Another distinct dehydrogenase, this time from aspen, was claimed to be specific for sinapyl alcohol (5)/syringyl lignin formation (143). This report from the Chiang laboratory was quite unexpected. This was because its actual broad substrate versatility for cinnamyl alde­hydes 19-23 eliminated it as being biochemically-specific for sinapyl alcohol (5)/syringyl lignin formation (77). That is, and as previously noted for all the CADs proper, this enzyme was also substrate versatile. Thus, any substrate specificity, if it exists at all, would presum­ably be a result of compartmentalization, i. e., where sinapyl aldehyde (23) and the so-called “sinapyl alcohol dehydrogenase” were co-localized. Since this has not been established, the physiological role of this dehydrogenase remains unknown at present. More recently, an X-ray crystal structure for “SAD” was also obtained (144), but which contained a vastly different substrate-binding pocket (in terms of both size and amino acid composition) to that of a bona fide CAD, i. e., AtCAD4/5 (133). Specifically, only 2 of the 12 amino acid residues which constitute the CAD substrate-binding pocket were conserved in “SAD,” with the substrate-binding pocket for the latter also being considerably larger (133). Taken to­gether, these data suggest an alternative biochemical function for “SAD,” as also provisionally suggested by Bomati and Noel (144).

Furthermore, in Arabidopsis, no evidence for any requirement for a “SAD” was obtained, since >94% of all monolignol 1-5 formation for lignification was carried out by AtCAD4/5, both of which share considerable homology (74-83% similarity) to that of bona fide CADs (56). Moreover, the dehydrogenases of highest similarity to the “sinapyl alcohol dehydroge­nases” (i. e., At4g37970 and At4g37980) did not reduce the p-hydroxycinnamaldehyde 19-23 to afford the monolignols 1-5 to any considerable extent in vitro (56). Thus, as for the other dehydrogenases described above, clarification of their biochemical and physiological roles need to be established as well.

A rice mutant, gold hull and internode (gh), was also first described as early as 1917, and characterized by a reddish-brown color in the hull and internode but not in the midrib. Since then a series of mutants were identified, i. e., gh1-gh4. Zhang et al. (145) recently characterized the gh2 mutant and showed that the GH2 gene encodes a CAD. The substrate versatile kinetic properties of the recombinant GH2 proteins were determined showing Kenz values for coniferyl (21) and sinapyl (23) aldehydes of ~289 000 versus ~ 162 000 M-1 s-1, respectively, these being in relatively good agreement with kinetic values for AtCAD5 in Arabidopsis (56). The recombinant mutant gh2 protein, however, lost the corresponding activities. Additionally, analyses of CAD activity in all tissues (i. e., panicle, hull, blade, midrib, sheath, internode, and root) of both wild type and gh2 mutant plants showed that the formation of coniferyl alcohol (3) was greatly reduced in the roots, internodes, hulls, and panicles of the gh2 mutant, with no detectable formation of sinapyl alcohol (5). On the other hand, the gh2 mutant had apparently little to no effect on overall lignification (estimated at a 5-6% reduction). Taken together there is no evidence thus far that “SAD” has the specific biochemical/physiological functions reported earlier (143).