You are on page 1of 19

View Online

REVIEW

www.rsc.org/npr | Natural Product Reports

Structural biology in plant natural product biosynthesisarchitecture of enzymes from monoterpenoid indole and tropane alkaloid biosynthesis
Joachim St ockigt*a ,b and Santosh Panjikarc
Received (in Cambridge, UK) 2nd August 2007 First published as an Advance Article on the web 24th October 2007 DOI: 10.1039/b711935f Covering: 1997 to 2007
Downloaded on 24 July 2011 Published on 24 October 2007 on http://pubs.rsc.org | doi:10.1039/B711935F

Several cDNAs of enzymes catalyzing biosynthetic pathways of plant-derived alkaloids have recently been heterologously expressed, and the production of appropriate enzymes from ajmaline and tropane alkaloid biosynthesis in bacteria allows their crystallization. This review describes the architecture of these enzymes with and without their ligands. 1 2 2.1 2.2 2.3 2.4 2.4 2.6 3 3.1 3.2 3.3 3.4 3.5 3.6 4 4.1 4.2 4.3 4.4 5 5.1 5.2 5.3 5.4 Introduction Strictosidine synthase (STR1) Biological PictetSpengler reaction and the role of STR1 for the entire indole alkaloid family Crystallization and structure determination of STR1, and the Auto-Rickshaw software pipeline The 6-bladed 4-stranded b-propeller fold of STR1the rst example from the plant kingdom Structures of STR1 in complex with its substrates and product Substrate specicity of STR1 and mutagenesis studies Structure-based alignments and evolution of STR1 Strictosidine glucosidase (SG) and its role Optimization of crystallization of SG Overall 3D-structure of SG Site-directed mutagenesis and SG Glu207Gln-strictosidine complex Recognition of the aglycone and glycone part of strictosidine Substrate binding site of SG compared to those of other plant glucosidases Prospects for the use of SG for alkaloid synthesis Polyneuridine aldehyde esterase (PNAE) PNAEone of the most substrate-specic esterases? Inhibition, primary structure and site-directed mutagenesis of PNAE The long route to PNAE crystals Homology modelling and structure of PNAE Vinorine synthase (VS) The role of VS in generating the ajmalan structure Inhibitors and mutants of VS Unusual crystallization conditions and instability of VS crystals Structure elucidation of VS 5.5 Important amino acids and mechanistic aspects 5.6 Signicance of the vinorine synthase structure for the BAHD enzyme family 6 Raucaffricine glucosidase (RG), a side route of the ajmaline pathway 6.1 Crystallization attempts and the 3D-structure of RG 7 Structure-based redesign of enzymes and applications in chemo-enzymatic approaches 8 Other structural examples from the alkaloid eld 9 Conclusions and future aspects of structural biology in the alkaloid eld 10 Acknowledgements 11 References

1 Introduction
Both their highly complex chemical structures and their pronounced pharmacological activities have made research on alkaloids, including elucidation of their biosynthetic pathways, very attractive for many decades. Major progress in the research into the biosynthesis of plant monoterpenoid indole alkaloids has been made by investigations of single enzymes, partial biosynthetic routes and entire pathways.13 This has been sporadically reviewed during the last 10 years together with the genetics of alkaloid biosynthesis.4,5 There are, however, only very few examples from the alkaloid eld, which deliver at present a coherent knowledge of biosynthetic routes at the enzyme level. These include: (a) The well investigated enzymatic steps connecting the Aspidosperma alkaloids tabersonine and vindoline, reactions occurring at the periphery of the basic Aspidosperma alkaloid skeleton.4,610 (b) The biosynthesis of camptothecin, our knowledge of which, despite some recent enzymatic work on the early steps,11 comes largely from Hutchinsons group, a long time ago.12,13 (c) The whole sequence between tryptamine plus secologanin and the heteroyohimbines ajmalicine and some of its isomers (Corynanthe alkaloids) was successfully investigated at the enzymatic level a few decades ago.1417 (d) Efcient alkaloid-producing cell suspension cultures of the Apocynaceae plant Catharanthus roseus, established at the This journal is The Royal Society of Chemistry 2007

a College of Pharmaceutical Sciences, Zijingang Campus, Zhejiang University, 310058, Hangzhou, China b Institute of Pharmacy, Johannes Gutenberg University Mainz, Staudinger Weg 5, D-55099, Mainz, Germany c European Molecular Biology Laboratory Hamburg, Outstation Deutsches Elektronen-Synchrotron, Notkestrasse 85, D-22603, Hamburg, Germany In memory of Professor Pierre Potier.

1382 | Nat. Prod. Rep., 2007, 24, 13821400

View Online

Joachim St ockigt received a PhD in organic chemistry from M unster University (Germany) with Professor Burchard Franck. He has worked with Professor Meinhart H. Zenk at the Faculty of Biology (Bochum University, Germany) and the Faculty of Pharmacy (Munich University, Germany) and is currently Full Professor at the Institute of Pharmacy (Mainz University, Germany; College of Pharmaceutical Sciences, Zhejiang University, Hangzhou, China). His research interests include natural products biosynthesis (phytochemistry, enzymology, molecular and structural biology). Santosh Panjikar earned a PhD in Biotechnology from Friedrich-Schiller University (Jena, Germany) and held a postdoctoral appointment at the EMBL Hamburg Outstation with Dr Paul A. Tucker. He is currently Staff Scientist at the EMBL Outstation. His research interests focus on structure-based drug design, method developments in structural biology and synchrotron instrumentation.

Downloaded on 24 July 2011 Published on 24 October 2007 on http://pubs.rsc.org | doi:10.1039/B711935F

Joachim St ockigt

Santosh Panjikar

beginning of the 1970s in Zenks laboratories, became the key for studies of this plant.18,19 (e) Cell suspension cultures of the Indian medicinal plant Rauvola serpentina Benth. ex Kurz from the same lab were also the major prerequisite for the isolation of some uncommon indole alkaloids20 and elucidation of the pathway for the formation of the antiarrhythmic ajmaline and structurally related alkaloids of the sarpagan- and ajmalan-type both by identication of many single enzymes2123 and in part by in vivo NMR.24,25 The number of enzyme-catalyzed reactions proven in that Rauvola cell system amounts to more than 15, representing at the moment one of the most detailed investigations in alkaloid biosynthesis. The pathway between tryptamine plus secologanin and ajmaline is illustrated in Scheme 1. Several of the cDNAs coding for enzymes of the abovementioned ajmaline route have been detected, isolated and functionally expressed, mostly by using the reverse genetic approach, applying polymerase chain reaction (PCR) and heterologous systems, especially Escherichia coli. For instance, most of the soluble Rauvola enzymes are now functionally expressed and characterized in detail,23 especially with regard to their substrate specicity and their major kinetic data and general properties, such as temperature and pH optimum, molecular size, and isoelectric points. For the rst time, cloning has also provided amounts of enzyme, in almost pure form, on the 2050 mg scale by fusionprotein techniques. This is the amount necessary to develop crystallization conditions of these proteins and to elucidate their three-dimensional structure by X-ray analysis. The following article summarizes in detail the achievements in the elucidation of the architecture of ve major enzymes of the ajmaline biosynthetic pathway in Rauvola,26 namely strictosidine synthase (STR1), strictosidine-O-b-D-glucosidase (SG), polyneuridine aldehyde esterase (PNAE), vinorine synthase (VS) and raucaffricine glucosidase (RG), together with two tropinone reductases (TR-I and TR-II) from tropane alkaloid biosynthesis.27,28 Other than a short general overview on 3DThis journal is The Royal Society of Chemistry 2007

analysis of Rauvola enzymes,26 this is the rst comprehensive and detailed review on structural biology in the alkaloid eld, providing a more direct insight into the mechanisms of alkaloid biosynthesis.

2 Strictosidine synthase (STR1)


2.1 Biological PictetSpengler reaction and the role of STR1 for the entire indole alkaloid family Strictosidine synthase (STR1, EC 4.3.3.2), which catalyzes the stereoselective PictetSpengler reaction of tryptamine and secologanin to form 3a(S)-strictosidine, has been described many times since its original detection.29,30,3134 It was the rst enzyme from alkaloid biosynthesis whose cDNA was functionally, heterologously expressed.3537 Later, the synthases from several plants other than Rauvola,35 such as Catharanthus38,39 and Ophiorhiza,40 were functionally expressed. The primary importance of STR1 is not only its precursor role for the biosynthetic pathway of ajmaline, but also because it initiates, in fact, all pathways leading to the entire monoterpene indole alkaloid family. Some prominent members are depicted in Fig. 1. Although STR1 is the crucial enzyme for all the 2000 monoterpenoid indole alkaloids, little information about its reaction mechanism and the critical amino acids for enzyme activity was known. 2.2 Crystallization and structure determination of STR1, and the Auto-Rickshaw software pipeline Expression and crystallization of STR1 was straightforward and typical of our crystallization of other Rauvola proteins, using pQE-2-plasmid and M15 E. coli cell line. The hanging-drop vapour-diffusion technique was used routinely and was the most efcient method in our hands. STR1 crystals were in space group R3 with a hexagonal unit cell containing two molecules of the enzyme. A set of selenium-labelled STR1s (4SeMet- and Nat. Prod. Rep., 2007, 24, 13821400 | 1383

View Online

Downloaded on 24 July 2011 Published on 24 October 2007 on http://pubs.rsc.org | doi:10.1039/B711935F

Scheme 1 Enzyme-catalysed biosynthesis of ajmaline in cell suspension culture of the medicinal plant Rauvola serpentina (L.) Benth. ex Kurz. Abbreviations are: STR1 = strictosidine synthase, SG = strictosidine glucosidase, SBE = sarpagan bridge enzyme, PNAE = polyneuridine aldehyde esterase, VS = vinorine synthase, VH = vinorine hydroxylase, CPR = cytochrome, P450 reductase, VR = vomilenine reductase, DHVR = dihydrovomilenine reductase, AAE = acetylajmalan esterase, NAMT = norajmalan methyltransferase, RG = raucaffricine glucosidine, VGT = vomilenine glucosyltransferase. Strictosidine aglycone has not been isolated. Enzymes shown in bold have been heterologously expressed, those marked with asterisks have been crystallised and their 3D-structures determined recently in our laboratories. They are described in this review.

Fig. 1 Formation of strictosidine and its biosynthetic role as central precursor of various monoterpenoid indole alkaloids.

