Laurence B. Davin, Ann M. Patten,

Michael Jourdes, and Norman G. Lewis

7.1 Lignin: molecular basis and role in plant adaptation to land

Life, as humanity understands it, has inextricably been tied since eons past to the successful evolutionary adaptation of aquatic flora to terrestrial environments. This conquest appar­ently first began with emergence of various “primitive” forms of, and/or forerunners to, tracheids (water-conducting elements) in the land-based plants during the late Ordovician to Silurian periods (>400 million of years) (1-3). Such land plant forms gradually became capable of efficient hydration and metabolism under “water-limited” conditions and hence attained an inherent ability to survive in a wide variety of habitats. Eventually, various other modified forms of plant cell walls also evolved, these containing — at least for the vascular apparatus — celluloses, hemicelluloses, lignins, and small amounts of proteins, as their main chemical/structural constituents. Ultimately, lignification provided the means by which large upright vascular plant forms could be produced, and which enabled some species to more successfully compete for photosynthetic energy. This, in turn, gave a molec­ular or structural basis for much of the plant biodiversity that humanity enjoys today in its many resplendent forms. Furthermore, in addition to competition for light, an upright growth habit (provided by a true vascular system), allowed for better spore/pollen dispersal, increasing the genetic variability and species range. Yet today, our knowledge of how plant cell wall assembly occurs is at the most rudimentary level.

This chapter focuses upon the lignins. Next to cellulose, they are Nature’s second most abundant organic substances, and are products of the phenylpropanoid pathway (Figure 7.1). Significantly, many of the monolignol/lignin-forming pathway steps apparently also evolved during transition of plants to the land habitat, providing broad adaptive advantage to some 350 000 or so distinct present-day vascular plant species (4). To put this pathway into the broader context of carbon allocation from photosynthesis, woody gymnosperm stems gen­erally have lignin contents ~28-30%, whereas those in woody angiosperms are of lower amount (~20-24%) (5). In both cases, this represents a significant commitment and des­ignation of the total carbon taken up during photosynthesis; indeed, the lignins are some of the most metabolically “expensive” of all plant products formed (6).

Studies of many different plant species (i. e., from gymnosperms to angiosperms) have established that their lignins proper, while often varying in monomeric compositions, are derived from the three monolignols 1, 3, and 5 (Figure 7.1); additionally, various grasses

Biomass Recalcitrance: Deconstructing the Plant Cell Wall for Bioenergy. Edited by Michael. E. Himmel © 2008 Blackwell Publishing Ltd. ISBN: 978-1-405-16360-6

СН2ОН	СН2ОН	СН2ОН	СН2ОН	СН2ОН
но у осн3 он 5-Hydroxyconiferyl alcohol (4)	н3со у осн3 он Sinapyl alcohol (5)
CAD	CAD
ОН 5-Hydroxyconiferyl aldehyde (22)	OH Sinapyl aldehyde (23)
CCR	CCR
OCH3
OH Sinapoyl CoA (18)
OH Sinapic acid (13)
Ferulicacid (11) 5-Hydroxyferulic acid (12)
Figure 7.1 Contemporary view of the phenylpropanoid pathway. pC3H, p-coumarate 3-hydroxylase; C4H, cinnamate 4-hydroxylase; CAD, cinnamyl alcohol dehydrogenases; CCOMT, hydroxycinnamoyl CoA O-methyl — transferases; CCR, cinnamoyl-CoA oxidoreductases; 4CL, hydroxycinnamoyl CoA ligases; COMT, caffeic acid O-methyltransferases; F5H, ferulate 5-hydroxylase; HCT, hydroxycinnamoyl CoA:shikimate hydroxycinnamoyl transferase; HQT, hydroxycinnamoyl CoA:quinate hydroxycinnamoyl transferase; PAL, phenylalanine ammonia lyase; TAL, tyrosine ammonia lyase. Note that most of the current understanding of biosynthetic steps results from in vitro analyses; how the hydroxycinnamic acids 10-13 are formed via, for example, CoA esters or aldehydes needs to be fully established.

contain small amounts (~10% or so) of p-hydroxycinnamate esters 30-32 linked to the monolignols (Figure 7.2A). Lignins are also generally designated as p-hydroxyphenyl (H), guaiacyl (G), and syringyl (S), based on their aromatic ring substitution pattern (Figure 7.2B). Regardless of monomeric compositions, lignins are considered to be complex, amor­phous, phenylpropanoid polymers within the lignocellulosic matrices of plant cell walls. This highly conserved monolignol deployment occurs irrespective of the remarkable dif­ferences encountered in both diverse body plans and architectures of the plant forms in existence today.

