) – the hydro-mineral sap conducting cells - has been an evolutionary breakthrough to enable long distance nutrition and upright growth of vascular land plants. To allow sap conduction, TEs form hollow laterally reinforced cylinders by combining programmed cell death and secondary cell wall formation. To ensure their structural resistance for sap conduction, TE cell walls are reinforced with the phenolic polymer lignin, which is deposited after TE cell death by the cooperative supply of monomers and other substrates from the surrounding living cells.

Xylem is the vascular tissue in land plants which transports water and minerals from the roots to the leaves.¹ To do so, specialized cells named tracheary elements (TEs) form conductive cellular structures which assemble end-to-end and laterally to establish a complex vascular network throughout the plant body. To fulfill their conductive role, TEs undergo 3 main morphological modifications including (i) programmed cell death (PCD) to empty the cell content, (ii) lateral secondary cell wall formation to strengthen the cell sides and (iii) thinning/perforation of the cell terminal end to provide access to the cell lumen.² Altogether these processes lead to the formation of a hollow cell, terminally perforated and with a reinforced lateral cell wall.³ This lateral reinforcement corresponds to the deposition of a secondary cell wall which is composed of cellulose (40–50%), hemicellulose (25–30%) and the phenolic polymer lignin (20–30%). Addition of lignin to TE cell walls provides an increased mechanical resistance (extra 25–75 MPa in tensile strength and 2.5–3.7 GPa in Young's modulus) and impermeability.^4,5 The formation of lignin results from the oxidative polymerization of at least 3 different 4-hydroxyphenylpropene monomers - p-coumaryl alcohol, coniferyl alcohol and sinapyl alcohol - which form the 3 main lignin subunits: H (hydroxyl), G (guaiacyl) and S (syringyl), respectively.⁶ Oxidizing enzymes, such as H₂O₂-dependent peroxidases and O₂-dependent laccases, generate monomer radicals which directly polymerize into lignin in the cell wall.⁶ Genetic or pharmacological disturbances of TE formation in plants result in dramatic defects including collapsed TEs, reduced growth, greater sensitivity to environmental stress and pathogen infections.⁶ To study TE formation without hindering plant development, simplified cell culture systems have been established such as the Zinnia elegans TE differentiating cell cultures. In this system, TE differentiation can be triggered hormonally from freshly isolated mesophyll cells: the differentiation is synchronous, morphological changes progress chronologically within 4 days, with 30–40% of the cells being fully differentiated into dead lignified TEs.⁷ The resulting in vitro TEs present all the morphological, genetic and biochemical characteristics associated to TEs in whole plants,^7,8 making the Zinnia elegans system an ideal tool to study TE formation.

Lignin deposition is controlled spatio-temporally

Lignin is differentially distributed between the different parts of the cell wall, which can be easily visualized by UV-autofluorescence coupled with confocal microscopy (Fig. 1A). Arabidopsis thaliana hypocotyl cross-sections exhibit the highest level of lignin autofluorescence in the cell corners (CC) and the middle lamella (CML) of the primary cell walls of the different xylem cell types (Fig. 1A and C). The secondary cell walls (SCW) generally display less lignin autofluorescence, and it seems that there is a stronger autofluorescence in the SCWs of the vessels than those of fibers (Fig. 1A and C). In isolated TEs of the Zinnia elegans cell culture system, lignin autofluorescence is restricted to the SCW and absent from the primary cell wall that is visible between the lateral SCW thickenings (Fig. 1B). The differential localization and abundance of lignin between the apoplast/primary cell wall and the SCW is therefore dependent on the cell type and suggests a tight spatial regulation of lignin deposition in plants. One of the mechanisms suggested to control the spatial deposition of lignin in TE secondary thickenings is the specific localization of oxidizing enzymes. The H₂O₂-dependent peroxidase ZPO-C (AB023959⁹) is co-regulated with SCW polysaccharide synthesis genes and localized to TE secondary cell walls.⁹ Similarly, O₂-dependent laccases 4 and 17 are also specifically localized in the secondary cell walls of Arabidopsis TEs.¹⁰ Lignification is therefore defined to specific areas of the cell wall at the time when the polysaccharide SCWs are synthesized but before deposition of the bulk lignin. During TE formation in Zinnia cell cultures, the timing of lignin formation is regulated as its deposition has been shown to occur after the deposition of the SCW polysaccharides.^11,12,13,14 Lignin synthesis in Zinnia TE SCWs directly depends on the proper synthesis of secondary polysaccharides as pharmacological disturbance of cellulose synthesis leads to dispersed and lower lignification of TEs.^15,16 In contrast, treatments with inhibitors of lignin synthesis do not affect the spatial organization of TE SCWs.^13,17,18 Thus a chronological sequence occurs during TE SCW formation, starting with polysaccharide synthesis and ending with lignin deposition. This sequence of cell wall polymer deposition is confirmed when disassembling TE cell walls as degradation of Zinnia TE secondary cellulose and xylan is enhanced when first removing lignin.^19,20

