As indicated in Figure 7.1, the phenylpropanoid pathway has several hydroxylation steps leading to the monolignols 1-5. In this context, all three enzymatic steps, in terms of historical discovery and recent progress made, were comprehensively discussed in Lewis etal. (31) and Anterola and Lewis (77). The first two of these, C4H and pC3H, are common to all known vascular plant species, whereas the third “ferulate-5-hydroxylase” (F5H) is restricted mainly to the angiosperms. An apparent exception to this is that some members of the “primitive” Selaginella family also form syringyl lignins (37-39), this being attributed to an F5H homologue (100).

All three conversions are now established to be cytochrome P-450 catalyzed (with each utilizing an associated NADPH reductase). Of these, C4H was discovered by Russell and Conn in 1967 (101), with the encoding gene reported in 1993 by Teutsch et al. (102) and Mizutani etal. (103). pC3H, on the other hand, was the subject of much literature confusion, since two other enzymatic processes were reported as being responsible for caffeic acid (10) formation: namely, a polyphenol oxidase (104, 105) and a FAD-dependent oxidase (106). These conversions have never been established though to have any physiological role as regards biosynthesis of caffeate derivatives leading to monolignol 2-5/lignin formation. On the other hand, the seminal contributions made by Heller and Kuhnl in 1985 (107) had established that, using parsley (Petroselinum crispum) cell suspension cultures, p-coumaroyl shikimate (25) was converted into the corresponding caffeoyl shikimate derivative (27) through the action of a cytochrome P-450. [By contrast, p-coumaric acid (9) was not converted into caffeic acid (10).] Yet, it was not until much later that the true significance of this discovery was fully recognized [see Lewis etal. (31)]. That is, they had, in fact, identified both the enzymatic nature of the hydroxylase, as well as the substrate(s) being utilized, as early as 1985. It was not until 2001, however, that a cytochrome P450 gene encoding pC3H was reported by Schoch etal. (108). In agreement with Heller and Kuhnl (107), the heterologously expressed protein was able to convert p-coumaroyl shikimate (25) into the corresponding caffeoyl (27) derivative, but not p-coumaric acid (9) into caffeic acid (10).

Since both C4H and pC3H are endoplasmic-reticulum (ER) membrane bound, it is quite interesting that the initial hydroxylation product, p-coumaric acid (9), needs first to be esterified prior to hydroxylation at the carbon 3 position. What that specifically means as to how the H versus G/S pathway segments are spatially and temporally kept distinct though still awaits further clarification. Yet, as for the pC3H discovery of Heller and Kuhnl (1985) (107), the formation of p-hydroxycinnamate shikimate esters had also long been known, albeit overlooked. These were considered to be possible phenylpropanoid pathway intermediates as early as in the 1970s, i. e., but again without initial recognition by many other researchers. Indeed, hydroxycinnamoyl CoA:shikimate hydroxycinnamoyl transferase (HCT) was first detected by Stockigt and Zenk (1974) (109). Only much later (110, 111) (Cardenas et al., this laboratory, unpublished results), were the genes encoding both HCT, and the closely related hydroxycinnamoyl CoA:quinate hydroxycinnamoyl transferase (HQT), cloned from tobacco, and their roles in monolignol metabolism defined. Interestingly, HCT preferentially utilizes shikimate (29) as an acyl group acceptor, whereas HQT prefers the quinate (28) moiety instead.

The final cytochrome P-450 hydroxylation step, F5H, was initially discovered by Grand in 1984 (112) and was considered to utilize ferulic acid (11) as substrate to afford 5-hydroxyferulic acid (12); the latter had previously been discovered in Zea mays and Hordeum vulgare (113). Later, the group of Fukushima et al. (114, 115) demonstrated that when [9-2H2,OC2H3] coniferyl alcohol (3) was administered to Magnolia kobus root tissue, it was metabolized into pentadeuterated-syringyl units with two 9-deuterium atoms via hy­droxylation at the C-5 position; moreover, the deuterium label was also present in the S-lignin that was formed. These seminal findings were subsequently confirmed by re-examination of the catalytic properties of a recombinant Arabidopsis F5H (116) where it was established that both coniferyl alcohol (3) and coniferyl aldehyde (21) served as preferred substrates over ferulic acid (11). In apparent agreement with possibly defining roles in carbon allocation in the pathway, it is of interest that in Arabidopsis there are only 1 C4H, 1 pC3H, and 1 HCT genes, as well as possibly 2 F5Hs, i. e., in contrast to the 6- and 4-multigene ADT and PAL families.