Reinvestigation of a Cyclic Dipeptide N‐Prenyltransferase Reveals Rearrangement of N‐1 Prenylated Indole Derivatives

In a previous study, a cyclic dipeptide N-prenyltransferase CdpNPT from Aspergillus fumigatus was found to catalyse the prenylation of tryptophan-containing cyclic dipeptides at position N-1 of indole moieties. The major products were identified as derivatives that carried 1’-(3’,3’-dimethylallyl)—or 1’-DMA— moieties, for example, cyclo-N1-(1’-DMA)-l-Trp-l-Trp (1; Scheme 1). The enzymatic reactions were terminated by addition of trichloroacetic acid (TCA) and the reaction mixtures were separated by HPLC under acidic conditions. However, reinvestigation of the reaction mixture of cyclo-l-Trp-l-Trp and dimethylallyl diphosphate (DMAPP) in the presence of CdpNPT showed that the peak area of 1 decreased significantly as the incubation time of the reaction mixture after addition of TCA was shortened (Figures 1A–C). Meanwhile, the area of the peak of 2, which eluted before 1, increased drastically. When the enzymatic reaction was terminated with the same volume of MeOH as that of the reaction mixture, 1 was detected only as a minor peak in the HPLC chromatogram (Figure 1D). Instead, the peak for 2 was found to be dominant. Therefore, 1 seems to be an artefact of an enzymatic product, for example, of 2. To prove this hypothesis, enzymatic reactions of eight cyclic dipeptides, cyclo-l-Trp-l-Trp, cyclo-l-Trp-l-Tyr, cyclo-d-Trp-lTyr, cyclo-l-Trp-l-Phe, cyclo-lTrp-l-Pro, cyclo-d-Trp-l-Pro, cyclo-l-Trp-l-Leu and cyclo-lTrp-Gly, were terminated by addition of TCA (final concentration 136 mm) as used in a previous study. The obtained mixtures had a pH value of 1 and were incubated at room temperature for 2 h. These mixtures were then neutralised to pH 7.0 by addition of NaOH and analysed by HPLC under the new conditions, which lacked acids in the elution solvents. Reaction mixtures terminated with MeOH were used as controls. With the exception of cyclo-d-Trp-l-Pro (Figure 2K), only one dominant product peak was detected for each of the tested substrates after termination with MeOH. In contrast, two or more peaks were found for almost all of the substrates after termination with TCA (Figure 2). These results show that rearrangement of the enzymatic products of CdpNPT had in fact taken place in the presence of TCA. For structural elucidation of the enzymatic products, each of the eight cyclic dipeptides (5 mmol of each) was incubated with DMAPP (5 mmol) and recombinant His6–CdpNPT (0.4 mg) for 16 h, and the mixtures were extracted with ethyl acetate immediately after the incubation period. The conversion rates were found to be from 30 to 70% for these substrates. The enzymatic products were subsequently isolated by HPLC without acids in the elution solvents and subjected to NMR spectroscopy and MS analysis. The H NMR spectroscopy and MS data are Scheme 1. Enzymatic reaction catalysed by CdpNPT and a hypothetical rearrangement mechanism of the product in the presence of acids; cyclo-l-Trp-l-Trp is used as an example.