摘要:

Natural Product Reports Editorial Board Professor T. J. Simpson (Chairman) Dr J. R. Hanson Dr R. B. Herbert Professor J. Mann Professor D. J. Robins Dr C. J. Schofield Dr D. A. Whiting Editorial Staff Dr Sheila R. Buxton Managing Editor Mrs Nicole Morgan Deputy Editor Miss Nicola P. Coward Production Editor Miss Carmel McNamara Technical Editor Miss Daphne E. Houston Miss Karen L. White Edito ria I Secretaries University of Bristol University of Sussex University of Leeds U n ive rs ity of Read in g University of Glasgow University of Oxford U n iversity of Notti ng ham Editorial Office The Royal Society of Chemistry Thomas Graham House Science Park Milton Road Cambridge UK CB4 4WF Telephone +44 (0)1223 420066 Facsimile +44 (0)1223 420247 E-mail rscl @rsc.org RSC Server http://c hem istry.rsc.org/rsc/ Natural Product Reports is a bimonthly journal of critical reviews. It aims to foster progress in the study of bioorganic chemistry by providing regular and comprehensive reviews of the relevant literature published during well-defined periods. Topics include the isolation structure biosynthesis biological activity and chemistry of the major groups of natural products -alkaloids terpenoids and steroids aliphatic aromatic and 0-heterocyclic compounds. This is augmented by frequent reviews of the wider context of bioorganic chemistry including developments in enzymology nucleic acids genetics chemical ecology primary and secondary metabolism and isolation and analytical techniques which will be of general interest to all workers in the area.Articles in Natural Product Reports are commissioned by members of the Editorial Board or accepted by the Chairman for consideration at meetings of the Board. Natural Product Reports (ISSN 0265-0568) is published bimonthly by The Royal Society af Chemistry Thomas Graham House Science Park Milton Road Cambridge UK CB4 4WF. 1996 Annual Subscription Price EEA f325.00 USA $615.00 Rest of World f333.00. Change of address and orders with payment in advance to The Royal Society of Chemistry The Distribution Centre Blackhorse Road Letchworth Herts. UK SG6 1HN. Air Freight and mailing in the USA by Publications Expediting Service Inc. 200 Meacham Avenue Elmont NY 11003.US Postmaster send address changes to Natural Product Reports Publications Expediting Service Inc. 200 Meacham Avenue Elmont NY 11003. Periodicals postage paid at Jamaica NY 11431-9998. All other despatches outside the UK are by Bulk Airmail within Europe and Accelerated Surface Post outside Europe. Printed in the UK. Members of the Royal Society of Chemistry should order the journal from The Membership Manager The Royal Society of Chemistry Thomas Graham House Science Park Milton Road Cambridge UK CB4 4WF. 0 The Royal Society of Chemistry 1996 All Rights Reserved No part of this publication may be reproduced stored in a retrieval system or transmitted in any form or by any means electronic mechanical photographic recording or otherwise without the prior permission of the publishers. Printed in Great Britain by the University Press Cambridge Subscription rates for 1996 EEA f325.00 USA $615.00 Rest of World f333.00

ISSN:0265-0568    

DOI:10.1039/NP99613FX021

出版商:RSC    

年代:1996

数据来源: RSC

ISSN:0265-0568    

DOI:10.1039/NP99613BX023

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

ISSN 0265-0568 NPRRDF 13(6) 469-535 (1996) Natural Product Reports A journal of current developments in bioorganic chemistry Volume 13 Number 6 CONTENTS ... 111 Hot off the Press Robert A. Hill and Andrew R. Pitt Reviewing the recent literature on natural products and bioorganic chemistry 469 The Glycopeptide Story -How to Kill the Deadly 'Superbugs' Dudley H. Williams 479 Catalytic Antibodies -Reaching Adolescence? Neil R. Thomas Reviewing the literature published up to the end of February 1996 513 Pigments of Fungi (Macromycetes) Melvyn Gill Reviewing the literature published between September 1992 and February 1996 529 The Sesterterpenoids James R. Hanson Reviewing the literature published between November 1991 and March 1996 Cumulative Contents of Volume 13 Number I 1 Modern Bioassays using Metal Chelates as Luminescent Probes Peter G.Sammes and Gokhan l'ahioglu 19 The DAP Pathway to Lysine as a Target for Antimicrobial Agents (icp to Sc~ptcviihcrIYYS) Russell J. Cox 35 The Biosynthesis of Plant Alkaloids and Nitrogenous Microbial Metabolites (1YY4) Richard B. Herbert 59 Diterpenoids (IYY4) James R. Hanson 73 Book Reviews Anticwic~t~r Dricg.~from Aninia/.v. Plants. rind Microorgrirrisrii.s. by George R. Pettit Fiona H. Pierson and Cherr! I.. Herald (reviewed by John Mann) O.\-it/titiixJ Strcss (inti Antio.\-itltrnt Dqfiwsrs in Biologj*.ed. S. Ahmad (reviewed by David J. Robins) -It/iwti(,c,\ in .Vitro,qw I~r.tcJro(:i.c,/c.s ( C'o/iiriir 1 ). ed. C. J. Moody (reviewed by Joseph P.Michael) Kumber 2 75 Marine Natural Products (IYY4) D. John Faulkner 197 /I-Phenylethqlamines and the Isoquinoline Alkaloids (Jrr/j. lYY4 to Jictic 19'95) K. W. Bentle) IS1 Triterpenoids (IYY4) Joseph D. Connolly and Robert A. Hill I7 1 Amaryllidaceae and Sw/cfiiirii Alkaloids (1YY4) John R. Lewis Yumber 3 177 The Biosqnthesis and Degradation of Thiamin (Vitamin B,) (Jcinirtir). IYXfi lo Jtinictir.i. lYY6) Tadhg P. Begley 1x7 Pyrroliidine Alkaloids (Ji,/.i. 19Y4 to Jimc 1Y95) J. Richard Liddell 195 Monoterpenoids (lYY1. IYY2 cintlptirt of IYY3) David H. Grayson 137 Steroids Reactions and Partial Synthesis (1YY4) James R. Hanson 341 Recent Progress in the Chemistry of Non-monoterpenoid Indole Alkaloids (Jir/j 1YY4 to JirnrJ 1YY5) Masataka lhara and Keiichiro Fukumoto 363 Book Review Etiz.i.tw C'titti/.\..yis in Orgritiic Sjxrhesis A Conip"c'hcti.rii,c HLititlbooX.by K. Draur and H. Waldmann ( re\.ieued bq Da\,id R. Kelly) Yumber 4 765 Dietary Antioxidants in Disease Prevention Michael H. Gordon 175 Recent Advances in Annonaceous Acetogenins (up to Jtrnuar~i.1Y96) Lu Zeng Qing Ye Nicholas H. Oberlies Guoen Shi. Zhe-Ming Gu Kan He and Jerry L. McLaughlin 307 Natural Sesquiterpenoids ( IYY4) Braulio M.Fraga 337 Recent Progress in the Chemistry of the Monoterpenoid Indole Alkaloids (Jir/.i.1YY4 to Drwwihcr 1Y95) J. Edwin Saston \umber 5 365 Decanolides. 10-membered Lactones of Natural Origin (1Y75 to 1Y95) Gerald Drager Andreas Kirschning Ralf Thiericke and Ylarion Zerlin 377 Natural Guanidine Derivatives (1YY4 tititi 19Y5) Roberto G.S. Berlinck 4 1 1 Oligomeric Proanthocyanidins Naturally Occurring 0-Heterocycles (Jrrnurirj,IYY.? to Dccw?iber lYY5) Daneel Ferreira and Riaan Bekker 43 5 M uscarine Imidazole. Oxazole. Thiazole and Peptide Alkaloids and Other Miscellaneous Alkaloids (Ju/.i. IYY3 to Jurw 1YY4) John R. Lewis Number 6 469 The Glycopeptide Story How to Kill the Deadly 'Superbugs' Dudley H. Williams ~ 479 Catalytic Antibodies -Reaching Adolescence? (upto Feb. 1996) Neil R. Thomas 513 Pigments of Fungi (Macromycetes) (1992 to 1996) Melvyn Gill 529 The Sesterterpenoids ( 199I to 1996) James R. Hanson Articles that will appear in forthcoming issues include Glycopeptide Alkaloids (Jtirrrrurj. 1985 to Drcw?rhrr 1995) D.C. Gournelis G. G. Laskaris and Robert Verpoorte Brassinosteroids Shozo Fujioka and Akira Sakurai Lignans Keolignans and Related Compounds (Jrrtiucirj 1994 to Dcwniht>r1995) Robert S. Ward Quinolinc. Quinazoline and Acridone Alkaoids (Ju/.i. 1YY4 to June 1995) Joseph P. Michael Indolizidine and Quinolizidine Alkaloids (Ji//). 1YY4 to Jictir. 1YY.5) Joseph P. Michael Recent Adunccs in Chemical Ecology (Ju/j /YY2 to Dcvewiber 1995) J. B. Harborne The Role of Carbohydrates in Biologically Active Natural Products Alexander C. Weymouth-Wilson The Biosynthesis of C,-C, Terpenoid Compounds (1993-1995) Paul M. Dewick Fatty Acids Fatty Acid Analogues and their Derivatives (1988 ro 1995) Marcel S. F. Lie Ken Jie Mohammed Khysar Pasha and M. S. K. Syed-Rahmatullah Diterpenoid and Steroidal Alkaloids (t~itl-lY44to the hcgintrittg of' lYY6) Atta-ur-Rahman and M. lqbal Choudhary Natural Sesquiterpenoids (IYYS) Braulio M. Fraga

ISSN:0265-0568    

DOI:10.1039/NP99613FP025

出版商:RSC    

年代:1996

数据来源: RSC

ISSN:0265-0568    

DOI:10.1039/NP99613BP027

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

The Glycopeptide Story -How to Kill the Deadly 'Superbugs' Dudley H. Williams Cambridge Centre for Molecular Recognition Department of Chemistry 1ensfield Road University of Cambridge Cambridge CB2 IEW UK Introduction The Rise and Partial Fall of Antibiotics Vancomycin and Glycopeptide Structure Determination Determination of the Binding Site of Glycopeptides for Bacterial Cell Wall Precursors New Twists in the Mechanism of Action of the Glycopeptides Simulating the Cooperative Binding of Antibiotics to Simple Models of Bacterial Cells The Operation of Targeting Devices in the Action of a New Anti biotic Comments on the Cooperativity and Conclusions The Gulliver Effect References and Footnotes I Introduction In 1967 the Surgeon General of the United States told a meeting in the White House that 'the time has come to close the book on infectious diseases.' The Surgeon General could perhaps be excused for his optimism as antibiotics and vaccines appeared to be sweeping all before them. In making this statement the Surgeon General failed to appreciate the power of natural selection and in particular he (and most others) did not appreciate how such selectional processes can work relatively quickly in bacteria. Bacteria are capable of passing genetic information between species through a process called conjugation. Suppose two bacteria can come into physical contact. If the first bacterium has a feature which renders it Dudley Williams received his bachelor and Ph.D. degrees in chemistry and organic chemistry respectively from the University of Leeds in 1958 and 1961.He sub- sequently studied as a research associate (with C.Djerassi) at Stanford University California from 1961-64 at which time his interest in the application of NMR and mass spectrometry to the solution of structural problems in organic chemistry developed.In 1964 he accepted an appointment at the University of Cambridge (where he is now Professor of Biological Chemistry) and became a Fellow of Churchill College. His work at Cam-bridge has included the development of shift reagents elucida- tion of the structure of a metabolite of vita- min D which is a human hormone and elucidation of the structure and the mode of action of the glycopeptide group of antibiotics.resistant to a given type of antibiotic then that feature can be transmitted to the second (of a different species) through the exchange of genetic material (Figure 1). This property and the relatively short generation time of bacteria have been important in leading to the current situation in which we are losing the fight against pathogenic bacteria; alarming numbers of people are losing their lives. It is against the above background that I wish to present the role that the glycopeptide antibiotics have played in therapy during the last 40 years and how in my research group in Cambridge we were able to determine the molecular basis for their mode of action. I will then discuss how in recent times resistance of bacterial pathogens against the most important member of the glycopeptide group vancomycin has emerged.Finally I will illustrate how the principles of action of a new semi-synthetic glycopeptide -which is showing remarkable activity against bacteria which currently cause lethal infec- tions -have emerged from our recent work. 2 The Rise and Partial Fall of Antibiotics Many antibiotics including the penicillins and the glyco- peptides interfere with enzymes of cell wall synthesis. Bacterial cell wall is built up on the outside of the cell in two main steps.' First disaccharide units with pendant peptides (mucopeptides) are exported from the cytoplasm to the outside of the cell where they are joined together by a transglycosylase enzyme (Figure 2).Second for mechanical strength these long polysaccharide chains are linked together through their peptide chains by a transpeptidase (Figure 3). The transpeptidase recognizes the sequence -D-alanyl-D-alanine at the end of the peptide chain cleaves off the terminal alanine and joins the remainder to a peptide chain from an adjacent polysaccharide. Penicillin is believed to mimic the structure of the transition state for the reaction of the -D-alanyl-D-alanine sequence when Because it is bound to the transpeptidase enzyme (Figure 4).2,3 of this structural similarity the enzyme binds penicillin and when it does so the strained and reactive peptide bond of the P-lactam ring is opened by the hydroxyl of a serine residue in the binding pocket of the enzyme (cf.Figure 5). The result of this reaction is that the opened penicillin is covalently bound to the active site of the enzyme which can take no further part in cell wall synthesis.l The bacteria die as a result of this cessation of cell wall synthesis. The success of penicillin led to its widespread use throughout the 1940s and early 1950s. During this time some pathogenic bacteria evolved such that they could release an enzyme when they were treated with penicillin. This enzyme dubbed penicillinase exploited the reactivity of the p-lactam ring opening it up before the antibiotic could get to the site of cell wall synthesis thus making it inactive. This was the driving force behind the development of semi-synthetic penicillins such as methicillin which retained the p-lactam ring but had structural changes elsewhere (R in Figures 4 and 5) such that the penicillinase enzyme became ineffective.However within just two years of the introduction of methicillin resistance to this antibiotic had emerged. As the bacteria evolved their penicillinases were tailored to inactivate methicillin and consequently these enzymes became known as the P-lactamases. In 1961 the outbreaks of methicillin resistant Staphylococcus 469 NATURAL PRODUCT REPORTS 1996 harmless bacterium Pathoaenic pathogenic resistant colony conjugation drug resistant bacteri um Figure 1 Drug resistance can be passed horizontally between species of bacteria. For example a non-pathogenic but antibiotic resistant bacterium can physically join with a disease causing bacterium and pass on the genetic information required for antibiotic resistance.With the advantage of drug resistance the pathogenic bacterium could then proliferate into an untreatable disease-causing colony. cytoplasm chain A precursor TRANSGLYCOSYLASE I I’ =I composition of precursor unit for cell wall synthesis G-G I G-G Figure 2 A disaccharide unit (mucopeptide) synthesized on the inside of the bacterium is exported through the membrane to the outside of the cell where it is joined to the growing polysaccharide chain of the cell wall by a transglycosylase enzyme. (NAG-NAM N-acetyl-glucosamine-N-acetylmuramate; the constitution of the pendant peptide is that found in Staphylococcus aureus.A = L-alanine; DA = D-alanine; DE = D-glutamic acid; G = glycine; K = lysine.) aureus (MRSA) were sporadic and the introduction of new antibiotics in subsequent years kept the resistant bacteria at bay. Since then MRSA have acquired resistance to virtually all of the antibiotics in clinical use including cephalosporins tetracyclines aminoglycosides erythromycin and the sulfon- amide~.~ MRSA is commonly found in hospitals and is a serious pathogen which causes a large number of deaths. One of the few groups of antibiotics that are effective against MRSA is the vancomycin group of glycopeptide antibiotics. Two members of the group are in clinical use today -vancomycin 1 and teicoplanin 2.Indeed along with gentamycin these glycopeptides are the antibiotics of last resort in our hospitals. cytoplasm TRANSPEPTIDASE (wlth the loss of the terminal -DAla) new peptide bond makes crosslink Figure 3 Mechanical strength is conferred to the cell wall through the crosslinking of the polysaccharide chains. The transpeptidase enzyme makes a new peptide bond between the pentaglycine of one saccharide chain and the penultimate -malanine of another chain (with the loss of the terminal -D-alanine). 3 Vancomycin and Glycopeptide Structure Determination Vancomycin was discovered in a soil sample from the jungles of Borneo during a research programme carried out by the American pharmaceutical giant Eli Lilly in the mid-1950s. It is produced by a microorganism as a secondary metabolite that is a compound which does not appear to be essential in the internal economy of the producing organism.It was first used clinically in 1959 although its early use was somewhat limited by side-effects. For example when vancomycin was admin- istered intravenously phlebitis (inflammation .of vein walls) would on occasion occur near the point of injection and also its use was not recommended in those with any history of hearing difficulties or kidney failure. These side-effects were NATURAL PRODUCT REPORTS 1996-DUDLEY H. WILLIAMS Figure 4 (a) The structure of penicillin (R is variable). (b) A conformation of the sequence -D-Ala-D-Ala which is mimicked by penicillin. R )co HN.0-lactamase * ? coo-Ser L+l C)-lactamase Figure 5 The hydroxy group of a serine residue in the binding pocket of the 0-lactamase opens up the reactive lactam ring thus deactivating the penicillin. much reduced when purer vancomycin became available in later years. The first structural work on vancomycin was carried out by Marshall and reported in 1965.5 The work indicated the presence of N-methylleucine glucose and chlorophenols. However progress was slow at this time because the methods needed to solve such a complex structure were not yet available. An important finding even before the structure of vancomycin was known was that of Perkins who showed that the antibiotic binds to bacterial cell wall mucopeptide precursors terminating in the sequence -L-lysyl-D-alanyl-D-alanine.6 At the time of this finding (1 969) little about the molecular basis of action of vancomycin could be inferred because the structure of the antibiotic was as yet unknown.My own group carried out intensive researches on the vancomycin structure in the period 1972-81. This appears to have been the first work to use the negative nuclear Overhauser effect (NOE) in proton NMR to derive extensive structural information on a relatively large secondary metabolite of truly unknown structure. The use of the NOE represented a major advance at the time because this method though not so precise as X-ray crystallography could be applied in solution and could give approximate distances between hydrogen atoms which were close together in the vancomycin structure.We were able to obtain a partial structure by 1977.$ Crucial information on the stereochemistry of the molecule was then derived from an X-ray study of a degradation product of vancomycin (CDP-l).' Early structures were incorrect insofar as they incorporated an iso-aspartic acid (rather than an 47 1 aspartic acid) residue and in misrepresenting the stereo-chemistry of a chlorinated aromatic ring. The former error was corrected by Harris and Harri~,~ and the latter by ourselves,1o to give the now accepted structure 1. In a shorter but overlapping period we were also able to elucidate the structure of a second glycopeptide ristocetin A.'l Although much less important than vancomycin this material is used in the diagnosis of von Willebrandt's disease (characterized by the absence of a blood clotting factor).By the early 1980s proton NMR spectroscopy had become so powerful that our third structure elucidation -that of teicoplanin -could be carried out relatively quickly.12 Although teicoplanin is not quite as important clinically as vancomycin it is extensively used in a similar manner and has the advantage of being cleared more slowly from the body. Its structure 2 incorporates a hydrocarbon sidechain attached to a glucos- amine residue on the fourth amino acid of teicoplanin. At the time of writing the structures of about 100 glycopeptides are established. H oH H HO.' "CH20H OH 2 teicoplanin 4 Determination of the Binding Site of Glycopeptides for Bacterial Cell Wall Precursors With the structure of vancomycin in hand it was then possible to elucidate the molecular basis for binding of the mucopeptide precursors terminating in -L-Lys-D-Ala-D-Ala to the antibiotic.Figure 6 The binding interaction between antibiotic (top) and bacterial cell wall analogue N-acetyl-D-Ala-D-Ala (below). Hydrogen bonds between the two are represented by dotted lines. We did this for ristocetin A in 1980,13 for vancomycin in 198314 and for teicoplanin in 1985.15 In all cases the method used was proton NMR spectroscopy carried out on the complex formed between the antibiotics and N-acetyl-D-Ala-D-Ala (or di-N- acetyl-L-Lys-D-Ala-D-Ala). Four key methods were used.First we examined the chemical shifts of the alanine methyl groups of N-acetyl-D-Ala-D-Ala when this bacterial cell wall analogue was both bound to the antibiotic and isolated from it. The changes in these chemical shifts indicated the location of these methyl groups over substituted benzene rings of the antibiotics. Second we used the changes in chemical shifts of amide NH resonances of the antibiotic to indicate which of these were involved in binding the carboxylate group of the cell wall analogue. Third we determined which amide NH groups of the antibiotic were protected from exposure to water in the antibiotic-cell wall analogue complex. This experiment allowed us to determine all of those NHs which were involved in binding the antibiotic to the cell wall analogue.Fourth we were able to determine which protons of the antibiotic were close in space to identified protons of the cell wall analogue. This involved using the NOE as described above although this time it was done to estimate distances between protons of the two different molecules of the complex (rather than inter-proton distances within one mol- ecule). These methods led to the model of the binding interaction which is reproduced in Figure 6. 5 New Twists in the Mechanism of Action of the Glycopeptides The work described at the end of the preceding section seemed to solve the problem of the molecular basis of the mode of action of the glycopeptides. However in 1989 we discovered that the antibiotic ristocetin A forms a dimer,16 and we were able to elucidate the structure of the peptide portion of the dimer (although not at this stage the orientation of the saccharide portions of the molecule with respect to peptide within this dimer).In the dimer 3 the peptide backbones of the two antibiotic molecules are indicated in bold outline and the binding sites are shown occupied by two molecules of the cell wall analogue di-N-acetyl-L-Lys-D-Ala-D-Ala (DALAA). The two antibiotic molecules are bound together by four amide- amide hydrogen bonds and the ammonium ion of the amino sugar (ristosamine) which is attached to residue six of the NATURAL PRODUCT REPORTS 1996 3 antibiotic forms a hydrogen bond to an amide carbonyl group in the other half of the dimer.Thus there is a total of six hydrogen bonds at the dimer interface.16-ls Most interestingly the process of dimerization brings the ammonium ion of the ristosamine residue into the proximity of the carboxylate ion of the cell wall peptide precursor as shown in 3. These groups respectively positively and negatively charged do not directly form a salt bridge but rather form such an interaction mediated through an amide group. The Coulombic attraction between these groups leads to the expectation that the dimer may bind the cell wall precursor peptide more strongly than the antibiotic monomer (since the carboxylate anion of the cell wall precursor will be ‘atttracted’ into its binding site by the opposite charge on the sugar in the other half of the dimer).This expectation is realised for all the antibiotics so far examinedlg (with the exception of ristocetin A where the binding of the bacterial cell wall analogue into the first binding site of the antibiotic dimer is cooperative but the binding into the second site is slightly anti-cooperative). Conversely antibiotic dimerization is promoted by binding of cell wall analogue. A specific example is the dimerization constant of the glycopeptide antibiotic eremomycin :19 K = 3 x lo6 and 3 x los dm3m01-’ in the absence and presence respectively of the cell wall analogue di-N-acetyl-L-Lys-D-Ala-D-Ala. Construction of a thermodynamic cycle shows that these values imply for two cell wall analogue binding sites of equal affinity in the dimer (a situation which appears to hold for eremomycin) that the antibiotic dimer binds this cell wall analogue with an affinity ten times greater than the antibiotic monomer .I9 The above findings strongly suggest that the dimer structure may be of physiological relevance.This possibility is also suggested by the observations that the sugars of the glyco- peptides are of such structures and located at such points on the peptide portions of the antibiotics that they promote dimerization. 