Plant species have long been known to contain three major monolignols, p-coumaryl alcohol, coniferyl alcohol, and sinapyl alcohol , and more recently 5-hydroxyconiferyl alcohol has been identified as an additional monolignol that can be incorporated into cell walls, particularly in COMT-deficient poplar (Populus tremula x alba) . However, 5-hydroxyconiferyl alcohol remains a relatively minor component of natural lignin, and these results indicated that monolignols other than the three major monolignols can be incorporated into lignin. Although 5-hydroxyconiferyl alcohol was not affected in the present study, its 4-O- and 5-O-glucosides were greatly accumulated (76-fold and 60-fold, respectively) in COMT down-regulated switchgrass lines. Here, we also demonstrate that the down-regulation of COMT loci in switchgrass additionally resulted in the accumulation of a novel monolignol-like metabolite, iso-sinapyl alcohol, its glucoside, iso-syringin, and related metabolites iso-sinapic acid and iso-sinapyl aldehyde. The presence of iso-sinapyl alcohol in non-pretreated samples of transgenic plants negates the potential argument that it is an artifact of the mild, hot water pretreatment as a breakdown product of benzodioxane (β-O-5, α-O-5) substructures, which can form from cross-coupling reactions involving radicals of 5-hydroxyconiferyl alcohol. If the latter were the case, iso-sinapyl alcohol should be evident in both wild-type and transgenic plants, given that both plant types have at least small quantities of 5-hydroxyconiferyl alcohol. iso-Sinapyl alcohol has previously been isolated from the roots of Ferula sinaica, the leaves of Croton xalapensis L. (Euphorbiaceae) , and the feces of the beetle Naupactus bipes. However, it has not been associated with lignin metabolism, it has not been synthesized previously, its biosynthesis in plants has not been investigated, and it has not been previously identified in any of the many previous COMT knockdown studies thus far. It is ironic that blocking of a major methylation step that generates a dimethoxycinnamyl alcohol leads to the appearance of an alternative, novel dimethoxycinnamyl alcohol. A simple explanation for the biosynthesis of iso-sinapyl alcohol would be that the reduced activity of COMT allows another methyltransferase to methylate the para (4-hydroxyl) position on the aromatic ring of the accumulating sinapyl alcohol precursors. Although this hypothesis lacks experimental support with respect to the presence of such a protein(s) being over-expressed, the accumulations of 5-hydroxyferulic acid, 5-hydroxyconiferaldehyde, and 5-hydroxyconiferyl alcohol glucosides provide metabolite evidence that the global COMT knockdown employed in this study resulted in the accumulation of metabolites that can all be methylated at the para position to produce all of the putative iso-sinapyl alcohol related precursors observed in this study. Given that native COMT specifically methylates the meta (3-hydroxyl and 5-hydroxyl) positions on the phenyl ring of aromatic acids/aldehydes and is precluded from substitution at the para position, it is possible that an alternative para-specific methyltransferase, similar to iso-eugenol 4-O-methyltransferase (IEMT; EC 22.214.171.124), described by , is able to para methylate the accumulating substrates, including the 5-hydroxyferulic acid, 5-hydroxyconiferaldehyde and/or 5-hydroxyconiferyl alcohol, or there may be an alternate pathway that emerges that generates iso-sinapic acid and results in the reductive formation of iso-sinapyl alcohol. It is generally accepted that COMT acts on 5-hydroxyconiferaldehyde/alcohol as substrates, which explains the reduced level of S-residues in mutants and transgenics with reduced COMT activity. Given that 4-coumarate-CoA ligase EC 126.96.36.199 (4CL) does not display much activity toward sinapic acid in grasses, there would be limited flux of sinapic acid towards sinapyl alcohol in grasses. Thus, iso-sinapyl alcohol synthesis from iso-sinapic acid was not anticipated, suggesting the aforementioned para-methylation of multiple substrates may be the more likely mode of synthesis of the iso-sinapyl alcohol related metabolites. Although not readily detected in wild-type plants, it cannot be unequivocally stated that this direct synthesis pathway of iso-sinapic acid to iso-sinapyl alcohol doesn’t exist in such plants at very low flux and metabolite concentrations. We have detected iso-sinapic acid (0.04 μg/ml) in another lignin pathway enzyme (ferulate-5-hydroxylase EC 1.14.-.-; F5H) knockdown line in switchgrass, but iso-sinapyl alcohol was not detected in that line. Furthermore, 4-O-methylation of monolignol precursors has been postulated in another monocot species, Vanilla planifolia, although no enzymatic basis for this conclusion has yet emerged . Additional tracer studies are needed to clarify metabolite flux and the primary pathway leading to the production of iso-sinapyl alcohol.
