The essential oil of cinnamon is a long-recognised anti-fungal agent ( Myers, 1927) and it is probable that the essential oil is produced to prevent plant infection by fungi. It can be speculated that cinnamic acid was at one time more-commonly produced in plants,

or that moulds such as Penicillium, Aspergillus and Trichoderma spp. have evolved from moulds infecting or recycling cinnamic-acid-producing plants. In a previous publication (Plumridge et al., 2010) we reported that Pad-decarboxylation in A. niger required activity by two genes, padA1 and ohbA1 and this was confirmed in S. cerevisiae ( Mukai et al., 2010). How these gene products interact is not yet known in fungi, but homologues are known to interact in prokaryotes. The E. coli homologues of padA1 and ohbA1 are UBIX and UBID respectively, and both are Afatinib in vitro required in the synthesis of coenzyme Q. Yeast PAD1 can functionally replace ubiX in E. coli ( Gulmezian et al., 2007); yeast deleted for pad1 has normal levels of coenzyme Q but lacks Pad-decarboxylation. The synthesis of coenzyme Q in many bacteria has been shown to require the presence of both Pad1p/UbiXp and Ohb1p/UbiDp ( Rangarajan et al., 2004 and Lupa et al., 2005), clearly showing a level of divergence in function. Whilst it has been reported that some prokaryotes can decarboxylate hydroxybenzoic and hydroxycinnamic

acids ( Cavin et al., 1998), these compounds are not substrates for Pad-decarboxylation

in the fungi studied here. Instead, they are rejected as enzyme substrates and fail to PLX4032 cost induce transcription of the relevant genes. It would appear that the two putative decarboxylases, Pad1p/UbiXp and Ohb1p/UbiDp are both required for coenzyme Q synthesis in bacteria (and are able to act on 2-hydroxybenzoic acid) and that both are required for Pad-decarboxylation in fungi (but not acting on 2-hydroxybenzoic acid and not affecting coenzyme Q). X-ray crystallography studies on the E. coli Pad1p/UbiXp ( Rangarajan et al., 2004) have shown that Pad-type proteins associate into oligomers with a trimer forming the common structural unit although 12-mer assemblies have been predicted for Pad from both E. coli ( Rangarajan et al., 2004; PDB entry found 1sbz) and Aquifex aeolicus (PDB entry 2ejb). The active site of the E. coli enzyme was identified at the interface of the three Pad monomers, although the interaction with Ohb/UbiD was not examined. The role of OhbA1 in fungi is clearly essential in Pad-decarboxylation but does not appear to involve 2-hydroxybenzoic acid decarboxylation. A structural role for the OhbA1 in a hetero-oligomeric OhbA1/PadA1 protein complex may be possible, such as maintaining PadA1 oligomers in the correct conformation, but a physical or mechanistic interaction between PadA1 and OhbA1 remains speculative.