Systematic approaches to C-lignin engineering in Medicago truncatula

Abstract

Background: C-lignin is a homopolymer of caffeyl alcohol present in the seed coats of a variety of plant species including vanilla orchid, various cacti, and the ornamental plant Cleome hassleriana. Because of its unique chemical and physical properties, there is considerable interest in engineering C-lignin into the cell walls of bioenergy crops as a high-value co-product of bioprocessing. We have used information from a transcriptomic analysis of developing C. hassleriana seed coats to suggest strategies for engineering C-lignin in a heterologous system, using hairy roots of the model legume Medicago truncatula.

Results: We systematically tested strategies for C-lignin engineering using a combination of gene overexpression and RNAi-mediated knockdown in the caffeic acid/5-hydroxy coniferaldehyde 3/5-O-methyltransferase (comt) mutant background, monitoring the outcomes by analysis of lignin composition and profiling of monolignol pathway metabolites. In all cases, C-lignin accumulation required strong down-regulation of caffeoyl CoA 3-O-methyltransferase (CCoAOMT) paired with loss of function of COMT. Overexpression of the Selaginella moellendorffii ferulate 5-hydroxylase (SmF5H) gene in comt mutant hairy roots resulted in lines that unexpectedly accumulated high levels of S-lignin.

Conclusion: C-Lignin accumulation of up to 15% of total lignin in lines with the greatest reduction in CCoAOMT expression required the strong down-regulation of both COMT and CCoAOMT, but did not require expression of a heterologous laccase, cinnamyl alcohol dehydrogenase (CAD) or cinnamoyl CoA reductase (CCR) with preference for 3,4-dihydroxy-substituted substrates in M. truncatula hairy roots. Cell wall fractionation studies suggested that the engineered C-units are not present in a heteropolymer with the bulk of the G-lignin.

Keywords: C-lignin; Co-product; Hairy roots; Medicago truncatula; Metabolic engineering; Transgenic plants.

Fig. 1

The monolignol biosynthesis pathway, showing all known available routes to H, C, G and S monolignols in dicots. C-lignin is made exclusively of caffeyl alcohol (C) units, biosynthesis of which requires prevention of the O-methylation of the 3-OH group by CCoAOMT and/or COMT. The reactions which are blocked during the phase of C-lignin biosynthesis in the Cleome seed coat are shown with a sign. The 3-OH group in caffeyl alcohol can be introduced by the direct action of C3H on coumarate, or by the combined activities of HCT, C3′H and CSE in the so-called “shikimate shunt”. S lignin is not made in the Cleome seed coat due to lack of expression of F5H. Enzymes are L-phenylalanine ammonia-lyase (PAL), cinnamic acid 4-hydroxylase (C4H), hydroxycinnamoyl CoA: shikimate hydroxycinnamoyl transferase (HCT), coumaroyl shikimate 3′-hydroxylase (C3′H), caffeoyl shikimate esterase (CSE), coumarate 3-hydroxylase (C3H), caffeic acid/5-hydroxyconiferaldehyde 3/5-O-methyltransferase (COMT), 4-hydroxycinnamate CoA ligase (4CL), caffeoyl CoA 3-O-methyltransferase (CCoAOMT), cinnamoyl CoA reductase (CCR), cinnamyl alcohol dehydrogenase (CAD), ferulic acid/coniferaldehyde 5-hydroxylase (F5H), laccase (LAC)

Fig. 2

Lignin content and composition in shoots of 10-day-old seedlings of wild-type and OMT mutant M. truncatula lines. A Total lignin thioacidolysis yield. B Overall % lignin monomer compositions. C % of H-units. D % of C-units. E % of G-units. F % of S-units. Data are shown as the mean ± sd (for n = 3 biologically independent samples); the different letters above the bars represent statistically significant differences determined by one-way analysis of variance (ANOVA; least significant difference (LSD), P < 0.0001). Stem, stem tissue from R108 wild-type background

Fig. 3

Monolignol pathway metabolite levels in shoots of 10-day-old seedlings of wild-type and OMT mutant M. truncatula lines. The figure shows metabolite levels superimposed on a scheme of the monolignol pathways. Metabolite levels were determined by LC–MS/MS transitions pre-determined for individual compounds and quantified by comparison to authentic standards. Structures are shown in Fig. 1. Data are means ± SD derived from three biological replicates. stem, stem tissue from R108 wild-type background