17+1s Additionally where measurements have been made those antibiotics which contain a chlorine atom attached to residue 2 dimerize more strongly than those lacking this chlorine. All these observations are consistent with the idea of a selectional pressure to produce antibiotics which dimerize and such a selectional pressure would of course only operate if the dimerization had a functional role in aiding the survival of the producing strain.It is interesting to look back and note that even before the above knowledge (relating structural features to dimerization propensity) was available we suggested in 1989 that the dimer may be of physiological significance.16 We suggested this since delivery of a dimer to the site of bacterial cell wall biosynthesis with which the glycopeptides interfere would simultaneously place two molecules of the antibiotic at the crucial site for NATURAL PRODUCT REPORTS 1996DUDLEY H. WILLIAMS (A) monomer Figure 7 Antibiotics binding as (A) monomer (B) dimer and (C) with a membrane anchor.Open or shaded circles represent saccharide units and the end portion of the attached bold line represents the sequence -L-Lys-D-Ala-D-Ala. The bold line not attached to the saccharide represents di-N-acetyl-L-Lys-D-Ala-D-Ala (tripeptide) in free solution; this acts as a ligand and antagonist to the antibiotic. (A) The binding of a single antibiotic molecule to the growing cell wall is a simple biomolecular association [such that externally added ligand can replace the cell wall peptide]. (B) The antibiotic dimer can benefit from an essentially intramolecular association at the surface of a bacterium [and it is more difficult to disrupt by the externally added ligand]. (C)The benefit from intramolecularity can also be exploited if the antibiotic has a membrane anchor [again the complex is more difficult to antagonize].action. By 1993-94 we had in fact shown that all the glycopeptides which we were able to examine with the exception of teicoplanin formed dimers ;measurement also had revealed very large dimerization constants for some of the antibiotics (see the horizontal scale of Figure 9 detailed discussion of which comes late^).^',^^ The generality of the phenomenon of dimerization led us to propose that if a dimer became bound to one -L-Lys-D-Ala-D- Ala fragment of a growing cell wall then the binding event in which a second fragment of the growing cell wall could bind to the same dimer would effectively be intramolecular (Figure 7B to be contrasted with Figure 7A).I9 Additionally we noted that since teicoplanin has a hydrocarbon sidechain then it should in principle be able to target itself to the site of biosynthesis with which it interferes by using this hydrocarbon sidechain as a membrane anchor (Figure 7C).l9 Thus in this latter case dimerization is unnecessary because both the growing cell wall fragment and the antibiotic are attached to the membrane making binding an intramolecular event.Hence in the mode of action of the vancomycin group of antibiotics either dimer- ization (Figure 7B) or membrane anchoring (Figure 7C) preferentially locate antibiotic at the crucial site for action. It was not until early 1995 that we were able to demonstrate experimentally that the hypotheses with regard to the above locating devices seemed to be correct.Our experiments involved trying to antagonize the action of the antibiotics in killing bacteria on agar plates by the addition of external di-N-acetyl- L-Lys-D-Ala-D- Ma. Figure 7 also shows the basis for these experiments. Where binding is simply bimolecular the probability of locating the antibiotic at the appropriate site is relatively low (Figure 7A). In contrast whilst the first antibiotic molecule of a dimer has been located to bind to cell wall precursor the second antibiotic molecule in this dimer is already in an appropriate place to bind another growing piece of cell wall. It will therefore be more difficult to antagonize this kind of binding event by externally added tripeptide (Figure 7B). Finally where teicoplanin is located at the site of biosynthesis through use of a membrane anchor the binding of membrane-bound cell wall precursors to antibiotic should be favoured because of proximity effects and again subject to more difficult antagonism by externally added tripeptide (Figure 7C).In fact approximately five times as much externally added tripeptide is required to antagonize the action of teicoplanin relative to teicoplanin lacking a membrane anchor (TA,- 1). More strikingly the strongly dimerizing antibiotic eremomycin requires roughly a thousand times as much externally added tripeptide to antagonize its action to a similar level as TA,-1.20 Representations of the actual inhibitions zones demonstrated by the various antibiotics strikingly confirm these remarkable effects (Figure 8).Indeed there is a striking correlation between the amount of externally added tripeptide which is needed to antagonize an antibiotic and its dimerization constant (Figure 9). 6 Simulating the Cooperative Binding of Antibiotics to Simple Models of Bacterial Cells The experiments described in the previous section give strong support to the promotion of glycopeptide antibiotic action by means of devices which preferentially locate the antibiotic at the site where it can interfere with cell wall biosynthesis. However it is desirable to demonstrate directly that when cell wall precursors bind to antibiotic in the previously described manner (Figure 7B) then the binding constant for the Figure 8 Three agar diffusion plates illustrating the effect of di-N-acetyl-L-Lys-D-Ala-D-Ala on the potencies of three representative antibiotics in inhibiting the growth of Bacillus subtilis.Each dark circle represents an area where the bacteria are killed by the antibiotic placed on the paper disk (white circles). The size of the dark circle is a measure of the potency of the antibiotic. Each paper disk contained 1 pg of antibiotic. (A) teicoplanin A,- 1 (does not dimerize or possess a membrane anchor) (B) vancomycin (dimerizes weakly) and (C)eremomycin (dimerizes strongly). The number superimposed on each disk is the amount of di-N-acetyl-L-Lys-D-Ala-D-Ala (pg) added to the antibiotics. 474 0 lo00 - 2 4 3 e m 0 5 '0 1 19 1 I 1 1 1 I I 0.1 10 lo00 105 107 Kdim /dm3 mol Figure 9 Plot of amount of di-N-acetyl-L-Lys-D-Ala-D-Ala required to give a 50% reduction in the potency of glycopeptide antibiotic versus dimerization constant Kdimof the antibiotic.Teicoplanin (entry no. 10) which possesses a membrane anchor and teicoplanin A,-1 (entry no. 9) without a membrane anchor show no evidence for dimerization (i.e. Kdinl< 1 dm3 mol-l). association is greater than in a simple bimolecular association (Figure 7A). This was shown by using a simple model system which can act as an analogue of a cell membrane. Molecules of the surfactant sodium dodecyl sulfate (SDS) aggregate in aqueous solution to give a structure known as a micelle. In this micelle the hydrocarbon tails of the molecule have a tendency to aggregate towards the centre of a roughly spherical structure in which the sulfate groups reside near the surface -where they can be solvated by water molecules.The structure thus generated (Figure 10A) is a crude model of a bacterial membrane. Next a fatty acid chain was attached to a cell wall mucopeptide precursor to give N-a-decanoyl-N-8- acetyl-L-Lys-D-Ala-D-Ala 4. This compound when in the presence of a micelle in aqueous solution can lower its free energy by inserting its hydrocarbon tail into the hydrocarbon portion of the micelle to give a crude mimic of a growing bacterium (Figure IOB). Several peptide structures can be incorporated into one pseudo-membrane in this way and since the hydrocarbon portion of the micelle has the fluidity characteristic of a liquid crystal two of these peptide structures can take up a geometry which might be appropriate to bind to a dimer without too much strain (Figure lOC cf.Figure 7B). NATURAL PRODUCT REPORTS 1996 If our hypothesis is correct then the binding affinity of a dimer glycopeptide to such an organized assembly should be greater than that of a simple bimolecular association in solution (Figure 7A). We were excited to find that when the binding of di-N-acetyl-D-Ala-D-Ala to ristocetin A in the presence of micelles is compared to the binding of N-a-decanoyl- N-&-acetyl-L-Lys-D-Ala-D-Ala4 to ristocetin A also in the presence of micelles the latter binding is greater by a factor of Thus the stronger binding of an antibiotic dimer to a simple analogue of the bacterial cell system has been directly demonstrated.ACHN' 4 In summary we have been able to demonstrate two distinct kind of cooperativity in binding of antibiotic dimers to bacterial cell wall precursor peptides. First a cooperativity that is expressed when both the binding components (cell wall analogue and antibiotic) are in free solution where at least part of the origin of this cooperativity is evident from the ionic interactions evident in structure 3. Second a cooperativity which is only expressed when at least two of the binding components have their relative motions reduced by sim-ultaneous attachment to the bacterial membrane. 7 The Operation of Targeting Devices in the Action of a New Antibiotic Illnesses due to methicillin resistant Staphylococcus aureus (MRSA) and vancomycin resistant enterococci (VRE) are on the These illnesses are often lethal.The VRE problem arises from bacteria which would not ordinarily be pathogenic to a healthy individual. However it is now commonplace for hospitalized individuals to be immuno-deficient. Immunodeficiency may arise due to cancer chemo- therapy AIDS or simply the impaired function that often goes hand-in-hand with weakness following operations or old age. The vancomycin resistant enterococci were first reported in 1989. It appears that these enterococci have been able to obtain genes from other bacteria such that the precursor from which their cell wall is built no longer terminates in -D-alanyl-D- alanine but rather terminates in -D-alanyl-D-lactate (-D-Ala-D- As a consequence the hydrogen bond which is normally made between the NH of the terminal D-alanine intramolecular association Figure 10 Schematic representation of (A) a micelle (B) N-cr-decanoyl-N-r-acetyl-L-Lys-D-Ala-D- Ala 4 using its membrane anchor and (C) an antibiotic dimer bound to the ligand at the surface of a micelle.NATURAL PRODUCT REPORTS 1996-DUDLEY H. WILLIAMS (A) Nacetyl-D-Ala-D-Ala antibotic binding pocket (B) Nacetyl-D-Ala-D-Lac c;l 9 Ye antibotic binding pocket Figure 11 In vancomycin resistant enterococci the terminal D-alanine of the immature cell wall has been replaced in part or essentially completely by D-lactate.This change drastically reduces the binding constant to vancomycin as an NH which can form a hydrogen bond is replaced by an 0 which cannot. group and a carbonyl group of the antibiotic (Figure 11A) can no longer be made. Instead it is replaced by a repulsive interaction between the oxygen of the C-terminal D-lactate group and the carbonyl group of the antibiotic (Figure 11B). The affinity of the glycopeptide antibiotics for the vancomycin resistant enterococci precursors which terminate in -D-Ala-D- Lac is hence decreased by a factor of the order of 1000(relative to the binding to precursors terminating in -D-Ala-D-Ala). Due to this markedly reduced binding constant vancomycin has little activity against bacteria using such precursors.Significant inhibition of VRE requires 10&1000 times as much vancomycin as is required in the treatment of sensitive organisms. Thus vancomycin is not a viable antibiotic in these cases and an initial conclusion might be that there is no hope for useful activity of glycopeptides against these serious pathogens. The above conclusion however reasonable it might seem at the outset is incorrect. Scientists at Eli Lilly in Indianapolis have recently taken the antibiotic chloroeremomycin,27 6 and added to this the hydrophobic tail which is shown in the resulting structure 5.28This molecule has remarkable activity against vancomycin resistant enter~cocci.~~ For example in a mouse model the effective dose in reducing colonization and clearing infection is typically 50 times lower than for vanco- my~in.~O The new compound also has good in vitro activity against MRSA being generally about eight times more active than vancomycin.28 It also has exceptional in vitro activity against penicillin resistant Streptococcus pneumoniae and in general its in vivo activity is commensurate with its in vitro potency.** The above facts regarding the activity of 5 raise a crucial question.How can it be that a compound that is not expected to bind strongly to -L-Lys-D-Ala-D-Lac precursors has re-markable activity? The apparent dilemma is made more clear by the observation that the new glycopeptide 5 and vancomycin both have low affinity for di-N-acetyl-L-Lys-D-Ala-D-Lac (both bimolecular binding constants lie in the range 500-1800 dm3 mol-l in contrast to vancomycin affinity of ca.lo6 dm3 mot1 for the cell wall analogue di-N-acetyl-L-Lys-D-Ala-D-Ala). One possible approach to try to understand this apparent anomaly would be to seek a completely new mechanism of action for the binding of glycopeptides which are active against bacteria utilizing D-Ala-D-Lac precursors. However it seemed 47 5 H& Me HO...A CI OH HO-OH 5 R=C1-i 6 R=H to us that we should first consider that the binding of glycopeptides to -D-Ala-D-Lac precursors might be strength- ened by subtle cooperative phenomena. This viewpoint was reinforced by our earlier work on the origins of binding affinities of cell-wall precursors to glycopeptides.This work indicates that the main source of binding affinity lies in the binding of the carboxyl group into a pocket containing 3 NHs (Figure 6). This conclusion is by no means self-evident. Acetate ion binds into the aforementioned pocket of 3 NHs with a binding constant of only ca. 10 dm3 mol-l and it is the further interactions with other portions of di-N-acetyl-L-Lys-D-Ala-D-Ala that increase the binding constant to ca. lo6 dm3 mol-'. Despite the fact that the latter number is dramatically larger than the former it is in the interaction of the carboxylate anion with the NH binding pocket (Figure 6) that much of the adverse entropy of binding is removed and it is therefore the intrinsic strength of this interaction that is greatest.Once this basic affinity is present then it may be increased by neighbouring interactions in two ways. First the neighbouring interactions add what is formally their own affinity to the binding. Second and crucially the neighbouring N-awl-KDADA 1oe 0 N-acetyl-DADA 0 N-awl-DA '1 100 acetate 0 free 1 0 I I I1,,,,,(,,,,,,,, ,,,( 7 8 9 10 111 Figure 12 Plot of the binding constant of peptide ligands to a vancomycin group antibiotic vs. the chemical shift of the NH proton wg. As the ligand becomes progressively longer the binding constant increases and the chemical shift of w2 appears further downfield -an indication that the hydrogen bond between w and the carboxylate is becoming stronger. (-DA = D-Ala; DADA = D-Ma-D-Ma; -KDADA = -L-Lys-D-Ala-D-Ala.interactions strengthen the binding of the carboxylate into its binding pocket. This effect is shown in Figure 12 each point on the graph is for a cell wall analogue starting with the simplest (CH,COO-) to have a terminal carboxylate (analogous to that of the terminal -D-Ala of a piece of bacterial cell wall) and proceeding along the series to N-acetyl-D-Ala thence to N-acetyl-D-Ala-D-Ala and finally to di-N-acetyl-L-Lys-D-Ala-D-Ala. In Figure 12 along the horizontal axis is plotted the chemical shift of one of the NH protons of the antibiotic (w2 that of the second amino acid from its N-terminus) which binds the carboxylate anion of the bacterial cell wall peptide (Figure 6). It is important to realize that the plotted chemical shift corresponds to fully bound antibiotic in each case such that in every antibiotic molecule the NH proton w forms a hydrogen bond to a carboxylate anion.Along the vertical axis is plotted the equilibrium constant for the binding of the cell wall analogues to the antibiotic as measured by UV spectrophotometry. It can be seen (Figure 12) that as the binding constants between the antibiotic and the cell wall analogues increase (as NATURAL PRODUCT REPORTS 1996 the cell wall analogues increase in length) the carboxylates of the various cell wall analogues hydrogen bond more and more strongly to the antibiotic (this conclusion is based simply on the assumption that a greater downfield shift of the NH reflects stronger hydrogen bond formation).Thus the experiments show how the tightness of binding of the carboxylate group into the antibiotic pocket which accepts it is increased by the cooperative binding provided by adjacent groups.31 The above demonstration that the binding of a functional group can be increased by other neighbouring groups which help to reduce the motion of the first in its binding site led us to the hypothesis that the affinity for -L-Lys-D-Ala-D-Lac precursors might be increased by alternative sources of cooperative interactions. It was felt that it should be possible to build on the strength of the weak interaction between the NH binding pocket and the carboxylate anion by dimerization of the antibiotic and by membrane attachment of the antibiotic as outlined earlier in this article.In particular it was felt that if these two features could operate simultaneously then the largest cooperative increase in binding would be observed. Figure 13 Schematic representation of compound 5 utilizing both a membrane anchor and dimerization to bind to cell wall terminating in D-lactate at the surface of a bacterium. The molecules constituting the bacterial cell membrane are shown as stick structures. The immature bacterial cell wall is shown in white and the two halves of the dimer in light and dark grey. Note the potential of the biphenyl groups (lower part of the antibiotic molecules) to act as a membrane anchor. Figure 14 Gulliver was tied down by many weak bindings. As with the association of molecules an array of weak interactions is an effective way to restrict motion and ensure strong net binding.NATURAL PRODUCT REPORTS 1996-DUDLEY H. WILLIAMS We hypothesize that 5 can bind to bacterial cell wall using the locating and partially immobilizing devices both of membrane anchoring and dimerization (Figure 13). 8 Comments on the Cooperativity and Conclusions The Gulliver Effect We have recently commented upon the fact that the decreased motion of a ligand in a binding site works to improve the electrostatic interactions that are formed in that binding In a reciprocal manner. features which improve the electro- statics of binding in a ligand-receptor interaction likewise reduce the degree of motion in the binding site. The two effects can be regarded as working iteratively on each other.For example if A binds in isolation to part of a binding site with an exothermicity X and B binds in isolation to another part of the binding site with an exothermicity Y then attachment of A to B so that they can simultaneously bind into the binding site in a cooperative manner is not only advantageous through the operation of the classical chelate effect but (in a strain-free system) leads to an exothermicity of binding greater than X+Y. A useful way of looking at this phenomenon is in terms of what one might call the ‘Gulliver effect’. If you wish to keep a close interaction between Gulliver’s shoulders and the ground then it is clear that an agitated Gulliver would be better constrained if he were also held at the waist and feet (Figure 14).The figure gives a crude but perhaps useful insight as to how the carboxylate group of a -D-Ala-D-Lac precursor can make a stronger interaction when other features that may reduce local motion (dimerization and membrane anchor) act at the same time. The work indicates how subtle locating devices can strengthen adjacent interactions and how important antibacterial activity can be recouped from a situation that may initially appear hopeless. Acknowledgements. The author wishes to thank EPSRC BBSRC and the Wellcome trust for financial support of this work at various stages. He also wishes to thank his research colleagues named in the references for their contributions; and Eli Lilly SmithKline Beecham Merill-Dow Lepetit and Abbott Laboratories for the samples of antibiotics.A recent exchange of information with scientists (Dr Norris Allen and Dr Thalia Nicas) at Eli Lilly is particularly acknowledged. 9 References and Footnotes 1 E. F. Gale E. Cunliffe P. E. Reynolds M. H. Richmond and M. J. Waring Molecular basis of antibiotic action Wiley London 1972. 2 D. J. Tipper and J. L. Strominger Proc. Natl. Acad. Sci. USA 1965 54 1133. 3 B. Lee J. Mol. Biol. 1971 61 464. 4 H. C. Neu Science 1992 257 1064. 5 F. J. Marshall J. Med. Chem. 1965 8 18. 6 H. R. Perkins Biochem. J. 1969 111 195. 7 D. H. Williams and J. R. Kalman J. Am. Chem. SOC. 1977 99 2768. 8 G. M. Sheldrick P. G. Jones 0. Kennard D. H. Williams and G.A. Smith Nature (London) 1978 271 223. 9 C. M. Harris and T. M. Harris J. Am. Chem. SOC. 1982 104 4293. 10 M. P. Williamson and D. H. Williams J. Am. Chem. SOC. 1981 103 6580. 11 J. R. Kalman and D. H. Williams J. Am. Chem. SOC. 1980,102 897. 12 J. C. J. Barna D. H Williams D. J. M. Stone T. C. Leung and D. M. Doddrell J. Am. Chem. SOC. 1984 106 4895. 13 J. R. Kalman and D. H. Williams J. Am. Chem. SOC. 1980 102 906. 14 D. H. Williams M. P. Williamson D. W. Butcher and S. J. Hammond J. Am. Chem. Soc. 1983 105 1332. 15 J. C. J. Barna D. H. Williams and M. L. Williamson J. Chem. SOC. Chem. Commun. 1985 254. 16 J. P. Waltho and D. H. Williams J. Am. Chem. SOC. 1989 111 2475. 17 J. P. Mackay U. Gerhard D. A. Beauregard R. A.Maplestone and D. H. Williams J. Am. Chem. SOC. 1994 116 4573. 18 U. Gerhard J. P. Mackay R. A. Maplestone and D. H. Williams J. Am. Chem. Soc. 1993 115 232. 19 J. P. Mackay U. Gerhard D. A. Beauregard M. S. Westwell M. S. Searle and D. H. Williams J. Am. Chem. SOC. 1994 116 4581. 20 D. A. Beauregard D. H. Williams M. N. Gwynn and D. J. Knowles Antimicrob. Agents Chemother. 1995 39 781. 21 M. S. Westwell B. Bardsley A. C. Try R. J. Dancer and D. H. Williams J. Chem. Soc. Chem. Commun. 1996 589. 22 P. Courvalin Antimicrob. Agents Chemother. 1990 34 2291. 23 G. M. Eliopoulos Eur. J. Clin. Microbiol. Infect. Dis. 1993 12 409. 24 A. P. Johnson A. H. C. Uttley N. Woodford and R. C. George Clin. Microbiol. Rev. 1990 3 280. 25 M. Arthur C.Molinas T. D. H. Bugg G. D. Wright C. T. Walsh and P. Courvalin Antimicrob. Agents Chemother. 1992 36 867. 26 C. T. Walsh Science 1994 261 308; see also C. T. Walsh S. L. Fisher I.-S. Park M. Prahalad and Z. Wu Chem. Biol. 1996 3 21. 27 Our nomenclature; also known as chloroorienticin and LY264826B. 28 T. I. Nicas J. E. Flokowitsch D. A. Preston D. L. Mullen J. Grissom-Arnold N. J. Snyder M. J. Zweifel S. C. Wilkie M. J. Rodriguez R. C. Thompson and R. D. G. Cooper in ICAAC Session 152 New glycopeptides San Francisco 1995 F248. 29 T. I. Nicas D. L. Mullen J. Grissom-Arnold N. J. Snyder M. J. Zweifel S. C. Wilkie M. J. Rodriguez R. C. Thompson and R. D. G. Cooper in ICAAC Session 152 New glycopeptides San Francisco 1995 F249. 30 C. J. Boylan T. I. Nicas D. A. Preston D. L. Zeckner B. J. Boyll R. A. Raab D. L. Mullen N. J. Snyder L. L. Zornes R. E. Stratford M. J. Zweifel S. C. Wilkie M. J. Rodriguez R. C. Thompson and R. D. G. Cooper in ICAAC Session 152 New glycopeptides San Francisco 1995 F255. 31 P. Groves M. S. Searle M. S. Westwell and D. H. Williams J. Chem. Soc. Chem. Commun. 1994 1519. 32 M. S. Westwell M. S. Searle and D. H. Williams J. Mol. Recog. 1996 9 88.