Figure 1.

Lignin distribution in xylem cell walls of Arabidopsis thaliana hypocotyls and in isolated TEs.¹⁷ (A) UV confocal microscopy of transverse sections of the hypocotyl; lignification in the secondary xylem is visualized by artificial color 8-bits intensity scale (from 0 to 256). vSCW, vessel secondary cell wall; fSCW, fiber secondary cell wall; CML, compound middle lamella; CC, cell corner. Bar = 30 µm. (B) UV confocal microscopy of a TE from differentiating cell cultures; lignification is visualized by artificial color 8-bits intensity scale. SCW, secondary cell wall; PCW, primary cell wall. Bar = 8 µm. (C) Quantification of the lignin UV autofluorescence intensity of hypocotyl transverse sections of Arabidopsis thaliana ecotypes Columbia-0 (Col-0) and Landsberg erecta (Ler). Error bars indicate ± SD.

TEs are lignified post-mortem

TE differentiation is completed by cell death and secondary cell wall lignification. In differentiating TE cell cultures, cell death is triggered once xylan and cellulose have been deposited in the secondary cell wall.^14,21,22 This process includes an influx of calcium (Ca²⁺)²³ and a change of the tonoplast permeability,²⁴ which leads to the inflation of the vacuole^23,24 and finally bursting of the tonoplast and release of the vacuolar hydrolytic contents (proteases and nucleases) into the cytoplasm.^3,23,25 Once cell death is accomplished, the gradual autolysis of the protoplast remnants by the released enzymes occurs rapidly; it takes less than 10 minutes to completely remove the nucleus and several hours for the chloroplasts.^21,25,26 Remarkably, lignification of TEs occurs mainly after cell death^17,27 and continues for several hours.^13,18,26 Accordingly, blocking cell death in TE cell cultures by silver thiosulfate also blocks lignification,¹⁷ while inhibiting lignification with piperonylic acid (PA) does not affect TE cell death.¹⁷ In order to demonstrate that TE corpses are able to fully lignify post-mortem, TEs were inhibited to lignify using PA until all TEs died. These TE corpses were then supplied with 60 µM coniferyl alcohol (G-OH) and/or sinapyl alcohol (S-OH). TE lignification was estimated at the single cell level using FT-IR microspectroscopy¹⁷ which showed that dead TEs were able to lignify only when supplied with external monomers. Multivariate analysis confirmed that the changes observed when adding external lignin monomers to PA-treated dead TEs were due to aromatic compounds built into the cell wall, based on bands at 1505 cm⁻¹ and 1595 cm⁻¹ (aromatic -C=C- vibrations, characteristic of lignin, Fig. 2A and B).²⁸ Hierarchical clustering of average FT-IR spectra revealed that dead PA-treated TEs achieved normal post-mortem lignification when supplied with a mixture of both lignin monomers (Fig. 2C). Moreover, estimation of the 1505 cm⁻¹ and 1595 cm⁻¹ band area (integrals) confirmed full post-mortem lignification of dead PA-treated TEs (Fig. 2D-E). To evaluate if post-mortem TE lignification also occurs in intact plants, FT-IR microspectroscopy was performed on proto- and metaxylem TEs at different internodes of 5-weeks old Zinnia plants and similarly confirmed an increase in lignin with increasing age of the TEs.¹⁷ TE post-mortem lignification was also shown to occur in Arabidopsis root TEs²⁹ suggesting that TE post-mortem lignification is a general event in angiosperms. However, the mechanisms controlling the triggering of lignin formation after TE cell death are still unknown.