6SeMet-STR1) needed to be prepared in order to solve the enzyme structure, because no protein structures with signicant sequence homology to strictosidine synthase were available.41,42 The STR1
1384 | Nat. Prod. Rep., 2007, 24, 13821400

structure was nally solved using the multiple wavelength anomalous dispersion (MAD) approach together with software package . Auto-Rickshaw,43 and rened to 2.4 A Structures of all enzymes from the ajmaline biosynthetic pathway discussed in this article were determined using AutoRickshaw. In fact, early evaluation of the software pipeline was performed on several multiple wavelength diffraction (MAD) datasets from crystals of various proteins, including MAD data of vinorine synthase (VS) and strictosidine synthase (STR1). At a later stage, native structures or ligand complexes of the various enzymes from the pathway were determined routinely using AutoRickshaw. Here we describe briey the software pipeline which provides a means of rapid structure solution of proteins. Crystal structure determination both by isomorphous replacement and by anomalous scattering techniques is a multi-step process in which each step, from substructure determination to model building and validation, requires certain decisions to be made. These decisions comprise the choice of the crystallographic computer programs that are most suitable to perform the specic tasks and the optimal input parameters for each of these programs. The interpretability of the map depends to a large extent on the success of the preceding steps and is generally limited by the resolution of the data and the quality of the phase information. Traditionally, each of the steps described was carried out by an experienced crystallographer, whose skill manifested itself in nding This journal is The Royal Society of Chemistry 2007

View Online

Downloaded on 24 July 2011 Published on 24 October 2007 on http://pubs.rsc.org | doi:10.1039/B711935F

the optimum, or at least a successful, path towards the completion of the structure determination. The automated crystal structure determination platform combines a number of macromolecular crystallographic software packages with several decision-making steps. The entire process in the pipeline is fully automatic. Each step of the structure solution is governed by the decision making module within the platform, which attempts to mimic the decisions of an experienced crystallographer. The role of computer-coded decision-makers is to choose the appropriate crystallographic computer programs and the required input parameters at each step of the structure determination. Various phasing protocols are encoded in the system. These are single anomalous diffraction (SAD), single isomorphous replacement with anomalous scattering (SIRAS), multiple wavelength diffraction (MAD), standard molecular replacement (MR), phased MR and combination of MR and SAD. A large number of possible structure solution paths for each phasing protocol are encoded in the system, and the optimal path is selected by the decision-makers as the structure solution evolves. Once the input parameters are given (number of amino acids, heavy atoms, molecules per asymmetric unit, probable space group and phasing protocol) and X-ray data have been input to AutoRickshaw, no further user intervention is required. It proceeds step by step through the structure solution using the decision makers. The Auto-Rickshaw server (http://www.embl-hamburg.de/AutoRickshaw) is available to EMBL-Hamburg beamline users, and the server will be made more widely accessible in the immediate future. 2.3 The 6-bladed 4-stranded b-propeller fold of STR1the rst example from the plant kingdom The overall structure of STR1 belongs to the six-bladed fourstranded b-propeller fold, where the blades are radially located around a pseudo-six-fold symmetry axis. Each blade consists of twisted four-stranded b-sheets. Although this particular fold has been detected in different organisms several times, this is the rst example from the plant kingdom. However, the other enzymes with the six-bladed fold have a completely different function.44,45 The active site of STR1 is near to the symmetry axis, as shown by the structure of the enzymetryptamine complex (Fig. 2).46 There are only three helices in the STR1 structure, of which two are connected by an SS bridge, which is essential for both the overall structure and the shape of the catalytic centre. 2.4 Structures of STR1 in complex with its substrates and product Tryptamine is located deep in the binding pocket, sandwiched between two hydrophobic residues (Tyr151 and Phe226) holding the molecule in the correct orientation for the PictetSpengler condensation. In addition, Glu309, which is important for the STR1 activity as proven by mutagenesis experiments, is coordinated with the amino group of tryptamine. There are (primarily) hydrophobic amino acids lining the binding pocket, and both positively charged and hydrophobic residues are located at the entrance to the binding site of STR1. The complex structure of STR1 bound with secologanin exhibits the monoterpene in the same binding pocket (Fig. 3a). The This journal is The Royal Society of Chemistry 2007

Fig. 2 Strictosidine synthase (STR1) from Rauvola in complex with its substrate tryptamine (N and C mark the N- and C-termini of the protein; the arrow points to the SS bridge).

hydrophilic glucose unit of secologanin points out of the catalytic pocket towards the solvent. This complex is illustrated in Fig. 3b, which is rotated by 84 along the x-axis to show better the accessibility of the hydrophilic part of the monoterpene to the solvent. The aldehyde group of secologanin points towards Glu309 and is in close proximity to the amino group of tryptamine (as observed in the previous structure), ready for the primary condensation reaction. The third complex was prepared by soaking crystals of STR1 in a solution containing strictosidine, and the structure illustrates how the product of the STR1 reaction is accommodated in the catalytic centre (Fig. 4). The location of the strictosidine is very similar to that found for both substrates (see above) but is not totally identical, showing that the way the ligands are accommodated is to some extent exible before and after the reaction. The electron density gives additional information about how the product is shielded from the residues of the active centre. Such shielding might have important implications for the substrate specicity of STR1 and should help in structure-based rational design of substrate acceptance. 2.4 Substrate specicity of STR1 and mutagenesis studies As far as the substrate specicity of STR1 is concerned, several studies with both STRs from Catharanthus and Rauvola have been reported earlier.31,4651 We have analyzed the substrate acceptance of STR1 again and focused especially on explanations of why particular tryptamine derivatives were not accepted by STR1 based on the available structural information. Table 1 highlights the comparison of substrate specicity of STR1, especially for tryptamines substituted differently at position 5. These tryptamines are not accepted in the case of 5-methyl- and 5-methoxy-derivatives by the wild-type enzyme.52 The complex structure demonstrates that the Val208 side-chain is shielding the Nat. Prod. Rep., 2007, 24, 13821400 | 1385

View Online

Downloaded on 24 July 2011 Published on 24 October 2007 on http://pubs.rsc.org | doi:10.1039/B711935F

Fig. 4 3D-structure of STR1 in complex with its product strictosidine. Table 1 Substrate acceptance of STR1 for differently substituted tryptamines at positions 5 and 6 (n.d. = not detectable, data taken in part from ref. 52) Substrates Tryptamine 5-Methyltryptamine 5-Methoxytryptamine 6-Methyltryptamine 6-Methoxytryptamine 5-Fluorotryptamine Fig. 3 (a) Strictosidine synthase (STR1) in complex with its substrate secologanin. (b) Fig. 3a rotated by 84 around the x-axis. 6-Fluorotryptamine 5-Hydroxytryptamine Enzymes Wild-type Val208Ala Wild-type Val208Ala Wild-type Val208Ala Wild-type Val208Ala Wild-type Val208Ala Wild-type Val208Ala Wild-type Val208Ala Wild-type Val208Ala K m /mM 0.072 0.219 n.d. 0.281 n.d. 3.592 0.393 0.762 0.962 0.307 0.259 1.302 0.136 0.356 2.255 0.844 (kcat /K m )/mM1 s1 147.92 246.99 23.35 22.18 5.90 14.37 5.53 54.27 144.63 16.31 171.84 38.29 249.29 21.47

5-position, preventing binding of 5-substituted indole bases. Replacement of this particular Val by the smaller Ala clearly broadens the substrate acceptance for 5-substituted tryptamines, delivering novel strictosidine analogues.52 Further structure-derived mutations based on this structure are expected to have signicant impact on future enzymatic synthesis of new strictosidines structurally modied at the tryptamine or secologanin parts. Moreover, if the enzyme strictosidine glucosidase (SG), which follows STR1 in the biosynthesis of all the monoterpene alkaloids, can be modied in a similar way to that described for the synthase, future chemo-enzymatic approaches might be developed for the generation of novel alkaloid libraries with biologically important compounds. 2.6 Structure-based alignments and evolution of STR1

STR1 has no functional homologies to other six-bladed bpropeller folds. Also, a structurally based sequence alignment of
1386 | Nat. Prod. Rep., 2007, 24, 13821400

these folds showed less than 16% sequence identity with most bpropellers such as diisopropyluorophosphatase (DFPase)45 from Loligo vulgaris, brain tumour NHL domain,53 serum paroxonase44 and low-density lipoprotein receptor YWTD domain. Analysis of the STR1 fold compared to the just-mentioned folds indicates that these structurally related proteins are also related from the point of view of their evolution.46 Most probably they all evolved from an ancestral b-sheet gene. The ancestral structure might have been modied by e.g. deletions or insertions during evolution, leading to a differently shaped reaction pocket in order to adopt the different functions of these propellers. These propellers keep a common sequence homology which maintains the overall structural features of the protein fold. It is typical of these six-bladed b-propellers that they are diverse in sequence and function and that they are of different phylogenetic origin. This journal is The Royal Society of Chemistry 2007

View Online

Downloaded on 24 July 2011 Published on 24 October 2007 on http://pubs.rsc.org | doi:10.1039/B711935F

A similar reaction to that of STR1 found in the plant Alangium lamarckii Thw. (Alangiaceae) is catalyzed by deacetylipecoside synthase (DIS). This synthase catalyzes the condensation between dopamine and secologanin, forming (1R)-deacetylipecoside which has the opposite conguration to strictosidine (at C-3).54 DIS and STR1 are similar with respect to their reaction type (Pictet Spengler condensation), their pH optimum, temperature optimum and their molecular size. Both enzymes exhibit high substrate specicity but they come from different plant species and families, and exhibit different substrate specicity and stereo-selectivity. It cannot be excluded, however, that both synthases evolved from the same ancestor but during evolution developed different substrate specicities leading to different alkaloid types. Cloning of DIS and comparison of the 3D-structures of both enzymes could help to support this hypothesis.

needle-shaped crystals (Fig. 5a), which were not useful for Xray experiments. However, in a large trial using a Cartesian syn QUADTM Microsys crystallization robot with 600 nl sitting drops, two conditions were found to produce rectangular crystal plates. These conditions gave, after extensive optimization with the hanging drop method, improved crystals for X-ray mea (Fig. 5b).63 surements with a nal resolution of 2.48 A

Strictosidine glucosidase (SG) and its role

Similarly to STR1, strictosidine glucosidase (SG, EC 3.2.1.105) has been described several times from Catharanthus and from Rauvola.5557 Here we refer mainly to the Rauvola enzyme because it was the second protein from the ajmaline biosynthetic pathway to be crystallized. The general and major importance of SG lies in its function to chemically activate strictosidine by deglucosylation. The aglycone thus generated is highly unstable and reactive. It enters all the pathways to the various structural types illustrated in Fig. 1, but the exact molecular stage at which it initiates all these pathways has not been claried so far. Although glycosides are involved in many metabolic processes such as regulation of plant hormone activity and biosynthesis58,59 and lignication,60 one of the main functions of glycosidases, especially in higher plants, is believed to be defence-related.61 For example, the generation by glycosidases of toxic products such as cyanide or isothiocyanates from the appropriate glycosides makes their ecological signicance as defensive agents immediately understandable.62 SG has also been discussed as being defencerelated since its product(s) were shown to have antibiotic activity.56 However, in view of the important role that the aglycone of strictosidine plays as the biogenetic precursor of all of the 2000 or so monoterpenoid indole alkaloids, a synthesis-related function of SG seems to be more important than its defence-related role. Reasons for the pronounced substrate specicity and details of the reaction mechanism of this enzyme remained mostly unknown until recently. Therefore, an efcient over-expression and purication system for strictosidine glucosidase in Escherichia coli was developed, followed by crystallization and preliminary X-ray analysis of the enzyme. 3.1 Optimization of crystallization of SG

Fig. 5 (a) Needle-type crystals of SG not suitable for X-ray analysis. (b) A rectangular prism of SG (images are from ref. 26).