ISSN:0265-0568    

DOI:10.1039/NP9961300469

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Catalytic Antibodies -Reaching Adolescence? Neil R. Thomas Department of Chemistry University of Nottingham University Park Nottingham UK NG7 2RD Reviewing the literature published up to the end of February 1996 1 1.1 Introduction The Origins of Protein-based Catalysis 19.5 19.6 Aggregates and Micelles Zeolites 1.2 2 3 Early Attempts to Produce Catalytic Antibodies Catalysis by Desolvation Antibodies as Entropic Traps 19.7 20 20.1 Comparison of Systems Applications and Future Directions Lessons from the Past Decade 3.1 3.2 3.3 3.4 Chorismate Mutase Antibodies Oxy-Cope Antibodies Diels-Alder Antibodies Retro[Z.;rr+2771 Dimerization 20.2 20.3 20.4 21 Catalytic Antibody Substrate and Reaction Specificity Catalytic Antibody Rate Enhancement Current and Future Applications References 4 Hydrolytic Antibodies 4.1 Stabilization of the Transition State by Hydrogen Bonding to the Oxyanion 4.2 Utilization of Nucleophilic Catalysis 1 Introduction 4.3 4.4 4.5 4.6 5 6 6.1 6.2 6.3 6.4 7 7.1 7.2 8 9 10 11 12 12.1 12.2 13 14 14.1 14.2 14.3 14.4 15 15.1 16 17 17.1 17.2 17.3 17.4 17.5 18 Utilization of General Acid-Base Catalysis Phosphate Ester Hydrolysis Antibodies for Glycoside Hydrolysis Ester and Amide Bond Formation in Water Non-hydrolytic Antibodies using General Acid-Base Chemistry Cofactor Assisted Catalysts Natural Cofactors Redox Reactions Semi-synthetic Catalytic Antibodies Use of Cheap Inorganic Cofactors Rerouting Chemical Reactions using Antibodies Formation of the Thermodynamically More Stable exo Diels-Alder Adduct Controlling Cyclization Reactions Multistep Reactions Substrate Attenuation and Mutagenesis Antibodies in Organic Solvents Polyclonal Catalytic Antibodies Autoimmune Antibodies Peptide Hydrolysing Autoantibodies DNA Hydrolysing Autoantibodies Anti-idiotypic Catalytic Antibodies Creation of Catalytic Antibodies -Novel Immunization Procedures Use of Autoimmune Mice In Vitro Immunization Heterologous Immunization Reactive Immunization Immortalization and Expression of Antibodies and Antibody Fragments Immunoglobulin Gene Expression Libraries and Phage Display Antibody Expression Screening Antibodies for Catalysis Chromogenic Assays Complementation of Auxotrophic Bacteria and Yeast catELISA Detection of Catalytic Activity through Irreversible Inhibition PCR Amplification of DNA Tagged Substrates Antibody Mutagenesis It is now ten years since the first examples of catalytic antibodies were reported in the literature.They heralded a new era of research centred on the creation and screening of large libraries of related compounds as a means of identifying specific chemical effector molecules (an example of com-binatorial biochemistry). Catalytic antibodies have also pro- vided the first vehicle for studying the evolution of a biological catalyst in realtime. The review highlights the significant advances that have been made in the development of antibody based catalysts (abzymes) as they enter their second decade. The developments in hapten (antigen immunogen) design antibody generation and methods of screening for catalysis that have been developed over the past decade are summarized.Antibodies have been generated that catalyse reactions corresponding to all of the major enzyme classes as defined by the International Union of Biochemistry with the exception of class 6 (nucleotide dependent ligases). Other developments have included several examples of antibody catalysed reactions with no known enzyme equivalent and the creation of antibodies which use simple inorganic redox reagents to conduct regio- and stereo- specific oxidation and reduction reactions so circumventing the requirement for expensive biological cofactors such as NADH. Detailed mechanistic studies have been conducted on several of the catalytic antibodies. These indicate that antibodies provide ideal vehicles for studying the fundamentals of protein based molecular recognition and catalysis and have allowed the experimental testing of many of the hypotheses formulated over the past century for how enzymes work.Whilst the overall cost and efficiency of catalytic antibodies currently precludes their use as preparative scale catalysts progress into the understanding of the methods of eliciting abzymes and cheaper methods of producing antigen binding fragments could soon make this a reality. Key developments in other aspects of immunology and molecular recognition pertinent to the future success of antibody based catalysts are also reviewed. A brief comparison of catalytic antibodies with other receptor based catalysts is presented to indicate the ‘State of the Art’ and highlight the relative strengths and weaknesses of the current systems. 19 19.1 Other Catalytic Biomimetic Systems Protein and Peptide Catalysts 1.1 The Origins of Protein-based Catalysis 19.2 19.3 19.4 Imprinted Polymers and Synzymes Ribozymes Macrocyclic Receptor Based Catalysts Catalytic antibodies have been developed as a consequence of our efforts to understand the processes involved in enzymatic catalysis.Primarily the observation that the rate enhancement 479 --------.I-..I-. ----.------..-.. ...-._-. E*P\ --I-I-. .--------__ E+P E+P uncatalysed II enzyme catalysed Figure 1 Energy profiles for uncatalysed and enzyme catalysed reactions of a substrate S imparted by enzymes is due in part to the high affinity of the enzyme (E) for a distorted form of the substrate (S) that occurs at a crucial point in its transformation to products (Figure 1).This idea was first recorded in discussions on catalysis by Michael Polanyi,l and developed by Linus Pauling2 who speculated that it should be possible to produce highly potent enzyme inhibitors that resembled this distorted substrate structure. In the late 1960s Jencks and Wolfenden inde-pendently revived this concept that inhibitors resembling the structure of the transition state (TS) for a reaction should bind more tightly to the active site of an enzyme than analogues of the substrate in the ground state. Jencks3identified many such inhibitors in the literature for which he coined the term ‘transition-state analogue ’,while Wolfenden4first developed a thermodynamic analysis that gave the concept a theoretical basis-see eqn (9)-and applied this to the development of triosephosphate isomerase inhibitors.The following is a thermodynamic cycle that interrelates substrate and transition state binding for catalysed and uncatalysed reactions KES? kES E-S Z [E-S]l E*P -+ Using eqn. (l) we can derive the dissociation constant for the enzyme-substrate complex K and that for the enzyme-transition state complex K, The equilibria between the substrate and transition state for the catalysed and uncatalysed reactions are represented by KEsI and KuS respectively (3) At equilibrium KEsTx K, = K x Kul or (4) and the rate of formation of product in the catalysed reaction is given by t K,, = Michaelis constant dissociation constant for enzyme substrate complex; K = dissociation constant for enzyme inhibitor complex; k,, = rate constant for reaction in presence of enzyme/antibody; k,,,, = rate constant for reaction in absence of enzymelantibody.NATURAL PRODUCT REPORTS 1996 where K is the Boltzmann constant T is absolute temperature and h is Planck’s constant. Now as [EeS]’ = KE:[E-S] from eqn. (3) then and then Provided that the catalysed and uncatalysed reactions pass through similar mechanisms and if the parameters are determined under conditions for which the Michaelis-Menten relationship is obeyed by the catalyst then kEs= kcat Ks = K, k = kuncat and K, = Ki (8)t so that (9) In other words differential binding = rate enhancement The boundary condition set by Wolfenden’s hypothesis was that the rate of an enzyme catalysed reaction could be faster than that of the non-enzymatic reaction to the extent and only to the extent that the binding of the protein to the distorted substrate in the transition state was tighter than binding to the substrate in the ground state.The rate enhancements observed for enzymes have been shown to fall in the range 5 x lo5 (human cyclophilin) to 10’’ (orotidine 5’-phosphate decarboxy-lase alkaline pho~phatase).~ Over the intervening twenty-five years a wealth of new specific enzyme inhibitors have been developed to exploit this observation6 and a significant amount of structural and kinetic data produced to substantiate its existence.’ 1.2 Early Attempts to Produce Catalytic Antibodies In the late 1960s many molecules had been identified as transition state analogues from (a) their potency as enzyme inhibitors and (b) their presumed structural similarity to intermediates in the rate determining transformations.Several groups used these molecules as antigens for the generation of antibodies with catalytic properties. It was anticipated that these antibodies would like enzymes selectively bind a reaction transition state structure in preference to the ground state of the substrate(s) and hence accelerate a reaction. Early attempts notably those of Raso and Stellar,* who attempted to produce antibodies capable of mimicking the action of the pyridoxal 5’-phosphate dependent enzymes tyrosine transaminase and tyrosine decarboxylase (Scheme I) and the 1983 report by Summers and Johnsong (Scheme 2) to create a catalyst for the enolization of diarylacetophenone used polyclonal sera.The isolation of a single catalytic species from such a mixture proved impossible and so it was not until the advent of monoclonal antibodies’O that the first catalysts were identified. In an initial study Kohen et d.ll were involved in the creation of monoclonal antibodies raised against active ester derivatives. These antibodies accelerated the hydrolysis of the ester haptens through the participation of reactive binding site amino acid side chains. However the antibody induced hydrolysis reaction was stoichiometric rather than catalytic with the reactive side chain of the antibody becoming permanently acylated.In December 1986 groups led by Lerner12 (Scheme 3) and Schultz13(Scheme 4) independently demonstrated that mono-clonal antibodies generated against tetrahedral phosphorus species could catalyse the hydrolysis of certain carboxylate esters and carbonates. NATURAL PRODUCT REPORTS 1996-N. R. THOMAS 48 1 HOmH:OZH + Me b 4 HO HAPTEN Scheme 1 HO HO' OH HO' Scheme 2 kcet = 0.027S-' KM= 1.9 pmol dm-3 kce4k-t = 960 KNl/& = 1 1.9 HOJyY0 Scheme 3 0 Scheme 4 In the sections below the key developments that have occurred in the field of catalytic antibodies in the ten years since these initial reports are described. It should be remembered that catalytic antibodies along with enzymes and all other catalysts obey several fundamental rules (1) They will not catalyse thermodynamically unfavourable reactions.(2) They will not reverse the direction of a reaction. (3) They will not change the equilibrium position of a reaction. (4) They remain unchanged at the end of the reaction. For an antibody to be a true catalyst it must obey all of these criteria. The examples below are presented in order of increasing complexity of catalytic mechanism. 2 Catalysis by Desolvation Catalytic antibodies have proved useful as vehicles for the testing of hypotheses concerning the methods employed by enzymes to accelerate reactions. One such area is the effect of a protein binding to a substrate and hence moving the substrate into a less polar environment.This obviously affects the amount of intra- and inter-molecular hydrogen bonding to the substrate and hence its reactivity. Kemp14 had demonstrated that 6-nitro-3-carboxybenzisoxazoleand related compounds underwent a decarboxylation reaction (Scheme 5) that was insensitive to general acid-base catalysis or conformational constraints. The large differences in the rate of decarboxylation observed in different media were thought to be due solely to the properties of the medium the benzisoxazole was placed in. Hilvert et d.15 have generated a number of antibodies that catalyse the decarboxylation of 6-nitro-3-carboxybenzisoxazole to 2-cyano-5-nitrophenol (haptens 1-3) and Kirby Hilhorst and coworkers16 have also explored a range of hapten structures (4-6)for this decarboxylation (Scheme 5).In Hilvert's case 25 out of 1200 hybridomas (2%) produced against a 1,5-naphthalenedisulfonate hapten 1 were found to be catalytic. Eight of these were further characterized with the best antibody (25ElO) providing a rate acceleration of 23200. This is somewhat slower than the decarboxylation of the benzisoxazole in neat hexamethylphosphoramide (rate enhancement of los over that in water) but is comparable to that found in other mixed media systems involving micelles macrocycles or polymers.l7 Interestingly and to the consternation of some of the catalytic antibody community the hapten structures 2 and 3 which were perceived to be the best mimics for the decar- boxylation reaction failed to produce ~ata1ysts.l~ It is now thought that the successful 1,5-naphthalene disulfonate hapten elicits a positively charged residue in the binding pocket.This attracts the carboxylate of the substrate and provides sufficient binding energy to overcome the loss of solvation energy that occurs when the carboxylate moves out of the aqueous environment into the non-polar binding pocket. NATURAL PRODUCT REPORTS 1996 S03H 1 H N-R NH 1 4 5 6 Scheme 5 It is presumed that the neutral haptens used in parallel studies which did not produce catalytic antibodies would have elicited hydrophobic pockets. These may be incapable of compensating for the desolvation of the carboxylate through binding interactions and hence have a lower affinity for the substrate making the antibody an ineffective catalyst.3 Antibodies as Entropic Traps In the case of many bimolecular or pericyclic reactions the entropy of activation (AS) can contribute significantly to the overall change in free energy of activation required for a reaction to occur. Enzymes are thought to catalyse entropically unfavourable reactions by binding and freezing out the rotational and translational degrees of freedom to form a ternary complex in which the substrates are held in reactive conformations and in the case of bimolecular reactions in sufficiently close proximity for a reaction to occur. This idea was first described by WestheimeP when he termed enzymes 'entropic traps'.Benkovic has estimated that an antibody forming a 'tight' strain-free complex with two substrates causes a loss of entropy of ca. 40 cal K-l mol-1 (1 cal = 4.184 J).19 Two early targets for antibody catalysis were the Claisen rearrangement as catalysed by the enzyme chorismate mutase and the Diels-Alder reaction ([4n +2n] cycloaddition) for which there is little evidence of an enzyme equivalent (a report of an exo-selective Diels-Alderase has recently appeared20). 3.1 Chorismate Mutase Antibodies Bartlett and Johnson21 synthesized a potent transition state analogue for the enzyme chorismate mutase (Ki= 75 nmol dm-3) and this was developed into a hapten by two groups (Scheme 6). Jackson et a1.22generated an antibody (1 lF1-2E11) which exhibits a rate enhancement of lo4.This is only two orders of magnitude lower than for the enzyme isolated from B. subtilis whilst Hilvert et al.23used a similar hapten to create an antibody (1F7) which displayed a lower rate enhancement of 250. Both antibodies displayed good stereo- selectivity in that they catalysed the Claisen rearrangement with the (-)-isomer of chorismic acid only. OH I H I OH I k4bcat 1R:kt= 1.2 x s-' KM= 51 pnol d~n-~ = 190 KdK = 85 11F1-2E11 &,,t=4.5x104s-' K~=260pmoldm-~ k&hncat = 10OOO Kd& = 29 Scheme 6 The crystal structure of antibody 1F7 with the oxobicyclic transition state analogue bound has been and transferred nuclear Overhauser effects between the chorismate substrate and the antibody in solution have been These indicate that the substrate is bound at the binding site in the reactive diaxial conformation necessary for the Claisen rearrangement to occur.A direct comparison between the active site of chorismate mutases from B. subtilis,26 and E. coli2' and the binding site of the catalytic antibody reveals that qualitatively the inhibitor-protein interactions are very similar (Figure 2). The main difference is that the number of interactions between the antibody binding pocket and its hapten are far fewer than between the two enzymes and the inhibitor. This suggests that the antibody has a more open binding pocket and is reflected in the tenfold difference in Ki values of the transi- tion state analogue for chorismate mutase from E.coli (75 nmol dm-3)21 and for antibody 1F7 (0.6 pmol dm-3).23 Several detailed comparisons of the active site of chorismate mutase antibodies and enzymes have recently appeared,28- 29 together with extensive mutagenesis studies of both enzymes.3o Whilst the antibody binding site has contacts with both the carboxylates at positions ClO and C11 the major difference between the antibody and enzyme structures appears to be a lack of a positively charged environment in the area around the ether oxygen of the hapten. This may be used to stabilize the negative charge developed on the ether oxygen of the substrate as it passes through its transition state with the cleavage of the C5-08 bond. As this feature was not designed into the transition state analogue there is no equivalent in the anti- body binding site.Redesign of the hapten or site specific mutagenesis of the antibody to incorporate positively charged groups at this position may result in the creation of more efficient catalysts. A mechanistic study comparing the mech- anism of action of antibody 1F7 with that of the enzymes has been p~blished.~~ Curiously the two antibodies appear to operate in different ways. Jackson et al.,32having ruled out the participation of NATURAL PRODUCT REPORTS 1996N. R. THOMAS (4 H2NYNH2 :i' ,/ HN-Arg 51 0-0Y I \Glu52 -TyrLW H2N&As#33 0 Figure 2 Schematic active sites of chorismate mutase from (a) E. cofi (b) B. subtifis and (c) antibody 1F7t general acid-base catalysis in the case of llF-2El1 then measured the differences in entropy and enthalpy of activation for the reaction in the presence and absence of 11F1-2E11.It was found that the enthalpy of activation was similar for both catalysed and uncatalysed reactions (20.5 kcal mol-I) whilst the change in the entropy of activation value was signifi-cantly lower in the antibody catalysed case (-1.2 eu = -1.2 cal K-l mol-') when compared to that of the uncata- lysed reaction (-12.9 cal K-l mol-'). This suggests that this antibody functions by freezing the substrate into a con-formation suitable for the Claisen reaction. Antibody 1F7 appears to operate by lowering the activation enthalpy to AH = 15 kcal mol-' while the activation entropy is more un-favourable than for the uncatalysed reaction (AS = -22 cal K-l m~l-l).~~ t Amino acid residues from catalytic antibodies are denoted as or ' meaning coming from the heavy chain (V,) and light chain (V,) variable regions of the F, fragment of immunoglobulin G (IgG).There have been attempts to improve the catalytic activity of antibody 1F7 by evolutionary election.^^ Details of this study are discussed in the section on catalytic screening (Section 17). 3.2 Oxy-Cope Antibodies Using a cyclohexane to mimic the ordered chair-like transition state of the reaction an antibody which catalyses an oxy-Cope ([3,3]-sigmatropic rearrangement) has been produced (Scheme 7).34This reaction has an estimated entropy of activation (AS:) of ca. 15 cal K-' m01-l.~~ J I6;lt= 0.026 min-' ,kt/Lat = 5300 Scheme 7 3.3 Diels-Alder Antibodies Several different examples of antibody catalysis of the Diels- Alder [4n+ 2771 cycloaddition reaction have been reported.At present there is no unequivocal proof that a Diels-Alderase enzyme exists although circumstantial evidence from the discovery of intermediate metabolites containing diene-dienophile functionality has been reported. 2o One problem anticipated with the creation of Diels-Alder antibody catalysts has been the fact that the close structural similarity of transition state and product structures would result in severe product inhibition in antibodies that functioned purely by transition state stabilization. To overcome this problem Hilvert developed an antibody that catalysed the Diels-Alder reaction between tetrachlorothiophene and N-eth~lmaleimide~~ resulting in an unstable adduct that spon- taneously extrudes sulfur dioxide and gives a planar product dissimilar to the transition state analogue used (Scheme 8).kcat = 0.07 s-' KM(dienophile)= 210 mmol dm-3 E.M. > 110 md dm-3 & = 126 nmol dms (E.M.= effective molarity) + Scheme 8 so2 NATURAL PRODUCT REPORTS 1996 309-1G7 NHPr NHPr I 0 0 kat3.37 x 10-'s-' KM (dienophile) = 3.11 mmol dms, = E.M. = 1100 mol dm-3 Scheme 9 Other Diels-Alder antibody catalysts have been reported by Braisted and Sch~ltz,~' Suckling et a1.,38whilst Meekel et aL3' have reported the first example of an antibody catalysed hetero Diels-Alder reaction (Scheme 9).A major breakthrough in antibody catalysis came with the report from Gouverneur et a1.40who generated antibodies capable of selectivelycatalysing the formation of the kinetically less favoured exo Diels-Alder adduct from 4-carboxyphenyl trans-buta-1,3-diene-1-carbarnate and N,N-dimethylacryl-amide using a bicyclo[2.2.2]octene system (Scheme 10). This antibody preferentially bound to the higher energy exo transition state resulting in the formation of the exo adduct. This reaction is discussed in more detail in Section 7 on rerouting reactions. 1 R GicoNMe,] -0-CONMe 0kl! NHC0,R * ex0 transition state -(= \NM% R HNYo 0 O bt= 3.2 x lo3 min-' KM (diene) = 700 pmol dm-3, 00C02H KM(dienophile) = 7.5 mmol dm-3 E.M = 18 md dms I r 1 "CONMe kRkFP -0 0 H CONMe NHCO,R endotransition state bt= 3.4 x 10-~min-' KM(diene) = 960 pmol dm4 KM(dienophile)= 1.7 mmol dm4 E.M.= 4.8 mol dm4 Scheme 10 Having shown that a rigid bicyclo[2.2.2]octene hapten was capable of eliciting Diels-Alder catalytic antibodies this group then took the novel step of using a conformationally flexible ferrocene derivative as a mimic for the same reaction.41It was anticipated that antibodies would be generated against specific conformations of the metallocene that bore a similarity in structure to the transition states of ortho-endo and ortho-exo Diels-Alder reactions (Scheme 11). The resulting antibodies have kinetic parameters equivalent to those for the bicyclo[2.2.2]octene suggesting that this is a valid approach for the creation of new catalysts.CONMe kt= 3.48 x 10" min-', 4- KMKM(diene) = 1.6 mmol dm-3, (dienophile) = 5.9 dm" & = 209 pmol dms E.M. = 4.9 mol dm4 x 13G5 I kt= 1.20 x 1o-~ min-' KM (diene) = 2.7 mmol dms ex0 KM(dienophile) = 10 mmol dm-3 & = 48 pmol dm-3 E.M. = 6.9 mot dm-3 Scheme 11 3.4 Retro[2n+24 Dimerization The retro [2n+2n] dimerization of thymine and uracil dimers has also been reported (Scheme 12).42The antibody 15F1-B1 catalysed the retrocycloaddition of the thymine dimer with a turnover of 1.2 min-l which is close to the value of DNA photolyase (k,, of 3.4min-l). This is an example of a light dependent reaction in which the quantum yield for the antibody (& > 0.75) is comparable with that of the enzyme (c$~= 1.0).Mechanistic studies on the antibody UD4C3.543which acts on the uracil dimer indicate that both bond making and bond HOZCCHZNH NHCHZCOZH HOZCCHpNH r 9 15F1-61 R = Me kt= 0.W 6' KM= 6.5 pnol dm-3 bv&t= 220 Kd& > 6.5 UD4C3.5 R = H bt= 7.8 x lo4 s-' KM= 280 pmol dm4 & = 54 nmol dm" bt'hcat = 380 Scheme 12 NATURAL PRODUCT REPORTS 1996-N. R. THOMAS breaking reactions are partially rate determining. Fluorescence quench experiments have implicated the involvement of a photoexcited tryptophan residue that transfers an electron to the dimer resulting in either a concerted cleavage of the cyclobutane ring once a radical anion has been formed or a stepwise cleavage in which the breaking of the first bond is reversible.4 Hydrolytic Antibodies Most of the initial studies towards the creation of antibody catalysts were directed towards producing hydrolytic antibodies capable of cleaving the carbon-oxygen or carbon-nitrogen bonds of carbonates esters or activated amides. This stemmed from the fact that previous work on the development of protease and esterase inhibitors had yielded a number of 'transition-state analogues ' for these reactions and a detailed knowledge of the mechanism of action employed by hydrolytic enzymes. Simple methods of screening for this type of catalytic activity were also readily available. The literature to the end of 1995 contains reports of over 80 hydrolytic antibodies.These have been generated against a variety of different 'transition- state analogue structures' (Figure 3) and in the cases where detailed mechanistic studies have been conducted it appears that a given hapten structure is capable of inducing catalytic antibodies that employ a variety of methods of stabilizing the reaction transition state. Before examining individual cases the general mechanism of ester-amide hydrolysis needs to be considered. It should be remembered that the hydrolysis reaction follows a reaction pathway that requires attack of water or a hydroxide ion at the carbonyl carbon to form a transient high energy tetrahedral intermediate followed by the collapse of this intermediate with the expulsion of the alcohol or amine leaving group to give the free acid.The transformation therefore passes through two chemical transition states; which of these is rate determining is dependent on the pK of the leaving group. In the case of p-nitrophenol and p-nitrophenylamine and other leaving groups capable of stabilizing a negative charge the first transition state involving the addition of solvent is rate determining. However in the case of simple aliphatic alcohols and amines the second phosphonate phosphonamidate (amidophosphonate) phosphate phosphorot hioat e (t hiophosphate) sulfonamide suifone HO OH H OH R1+P2F R'XO-R2 hemiketal Figure 3 Haptens used to generate hydrolytic antibodies 485 102.4" 113.4O ri f?@( .'ocy 124.7O H3 transition state tetrahedral intermediate (CT/PMB) (mp2/6-31 +G') 125.4' 7.0' transition state analogue (mp2/6-31 + G') Figure 4 Transition state tetrahedral intermediate and transition state analogue structures for the hydrolysis of methyl acetate as predicted by computer modelling (from ref 45) transition state (protonation of the leaving group) may well be rate determining. Most of the transition state analogues employed are based on tetrahedral sulfur or phosphorus derivatives (sulfonamides phosphonates phosphates and phosphonamidatest) that mimic the rate determining transition state. Both transition state45 and tetrahedral intermediate for nucleophilic attack on esters have been modelled using ab initio calculations.Using the .,3 method a reasonable geometry for the alkaline hydrolysis of methyl acetate was obtained which was then modified under the conditions imposed by the Cramer-Truhlar SM3solvation model. This indicated that as the phosphorus- oxygen or sulfur-oxygep bond distances of the transition state mimics are 0.2-0.3 A longer than those calculated for the tetrahedral intermediate but the bond between the nucleophile (e.g.oHO-) and the carbonyl carbon at the transition state is 0.6 A longer than the phosphorus-oxygen bond used to mimic it (Figure 4). It is interesting to note that phosphorus based transition state analogues have been considerably more successful than their sulfur counterparts in the elicitation of catalytic antibodies.This is probably due to the increased acidity and size of phosphonates and phosphonamidates. For almost all of the hydrolytic antibodies elicited against phosphorus based anti- gens the binding pocket contains a key basic residue that interacts with the negative charge on the hapten. This can participate in the catalytic cycle by acting as a proton donor or electrostatically in the stabilization of the oxyanion formed in the key transition state or in protonation of the leaving group. 4.1 Stabilization of the Transition State by Hydrogen Bonding to the Oxyanion Antibodies 6D9 4B5,8D11 and 9ClO raised against the same phosphonate hapten (Scheme 13) have been shown to obey the Wolfenden relationship discussed above (Section 1.l) which equates differential binding with rate enhan~ement.