The native Taxus N-debenzoyl-2<&rsquo>-deoxypaclitaxel:N-benzoyltransferase (NDTBT) from the paclitaxel pathway transfers a benzoyl group from the corresponding CoA thioester to the amino group of the <&beta>-phenylalanine side chain of N-debenzoyl-2<&rsquo>-deoxypaclitaxel. To elucidate the substrate specificity of NDTBT with a His6-fusion epitope was overproduced in Escherichia coli. The purified enzyme was incubated with semisynthetically derived N-debenzoyltaxoid substrates and aroyl CoA donors (benzoyl; ortho-, meta-, and para-substituted benzoyls; various heterole carbonyls; alkanoyls; and butenoyl). Several unnatural N-aroyl-N-debenzoyl-2<&rsquo>-deoxypaclitaxel analogues were biocatalytically assembled with catalytic efficiencies (Vmax/KM) ranging between 0.15 and 1.74 nmol/min/mM. In addition, several N-acyl-N-debenzoylpaclitaxel variants were biosynthesized when N-debenzoylpaclitaxel and N-de(tert-butoxycarbonyl) docetaxel (i.e., 10-deacetyl-N-debenzoylpaclitaxel) were used as substrates. The relative velocity (vrel) for NDTBT with the latter two N-debenzoyl taxane substrates ranged between ~1% and 200% for the array of aroyl CoA thioesters compared to benzoyl CoA. Interestingly, NDTBT transferred hexanoyl, acetyl, and butyryl moieties more rapidly than butenoyl or benzoyl groups from the CoA donor to taxanes with isoserinoyl side chains, whereas N-debenzoyl-2<&rsquo>-deoxypaclitaxel was more rapidly converted to its N-benzoyl derivative than to its N-alkanoyl or N-butenoyl congeners. Biocatalytic N-acyl transfer of novel acyl groups to the amino functional group of N-debenzoylpaclitaxel and its 2<&rsquo>-deoxy precursor reveal the surprisingly indiscriminate specificity of this transferase. This feature of NDTBT potentially provides a tool for alternative biocatalytic N-aroylation/alkanoylation to construct next generation taxanes or other novel bioactive diterpene compounds.To elucidate the basis of the broad substrate specificity and to direct mutational analysis, the crystal structure for NDTBT was sought. Therefore, several chromatographic methods were applied to increase the purity of the protein from 70%, obtained in previous attempts, to >95%, without losing activity. This sequence of chromatographic steps was also used to purify 10-deacetylbaccatin III: 10<&beta>-O-acetyltransferase (DBAT) and a modified taxane-2<&alpha>-O-benzoyltransferase (mTBT) to 90-95% purity. Preliminary screens for crystals of purified NDTBT (at 5 mg/mL) with and without benzoyl CoA in a matrix of crystallization buffers were negative. The histidine tag was hypothesized to inhibit crystallization and thus is currently being removed; attempts to cleavage of the His6-tag from NDTBT by thrombin-catalyzed scission yields only ~10% of the liberated protein.Meanwhile, a mutagenesis study was constructed to examine potential structural components of the Taxus acyltransferase family responsible for substrate selectivity. Regional and domain swapping between pairs of enzymes: NDTBT/DBAT and DBAT/mTBT were employed to evaluate these structural hypotheses. Vinorine synthase and anthocyanin malonyltransferase, homologousenzymes in the same BAHD family whose structures have been solved, were used as models to identify mutagenic sites for constructing the chimeras of the Taxusenzymes. The so-derived mutant chimeras were purified from a nickel affinity column at 0.24 mg/L yield. The mutants were incubated with the parent substrates and none of the mutants made products detectable by LC-ESI/MS analysis. The chimeras are potentially inactive because incorrect splicing sites were perhaps unexpectedly identified to construct the mutant enzymes.
Read
