Handbook of Vanilla Science and Technology

18.3 THE WAY FORWARD?

What did it take to establish the above pathway for ferulate formation? Interestingly, labeling experiments played only a small part. Rather, the major paradigm shifts came about through the application of genetic approaches in Arabidopsis, coupled with substrate specificity studies with recombinant enzymes. Unfortunately, V. planifolia does not appear to be a genetically tractable system at this stage. However, various tools of functional genomics that are now commonly applied to other systems might help throw light on the biosynthetic pathway, and some progress has already been made in this area (Pak et al. 2004; Havkin-Frenkel and Belanger 2007). The potentially applicable approaches center on gene expression profiling, both temporally and spatially.

The so-called “next generation” sequencing techniques (454 and Solexa/Illumina; www.454.com;www.solexa.com) have made it relatively simple to obtain massive expressed sequence tag datasets from relatively small amounts of tissue. Such EST datasets could easily be obtained from dissected tissues from vanilla pods throughout their period of development. As a control, similar datasets should be obtained from tissues shown not to accumulate significant amounts of vanillin, such as stems, roots, and leaves. After assembly and initial annotation of the sequences, the data can be mined for sequences matching the enzyme types predicted for involvement in vanillin biosynthesis based on all potential pathway models in Figure 18.1. Apart from the side chain shortening reaction (the most problematical part, as it is not immediately clear what the chain-shortening enzyme might look like), these will include aromatic hydroxylation and subsequent O-methylation, and possibly CoA ester reduction (analogous to CCR). It is more than likely that the hydroxylation reaction will be catalyzed by a cytochrome P450 enzyme, and that this will exhibit a significant degree of substrate specificity (Chapple 1998). Plantphenolic O-methyltransferases fall into two major classes, the type members being the so-called caffeic acid 3-O-methyltransferase (COMT, type I), which should properly be referred to as 5-hydroxyconiferaldehyde 3-O-methyl-transferase based on its preferred substrate in the lignin pathway, and the type II CCoAOMT that is also involved in monolignol biosynthesis (Noel et al. 2003). Either type could potentially be involved in vanillin biosynthesis.

In contrast to most plant biosynthetic P450 enzymes, COMT is relatively promiscuous. In fact, the enzyme from alfalfa shows high activity against 3,4-dihydroxybenzaldehyde to form vanillin (Kota et al. 2004), although this is unlikely to be a function for the enzyme in alfalfa. Because vanillin accumulation occurs over a long time period, high activity may not be critical for candidate enzymes. For example, the formation of a major strawberry aroma compound involves the activity of a COMT, even though this enzyme is much more active with monolignol precursors than it is with the percursor of the 2,5-dimethyl-4-methoxy-3 (2H)-furanone flavor compound (Wein et al. 2002). Thus, in vitro biochemistry will ultimately need to be confirmed by either genetic approaches or detailed flux analysis measurements. Rapid techniques for reverse genetics based on virus-induced gene silencing are now being developed, and work well in some monocot systems (Lu et al. 2003; Ding et al. 2006). Likewise, techniques for precursor labeling and metabolic flux analysis are becoming increasingly sophisticated (Boatright et al. 2004).

Two factors are currently limiting the final assault on the vanillin pathway; the lack of a good experimental system (e.g. a highly inducible cell or tissue cultures) to simplify labeling experiments, and the lack of economic drivers to stimulate funding for this type of work. Pure vanillin is very cheap to produce synthetically but, at the same time, high value natural vanilla flavor has to be extracted from the pods and is a complex mixture of natural products, among which vanillin predominates. There is currently no clear economic benefit from understanding how the vanillin molecule is assembled, since the idea of using such information to engineer the pathway, at least in V. planifolia, goes against the concept of natural vanilla, and synthetic vanillin is so cheap that introducing this molecule alone into other plants as a flavor component also does not make much economic sense. These factors should not, however, be used to argue against supporting research on vanillin biosynthesis. The pathways and mechanisms uncovered could in the future prove critical for the development of more complex bioactives in plants with applications in agriculture, food science, and biomedicine.

REFERENCES