~~ This suggests that the entire binding energy of these four antibodies is used for rate enhancement through transition state stabil- t R'-PO(0-) (NHR2) as derivatives of phosphonic acid are properly named phosphonamidates or amidophosphonates.Some literature references may incorrectly refer to these as phosphoramidates. 486 u 0mNfF3 0 b 4 H 6D9 &OH 02N 0 HNKCHC’2 kat = 1.5 min-’ KM = 50 pmol dm-3 & = 0.06 pmol dm-3 &a&,ncat = 900 KdK(= 830 H HOr n N T C F 3 + YH 6D9 binding site predicted by computer modelling H1OObArg h H 0 -10 .dS0 ’ TyrL32 NATURAL PRODUCT REPORTS 1996 ization.Evidence from pH-rate studies and chemical modifi- cation suggests that the catalytic mechanism involves attack of hydroxide at the ester carbonyl. The side chain of in the antigen binding site may stabilize the transition state by forming a hydrogen bond with the oxyanion of the hydrated ester -a similar role has been postulated for a histidine residue in therm~lysin.~’ Two other antibodies generated against the same phos- phonate hapten are thought to operate slightly differently. In the case of antibody 3G6 a tyrosine residue substitutes for the histidine conserved in the above four antibodies and it is thought that the phenol of this residue acts as a proton donor. The other catalytic antibody produced 7C8 shows a higher activity than would be expected from its differential binding ability to substrate and hapten suggesting that the catalysis may be enhanced by general acid-general base and/or nucleophilic amino acid side chain involvement.However 7C8 displays a linear pH dependence of k,, (although only the range pH 6-8.5 has been investigated) that would appear to contradict this argument. The Fa fragment of antibody 6D9 has been crystallized and a preliminary report on its structure published.48 More recently the structure of the Fa of antibody CNJ206 bound to its p-nitrophenyl phosphonate antigen has been solved (Scheme 14).49This antibody is thought to operate almost solely by stabilization of the negatively charged transition states as all of the polar residues TyrHB0 TyrHloZand TyrLg6) in the binding site are further than 4.1 A from the phosphorus of the hapten precluding their involve- ment as nucleophiles.The oxyanion stabilization is thought to occur via hydrogen bonds to the backbone NHs of residues AspHloo and TyrHlo1 or with the N‘ of An additional contribution to the rate enhancement may arise from general base catalysis involving Patten et have taken the structure-activity relationship for catalytic activity one stage further by obtaining the crystal structure of th chimeric Fa of antibody 4867 with its hapten bound at 2.0 A resolution (Scheme 15) and the concurrent CNJ206 binding site TyrH’O1 0 <)( Asp”’” N 20; jsH35 1’PheL9* H Ab 6D9 485 8D11 9c10 - 0 I02N~O-~-CH,C02HOH I bt= 0.41s-’ KM = 0.90 mmol dm” Scbeme 13 Scheme 14 NATURAL PRODUCT REPORTS 1996-N.R. THOMAS 0 MeOH + COP kat = 0.023 s-l,KM = 660pmol dm-3 K = 3.3 pnol dms &Jhncat = 810 KdK = 200 Scheme 15 creation by mutagenesis of the germline antibody that has undergone affinity maturaiion to give rise to this antibody. The antibody has a 10 A deep binding pocket which has a surface calyx of tyrosine residues (TyrL32 TyrLgl TyrLg4 TyrH33 TyrHg9 TyrHIOO). The anionic phosphonate group is bound by a salt bridge between ArgLg6 and the Pro-S phosphonyl oxygen of the hapten a hydrogen bond between TyrH33 and the same phosphonyl oxygen and a hydrogen bond between the e-imino group of and the pro-R phosphonyl oxygen. From the crystal structure it is predicted that residues TyrH33and ArgLS6 stabilize the oxyanion formed in the reaction or alternatively could function as a general base to activate the substrate water molecule.There are no residues sufficiently close to the phosphonyl moiety to act as nucleophiles which is in keeping with the information gained from chemical modification and kinetic studies. A comparison of the binding and catalysis data of 4867 and its germline antibody show that the Kd has dropped by 3 x lo4 and the overall catalytic efficiency (kcat/Ksf)improved by almost a factor of lo2 in the antibody that has undergone somatic mutation. Perhaps more interestingly none of the nine residues that have been mutated to create 4867 are in direct contact with the antigen.They are involved in complex hydrogen bonding interactions which 'firm up' the binding pocket of the antibody and remove some of the flexibility of the hypervariable loops. Further insight into this study of antibody evolution will be gained if a germline or germline-antigen crystal structure can be produced. Similar interactions can be seen in another esterolytic antibody for which a crystal structure is available (Scheme 16) the anti-phosphocholine antibody M~PC603.~l In this case the negatively charged phosphate is stabilized by hydrogen bonds from TyrH33 and ArgHS2. 4.2 Utilization of Nucleophilic Catalysis Mechanisms involving nucleophilic catalysis are common in hydrolytic enzymes as exemplified by the cysteine and serine protease families.There is now significant evidence that several hydrolytic antibodies operate by a pathway involving nucleo- philic attack at the carbonyl carbon by a binding site residue. McPC603 binding site ArgHd &!$.,-a TyrH33/ bt= 0.045 rnin-' KM = 1.3 mmol dm4 btlkt = 83 Scheme 16 For one of these antibody 17E8 reported by Zhou et uI.~* the crystal structure of the antibody wit@ the transition state analogue bound has been solved to 2.5 A resolution (Scheme 17). This crystal structure illustrates a key feature of antibody 17E8 binding site TyrH'O' PheLQ8t-' HY!$0 17E8 0/o to + PhOH kat223 min-' KM = 457 pmol d~n-~, = K = 500 nmol dm" &Jk,,ncat = 8300,KdK = 910 Scheme 17 488 recognition in that only the R-enantiomer of the transition state analogue is bound to the antibody even though the hapten used to elicit the antibody was racemic.This factor is borne out in the substrate specificity of the antibody which is highly enantioselective for the hydrolysis of S-amino acids. The antibody has two well defined binding pockets for the phenyl ring and butyl side chain whilst the succinamide tether extends out of the binding site. LysineHg7 forms a salt bridge with the pro-S diastereotopic oxygen of the phosphonate suggesting that the protonated e-amino group is used to electrostatically stabilize the oxyanion in the esterolysis transition state. Mechanistic studies on the antibody have suggested that two ionizable groups of pKa 9.1 and 10.0 are involved in the catalytic mechanism.Moreover the pH-rate profile is bell shaped with a maximum at pH 9.5 implying that to be most effective the group with pKa of 9.1 is deprotonated and that with a pKa of 10.0 is protonated. From the structure of the binding pocket it has been proposed that the e-amino group of LysHg7 is the residue with a pKa of 10 which is involved in oxyanion stabilization (possibly along with ArgLg6) and can only accomplish this in its protonated form. The phenolic group of tyrosine residues typically have a pK ca. 9.0. Whilst there is no tyrosine within direct bonding distance of the hapten TyrHlol is close enough to hydrogen bond to SerHg9. It is proposed that SerHgs forms part of a catalytic dyad with This resembles the catalytic triad of serine proteases such as trypsin and subtilisin and it has therefore been proposed that the antibody 17E8 operates by a double displacement mechanism with SerHg9 acting as a nucleophile forming an acyl intermediate with the substrate.The obser- vation that the hydrolysis of norleucine phenyl ester in the presence of 17E8 and hydroxylamine gives rise to a mixture of amino acid and amino hydroxamic acid products lends credence to the involvement of an acyl antibody intermediate.53 In this scenario the phenolic group of TyrHlol is thought to act as a proton donor and hence a pH-sensitive switch that immobilizes the nucleophilic serine in the ‘off’ state as the pH of the medium decreases (Scheme 18). If antibody 17E8 closely follows a serine protease mechanism it is predicted that site specific mutagenesis to introduce a correctly positioned Asp or Glu residue to complete the catalytic triad could give an increase in rate (kcat)of between lo3 and lo4 (the difference in TyrH1’’ “97LyS\/ deactivated TyrH1O’ H97 activated Scheme 18 NATURAL PRODUCT REPORTS 1996 turnover number seen for trypsin wild type and a trypsin D 102N mutant with the unnatural substrate Z-GPR-SBn54).Alternatively the maximum activity of the antibody could be shifted closer to neutral pH which would increase the catalytic efficiency by up to two orders of magnitude. A similar phenomenon could be observed if the ‘deactivating’ tyrosine is mutated to a phenylalanine residue.Given that antibodies CNJ206 and 17E8 were elicited against haptens that are only similar in that they both contain 0-linked phosphonate and aryl groups their binding pockets share many common residues illustrating that only a small subset of the available antibody repertoire is selected after initial B-cell stimulation. The catalytic antibody NPN43C9 was generated against a phosphonamidate hapten (Scheme 19) and was found to hydrolyse both amide and ester bonds. The antibody has since been cloned and expressed as a single-chain antigen-binding fragment (scF,) and has been extensively studied by steady state and pre-steady state kinetics,55 chemical modification studies substrate structure-activity relationship^^^ and muta- gene~is.~’ A computer model of the antigen binding site (F,) has been produced based on information obtained from the above studies (Scheme 19).This uses the McPC603 V and D1.3 V crystal structures as templates for the respective chains.58 NPN43C9 binding site ?yrHg6 I Tyt .. X=O,NH I* c NPN43C9 NO2 H X = NH bt= 0.049 min-’ KM = 0.37 mmol dms = 1.5 x lo5,Kd& = 3.7x 10’ Scheme 19 These studies suggest that the imidazole of HisLg1 is participating as a nucleophile in the catalytic cycle with the formation of an acyl intermediate. Key evidence for this is the direct observation of an acyl-imidazole intermediate by electrospray ionization mass spectrometry studies and the absence of this covalent complex in a HisLgl+Gln mutant which was also catalytically inactive but bound to the transition state analogue with a similar affinity.59 Other features of the NATURAL PRODUCT REPORTS 1996-N.R. THOMAS NPN43C9 catalytic mechanism are the participation of ArgLg6 through electrostatic stabilization of the oxyanion in the two transition states (cf. the same residue in 17E8 and 4867). This arginine at the V,-J junction appears to be strongly selected in cases where it is necessary to bind to a negatively charged aromatic antigen such as the above haptens or the widely studied phenylarsonates.60 There is also strong evidence that antibody 20G9 catalyses the hydrolysis of phenyl acetate via an 0-acetyltyrosine intermediate,61 and the transesterification reaction catalysed by antibody PCP21H3 is also thought to occur via an 0-acyltyrosine intermediate.62 For this second antibody the reaction displays ping-pong kinetics and exhibits a pre-steady state burst of p-nitrophenol equal to the number of antigen binding sites.4.3 Utilization of General Acid-base Catalysis Suga et took an alternative approach to hapten de- velopment and attempted to generate ‘ aspartyl protease ’ like catalytic antibodies by a ‘bait and switch’ mechanism which uses highly polar groups in the hapten to elicit complementary charges in the antibody binding pocket. Three different hapten structures were synthesized a 1 ,2-amino alcohol ; a 1,2-trimethylammonium alcohol and a 1,2-trirnethylammonium phosphate (Scheme 20). It was intended that these would induce both a group that would act as a general base mediating H6-32 bt= 0.71 min-’ KM= 850 pmol d~n-~ 4 = 360 pmol dms ktlkt = 2.4 = 2400 KMI~ \N+’ I proton transfer from a water molecule to the alcohol or amine product and a neutral/basic group that would stabilize the oxyanion generated in the rate determining transition state.These are roles performed by two aspartyl side chains in aspartyl proteases such as renin. Interestingly all three haptens generated displayed esterolytic antibodies with k,,,/k,,,, values of ca. 3 x lo3 and similar specificity constants. The major difference displayed between the haptens was the number of catalytically active antibodies generated from their respective immunizations 16 out of 34 in the case of the 1,2-amino alcohol; 7 out of 48 in the case of the 1,2-trirnethylammonium alcohol; and 4 out of 20 in the case of the 1,2-trimethylarnmonium phosphate.A detailed mechanistic study of these antibodies has yet to be reported. 4.4 Phosphate Ester Hydrolysis There are several reports of phosphate ester hydrolysing monoclonal antibodies. The ultimate aim of this work is either to generate specific phosphorylation-dephosphorylationcata-lysts or to develop catalysts capable of degrading insecticides or biological warfare agents such as sarin soman and tabun. Given the unique hydrolytic stability of many phosphate mono- di- or tri-esters these reactions provide a testing target for catalytic antibodies. Examples of phosphate monoester and phosphate triester hydrolysing antibodies have appeared.Molecules that mimic the pentacoordinate transition state for base promoted phosphate ester hydrolysis such as vanadate esters and pentacoordinate phosphorus species are either toxic or chemically or conformationally unstable under typical immunization conditions ;64 hence other approaches to creating binding pockets have been employed. Antibodies for phosphate monoester hydrolysis have been generated using an a-hydroxy phosphonate hapten (Scheme 21),65 whilst a cyclic amine oxide has been used to generate antibodies capable of hydrolysing phosphate triesters (Scheme 22).66 One poorly characterized antibody has been reported which utilizes a pentacoordinate phosphorus species (Scheme 23).67 However due to the small quantity of antibody available a full set of kinetic data for the hydrolysis of pinacolyl methylphosphonofluoridic acid (soman) which would rule out the possibility of enzyme contamination perturbing the results has yet to be obtained.None of these antibodies offers large rate enhancements for their respective phosphate ester hydro- lysing reactions. 02N NlJOH I H H7-38 bt= 1.Omin-’ KM= 870 pmol dm4 4 = 5.0 prnd dm4 &&,,,cat = 3300 = 174 OH H7-59 kt= 0.49 min-’ KM= 1100 pmol dm-3 6 = 230 pmol dm-3 k&k+,ncat= 1600 KdK = 4.8 Scheme 20 0 38E 1 Q + H3P04 NO2 kat = 0.0012 min-’ KM = 155 pmol d~n-~ = 8000 K~lq= 4.6 Scheme 21 4.5 Antibodies for Glycoside Hydrolysis Several groups have developed model systems for the creation of antibodies with glycosidase activity.These could find a use in sequence specific oligosaccharide cleavage or in the synthesis of new a- or P-glycoside bonds. The first example was reported by Reymond et a1.68in 1991 (see Scheme 37e). In this case the hapten was based on a piperidinium ion as a mimic of the NATURAL PRODUCT REPORTS 1996 Q ? 0 T~1-4C6 t I 7 0 I X = 0-,Me I Substrate A kaI= 1.O x min-' KM= 3.57 pmol dm-3 & = 0.67 pmol dm3 katlknat= 361 Substrate B kat= 1.85 x min-' KM= 18.7 mmol dms kad&t= 21 Scheme 22 2 F-y-0 -F-7-OH + HO Me \ J bt= 4 min-' KM= 0.33 mmol dm-3 k&kmI=5500 Scheme 23 cationic transition state involved in the acid catalysed cleavage of an acetal.In a second example Suga et al.69generated a hapten which contained a cyclohexane ring which is locked in a half-chair conformation due to the presence of an azetidinium group (Scheme 24). This group also mimics the protonated form of the exo-oxygen and may induce the presence of a carbonium ion stabilizing group in the binding pocket. The antibodies produced against this hapten were around fivefold better catalysts than those reported by Reymond et a/.68 Schultz's group7o investigated the efficacy of three different transition state structures based on well known glycosidase inhibitors (Scheme 25). A hapten 7 containing a cyclic amine OH HO+H -H + H kat= 1.5 x min-' KM= 1.16 mmol dm-3 Kd = 10 pmol dm3 k-atlhncat= 375 Scheme 24 oA0-AA71.17 kcat= 0.904 h-' KM= 324 pmol dn~-~ 6= 35 pmol dm-3 katlknc-1 = 17700 KM/&= 9.3 N=C=S Scheme 25 group and hence similar to the natural product nojirimycin produced catalytic antibodies.However haptens based on a cyclic amidine 8 or a D-galactal derivative 9,were unsuccessful. In the case of 7,an even more efficient catalyst with a rate enhancement of 17600 was obtained. NATURAL PRODUCT REPORTS 1996-N. R. THOMAS Several other groups71have reported the synthesis of haptens for the generation of antibodies capable of catalysing the hydrolysis or formation of glycoside bonds but details of the catalytic activity of antibodies produced against these haptens have yet to appear.4.6 Ester and Amide Bond Formation in Water There are several examples of antibodies capable of catalysing the formation of new ester or amide bonds under aqueous conditions. These include antibody 24Bll that is capable of both lactonization and amide formati~n,~~ antibody 18R.136.1 which catalyses the transesterification reaction of a benzyl ester,lg8and antibody 17G81g9which catalyses the formation of a benzyl amide. Selective peptide bond formation has been found to be catalysed by antibodies 16G3200and 9B5.lZo1which use substrates containing p-nitrobenzoate and azide activated acyl groups respectively. The haptens used to elicit these antibodies contain phosphonamidate or neutral phosphonate diester groups as transition state mimics.5 Non-hydrolytic Antibodies using General Acid-Base Chemistry There are several reports in the literature of antibodies that catalyse elimination reactions and employ haptens that have been functionalized to induce basic functionality at the antigen binding The most recent case is of an antibody (1D4) that is capable of catalysing a 'disfavoured ' syn elimination through an eclipsed transition state to give a cis (2)olefin product (Scheme 26).72It has been estimated that the energy difference between the anti elimination via a staggered transition state and the syn elimination via an eclipsed transition state is ca. 5 kcal mol-1 and hence the reaction is accessible to a pathway change through antibody binding. In the case of the uncatalysed reaction the only product detected is the trans isomer.The aminobicyclo[2.2.llheptane hapten is derivatized with an amino group in the position occupied by the proton of the substrate that has to be removed. A general base can then be induced close to this position in the binding pocket. Me .. .. .. .. kat = 2.95x lo4 min-' KM= 212 pmol dms Scheme 26 Shokat et ~1.~~ have carried out a detailed mechanistic study on antibody 43D4-3D12 that catalyses the elimination of HF from (4R) and (4S)-4-fluoro-4-(4'-nitrophenyl)butan-2-one (Scheme 27). The pH dependence of k,, gives the classicalprofile for catalysis involving one titratable group with a pK ca. 6.2. By using a tritiated epoxide analogue of the substrate it was possible to label G ~ onu the~ antibody implying that ~~ it was the carboxylate side chain of this amino acid that was responsible for removal of the proton in the elimination reaction.As the substrate contains a carbonyl group a to the proton removed two possible mechanisms were considered 49 1 FO 0 bt= 0.193 min-' KM= 182pmol dm" 4 = 290 nmol dm-3 &aik,,ncat = 1.45x lo3,KMlK = 630 Scheme 27 either an Elcb or E2 mechanism could operate depending on which of the C3 prochiral hydrogens is abstracted by G~u~~~. The four C3 monodeuterated diastereomeric analogues of the substrate were synthesized and incubated with the antibody. The antibody displayed a stereochemical preference for anti rather than syn elimination but exhibited little stereofacial selectivity for either the pro-R or pro-S C3 hydrogens suggesting that both Elcb and E2 mechanisms may be operating.Thorn et ~1.'~ have reported the creation of an antibody capable of catalysing an E2 elimination of 5-nitrobenzisoazole to give 2-cyano-4-nitrophenol by the use of a positively charged benzimidazole hapten (Scheme 28). The pH dependence of the reaction and chemical modification studies suggest that a carboxylate group with an elevated pK (60) is responsible for the deprotonation which is the rate determining step of the reaction (as determined from the kinetic isotope effects observed when 3-2H-5-nitrobenzisoxazole is used as the substrate). The antibody in this case appears to use a combination of desolvation (substrate binding to a generally hydrophobic binding pocket) and general base catalysis resulting in a rate increase of > los above background for the proton transfer.H ME4 I H I kat = 0.66s-' KM= 120 pmd dm4 & = 1 nmol dm-3 = 41 OOO Scheme 28 6 Cofactor Assisted Catalysts To extend the repertoire of antibody catalysts several groups have followed nature's lead and attempted to develop catalysts that utilize cofactors to provide additional chemical function-ality at the antibody binding site. These studies fall into two main groups those involving cofactors found in nature (pyridoxal phosphate porphyrins metals) and those that involve simple inorganic reagents. The latter are important as they demonstrate that it is possible to generate catalysts for redox chemistry which avoid the need for expensive biological redox cofactors such as NAD(P)H.6.1 Natural Cofactors In many cases a cofactor or coenzyme is able to conduct the reaction independently but the antibody provides a template that controls the substrate selectivity and regio-and stereo- specificity of the reaction. The literature features several enzyme mimics including a pyridinium based decarboxylase (mim- icking pyridoxal 5’-phosphate dependent enzymes)75 and several metal dependent hydrolytic antib~dies.~~ Attempts to make antibodies with binding sites for protein coordinated metals by mutagenesis of existing catalytic antibodies in which the metal can function as an electrophilic catalyst have also been rep~rted.~‘ These however have yet to exhibit catalytic properties.6.2 Redox Reactions Several redox active antibody systems that involve flavin or porphyrin cofactors have been rep~rted.’~ In these cases the antibody provides a less polar environment and some substrate specificity the most recent example being the creation of peroxidase antibodies dependent on metal porphyrins (meso- haemin or TCPP).79 A Japanese group has shown that the light chain of antibody 13-1 generated against 5,1O7l5,20-tetrakis(4- carboxypheny1)porphyrin (TCPP) when re-expressed in E. coli is capable of oxidizing both pyrogallol and ABTS (2,2-azinobis- 3-ethylbenzthiazolin-6-sulfonicacid) in the presence of hy- drogen peroxide (Scheme 29).Unusually the light chain of 13- 1 is reported to have much higher catalytic activity and thermal stability than the parent antibody. Given the considerable number of porphyrin model systems (capped picket fence etc.) that have been investigated the creation of novel antibody- porphyrin catalysts is one that offers considerable opportunities for future exploitation. pyrogallol Fe%-TCPP Antibody 13-1 FH 6 C02H Scheme 29 6.3 Semi-synthetic Catalytic Antibodies Pollack and Schultzso reported the creation of a ‘semi-synthetic’ catalytic antibody in which cofactors are covalently linked to the antibody molecule in close proximity to the antigen binding site (Scheme 30). A cleavable affinity label was used to introduce a free thiol group at the periphery of the binding site.There was a possibility that this could then act as a nucleophile or be derivatized so that another reactive group (an imidazole or pyridine derivative) could be introduced into the antigen binding site. In this case the catalytic activity of an anti-phosphocholine IgA MOPC3 15 was enhanced giving a rate enhancement of 6 x lo4 for ester hydrolysis. Another recent study along similar lines is work reported by Luo et dS1 into the generation of a glutathione peroxidase antibody. This required the introduction of a selenium into the binding site of an antibody that had been generated against a dinitrophenyl glutathione hapten (Scheme 31). One of the NATURAL PRODUCT REPORTS 1996 02N)$ I kat= 0.0145 s-’ KM = 1.2 pmol drne3 02N* + 0 NO* Scheme 30 anti-dinitrophenyl glutathione antibody seleno-4A4 H Se Scheme 31 antibodies specific for this hapten 4A4 was found to contain a binding site serine residue which was chemically mutated to selenocysteine using first toluene-p-sulfonyl fluoride and then hydrogen selenide.In a comparison of seleno-4A4 and a similarly treated rabbit IgM raised against human IgG it was found that the seleno-4A4 had 15 times the activity of the mutated rabbit antibody (1097 U mo1-I cf. 71.4 U mol-l) probably due to the higher binding affinity of 4A4 for glutathione. 6.4 Use of Cheap Inorganic Cofactors One of the main drawbacks with preparative biotrans-formations has been the requirement for expensive redox cofactors such as NAD(P)H.82 Even though efficient methods of recycling these cofactors have been developed it is still economically beneficial to utilize alternative inorganic redox reagents provided their reactivity can be controlled by the antibody (template).Several catalytic antibodies have been developed that make use of such cofactors. Antibody 37B.39.3 activates a specific ketone such that it can be reduced by sodium cyano- NATURAL PRODUCT REPORTS 1996-N. R. THOMAS Scheme 32 + NaI04 2804.2 01 OH &t = 8.2S-’ KM (sulfide) = 43 pmol dm-3 KM (periodate) = 250 pmol dm-3 Scheme 33 borohydride. This antibody catalyses the stereospecific re- duction of one of a pair of chemically similar ketones generating an alcohol in 96 % ee demonstrating the good regioselectivity that can be obtained with antibody catalysts (Scheme 32).83 Importantly the antibody was able to catalyse more than 25 turnovers without undergoing deactivation demonstrating that sodium cyanoborohydride does not significantly modify this anti body.r 1* Uncatalysed ‘0 5exo-tef INoMe As a counterpoint to this reaction Schultz et ~l.*~ have demonstrated that antibody 28B4.2 raised against an am-monium phosphonate is capable of mediating the oxidation of a sulfide to a sulfoxide using sodium periodate (Scheme 33). The catalytic efficiency in this case [(k,,,/K,)/k,,,,,] was 2.2 x lo5 and the antibody possesses similar turnover numbers to that of P-450 enzymes.Its activity was unaffected by extended exposure to the oxidant. 7 Rerouting Chemical Reactions using Anti bodies One of the key areas that catalytic antibodies have shown a distinct advantage over other forms of rate enhancers is in rerouting a reaction to give a kinetically less favoured product or to direct a reaction down a particular pathway instead of resulting in the formation of a manifold of products. Two types of reaction have exploited antibody rerouting one involving the selective preparation of an exo Diels-Alder adduct and the other a cyclization+poxide opening reaction to give a less favoured product as predicted by Baldwin’s rules for cyclization reactions.*’ 7.1 Formation of the Thermodynamically More Stable exo Diels-Alder Adduct Gouverneur et aL40 have reported the creation of antibodies capable of selectively catalysing the formation of the less favoured exo Diels-Alder adduct from 4-carboxyphenyl truns-buta- 1,3-diene- 1-carbarnate and N,N-dimethylacrylamide using a bicyclo[2.2.2.]octene hapten (Scheme 10).In the absence of a catalyst the diene and dienophile react to form an 85 15 mixture of a ortho'-endo adduct e ortho'-exo adduct which is in agreement with ub initio calculations. At the RHF/6-3 1G* level the relative differences in energies are ortho-endo 40.80 kcal mol-’ ortho-exo 42.70 kcal mol-’ (difference = 1.9 kcal mol-’) metu-endo 42.88 kcal mol-1 (2.1) metu-exo 43.94 kcal mol-’ (3.1) The difference in activation energy for the favoured endo and disfavoured exo processes is only 1.9 kcal mol-’ ;thus the use of the available antibody binding energy (estimated at 20 kcal mol-1 85) to reroute the reaction by stabilizing the exo transition state results in the formation of solely the exo adduct if sufficient antibody is used.7.2 Controlling Cyclization Reactions Janda et generated antibodies against a piperidinium N-oxide (Scheme 34). This favoured the ring closure of an epoxy alcohol to form a tetrahydropyran rather than the preferred tetrahydrofuran product as predicted by Baldwin’s rules,” “2 0 r l* L -1 OH Me06” ‘ bt= 0.91 min-’ KM= 356 pmol dm” Scheme 34 t trans kcat= 0.86rnin-' KM= 190 iirnol d~n-~ & = 16 prnol dm-3 cis bt= 0.64 rnin-' KM= 170 prnol d~n-~ 4 = 15 pmol drn-3 kcat = 0.81 min-' KM= 180pmol drn-3 4 = 17 prnol drn-3 1 Scheme 35 NATURAL PRODUCT REPORTS 1996 which is formed in the absence of the catalyst.Molecular modelling in this case indicates a difference of ca. 1.8 kcal mo1-l for the activation energies to the two transition states in aqueous solution and a difference in the entropic energy for the two transition states of 0.3 cal K-l mo1-1.88 Another antibody rerouted cyclization is that reported by Kitazume and Takeda.89 In this case antibodies could be generated that would catalyse either the exo or endo lacto-nization of a-trifluoromethyl y,J-unsaturated acids to form the corresponding y-or b-lactones (Scheme 35). 