In spite of their overall abundance, the lignin biopolymers are quite challenging to work with. This is because of their aromatic and hence very hydrophobic character; their relative intractability due to difficulties to efficiently solubilize and characterize derivatives thereof and then only under quite drastic chemical degradation conditions (7); and their striking abilities to self-associate (8, 9) due to the strong electronic stabilization energies between the various subunits in the adjacent biopolymeric chains. In terms of their molecular ar­chitecture, the lignins are considered to be products of phenoxy-radical-derived coupling reactions as originally proposed by Erdtman in 1933 (10) in model studies using isoeugenol (33, Figure 7.2C); Figure 7.2D also illustrates several of the most abundant interunit linkage types known in lignins proper that can be at least partially quantified (discussed below).

It needs to be emphasized that the artistic depictions of lignin macromolecular config­uration have changed rather enormously over a time span of nearly five decades, i. e., from originally being envisaged as a complex three-dimensional (Bakelite-like) phenolic poly­mer (11, 12) (Figure 7.3A), to that amenable for a computer-programmable simulation (Figure 7.3B) (13,14), to that now more recently of essentially more linear macromolecular entities (e. g., Figure 7.3C) (15). All of these models though represent simply artistic and consequently quite artificial depictions of native lignin structure(s). Indeed, several quite abundant substructures (e. g., 8-1′ or their forerunners, dibenzodioxocin, etc.) are even absent in Figures 7.3A and 7.3B; moreover, the very changing nature of such depictions un­derscores the fact that little truly systematic research has yet been carried out to investigate (and develop methodologies for) the study of lignin primary structure(s). Furthermore, there maybe no other biopolymeric entities whose structure(s) has (have) been depicted in such quite arbitrary ways, particularly since experiments had neither been conducted nor devised to investigate both the actual biochemical mode of assembly and the structure(s) so obtained. Yet this has been the situation for lignins for almost a century now.

Interestingly, in the 1970s and 1980s, there was also much enthusiasm in identification of lignin-degrading enzymes, with various laccases and peroxidases (specifically manganese and lignin peroxidases) reported as the main degradative enzymes involved (16-19). Today, it is doubtful that either class has such a function, and neither has fulfilled the optimistic promise for industrial application once envisaged over two decades ago (20). Moreover, our understanding today of how lignin deconstruction (so-called biodegradation) occurs in vivo is still at a most rudimentary level. For this reason, research activities are now being directed and/or initiated toward identifying the nature of true lignin degrading enzymes, e. g., putative lignin depolymerases acting on specific linkage types (21).

Nevertheless, the fantastic diversity of the extant vascular plant species — in terms of not onlytheir remarkable differences in size, shape, and growth rates, but also as to whether they have woody or non-woody character, etc. — has all depended upon the successful formation of the lignified vascular apparatus.

The maize bm3 mutant (61), now known to affect COMT activity (66), is a plant line with reduced stem stalk strength and prone to stalk lodging (falling and collapsing) relative to wild type, thereby currently precluding its commercial utilization (250) [see discussion in Anterola and Lewis (77)]. Not having a rate-limiting function in the pathway in terms of carbon allocation, modulation of COMT activity does not affect overall lignin contents. Instead, the monolignol, 5-hydroxyconiferyl alcohol (4), is formed rather than sinapyl alcohol (5); the former is then used in place of sinapyl alcohol (5), but whereby a polymeric structure is generated whose properties are such that it results in a weakening of overall plant vascular integrity due to the incorporation of (4) into the lignifying matrix. This is again considered to reflect a tightly programmed lignification response, albeit where the outcome is affected by a limited substrate degeneracy during template polymerization. This, however, results in diminished cell wall properties and helps explain why 5-hydroxyconiferyl alcohol (4) never evolved to be an effective lignin precursor.