3.2 Overall 3D-structure of SG From initial alignment studies SG unequivocally belongs to the glycosyl hydrolase (GH) family 1, which is grouped together with 16 other families in clan GH-A, representing the biggest of the 13 clans of glycosidases. There are only seven three-dimensional structures of glucosidases from family 1 known from eukaryotic sources, and six of these are of plant origin. The expression, purication, crystallization and preliminary X-ray analysis of SG has recently been reported.63 In Fig. 6 the overall fold of SG

When our routine expression system (E. coli, M15 [pREP4]) and purication method (fusion protein with N-terminal His6 tag), followed by removing the tag with dipeptidyl aminopeptidase (DAPase) and nal enzyme purication by MonoQ ion exchange chromatography was used, about 10 mg of pure SG were obtained from a 5 litre culture of E. coli. Such an amount of enzyme is acceptable for a large screening of crystallization conditions. More than 10 commercially available crystallization kits were tried. Many conditions resulted in formation of rather at- or This journal is The Royal Society of Chemistry 2007

Fig. 6

). Tim barrel overall fold of SG (2.48 A

Nat. Prod. Rep., 2007, 24, 13821400 | 1387

View Online

is shown, representing the well known (b/a)8 barrel fold. The structure is built from 13 a-helices and 13 b-strands; the core of the structure consists of 8 parallel b-strands that form a b-barrel. This barrel is surrounded by eight helices and hosts the catalytic binding site of SG for the substrate strictosidine. 3.3 Site-directed mutagenesis and SG Glu207Gln-strictosidine complex Complete conservation of amino acid residues Glu207, Glu416 and His161 is indicated by sequence alignment of SG with glucosidases of various origins. Glu207 is the proton donor that assists nucleophilic attack of Glu416 at the anomeric carbon C-1.64 Hydrolysis of the glucoside bond is catalyzed in concert by both glutamic acids.65 For SG no enzyme activity was obtained after mutations Glu207Gln, Glu207Asp, Glu416Gln or Glu416Asp.66 This clearly points to the crucial role of both glutamates for the deglucosylation of strictosidine. In the reaction catalyzed by SG, His161 also plays an important role. When it is replaced by Asn or Leu, enzyme activity is decreased to <0.1%. These results were the basis for the assumption that the three mentioned residues must be involved in the reaction mechanism of strictosidine deglucosylation. For this reason, they should be located near the binding pocket or even belong directly to the catalytic centre of SG, which is, in fact, what is shown by the 3D-structure. The structure of the complex of inactive mutant SG-Glu207Gln with strictosidine was generated (Fig. 7) to obtain additional information on the role of other individual amino acids of the binding pocket.66 Since there was no detectable catalytic activity of the Glu207Gln mutant, soaking was possible with the natural substrate strictosidine and resulted in crystals exhibiting a .66 resolution of 2.82 A

catalytic centre of SG in the substrate binding state. The binding pocket is located at the top of the barrel. The base of the pocket contains several charged residues whereas the gate to the pocket is hydrophobic in nature. The glycosidic part of strictosidine is located deep in the catalytic pocket, and the indole part points towards the solvent even though this part of the substrate is strongly hydrophobic. The aglycone part is surrounded by mostly hydrophobic residues. The sugar unit of strictosidine interacts with many hydrophilic amino acids. His161 is not a catalytic amino acidit is located too far away from the anomeric C-atom of the glucose unitbut it helps to bind strictosidine in the correct orientation for the deglucosylation process.

Downloaded on 24 July 2011 Published on 24 October 2007 on http://pubs.rsc.org | doi:10.1039/B711935F

3.5 Substrate binding site of SG compared to those of other plant glucosidases The structure of SG-Glu207Gln from R. serpentina in complex with strictosidine can be compared with the structures of two other plant glucosidasesubstrate complexes, ZmGlu1-Glu191Asp from Zea mays in complex with DIMBOA-b-D-glucoside (PDB code 1E56), and Dhr1-Glu189Asp from Sorghum bicolor in complex with dhurrin, (PDB code 1V03) (Fig. 8).6668 Ten out of the eleven amino acids that participate in the binding of the glucose unit are conserved in the sequences of the glucosidases in family 1. In some glucosidases only, Tyr481 is replaced by Phe, indicating the pronounced conservation which seems to keep the different glycosylated compounds in the correct location for the hydrolysis reaction. The aglycone binding site of family 1 members is much more variable. The most striking difference in their active sites is observed for the amino acid Trp388. This residue is conserved within the family 1 and its role in substrate recognition has been discussed previously.69 Tryptophan shows in all known plant glucosidases its v-angle (the angle about the side chain CaCb bond) of 60 . That is obviously the typical conformation for plant glucosidases. In SG, this v-angle is changed to 180 . Such a change in conformation results in more space within the binding pocket of SG. This additional space is needed to accommodate the large substrate strictosidine, compared to the smaller substrates of the other b-glucosidases. Moreover, Trp388 forms a stacking interaction with the indole part of strictosidine. Due to the conformational change of Trp388, some amino acids (Gly386, Met275, Thr210 and Met297) can better interact with strictosidine. Together with Phe221 and Trp388, all these amino acids form the substrate binding site of SG for the indolic part of the substrate strictosidine. The specic conformation of Trp388 of SG has been until now an exception amongst b-glucosidases. Structural comparison with fteen glucosidases revealed three different conformations for this amino acid: (a) Trp points directly into the binding pocket, which is found for glycosidases from bacteria and archaea; (b) Trp points in the direction of the N-terminus (which is the plant-typical conformation); (c) Trp points in the direction of the C-terminus, which is the SG Trp conformation.66 All plant-derived glucosidases are folded in a similar way and the major difference is the conformation of Trp388. If additional examples of this new Trp388 conformation are discovered in the future, b-glucosidases might be sub-divided based on this feature. This journal is The Royal Society of Chemistry 2007

Fig. 7 Complex structure of inactive mutant SG-Glu207Gln with strictosidine.

3.4

Recognition of the aglycone and glycone part of strictosidine

For both the glycone part and aglycone part of the substrate molecule, the electron density allowed detailed description of the
1388 | Nat. Prod. Rep., 2007, 24, 13821400

View Online

Downloaded on 24 July 2011 Published on 24 October 2007 on http://pubs.rsc.org | doi:10.1039/B711935F

on detailed biosynthetic knowledge, on rational molecular engineering of the enzymes, their synthetic application and metabolic engineering of the medicinal plants.70,71 Because of the special and multiple role of the aglycone of strictosidine, the enzyme SG may serve as an attractive candidate for such directed biosynthesis approaches. Such approaches only make sense if the substrate specicity of SG can be modied. Because the three-dimensional structure of strictosidine synthase has become available, initial steps were successful to engineer and to change the substrate acceptance of STR1.52 Generation of large alkaloid libraries by structure-based enzyme re-design of STR1 and SG in combination with biomimetic approaches may therefore become possible in the near future.47,51,52

4 Polyneuridine aldehyde esterase (PNAE)


4.1 PNAEone of the most substrate-specic esterases? Polyneuridine aldehyde (PNA) is the rst alkaloid in the pathway to ajmaline, exhibiting a sarpagan structure.72 The aldehyde is biosynthetically generated by the so-called sarpagan-bridge enzyme (SBE),73 whose mechanism has not been elucidated so far. The esterase PNAE (EC 3.1.1.78) converts PNA after ester hydrolysis and decarboxylation into 16-epi-vellosimine.71,74 The substrate specicity of this enzyme has been investigated several times at various stages of enzyme purity,72,74,75 and was nally determined with the overexpressed homogenous His6 -tag PNAE. From a series of 13 structurally different esters only the natural substrate (PNA) was hydrolyzed (Fig. 9).7678 The esters included ten methyl esters, of which eight were monoterpenoid indole alkaloids and two more simple methyl esters, and additionally three acetate esters. The results demonstrated an exceptionally high substrate specicity of the over-expressed enzyme and indicated that PNAE might be one of the most substrate-specic esterases. 4.2 Inhibition, primary structure and site-directed mutagenesis of PNAE Inhibition studies with selective serine or cysteine inhibitors and unselective histidine/cysteine inhibitors provided insight into reactive residues of the enzyme (Table 2). They showed complete inhibition with diethyl pyrocarbonate, indicating the importance of histidine for enzyme activity, and partial inhibition was observed with serine/cysteine inhibitors. However, these experiments did not allow PNAE to be dened as a serine or cysteine hydrolase. Site-directed mutagenesis was applied to get more detailed information on particular amino acids that seemed likely to be involved in the catalytic process of PNAE from a rigorous analysis of the primary structure of the enzyme. Comparison of

Fig. 8 Comparison of the substrate complex of SG (a) with those of two other b-glucosidases, from Zea mays (b) and from Sorghum bicolor (c),67,68 displaying the different conformation of tryptophan (W376) in SG. This illustration was partly taken from ref. 66 (hydrophobic residues in grey, hydrophilic in green, acidic residues (Glu, Asp) in red, positively charged residues in blue).

3.6

Prospects for the use of SG for alkaloid synthesis

A future demand for natural products, in this case of complex plant alkaloids with high structural diversity, will probably depend
Table 2 Inhibition of PNAE by serine, cysteine, and histidine inhibitors Inhibitor AEBSF E-64 TPCK PMSF DEPC Hg2+ Type of Inhibition Selective Ser Selective Cys Ser-Cys Ser-Cys Unselective His Unselective Cys Concentration 4.0 mM 25 lM 200 lM 1.0 mM 1.2 mM 200 lM

Incubation time/min 60 60 60 60 60 60

Relative inhibition (%) 0 0 12 20 100 100

This journal is The Royal Society of Chemistry 2007

Nat. Prod. Rep., 2007, 24, 13821400 | 1389

View Online

Downloaded on 24 July 2011 Published on 24 October 2007 on http://pubs.rsc.org | doi:10.1039/B711935F

Fig. 9

Structurally different esters as putative substrates of PNAE. Only the natural substrate, polyneuridine aldehyde, displayed any activity.78

the overall amino acid sequence of PNAE showed relatively high identity (up to 50%) to some putative lyases e.g. from Arabidopsis thaliana.79 A 4043% identity was found for two well characterized hydroxynitrile lyases (Hnls), one from Hevea brasiliensis (HbHnl) and one from Manihot esculenta (MeHnl)8087 with known 3Dstructure, demonstrating a relatively close relationship of PNAE to Hnls (Scheme 2). The above enzymes belong to the a/b-fold hydrolases, indicating that PNAE is a newly-detected member of that enzyme superfamily. The most important features of that family are a catalytic triad formed by a nucleophilic residue, an acidic amino acid and a histidine, which appear in the same order in PNAE (Ser87, Asp216, His244), and two additional strongly related and conserved motifs. Indeed, mutations in PNAE showed

the presence of the catalytic triad, since replacement of each of the three residues by alanine resulted in completely inactive mutants,76 classifying the enzyme as a novel a/b-fold hydrolase. Structural analysis of PNAE (see Section 4.4) supported that classication. 4.3 The long route to PNAE crystals The key to the crystallization of an enzyme is its purity. Crystallization also depends on the availability of homogenous, ideally monodisperse solutions of protein molecules with a uniform, rigid three-dimensional fold. Protein crystallization can be divided into two steps: coarse screening to identify initial crystallization conditions and then optimization of these conditions to produce

Scheme 2 Alignment of the primary sequence of PNAE with hydroxynitrile lyases, showing the close relationship of the three enzymes, the conserved catalytic triad (marked with asterisks) and two typical motifs (black on red) around His17 and Ser87.