8 Multistep Reactions One key area that requires further exploration is the catalysis of reactions with two or more rate limiting transition states with widely differing structures.The best studied example in this area is the antibody catalysed conversion of an aspara-ginylglycine N-phenethylamide to an intermediate succinamide and the subsequent hydrolysis of the latter to both aspartate and isoaspartate products (Scheme 36).'O Other examples of antibodies which catalyse consecutive reactions include anti- body 24B11 which catalyses a lactonization followed by an amidation reaction,g1 and antibody 78H6 which catalyses an aldol addition followed by an elimination reaction (Scheme 374.92 9 Substrate Attenuation and Mutagenesis There are two aspects to substrate attenuation or mutagenesis.The first is the minor modification of a substrate to overcome product inhibition and to focus the binding energy of the antibody on the interactions intimately involved in transition state stabilization. This results in improved catalytic charac- teristics as exemplified in the work of Jandag3 and Ta~fik.~~ + 0 asparlate isoaspartate Scheme 36 NATURAL PRODUCT REPORTS 1996N. R. THOMAS ; (c) \Ar -s,-‘3 MeCN + H,O H Ar MeCONH OMe Ar $. dienone-phenol [1,2]-rearrangement kat= 1.22x lo4 s-’ KM= 670pmol dm4 \ 0‘A r >98% ee epoxidation aldol condansation/~-~llminatlon i (haptenof opposite stereochemistry) \ kaI= 5 x s-’ KM = 85 pmol dm-3 14D9 I. enol ether hydrolysis/ enantioselective protonation R=H,Me acctal hydrolysis 1 pmol dm-3 kat= 7.8 x lo4 s-’,KM= 100 pmol dms AcH N inCH0 7204 ___t + AcHN Me >95% de k, J + AcHN enamine mediated aldol condensation@-elimination Scheme 37 The second aspect is the manipulation of substrates such that the antibody binding pocket can be used to catalyse a wide variety of different reaction types which possess a common rate determining transition state structure.The best example of this is the research that has been performed with antibodies generated against the enantiomers of an N-alkyl-N-methyl-3- (glutarylamidomethy1)piperidinium hapten. It has been demon- strated that one of these antibodies 14D9 is capable of acid promoted cleavage of cyclic acetals to give disfavoured products (Scheme 37e),94 as discussed above in Section 7; enol ether hydrolysis and enantioselective protonation (Scheme 374 ;95 and enantioselective hydrolysis of epoxide~.~~ Three other antibodies against the same hapten exhibit other catalytic properties antibody 20B 11 catalyses an enantioselective epoxi- dation (Scheme 37b);” antibody 62C7 an acid promoted dienone-phenol [ 1,2]-rearrangement (Scheme 37a) ;98 and anti- body 78H6 catalyses consecutive aldol condensation and p-elimination reactions (Scheme 37~).~~ More recently related hapten structures have been used to create antibodies capable of catalysing cationic cyclization NATURAL PRODUCT REPORTS 1996 4C6 0-0 I o=s=o cationic cyclisation Q bt= 0.02min-' KM = 230 mmol dm-3 NHCOMe I PH TM1-87D7 * -Si-+ o=s=o j 0 1 o=s=o QWH Me % 0 1 cyclopropanation 63% NHCOMe i kt= 0.021min-' KM= 102 pmol drn-3 j I NHCOMe i Scheme 38 Q -?i-* (98%) (2%) OH I o=s=o Q NHCOMe ?H -Si-Si-c -c T 1605 Nal 0 I SN1reaction I NHCOMe o=s=o kt= 0.028 min-' ,KM(NaI) = 150 mmol dm-3 KM (sulfonate) = 130 prnd dm4 E.M.= 580 rnol dm-3 NHCOMe 4 = 10prnol dm-3 (Scheme 38a),loo nucleophilic substitution (S 1) (Scheme 38c) and cationic cyclopropanation reactions (Scheme 38b).lo1 In these cases the antibody binding site favours the formation of a cationic intermediate which can then be trapped in a variety of ways depending on the substrate structure.10 Antibodies in Organic Solvents There have been several reports of modified and unmodified catalytic antibodies being used in organic solvents either in aqueous-organic biphasic systems;lo2 reversed micellar as immobilized or with lipid coated catalytic antibodies (Scheme 39).lo5 As many of the antigens and hence substrates processed by catalytic antibodies are hydrophobic it is normally beneficial to use non-aqueous or mixed reaction conditions. As has been found with other H20-acetone / lipid coated antibody -Scheme 39 proteins antibodies change their affinity and structural integrity under such conditions; however they still appear to function as catalysts. The exclusion of water offers the opportunity of developing ester amide and glycosyl transferases from hy- drolytic antibodies that form acyl intermediates.11 Polyclonal Catalytic Antibodies Several groups attempted to produce catalytic polyclonal antibodies in the 1960s and 70s prior to the development of the hybridoma methodology by Kohler and Milstein.lo These include (a) the research of Slobinlo6 who found that polyclonal antibodies generated against a p-nitrocarbobenzoxy conjugate protected p-nitrophenylacetate and p-nitrophenyl-s-amino caproate from hydrolysis when [S] < [Ab] (as would be expected from a hapten which was a ground state analogue) and (b) Raso and Stollar's attempt8 to produce polyclonal antibodies with tyrosine decarboxylase activity as discussed earlier (Section 1). Initial studies in the use of polyclonal antibodies with haptens that had been successful in generating monoclonal antibodies met with little success.1o7 This may have been due to any catalytic antibodies being a very minor component of the antibody mixture.More recently Gallacher et al.lo8have shown that it is possible to isolate polyclonal sheep antibodies raised against a phosphate hapten which display significant hydrolytic activity with both carbonate and amide substrates. Other groups have also reported polyclonal antibodies with catalytic activity.log The advantages of using polyclonal antibodies are the ease and low cost of their preparation relative to monoclonal antibody production and the fact that the animal's entire NATURAL PRODUCT REPORTS 1996-N.R. THOMAS immune repertoire is effectively being screened in one ex- periment.ll0 However a recent experiment by Iversen et al.lll suggests that against haptens the immune response generates a poly- clonal mixture with a fairly homogeneous response in terms of photophysical behaviour and by implication binding affinity and kinetics. One of the added complications with polyclonal antibodies is that it has been impossible to separate the antibody fraction that is catalytically active from the total polyclonal sera. It has been estimated by Wilmore and Iversen112 that the catalytically active fraction could be between 1 and 10% of the antibody population specific to the antigen. Until an effective protocol ,based on the hapten structure or a catalytic screen based around mechanism based substrates has been developed (see Section 14.3),there will be some uncertainty in the kinetic parameters given for polyclonal antibody preparations.Other disadvantages of handling a heterogeneous mixture include the fact that catalytic activity may be masked by the large number of antibodies that purely bind the substrate and hence perturb the kinetic measurements obtained. There is also the possibility of losing catalytic activity as the antibodies are subjected to further hypermutation in vivo or through death of the donor. 12 Autoimmune Antibodies 12.1 Peptide Hydrolysing Autoantibodies There are now several reports of autoantibodies (antibodies generated against molecules naturally found in the organism) with catalytic activity.Paul et aL1l3identified a human anti-VIP (vasoactive intestinal peptide) antibody that catalysed the cleavage of a VIP peptide between residues 16 and 17 glutamine and methionine. This polyclonal antibody was found in around 17YOof healthy subjects and at elevated levels in the asthma patients screened. As VIP peptides are involved in broncho- dilation it is possible that reduced levels of VIP are responsible for the airway hyper-responsiveness experienced by asthmatics. Recent studies have shown that polyclonal IgG antibody mixtures from three different patients contained antibodies that were capable of cleaving the VIP at seven different sites. An extensive study of the occurrence and catalytic mechanism of the VIP hydrolysing antibody has been undertaken.The antigen responsible for eliciting the catalytic antibody is currently unknown. However it appears that it could be VIP itself or a closely related peptide. Paul et al.,14 have demon- strated that a murine monoclonal antibody generated against a VIP-keyhole limpet haemacyanin conjugate was capable of catalysing the hydrolysis of VIP at low protein concentrations and that the recombinant light chain subunit displayed catalytic activity on its own towards both VIP and a tripeptide substrate (Pro-Phe-Arg-methyl coumarinamide). Molecular modelling of the light chain sequence suggests the possibility of the involvement of a catalytic triad like that seen in serine protease (SerZ8 Hiss8 Asp') in the catalytic The same group has also reported that Bence Jones proteins (monoclonal antibody light chains found in the urine of ca.60 YO of multiple myeloma patients) display hydrolytic activity with several tripeptides and tetrapeptides containing arginine residues. 116 There are also autoantibodies to thyro- globulin (a precursor to thyroid hormones) which are found in patients suffering from Hashimoto's thyroiditis contain a component that is capable of catalysing the hydrolysis of thyroglobulin into several smaller fragment^.^,' 12.2 DNA Hydrolysing Autoantibodies Gabibov and coworkers118 have isolated IgG and IgM fractions of autoantibodies from the sera of patients with systemic lupus erythematosus ;these autoantibodies display DNA hydrolysing activity.A second part of DNA-cleaving anti-idiotypic anti- bodies has recently appeared.llg In this case polyclonal anti- idiotypic antibodies were generated against an antibody that had been raised in rabbits and shown to inhibit competitively the catalytic action of DNase suggesting that it possessed an internal image of the enzyme's active site. As with the peptide hydrolysing antibodies discussed above these polyclonal anti- bodies appear to have a broad substrate specificity in this case for both supercoiled and linear double-stranded DNA. 13 Anti-idiotypic Catalytic Antibodies The name given to the unique amino acid sequence found on the V (variable) domains of an antibody molecule is called its idiotype. It has been shown that this can act as an antigenic determinant for a second class antibody.In this case the 'epitopes' on the first antibody Ab that are bound by a second antibody Ab are called idiotopes and hence Ab is called an anti-idiotypic antibody. There are three classes of anti-idiotypic antibodies. Ab,a are antibodies that recognize idiotopes that are outside the antigen- binding site. Ab,P recognizes the binding site of the Ab antibody and resembles the antigen epitope bound by Ab, hence it is said to be an 'internal image' of the antigen (Scheme 40). The third class of anti-idiotypic antibody Ab,y recognizes the antigen binding site of Ab but does not function as an internal image. human erythrocyte acetylcholine esterase antiidiotypic ) u kat= 81 s-' KM = 0.6 mmol dms and bdkt= 4.15 x lo8 Scheme 40 Anti-idiotypic antibodies are thought to be involved in regulation of the immune system either by neutralizing their target antibodies (competitive inhibition of the Ab binding sites) or by binding to the surface immunoglobulin receptors on a B cell to modulate antibody production and act as surrogate antigens.It has been shown that anti-idiotypic antibodies can be produced against anti-enzyme active site antibodies and several groups have reported that these antibodies contain an image of the catalytic site of the original enzyme and hence exhibit similar catalytic activity. Izadyar et al. have produced both polyclonal antibodieslZ0 and a monoclonal IgM anti-idiotypic antibody 9A8 (Scheme 40)12 against antibody AE-2 a monoclonal antibody directed against the active site of acetylcholine esterase.It has been shown that both the polyclonal and monoclonal antibodies were capable of hydro- lysing acetylthiocholine and related esters in a very efficient manner when compared with antibodies generated against transition state analogues and other low molecular mass antigens. Selective chemical modification of 9A8 with diiso- propyl fluorophosphate and echothiopate suggests that the antibody utilizes a catalytic mechanism involving a nucleophilic serine as is the case with its progenitor acetylcholinesterase. Interestingly it has been speculated that catalytic human autoantibodies with DNA nicking activity found in the sera of 49 8 patients with systemic lupus erythematosus are anti-idiotypic antibodies to topoisomerase 1 as discussed in Section 12.2.14 Creation of Catalytic Antibodies -Novel Immun izat ion Proced ures Whilst most of the catalytic antibodies generated have used a standard immunization procedure first described by Kohler and Milstein,lo there have been several studies where modifi- cations in this procedure have led to the generation of antibodies with improved catalytic activity. 14.1 Use of Autoimmune Mice Tawfik et a1.122 have tested the ability of several strains of mice to elicit esterolytic antibodies. It was found that MRL/lpr and SJL autoimmune mice produce much greater numbers of catalytic antibodies than either the parent strain MRL/ + + or BALB/c mice.In the early stages of the immune response (21 days after immunization) 576 out of the 640 clones (87%) produced were found to be catalytic in the case of the MRLllpr mouse. Interestingly the number of catalytic antibodies declined to the levels seen in normal mice after multiple immunizations suggesting that there may be a deselection for catalytic antibodies that is less effective in the autoimmune mice. 14.2 In vitvo immunization StAhl et al.123have reported the use of an in vitro immunization protocol for the creation of murine catalytic antibodies. This approach offers both experimental simplicity and a means of generating antibodies against haptens that have failed to give a significant immune response in vivo.Whilst there is evidence that the amount of somatic mutation that occurs in an in vitro immunization is limited the carbonate hydrolysing monoclonal antibodies produced in the reported study have similar kinetic characteristics to those generated using similar haptens and in vivo immunization. The in vitro immunization approach has homologous-H7-38:ht= 0.79 rnin-’ KM= 1430 prnol drn-3 4 NATURAL PRODUCT REPORTS 1996 been used by several groups to produce human monoclonal antibodies for use in therapy and one example of a human monoclonal catalytic antibody has been reported.124 14.3 Heterologous Immunization Suga et have pioneered the application of heterologous immunization for the creation of catalytic antibodies. This involves immunizing an animal with two different but structurally related antigens in succession (Scheme 41).It is anticipated that the host’s immune system then responds by mutation of some of the antibodies generated against the first antigen so that they have an affinity for both haptens. Suga et al.125 have used this approach to attempt to introduce both acidic and basic groups in the antibody binding site for participation in catalysis. This would have involved a com- plicated synthesis for inclusion of both haptenic elements in a single molecule. Instead two simple readily available haptens were used one with a positively charged quaternary ammonium alcohol 10 and the second with a negatively charged phosphonamidate 11. Given the immunization protocol used it is probable that the booster injection with the phosphonamidate resulted in an increased expression of some of the antibodies stimulated by the first immunization rather than significant further somatic mutation.However the antibodies produced by this approach had superior catalytic properties to those generated solely against the positively charged hapten the best H5H2-42 exhibiting a rate enhancement of 1.5 x lo5; this was two orders of magnitude better than that observed for antibody H5-67 generated in a homologous immunization. 14.4 Reactive Immunization Winching et al.lZshave taken the heterologous immunization idea one stage further in a process they term ‘reactive immunization’. In this case a hapten containing a bis[p-(methylsulfonyl)phenyl] phosphonate which has a half-life of several days under physiological conditions was used as an = 5.0pmol dm-3 (lo) = 2700,KM’~= 286 heterologous-H7-38:kt= 12.5min-’ KM= 240 pnol dm4 4 = 300 (10)and 22 prnol drn-3 (1l) katlkncat = 68000 homologous immunise (100pg 10) 2 weeks -boost with 10 3 weeks -boost with 10 3 days -harvest spleen 3 days heterologous irnmunise (100 pg 10) 2 weeks boostwith 10 ‘Iweeks imrnunise (100 pg 11) harvest spleen t + Scheme 41 NATURAL PRODUCT REPORTS 1996-N.R. THOMAS HO S02Me kt= 31 min-' KM= 300 pmol dms & (12) = 0.11 pnol dm-3 Scheme 42 immunogen (Scheme 42). Because of its inherent reactivity it was expected that this hapten would stimulate B cells producing antibodies that have reactive tyrosine histidine serine or cysteine residues in their binding pockets and hence form phosphonyl-antibody intermediates.Whilst this was occurring most of the diary1 ester in solution would be hydrolysed to the negatively charged monoaryl ester 12. This would stimulate further mutation of the antibody binding site to give better transition state recognition and induce residues capable of stabilizing the oxyanion generated in the course of substrate hydrolysis. Of the 19 monoclonal antibodies analysed almost all displayed a higher affinity for the monoaryl ester and several exhibited burst kinetics indicative of the formation of acyl- antibody intermediates. In terms of catalytic ability relative to antibodies generated by using the standard approach of immunizing purely with a transition state analogue the most efficient antibody SP049H4 had a specificity constant (kcat/&) of 1 x lo5 dm3 mo1-' min-l that puts it among the best two or three hydrolytic antibodies so far discovered.This work suggests that compounds which can function as hybrid mechanism based and transition state mimic inhibitors may elicit more potent catalytic antibodies than those presently available. One other example of a reactive immunization has been rep~rted.'~'In this case a diketone was used as the immunogen and this appears to have selected several antibodies with lysine residues at their binding site (Scheme 43). These are capable of forming Schiff bases with one of the carbonyl groups of both the hapten and substrate and hence mimic the enamine mechanism used by Class 1 aldolases.The substrate and regioselectivity of two of the antibodies generated by this immunization have been examined and found to have a wider substrate specificity than existing enzymes and also to exhibit a reasonable degree of stereocontrol over the aldol reaction even though the hapten is achiral. 15 Immortalization and Expression of Antibodies and Antibody Fragments Whilst polyclonal sera are simple and cheap to obtain the use of polyclonal antibodies as catalysts suffers from several major disadvantages as described in Section 11. Kohler and Milstein's major breakthrough,1° the development of the hybridoma methodology for the production of monoclonal antibodies of predefined specificity led to the identification of monoclonal antibodies that underwent both single turnover and more Scheme 43 NATURAL PRODUCT REPORTS 1996 recently multiple turnover events i.e.catalysis. For most catalytic antibodies a combination of in vivo immunization and cell line immortalization through hybridoma generation has remained the method of choice. However the creation of catalytic antibodies by in vitro immunization (Section 14.2) and the cloning of the murine and human immune repertoires (Section 15.1) have recently been rep~rted.~~~-~~~ These provide alternative routes to the gen- eration of antibodies or functional antigen binding fragments and potentially offer significant advantages over the existing in vivo immunization-hybridoma route in terms of rapidity of antibody isolation and the possibility of identifying antibodies that bind molecules that are toxic in vivo even when conjugated.15.1 Immunoglobulin Gene Expression Libraries and Phage Display The majority of the research into the development of immunoglobulin expression libraries has been conducted by the groups of Winter in the UK and Burton Barbas and Lerner in the USA. Several comprehensive reviews of this work have recently appeared.128 By using a combination of reverse transcriptase and the polymerase chain reaction (PCR) it is possible to obtain F (heavy chain V,) and light chain (V,) gene libraries from a mouse spleen as complementary DNA (cDNA).Using this approach a naive (unimmunized) mouse combinatorial DNA library containing 2.5 x lo7 clones can be produced in several days. If a specific antigenic reactivity is desired it appears to be beneficial for an immunized mouse spleen to be used. In this case the B cells will have undergone somatic hypermutation to give higher affinity antibodies whilst B cells producing weak binders will have been deselected reducing their appearance in the DNA library.129 The affinity of single chain antibody fragments isolated from naive combinatorial libraries are generally found to be in the range of lo6 to lo7 dm3 mol-’ with those from immunized spleens reaching lo9 dm3 mol-’. - z age DNA phage DNA with Vgene insertion SCF U OR phage plll fusion phage pVlll fusion Figure 5 The steps required to modify a filamentous bacteriophage particle to generate a phage display library in which the F region of an antibody is expressed as a fusion with one of the major coat proteins on the surface of the bacteriophage Initial screening of cDNA libraries is accomplished by expressing the antigen binding fragments as coat proteins on the surface of a phage.This has the advantage of maintaining a link between the DNA of the heavy and light chain sequences and the protein for which they code (Figure 5). The proportion of antigen specific proteins in the library can then be enriched by a process known as ‘panning’ in which the library undergoes affinity chromatography for the antigen based on an ELISA (enzyme-linked immunosorbent assay) system.Several in vitro methods of mimicking somatic hyper- mutation-affinity maturation as a means of increasing the proportion of antigen reactive proteins have been developed. These include chain shuffling in which a heavy chain F that is known to bind the antigen is combined with a library of light chains or vice versa. It has been shown that when heavy or light chain genes from antigen reactive Fa,fragments are recombined with a library of light or heavy chains respectively the number of antigen reactive Fa fragments that were generated was significantly Griffiths et al.131have produced a large combinatorial library using human Y gene segments as building blocks by combining > lo8 heavy chains and > 8 x lo5 light chains to give 6.5 x 1O1O viable F, fragments.One important observation resulting from the selection of antigen binders from this library is that the highest affinity Fa,s selected have dissociation constants in the 4-60 nmol dm-3 range while Fa,s from a combinatorial library with a repertoire of 1 x lo7members had a lowest K of 0.82 pmol dm-3. More recently a Fa fragment of an anti-HIV1 antibody has been selected from a library and its affinity for the antigen improved into the picomolar range by saturation mutagenesis of the CDRs (complementarity determining regions). Also human single-chain F (scF,) fragments with sub-nanomolar affinities for fluorescein have now been isolated from a large (1.4 x lolo) non-immunized phage display library.132 16 Antibody Expression Improvements in the large-scale production of antibodies is an area of intense academic and industrial research interest.If catalytic antibody fragments are to be used for chemical transformations methods of generating large quantities of the F region that are correctly folded will need to be found. Once cloned the F region can be expressed in several different forms (Figure 6). These include the traditional F and Fa constructs as well as the single-chain antigen binding fragment scF which uses a synthetic peptide linker typically of 10-14 amino acids to prevent the F regions from dissociating. For therapeutic purposes bispecific constructs such as ‘diabodies ’ and chelating recombinant antibodies (CRAbs) have also been generated.133 The choice of expression system is also of importance with examples of whole antibodies and antibody fragments being generated in bacteria fungi plants,134 and insect cells infected with baculo~irus.~~~ The yield of correctly folded antibody varies with each antibody and expression system.There are currently insufficient data to rationally choose the best system for a given antibody. Pliickthun has shown that by mutating several of the residues in the framework region of the Fv of McPc603 an anti-phosphocholine antibody it is possible to increase the in vivo and in vitro yields of isolated and purified functional F by an order of magnitude (per litre of cell culture) relative to the wild type.136 Hein et have shown that a fully functioning IgG murine catalytic antibody could be expressed in tobacco plants whilst a chorismate mutase antibody (1F7) has been expressed as a functional Fa in the cytoplasm of yeast.138j139 Research by Sch~ltzl~~ has shown that the level of expression of eight different catalytic antibodies expressed as Fa,fragments in bacteria can be improved if they are ‘humanized’ by grafting the CDRs of the murine antibody onto the constant region framework of a human antibody.NATURAL PRODUCT REPORTS 1996-N. R. THOMAS 501 ImmunoglobulinG (IgG) 150 kDa single-chain antigen-binding fragment (scF,) (27 kDa) Figure 6 The structures of an intact immunoglobulin G antibody and the artificially produced Fa, F, scF, diabody and CRAb fragments that can be derived from it of monoclonal antibodies in hybridoma supernatants and the 17 Screening Antibodies for Catalysis small quantities of specific Fa or scFv contained within a single Now that it is possible to create libraries of Fa,s or Fvs with colony from a phage display library pose problems when anything up to 6.5 x loLodifferent members and a single attempting to screen directly for catalysis.Several different immunization could result in 103-104 viable hybridomas one of methods of tackling this problem have been reported in the the major logistical and financial problems is the screening of all literature and an estimation of the number of antibodies needed of the candidate proteins for catalytic activity. Both the purity to be screened to identify a catalyst has been estirnated.141 NATURAL PRODUCT REPORTS 1996 17.1 Chromogenic Assays Most of the catalytic antibodies generated have been discovered by screening a limited number (25-100) of different hybridomas by following the release of a chromogenic group.A system which could be applied to a wide variety of reactions is that reported by Gong et al.142in which indigo is precipitated after the antibody has catalysed the formation of 3-hydroxyindole (Scheme 44). 