As indicated earlier, the role of COMT in the second methylation step of syringyl lignin biosynthesis was first demonstrated by Atanassova et al. (136) in tobacco, and later fur­ther confirmed by others in poplar (182), aspen (251), and alfalfa (252). Downregulation/ mutation of COMT though had no apparent adverse effect on overall lignin amounts as indicated using Klason lignin estimations (77), with the striking exceptions reported from the Dixon laboratory (186, 212). These latter researchers had again used the unreliable thioglycolic acid lignin (186) and the neutral detergent fiber (NDF)/Klason lignin methods (212), the results of which appeared to indicate that lignin contents had been (significantly) reduced (by circa 60%). While this interpretation can now be viewed as incorrect, this was the same methodology used for the PAL (207, 212) and C4H (207) studies; hence, the over­all findings from these studies should also be reexamined with more robust and reliable technologies.

Additionally, in preliminary studies, COMT downregulation had no apparent significant adverse effect on the amounts of thioacidolysis-releasable G-monomers in the lignins from either tobacco or poplar, given that the total amounts, relative to lignin contents, were very similar to the wild-type lines (77). On the other hand, little to no S-units could be detected. To account for this, we can provisionally propose that in the COMT downregulated lines, the 5-hydroxyconiferyl alcohol (4) moieties were undergoing homo-coupling, whereas in the wild-type lines, sinapyl alcohol (5) was also doing the same. However, instead of 8- O-4′ bond formation, etc. occurring as in the wild-type line (see Figure 7.15A), the presence of the flanking 5-hydroxyl group (from 4) has resulted in an apparently near-quantitative replacement with the benzodioxane substructure (Figure 7.15B), rather than the “simple” 8-O-4′ interunit linkage present in the wild-type line. That is, the main difference be­tween the 8-O-4′ linkage (Figure 7.15A) and the benzodioxane substructure formation (Figure 7.15B) is only in “trapping” of the quinone methide intermediate during template polymerization. [By contrast, if the incoming monolignol radical was a p-coumaryl (1), coniferyl (3), or sinapyl (5) alcohol moiety, the “trapping” would be performed by an exter­nal water or possibly carbohydrate molecule.] Nevertheless, formation of the benzodioxane substructure can thus still be considered as an 8-O-4′ interunit linkage, even if the overall substructure generated differs.

Furthermore, with the Arabidopsis COMT-downregulated line, there provisionally appears to be a near equivalent reduction in amounts of G-releasable monomers to that of the original S levels (see Figure 7.15C which plots G and S 8- O-4′ interunit cleavage, leading to monomer release, at different stages of Arabidopsis growth and development; Jourdes et al., manuscript in preparation). In this case, the near equivalent reduction of both G and 5-OHG (previously S) moieties at different stages of growth and development suggests that G-S hetero-coupling was mainly occurring in the wild-type line, i. e., via hetero-coupling of coniferyl (3) and sinapyl (5) alcohol-derived monomers, thereby affording cleavable 8-O-4′ interunit linkages. This is now apparently replaced by an equivalent level of hetero ­coupling between coniferyl (3) and 5-hydroxyconiferyl (4) alcohol-derived radicals to afford the mixed benzodioxane substructure (Figure 7.15B). This near 1:1 reduction in releasable G:S moieties thus places another considerable constraint on how lignin macromolecular configuration is actually being achieved. It could, for example, possibly indicate that G and S monomers are being alternately laid down during macromolecular lignin assembly;

such observations once again underscore the urgent need for technology development to determine lignin primary structure(s).

In alfalfa, however, COMT downregulation resulted in reduction of both G-S and S-S amounts, this in turn presumably reflecting differing cell-type lignification processes. In this regard, although the incorporation level of 5-hydroxyconiferyl alcohol (4) moieties into the

lignin was not quantitatively estimated using chemical lignin degradative analysis, attempts to do this indirectly were carried out, i. e., by attempting to estimate the level of benzodioxane substructures using NMR spectroscopic analyses of the extracted lignin-enriched isolates by measuring volume integrals in the 2D HMQC spectra (253). Reassessment of this data herein also results in the provisional conclusion that overall levels of 8— O—4′ interunit linkages were actually conserved in both wild-type and COMT-deficient alfalfa lines. That is, the wild-type line apparently had circa 81% of 8— 0—4 linkages, whereas the COMT-deficient alfalfa had ~38% of benzodioxane and 44% of 8—0—4′ linkages. Together, these represent ~82% of substructures with 8— O—4′ linkages, and thus represent a high degree of conservation of the overall 8— O—4′ linkage type in both wild type and COMT lines.