iso-Sinapyl alcohol is a monolignol based on its molecular structure, but a key question remains as to whether it is incorporated into the plant cell wall. Quantum chemical calculations demonstrate a reduced number of conjugation sites for iso-sinapyl alcohol compared to sinapyl alcohol. The most likely homodimeric lignans formed from single dehydration reactions were predicted to be coupled at C6–C2′ (based on atom numbering of sinapyl alcohol), C2–C2′, C6–C6′, and C6–O5′, and these were confirmed by organic synthesis, but none was detected in plant samples. However, two lignan-like metabolites were detected only in COMT down-regulated plants that may be iso-sinapyl alcohol-based lignans, but they remain unidentified. A 5-hydroxconiferyl alcohol-coniferyl alcohol heterodimeric benzodioxane structure in the lignin of COMT-deficient Populus has been identified . An analogous (benzodioxane) metabolite (5-hydroxconiferyl alcohol-sinapyl alcohol), reported by [18, 31], may be the lignan RT 15.09 min (molecular ion (M+) 620, key m/z 510 420 235), which co-elutes with another lignan that is unique to COMT-deficient plants with key m/z 620 239 354 323 265, the latter three m/z are typical of iso-sinapyl alcohol/sinapyl alcohol and suggest the peak may be an iso-sinapyl alcohol heterodimeric lignan. However, this has yet to be verified. Another COMT-deficient unique lignan occurred at RT 15.18 min (M+ 530 219 354) and is likely an iso-sinapyl alcohol-phenolic acid conjugate. The generation of such lignans following pretreatment suggests that iso-sinapyl alcohol may be a wall component, but we have not found any evidence to support this. Furthermore, the hypothesis that incorporation of the novel monolignol may result in a lower degree of polymerization of the lignin molecule, was not supported from the GPC analysis. It can be concluded that the presence of iso-sinapyl alcohol did not affect the molecular weight of lignin produced by horseradish peroxidase catalyzed dehydrogenative polymerization of either coniferyl alcohol or sinapyl alcohol. In addition, the yield and degree of polymerization (DPn) are slightly lower than the literature report , in which a larger scale of HRP-catalyzed DHP of sinapyl alcohol in the presence of sodium azide was carried out (0.5 mmol sinapyl alcohol: isolated yield: 54.2%;
: 1.3; DPn : 4.4). Furthermore, a follow-up analysis of the presence of iso-sinapyl alcohol in transgenic COMT-deficient switchgrass biomass that had water-soluble constituents removed, followed by sequential enzymatic saccharification with fungal (Trichoderma reesei) enzymes, followed then by exposure to cellulolytic microbes Caldicellulosirupter bescii, C. obsidiansis, and C. thermocellum, indicated that no iso-sinapyl alcohol was detected in the culture supernatants, whereas sinapyl alcohol, coniferyl alcohol, and 5-hydroxyconiferyl alcohol were present. We conclude that iso-sinapyl alcohol is not a major cell wall constituent and should be considered a monolignol analog, given its structure and coupling propensities. This explains the lack of evidence of cell wall structures derived from iso-sinapyl alcohol in the present study, whereas we were able to detect benzodioxane substructures by HSQC NMR of internode 1 biomass of switchgrass, as has often been reported in COMT-deficient plants [18, 29, 32]. Although not detectable in wild-type plants, these substructures constituted 11% of the total lignin linkages, similar to the 12% observed in COMT-deficient Arabidopsis, and 10% in COMT antisense Populus. It should be noted that despite the number of previous studies of various plants species with reduced COMT activity e.g., [15, 18, 30, 33], iso-sinapyl alcohol has not been previously identified in such plants, nor has it been identified with the