0 H indigo precipitate Scheme 44 17.2 Complementation of Auxotrophic Bacteria and Yeast Two catalytic antibody activities have been identified by using organisms with blocked biosynthetic pathways as selectors. A chorismate mutase antibody (1F7) has been expressed as a functional Fa in the cytoplasm of yeast,138 and it has been demonstrated that this antibody could assume the functional role of chorismate mutase in vivo conferring a significant growth advantage to a permissive yeast strain lacking the natural enzyme.These experiments demonstrate the potential of being able to use catalytic antibodies to complement metabolic defects in eukaryotic cells and the possibility of identifying antibody molecules with enhanced catalytic activity by observing the growth advantage they confer on the cell.139 An orotate decarboxylase antibody143 was detected using a pyrimidine auxotrophic (pyrF) strain of E. coli by growing recombinant F,,s from a combinatorial cDNA library of 16 000 members.In this case six of the cultures produced plasmids which expressed intact Fa genes that conferred a growth advantage to the original mutant. One of the Fa sequences was re-expressed as a single-chain antigen binding protein and this is currently being characterized in more detail. In a related genetic approach Lesley et al.144have proposed a screen for hydrolytic catalytic activity by using substrates that are hydrolysed to release biotin or 4-methyl-5-thiazole ethanol. Expressing recombinant antibodies in a strain of E. coli deficient in the biotin biosynthetic pathway (Abio-gal)should allow the selection of bacteria which contain antibody catalysts and so conferred a growth advantage on their hosts. 17.3 CatELISA Tawfik et reported the development of an immunoassay method for screening large numbers of potential antibody substrate on microtitre plate catalyst 1 enzyme labelled product specific substrate anti bodies coloured product Figure 7 The key steps in a catELISA assay catalysts by modifying the conventional enzyme-linked immuno- sorbent assay (ELISA) approach (Figure 7).In their method the potential substrate for the reaction is immobilized in the wells of a microtitre plate which is then exposed to the antibody library resulting in the cleavage of the substrate in the wells containing catalytically active antibodies. The supernatant is then removed from the wells and an antibody that specifically binds the product is added. This is detected using a second enzyme-linked antibody.Using this method Tawfik et al. were able to screen 1570 clones for esterolytic activity in the cleavage of a p-nitrobenzyl ester. MacBeath and Hil~ertl~~ have recently shown that this method can be successfully extended to reactions involving bond formation rather than cleavage in this case a Diels-Alder reaction. The major requirement for catELISA to be an efficient method of detecting new catalysts is that the anti-product antibody must be able to discriminate between the substrate and product structures. Given that ELISA measurements have an accuracy of 10% the catalytic reaction must be > 10 % of the background rate to be confidently detected. To produce a highly sensitive system care must be taken with the con-centration of detection antibody relative to antigen and the relative dissociation constants for the binding of the detection antibody to the substrate and products.In principle 0.1 femto-mol of antigen can be detected by ELISA in a microtitre plate well. However because of the discrimination necessary between immobilized substrate and product Hilvert et al. estimate that a reasonably efficient catalyst of a reaction with a fair background rate would have to be present at a concentration of between 1 and 10 nmol dmP3. Hybridoma supernatants typically contain antibody concentrations between 30 and 300 nmol dm-314i and so are amenable to this screening method. 17.4 Detection of Catalytic Activity through Irreversible Inhibition Janda et al.148have shown that it is possible to screen for antibodies that have highly nucleophilic residues at their antigen binding sites and hence could display hydrolytic or NATURAL PRODUCT REPORTS 1996N.R. THOMAS transesterification activity (see also Section 14.1). They pro- duced a semi-synthetic combinatorial antibody library that expressed F,,s on the surface of a filamentous phage. Phages displaying Fahswith nucleophilic cysteine residues were detected by panning the phages in microtitre plates that had been derivatized with an a-phenethyl pyridyl disulfide. The Fa fragments with nucleophilic cysteines underwent a disulfide exchange and hence became covalently attached to the microtitre plate. After washing the wells to remove any unbound phage the disulfide bond was cleaved by addition of dithio- threitol and then E.coli were reinfected with the phage. The Fa from one of the colonies selected was then overexpressed purified and shown to catalyse the hydrolysis of a thioester substrate with the following kinetic parameters k,, = 0.030 min-’ K, = 100 pmol dm-3 k,a,/k,,,,t = 30. Fastrez and have shown that it is possible to screen for more sophisticated catalytic activities in selecting a catalytically active /3-lactamase displayed on a filamentous bacteriophage in the presence of a correctly folded Ser 70 mutant which was catalytically inactive. In this case the library was first incubated with a biotinylated P-lactamase mechanism based inhibitor and then passed down a streptavidin column.Several other approaches that utilize biotin-streptavidin technology for the detection of proteolytic activity at extremely low levels have also been reported. It was shown that chymotrypsin could be detected at 10 nmol dm-3 levels,15o whilst using a slightly different approach trypsin at a concentration of 0.5 ng 50 1.11-1 could also be identified.l5I 17.5 PCR Amplification of DNA Tagged Substrates Fenniri et al.15phave developed a system for screening the synthesis or cleavage of peptide bonds. This involves a construct in which a substrate (in this case a pentapeptide) is linked to a 45-mer DNA peptide which contains information on the sequence of the peptide and two primer sequences; this is tethered to a solid matrix.If the substrate is cleaved the polynucleotide tag can be isolated amplified by polymer chain reaction (PCR) to indicate the substrate specificity of protease. This system has been tested with a-chymotrypsin and can detect ca. 1 picomol of the enzyme. In a simplified version of this a-chymotrypsin was used to catalyse the formation of a new amide bond between a matrix supported phenylalanine carboxyami-domethyl ester and a HLeu-cystamine-LeuH group. The disulfide linkage of this group could then be cleaved by DTT and a new disulfide formed with an oligonucleotide tag (the pCantab 5 vector). Hybridization of this tag with the pCantab 5 vector and PCR amplification allowed a-chymotrypsin activity to be detected at 10 picomol levels. I8 Antibody Mutagenesis With the development of site specific mutagenesis of proteins in the early 1980s the possibility of rationally redesigning proteins to give them new properties became a reality.Controlled mutagenesis has resulted in the creation of a large number of examples of proteins with increased thermostability and to a lesser extent increased stability to oxidation heavy metal cleavage and pH changes. However attempts to improve catalytic properties (changes in substrate specificity increased turnover number or changes in pH profile) have met with very limited success. Many consider enzymes to have been com- pletely evolutionary optimized for their biological role by evolutionary pressures.154 Several examples of changes in substrate specificity have been reported for which the specificity optimized by evolution and the transition state analogues used to produce them are in all cases non-ideal site specific mutagenesis potentially offers a method of significantly improving the catalytic efficiency of a given antibody.Several different approaches can be taken. These include substituting amino acids to lower the affinity of the antibody for reaction products so minimizing product inhibition as has been attempted with the hydrolytic antibody NPN43C9 described above. Alternatively mutagenesis can be used to incorporate reactive groups at the active site of the antibody which function as acids bases or nucleophiles; or several mutations could be made to introduce a metal or coenzyme binding sequence.Several groups have created metalloantibodies,16’ but not improved catalysts. 19 Other Catalytic Biomimetic Systems Having looked in depth at the development in antibody catalysis over the past decade this should now be considered in context with other ‘receptor-based ’ catalytic systems both homogeneous and 203 19.1 Protein and Peptide Catalysts Catalytic antibody systems exhibit substrate specificity and reaction regio- and stereo-selectivity comparable to that of enzymes which are now finding increasing use in industry.163 Significant advantages catalytic antibodies have over enzyme systems are (i) the ability to generate antibody catalysts with a preselected reaction and substrate specificity for which there is no natural counterpart (ii) the development of catalysts which avoid expensive cofactor requirements (iii) the catalyst can be produced as a small protein of 25-27 kDa However as can be seen from Figure 8 the rate enhancements imparted by antibodies are orders of magnitude below those of their natural brethren.5 This is because none of the haplens employed so far is ideal.We are still learning how to generate specific catalytic activities and have a considerable way to go before we can fully comprehend the molecular recognition features necessary to generate a fully optimized catalyst without extensive trial and error. Whilst there has been some success in modifying existing enzyme activity either by direct chemical means,164 genetically through mutagenesis as discussed above in Section 18 and most recently by ‘bi~imprinting’,~~~ the success in these areas is limited.Catalytic antibodies provide a viable alternative to these approaches. Other peptide based catalysts that have been reported include the ‘pepzyme’ reported by Johnsson et which consisted of a 14 residue peptide based on leucine and lysine that was capable of forming an a-helical structure which would bind to and catalyse the decarboxylation of oxalylacetate. This peptide displays Michaelis-Menten saturation kinetics and imparts a rate acceleration of lo3 above that observed with simple amines. Atassi and Manshouril6’ synthesized two peptides based on the active sites of trypsin and chymotrypsin. These were reported to catalyse the hydrolysis of amino acid ethyl esters.However independent studies by other research groups have shown these results to be irreprod~cib1e.l~~ One further example of a polypeptide catalysis system is that constant of the enzyme is maintained :lactate dehydrogena~e;’~~ subtilisin ;156 alcohol dehydrogenase ;I5’ a-amyla~e;~~~ gluta-thione reducta~e’~~ and papain,160 but in most cases the mutant enzymes display lower specificity and/or turnover numbers than the wild type enzymes. Because catalytic antibody binding pockets have not been reported by JulialfiS with polyleucine and polyalanine systems which were capable of catalysing the hydrogen peroxide based epoxidation of a,P-unsaturated carbonyl compounds with enantioselectivities higher than 90 % in some cases.A polymer supported version of this catalyst has also been The mechanism of action of these catalysts and whether they have specific binding sites is currently unknown. NATURAL PRODUCT REPORTS 1996 k,4KE (dm3 mol-I s-1)l I o cataiysed (dm3 mol-l s-1) uncatahsed (5'') Antibodbs Figure 8 Rates for uncatalysed [k,,,, (s-')I and enzyme and antibody catalysed [k,,,/K (dm3 mol-' s-')] reactions. The length of the vertical bar provides a measure of binding affinity in the transition state. Enzymes OMP decarboxylase (ODC) staphylococcal nuclease (STN) adenosine deaminase (ADA) carboxypeptidase A (CPA) triosephosphate isomerase (TIM) chorismate mutase (CMU) cyclophilin (CYC) carbonic anhydrase (CAN).For numerical data on these reactions see Tables 2 and 3 19.2 Imprinted Polymers and Synzymes The term 'synzymes' for synthetic enzymes was first coined by O~erberger~~~ to indicate that compounds such as poly(4-vinylpyridine) and poly(N-vinylimidazole) could participate in ester hydrolysis reactions. There is little or no substrate selectivity in these catalysts forerunners of the imprinted catalysts described below. The area has recently been reviewed17* with hydrolytic decarboxylative and redox syn- zymes having been described in the literature. This approach has recently been updated by Menger and co-w~rkers~~~ who have shown that a poly(ally1amine) modified with a combination of different functionalized carboxylates [imidazole (1 5 YO), phenyl (10 %) octyl (7.5 Yo)]hydrolysed bis(p-nitrophenyl) phosphate with a rate enhancement of 31 400 in the presence of 5 '10 Fe3+.Another polymer [oxalyl (10 YO), imidazole (10 YO),thiol(l0 YO) Zn2+ (10 YO)] hydrolysed p-nitrophenyl phosphate around five- fold faster than an antibody.65 This approach offers the potential to be a very cheap simple method of creating catalysts if chiral polymer backbones such as poly(L-lysine) are used. Imprinted polymers were first reported in 1972 by Wulff and co-workers and the area has recently been comprehensively reviewed.174 The basic idea is to create shape specific cavities in polymeric networks by suspending a template molecule in a simple polymerizable mixture containing a reasonable amount of crosslinking reagent and then polymerize this mixture.This generates a macroporous polymer which on removal of the template molecules leaves a selection of non-identical cavities which have a degree of shape discrimination for the template molecule (Figure 9). In more advanced imprinted polymers these cavities can be functionalized by incorporating chemically labile groups in the template molecule or by doping the polymerizable mixture with monomers that interact non-Figure 9 Schematic representation of the formation of microcavities with functional groups using template molecules (T) covalently with the template and form a pre-organized system that is then polymerized. Imprinted polymers have been created which exhibit good molecular discrimination properties and which under the correct conditions can rival antibodies in molecular separation app1i~ations.l'~ Hence the same principle of using transition state analogues to generate binding pockets that bind tighter to transition states than ground states can be applied to create catalytic imprinted polymers.Reactions catalysed by imprinted polymers include ester hydroly~is,"~ elimination177 and decar- boxylation reactions. 17* Bystrom el have reported an example of an imprinted polymer that was capable of directing the regio- and stereo- NATURAL PRODUCT REPORTS 1996-N. R. THOMAS LAIH,. THF imprintedpolymer components 9 Scheme 45 specific reduction of androstane-3,17-dione. In the presence of the polymer and lithium aluminium hydride the carbonyl group at position 17 is selectively reduced the reaction being mediated by a polymer bound hydroxy group (Scheme 45).19.3 Ribozymes Ribozymes are RNA molecules which use nucleotide base pairing to control substrate recognition and have been shown to catalyse the hydrolysis of phosphate esters in other RNA/DNA molecules in the presence of magnesium. Recently there has been a report of the cleavage of an unactivated amide bond in a modified RNA substrate.ls0 Schultz and co-workersl*l have shown that a transition state analogue for a carbon-carbon bond isomerization reaction could be used to select specific RNA molecules from an artificial library. After seven rounds of selection-amplification a RNA sequence which exhibited a k,,,/k,,,, value of 88 for the isomerization was isolated.More recently antibodies that catalyse the same reaction have been reported.201 In a similar study Hilvert et al. searched a RNA library for polynucleotide strands that bound a Diels-Alder transition state analogue.lg2 In this case they identified a sequence with a K of 350 pmol dmP3 for the transition state analogue from a library of ca. 1014 fragments but did not detect any catalysis. The major advantage of ribozymes is that the catalyst carries its own information for replication making the creation of large libraries (> 10l2 molecules) significantly easer than is the case with antibodies where poor protein expression and incorrect folding can severely limit the yields and diversity of an antibody library.A major limitation of ribozymes is their restricted substrate recognition. Currently this is restricted to Watson-Crick and Hoogsten base-pairing and hydrogen bonding to the hydroxy groups of the ribose. 19.4 Macrocyclic Receptor Based Catalysts Biomimetic catalysts have seen many false dawns in which rate acceleration but not true catalysis has been observed. Smithrud .~ and Benkovic,lR3 Kirbyzo3 and Murakami et ~ 1have recently ~ reviewed this area and I will only mention a few examples with relevance to catalytic antibodies here. The Diels-Alder reaction is one example where there has been an extensive amount of research into the use of 'synthetic receptor' catalysts and it is useful to compare the results obtained in these systems with those of the Diels-Alderase antibodies described above.Much of the early research was conducted by Breslow and co-worker~~~~ using cyclodextrin based systems. More recently Walter et al.ls5 developed a trimeric porphyrin host that accelerates the Diels-Alder reaction between pyridine-sub- stituted furan and maleimide derivatives in 1,2-dichloroethane to give the normally disfavoured exo adduct as the only detectable product (Scheme 46). In this case the reaction reaches equilibrium at 65 % of the trimer concentration as the adduct binds ca. 100 times tighter to the trimer than either the diene or dienophile (and the exo adduct ca. 15 times tighter than the endo adduct) and hence causes significant product inhibition which prevents true turnover.N' 0 Toy + c t 0 Scheme 46 The porphyrin trimer has also been used as a true catalyst in an acyl transfer reactionls6 between N-acetylimidazole and pyrid-4-ylmethanol (Scheme 46). It has been estimated in this case that the k,, within the trimer is 6 x ssl corresponding ~ to an effective molarity of ca. 2 mol dmP3. A major problem with this system is that there is a statistical distribution of porphyrin bound substrates inside and outside the binding pocket suggesting that only a small fraction of the trimer will contain an imidazole and a pyrid-4-ylmethanol within the pocket at any given time. Other examples of catalysts include Bender's cyclodextrin based protease,'*' and the rigid cleft based on acridine yellow and Kemp's triacid reported by Wolfe et ul.lSs which catalyses the dissociation of a glycolaldehyde dimer into its monomer with a rate enhancement of cu.360 (Scheme 47).-2 ,Po HO 0 HO f-jH- ~~~~~ 360-fold rate enhancement Scheme 47 19.5 Aggregates and Micelles This in many ways is potentially another 'combinatorial' approach to creating catalysts. A series of long-chain carboxylic acids amines and alcohols are mixed together to form aggregates (clumps) or micelles. These mixtures are then screened for catalytic activity. In the first example Menger and Feilss reported that a long-chain ammonium anilide was hydrolysed at least lo8 times quicker in the presence of hexadecanoate than a similar concentration of acetate.How- ever it has since been shown by Fife and Liu2l0 that the measured rates were due not to the hydrolysis of the anilide but to a kinetically slow anionic-cationic surfactant aggregation process. A different approach has been taken by Suh and Oh211 who have shown that coordinatively polymerized bilayer membranes held together by Co'I' or Ferrl ions were capable of efficient non-specific protein hydrolysis. This preliminary communi- cation has little kinetic detail and the catalysts require further characterization so that their mechanism of action can be identified. 19.6 Zeolites Microporous microcrystalline catalysts based on naturally occurring zeolites (three-dimensional networks of corner-sharing SiO tetrahedra) have been shown to operate as shape- selective heterogeneous catalysts for a number of reactions.lgo Acidic zeolites such as H+-ZSM-5 in which aluminate (A10,) groups replace between one in ten and one in fifty of the silicates function as Brarnsted acids capable of catalysing the conversion of ethene and benzene to ethyl benzene and other alkane conversions.The zeolites offer some control of substrate and reaction regioselectivity due;o the size of the pores. H+- ZSM-5 has a small pore size (5.5 A diameter) which restricts its reactions to single ring systems whilst mordenite derivatives Table 1 A summary of catalyst properties NATURAL PRODUCT REPORTS. 1996 have pores large enough to %ccommodate naphthalene and other polynuclear arenes (7.5 A diameter).19.7 Comparison of Systems Table 1 gives an indication of the primary differences between the various catalytic systems discussed in this review illustrating their individual strengths and weaknesses. 20 Applications and Future Directions 20.1 Lessons from the Past Decade Now that well over 150 primary research papers have been published on catalytic antibodies it is possible to identify certain principles and trends from the observed results. Several critical assessments of the catalytic abilities of antibodies have appeared.lgl.lg3.203 Tawfik discusses the shortcomings of presen- ting the kinetic data concerning antibodies only in terms of rate enhancement (kCat/kuncat).lg1 This represents the extreme and practically improbable conditions with a catalyst operating at its maximum velocity (V,,,) and being present at 1 mol dm-3 concentration and saturated with substrate.Also the data reported in most papers are only for the initial rate of the reaction. They do not indicate if there is true turnover of substrate or any limitations brought about by product inhibition or chemical modification of the catalyst. It is expected that these would be detrimental to the catalytic activity after prolonged reaction periods. In examples where an antibody is involved in an enantioselective resolution it would be useful also to include the enantioselectivity value E that relates the extent of conversion to the enantiomeric excess of the resulting product;lg2 this is now common practice when preparative scale reactions are reported using enzymes.Jacobsenlg3 compared the catalytic abilities of organo-metallic enzyme and antibody catalysts for epoxidation reactions. His data demonstrate that for this particular reaction the use of a manganese salen complex is favoured in terms of cost and catalytic efficiency. He does however point out that the enzyme chloroperoxidase catalyses epoxidations with alkenes that are unreactive with organometallic catalysts. Catalytic antibodies cannot currently compete with existing enzyme or organometallic catalysts in many reactions. However they do offer a means of rationally creating a catalyst for which no suitable alternative has been identified and as a test system for accentuating the effect of a particular form of catalysis on a reaction.20.2 Catalytic Antibody Substrate and Reaction Specificity Catalytic antibodies have been produced which demonstrate regio- and stereo-selectivity equal to and in some cases better than that observed with enzymes.2o3 One area only recently explored which is of relevance to the synthetic chemist who will not want to generate new antibodies for every new substrate is the creation of antibodies with a broad substrate specificity.202 The lack of research in this area is in part due to the current protocols used to identify catalytic antibodies. These would normally result in a promiscuous catalyst being discarded and Imprinted synthetic Macrocyclic Ribozymes polymers receptors Yes No Yes 10' (DNA) 1o2 105 lo3 (amide bond) Catalytic Enzymes Small peptides antibodies Yes Yes Yes lo" 103 1O8 Isomorphous binding sites? Maximum rate enhancement (observed so far) Molecular mass Asymmetric catalysis? > 2000 1000-2000 25 000-75 000 10000-15 000 > 5000 100&3000 Yes Yes Yes Not yet per binding site Yes Not yet NATURAL PRODUCT REPORTS 1996-N.R. THOMAS Table 2 Reactivity and transition state binding data for enzyme reactions (modification of data presented in ref. 5). These data are represented diagrammatically in Figure 8 Reaction Rate specificity Catalytic efficiency Reaction Non-enzymatic t; kUncat/s-' kcat/S-' enhancement kc,,/KMkcat/kuncat /dm3 mol-' s-' (kcatlKm)/kuncat/dm3 mol-' Ksf/Ki Ref.OMP decarboxylase (ODC) Adenosine deaminase (ADA) Carboxypeptidase A (CPA) Triosephosphate isomerase (TIM) Chorismate mutase (CMU) Cyclophilin (CYC) Staphylococcal nuclease (STN) Carbonic anhydrase (CAN) 78 000000 y 120 y 7.3 y 1.9 d 7.4 h 23 s 130000 y 5s 2.8 x lo-'' 1.8 x lo-'' 3.0~lo-' 4.3 x 2.6 x lop5 2.8 x 1.7x 10-13 1.3 x lo-' 39 370 578 4300 50 13000 95 1O6 1.4 x lo'' 5.6 x 10' 2.1 x 1Ol2 1.4 x 10' 1.9 x 10" 6.6 x lo6 1.9~lo6 1.1 x 10' 4.6 x 10' 1.5 x 10' 7.7 x 10' 1.2 x los 5.6x 1014 1 x 10' 1.0 x 109 2.4x 10s 2 x 1023 7.8 x 10l6 2.2 x 1015 5.6 x 1013 4.2 x lolo 5.3 x lo8 9.2 x los 5.9 x 1019 6.9 x lo4 206 1 x lo8 207 2670 208 750 209 450 21 --- Table 3 Reactivity and transition state binding data for catalytic antibody reactions. These data are represented diagrammatically in Figure 8 Reaction Decarboxylation 32A11 Phosphate hydrolysis 38E1 (Scheme 2 1 ) Amide hydrolysis 43C9 (Scheme 19) HF elimination 43D4-3D12 (Scheme 27) Claisen rearrangement 1 1 F 1-2E11 (Scheme 6) Decarboxylation 2 1D8 (Scheme 5) Proton transfer 34E4 (Scheme 28) Ester hydrolysis.43C9 (Scheme 19) Reaction Rate specificity Non-enzymatic enhancement kca,/KM 1; kuncat/S-' kCat/s-' kCat/kUnCat /dm3 mol-' s-' 169 y 1.30 x lo-'' 2.5 x lo-' 1.92 x lo5 3.6 x lop4 8.8 y 2.5 x 10-9 2.0 x 10-5 8.0 x 103 0.130 3.9 y 5.6 x 10-9 8.3 x 10-4 1.48 x 105 2.2 88 d 2.17 x 3.2 x 1.45 x lo3 18.0 18 d 4.5 x 4.5 x 1.0 x lo4 170 12.8 h 1.5 x lo-' 0.28 1.9 x 10* i.4x 103 12 h 1.6 x 0.66 2.1 x lo4 5.5 x 103 23 min 5.1 x 0.87 1.71 x lo3 290 Catalytic efficiency (kcat/Km)/kuncat /dm3 mol-' K,/K Ref.2.7 x lo6 7.0~lo6 75 5.2 x 107 4.56 65 4.0 x 10' 3.7x lo5 55 8.2 x lo6 1.45 x 73 3.8 x 10' 28.9 22 1.1 x los 2.7~lo4 15 3.4 x 108 1.3x lo5 74 5.7 x 105 3x lo6 56 its catalytic activity attributed to a contaminating enzyme. Another factor controlling substrate specificity is the narrow differential binding affinity exhibited by the catalytic antibodies (Figure 8). This means that small decreases in substrate affinity will have a significantly deleterious effect on the catalytic ability of the antibody. One set of encouraging results (discussed in Section 7) has been the ability of researchers to design haptens and elicit antibodies that are capable of selectively catalysing the formation of one of a manifold of products.This is an area in which catalytic antibodies have few rivals and should be exploited further in the future. 