Figure 1.
(A) Biosynthetic pathway of hydroxymalonyl-ACP. The final FAD dependent oxidation step catalyzed by ZmaE may proceed through an endiol intermediate (red), resulting in the loss of stereospecificity at C2 of the final product, hydroxymalonyl-ACP. (B) ZMA PKS/NRPS. Nine extender units are utilized to form the precursor of metabolite A (green), zwittermicin A (red), and metabolite B (blue). Hydroxymalonyl-ACP is recognized by ZmaA (dotted line). Each circle represents a catalytic domain of the PKS/NRPS: C, condensation; A, adenylation; PCP, peptide carrier protein; E, epimerization; KS, ketosynthase; AT, acyltransferase; KR, ketoreductase; ACP, acyl carrier protein; Pr, protease; TE, thioesterase. (C) Natural prodrug activation. ZmaL is proposed to catalyze the cleavage of the ZMA precursor molecule from ZmaB-bound alanine, which is further condensed to leucine and methionine to form metabolite B (blue). ZmaM is proposed to catalyze the separation of metabolite A (green) from ZMA (red).
Table 1.
Data Collection and Refinement Statistics.
Figure 2.
The N-terminal KS-AT linker (green), α/β-hydrolase large subdomain (blue), small subdomain (gray), and post-AT linker (red) make up the complete asymmetric unit. The active site of ZmaA-AT (inside solid red box) is bounded on the left by the substrate pocket lid (containing the YASH motif, which in ZmaA-AT is GAAH) and on the top by the RVDVVQ motif (yellow) and is occupied by formate (spheres); cf. Figure 3. Residues E293-G294-A295 are not observed and are indicated with a dashed line. The proposed substrate ACP binding surface M286-E293 contains the methionine residues of the RXR motif (MCM in ZmaAT) (dotted red box) which correspond to the inchoate β-strand of the ferredoxin fold in the smaller subdomain of other ATs; cf. Figure 4.
Figure 3.
YASH and GHSXG motifs of ZmaA-AT compared to a methylmalonyl-CoA specific AT.
The substrate binding-pocket amino acid residues (290–300 and 194) of ZmaA-AT (blue, with white span for disordered 293–295) are superimposed on those of AT from the DEB PKS module 3 (wheat). Bulky F193 is found next to the active site S192 in ZmaA-AT, instead of the glutamine residue found in methylmalonyl-CoA specific ATs. The catalytic H297 is positioned similarly to other ATs, despite its proposed steric hindrance to extender units with (2R) conformations. Despite high mobility for the substrate pocket lid YASH motif, we conclude based on the positions of well-ordered flanking residues that they must wander within the substrate binding pocket of ZmaA-AT, which holds co-crystallized formate (spheres). The red box, with its marked corner, can be compared to the same box in Figure 2 in order to orient the reader.
Table 2.
The GHSXG and the YASH Motifs of Select Acyltransferases are Responsible for ACP vs CoA Discrimination.
Figure 4.
Proposed AT-Domain Interaction with ACP Substrate Carrier.
Approximate protein contact potential calculated using PyMOL vacuum electrostatics function. The colors represent potentials ranging from −70 mV (red) to +70 mV (blue). (A) Proposed AT/ACP interface of ZmaA-AT. FabD was aligned to the structure of ZmaA-AT to show the relative position of CoA (dots, FabD not shown). (B) AT/CoA interface of E. coli FabD (PBD ID: 2G2Z, see Methods [18]). CoA is shown as spheres.
Table 3.
The RXR Motifs of Select Acyltransferases Control Extender Unit Specificity.
Figure 5.
Transacylase assay of ZmaA-AT Distinguishes ACP from Acyl Unit Recognition.
SDS-PAGE of reaction mixtures and corresponding phosphorimage. Lane 1: Molecular mass markers (Prestained Broad-range, Biorad). Lane 2: ZmaA-AT. Lane 3: ZmaD (ACP). Lane 4: Sfp (4′-phosphopantetheinyl transferase). Lane 5: ZmaA-AT, Sfp, and *Malonyl-CoA. Lane 6: ZmaA-AT, Sfp, ZmaD, and *Malonyl-CoA. Lane 7: ZmaA-AT, Sfp, and *(2-RS)-methylmalonyl-CoA. Lane 8: ZmaA-AT, Sfp, ZmaD, and *(2-RS)-methylmalonyl-CoA.
Figure 6.
Proposed movement of the substrate pocket lid induced by ZmaD binding.
(A) Based on the crystal structure of ZmaA-AT, the substrate pocket lid (blue) is shown in the closed position, restricting the entry of the extender unit, in the absence of substrate carrier protein. (B) Model structure of ZmaA-AT bound to substrate carrier protein, ZmaD (blue spheroid). The binding of the substrate carrier protein to the RXR motif (M286-C-M288 in ZmaA-AT; gray sticks) in the small subdomain of ZmaA-AT is proposed to cause the formation of the β-strand (red), resulting in the opening of the substrate pocket lid (blue).
Figure 7.
Possible difference in substrate entry angles between DEB PKS AT-5 and ZmaA-AT.
(A) In DEB PKS-AT5, Q643 has been proposed to orient the incoming (2S)-methylmalonyl-CoA so that Y742 makes a hydrophobic interaction with the methyl-group and H745 sterically hinders the entry of (2R)-methylmalonyl-CoA [26]. (B) In ZmaA-AT, F193 is not positioned to orient the incoming substrate, which may allow hydroxymalonyl-ACP with (2R)-stereochemistry to enter the substrate pocket unhindered.