Fig. 4

Engineering lignin composition in M. truncatula hairy routes by transformation with MtCOMT-MtCCoAOMT RNAi and ChLAC8 overexpression constructs in the comt mutant background. A Lignin biosynthesis pathway designed for accumulation of C-lignin. The reactions which are blocked during the phase of C-lignin biosynthesis in the Cleome seed coat are shown with a sign. B–E Monolignol compositions of transgenic lines determined by thioacidolysis. Data show % of H-lignin monomer (B), C-lignin monomer (C), G-lignin monomer (D) and S-lignin monomer (E). cGUS, GUS control in comt mutant background; RGUS, GUS control in R108 wild-type background. Stem, mature R108 stem. −/−, no transgenic event; −/+, transgenic roots only harboring ChLAC8 overexpression construct; ±, transgenic roots only harboring MtCOMT-MtCCoAOMT RNAi construct; +/+, transgenic roots harboring both MtCOMT-MtCCoAOMT RNAi and ChLAC8 overexpression constructs. CCL, RNAi construct for MtCOMT (C) and MtCCoAOMT (C) and overexpression construct for ChLAC8 (L). Data are means ± SD derived from three biological replicates

Fig. 5

Monolignol pathway metabolite levels in M. truncatula MtCOMT-MtCCoAOMT RNAi and ChLAC8 overexpression hairy roots. The engineered lines, in the comt mutant background, are a selection of those shown in Fig. 4. The Figure shows metabolite levels superimposed on a scheme of the monolignol pathways. Metabolite levels were determined by LC–MS/MS transitions pre-determined for individual compounds and quantified by comparison to authentic standards. Structures are shown in Fig. 1. CCL, RNAi construct for MtCOMT (C) and MtCCoAOMT (C) and overexpression construct for ChLAC8 (L). c_GUS, GUS control in comt mutant background; R_GUS, GUS control in R108 wild-type background; stem, mature R108 stem (controls, right-hand 4 bars in all panels). −/−, no transgenic event; −/+, transgenic roots only harboring ChLAC8 overexpression construct; ±, transgenic roots only harboring MtCOMT-MtCCoAOMT RNAi construct; +/+, transgenic roots harboring both MtCOMT-MtCCoAOMT RNAi and ChLAC8 overexpression constructs. Data are means ± SD derived from three biological replicates

Fig. 6

Engineering lignin composition in M. truncatula hairy routes by transformation for SmF5H, MtCOMT-MtCCoAOMT RNAi and ChLAC8-ChLAC15 expression in the comt mutant background. A Lignin biosynthesis pathway designed for accumulation of C-lignin. B–E Monolignol compositions of transgenic lines determined by thioacidolysis. Data show % of H-lignin monomer (B), C-lignin monomer (C), G-lignin monomer (D) and S-lignin monomer (E). c_GUS, GUS control in comt mutant background; R_GUS, GUS control in R108 wild-type background. Stem, mature R108 stem. −/−/−, no transgenic event; ±/−, transgenic roots only harboring SmF5H overexpression construct; -/±, transgenic roots only harboring MtCOMT-MtCCoAOMT RNAi construct; −/−/+, transgenic roots harboring only ChLAC8-ChLAC15 construct; −/+/+, transgenic roots harboring both MtCOMT-MtCCoAOMT RNAi and ChLAC8-ChLAC15 overexpression constructs.. FCCLL, overexpression construct for SmF5H (F), RNAi construct for MtCOMT-MtCCoAOMT RNAi (CC) and overexpression construct for overexpression for ChLAC8-ChLAC15 (LL). Data are means ± SD derived from three biological replicates

Fig. 7

Engineering altered lignin composition in M. truncatula hairy roots by down-regulation of COMT and HCT, and overexpression of Selaginella moellendorffii F5H (SmF5H) and Cleome hassleriana CAD5 (ChCAD5). A Scheme showing the predicted pathway. B–G Monolignol content and composition of selected lines. (B) Total lignin thioacidolysis yields; (C) overall % lignin monomer compositions; (D) % of H-lignin monomers; (E) % of C-lignin monomers; (F) % of G-lignin monomers; (G) % of S-lignin monomers. Hairy root lines designated OR(h) are engineered lines for SmF5H/ChCAD5-OX and HCT/COMT RNAi in the comt mutant background. Subsequent genotyping led to selection of plants with no transgenes (−/−), RNAi but not OX transgene (−/+), OX but no RNAi transgene (±), and the presence of both transgenes (+/+). Wild-type R108 and comt mutant, both transformed with GUS, served as additional negative controls, and wild-type stems were included for comparison. Transgene constructs are shown in Additional file 1: Fig. S2, and transcript levels of the targeted genes in each line shown in Additional file 1: Fig. S8. FAHC, transformed with overexpression construct for SmF5H (F) and ChCAD5 (A) and RNAi construct for MtHCT (H) and MtCOMT (C). Data are means ± SD derived from three biological replicates