20.3 Catalytic Antibody Rate Enhancement As indicated in the introduction enzymes are capable of imparting rate enhancements of up to lo1' through a combination of transition state stabilization and destabilization of bound substrates (Table 2 Figure S).5 Included in Table 2 are the differential binding data (KM/Ki)for the best competitive inhibitors reported for these enzymes that reasonably qualify as 'transition state analogues'. It is interesting to note that in most cases the rate enhancement imparted by the enzymes is several orders of magnitude higher than these values illustrating that we still have a considerable way to go to obtaining the perfect transition state analogue for enzyme inhibition or catalytic antibody generation.Two detailed studies examining differential binding versus rate enhancement in catalytic antibodies have been reported.lg4-*03 Interestingly the relationship between hapten or substrate binding and rate enhancement appears to be obeyed by both antibodies which are thought to operate purely by transition state stabilization and those that utilize an alternative mechanism such as nucleophilic catalysis (Sections 4.2 4.3 and 5) with few exceptions. Most examples of antibodies in the literature display differential binding (K,/KJ and hence rate enhancements (k,a,/kunca,)in the 103-104 region (Table 3 Figure 8).Antibodies are often thought to display very high affinities for their antigens but is this really the case? Recently there has been a significant amount of discussion concerning the absolute limits of antibody-antigen binding kinetics and affinity. The consensus appears to be that in vivo there is little biological benefit in the immune system generating antibodies with affinities higher than 1O'O dm3 mol-' a value derived from the diffusion limited on-rate (ca.lo6 dm3 mol-I s-') and the resi- dence time on the receptor for signal transduction which gives an off-rate lower limit of -lo4 s-l. There are several antibodies that have reported affinities in the 10'1-10'2 dm3 mol-1 s-l range but none that binds its antigen as tightly as avidin binds biotin (10'j dm3 mol-l) or as certain enzymes apparently bind their transition states (Table 2).It could be suggested therefore that the ceiling for the rate enhancement imparted by differential binding of an antibody isolated directly from a normal immunization would be ca. lo6 (K,,/Kl = lop3mol dm-3/10-9 mol dm-3) due to the fact that no antibody exhibits an affinity of greater than lo9dm3 mol-1 for its antigen. This limitation may soon be removed as in vitro it is now possible to select scF\ fragments from phage libraries with sub- nanomolar affinities for small antigens. Therefore it should be possible through the use of mutagenesis to modify further the binding sites of these fragments and optimize their differential binding and hence improve their catalytic ability.As mentioned above the transition state analogues used to create catalytic antibodies are in many cases far from ideal. Their bond lengths molecular geometries and charge distributions are substantially different from those predicted for the transition states them- selves by molecular modelling as indicated in Figure 4. The design of better transition state analogues will result in both better catalytic antibodies and also more potent enzyme inhibitors. The possibility of creating a dynamic binding pocket that is capable of stabilization at multiple points along a reaction coordinate and hence creating a truly accurate enzyme mimic is one area that researchers are only now addressing through the use of heterologous immunization and other techniques which manipulate the binding pocket environment.20.4 Current and Future Applications At the current time no catalytic antibodies are used in laboratory or industrial scale transformations. This is due to both their inefficiency as catalysts and the time-scale and cost of creating and purifying the antibodies concerned. However over the past decade significant advances have been made in both our understanding of hapten design for the creation of catalytic antibodies and concurrently in the yields of correctly folded antibody fragments that can be expressed and harvested in bacteria and plants. Providing similar levels of improvement are forthcoming in the next ten years catalytic antibodies will find a place as a tool for the synthetic chemist.Given the current limitations catalytic antibodies have been developed for use in biosen~ors~~~ and as delivery and activation systems (ADAPT; antibody-directed antibody prodrug ther- apy) for the treatment of cancer (Scheme 48).lg6 There have been promising in vitro results against human colonic carcinoma (LoVo) cells for this system which is a modified version of the antibody-enzyme conjugate system (ADEPT) used for drug targeting that is currently in late clinical trials. CI e h02H c1 I -OH h02H J kt= 1.88 min-' K,, = 201 pmol dm-3 Scheme 48 The general concept of active immunization with a hapl capable of generating antibodies that have a beneficial catalytic role in the body has been explored by several groups.These have concentrated on creating antibodies capable of hydro-lysing cocaine to either of the biologically inactive monoester products benzoylecgonine or ecgonine methyl ester (Scheme 49).lg7 This would be used as a means of curing people of cocaine addiction with antibodies hydrolysing cocaine before it can enter the central nervous system. In vitro studies with mouse polyclonal antibodies indicate some promise for this type of therapy provided we can design haptens that elicit NATURAL PRODUCT REPORTS 1996 slightly more potent antibodies. Preliminary in uivo studies using mice have indicated that active immunization does reduce cocaine induced locomotor activity205 suggesting that the narcotic is being efficiently hydrolysed by antibodies in the immune system.389 &t = 0.0018S-' KM = 490 pmol dms &at/kncat = 540 Ph t + MeOH Ph Scheme 49 Catalytic antibody research has led to a better understanding of protein based catalysis and molecular recognition and has given us an insight into epitope selection and antigen processing. Significant improvements in antibody creation expression and characterization have also resulted as a consequence of the interest in producing antibody catalysts over the past decade. Acknowledgements. I wish to thank the Royal Society for a University Research Fellowship and Dominique Maffre for proof reading this paper. 21 References 1 M.Polanyi Z. Electrochem. 1921 27 143. 2 L. Pauling Chem. Eng. News 1946 24 1375. 3 W. P. Jencks in Current Aspects of Biochemical Energetics ed. N. 0.Kaplan and E. P. Kennedy Academic Press New York 1966 pp. 273-298. 4 R. Wolfenden Nature (London) 1969 223 704. 5 A. Radzicka and R. Wolfenden Science 1995 267 90; J. P. Gutherie J. Am. Chem. SOC. 1977 99 3391. 6 A. Radzicka and R. Wolfenden Methods Enzymol. 1995 249 284; R. Wolfenden and L. Frick in Enzyme Mechanisms ed. M. I. Page and A. Williams The Royal Society of Chemistry London 1987 pp. 97-122. 7 J. Kraut Science 1988,242,533;E. Lolis and G. A. Petsko Annu. Rev. Biochem. 1990 59 597. 8 V. Raso and B. D. Stollar Biochemistry 1975 14 591. 9 R. Summers Catalytic Principles of Enzyme Chemistry PhD Thesis Harvard University 1983.10 G. Kohler and C. Milstein Nature (London) 1975 256 495. 11 F. Kohen J. B. Kim H. R. Lindner Z. Eshhar and B. Green FEBS Lett. 1980 111 427. 12 A. Tramontano K. D. Janda and R. A. Lerner Science 1986 234 1566. NATURAL PRODUCT REPORTS 1996-N. R. THOMAS 13 S. J. Pollack J. W. Jacobs and P. G. Schultz Science 1986 234 1570. 14 D. S. Kemp D. D. Cox and K. G. Paul J. Am. Chem. SOC. 1975 97 7312. 15 C. Lewis T. Kramer S. Robinson and D. Hilvert Science 1991 253 1019; C. Lewis P. Paneth M. H. O’Leary and D. Hilvert J. Am. Chem. Soc. 1993 115 1410; J. Grate R. A. McGill and D. Hilvert J. Am. Chem. SOC. 1993 115 8577; T. M. Tarasow C. Lewis and D. Hilvert J. Am. Chem. SOC. 1994 116 7959.16 M. V. Sergeeva V. Yomtova A. Parkinson M. Overgaauw R. Pomp A. Schots A. J. Kirby and R. Hilhorst Isr. J. Chem. 1996 in the press. 17 J. Smid A. Varma and S. C. Shah J. Am. Chem. SOC. 1979,101 5764. 18 F. H. Westheimer Adv. Enzymol. 1962 24 441. 19 J. D. Stewart L. J. Liotta and S. J. Benkovic Acc. Chem. Res. 1993 26 396. 20 See for example H. Oikawa K. Katayama Y. Suzuki and A. Ichihara J. Chem. SOC. Chem. Commun. 1995 1321 ;S. Laschat Angew. Chem. Int. Ed. Engl. 1996 35 289. 21 P. A. Bartlett and C. R. Johnson J. Am. Chem. SOC. 1985 107 7792. 22 D. Y. Jackson J. W. Jacobs R. Sugasawara S. H. Reich P. A. Bartlett and P. G. Schultz J. Am. Chem. SOC. 1988 110 4841. 23 D. Hilvert S. H. Carpenter K. D. Nared and M. T. M. Auditor Proc.Natl. Acad. Sci. USA 1988 85 4953. 24 M. R. Haynes E. A. Stura D. Hilvert and I. A. Wilson Science 1994 263 646; M. R. Haynes E. A. Stura D. Hilvert and I. A. Wilson Proteins 1994 18 198. 25 A. P. Campbell T. M. Tarasow W. Massefski P. E. Wright and D. Hilvert Proc. Natl. Acad. Sci. USA 1993 90 8663. 26 T. M. Chook H. Ke and W. N. Lipscomb Proc. Natl. Acad. Sci. USA 1993 90,8600. 27 A. Y. Lee P. A. Karplus B. Ganem and J. Clardy J. Am. Chem. Soc. 1995 117 3627. 28 A. Y. Lee J. D. Stewart J. Clardy and B. Ganem Chem. Biol. 1995 2 195; D. Hilvert Curr. Opin. Struct. Biol. 1994 4 612. 29 0.Wiest and K. N. Houk J. Am. Chem. SOC. 1995,117 11628. 30 S. T. Cload D. R. Liu R. M. Pastor and P. G. Schultz J. Am. Chem. SOC., 1996,118 1787; D. R.Liu S. T. Cload R. M. Pastor and P. G. Schultz J. Am. Chem. SOC. 1996 118 1789. 31 J. A. Shin and D. Hilvert Bioorg. Med. Chem. Lett. 1994,5,2945. 32 D. Y. Jackson M. N. Liang P. A. Bartlett and P. G. Schultz Angew. Chem. Int. Ed. Engl. 1992 31 182. 33 Y. Tang J. B. Hicks and D. Hilvert Proc. Natl. Acad. Sci. USA 1991 88 8784. 34 A. C. Braisted and P. G. Schultz J. Am. Chem. SOC. 1994 116 221 1. 35 W. V.-E. Doering and W. R. Roth Tetrahedron 1962 18 67. 36 D. Hilvert K. W. Hill K. D. Nared and M.-T. M. Auditor J. Am. Chem. SOC. 1989 111,9261. 37 A. C. Braisted and P. G. Schultz J. Am. Chem. SOC. 1990 112 7430. 38 C. J. Suckling M. C. Tedford L. M. Bence J. I. Irvine and W. H. Stimson J. Chem. SOC. Perkin Trans. 1 1993 1925; see also the discussion of this research by T.Tarasow and D. Hilvert in Concepts in Protein Engineering and Design ed. P. Wrede and G. Schneider Walter de Gruyter Berlin 1994 pp. 237-262. 39 A. A. P. Meekel M. Resmini and U. K. Pandit J. Chem. SOC. Chem. Commun. 1995 571. 40 V. E. Gouverneur K. N. Houk B. de Pascual-Teresa B. Beno K. D. Janda and R. A. Lerner Science 1993 262 204. 41 J. T. Y. Li-Knohalooma J. A. Ashley C.-H. Lo L. Tucker M. M. Wolfe and K. D. Janda J. Am. Chem. SOC. 1995,117,7041. 42 A. G. Cochran R. Sugasawara and P. G. Schultz J. Am. Chem. SOC.,1988 110 4841. 43 J. R. Jacobsen A. G. Cochran J. C. Stephans D. S. King and P. G. Schultz J. Am. Chem. SOC. 1995 117 5453. 44 E. C. Sherer C. M. Turner and G. C. Shields Int. J. Quantum Chem.1995 1 83. 45 K. Teraishi I. Fujii and H. Nakamura Tetrahedron Lett. 1992 33 7153. 46 I. Fujii F. Tanaka H. Miyashita R. Tanimura and K. Kinoshita J. Am. Chem. SOC. 1995 117 6199. 47 A. F. Monzingo and B. W. Matthew Biochemistry 1984,23,5724. 48 0.Kristensen H. Miyashita D. G. Vassylyev F. Tanaka I. Fujii and K. Morikawa Protein Pept. Lett. 1995 1 252. 49 J.-P. Charbonnier E. Carpenter B. Gigant B. Golinelli-Pimpaneau Z. Eshhar B. S. Green and M. Knossow Proc. Natl. Acad. Sci. USA 1995 92 1171; P. S. Tawfik R. S. Zemel R. Arad-Yellin B. S. Green and Z. Eshkar Biochemistry 1990 29 9916. 50 P. A. Patten N. S. Gray P. L. Yang C. B. Marks G. J. Wedemayer J. J. Boniface R. C. Stevens and P. G. Schultz Science 1996 271 1086; M. M. Davis Science 1996 271 1078.51 E. A. Padlan G. H. Cohen and D. R. Davies Ann. Inst. PasteurlImmunol. 1985 136 271. 52 G. W. Zhou G. Guo W. Huang R. J. Fletterick and T. S. Scanlan Science 1994 265 1059. 53 J. Guo W. Huang and T. S. Scanlan J. Am. Chem. SOC. 1994 116 6062. 54 D. R. Corey and C. S. Craik J. Am. Chem. SOC. 1992,114 1784. 55 S. J. Benkovic J. A. Adams C. L. Borders K. D. Janda and R. A. Lerner Science 1990 250 1135. 56 R. A. Gibbs P. A. Benkovic K. D. Janda R. A. Lerner and S. J. Benkovic J. Am. Chem. SOC. 1992 114 3528. 57 J. D. Stewart V. A. Roberts N. R. Thomas E. D. Getzoff and S. J. Benkovic Biochemistry 1994 33 1994. 58 V. A. Roberts J. D. Stewart S. J. Benkovic and E. D. Getzoff J. Mol. Biol. 1994 235 1098. 59 J. F. Krebs G.Siuzdak H. J. Dyson J. D. Stewart and S. J. Benkovic Biochemistry 1995 34 720 60 D. J. Jeske J. Jarvis C. Milstein and J. D. Capra J. Immunol. 1984 133 1090. 61 M. T. Martin A. D. Napper P. G. Schultz and A. R. Rees Biochemistry 1991 30 9757. 62 P. Wirsching J. A. Ashley S. J. Benkovic K. D. Janda and R. A. Lerner Science 199 1 252 680. 63 H. Suga 0.Ersoy T. Tsumuraya J. Lee A. J. Sinskey and S. Masamune J. Am. Chem. SOC. 1994 116 487. 64 F. Hillerns and D. Rehder Chem. Ber. 1991 124 2249. 65 T. S. Scanlan J. R. Prudent and P. G. Schultz J. Am. Chem. SOC. 1991 113 9397. 66 J. S. Rosenblum L.-C. Lo T. Li K. D. Janda and R. A. Lerner Angew. Chem. Int. Ed. Engl. 1995 34 2275. 67 A. A. Brimfield D. E. Lenz D. M. Maxwell and C. A. Broomfield Chem.-Biol.Interact. 1993 87 95. 68 J.-L. Reymond K. D. Janda and R. A. Lerner Angew. Chem. Int. Ed. Engl. 1991 30 1711. 69 H. Suga N. Tanimoto A. J. Sinskey and S. Masamune J. Am. Chem. Soc. 1994 116 11 197. 70 J. Yu L. C. Hsieh L. Kochersperger S. Yonkovich J. C. Stephans M. A. Gallop and P. G. Schultz Angew. Chem. Int. Ed. Engl. 1994 33 339. 71 C. Barbaud M. Bols I. Lundt and M. R. Sierks Tetrahedron 1995,51,9063; J. Lehmann and B. Rob Tetrahedron Asymmetry 1994 5 2255. 72 B. F. Cravatt J. A. Ashley K. D. Janda D. L. Boger and R. A. Lerner J. Am. Chem. SOC. 1994 116 6013. 73 K. Shokat T. Uno and P. G. Schultz J. Am. Chem. SOC. 1994 116 2261. 74 S. N. Thorn R. G. Daniels M. -T. M. Auditor and D. Hilvert Nature (London) 1995 373 228; see also D.S. Kemp Nature (London) 1995 373 196. 75 J. A. Ashley C.-H. L. Lo G. P. McElhaney P. Wirsching and K. D. Janda J. Am. Chem. SOC. 1993 15 2515. 76 B. L. Iverson and R. A. Lerner Science 1989 243 1184; W. S. Wade J. A. Ashley G. K. Jahangiri G. McElhaney K. D. Janda and R. A. Lerner J. Am. Chem. SOC. 1993 115 4906. 77 M. W. Crowther J. D. Stewart V. A. Roberts C. J. Bender E. Tevelrakh J. Peisach E. D. Getzoff B. J. Gaffney and S. J. Benkovic J. Am. Chem. Soc. 1995 117 5627. 78 N. Janjic and A. Tramontano J. Am. Chem. SOC.,1989 111 9109; A. G. Cochran and P. G. Schultz J. Am. Chem. SOC. 1990 112 9414. E. Keinan S. C. Sinha A. Sinhabagchi E. Bendry M. C. Gnozi Z. Eshhar and B. S. Green Pure Appl. Chem. 1990 62 2013. 79 Y. Feng Z.Liu G. Gao S. J. Gao X. Y. Liu and T. S. Yang Ann. N. Y. Acad. Sci. 1995 750 271; M. Takagi K. Kohda T. Hamuro A. Harada H. Yamaguchi M. Kamachi and T. Imanaka FEBS Lett. 1995,375,273;T. Imanaka and M. Takagi in Antibody Expression and Engineering ed H. Y. Wang and T. Imanaka ACS Symposium series no. 604 American Chemical Society Washington DC 1995 pp. 138-143. 80 S. J. Pollack and P. G. Schultz Science 1989 111 1929. 81 G.-M. Luo L. Ding Z.-Q. Zhu G. Guo Q.-A. Sun Z. Liu T.-S. Yang and J.-C. Shen Ann. NY Acad. Sci. 1995 750 277. 82 C.-H. Wong and G. M. Whitesides in Enzymes in Organic Synthesis Elsevier Oxford 1994 pp. 13 1-194. 83 L. C. Hsieh S. Yonkovich L. Kochersperger and P. G. Schultz Science 1993 260 337. 84 L. C. Hsieh J.C. Stephans and P. G. Schultz J. Am. Chem. SOC. 1994 116 2167. 85 R. W. Watt J. N. Herron and E. W. Voss Mol. Zmmunol. 1980 17 1237. 86 K. D. Janda C. G. Shevlin and R. A. Lerner. Science 1993 259 490. 87 J. E. Baldwin and M. Silusik Tetrahedron 1982 19 2939. 88 J. Na K. N. Houk C. G. Shevlin K. D. Janda and R. A. Lerner J. Am. Chem. SOC. 1993 115 8453. 89 T. Kitazume and M. Takeda J. Chem. Soc. Chem. Commun. 1995 39. 90 L. J. Liotta R. A. Gibbs S. D. Taylor P. A. Benkovic and S. J. Benkovic J. Am. Chem. Soc. 1995 117 4729. 91 S. J. Benkovic A. D. Napper and R. A. Lerner Proc. Natl. Acad. Sci. USA 1988 85 5355. 92 J.-L. Reymond Ange)c. Chem. Int. Ed. Engl. 1995 34 2285. 93 K. D. Janda S. J. Benkovic D. A. Mcleod D. M. Schloeder and R.A. Lerner Tetrahedron 1991 47 2503. 94 J.-L. Reymond K. D. Janda and R. A. Lerner Angew. Chem. Int. Ed. Engl. 1991 30 1711. 95 J.-L. Reymond K. D. Janda and R. A. Lerner J. Am. Chem. Soc. 1992 114 2257. 96 S. C. Sinha E. Keinan and J.-L. Reymond J. Am. Chem. Soc. 1993 115 4893. 97 A. Koch J.-L. Reymond and R. A. Lerner J. Am. Chem. Soc. 1994 116 803. 98 Y. Chen J.-L. Reymond and R. A. Lerner Angeic. Chem. Znt. Ed. Engl. 1994 33 1607. 99 T. Koch J.-L. Reymond and R. A. Lerner J. Am. Chem. SOC. 1995 117 9383. 100 T. Li K. D. Janda S. Hilton and R. A. Lerner J. Am. Chem. Soc. 1995 117 2367. 101 T. Li K. D. Janda and R. A. Lerner Nature (London) 1996,379 326. 102 J. A. Ashley and K. D. Janda J. Org. Chem. 1992 57 6691. 103 C.W. Dufor R. J. Bolin R. J. Sugawara R. J. Massey J. W. Jacobs and P. G. Schultz J. Am. Chem. Soc. 1988 110 8713. 104 K. D. Janda K. J. Ashley T. M. Jones D. A. Mcleod D. M. Schloeder and M. J. Weinhouse J. Am. Chem. Soc. 1990 112 8886; T. M. Spitznagel J. W. Jacobs and D. S. Clark Enzyme Microb. Technol. 1993 15 916. 105 Y. Okahata M. Yamaguchi F. Tanaka and I. Fujii Tetrahedron. 1995 51 7673. 106 L. I. Slobin Biochemistry 1966 5 2836. 107 P. G. Schultz and J. W. Jacobs in Environmental Influences and Recognition in Enzyme Chemistry ed. J. F. Liebman and A. Greenberg VCH New York 1988 pp. 303-335. 108 G. Gallacher C. S. Jackson M. Searcey R. Goel G. W. Mellor C. Z. Smith and K. Brocklehurst Eur. J. Biochem. 1993,214 197. 109 D. Stephens and B.L. Iverson Biochem. Biophys. Res. Commun. 1993 192 1439; D. B. Stephens B. W. Wilmore and B. L. Iverson Bioorg. Med. Chem. 1994 2 653; A. Tubul R. Michel B. Gharib and M. De Reggi Tetrahedron Lett. 1994 35 5865. 110 B. L. Iversen CHEMTECH 1995 17. 111 K. Shreder A. Harriman and B. L. Iverson J. Am. Chem. Soc. 1995 117 2673; M. B. Wallace and B. L. Iversen J. Am. Chem. Soc. 1996 118 251. 112 B. H. Wilmore and B. L. Iverson J. Am. Chem. Soc. 1994 116 2181. 113 S. Paul D. J. Volle C. M. Beach D. R. Johnson M. J. Powell and R. J. Massey Science 1989 244 1158. 114 S. Paul W. Sun R. Mody H. K. Tewary S. Mehrotra T. Gianferrara M. Meldal and A. Tramontano J. Biol. Chem. 1992 267 13 142. 115 Q.-S. Gao M. Sun S. Tyutyulkova D. Webster A. Rees A.Tramontano R. J. Massey and S. Paul J. Biol. Chem. 1994,269 32389. 116 S. Paul L. Li R. Kalaga P. Wilkins-Stevens F. J. Stevens and A. Solomon J. Biol. Chem. 1995 270 15257. 117 L. Li S. Paul S. Tyutyulkova M. D. Kazatchkine and S. Kaveri J. Zmmunol. 1995 33298. 118 A. M. Shuster G. V. Gololobov 0.A. Kvashuk A. E. Bogo- molova I. V. Smirnov and A. G. Gabibov Science 1992 256 665; A. G. Gabibov G. V. Gololobov 0.I. Makarevich D. V. Schourov E. A. Chernova and R. P. Yadav App. Biochem. Biotechnol. 1994 47 293. 119 H. Crespeau A. Laouar and D. Rochu C. R. Life Sci.. 1994,317 819. NATURAL PRODUCT REPORTS 1996 120 L. Joron L. Izadyar A. Friboulet M.-H. Remy G. Pancino A. Roseto and D. Thomas Ann. N. Y. Acad. Sci. 1992 672 216. 121 L.Izadyar A. Friboulet M.-H. Remy A. Roseto and D. Thomas Proc. Natl. Acad Sci. USA 1993 90 8876. 122 D. S. Tawfik R. Chap B. S. Green M. Sela and Z. Eshhar Proc. Natl. Acad. Sci. USA 1995 92 2145. 123 M. StHhl B. Goldie S. Bourne and N. R. Thomas J. Am. Chem. SOC.,1995 117 5164. 124 C. J. Suckling W. H. Stimson G. R. Proctor L. H. Bence L. Brooks A. L. Khalaf C. M. Tedford C. M. Linaza I. McGilp R. Maguire and J. Irvine Biochem. Soc. Trans. 1993 21 1099. 125 H. Suga 0. Ersoy S. F. Williams T. Tsumuraya M. N. Margolies A. J. Sinskey and S. Masamune J. Am. Chem. Soc. 1994 116 6026 8432. 126 P. Wirsching J. A. Ashley C.-H. L. Lo K. D. Janda and R. A. Lerner Science 1995 270 1775. 127 J. Wagner R. A. Lerner and C. F. Barbas 111 Science 1995,270 1797.128 G. Winter A. D. Griffiths R. E. Hawkins and H. R. Hoogen- boom Annu. Rev. Zmmunol. 1994 12 433; D. R. Burton and C. F. Barbas Adv. Zmmunol. 1994 57. 191. 129 E. S. Ward D. Gussow A. D. Griffiths P. T. Jones and G. Winter Nature (London) 1989 341 544. 130 A. S. Kang T. M. Jones and D. R. Burton Proc. Natl. Acad. Sci. USA 1991. 88 11 120. 131 A. D. Griffiths S. C. Williams 0. Hartley I. M. Tomlinson P. Waterhouse W. L. Crosby R. E. Kontermann P. T. Jones N. M. Low T. J. Allison T. D. Prospero H. R. Hoogenboom A. Nissum J. P. L. Cox J. L. Harrison M. Zaccolo E. Gherardi and G. Winter EMBO J. 1994 13 3245. 132 W. P. Wang K. Green S. Pinz-Sweeney A. T. Briones D. R. Burton and C. F. Barbas J. Mol. Biol. 1995 254 392; T. J. Vaughan A.J. Williams K. Pritchard J. K. Osbourn A. R. Pope J. C. Earnshaw J. McCafferty R. A. Hodits J. Wilton and K. S. Johnson Nut. Biotechnol. 1996 14 309. 133 K. A. Chester and R. E. Hawkins Trends Biochem. Sci. 1995,13 294. 134 A. Pluckthun in The Pharmacology of Monoclonal Antibodies ed. M. Rosenberg and G. P. Moore Springer Verlag Berlin Germany 1994 pp. 269-315; D. Filupa M. Rollence N. Essig J. Nagle A. Achari and T. Lee in Antibody Expression and Engineering ed. H. Y. Wang and T. Imanaka ACS Symposium Series no. 604 American Chemical Society Washington DC 1995 pp. 7&85; J. K.-C. Ma and M. B. Hein Trends Biotechnol. 1995 13 522. 135 K. N. Potter Y. Li and J. D. Capra in Antibody Expression and Engineering ed. H. Y. Wang and T. Imanaka ACS Symposium Series no.604 American Chemical Society Washington DC 1995 p. 41. 136 A. Knappik and A. Pluckthun Protein Eng. 1995 8 81. 137 M. B. Hein Y. Tang D. A. Mcleod K. D. Janda and A. Hiatt Biotechnol. Prog. 1991 7 455. 138 K. Bowdish Y. Tang J. B. Hicks and D. Hilvert J. Biol. Chem. 1991 266 11901. 139 Y. Tang J. B. Hicks and D. Hilvert Proc. Natl. Acad. Sci. USA 1991 88 8784. 140 H. D. Ulrich P. A. Patten P. L. Yang F. E. Romesberg and P. G. Schultz Proc. Natl. Acad. Sci. USA 1995 92 11907. 141 B. Posner J. Smiley I. Lee and S. J. Benkovic Trends Biol. Sci. 1994 145; D. S. Tawfik B. S. Green and Z. Eshhar Anal. Biochem. 1992 202 35. 142 B. Gong S. A. Lesley and P. G. Schultz J. Am. Chem. Soc. 1992 114 1486. 143 J. A. Smiley and S.J. Benkovic Proc. Natl. Acad. Sci. USA 1994 91 8319. 144 S. A. Lesley P. A. Patten and P. G. Schultz Proc. Natl. Acad. Sci. USA 1993 90 1160. 145 D. S. Tawfik B. S. Green R. Chap M. Sela and Z. Eshhar Proc. Natl. Acad. Sci. USA 1993 90 373. 146 G. MacBeath and D. Hilvert J. Am. Chem. SOC. 1994 116 6101. 147 J. W. Goding Monoclonal Antibodies Principles and Practice Academic Press Orlando Florida 1986. 148 K. D. Janda C.-H. L. Lo T. Li C. F. Barbas 111 P. Wirsching and R. A. Lerner Proc. Natl. Acad. Sci. USA 1994 91 2532. 149 P. Soumillion L. Jespers M. Bouchet J. Marchand-Bynaert G. Winter and J. Fastrez J. Mol. Biol. 1994 237 415. 150 J. W. Lane X. Hong and A. W. Schwabacher J. Am. Chem. Soc. 1993 115 2078. 151 V. M. Koritas and H.J. Atkinson Anal. Biochem. 1995,227 22. NATURAL PRODUCT REPORTS 1996-N. R. THOMAS 51 1 152 H. Fenniri K. D. Janda and R. A. Lerner Proc. Natl. Acad. Sci. USA 1995,92 2278. 153 T. D. Hua V. Rolland-Fulcrand R. Lazaro P. Viallefont M.-P. Lefranc and M. Weill Tetrahedron Lett. 1996 37 175. 154 J. R. Knowles Phil. Trans. R. SOC. Lond. Ser. B 1991 332 115. 155 H. M. Wilks D. J. Halsall T. Atkinson W. N. Chia A. R. Clarke and J. J. Holbrook Biochemistry 1990 29 8587; C. R. Dunn H. M. Wilks D. J. Halsall T. Atkinson A. R. Clarke H. Muirhead and J. J. Holbrook Philos. Trans. R. SOC. London B 1991 332 177. 156 R. J. Siezen W. M. de Vos J. A. M. Leunissen and B. W. Dijkstra Protein Eng. 1991 4 719. 157 E. H. Creaser C. Murali and K.A. Britt Protein Eng. 1990 3 523. 158 K. Ishikawa and H. Hirata Arch. Biochem. Biophys. 1989 272 356. 159 R. N. Perham N. S. Scrutton and A. Berry Bioessays 1991 13 515. 160 H. E. Khouri T. Vernet R. Menard F. Parlati P. Laflamme D. C. Tessier B. Gour-salin D. Y. Thomas and A. C. Storer Biochemistry 1991 30 8929. 161 V. A. Roberts and E. D. Getzoff FASEB J. 1995 9 100; W. S. Wade J. S. Koh D. M. Hoekstra and R. A. Lerner J. Am. Chem. SOC.,1993 115 4449; see also L. Regan Trends Biol. Sci. 1995 20 280. 162 J. M. Thomas Angew. Chem. Int. Ed. Engl. 1994 33 913; J. M. Thomas R. G. Bell and P. A. Wright Bull. SOC. Chim. Fr. 1994 131 463 ; Catalysis An Integrated Approach to Homogeneous Heterogeneous and Industrial Catalysis ed. J. A. Moulijn P.M. N. M. van Leeuwen and R. A. van Santen Elsevier Amsterdam 1993. 163 Chirality in Industry. ed. A. N. Collins G. N. Sheldrake and J. Crosby Wiley & Sons Chichester UK 1992. 164 E. T. Kaiser. Acc. Chem. Res. 1989 22 47. 165 M. Stihl. U. Jeppsson-Wistrand M.-0. Mansson and K. Mosbach J. Am. Chem. SOC. 1991 113 9366. 166 K. Johnsson. R. K. Allemann H. Wjdmer and S. A. Benner Nature (London) 1993 365 530. 167 M. Z. Atassi and T. Manshouri Proc. Natl. Acad. Sci. USA 1993 90 8282. 168 B. W. Matthews C. S. Craik and H. Neurath Proc. Natl. Acad. Sci. USA 1994 91 4103 ;D. R. Corey and M. A. Phillips Proc. Nutl. Acad. Sci. USA 1994 91 4106; J. A. Wells W. J. Fairbrother J. Otlewski and J. Burnier Proc. Natl. Acad. Sci. USA 1994,91,4110;T.J. Marrone and J. A. McCammon J. Am. Chem. Soc.. 1994. 116 6987. 169 S. Banfi S. Colonna H. Molinari S. Julia and G. Guixer Tetrahedron 1984,40 5207; M. E. Sanchez and S. M. Roberts J. Chem. SOC.. Perkin Trans. I 1995 1467. 170 S. Itsuno M. Sakakura and K. Ito J. Org. Chem. 1990 55 6047. 171 C. G. Overberger and J. C. Salamone Acc. Chem. Res. 1969 2 217. 172 I. M. Klotz in Enzyme Mechanisms ed. M. I. Page and A. Williams The Royal Society of Chemistry London 1987 pp. 14-34. 173 F. M. Menger A. V. Eliseev and V. A. Migulin J. Org. Chem. 1995 60 6666. 174 G. Wulff Angew. Chem. Int. Ed. Engl. 1995,34 1812; J. Steinke D. C. Sherrington and 1. R. Dunkin Adv. Polym. Sci. 1995 123 82. 175 I. A. Nicholls L. 1. Anderson K. Mosbach and B.Ekberg Trends Biotechnol. 1995 13 47. 176 D. K. Robinson and K. Mosbach J. Chem. Soc. Chem. Commun. 1989,969; B. Sellegren and K. J. Shea Tetrahedron Asymmetry 1994 5 1403. 177 J. V. Beach and K. J. Shea J. Am. Chem. SOC. 1994 116 379. 178 K. J. Shea unpublished results. 179 S. E. Bystrom A. Boerje and B. Akermark J. Am. Chem. SOC. 1993 115 2081. 180 X. Dai A. DeMasmaeker and G. F. Joyce Science 1995 267 237. 181 J. R. Prudent. T. Uno and P. G. Schultz Science 1994 264 1924. 182 K. N. Morris T. M. Tarasow C. M. Julin S. L. Simons D. Hilvert and L. Gold Proc. Natl. Acad. Sci. USA 1994,91 13028. 183 D. B. Smithrud and S. J. Benkovic in Perspectives in Supra- molecular Chemistry The Lock-and-Key Principle ed. J.-P. Behr 1994 vol.1 pp. 149-172; see also Y. Murakami J.-I. Kikuchi Y. Hisaeda and 0. Hayashida Chem. Rev. 1996 96 721. 184 R. Breslow Acc. Chem. Res. 1995 28 146. 185 C. J. Walter H. L. Anderson and J. K. M. Sanders J. Chem. Soc. Chem. Commun. 1993 458. 186 L. G. Mackay R. S. Wylie and J. K. M. Sanders J. Am. Chem. SOC.,1994 116 3141. 187 V. T. D. Souza K. Hanahusa T. O’Leary R. C. Gradewood and M. L. Bender Biochem. Biophys. Res. Commun. 1985 129 727. 188 A. Wolfe D. Nemeth A. Costero and J. Rebek Jr. J. Am. Chem. Soc. 1988 110 983. 189 F. Menger and Z. X. Fei Angew. Chem. Int. Ed. Engl. 1994,33 346. 190 M. E. Davis A. Katz and W. R. Ahmad Chem. Muter. 1996 8 1820. 191 D. S. Tawfik Z. Eshhar and B. S. Green Mol. Biotechnol. 1994 1 87. 192 C.-S. Chen Y.Fujimoto G. Girdaukas and C. J. Sih J. Am. Chem. SOC. 1982 104 7294. 193 E. N. Jacobsen and N. S. Finney Chem. Bid. 1994 1 85. 194 J. W. Jacobs BiolTechnology 1991,9,258;J. D. Stewart and S. J. Benkovic Nature (London) 1995 375 388. 195 G. F. Blackburn D. B. Talley P. M. Booth C. N. Dufor M. T. Martin A. D. Napper and A. R. Rees Anal. Chem. 190,62,2211. 196 P. Wentworth A. Datta D. Blakey T. Boyle L. J. Partridge and G. M. Blackburn Proc. Natl. Acad. Sci. USA 1996 93 799; D. A. Campbell B. Gong L. M. Kocheresperger S. Yonkovich M. A. Gallop and P. G. Schultz J. Am. Chem. SOC., 1994 116 2165. 197 G. P. Basmadjian S. Singh B. Sastrodjojo B. T. Smith K. S. Avor F. Chang S. L. Mills and T. W. Searle Chem. Pharm. Bull. 1995 43 1902; D. W.Landry K. Zhao G.X.-Q. Yang M. Glickman and T. M. Georgiadis Science 1993 259 1899; V. Morell Science 1993 259 1828; N. S. Chandrakumar C. P. Carron D. M. Meyer P. M. Beardsley S. A. Nash L. L. Tam and M. Rafferty Bioorg. Med. Chem. Lett. 1993 3 309; E. C. Shere G. M. Turner T. N. Lively D. W. Landry and G. C. Shields J. Mol. Model. [Electronic Publication] 1996 2 62. 198 J. R. Jacobsen J. R. Prudent L. Kochersperger S. Yonkovich and P. G. Schultz Science 1992 256 365. 199 K. D. Janda R. A. Lerner and A. Tramontano J. Am. Chem. SOC.,1988 110 4835. 200 R. Hirschmann A. B. Smith 111 C. M. Taylor P. A. Benkovic S. D. Taylor K. M. Yager P. A. Sprengler and S. J. Benkovic Science 1994 265 234. 201 J. R. Jacobsen and P. G. Schultz Proc. Natl. Acad.Sci. USA 1994 91 5888. 202 F. Tanaka K. Kinoshita R. Tanimura and I. Fujii J. Am. Chem. SOC.,1996 118 2332. 203 A. J. Kirby Acta Chem. Scand. 1996 50 203; A. J. Kirby Angew. Chem. Int. Ed. Engl. 1996 35 707. 204 J. Foote and H. N. Eisen Proc. Natl. Acad. Sci. USA 1995 92 1254 and references therein. 205 M. R. A. Carrera J. A. Ashley L. H. Parsons P. Winching G. F. Koob and K. D. Janda Nature (London) 1995 378 727. 206 S. A. Acheson J. B. Bell M. E. Jones and R. Wolfenden Biochemistry 1990 29 3198. 207 W. Jones and R. Wolfenden J. Am. Chem. Soc. 1986 108,7444. 208 N. E. Jacobsen and P. A. Bartlett J. Am. Chem. SOC. 1981 103 654. 209 F. C. Hartman G. M. La Muraglia Y. Tomozawa and R. Wolfenden Biochemistry 1975 14 5274. 210 W. K. Fife and S.Liu Angew. Chem. Int. Ed. Engl. 1995 34 2718. 211 J. Suh and S. Oh Bioorg. Med. Chem. 1996 6 1067. 212 Y. Murakami J.-I. Kikuchi Y. Hisaeda and 0.Hayashida Chem. Rev. 1996 96 721.