These data are thus also in apparent harmony with a proposed limited template degeneracy in assembly of the lignin macromolecule — but where both the chemistry on the proposed template is altered in the mutation (to involve catechol, rather than phenolic, chemistry). It is worthwhile reflecting that the alfalfa data has been known since 2003 (253), and yet has been interpreted by others as evidence for random coupling/combinatorial chemistry leading to 1066 isomers, etc. (175). On the other hand, while the COMT mutation results — in a presumably ordered way — in formation of benzodioxane substructures (Figure 7.15B), the biophysical effects, nevertheless, resulted in weakening of the overall vasculature.

The previous section has provided a very basic introduction to the concept of molecular dynamics. However, up to this point we have only covered what is required to run a simulation and obtain the variation in atomic positions as a function of time. The sobering fact is that even with computers many orders of magnitude faster than today’s machines we will still not be able to achieve biological time scales for all but the very fastest enzymes. Thus, there is little to be gained scientifically in just running an MD simulation and waiting to see the system react. Firstly, the force field prevents covalent reactions from occurring and even if a method that supports bond breaking is used, such as QM/MM, the simulation time scales are too short for direct observation. Thus, it is necessary to make use of a large number of different techniques and analysis methods in order to uncover scientifically meaningful information. There are a huge number of different analysis techniques that can be used with MD methods and more are being devised every day. Here we list some of the more common and useful techniques for studying cellulose and cellulose hydrolysis but this represents a very small subset of what is possible.

Clustering. Clustering is a method of postprocessing the sampling results of an MD sim­ulation to quantify the number of more highly populated states the system visits, and can be used to reprocess a dynamics trajectory to map the essential transitions between states. In particular, when multiple structural states are observed in MD simulations, the identity of the states and the populations in the MD simulation are quantified in a cluster analysis. Once one knows the states, the sequence of transitions between states can be reconstructed from the same trajectory, revealing the nature and frequency of transitions and the paths through intermediate states if there are any. Cluster analysis can be used to analyze any trajectory, whether it is constrained in any way or not, so care must be taken to interpret the results within the framework of any biasing potentials. In cellulose modeling, this kind of analysis is essential for characterizing crystalline and amorphous states and the transitions between them. For example, it is possible to use cluster analysis to test the hypothesis that a system goes straight from state A to state B, the transition being observed in either experiment or simulation. The analysis may show instead, that there is a highly populated state C that the system always, in simulation, goes through on the way, and never goes straight from A to B or B to A.

Normal mode analysis. Normal mode analysis is a powerful tool for extracting the charac­teristic motions of a macromolecule or complex such as a cellulose fiber, cellulose/lignin complex or protein/cellulose system. With this method, the high-frequency modes can be separated out allowing the low frequency and larger displacement motions to be identi­fied. Often, those are the motions that define the behavior and the biological function of macromolecules. The method involves finding a structure of the complex which is at an energy minimum and subsequent diagonalization of the mass-weighted Hessian matrix from which the frequencies of motion and normal mode vector (eigenvectors) can be extracted. However, for larger molecules, the diagonalization of the matrix can quickly become too large a problem for most computers. There are other methods of treating larger molecules and systems of molecules that rely on simple assumptions. One method of treating the larger systems is to minimize the structure to a minimum energy state using the all-atom model, then switch a more approximate method such as the elastic network model, where selected atoms such as alpha carbons are all connected together by a series of harmonic springs, or the Rotations Translations of Blocks (RTB) model which uses an approximate diagonalization method by combining multiple residues into rigid blocks, and then apply the normal mode method (25). Quasiharmonics is a variation of the normal mode analysis in which the effective modes of vibration are calculated from fluctuations which are determined from a MD simulation. Since the fluctuations in an MD simulation contain anharmonic contributions, the quasiharmonic vibrational

modes may differ from the normal modes calculated from the energy minima. The nor­mal mode method has been extended to follow troughs in the potential energy surface, and give information about essential modes of extrema besides minima such as saddles, or transition states, and maxima.