Figure 2.

Post-mortem lignification of Zinnia elegans TEs. (A) Principle component analysis (PCA) of FT-IR spectra of differently treated TEs from Zinnia elegans cell cultures. Samples include TEs without 50 µM piperonylic acid (PA) (white squares), TEs with 50 µM PA and 60 µM coniferyl alcohol (G-OH) (gray), TEs with 50 µM PA (black). (B) Correlation scaled loadings of the first principal component, showing lignin characteristic bands (1505 cm⁻¹ and 1595 cm⁻¹), accumulating in PA untreated samples as well as in PA treated samples with G-OH compared to PA treated samples only. (C) Hierarchical clustering of average FT-IR spectra of Zinnia non-TEs with or without 50 µM PA, TEs with 50 µM PA and 60 µM G-OH or sinapyl alcohol (S-OH), TEs with or without 50 µM PA and TEs with PA and 60 µM G-OH and 60 µM S-OH. (D) Average peak area of the 1505 cm⁻¹ FT-IR band (aromatic –C=C– vibration associated to G-type lignin in au, arbitrary units) of non-TE and TE samples mentioned in (C). (E) Average peak area of the 1595 cm⁻¹ FT-IR band (aromatic –C=C– vibration associated to S-type lignin in au, arbitrary units) of non-TE and TE samples mentioned in (C). The asterisks indicate statistically significant difference from TEs treated with PA by t-test with Welch correction (*P < 0.05, **P < 0.01, and ***P < 0.001). Error bars indicate ± SD.

Lignification of TEs is non-cell autonomous

TE post-mortem lignification implies that the substrates necessary for lignin formation (monomers and/or H₂O₂) are either released in the extracellular medium when TEs die and/or secreted by the surrounding living parenchyma cells. In Zinnia TE differentiating cell cultures, about 30–40% of the cells become TEs and die while the rest of the cells remain alive.⁷ Although these parenchyma cells do not exhibit distinct morphological features, they are differentiated cells which express specific genes that are also expressed in xylem parenchyma of whole plants.³⁰ During TE differentiation in Zinnia cell cultures, the gene expression of the lignin monomer synthesis genes PAL (phenylalanine ammonia-lyase), C4H (cinnamate-4-hydroxylase), CCR (cinnamoyl-CoA reductase) and CAD (cinnamyl-alcohol dehydrogenase) continues long after the TEs have committed cell death.^12,17,31,32 Similarly, the enzymatic activities of C4H and CAD can also be detected several days after TEs have died.^31,32 The extended gene expression and protein activity beyond the TE lifespan suggests that the remaining parenchyma cells are involved in the non-cell autonomous supply of lignin monomers. Although no significant increase of intra- and extracellular total phenolics is visible during TE formation in Zinnia cell cultures,^12,18 the extracellular medium accumulates in TE differentiating conditions known lignin monomers such as coniferyl and sinapyl alcohols beyond the TE lifespan.^13,17,18,27 The fact that dead, non-lignified TEs were able to lignify after washing away the lignin-biosynthesis inhibitor PA demonstrates that the living parenchyma cells present enable the post-mortem lignification of TEs by directly exporting lignin monomers into the extracellular medium.¹⁷ TE non-cell autonomous lignification also concerns the production of H₂O₂ (necessary for peroxidase activity) which is produced by both differentiated parenchyma cells in Zinnia cell cultures and xylem parenchyma in Zinnia plants.^33,34,35 Interestingly, pharmacological inhibition of H₂O₂ production in differentiated parenchyma cells effectively reduced post-mortem lignification of TEs in cell cultures, suggesting that the living parenchyma cells provide other substrates, such as H₂O₂, also for lignin polymerization.¹⁷