1390 | Nat. Prod. Rep., 2007, 24, 13821400

This journal is The Royal Society of Chemistry 2007

View Online

single, diffraction-quality crystals. Currently, there are no systematic methods to ensure that ordered three-dimensional crystals will be obtained.88 Sitting-drop and hanging-drop methods of vapour diffusion are the most commonly used techniques. Microbatch crystallization under oil is mainly used when these crystallization methods have failed.89 A recent development in protein crystallization has been the use of microuidic systems for crystallizing proteins using the free-interface diffusion method on the nanolitre scale, but this has one major drawback in that it is often difcult to translate hits into higher volume solutions in order to grow crystals for diffraction experiments. Crystal optimization aims to turn poor quality crystals into diffraction-quality crystals that can be used for structure determination. There are a variety of methods that can be used to improve crystal quality, including crystal seeding. The nucleation event in protein crystallization is a poorly understood process. In many crystallization experiments, it is not possible to reach sufciently high levels of saturation for nucleus formation. In many cases, the introduction of a crystal or crystal seed stock at lower levels of saturation can facilitate nucleation and crystal growth. Micro-, macro-, heterogeneous and in situ seedings represent different seeding techniques. High-throughput crystallization and visualization platforms have been widely established and are commonly used by high-throughput structural genomics initiatives.9093 Crystallization experiments can be monitored using an imaging robot and the images can either be analyzed manually or using automatic crystal recognition systems. Such a kind of system has been installed at EMBL Hamburg, and the high-throughput crystallization facility has been open to the general scientic community.92 The facility covers every step in the crystallization process from the preparation of crystallization cocktails for initial or customized screens to the setting up of hanging-drop vapourdiffusion experiments and their automatic imaging. Obtaining useful PNAE crystals was our most tedious example of plant protein crystallization because it took in total four years to obtain good X-ray-diffracting crystals, although the entire time period was not invested in this particular enzyme. The two major obstacles were: (a) precipitation of PNAE in crystallization solution and (b) the complete lack of crystal (nucleus) formation when nearly all of the commercially available crystallization kits were applied. Most of the typical crystallization procedures mentioned above were used. In a total about 6000 crystallization conditions were tried, with very few conditions resulting in crystal formation. Only 2-dimensional crystals were obtained, which were useless for X-ray analysis. These conditions then were optimized in timeconsuming trials by hand-pipetting of 48 ll hanging drops. From about 1000 drops, 10 crystals were nally detected, from and allowed immediate which the best gave a resolution of 2.0 A structure elucidation of the enzyme. Meanwhile, the success rate in crystallizing this particular enzyme increased and probably will also become routine soon. 4.4 Homology modelling and structure of PNAE

Downloaded on 24 July 2011 Published on 24 October 2007 on http://pubs.rsc.org | doi:10.1039/B711935F

models with the software MODELLER (version 4).94 The model with the best overall steric properties was chosen for comparison with the 3D-structure of PNAE generated from X-ray data. ) was The 3D-structure of His6 -PNAE (resolution of 2.1 A elucidated by the molecular replacement method with the software Auto-Rickshaw.43 The model of the salicylic acid binding protein 2 (SABP2, pdb code: 1XKL) from Nicotiana tabacum95 with identity of the amino acid sequence to PNAE of 54% was successfully used as search model for molecular replacement. Sufcient electron density was not observed for the His6 -tag or for the rst eight amino acids from the N-terminal end of the enzyme because of high exibility. The fold of PNAE belongs to the a/b-hydrolase family, which is one of the biggest and fastest growing enzyme families with about 4250 members.96,97 Based on the 3D-structure, PNAE indeed represents a novel member of that family. The enzyme consists of eight b-strands and eight a-helices (Fig. 10).

Fig. 10 Overall structure of PNAE highlighting the cap-region, which is shown by the arrow.

The structure of PNAE was modelled76,77 based on the 3Dstructure of hydroxynitrile lyase (EC 4.2.1.39) from Hevea resolution (see Protein brasiliensis, which was available at 1.9 A databank 1YAS) and which has an amino acid sequence identity to PNAE of 43%. Homology modelling gave a range of PNAE This journal is The Royal Society of Chemistry 2007

The two anti-parallel strands and the three helices (see secondary structure above, the straight line in Fig. 10) form a domain which is named the cap region. The connection of all these structural elements is illustrated by the 2D-topological diagram (Fig. 11). The three amino acids (Ser87, Asp216 and His244) that form the catalytic triad and are essential for enzyme activity, as shown by mutation experiments, are included in Fig. 11.77 The enzyme structure illustrates that these residues are located directly in the reaction channel. Ser87 is positioned in the socalled nucleophilic elbow which is located between the b3 sheet and a3 helix. The narrow reaction channel connects the binding site with the surface of the enzyme, and the channel is anked by predominantly hydrophobic residues of the cap-domain. This domain consists of a total of 71 amino acids from Asp116 to Phe187. In accord with the non-polar structure of the substrate PNA, about 50% of the cap consists of hydrophobic residues. Most of the a/b-hydrolase enzymes exhibit this domain, which is involved in substrate recognition and interfacial activation.98 The cap-region has, in contrast to the core-region of these enzymes, Nat. Prod. Rep., 2007, 24, 13821400 | 1391

View Online

models of PNAE, HbHnl and the salicylic acid binding protein. The entrance to the channel is represented in Fig. 12 and shows clear differences between the modelled and the X-rayderived structure of PNAE, indicating the necessity of the X-ray analysis. There is a future need to generate appropriate substrate enzyme complexes to understand more precisely the extraordinary substrate specicity of the enzyme and especially to get to know details about the amino acids involved in the recognition of the substrate PNA.

5 Vinorine synthase (VS)


Downloaded on 24 July 2011 Published on 24 October 2007 on http://pubs.rsc.org | doi:10.1039/B711935F

5.1 The role of VS in generating the ajmalan structure On the pathway to the ajmalan basic skeleton, vinorine synthase (VS, EC 2.3.1.160) plays a central role by nalizing the synthesis of the six-ring ajmalan system. The enzyme converts the sarpagantype alkaloid epivellosimine, which is the reaction product of the preceding enzyme PNAE (see Section 4), to the ajmalantype in a coenzyme A-dependent reaction. VS has been identied from R. serpentina cell suspensions and its properties and were preliminarily described a long time ago.99 Later, the cDNA of the enzyme was isolated and could successfully be expressed in E. coli. Again, the reverse genetic approach was the successful strategy to get the full length cDNA clone:100,101 after purication of VS from the plant cells, it was identied by SDS gel electrophoresis and partial sequencing. The primary structure of the encoded protein and sequence alignment studies demonstrated that it was a novel enzyme. The pQE-2 vector and E. coli M-15 cell line served for good overexpression with mg amounts of enzyme per litre

Fig. 11 2D-topological diagram of PNAE illustrating the connection of the structural elements (cap domain at bottom).

a spectacular variability in topology, leading to the surprising variability in functions and to the substrate selectivity, which is especially pronounced for PNAE. Comparison of the overall structure of modelled PNAE with the X-ray structure gave a root mean square deviation (rmsd) . For the cap region it was 1.045 A , and the value of 0.916 A core region exhibited a rmsd of 0.812 A. The core region, the conserved region in the a/b-hydrolase family, shows the lower rmsd, whereas the exible and variable cap domain differs more ). between the two models (rmsd of 1.045 A Comparison of the catalytic amino acids Ser87, Asp216 and His244 shows only slight differences between the three X-ray

Fig. 12 Geometry of the entrance of the reaction channel (A) derived from X-ray analysis and (B) from modelling of PNAE; and zoomed regions (C) and (D), respectively (images provided by Dr M. Hill).

1392 | Nat. Prod. Rep., 2007, 24, 13821400

This journal is The Royal Society of Chemistry 2007

View Online

Downloaded on 24 July 2011 Published on 24 October 2007 on http://pubs.rsc.org | doi:10.1039/B711935F

of bacterial suspension. The soluble enzyme was puried, as for the previous proteins (Sections 24), by Ni-NTA chromatography, always an important step in getting homogenous enzyme preparations for biochemical characterization as well as for crystallization. The functional identity of VS was proven by showing formation of the product of the reaction (the methoxylated indolenine alkaloid vinorine) by electron impact mass spectrometry.101 Sequence alignment studies supported classication of VS within the so-called BAHD enzyme superfamily.102 BAHD is the abbreviated name from the rst four enzymes of this family isolated from plant species. Primarily this classication was based on two highly conserved motifs in that family, 362 Asp-PheGly-Trp-Gly366 and 160 His-xxx-Asp164 , respectively. As deduced from X-ray structures of chloramphenicol acetyltransferase and dihydrolipoamide acetyltransferase, these motifs participate in the catalytic reaction of acetyl transfer.103,104 A catalytic triad (Lys-His-Asp) was discussed for the activity of arylamine N acetyltransferase.105,106 The primary structure of VS contains several of these residues, suggesting an important function of these particular three amino acids.101 5.2 Inhibitors and mutants of VS

enzyme mutants, which provided the rst clear evidence for the functional and maybe exclusive role of the BAHD-typical domain His-xxx-Asp in this particular enzyme family. The second motif, 362 Asp-Phe-Gly-Trp-Gly366 (DFGWG), is completely conserved in the BAHD enzymes. Mutation of the rst Asp reduced the VS activity by about one-third, a result matching with mutation of another BAHD enzyme involved in anthocyanin glucoside malonylation.107 The 3D X-ray structure of VS discussed below will explain the signicance of this motif. Based on sequence alignments, acyltransferases are divided into four evolutionary sequence clusters.108 Adopting this classication, VS falls into cluster C enzymes catalyzing esterication of hydroxyl groups of metabolically unrelated secondary metabolites. From the above results only two probable catalytically active amino acids were identied for VS; His160 and Asp164. Their function was proven by the 3D-structure of VS (discussed in Section 5.4), after nding appropriate conditions of crystallization of the recombinant enzyme in long crystallization trials.