Local water density. Local water density concerns are important when considering many biological models, especially in cases where complex structures compartmentalize regions of water away from bulk water. Considering the complexity of these situations, readers are encouraged to consult specific examples in the literature.

David B. Wilson

11.1 Introduction

An important step in the global carbon cycle is the degradation of cellulose, the most abun­dant form of fixed carbon with 1011 tons produced by plants each year (1). Most terrestrial cellulose is degraded by cellulolytic microorganisms, primarily fungi and bacteria, although some cellulose is recycled by fire and by photodegradation (2). Aerobic microorganisms are responsible for much of the cellulose degradation in soils, but there also are many species of cellulolytic anaerobic soil bacteria such as Clostridium thermocellum, C. cellulovorans, and Acetivibrio cellulosolvens (3-5). Termites and some other insects are very important in cellu­lose degradation, especially in tropical regions, and most cellulose degrading insects contain symbiotic cellulolytic microorganisms, even though many termites and other insects pro­duce cellulases (6, 7). Some very cellulolytic termites utilize aerobic symbiotic cellulolytic fungi to breakdown plant material in their nests and then eat the fungi and residual plant material (8). Ruminants, such as cows, sheep, and deer, also are important cellulose de­grading organisms but all of the cellulose that they utilize is degraded by symbiotic rumen microorganisms, primarily bacteria. The rumen is an extremely anaerobic environment and this area has been reviewed recently (9, 10), so that these organisms will not be discussed further here.

Microorganisms catalyze cellulose degradation by producing enzymes called cellulases, which hydrolyze the (3 -1-4 linkages present in cellulose. Almost all cellulytic microorganisms secrete their cellulases outside their cell wall, as bacteria and fungi are unable to transport insoluble materials, like cellulose, inside the cell. The soluble sugars produced by cellulase digestion of cellulose are transported inside the cell and metabolized. Native cellulose is very resistant to hydrolysis because it is insoluble and contains crystalline regions in which the adjacent cellulose molecules have strong interactions, such as hydrogen bonds and hydrophobic stacking. Thus, the specific activities of individual cellulases are much lower than those of most enzymes. However, in terms of catalytic enhancement, cellulases are very active enzymes, as the half-life of crystalline cellulose in water at neutral pH is estimated to be about 100 million years. It takes concentrated sulfuric acid at 125°C to hydrolyze native cellulose at a reasonable rate. When cellulases are assayed on low molecular weight soluble substrates, they show normal Michaelis-Menten kinetics and some have high specific activities, showing that they are basically similar to other enzymes. However, when cellulases are assayed on insoluble substrates, they have very different properties, as the assays are

Biomass Recalcitrance: Deconstructing the Plant Cell Wall for Bioenergy. Edited by Michael. E. Himmel © 2008 Blackwell Publishing Ltd. ISBN: 978-1-405-16360-6

usually nonlinear with time and with the amount of enzyme. Different studies have produced different explanations for this behavior. For the endocellulase Thermobifidafusca Cel6A, the nonlinearity was shown to be due to substrate heterogeneity (11).

14.5.3.1 Liquid hot water batch pretreatment

In addition to uncatalyzed steam explosion pretreatments, other uncatalyzed pretreatment processes using pressurized liquid hot water without rapid decompression have been inves­tigated. Process conditions have been developed for cellulose hydrolysis at very high temper­atures of about 260°C (10, 43) and as a pretreatment approach for achieving hemicellulose hydrolysis at lower temperatures of about 200-230°C (44, 45). High yields of soluble sugars from the hemicellulose fraction of some biomass types (primarily herbaceous crops and agricultural residues) can be achieved, but liquid hot water processes generally liberate the sugars in an oligomeric form and thus require a secondary acidic or enzymatic hydrolysis step to produce fermentable monomeric sugars.

Another approach for liquid hot water pretreatment uses some chemicals as agents to control the pH in the range of 4-7 (10, 46). With some feedstocks, such as corn stover, there may be enough inherent buffering capacity from the feedstock that the target pH range is achieved without any requirement of pH-controlling chemicals. The reason for controlling the pH is to retain the released hemicellulose sugars in oligomeric form as a means of controlling sugar degradation losses and fermentation inhibitor formation (10). However, recent data suggests that this approach may not be effective in achieving high enzymatic digestibility of the cellulose (>80% glucose yield from the available cellulose) in pretreated corn stover (17). In general, liquid hot water pretreatments are attractive from a process cost-savings potential (no catalyst usage, low-cost reactor construction due to low-corrosion potential), but these cost savings can be offset by lower overall sugar yields and the need for enzymatic hydrolysis conversion steps (21).

