cis or trans with class II diterpene cyclases

Page 3160 Further investigation indicates that our assignment of the product resulting from AgAS:D621A reacting with the alternative substrate (Z,Z,Z)-nerylneryl diphosphate (NNPP, 2), produced by a previously identified NNPP synthase (NNPS),¹ was incorrect. Specifically, upon further reacting this product with the subsequently acting sclareol synthase from Salvia sclarea (SsSS), which promiscuously reacts with a side variety of such precursors, catalyzing heterolytic lysis of the allylic diphosphate ester and addition of water to the ensuing tertiary carbocation to yield a tertiary hydroxyl derivative,^{2, 3} the final product was found to be manool (Figure 1). This was further verified by NMR, with polarized optical spectra demonstrating that this manool was of ‘normal’ (+) configuration (data not shown). Given that manool is produced by SsSS from normal labda-8(17),13E-dienyl diphosphate,² and the configuration of the carbon-carbon double bond at carbon-13 (C13) is lost during the SsSS catalyzed reaction, this result strongly indicates that AgAS:D621A reacts with NNPP to produce normal labda-8(17),13Z-dienyl diphosphate in which the produced decalin bicycle has trans substituents across the C5-C10 bridgehead carbons. Accordingly, Scheme 2 in the published Communication should be replaced with Scheme 1 shown here. We apologize for this erroneous assignment.

Fig. 1.

The combined activity of NNPS+AgAS:D621A+SsSS produces manool. GC-MS chromatograms of (top) an extract from a culture of E. coli engineered to co-express all three enzymes and (bottom) an authentic sample of manool (as indicated).

Scheme 1.

Reaction catalyzed by AgAS:D621A with 2.