5.3 Unusual crystallization conditions and instability of VS crystals Beginners luck is denitely a factor in nding conditions for crystallizing a protein the rst time. This is partly because beginners are more willing to try new conditions and will often do na ve things to the sample, thus nding novel conditions for crystal growth. This is also because no one can predict the proper conditions for crystallizing a new protein.109 In this light, VS represents a good example of successfully using rather uncommon crystallization conditions. Similarly to the other enzymes described in this article, on the one hand the purication procedure of VS after over-expression in E. coli was crucial for obtaining the rst crystals, but on the other hand, removal of the His6 -tag by DAPase was the second requirement to succeed. Each experiment to crystallize the His-tagged enzyme resulted in precipitation or in undetectable nucleation of the enzyme. After NiNTA chromatography, two additional chromatographic steps (ion exchange on MonoQ followed by Sephacryl S-100) were essential and gave highly pure VS as judged by SDS gel chromatography and Coomassie staining. Also in this case the hanging drop method was used and initially small, clustered crystals were observed on applying Crystal Screen and Crystal Screen2 kits from Hampton Research110,111 at 22 C and using ammonium sulfate as the precipitant in the presence of polyethylene glycol (PEG) 400. Optimum crystallization conditions were found after systematic changes of precipitant concentrations, types of PEG, buffers, pH, temperature and enzyme concentrations. VS crystals obtained only at the relatively high temperature of 32 C and low protein concentration of 2 mg ml1 , proved to be the best ones for X-ray measurements. These conditions are exceptional, because such a combination of temperature and enzyme concentration was not found in the Biological Macromolecule Crystallization Database (BMCD),112 and was just based on trying unusual conditions. The well diffracting VS crystals proved to be very sensitive to temperature changes. In contrast to all other crystals mentioned in this article, it was not possible to transport VS crystals for synchrotron measurements without careful freezing procedures, and such procedures were always essential. Nat. Prod. Rep., 2007, 24, 13821400 | 1393

After expression of the VS cDNA in E. coli and its purication to near homogeneity, inhibition studies gave the rst clue as to the catalytically active amino acids.101 Inhibitors acting on Ser, His and Cys such as chloromethyl ketones, the unselective Hisdirected diethyl pyrocarbonate, Hg2+ -ion as a SH-group modier, the Ser- and Cys-selective benzenesulfonyluroide (AEBSF) and the agmatine derivative E-64 were all tested. All these compounds inhibited the VS activity.101 The results clearly indicated that Ser, His and Cys are essential for the VS-catalyzed reaction. Together with results of sequence alignment studies, a series of mutation experiments was performed101 and gave some detailed information on the putative catalytic residues (see Table 3). Five conserved Ser residues were replaced by Ala and all the derived mutant enzymes still showed catalytic activity. This made involvement of a SerHis-Asp triad in the binding site unlikely. Also, when Cys89 and Cys149 were mutated to Ala, the VS activity was not completely knocked out, making a Cys-His-Asp triad also unlikely. But replacement of His160 and Asp164 by Ala resulted in inactive

Table 3 Recombinant His6 -tagged VS mutants and their relative activity (n.d. = not detectable, data are from ref. 101) Enzyme Wild-type S68A C89A S413A D360A S16A N293A D362A S29A S243A D32A C149A H160A D164A Relative activity (%) 100 100 100 100 100 71 68 35 25 17 14 10 n.d. n.d.

This journal is The Royal Society of Chemistry 2007

View Online

5.4

Structure elucidation of VS

Downloaded on 24 July 2011 Published on 24 October 2007 on http://pubs.rsc.org | doi:10.1039/B711935F

Up to the time when this review was written, there were no crystal structures known in the BAHD enzyme family which could be helpful for 3D-structure elucidation of VS. By inhibition of methionine biosynthesis,113 the recombinant E. coli produced (in the presence of selenomethionine) SeMet-VS, which was puried and crystallized similarly to the wild-type VS. The multiwavelength anomalous diffraction (MAD) approach together resolution) with the diffraction data of the wild-type VS (2.6 A allowed for the rst time determination of the VS structure, for which the selenium sites were taken as marker residues to place the amino acid side chains into the appropriate electron densities. After several rounds of renement, the nal three-dimensional structure of VS was obtained.114116 The structure contains 14 bstrands (b1b14) and 13 a-helices (a1a13) and consists of two domains, A and B, of about the same size. The two domains are connected by a long loop. Both have a similar backbone fold, but their topology is different. Between the domains a solvent channel is formed, running through the entire VS molecule (Fig. 13). The His-xxx-Asp motif is located at the interface of the two domains. His160 is situated directly in the centre of the solvent channel and is accessible from both ends of the channel. Such an arrangement allows both reaction partners, 16-epi-vellosimine and acetyl-CoA, to approach the active site. Kinetic data obtained with a partially puried VS preparation from Rauvola cells had earlier suggested a ternary complex between both ligands and the enzyme with independent binding of the ligands.99 The mechanism

of the reaction is illustrated in Scheme 3. The 3D-structure of VS strongly supports this proposal. 5.5 Important amino acids and mechanistic aspects The mutagenesis studies indicated the indispensability of His160 for VS activity, and the 3D-structure of the enzyme conrms the functional importance of this residue. It is this amino acid which is located directly in the centre of the solvent channel (Fig. 14). This is in agreement with the above-mentioned kinetic data, that the two substrates may approach the binding pocket from different directions (from which direction still needs to be analyzed). Based on His160, the acetyl-transfer reaction would occur as illustrated in Scheme 3. The amino acid acts as a base, taking off the proton of the hydroxy group of 17-deacetylvinorine. The 17-oxygen will then attack the carbonyl carbon of acetylCoA, resulting in its acetylation and release of CoA. The reaction probably proceeds without formation of an acetylated enzyme intermediate. The second conserved residue (Asp164) is also located directly at the binding site and belongs to the essential His-xxx-Asp motif. Asp164 has no catalytic function. Its side chain points away from His160, not allowing hydrogen-bond formation between the residues. Such an arrangement excludes an amino acid dyad in the catalytic process as found for e.g. human carnitine acetyltransferase.117 Asp164 forms a salt bridge to Arg279, which is also a conserved residue in the BAHD family. This interaction of the two residues appears to be of structural, not of catalytic, importance. Exchange of Arg279 against Ala results in loss of enzyme activity. Most probably it is the geometry of the binding pocket which is signicantly changed when the salt bridge is interrupted. It must be demonstrated in future whether the importance of the His-xxx-Asp motif in other enzymes of the BAHD family is based on the same reasons as shown for VS. The DFGWG motif completely conserved in that family is not localized in the binding pocket of the enzyme. In contrast, it is far away from the region of catalysis, indicating its structural importance. It is not involved in substrate binding and might maintain the conformation of the enzyme structure. Mutation of the Asp residue to Ala therefore caused decrease of VS activity (Table 4). Such mutation also resulted in complete loss of reactivity of another BAHD enzyme participating in anthocyanin glucoside malonylation.107 A more detailed insight into the binding pocket of VS requires more structural information. In particular, the crystal structure of vinorine synthase with substrate or product bound should provide much deeper understanding of the catalytic process in this enzyme family.

Fig. 13 Image of VS illustrating the two-domain structure (A and B), the reaction channel and the catalytic His in the binding pocket.

Scheme 3 Proposed reaction mechanism of VS; scheme from ref. 116.

1394 | Nat. Prod. Rep., 2007, 24, 13821400

This journal is The Royal Society of Chemistry 2007

View Online

Downloaded on 24 July 2011 Published on 24 October 2007 on http://pubs.rsc.org | doi:10.1039/B711935F

recent examples. Rosmarinic acid synthase is one of the most recent members of the BAHD family detected by reverse genetics.121 Prediction of the reactions catalyzed or the substrates accepted by the enzymes, based on sequence alignments, is difcult.122 But mutation experiments, such as those described above, combined with sequence comparison might help to elucidate the functionalities of other members of the BAHD family. The structure of VS could also be helpful in solving other structures by molecular replacement approaches, or at least could provide opportunities for homology modelling. This is particularly important for those family members for which the cDNAs have functionally been cloned and which take part in the biosynthesis of important plant natural products. Prominent examples are salutaridinol acetyltransferase in morphine biosynthesis,123 deacetylvindoline acetyltransferase124 on the way to vinblastine, and the acyltransferases acting on the skeleton of taxol,125 just to mention BAHD members involved in alkaloid biosynthesis.

6 Raucaffricine glucosidase (RG), a side route of the ajmaline pathway


As a part of a metabolic network, the ajmaline biosynthetic pathway has at several intermediate stages side routes, of which the formation of raucaffricine is one of the most important. Raucaffricine is the glucoside of the intermediate vomilenine. This glucoalkaloid has previously been detected as the main alkaloid of Rauvola cell cultures, exceeding signicantly the amounts in differentiated Rauvola cells.126 The enzyme, raucaffricine glucosidase (RG, EC 3.2.1.125), converts the glucoside back to vomilenine. It has been identied, characterized and its cDNA has been cloned in E. coli.127,128 RG represents the second glucosidehydrolyzing enzyme in Rauvola alkaloid biosynthesis, the rst being SG. The primary structure and the substrate specicity of the two glucosidases are different; RG accepts strictosidine, the substrate of SG, but SG does not hydrolyze raucaffricine.127,128 To compare the two glucosidases, it was important to determine the 3D-structure of RG in order to evaluate substrate recognition and better understand the deglucosylation process. Over-expression yielded enough pure RG enzyme for crystallization experiments when the routine strategy, applied for the other Rauvola enzymes, was followed. 6.1 Crystallization attempts and the 3D-structure of RG His6 -tag RG was crystallized by the hanging-drop vapour diffusion technique using similar conditions as for strictosidine glucosidase. Crystals reached maximum dimensions of about 0.2 0.15 0.05 mm. The crystals belong to space group I 222 and diffract to .129 2.30 A The structure of RG was determined by the molecular replacement method (unpublished data). The structure of Rauvola SG served as a search model, because RG shares a sequence identity with SG of 56%. RG belongs to the family 1 of the glycosyl hydrolases. Crystal packing showed two enzyme molecules for each crystallographic asymmetric unit, but the enzyme is active as the monomer.127 The rened model of RG consists of 13 a-helices and 13 bstrands. RG adopts the expected topology of a single (b/a)8 barrel fold (TIM barrel) (Fig. 15). The barrel hosts a binding site for the Nat. Prod. Rep., 2007, 24, 13821400 | 1395

Fig. 14 Surface representation, reaction channel and the localization of the catalytic residue His160 (in red) in VS (images kindly provided by Dr M. Hill).

Very recently, structural and mutational studies on anthocyanin malonyltransferase were reported.118 This enzyme is also a BAHD member and now represents the second example where the 3Dstructure has been solved. The complex of this malonyltransferase with its ligand malonyl-CoA together with a series of site-directed mutations allowed the acyl-acceptor binding site to be identied. These results together with those on VS are very helpful for the understanding in more detail of the diversity of the acyl-acceptor specicity within the BAHD family.118 5.6 Signicance of the vinorine synthase structure for the BAHD enzyme family The BAHD enzymes belong to a constantly growing family with an increasing number of functionally expressed members. At the DNA level it is suggested that in two organisms alone 180 genes occur that might code for BAHD acyltransferases; in Oryza sativa (rice) 119 genes of this family have been identied, but none have been investigated for their biochemical function. The Arabidopsis thaliana genome delivered 64 genes of the family, but very few have been functionally described so far. The (Z)-3-hexen-1-ol O-acetyltransferase (producing the acetylated hexenol)119 or another anthocyanidin 5-O-glucoside-O-malonyltransferase120 are This journal is The Royal Society of Chemistry 2007

View Online

natural substrate raucaffricine. The groove leading to the catalytic centre is formed mainly by irregular loops between the secondary structures on top of the enzyme. Similarly to SG, RG contains the catalytic residues E186 and E420. The most striking difference is at the catalytic centre of the RG and SG. An identical position of Trp392 but a different orientation of the tryptophan in the two glucosidases helps the protein to recognize its substrate.

Downloaded on 24 July 2011 Published on 24 October 2007 on http://pubs.rsc.org | doi:10.1039/B711935F

and 5-methoxytryptamine into the corresponding novel substituted strictosidines, indicating the importance of knowing the 3Dstructure of STR1 for redesigning its enzyme activity.51,52 Based on such information, a more systematic mutation approach will in the future show whether a further expanded substitution pattern at the indole moiety can be achieved. A biomimetic approach which uses strictosidine (or derivatives), strictosidine glucosidase and an excess of primary amines (with obviously any residues R followed by reduction) (Scheme 4), results in formation of novel N-analogous heteroyohimbine alkaloids.47,52 After producing STR1 mutants (allowing synthesis of a range of novel strictosidines) and optimizing the second enzyme strictosidine glucosidase by mutation, such a chemo-enzymatic approach may be an excellent tool for the generation of new alkaloid libraries containing thousands of alkaloids. Knowledge of biosynthesis combined with structural biology might be an excellent tool in future to develop such combinatorial enzyme-mediated approaches.