lignin biosynthetic pathway. However, given the associated occurrence of iso-sinapic acid, iso-sinapyl aldehyde, upstream precursors from the lignin pathway, including 5-hydroxyferulic acid and 5-hydroxyconiferaldehyde, accumulation of glucosides of 5-hydroxyconiferyl alcohol, we conclude that the metabolite flux associated with 5-hydroxyconiferyl alcohol production and subsequent metabolism differs in switchgrass from the other species previously characterized. The accumulation of 5-hydroxyconiferyl alcohol related precursors and glucoside conjugates, provide the substrates that can then be methylated at the para- position on the aromatic ring to generate iso-sinapyl alcohol and related metabolites. The production of iso-sinapyl alcohol and its glucoside, iso-syringin, may be non-specific detoxification processes. Other species that have lower rates of production of 5-hydroxyconiferyl alcohol or greater flux of 5-hydroxyconiferyl alcohol into cell walls don’t permit the accumulation of the substrates that would lead to iso-sinapyl alcohol production, and hence, a possible explanation for the lack of their detection in previous studies.
The evidence of reduced recalcitrance to deconstruction processes recently reported by  may be related to the additional metabolite responses associated the appearance of iso-sinapyl alcohol, namely the increased incorporation of phenolic acids of the lignin pathway, particularly ferulic acid, 5-hydroxyferulic acid, and ferulic acid-glycoside conjugates. These changes result in the reduced p-coumaric acid to ferulic acid ratio that has been associated with increased forage digestibility in six barley lines , but, paradoxically, increased recalcitrance in switchgrass . In monocots, such as corn, the p-coumaric acid in secondary cell walls is thought to be bound to lignin, whereas ferulic acid serves as a bridge between lignin and hemicellulose . Sophisticated re-engineering of the cell walls by monolignol substitution with methyl caffeic acid, caffeoylquinic acid, and feruloylquinic acid, has succeeded in creating cell walls that have less lignin and are more easily deconstructed . The COMT-deficient switchgrass was reported to have an increased dry matter digestibility . Similar responses were reported for COMT-deficient tall fescue [6, 15]. Although increased incorporation of ferulic acid into cell walls may lower cell wall recalcitrance, ferulic acid is thought to be one of the most inhibitory factors contributing to the biodegradability of biomass . Phenolic acids and aldehydes derived from cell wall biodegradation are known fermentation inhibitors [38–40]. These cumulative responses are likely correlated with the metabolic block in the lignin pathway plus the observed lowered recalcitrance, yielding increased enzymatic sugar release from cell walls during deconstruction. Therefore, the sum of the increase of many phenolic constituents in COMT-deficient plants, including ferulic acid, its many conjugates, and the phenolic aldehydes, may explain, at least in part, the observed increase in the inhibitory nature of these plants relative to wild-type controls. When added separately to media, iso-sinapyl alcohol (up to 50 μg/ml) and iso-sinapic acid (up to 25 μg/ml) were not inhibitory to the growth of C. thermocellum cultures (data not shown). The complex changes in the cell walls of transgenic biomass that include the greater release of phenolic acids and aldehydes must be tolerated by cellulolytic microbes. However, given the significant increase in the mass yield of fermentation products with the COMT transgenic switchgrass and the observation that simple washing allows efficient fermentation by yeast and C. thermocellum, these transgenic biomass sources remain valuable and viable future resources for biofuels.