HG may be partially O-acetylated at C-2 or C-3 of GalA (198, 199). No gene for HG acetyltransferase has been identified; however, O-acetyltransferase activity in microsomes from suspension-cultured potato cells (339) has been shown to transfer [14C]acetate from [14C]acetyl-CoA onto endogenous acceptor in the microsomes to yield a salt/ethanol pre — cipitable product from which approximately 8% of the radioactivity could be solubilized by treatment with endopolygalacturonase and pectin methylesterase. Such results could indicate the presence of HG-AT, although the possibility that the radiolabeled acetate was transferred either onto RG-II or RG-I that was solubilized by the glycanase treatments, and thus represents an enzyme that acetylates one of the other pectic polysaccharides that may be covalently linked to HGA, cannot be ruled out.

UDP-Apiose Synthase (UAS, AXS) converts UDP-GlcA in the presence of NAD+ to UDP — Api. The enzyme decarboxylates UDP-GlcA to form a UDP-4ketopentose intermediate and the release ofCO2, and then catalyzes rearrangement ofthe sugar skeleton to form UDP-Api

(476,477). In vitro, the enzyme forms both UDP-Xyl and UDP-Api (403). We believe that the formation of UDP-Api was not confirmed satisfactorily since the product is readily degraded to cyclic apiose1,2-phosphate. Functional genes encoding AXS were isolated from tobacco (358), Arabidopsis and potato (Guyet and Bar-Peled, unpublished) (478). Two isoforms exist in Arabidopsis (At1g08200 and At2g27860) and are predicted to be in the cytosol. The “dual function” of the enzyme in generating both UDP-Xyl and UDP-Api was assayed by NMR spectroscopy (Guyet and Bar-Peled, unpublished). NMR time course assays for the conversion of UDP-GlcA into UDP-pentose using recombinant potato UAS, confirms explicitly that UDP-Api is made first. The analysis indicates that in vitro, UDP-Xyl synthesis lags behind UDP-Api. Mutation in UDP-Api synthase in Nicotiana benthamiana is lethal as a consequence of the lack of RG-II (358). To unambiguously determine if the enzyme is bifunctional and contributes to UDP-Xyl synthesis, a mutation in the three cytosolic UXS (Type C) genes must be carried out.

7.3.8.1 CAD1

In addition to the bona fide CADs [such as from tobacco (126), loblolly pine (35) and Arabidopsis (56, 57)], there continue to be a number of reports proposing that another class of dehydrogenases (so-called CAD1) is involved in monolignol/lignin biosynthesis (140, 141). These enzymes, by contrast, bear little homology to bona fide CADs and lack both catalytic and structural zinc metal ions (56). For example, a dehydrogenase, annotated as CAD1, was purified from Eucalyptus gunnii in parallel with a bona fide CAD, then called CAD2 (142). CAD1 was shown to have lower affinity for p-coumaryl (19)/coniferyl (21) aldehydes (Km = 70 and 25 ^M) as compared to CAD2 (Km = 1.2 and 1.7 ^M), with sinapyl aldehyde (23) not serving as a substrate; CAD1, however, was not characterized further. It was also presumed to be a monomer (~38 kDa), as compared to CAD2, which is a dimer (~83 kDa) (142). However, the encoding gene (CAD1-5) (140) had only 30/15% similarity/identity to bona fide CADs [e. g., AtCAD5 (56)].