8 Other structural examples from the alkaloid eld


Despite the fact that alkaloids have been an excellent source of biologically active compounds, e.g. anti-infectious drugs,130 only limited efforts have been focussed on understanding the mode of action of alkaloids and their interactions with host proteins or enzymes from other organisms at the molecular level. These include crystallographic studies of tropinone reductase, phospholipase, transcriptional repressor, acetylcholine esterase, calmodulin and tyrosine kinase and their complexes with various alkaloids, listed in Table 4. Below the only other structural examples from alkaloid biosynthesis, the tropinone reductases, are discussed in some detail. Tropinone reductase (TR) comprises a branching point in the biosynthetic pathway of tropane alkaloids which includes such medicinally important compounds as hyoscyamine, scopolamine and cocaine (Scheme 5). All the tropane-alkaloid-producing plants species so far examined have two TR activities and their amino acid sequence is known from Datura stramonium,131,132 Hyoscyamus niger,133 Atropa belladonna134,135 and Solanum tuberosum.136 In these species, there are two types of TR, TR-I and TR-II. These

Fig. 15 The (b/a)8 barrel fold of RG, illustrating the binding pocket and the catalytic acids E186 and E420.

7 Structure-based redesign of enzymes and applications in chemo-enzymatic approaches


The enzymes described in this review are of relatively high substrate specicity. This is a great disadvantage in terms of using them in the future as biocatalysts e.g. for the enzymatic synthesis of structurally novel alkaloids. For instance, the best substrate for STR1 is tryptamine, and benzene-ring-substituted tryptamines generally react at less than 10% of the rate for tryptamine. The STR1 mutant Val208Ala, however, converts 5-methyltryptamine

Scheme 4 A biomimetic approach to indole alkaloid diversity based on rational enzyme design (STR1, strictosidine synthase; SG, strictosidine glucosidase; X, various substituents).

1396 | Nat. Prod. Rep., 2007, 24, 13821400

This journal is The Royal Society of Chemistry 2007

View Online

Table 4 3D-structures of enzymealkaloid complexes from the PDB data bank PDBcode 2AE2 1IPF 1ZR8 1JUM 1VOT 1ZGB 1DX6 1H22 1XA5 Protein Tropinone reductase-II Tropinone reductase-II Phospholipase Transcriptional repressor Acetylcholine-esterase Acetylcholine-esterase Acetylcholine-esterase Acetylcholine-esterase Calmodulin Complex with alkaloid Pseudotropine142 Tropinone143 Ajmaline144 Berberine145 Huperzine A146 (R)-Tacrine-C10 H20 -hupyridone147 Galanthamine148 (S)-Hupyridone-C10 H20 -(S)-hupyridone149 A bis-indole alkaloid150

Downloaded on 24 July 2011 Published on 24 October 2007 on http://pubs.rsc.org | doi:10.1039/B711935F

Scheme 5 Biosynthetic connection of tropane alkaloids through the intermediate tropinone (ODC, ornithine decarboxylase; PMT, putrescine methyltransferase; TR-I, tropinone reductase I; TR-II, tropinone reductase II).

TRs share 64% sequence identity and belong to the short-chain dehydrogenase/reductase (SDR) family. TRs catalyze NADPHdependent reductions of the 3-carbonyl group of their common substrate, tropinone, to hydroxy groups with different diastereomeric congurations: TR-I (EC 1.1.1.206) produces tropine (3a-hydroxytropane), and TR-II (EC 1.1.1.236) produces pseudotropine (W -tropine, 3b-hydroxytropane). These enzymes have different K m values for tropinone and its analogues but have similar K m values for NADPH, and both catalyze transfer of the pro-S hydrogen atom of NADPH to tropinone.137 This indicates that the two TR enzymes have different binding sites for tropinone but have similar ones for NADPH, and that their different stereospecicities result from the different binding modes of tropinone. Both TRs from Datura stramonium have been cloned and expressed in E. coli, and puried and crystallized using hanging drop vapour diffusion techniques with sodium This journal is The Royal Society of Chemistry 2007

citrate and 2-methyl-2,4-pentenediol as buffer and precipitant respectively.138,139 The structure of each TR was solved using the isomorphous replacement method.140 The two structures are almost indistinguishable from each other in both subunit folding and their association into dimers. Both TR subunits consist of a core domain that includes most of the polypeptide and a small lobe that protrudes from the core. In the centre of the core domain is a sevenstranded parallel b-sheet, anked on each side by three a-helices, which constitutes the Rossmann-fold topology. This core structure is highly conserved among the SDR family members, despite relatively low residue identity between these enzymes (30%).141 The structure of TR-I was also determined in the presence of NADP+ . The cofactor was found to be located at the bottom of the cleft between the core domain and the small lobe. The carboxamide group of the nicotinamide ring is anchored by the mainchain nitrogen and oxygen atoms of Ile204 and the side-chain Nat. Prod. Rep., 2007, 24, 13821400 | 1397

View Online

Downloaded on 24 July 2011 Published on 24 October 2007 on http://pubs.rsc.org | doi:10.1039/B711935F

oxygen of Thr206. This tight binding of the carboxamide group to the protein directs the B-face of the nicotinamide ring toward the void of the cleft, consistent with the observed specicity for the pro-S hydride transfer of both TRs.137 TR-II catalyzes the one step chemical reaction via the following states: (1) TR-II (E) + NADPH (S1 ) + tropinone (S2 ) (2) ES1 S2 (3) EP1 P2 (4) TR-II + NADP+ (P1 )+ W -tropinone (P2 ), where the states are: (1) E, the free enzyme; (2) ES1 S2 , the enzyme with the substrates [NADPH (S1 ) and tropinone (S2 )] bound prior to reaction initiation; (3) EP1 P2 , the enzyme with the products [NADP+ (P1 ) and W -tropine (P2 )] bound just after the reaction is completed; and (4) the free enzyme again after releasing the products. The crystallographic structures of TR-II have been determined in an unliganded form,140 as the complex with NADPH,141 as the complex with NADPH and tropine,141 and as the complex with NADP+ and W -tropine.142 These structures provide a great deal of insight into each state of the reaction in the biosynthesis of tropane alkaloids (Fig. 16).

details of the enzymatic biosynthesis of monoterpenoid indole alkaloids, particularly of the pathway to the antiarrhythmic ajmaline. For ve major enzymes, the fold-families, their overall structures and substrate binding sites are now known, including identication of amino acids of catalytic and structural importance. This will facilitate elucidation of the corresponding enzyme mechanisms. For enzymes of more general signicance, such as strictosidine synthase, crystal structures of substrates and product complexes have been analyzed; this allows understanding of ligand recognition and permitted the rst successful rational enzyme engineering for the synthesis of novel alkaloids. Structure-guided combinatorial approaches with redesigned enzyme mutants together with biomimetic strategies will help to generate large alkaloid libraries with hopefully important and novel biological activities.52

10 Acknowledgements
We thank Mrs Yang Liuqing for kind help in preparing the diagrams for this article. The described research was continuously supported by the Deutsche Forschungsgemeinschaft (Bonn, BadGodesberg, Germany), the Fonds der Chemischen Industrie (Frankfurt/Main, Germany) in cooperation with the Bundesmin Bildung und Forschung (BMBF, Bonn, Germany), the isterium fur DLR Ofce Bonn of BMBF, and the Deutscher Akademischer Austauschdienst (DAAD), Bonn, Germany. The work was also supported by European Community Access to Research Infrastructure Action of the Improving Human Potential Programme to the European Molecular Biology Laboratory (EMBL) at Hamburg Outstation (contract No. HPRI-CT-1999-100017) and the FP6 Programme (No. RII3/CT/2004/5060008), and the Berliner Synchrotronstrahlung Elektronenspeicherring-Gesellschaft fur (Berlin, Germany). We also acknowledge staff members at the DORIS storage ring (DESY, Hamburg, Germany), Prof. Lottspeichs team at the Max-Planck-Institute of Biochemistry (MPI, Martinsried, Germany) for protein sequencing, Prof. H. Michels group at MPI (Frankfurt/Main, Germany) and Prof. T. Kutchan (St Louis, USA) for introduction into structural and molecular biology techniques, respectively, and the ESRF beamline (Grenoble, France) for technical help. Present co-workers (in alphabetical order: Dr Leif Barleben, Dr Marco Hill, Petra Kercmar, Elke Loris, Kerstin Oelrich, and Dr Martin Ruppert) and former coworkers of various nationalities are very much appreciated for their enthusiastic performing of chemical, enzymatic and crystallization experiments. Both Dr E. Mattern-Dogru and Dr Xueyan Ma originally initiated structural biology research in our group. We thank Dr F. Leeper (Cambridge, UK) and Dr P. Tucker (EMBL Hamburg, Germany) for advice and help in correcting the manuscript.

11 References
Fig. 16 (a) Complex structure of tropinone reductase II (TR-II) with tropinone in black and NADPH in blue+grey. (b) Complex structure of tropinone reductase I (TR-I) with NADPH illustrated in blue+grey. 1 E. J. Saxton, Nat. Prod. Rep., 1997, 14, 559590. 2 J. Leonhard, Nat. Prod. Rep., 1999, 16, 319338. 3 J. Stoeckigt, Biosynthesis in Rauvola serpentinaModern Aspects of an Old Medicinal Plant, in: The Alkaloids, ed. G. A. Cordell, Academic Press, San Diego, California, vol. 47, pp. 115172. 4 S. OConnor and J. J. Maresh, Nat. Prod. Rep., 2006, 23, 532547. 5 T. M. Kutchan, Recent Adv. Phytochem., 2002, 36, 163178. 6 V. De Luca, Recent Adv. Phytochem., 2003, 37, 181202. 7 T. Hashimoto and Y. Yamada, Curr. Opin. Biotechnol., 2003, 14, 163 168. 8 J. Murata and V. De Luca, Plant J., 2005, 44, 581594.

