#5

Caffeic acid O-methyltransferase (COMT)

Originally, this particular enzyme was considered to be exclusively involved in all O-methylation reactions during monolignol biosynthesis. It was thought to be bifunctional in angiosperms, in terms of phenylpropanoid substrate specificity, since various crude plant extracts (20,21) were reportedly capable of converting both caffeic and 5-hydroxyferulic acids into ferulic and sinapic acids, respectively; on the other hand, with the corresponding gymnosperm plant crude extracts the preferred substrate was caffeic acid, and it was thus initially concluded that gymnosperm COMTs were monofunctional.

Analysis of the Arabidopsis genomic sequence indicates that some seventeen genes are annotated as putative COMTs. Of these Arabidopsis homologues, COMT1 has the highest similarity (78.8 and 84.0%) and identity (75.5 and 79.6%) to that of the COMTs reportedly involved in sinapyl alcohol biosynthesis in tobacco (22) and aspen (23). Other than the COMT1 gene, however, none of the other Arabidopsis COMT homologues have been characterized in terms of their biochemical conversions in vitro, and none (including COMT1) have been studied to delineate either patterns of gene expression and/or their true metabolic contexts (roles) in vivo.

Interestingly, the analysis of the EST database entries indicated that COMT1 is expressed in all organs (particularly root tissues), whereas all others were either undetected (i.e. COMT2, 3 ,5 10, 12 and 16) and/or showed quite variable patterns of expression. As before for 4CL and PAL homologues, there is thus an urgent need to now delineate the specific roles of each COMT homologue in vivo, paying particular attention to the patterns of gene expression as dictated by the corresponding native promoters, and the corresponding metabolic context, for each. The data obtained, however, again demonstrate thus far the very limited utility of the EST databases in identifying possible physiological functions, as well as highlighting the dangers in drawing conclusions on function on the basis of homology in the absence of demonstrable biochemical conversion data.

As for the other phenylpropanoid pathway genes, there are thus three ongoing subprojects: (i) cloning of Arabidopsis COMT genes and characterization of the corresponding recombinant proteins; (ii) analysis of COMT gene “knockouts”; and (iii) identifying temporal and spatial patterns of expression of the four COMT promoters when fused to GUS/GFP reporter genes. Progress is shown in Tables 9 and 10.