Since then, two other alcohol dehydrogenase genes (so-called NtCAD1-1 and NtCAD1-7) present in tobacco were also obtained by screening a cDNA library with the E. gunnii CAD1-5 coding sequence. Curiously, these were reported as a new step for the biochemical formation of coniferyl alcohol (3) (141). Both dehydrogenases again lacked the Zn catalytic center and the Zn-binding signature found in bona fide CADs, and also had ~30% similarity and ~ 15% identity compared to AtCAD5. The kinetic parameters of both NtCAD1-1 and NtCAD1- 7 were nevertheless investigated using coniferyl aldehyde (21)/alcohol (3) and sinapyl alcohol (5) as potential substrates. Both proteins were apparently capable of catalyzing the forward reaction of reduction of coniferyl aldehyde (21), as well as the oxidation of coniferyl alcohol (3), but they were apparently unable to convert sinapyl alcohol (5) to sinapyl aldehyde (23); the forward reaction was not examined. The Kenz values for both

NtCAD1-1 and NtCAD1-7 were very modest (low) relative to bona fide CADs, i. e., 3230 and 7650 M-1 s-1 versus 348 000 M-1 s-1 for AtCAD5. We would suggest that this family of dehydrogenases, even though they may be able to inefficiently reduce coniferyl alde­hyde (21) to coniferyl alcohol (3) in vitro, have biochemical functions unrelated to that of monolignol/lignin biogenesis. Indeed, as discussed below, a recent study has also reported that genes highly homologous to “CAD1” are instead 2-phenylacetaldehyde reductases. As a further caution, numerous dehydrogenases are known to exhibit broad substrate versatility of varying levels of efficacy.

As indicated above, none of the various “representations” of lignin structure have ade­quately reflected native macromolecular configuration. Indeed, there may not be any other natural product whose “structure(s)” have been approximated through attempts to de­termine subunit linkage type and frequency, but not through obtaining sequences of the interunit linkages. However, the trends noted through attempts to obtain precise, rigorous, quantification and identification of subunit type, and their frequencies, in lignins present in Arabidopsis and alfalfa are considered indicative, at least by ourselves, of non-random assembly. This is not to say though that all interunit linkages can yet be accounted for — currently, ~40-45% can at best be fully quantified using these methods. In addition, lignin analyses suffer from another limitation: Technologies have not yet been developed to probe precise lignin structures in different cell wall layers and/or in distinct cell wall types (e. g., xylem versus fibers).

These calculations are computer simulations of the time evolution of atoms based upon their velocities and potential energies due to interactions with the other atoms in the system. In Car-Parrinello molecular dynamics (CPMD) (23,24) approach the potential energy due to the interactions of the atoms in the system is determined by conducting quantum mechanical
calculations of the valence and semi-core electrons in each atom. The CPMD code is based on density functional theory and is capable of simulating chemical reaction pathways for systems of up to several thousand atoms. The valence electrons were treated using the density functional developed by Becke (25) and Lee et al. (26), BLYP, and were assumed to exist in a psuedopotential exerted by the nuclei and the core electrons. The BLYP functional was shown to be appropriate to describe the liquid water (27, 28). The valence and semi-core electrons were treated with the Troullier-Martin norm-conserving pseudopotential (20). Plane waves were used as the basis functions in these calculations and the simulations were conducted using a time step of 0.125 fs in our calculations. The plane-wave basis set cut-off used was 70 Ry, which was shown to be sufficient for biomolecular simulations in aqueous solution (28). Three-dimensional periodic boundary conditions were applied. During simulations of sugar decomposition reactions in vacuum, the calculations were carried out with a (12 x 12 x 12 A3) unit cell containing one protonated sugar molecule surrounded by sufficient vacuum space to separate the interactions between the neighboring sugar molecules. Protonation was assumed to initiate the sugar degradation reactions. All probable protonation sites of the sugar molecule were investigated including the hydroxyl groups on the sugar ring and the ring oxygen. The simulations were carried out at 500 K. A total of 2 ps simulation time was carried out on a high-speed Linux PC machine. Each step took about 60 seconds of CPU time. The initial atomic coordinates of the xylose molecule were taken from the optimized structure of the CPMD calculations without the proton. The events simulated are qualitative, not quantitative in nature.

Simulations of xylose in water were carried out with each sugar molecule surrounded by 32 water molecules in a unit cell of with a lattice parameter of 11.5 A. Each sugar molecule had approximately two hydration shells. In addition, one proton was added to the system to mimic the acidic medium. Ab initio MD was carried out at constant temperature of500 K, which is at the higher end of the pretreatment temperature. Because these calculations sampled a large portion of the conformational space, they were ideally suited for finding low energy structures in sugars and in identifying likely reaction pathways.