9 Conclusions and future aspects of structural biology in the alkaloid eld


Elucidation of the three-dimensional structures of several of the Rauvola enzymes has extensively broadened our knowledge on
1398 | Nat. Prod. Rep., 2007, 24, 13821400

This journal is The Royal Society of Chemistry 2007

View Online

9 F. Vazquez-Flota, E. De Carolis, A. Alarco and V. De Luca, Plant Mol. Biol., 1997, 34, 935948. 10 B. St. Pierre, P. Laamme, A. Alarco and V. De Luca, Plant J., 1998, 14, 703713. 11 Y. Yamazaki, A. Urano, H. Sudo, M. Kitajima, H. Takayama, M. Yamazaki, N. Aimi and K. Saito, Phytochemistry, 2003, 62, 461470. 12 A. H. Heckendorf and C. R. Hutchinson, Tetrahedron Lett., 1977, 48, 41534154. 13 C. R. Hutchinson, A. H. Heckendorf, J. L. Straughn, P. E. Daddona and D. E. Cane, J. Am. Chem. Soc., 1979, 101, 33583369. 14 A. I. Scott and S.-L. Lee, J. Am. Chem. Soc., 1975, 97, 69066908. 15 J. Stoeckigt, J. Treimer and M. H. Zenk, FEBS Lett., 1976, 70, 267 270. 16 J. Stoeckigt, T. Hemscheidt, G. Hoee, P. Heinstein and V. Formacek, Biochemistry, 1983, 22, 34483452. 17 J. Stoeckigt, The Biosynthesis of Heteroyohimbine-Type Alkaloids, in: Indole and biogenetically related alkaloids, ed. J. D. Philipson and M. H. Zenk, Academic Press, London/New York, 1980, pp. 113141. 18 M. H. Zenk, H. El-Shagi, H. Arens, J. Stoeckigt, E. W. Weiler and B. Deus, Formation of the Indole Alkaloids Serpentine and Ajmalicine in Cell Suspension Cultures of Catharautus roseus, in: Plant Tissue Culture and its Bio-technological Application, ed. W. Barz, E. Reinhard and M. H. Zenk, Springer Verlag, Berlin/Heidelberg/New York, 1977, pp. 2743. 19 J. Stoeckigt and H. J. Soll, Planta Med., 1980, 40, 2230. 20 J. Stoeckigt, P. Obitz, H. Falkenhagen, R. Lutterbach and S. Endress, Plant Cell, Tissue Organ Culture, 1995, 43, 97109. 21 R. B. Herbert, The Biosynthesis of Terpenoid Indole Alkaloids, in: Monoterpenoid Indole Alkaloids, The Chemistry of Heterocyclic Compounds, Vol. 25 (supplement), ed. J. E. Saxton, John Wiley and Sons, Chichester/New York/Brisbane/Toronto/Singapore, 1994, pp. 113. 22 J. Stoeckigt, Alkaloid Metabolism in Plant Cell Culture, in Natural Product Analysis, ed. P. Schreier, M. Herderich, H. U. Humpf and W. Schwab, Vieweg, Braunschweig/Wiesbaden, Germany, 1998, pp. 313325. 23 M. Ruppert, X. Y. Ma and J. Stoeckigt, Curr. Org. Chem., 2005, 9, 14311444. 24 C. Hinse, Y. Sheludko, A. Provenzani and J. Stoeckigt, J. Am. Chem. Soc., 2001, 123, 51185119. 25 C. Hinse, C. Richter, D. Moskau, A. Provenzani and J. Stoeckigt, Bioorg. Med. Chem., 2003, 11, 39133919. 26 J. Stoeckigt, S. Panjikar, M. Ruppert, C. Barleben, X. Y. Ma, E. Loris and M. Hill, Phytochem. Rev., 2007, 6, 1534. 27 E. Leete, Planta Med., 1990, 56, 339352. 28 T. Hashimoto and Y. Yamada, Annu. Rev. Plant Physiol. Plant Mol. Biol., 1994, 45, 257285. 29 J. Stoeckigt and M. H. Zenk, J. Chem. Soc., Chem. Commun., 1977, 646648. 30 J. Stoeckigt and M. H. Zenk, FEBS Lett., 1977, 79, 233237. 31 J. F. Treimer and M. H. Zenk, Eur. J. Biochem., 1979, 10, 225233. 32 A. de Waal, A. H. Meijer and R. Verpoorte, Biochem. J., 1995, 306, 571580. 33 L. H. Stevens, C. Giroud, J. M. E. Pennings and R. Verpoorte, Phytochemistry, 1993, 33, 99106. 34 H. Wang, R. Chen, M. Chen, M. Sun and Z. H. Liao, Acta Bot. Boreali Occident. Sin., 2006, 26, 900905. 35 T. M. Kutchan, N. Hampp, F. Lottspeich, K. Beyreuther and M. H. Zenk, FEBS Lett., 1988, 237, 4044. 36 T. M. Kutchan, FEBS Lett., 1989, 257, 127130. 37 T. M. Kutchan, H. Dittrich, D. Bracher and M. H. Zenk, Tetrahedron, 1991, 47, 59455954. 38 T. D. McKnight, C. A. Roessner, R. Devagupta, A. I. Scott and C. L. Nessler, Nucleic Acids Res., 1990, 18, 4939. 39 C. Canel, M. I. Lopes-Cardoso, S. Whitmer, L. von der Fits, G. Pasquali, R. von der Heijden, J. H. Hoge and R. Verpoorte, Plant Mol. Biol., 1999, 39, 12991310. 40 Y. Yamazaki, H. Sudo, M. Yamazaki, N. Aimi and K. Saito, Plant Cell Physiol., 2003, 44, 395403. 41 X. Y. Ma, J. Koepke, G. Fritzsch, R. Diem, T. M. Kutchan, H. Michel and J. Stoeckigt, Biochim. Biophys. Acta, 2004, 1702, 121124. 42 J. Koepke, X. Y. Ma, G. Fritzsch, H. Michel and J. Stoeckigt, Acta Crystallogr., Sect. D, 2005, 61, 690693. 43 S. Panjikar, V. Parthasarathy, V. S. Lamzin, M. S. Weiss and P. A. Tucker, Acta Crystallogr., Sect. D, 2005, 61, 449457.

44 M. Harel, A. Ahoroni, L. Gaidukov, B. Brumshtein, O. Khersonsky, R. Meged, H. Dvir, R. B. G. Ravelli, A. McCarthy, L. Toker, L. Silman, J. L. Sussman and D. S. Tawk, Nat. Struct. Mol. Biol., 2004, 11, 412419. 45 E. I. Scharff, J. Koepke, G. Fritzsch, C. Luecke and H. Rueterjans, Structure, 2001, 9, 493502. 46 X. Y. Ma, S. Panjikar, J. Koepke, E. Loris and J. Stoeckigt, Plant Cell, 2006, 18, 907920. 47 J. Stoeckigt, M. Ruppert, Strictosidine: The biosynthetic key to monoterpenoid indole alkaloids, in: Comprehensive Natural Products Chemistry: Amino Acids, Peptides, Porphyrins and Alkaloids, ed. D. H. R. Barton, K. Nakanishi, O. Meth-Cohn and J. W. Kelly, Elsevier, Amsterdam, 1999, vol. 4, pp. 109138. 48 N. Hampp and M. H. Zenk, Phytochemistry, 1988, 27, 38113815. 49 L. H. Stevens, G. Giroud, E. J. M. Pennings and R. Verpoorte, Phytochemistry, 1993, 35, 353360. 50 E. McCoy, M. C. Galan and S. E. OConnor, Bioorg. Med. Chem. Lett., 2006, 16, 24752478. 51 S. Chen, M. C. Galan, C. Coltharp and S. E. OConnor, Chem. Biol., 2006, 13, 11371141. 52 E. Loris, S. Panjikar, L. Barleben, M. Unger, H. Schubel and J. Stoeckigt, Chem. Biol., 2007, 14, 979985. 53 T. A. Edwards, B. D. Wilkinson, R. P. Wharton and A. K. Aggarwal, Genes Dev., 2003, 17, 25082513. 54 W. De-Eknamkul, N. Suttipanta and T. M. Kutchan, Phytochemistry, 2000, 55, 177181. 55 T. Hemscheidt and M. H. Zenk, FEBS Lett., 1980, 110, 187 191. 56 A. Geerlings, J. I. Memelink, R. von der Heiden and R. Verpoorte, J. Biol. Chem., 2000, 275, 30513056. 57 I. Gerasimenko, Y. Sheludkov, X. Y. Ma and J. Stoeckigt, Eur. J. Biochem., 2002, 269(2), 2042213. 58 G. Haberer and J. J. Kieber, Plant Physiol., 2002, 128, 354362. 59 K. Ljung, A. Oestin, C. Lioussanne and G. Sandberg, Plant Physiol., 2001, 125, 464475. 60 D. P. Dharmawardhana, B. E. Ellis and J. E. Carlson, Plant Physiol., 1995, 107, 331339. 61 M. Zagrobelny, S. Bak, A. V. Rasmussen, B. Jorgensen, C. M. Naumann and B. L. Moller, Phytochemistry, 2004, 65, 293306. 62 J. E. Poulton, Plant Physiol., 1990, 94, 401405. 63 L. Barleben, X. Y. Ma, J. Koepke, G. Peng, H. Michel and J. Stoeckigt, Biochim. Biophys. Acta, 2005, 1747, 8992. 64 J. D. McCarter and S. G. Withers, Curr. Opin. Struct. Biol., 1994, 4, 885892. 65 J. Sanz-Aparicio, J. A. Martinez-Ripoll, M. Lequerica and J. L. Polaina, J. Mol. Biol., 1998, 275, 491502. 66 L. Barleben, S. Panjikar, M. Ruppert, J. Koepke and J. Stoeckigt, Plant Cell, 2007, DOI: 10.1105/tpc.106.045682. 67 L. Verdoucq, M. Czjzek, J. Moriniere, D. R. Bevan and A. Esen, J. Biol. Chem., 2003, 278, 2505525062. 68 L. Verdoucq, J. Moriniere, D. R. Bevan, A. Esen, A. Vasella, B. Henrissat and M. Czjzek, J. Biol. Chem., 2004, 279, 3179631803. 69 M. Czjzek, M. Cicek, V. Zamboni, D. R. Bevan and A. Esen, Proc. Natl. Acad. Sci. U. S. A., 2000, 97, 1355513560. 70 M. H. Zenk, Chasing the Enzymes of Alkaloid Biosynthesis, in Organic Reactivity: Physical and Biological Aspects, ed. B. T. Golding, R. J. Grifn and H. Maskill, The Royal Society of Chemistry, Cambridge, UK, 1995, pp. 89109. 71 T. M. Kutchan, Plant Cell, 1995, 7, 10591070. 72 A. Ptzner and J. Stoeckigt, Planta Med., 1983a, 48, 221227. 73 D. Schmidt and J. Stoeckigt, Planta Med., 1995, 61, 254258. 74 A. Ptzner and J. Stoeckigt, J. Chem. Soc., Chem. Commun., 1983, 459460. 75 H. Warzecha and J. Stoeckigt, New Enzymes of Indole Alkaloid Biosynthesis, in: Proceedings of the 3rd International Symposium on Frontiers in Protein Chemistry and Biotechnology, ed. Z. Wang and W. Li, Changchun Publishing House, China, 1999, pp. 3843. 76 E. Dogru, H. Warzecha, F. Seibel, S. Haebel, F. Lottspeich and J. Stoeckigt, Eur. J. Biochem., 2000, 267, 13971406. 77 E. Mattern-Dogru, X. Y. Ma, J. Hartmann, H. Decker and J. Stoeckigt, Eur. J. Biochem., 2002, 269, 28892896. 78 E. Mattern-Dogru, Molecular Analysis of Polyneuridine Aldehyde Esterase by Functional Expression: Site-directed Mutagenesis and Homology-Modelling, Der Andere Verlag, Osnabrueck, 2002, pp. 1 116.