Genetic evidence for the TE non-cell autonomous post-mortem lignification was presented by the identification of Arabidopsis genes that were specifically expressed in xylem parenchyma and that affected xylem lignification in a reverse genetic analysis.¹⁷ These included the lignin monomer biosynthesis gene C4H (AT2G30490), the RADICAL-INDUCED CELL DEATH 1 RCD1 (AT1G32230) and the transcription factor MYB13 (AT1G06180).¹⁷ Cell specific expression was analyzed using the promoter driven β-glucuronidase (GUS) reporter system in 7-d old seedlings. C4H was expressed along the entire root except for the root apical meristem as well as in the vascular system of the leaves (Fig. 3A). RCD1 was highly expressed in the root apical meristem and in the youngest leaf primordia and to a lower extent in the entire root and in the vasculature of the leaves (Fig. 3B). MYB13 was expressed in the hypocotyl, the shoot apical meristem and to a lower extent in the vasculature of the leaves and the root apical meristem (Fig. 3C). GUS analysis in stem and hypocotyl cross sections confirmed the expression of C4H, RCD1 and MYB13 in the xylem parenchyma cells surrounding TEs (Fig. 3A-C). Hypocotyl lignin analysis of the corresponding loss-of-function mutants using pyrolysis-GC/MS showed significant changes in both lignin quantity and composition. The c4h-3 and myb13 mutants exhibited a significant reduction in lignin,¹⁷ essentially due to a decrease in lignin G units (Fig. 3D) whereas the rcd1-1 mutant showed an increase in lignin¹⁷ due to a higher amount of G and H units while S units were not affected (Fig. 3D). Taken together, these analyses elucidate the action of the monomer biosynthetic machinery, including C4H, in cooperative lignin biosynthesis and reveal novel proteins, such as RCD1 and MYB13, in this process as well.

Figure 3.

Xylem parenchyma expression of C4H, RCD1 and MYB13 and mutational impact on lignin quantity and composition. Histochemical β-glucuronidase GUS reporter gene expression in 7d old Arabidopsis thaliana seedlings (representative picture of n = 12 independent seedlings), stem sections (upper image) and hypocotyl sections (lower image) of C4H (A), RCD1 (B) and MYB13 (C); TE, tracheary element; arrows indicate expression in xylem parenchyma; bars (seedling) = 200 µm; bars (stem/hypocotyl) = 50 µm. (D) Pyrolysis-GC/MS analysis of 2-month old Arabidopsis hypocotyls of Columbia-0 (Col-0) and Landsberg erecta (Ler) wild-type plants and c4h-3, myb13, rcd1-1 and rcd1-2 mutant plants (n = 3 replicated experiments). All mutants are in Col-0 background except for myb13, which is in Ler background. Pyrolysis-GC/MS profiles of the different mutants and wild-type controls (Col-0 and Ler) were compared using heat-map hierarchical clusterization of 54 cell wall-related pyrolysis product peaks according to.³⁶ Heatmap color scale indicates fold changes in specific peak accumulation compared to wild-type (WT).

Conclusion

The differential distribution of lignin in the cell wall and the apoplast of specific cell types is a tightly controlled process. Among the lignifying cells in plants, TEs undergo post-mortem secondary cell wall lignification enabled by a non-cell autonomous supply of lignin monomers and H₂O₂, provided by the surrounding living parenchyma cells. The quantity and composition of lignin in TEs depend on a tight coordination and cooperation between TEs and the surrounding parenchyma cells, perhaps to enable optimal sap conduction as the plant grows.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

We thank Junko Takahashi-Schmidt and the plant cell wall and carbohydrate analytical facility at UPSC, supported by Bio4Energy and TC4F project, for the pyrolysis-GC/MS analysis.

Funding

This research was supported by a Vetenskapsrådet (VR) research grant 2010-4620 (to E.P.) and the Gunnar Öquist Fellowship from the Kempe Foundation (to E.P.).