Downloaded on 24 July 2011 Published on 24 October 2007 on http://pubs.rsc.org | doi:10.1039/B711935F

This journal is The Royal Society of Chemistry 2007

Nat. Prod. Rep., 2007, 24, 13821400 | 1399

View Online

79 The Arabidopsis Initiative, Nature, 2000, 408, 796815. 80 M. Hasslacher, M. Schall, M. Hayn, H. Griengl, S. D. Kohlwein and H. Schwab, J. Biol. Chem., 1996, 271, 58845891. 81 M. Hasslacher, C. Kratky, Z. H. Griengl, H. Schwab and S. D. Kohlwein, Proteins, 1997, 27, 438449. 82 U. Hanefeld, A. J. J. Straathof and J. J. Heijnen, Biochim. Biophys. Acta, 1999, 1432, 185193. 83 K. Gruber, M. Gugganig, U. G. Wagner and C. Kratky, Biol. Chem., 1999, 380, 9931000. 84 J. Hughes, F. J. P. Carvalho and M. A. Hughes, Arch. Biochem. Biophys., 1994, 311, 496502. 85 H. Wajant and K. Pzenmaier, J. Biol. Chem., 1996, 271, 25830 25834. 86 H. Lauble, S. Foerster, B. Miehlich, H. Wajant and F. Effenberger, Acta Crystallogr., Sect. D, 2001, 57, 194200. 87 H. Lauble, B. Miehlich, S. Foerster, C. Koblar, H. Wajant and F. Effenberger, Protein Sci., 2002, 11, 6571. 88 N. E. Chayen, Methods Mol. Biol., 2007, 363, 175190. 89 N. E. Chayen, Structure, 1997, 5, 12691274. 90 I. M. Berry, O. Dym, R. M. Esnouf, K. Harlos, R. Meged, A. Perrakis, J. L. Sussman, T. S. Walter, J. Wilson and A. Messerschmidt, Acta Crystallogr., Sect. D, 2006, 62, 11371149. 91 B. Hazes and L. Price, Acta Crystallogr., Sect. D, 2005, 61, 1165 1171. 92 J. Mueller-Dieckmann, Acta Crystallogr., Sect. D, 2006, 62, 1446 1452. 93 L. J. De Lucas, D. Hamrick, L. Cosenza, L. Nagy, D. McCombs, T. Bray, A. Chait, B. Stoops, A. Belgovskiy, W. W. Wilson, M. Parham and N. Chernov, Prog. Biophys. Mol. Biol., 2005, 88, 285309. 94 A. Sali and T. Blundell, J. Mol. Biol., 1993, 234, 779815. 95 F. Forouhar, Y. Yang, D. Kumar, Y. Chen and E. Fridman, Proc. Natl. Acad. Sci. U. S. A., 2005, 102, 17731778. 96 Esther database, 02.2007: T. Hotelier, L. Renault, X. Cousin, V. Negre, P. Marchot and A. Chatonnet, Nucleic Acids Res., 2004, 32, D145 D147. 97 N. Siew, H. K. Saini and D. Fischer, FEBS Lett., 2005, 579, 3175 3182. S. A. Patkar and M. A. Alsina, 98 Y. Cajal, A. Svendsen, J. De Bolos, Biochimie, 2000, 82, 10531061. 99 A. Ptzner, L. Polz and J. Stoeckigt, Z. Naturforsch., C: Biosci., 1986, 41, 103114. 100 I. Gerasimenko, X. Y. Ma, Y. Sheludko, R. Mentele, F. Lottspeich and J. Stoeckigt, Bioorg. Med. Chem., 2004, 12, 27812786. 101 A. Bayer, X. Y. Ma and J. Stoeckigt, Bioorg. Med. Chem., 2004, 12, 27872795. 102 B. St. Pierre and V. De Luca, Evolution of actyltransferase genes: origin and diversication of the BAHD superfamily of acyltransferases involved in secondary metabolism, in: Recent Advances in Phytochemistry: Evolution of Metabolic Pathways, ed. J. T. Romeo, R. Ibrahim, L. Varin and V. De Luca, Elsevier Science, 2000, vol. 34, pp. 285315. 103 A. G. Leslie, Mol. Biol., 1990, 213, 167186. 104 T. A. Keating, C. G. Marshall, C. T. Walsh and A. E. Keating, Nat. Struct. Biol., 2002, 9, 522526. 105 J. C. Sinclair, J. Sandy, R. Delgoda, E. Sim and M. E. M. Noble, Nat. Struct. Biol., 2000, 7, 560564. 106 A. Upton, N. Johnson, J. Sandy and E. Sim, Trends Pharmacol. Sci., 2001, 22, 140146. 107 H. Suzuki, T. Nakayama and T. Nishino, Biochemistry, 2003, 42, 17641771. 108 L. Hoffmann, S. Maury, F. Martz, P. Geoffroy and M. Legrand, J. Biol. Chem., 2003, 278, 95103. 109 D. E. McRee, Practical Protein Crystallography (2nd edn), Academic Press, San Diego, California, 1999, p. 9. 110 J. Jansarik and S.-H. Kim, J. Appl. Crystallogr., 1991, 24, 409 411. 111 R. Cudney, S. Patel, K. Weisgraber, Y. Newhouse and A. McPherson, Acta Crystallogr., Sect. D, 1994, 50, 414423. 112 G. L. Gilliland, M. Tung, D. M. Blakeslee and J. Ladner, Acta Crystallogr., Sect. D, 1994, 50, 408413. 113 G. D. Van Duyne, R. F. Standaert, P. A. Karplus, S. L. Schreiber and J. Clardy, J. Mol. Biol., 1993, 229, 105124. 114 X. Y. Ma, J. Koepke, A. Bayer, V. Linhard, G. Fritzsch, B. Zhang, H. Michel and J. Stoeckigt, Biochim. Biophys. Acta, 2004, 1701, 129 132.

115 X. Y. Ma, J. Koepke, A. Bayer, G. Fritzsch, H. Michel and J. Stoeckigt, Acta Crystallogr., Sect. D, 2005, 61, 694696. 116 X. Y. Ma, J. Koepke, S. Panjikar, G. Fritzsch and J. Stoeckigt, J. Biol. Chem., 2005, 280, 1357613583. 117 D. Wu, L. Govindasamy, W. Lian, Y. Gu, T. Kukar, M. AgbandjeMcKenna and R. McKenna, J. Biol. Chem., 2003, 278, 1315913165. 118 H. Unno, F. Ichimaida, H. Suzuki, S. Takahashi, Y. Tanaka, A. Saito, T. Nishino, M. Kusunoki and T. Nakayama, J. Biol. Chem., 2007, 282, 1581215822. 119 J. C. DAuria, E. Pichersky, A. Schaub, A. Hansel and J. Gershenzon, Plant J., 2007, 49, 194207. 120 J. C. DAuria, M. Reichelt, K. Luck, A. Svatos and J. Gershenzon, FEBS Lett., 2007, 581, 872878. 121 A. Berger, J. Meinhard and M. Petersen, Planta, 2006, 224, 15031510. 122 J. Beekwilder, M. Alvarez-Huerta, E. Neef, F. Verstappen, H. Bouwmeester and A. Aharoni, Plant Physiol., 2004, 135, 18651878. 123 T. Grothe, R. Lenz and T. M. Kutchan, J. Biol. Chem., 2001, 276, 3071730723. 124 B. St-Pierre, P. Laamme, A. M. Alarco and V. De Luca, Plant J., 1998, 14, 703713. 125 K. Walker, R. Long and R. Croteau, Proc. Natl. Acad. Sci. U. S. A., 2002, 99, 91669171. 126 H. Schuebel and J. Stoeckigt, Plant Cell Rep., 1984, 3, 7274. 127 H. Warzecha, P. Obitz and J. Stoeckigt, Phytochemistry, 1999, 50, 10991109. 128 H. Warzecha, I. Gerasimenko, T. M. Kutchan and J. Stoeckigt, Phytochemistry, 2000, 54, 657666. 129 M. Ruppert, S. Panjikar, L. Barleben and J. Stoeckigt, Acta Crystallogr., Sect. F, 2006, 62, 257260. 130 C. Mahidol, S. Ruchirawat, H. Prawat, S. Pisutjaroenpong, S. Engprasert, P. Chumsri, T. Tengchaisri, S. Sirisinha and P. Picha, Pure Appl. Chem., 1998, 70, 20652072. 131 K. Nakajima, T. Hashimoto and Y. Yamada, Proc. Natl. Acad. Sci. U. S. A., 1993, 90, 95919595. 132 K. Nakajima and T. Hashimoto, Plant Cell Physiol., 1999, 40, 1099 1107. 133 K. Nakajima, T. Hashimoto and Y. Yamada, Plant Physiol., 1993, 103, 14651466. 134 U. Richter, G. Rothe, A. K. Fabian, B. Rahfeld and B. Draeger, J. Exp. Bot., 2005, 56, 645652. 135 B. Draeger, Phytochemistry, 2006, 67, 327337. 136 R. Keiner, H. Kaiser, K. Nakajima, T. Hashimoto and B. Draeger, Plant Mol. Biol., 2002, 48, 299308. 137 T. Hashimoto, K. Nakajima, G. Ongena and Y. Yamada, Plant Physiol., 1992, 100, 836845. 138 K. Nakajima, T. Hashimoto and Y. Yamada, J. Biol. Chem., 1994, 269, 1169511698. 139 A. Yamashita, K. Nakajima, H. Kato, T. Hashimoto, Y. Yamada and J. Oda, Acta Crystallogr., Sect. D, 1998, 54, 14051407. 140 K. Nakajima, A. Yamashita, H. Akama, T. Nakatsu, H. Kato, T. Hashimoto, J. Oda and Y. Yamada, Proc. Natl. Acad. Sci. U. S. A., 1998, 95, 48764881. 141 H. Joernvall, B. Persson, M. Krook, S. Atrian, R. Gonzalez-Duarte, J. Jeffey and D. Ghosh, Biochemistry, 1995, 34, 60036013. 142 A. Yamashita, H. Kato, S. Wakatsuki, T. Tomizaki, T. Nakatsu, K. Nakajima, T. Hashimoto, Y. Yamada and J. Oda, Biochemistry, 1999, 38, 76307637. 143 A. Yamashita, M. Endo, T. Higashi, T. Nakatsu, Y. Yamada, J. Oda and H. Kato, Biochemistry, 2003, 42, 55665573. 144 M. Mahendra, N. Singh, P. Kaur, S. Sharma and T. P. Singh, to be published. 145 M. A. Schumacher, M. C. Miller, S. Grkovic, M. H. Brown, R. A. Skurray and R. G. Brennan, Science, 2001, 294, 21582163. 146 M. L. Raves, M. Harel, Y. P. Pang, I. Silman, A. P. Kozikowski and J. L. Sussman, Nat. Struct. Biol., 1997, 4, 5763. 147 H. Haviv, D. M. Wong, H. M. Greenblatt, P. R. Carlier, Y. P. Pang, I. Silman and J. L. Sussman, J. Am. Chem. Soc., 2005, 127, 1102911036. 148 H. M. Greenblatt, G. Kryger, T. Lewis, I. Silman and J. L. Sussman, FEBS Lett., 1999, 463, 321326. 149 D. M. Wong, H. M. Greenblatt, H. Dvir, P. R. Carlier, Y. F. Han, Y. P. Pang, I. Silman and J. L. Sussman, J. Am. Chem. Soc., 2003, 125, 363373. 150 I. Horv ath, V. Harmat, A. Perczel, V. Pal, L. Nyitray, A. Nagy, E. and J. Ov Hlavanda, G. Naray-Szabo adi, J. Biol. Chem., 2005, 280, 82668274.

Downloaded on 24 July 2011 Published on 24 October 2007 on http://pubs.rsc.org | doi:10.1039/B711935F

1400 | Nat. Prod. Rep., 2007, 24, 13821400

This journal is The Royal Society of Chemistry 2007

You might also like