ISSN:0265-0568    

DOI:10.1039/NP9961300479

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Pigments of Fungi (Macromycetes) Melvyn Gill School of Chemistry University of Melbourne Parkville Victoria 3052 Australia ~~ ~~~ ~ ~~~ ~ ~ ~ ~ ~~ ~~~ ~~~ ~~~ ~ ~~ ~ ~ ~ ~ ~ Reviewing the literature published between September 1992 and February 1996 (Continuing the coverage of literature in Natural Product Reports 1994 vol. 11 p. 67) 1 Introduction 2 Pigments from the Shikimate-Chorismate Pathway 2.1 Compounds Derived from Arylpyruvic Acids 2.1.1 Terphenylquinones 2.1.2 Pulvinic Acids and Related Butenolides 2.2 Compounds Derived from Cinnamic Acids 2.3 Compounds Derived from 4-Hydroxybenzoic Acid 3 Pigments from the Acetate-Malonate Pathway 3.1 Pentaketides 3.2 Hepta ke tides 3.3 Octaketides 3.3.1 Anthraquinones and Anthraquinone Carboxylic Acids 3.3.2 Pre-anthraquinones 3.3.3 Coupled Pre-anthraquinones 3.3.4 Pyranonaphthoquinones 3.4 3.5 4 5 5.1 5.2 6 Nonaketides Further Polyketides and Compounds of Fatty Acid Origin Pigments from the Mevalonate Pathway Nitrogen Heterocycles Indole Pigments Miscellaneous N-Heterocyclic Pigments References 1 Introduction The aim of this review like its is to survey the chemical biological and mycological literature dealing with the isolation characterisation and chemistry of colouring matters produced by those fungi that form conspicuous fruit bodies (Macromycetes).Also included as before are pigments from slime moulds (Myxomycetes) and in certain circumstances pigments produced by macromycetes grown in mycelial culture and some colourless metabolites where these are chemically relevant.The Report covers work that has appeared in the literature between September 1992 and the present time (February 1996). As usual pigments are classified according to their probable biosynthesis although experimental evidence in support of this is available in only a few cases. 2 Pigments from the Shikimate-Chorismate Pathway 2.1 Compounds Derived from Arylpyruvic Acids 2.2.2 Terphenjlquinones The isolation of several leucoacetyl derivatives of the terphenyl- quinone cycloleucomelone 1 from the fruit bodies of Boletopsis leucornelaswas described in the earlier reviews.' More recently several additional members of this group have been found during a screening programme that detected 5-lipoxygenase- the aromatic region which allowed location of the hydroxy and acetoxy groups in the terphenyl nucleus.The structure of 6 on the other hand could not be unambiguously assigned in this way. Instead by using 2D NMR and INADEQUATE techniques together with NOE difference and deuterium-induced differential isotopic shift 13C NMR spectroscopy the positions of the hydroxy and acetoxy groups could be ascertained. The leucotetraacetate 4 in particular displayed potent inhibitory activity (IC5, 0.35 PM) against 5-lipoxy- gena~e.~ Extraction of the fresh fruit bodies of Paxillus atroto-mentosus collected from decaying pine in Tokushima Pre- fecture Japan with ethyl acetate afforded the purple spiro- mentins B 7 and C 8,* previously known from European HO OH 0 1 OAc inhibiting activity in the methanolic extracts of B.le~comelas.~ The compounds responsible were identified as the cyclo- leucomelone leucoacetates 2-6 of which only the leuco-hexaacetate 2 and the leucotriacetate 5 were known pre-viously.'* The structures 3-5 followed unequivocally by comparison of their 'H NMR spectra with that of 1. Thus unambiguous acylation shifts were observed for the signals in HO 7 8* * Spiromentin C 8 is erroneously depicted as lacking one of the three lactone-acetal oxygen atoms in ref. 1. 513 collections of the same fungus together with six new terphenylquinol derivative^.^ The structures 9-14 of the new colourless spiromentins E-F respectively were determined by spectroscopic methods particularly high-field NMR experi- ments.Significant from a stereochemical viewpoint is the magnetic equivalence of the four protons at C-2’ (and also of those at (2-3’) in both para-disubstituted benzene rings in spiromentin E 9 compared to their non-equivalence in spiro- mentin F 10. This points to the fact that the spiromentins E 9 and F 10 differ only in the relative configuration at one of the two spiro-carbons. Analogous relationships could be deduced between the spiromentins G 11 and H 12 and between the spiromentins I 13 and J 14 by careful analysis of the NMR spectra of the respective mixtures. 9 10 12 HO 13 It is noteworthy that no pigments of the leucomentin5 or flavomentin6 types were observed during work on the Japanese material.Interestingly the predominant component (ca. 23%) of the extract of Japanese P.atrotomentosus was identified as (4S,5 R)-osmundalactone 15 which was isolated previously along with its 4-0-acetyl derivative after acid catalysed acetylation of leucomentin 3 16.5 Minor constituents of Japanese P.atrotomentosus include the (4R,SS)-ylactone 17 which points to a common origin for the lactones 15 and 17 and the lactone-acetal moieties of the spiromentins E-J 9-14 in the NATURAL PRODUCT REPORTS 1996 form of (2Z74S,5S)-4,5-epoxyhex-2-enoic acid [cf.the ester side chain in leucomentin 3 16].1*2 2.1.2 Pulvinic Acids and Related Butenolides The gasteromycete genera Pisolithus and Scleroderma have been closely linked with members of the Order Boletales on chemotaxonomic grounds.2* ’ This relationship has recently received further support by the isolation of several methylated and chlorinated derivatives of atromentic acid from the tropical fungus Scleroderma sinnamariense and a related species from Malaysia.* Methyl 4,4’-di-@methylatromentate 18 and methyl per-0-methylatromentate 19 were identified from the spec- troscopic data and by comparison with published properties.The enol ester 18 had earlier been described from Pulveroboletus aur~Jlammeus,~ while the permethyl derivative 19 a new natural product was known from the early work on pulvinic acids by Kogl.Io Methyl 2’,5’-dichloro-4,4’-di-O-methylatromentate 20 could be assigned the structure shown from the ‘H NMR spectrum in which the resonances from the protons of the dihalogenated aromatic ring appear separately each as a singlet.8 Chlorinated pulvinic acids have been reported pre- viously from P.a~rlfammeus~ and Xerocomus chrysenteron,ll both of which belong in Boletales.A full account of the isolation and structural elucidation of pisoquinone 21 from a geographical variant of Pisolithus arhizus which was foreshadowed in the previous Report has now appeared.12 The retipolides A-E are a unique group of butenolides isolated from the fruit bodies of the North American bolete Boletus retipes.13.14 Retipolide A 22 together with a small amount of the 5’-hydroxy analogue retipolide B 23 are OR OH HO Me OH 15 MeOH 0 17 R2 18 R’ =R~=H 19 R‘ =Me; R2=H Me0 20 R’ = H; R2 = CI HO OH NATURAL PRODUCT REPORTS 1996-M.GILL OH OH R 0 OH 0 0- 0 0 OH 0 25 R=H 0 24 R=HandOH 26 R=OH 27 responsible for the bitter taste of the flesh of B. retipes while The structures of the retipolides A 22 C 25 and E 27 suggest the corresponding resonance stabilised anions 24 give rise to an origin from three molecules of 4-hydroxyphenylpyruvic acid the intense yellow colour of its flesh. The isolation of the according to the pathway proposed in Scheme 2.14Additional retipolides C 25 and D 26 suggested that the biosynthesis of the support for these ideas was obtained by HPLC detection both fused heterocyclic moiety in retipolide A 22 for example of the methyl ester analogue of the 'dimerisation' product 28 involves enzymatic oxidation of the 4-hydroxyphenyl ring of a and of secoretipolide E 29 in the fungal extract.Again precursor such as 27 followed by appropriate reorganisation chromatography was guided by the properties of synthetic (see below). racemic materials. l4 In order to detect and subsequently isolate retipolide E 27 it was first necessary to synthesise the molecule in racemic form which was achieved according to the chemistry depicted in Scheme 1.14 Guided by the chromatographic properties of the racemic material optically active retipolide E 27 could then be isolated from the crude methanolic extract of B.retipes by means of HPLC. CO.SCoA 0 28 OMe 1 ii Toy H02C OSiMe2But OH 0 Ii OMe t 0 29 t$ t3 __t ii iii I iii ?H OSiMe2Bu' 0 0 0 J iv OMe ToypoMe 27 Iv 0 OH 0 5p 0 vi __c Jv OH 0 25 I vii yo9@JoH 5fio 0 0 OH OH 0 0 (+)-27 22 Reagents i DEAD Ph,P toluene under precisely defined conditions Steps i self-condensation; ii tyrosol; iii oxidative coupling; iv (see Ref. IS) 73 YO; ii Bu,NF THF; iii PCC 91 YOover two steps; monooxygenase; v electrocyclic or acid-catalysed rearrangement; iv 4-(Me0)-C,H,CH,COCO,Me NEt, 25 "C 90 'XI ; v BI, 47 YO vi 0 -t C acyl shift; vii double bond migration and hemiacetal formation Scheme 1 Scheme 2 2.2 Compounds Derived from Cinnamic Acids A bioassay directed fractionation of ethanolic extracts of Inonotus hispidus gave two phenolic compounds that show interesting immunomodulatory and antiviral activity.16These have been identified from the spectroscopic properties as the well known styrylpyrone hispidin 302(0.78 YOdry mass) and a new closely related ketone hispidon 31 (0.06% dry mass).Hispidon 31 [A,, (MeOH)/nm 373 (log e/dm3 mol-' cm-l4.35) and 258 (3.69)] was also detected by using HPLC and TLC in the ethanolic extracts of both fresh and freeze dried specimens which suggests that it is not an artefact. Hispidon 31 is obviously closely related to the P-keto ester 32 that was isolated from I. hispidus some years ago.2 HO do \ 30 OH 0 HoeMe HO 31 OH 0 C02Me HO 32 Hispidin 30 has been synthesised by a convenient three-step sequence in 24 YO overall yield starting from commercially available piperonal." The process which is summarised in Scheme 3 overcomes several perceived drawbacks associated with earlier methods2 OMe + NATURAL PRODUCT REPORTS 1996 Some fungi belonging to the tropical genus Flavolaschia produce small bright orange fruiting bodies.From the cultured mycelium established from the orange fruit bodies of a yet-unidentified species of Flavolaschia collected in Kolobu Ethiopia have been isolated the biologically active 9-methoxy-strobilurins 33 and 34.18The significance of these compounds which are essentially colourless lies in their position in-termediate between the strobilurins e.g.strobilurin A 35 and the oudemansins e.g. oudemansin A 36 which are (E)-P-methoxyacrylate derivatives first isolated from fungi belonging to the genus O~demansiella.~~ The strobilurins and oudemansins display potent antifungal activity and their discovery has stimulated considerable synthetic activity particularly in the agrochemical industry.ls Extensive analogue synthesis and testing has led to a broad-spectrum systemic fungicide (ICIA5504) that is now in commercial A compound of the strobilurin-oudemansin type has also been isolated from the fruit bodies of FZavolaschia calocera a Southern Pacific species introduced into New Zealand.20 OMe 33 M e b o d Me ? ' Me0& \ OMe *Me Me 34 -0.e Me02C 35 OMe 36 HO&I2.3 Compounds Derived from 4-Hydroxybenzoic Acid A synthetic approach towards the ansa-bridged benzofuran-quinone tridentoquinone 37 a constituent of Suillus tridentus has been pubished.21 HO \ The reaction sequence (Scheme 4) begins with commercially available 2,2-dimethy1-2,3-dihydrobenzofuran-7-o138 and pro- ceeds through the key intermediate 39 to which C and iii dimethyl-C, side chains were added sequentially by means of the C-3 carbonyl group and the bromo substituent at C-6 respectively.Unfortunately the exocyclic alkene 40 could not P HOGo pJ \ Me HO Me 30 Reagents i 4-methoxy-6-methyl-2-pyrone, Mg(OMe), MeOH reflux 7 h 52%; ii BCl, CH,Cl, 40"C 22 h 74%; iii EtSH NaH DMF reflux 1 h 64% Me a Scheme 3 37 NATURAL PRODUCT REPORTS 199cM.GILL OMe OMe OMe OMe -&Me Me i ii iii VI __t Me Me0 Me0 Me0 Me0 OMe OMe 38 39 40 E:Z = 217 Me OMe I Me0 Me Me0 Me Me0 Me Me OMe Me Me 0 0 McMurw Me coupling 46 42 R = i-CH=CMe2 41 i xvii t 43 R = i-CH-CMe2 I Br I OH Me Me (9-37 poor yield uncertain stereochemistry Reagents i Fremy's salt; ii Thiele acylation -10 "C; iii KOH DMSO MeI; iv NBS (3.5 equiv.) CCl, reflux; v H,O acetone; vi 3,3- ethylenedioxybutylmagnesium chloride <25 "C; vii SOCl, pyridine -10 "C; viii 2 M H,SO, acetone; ix H, PtO, EtOAc 80 "C 1100 psi; x HOCH,CH,OH PTSA; xi BuLi Et,O -78 "C; xii farnesyl bromide Li,CuCl, -78 -+ 20 "C; xiii NBS H,O THF 12 "C; xiv K,CO, MeOH; xv HIO;2H,O THF; xvi 2 M H,SO, acetone; xvii C,K TiCl, dimethoxyethane reflux 3 h Scheme 4 be reduced directly to the alkane 41 and temporary removal of the acetal protecting group proved essential for success.Subsequent alkylation of the 6-lithio derivative of the halide 41 with farnesyl bromide followed by oxidative cleavage of the terminal double bond in the product 42 via the bromohydrin 43 and the epoxide 44 then gave the aldehyde 45. It had been anticipated from the outset that the ansa-bridge of tridento-quinone 37 might be closed by McMurry coupling of a suitable dicarbonyl precursor [see bond a in formula 371. Consequently the keto aldehyde 46 was exposed to potassium graphite (C,K) and titanium(II1) chloride in dimethoxyethane.Unfortunately this step occurred in only very poor yield and with uncertain stereochemistry and the synthesis of tridentoquinone was not completed. 3 Pigments from the Acetate-Malonate Pathway 3.1 Pentaketides The chemistry of the xylariaceous wood fungus Daldinia concentrica is dominated by metabolites produced by oxidative coupling of naphthalene-1,S-diol.' Recently the ethyl acetate extractives of dried D. concentrica collected in Tokushima Japan have afforded the new binaphthyl ether 47 along with three new benzophenone derivatives.22 The occurrence of 47 in D. concentrica is not surprising since the corresponding monomer 8-methoxy- 1-naphthol and the parent 1,1',8,8'-tetrahydroxy-4,4'-binaphylare both known constituents of this fungus.The novel naphthalenone 48 was isolated from an as yet unidentified species of Daldinia where it occurs along with several new cyto~halasins.~~ It is worth noting the abundance and variety of cytochalasins that have been found in xylari- aceous fungi including such genera as Daldinia," 24-25 Xylaria26s27 and Hypoxylor~.~~ The (colourless) cytochalasins show remarkable cytostatic effects on mammalian cells in tissue culture. Me0 OH Me0 OH 47 OH OMe 48 NATURAL PRODUCT REPORTS 1996 3.2 Heptaketides Me0 0 OH Me0 0 OH Hypocrellin A 49 and several related perylenequinones from fungi growing on bamboo are phototoxic to human immuno- Me0 deficiency virus 49 itself being almost as effective as the plant Me0 and fungal pigment hypericin (see Section 3.3.3).29 There is in fact a rapidly expanding literature on the use of photodynamic compounds in cancer ~hemotherapy~~ and to inactivate virus- infected cells.,l An isomer of hypocrellin A 49 has been detected in situ and with minimal specimen preparation from two tropical lichen species (fungal symbiont?) by using FT laser microprobe mass ~pectrometry.~~ o,H...o 49 3.3 Octaketides 3.3.1 Anthraquinones and Anthraquinone Carboxylic Acids The 8-0-P-~-gentiobioside 50 of xanthorin 1 -0-methyl ether and the 8-O-/3-~-glucopyranoside 51 of o-hydroxyxanthorin 1-0-methyl ether have been isolated from the water-soluble fraction of the ethanolic extract of Dermocybe sp.WAT 22963 and characterised as their peracetyl derivative^.^^ The aglycones corresponding to 50 and 51 were known previously from the ethyl acetate-soluble fraction of the total extract of this Australian fungus.This fraction also contains emodin 1-0-methyl ether 52 a rare natural product known previously only from cultures of the aspen fungus Phialophora ~lba.~~ The isolation from Dermocybe canaria of physcion and erythroglaucin as their 8-0-P-~-glucopyranosides53 and 54 respectively which was mentioned in the earlier Report' has now been published in full.35 OH HO Me0 Me OH 0 50 OH OH 0 51 A0 52 53 R=H 54 R=OH 0 0 55 R=H 57 56 R=Me jJyqc Me0 0 OH HOyypJC02HOH OH 0 Me0 Me0 Me 0 0 0 58 59 Full details of the synthesis of austrocorticinic acid 55 racemic austrocorticin 57 and the methyl ester 58 of austro- corticone by using intramolecular Friedel-Crafts chemistry to assemble the appropriately substituted anthraquinone nucleus have appeared.36 An alternative approach employing Diels- Alder cycloaddition methodology that leads to austrocorticinic acid 55 and to the unnatural (S)antipode of (R)-austrocorticin 57 has also been p~blished.~' Both of these methods were discussed in the previous Report' and details need not be repeated here.The crystal structure of methyl austrocorticinate 56 was determined38 in order to resolve a discrepancy between the melting point of the naturally derived ester (mp 210-212 "C) and that of its entirely synthetic counterpart (mp 189-190 0C).36 The results confirmed the identity of the materials but revealed that crystals of the 'natural' ester but not the synthetic compound were twinned.The toadstool Dermocybe cardinalis from New Zealand contains in addition to a series of unique pyranonaphtho- quinones (see Section 3.3.4) the new anthraquinone carboxylic acid cardinalic acid 59.33The structure 59 for cardinalic acid followed from the spectroscopic data of the natural product and of several methylated and acetylated derivatives and was confirmed by synthesis of the methyl ester permethyl ether 60 by the route outlined in Scheme 5.33 1 ii Me0 0 OMe Me0 0 II iii iv c- Me0 Me Me0 0 0 60 MeO,f::Ms Meo5C02Me TMSO Me A B Reagents i A benzene reflux 4 days 67 % ; ii Ag,O MeI CHCI, 100%; iii B benzene reflux 2 h; iv 150-155 "C 3.5 h 71 % over two steps Scheme 5 NATURAL PRODUCT REPORTS 1996-M.GILL 519 Table 1 Ratio of torosachrysone enantiomers 62 and 63 from Dermocybe and Cortinarius toadstools" Entry Species (S)-62 (%) (R)-63(YO) ee (Oh) Dermocybe splendida 98.0 2.0 96.0 Cortinarius basirubescens 86.9 13.1 73.8 61 WAT 20880 82.3 17.7 64.6 WAT 20880b 82.2 17.8 64.4 WAT 20880" 82.9 17.1 65.7 OH OH 0 WAT 26640 74.2 25.8 48.4 Derrnocybe sp. B 70.7 29.3 41.4 WAT 26641 68.0 32.0 36.0 WAT 24723 62.2 37.8 24.4 Me0 WAT 20881 61.5 38.5 23.0 62 R' =OH; R2= Me WAT 20881 62.3 37.7 24.6 63 R' =Me; R2= OH WAT 24723 57.0 43.0 14.0 WAT 24274f 55.3 44.7 10.6 54.2 45.8 8.4 Clavorubin 61 has been isolated from the Australian WAT 24274e WAT 20934g 15.9 84.1 68.2 Cortinarius sp.WAT 24723 and characterised both as its WAT 20934e 16.2 83.8 67.6 methyl ester and permethyl ether by spectroscopic analysis and by comparison with synthetic material^.^^ Clavorubin 61 was Determined by chiral HPLC for conditions see Figure 1. Sample obtained by known previously only from the ascomycete Claviceps purpurea acid hydrolysis of the corresponding gentiobioside'. Sample from entry 3 after repeated subjection to the isolation procedure. Sample from fungus collected in (ergot) .2 1989 and stored at -20 "C. Sample isolated from fresh material (1 993). fSample of torosachrysone stored at 4 "C for 1 year. @Sample isolated from fungus collected in 1992 and stored at -20 "C.3.3.2 Pre-an thraqiiinones The dihydroanthracenone torosachrysone occurs as an aniso- chiral mixture of the (S) and (R) antipodes 62 and 63 respectively in toadstools belonging to Cortinarius and Derrnocybe. Degradative and spectroscopic methods developed to determine the chirality of the predominant enantiomer and chemical methods for estimating the enantiomeric excess (ee) in certain specific cases were discussed ear1ier.l Recently it has been found that HPLC analysis of samples of torosachrysone isolated from a variety of related species of Cortinarius and Dermocybe using a chiral stationary phase can conveniently answer both questions (Figure l).40The results (Table 1) reveal somewhat disconcertingly that the stereochemical integrity of this pivotal natural product varies considerably even among a closely related group of toadstools.It is interesting to speculate as to whether or not the ratio of the enantiomers 62 and 63 will prove to be of chemotaxonomic or other significance. Dihydro- anthracenones with both (S) and (R) configurations e.g. the torosachrysones 62 and 63 and aloechrysone 64 have also been found in The toadstool Dermocybe splendida is a particularly rich source of anthraquinones and tetrahydroanthraquinones and their 8-0-/3-~-gentiobiosides.'**A minor red tetrahydroanthra- quinone that has recently been isolated from this toadstool has been identified as (1 S,3R)-austrocortirubin 65 the C-3 epimer of the major red pigment.42 The trans relationship between the hydroxy groups in the tetrahydroaromatic ring in 65 was established from the lH NMR spectrum while the absolute stereochemistry followed from catalytic hydrogenolysis of 65 to the known dextrorotatory 1-deoxy derivative 66.Me0 OH 0 mMe OH 64 0 OH R I I I 0 10 20 30min Me0WGe Figure 1 HPLC traces from a Chiralpak-AD column with toro- sachrysone 62 and 63 from (a) WAT 20880 and (b) WAT 20934. 0 OH Solvent ethanol; flow rate 0.5 ml min-'; detector UV at 280 nm; 65 R=OH column 0.46 x 25 cm. 66 R=H NATURAL PRODUCT REPORTS 1996 A€ 3.3.3 Coupled Pre-anthraquinones The icterinoidins A 67 B 68 and C 69 are 5,S-coupled pre- anthraquinones isolated by careful chromatography from extracts of the New Zealand fungus Dermocybe icterin~ides.~~ The atropisomeric relationship between the icterinoidins A 67 and B 68 was evidenced by the respective CD spectra (Figure 2) from which the axial chirality could also be deduced by comparison with bianthranyls of known configuration.' The pigments 67 and 68 display distinct differences in their lH NMR spectra particularly in the methylene proton region; they must therefore possess the same central stereochemistry although this has not yet been defined.Icterinordin C 69 is a severely sterically-congested dehydro dimer of dehydroicterinoidin B amounting to a ' tetrameric' octaketide. Molecular modelling suggests that the C-10 bianthrone bond can be accommodated only when the C-10 proton and the rings A and B of the adjacent anthraquinone substituent are located on the same face of the anthrone ring.OH OH 0 OH OH 0 HOWO Me H HO@OH Me 67 68 OH 0 OH Me0 Me 0 70 This leads to (S) stereochemistry at the C-10 stereogenic centres and a bianthrone axis that has no choice other than to assume a conformation in which the C-10 hydrogens are gauche- rather than anti-staggered as shown in Figure 3. From the bright orange mycelium and diminutive purple-red fruit bodies of the Australian Dermocybe sp. WAT 26641 have been isolated the 7,7'-linked tetrahydroanthraquinones 70 and '0" Figure 3 Computer simulation of icterinoidin C 69 showing the OH 0 OH conformation of the anthrone rings. Key Filled spheres = carbon; 69 open spheres = hydrogen; hatched spheres = oxygen.NATURAL PRODUCT REPORTS 1996M. GILL 52 1 0 0 .:-: OH 0 HO . OH Me0 Me OH o% 71 0 74 0 0 0 OH 0 OH I II 1 Me0 0 03 OH 72 R1=OH; R2=Me 75 0 73 R' = Me; R2 = OH MeO@4Me0,g$&,Me II II 71.39The structure and relative disposition of the hydroxy groups within the individual tetrahydroaromatic rings in the dimer 70 followed from direct comparison of the lH NMR data with those of the (1S,3S)- and (lS,3R)-diastereoisomers 72 and Me0 0 Me OH 73 respectively of austrocorti1utein.l The absolute configur- I OH 0% ation at the various chiral centres in the pigments 70 and 71 76 0 remains to be determined but the axial stereochemistry is 0 thought to be (R)in both cases as shown from the CD spectra.Tetrahydroanthraquinones of the austrocortilutein type are known in all four stereochemical modifications from various Australian Dermocybe species but the isolation of 70 and 71 was the first occasion on which dehydro dimers of these quinones had been found in the higher fungi.44 Pigments of the unique dermocanarin class were mentioned in the previous report.' A full account of the isolation and structural elucidation of the first three members of this new troscopic data with particular use being made of those regions type of biaryl the dermocanarins 1 74,2 75 and 3 76 has now in the lH NMR spectrum associated with the protons of the been p~blished.~~ The dermocanarins 1-3 are not pre-anthra- methyl and methylene groups of the lactone bridge.quinones in a proper sense but are included here for convenience Dermocanarin 4 77 was found subsequently in the fruit and because of their obvious close relationship to other bodies of Cortinarius sinapicolor a toadstool that has proved to members of the same group (vide infra) in which the tricyclic be useful in probing the biosynthesis of the dermocanarin~.~~ nucleus is indeed only at the dihydroanthracenone level of Thus by feeding [Me-13C]methionine and the sodium salts of development. Interestingly the pigments 74-76 were isolated [1-13C] [2-13C] and [1,2-13C,]acetate separately to young fruit not from the fruit bodies of Dermocybe canaria a common bodies growing in their natural habitat and thereafter examining toadstool in Nothofugus forests in Tasmania and New Zealand the 'H and 13C NMR spectra of the dermocanarin 4 77 that but rather from the subterranian mycelium.The structures resulted it was possible to deduce that the biogenesis of 74-76 including the relative stereochemistry between the dermocanarin 4involves intermediates that are interrelated as stereogenic axis and centre were deduced from the spec- is summarised in Scheme 6. 0 8xMe-CO2Na -CO.SEnz OH OH 0 000 70 cleavage of bond a,& -' " ' O y qMe pOH H cleavage of bond c OH OH OH OH 0 79 62/63 0 I OH tFrom S-adenosylrnethionine 77 0 Incorporation of sodium [1,2-l3Cz]acetate into dermocanarin 477 Scheme 6 Thus from the experiments involving singly labelled pre- cursors it was apparent that (with the exception of the two 0-methyl carbons which come from methionine) the dihydro- anthracenone moiety on the one hand and the naphthoquinone together with the lactone bridge on the other both arise independently by way of regular octaketide assembly.A 2D INADEQUATE experiment on 77 isolated after feeding C. sinapicolor with sodium [1,2-13C,]acetate showed clearly that the tricyclic nucleus is formed as in torosachrysone 62/63 itself by cleavage of bond b in the hypothetical P-keto carboxylic acid 78. However the naphthylbutanoic acid building block 79 arises by retro-Dieckman type cleavage of bond a in 78 accompanied by loss of CO,. The crucial observation that supports this suggestion is that of strong I3C-l3C coupling between C-1' and C-2' in 77 showing that both atoms originate from the same acetate unit.This precludes the alternative route involving cleavage of bond c in 62/63. The (R) stereochemistry at the biaryl axis in 77 was implied from the CD spectrum which displays a strong A-type* Cotton effect couplet. It follows then that the configuration at C-3' must be (R) but the stereochemistry at C-3 remains unknown. The dermocanarins 5 and 6 80 mentioned in the previous report,l also exhibit A-type CD curves and since their 'H NMR spectra are different in detail they must be epimeric at C- 3.3g Dermocanarin 8 81 is a cometabolite of dermocanarin 5 in Dermocybe sp. WAT 24273 while dermocanarin 9 82 is found alongside dermocanarin 6 in the toadstool WAT 24723.39 It is notable that dermocanarin 9 82 possesses (S)chirality both at the axis and at C-3'.Since the atropisomeric dermocanarins 8 and 9 display different 'H NMR spectra they must have the same stereochemistry at C-3.3g Me0 HO Me OH 80 0 '0 81 0 Me OH Me0 0 I I OH "-& 0 82 * For an explanation of this terminology as it applies to coupled pre-anthraquinones see ref. 1. NATURAL PRODUCT REPORTS 1996 The most recent addition to the group is dermocanarin 10 83 which has been isolated from the diminutive Derrnocybe species WAT 2664046 and WAT 24274.47 Dermocanarin 10 83 is an isomer of dermocanarin 1 74 in which the biaryl bond and the oxygen terminus of the lactone bridge span C-5 and C-6 (rather than C-7 and C-8) in the anthraquinone nucleus.3.3.4 Pyranonaphthoquinones The majority of quinonoid and pre-quinonoid pigments from Dermocybe and Cortinarius are octaketides that have structures that accord with the folding pattern shown for structure 78 (OH in place of OMe in some cases) in Scheme 6. This mode of biogenesis leads ultimately to anthraquinones with emodin and endocrocin frameworks or as we have seen so far in this section to related pre-anthraquinones or their dehydro dimers.' Recently the first exceptions to this pattern have been found during an investigation of the New Zealand toadstool Dermo-cybe cardinalis ethanolic extracts of which show activity at low concentration (ICs0 < 0.5 pug ml-l) against murine leukaemia P388.The constituents responsible not only for the activity but also for the striking yellow and purple-red pigmentation of the fruit bodies are a group of naphthoquinones 84-92 of the 0 Me Me OH 84 0 Me Me OH 85 0 Me OH 0 Me Me 0 86 OH 0 Me Me 0 87 R=a-H 88 R=P-H OH0 Me Me 0 89 NATURAL PRODUCT REPORTS 1996-M. GILL Me0 Me H 90 0 Me Me 91 0 Me Me H 92 benzoisochromanquinone type." 49 These compounds the cardinalins are stereochemically complex dehydro dimers related to plant and mould metabolites such as the eleutherins ventiloquinones nanaomycins and the actinorhodins. The structures of the cardinalins 1-6 (84-89) and 8-10 (90-92) were deduced from spectroscopic data of which lH and 13C NMR experiments proved particularly informative.Some selected data that lead to the structure and relative stereochemistry 84 for cardinalin 1 are collected in Figure 4. The biaryl axis in the cardinalins 1-3 (84-86) and 8-10 (90-92) is chiral by virtue of restricted rotation. The absolute configuration at the axis in cardinalin 3 86 [and also presumably in the related compounds 84 85 and 90-921 is thought to be (S),as shown in formula 86 from the presence of a strong B-type bisignate Cotton effect in the CD spectrum. The biologically most potent constituents of D. cardinalis are the deep red cardinalins 4 87 and 5 88 in which the quinonoid moieties are further linked by a furan bridge.Cardinalin 4 87 for example exhibits an IC, value of 0.28 pg ml-l against the P388 cell line. Certain features of the cardinalins still remain undefined at this time. For example the structure of cardinalin 7 (which seems to have suffered contraction of one of the peripheral rings) the absolute stereochemistry of the compounds and the folding pattern in the putative octaketide biosynthetic precursor to the cardinalins are not yet known. 2.17 ddd 13.8,4.2 1.8 Hz 1 3.4 Nonaketides Full details of the synthesis of (S)-(+)-dermolactone 93 beginning with ethyl (S)-lactate have been published.50 An account of this synthesis during which the tetracyclic nucleus was assembled by way of a regiospecific cycloaddition between the known naphthoquinone 94 and the novel highly functional- ised chiral butadiene derivative 95 was given in the previous report' and details need not be repeated here.Comparison between dermolactone as it is isolated from the Australian toadstool Derrnocybe sanguinea (sensu Cleland) and the synthetic quinone 93 revealed that the natural product consists of an anisochiral mixture of 93 and its antipode in which the former predominates. The natural enantiomeric excess of 93 was determined to be 28.6% by 'H NMR shift experiments of the corresponding permethyl ether derivative using [( + )-Eu(hfc),]. 93 mc,OMe Mey; Me0 0 Me OSiBu'Me 94 95 A reinvestigation of the water soluble fraction from D. sanguinea (sensu Cleland) has led to the isolation of the intensely fluorescent 8-0-P-~-gentiobioside96 of dermo-chrysone.The yellow pigment 96 was characterised from the spectroscopic properties of the corresponding hepta-0-acet- ate.40 The aglycone was known previously from the organic soluble extractives of the same toadstool. Chemical studies on Australian and New Zealand species of Cortinarius and Derrnocybe have been reviewed recently both from chemica151 and taxonomic ,OH HO Me0 OH OA Me I 96 3.07ddds 4.2 11.5 13.2 Hz 1.62 Id 13.8 11.5 Hz \ 1.29 d 6.2 Hz 'Ir 201.5 MeO' 1.56 d. 5.9 Hz '1 A 12.031 2.69 dd.9.2 13.2Hz 3.83 ,,,/ HO'I 70 I 1.67 ddd 1.8 4.0 12.6 Hz 4,62 brd 2.6 Hz/ 1.77. q 12.6 Hz 524 3.5 Further Polyketides and Compounds of Fatty Acid Origin Echinotinctone 97 the first naturally occurring pigment with a simple fluorone chromophore has been isolated as a minor orange constituent from the fruit bodies of the 'Indian Paint Fungus ' Echinodontium tinctorium collected in the western United States and from Pyrofomes albomarginatus found in Malaysia.53 With diazomethane echinotinctone 97 [A,, (MeOH)/nm 242 (loge/dm3 mol-1 cm-' 3.54) 442 (3.30) and 468 (3.22)] forms the lemon yellow monomethyl ether 98 which is itself reduced by using sodium borohydride to the dihydro derivative 99.lH and 13C NMR experiments on the natural product 97 and its derivatives 98 and 99 and comparison of these data with those obtained from trametin pentamethyl ether 100 (trametin dimethyl ether was obtained from the methanolic extracts of Gloeophyllum odoratum and G.sepi- arium2) led to the fluorone structure shown. The major pigments of Echinodontium tinctorium and Pyrofomes albo- marginatus have so far resisted effective purification ; in admixture they change colour from yellow-brown to purple- violet on exposure to ammonia. Me Me Me Me 0 OR HO OMe 97 R=H 99 98 R=Me "0 " O W OMe 100 The yellow plasmodia1 pigment ceratioflavin A 101 and four closely related colourless 6-alkyl-4-methoxypyran-2-ones, the ceratiopyrones A-D 102-105 have been isolated from the slime mould Ceratiomyxa fruticulosa (Myxomycete~).~~ The degree of conjugation in ceratioflavin A 101 was evident from the electronic spectrum [A,, (MeCN-H,O)/nm 4171 and the non-conjugated double bond was located from the frag-mentation pattern in the mass spectrum.Mass spectrometry also helped to define the pattern of unsaturation in the ceratiopyrones C 104 and D 105. To determine the position of the diene unit in ceratiopyrone B 103 the compound was exposed to osmium tetroxide and the tetrol so formed was trimethylsilylated. The mass spectrum of the resulting tetrakis- (trimethylsilyloxy) derivative then revealed the location of the silyloxy groups and consequently the double bonds in 103. OMe I 101 OMe 102 OMe Me 1 03 NATURAL PRODUCT REPORTS 1996 OMe I Me 104 OMe Me 105 Very little is known about the chemistry of marine-derived heterotrophic fungi.Although not strictly within the terms of reference of this Report it is worth a note in passing that developments in this area are taking place most recently with the discovery of the a-pyrones 106 and 107.55 A series of benzoquinones with singly and doubly unsaturated carboxylate side chains have been isolated from the Japanese fungus Phlebia chrysocrea. The principal pigment has been assigned the structure OMe 106 107 0 Me0 0 108 4 Pigments from the Mevalonate Pathway Vitamin B, and other highly polar pteridine derivatives that are biosynthetically related to this primary metabolite are responsible for the pigmentation of many fungi belonging to the genus Russula.2 However in the bright orange-yellow outer skin of the North American toadstool Russula jlavida have been found a pair of lipophilic pigments the russulaflavidins A 109 and B 110 which are quinone methides based on a shionane (D :Afriedo-18,19-secolupane) ~keleton.~' The relative stereochemistry in 109 and 110 followed from difference NOE experiments while the absolute configuration was deduced by comparison of the CD spectrum of russulaflavidin B 110 with that of the plant product pristimerin.R. JEavida collected in Japan displays identical TLC and HPLC chromatograms to its North American counterpart indicative of remarkable genetic stability over a considrable geographical area. Me HO the 109 he HO Me 110 NATURAL PRODUCT REPORTS 1996-M. GILL 5 Nitrogen Heterocycles 5.1 Indole Pigments The simple indolone 111 was isolated initially as a glycoprotein conjugate from the edible Japanese mushroom Pleurotus salmoneostramineus by aqueous extraction followed by repeated gel filtration.Precipitation of the colourless glycoprotein and its associated metal ions with acetone left a pink solution from which the indolone 111 was isolated by cry~tallisation.~~ $8 N o N H H 0 0 111 112 115 A much more elaborate indole-based pigment haematopodin 112,has been isolated from the red-violet methanolic extracts of the wood-rotting fungus Mycena haematopu~.~~ The structure of haematopodin 112 which is believed to be an artefact formed by breakdown of a very sensitive native precursor was elucidated from analysis of the spectroscopic data and was confirmed including the (6R) absolute stereochemistry by a single crystal X-ray analysis.(&)-Haematopodin 112 has been synthesised from 6,7-dibenzyloxyindole 113 according to the chemistry depicted in Scheme 7.60It is interesting to note that Me / rN PhCH20 PhCH20 PhCH20 .. PhCH20 1 113 ii,iii rCHO rCN PhCH20qhPhCH20jp PhCH20 Boc PhCH2O Ivi t ii.viii PhCH20 6 Boc PhCH20 Boc 0 BOC c-- %o: 0 0 H H 0 0 (*)-112 DMB 2,6dimet hoxybenzyI Reagents i CH,=NMe,+ C1- CH,Cl,; ii MeI; iii NaCN; iv Boc,O DMAP; v DIBAL-H benzene; vi (MeO),CH MeOH PTSA; vii H, 10% Pd-BaSO,; viii DDQ; ix 114; x CF,CO,H neat Scheme 7 a biomimetically designed approach to 112 that involved intramolecular cyclisation of a tryptamine-6,7-quinone to yield the putative precursor 115 foundered when 115 could not be induced to undergo oxidative cyclisation.Nevertheless the concept led to a successful synthesis of other natural products e.g. marine pyrroloquinones of the damirone class.59 U H OVNV0 H H H H 116 117 118 R’=R2=R3=H 119 R’=Me; R2=R3=H 120 R’ =Me; R2=OH; R3=H N 121 R’=Me; R2=R3=OH H H 122 Me I Me02CyNyC02Me \I Me Me 123 The myxomycete Arcyria denudata and several related slime moulds produce bisindolylmaleimides such as arcyriarubin A 116 and arcyriaflavin A 117 during their rapid transition from the colourless slimy plasmodia1 form to the delicate and brightly coloured sp0rangia.l.A closely related type of indole derivative in the form of lycogalic acid A 118 its dimethyl diester 119 and the hydroxy derivatives 120 and 121 has recently been isolated from the common slime mould Lvcogala epidendron by Steglich and coworkers who also detected traces of the bisindolylmaleimides 116 and 117 and staurosporinone 122 in the same organism.61 Asakawa et al. independently examined L. epidendrum and also found the esters 119-121 (initially termed the ‘lycogarubins ’) the structures of which were confirmed by X-ray analysis of a single crystal of the permethyl derivative 123 of dimethyl lycogalic ester C.62The Japanese group have also reported the presence in L. epidendrum of a series of novel polyacetylenic tri-and di-glycerides the lycogarides.63 64 The symmetrical structures of the lycogalic esters suggests a biosynthesis perhaps shared with the arcyriarubins and other bisindolylmaleimides that proceeds via oxidative dimerisation of a 3-(indol-3-yl)pyruvate e.g.124 followed by hetero-cyclisation of the resulting 1’4-dicarbonylintermediate with an ammonia eq~iva1ent.l~-61 This idea has found strong support in the efficiency of the biomimetic synthesis of dimethyl lycogalic ester 119 that is shown in Scheme 8.61This approach has also been applied to the synthesis of other 3,4-diarylpyrole-2,5-dicarboxylates and thence by oxidative decarboxylation to 3,4-bisarylmaleimides such as the marine alkaloid polycitrin A.65 The ability of some bisindolylmaleimides to inhibit protein kinase C and their therapeutic potential as drugs against autoimmune diseases has stimulated synthetic activity in the pharmaceutical sector towards the arcyriarubins e.g.116,and staurosporinone 122. Workers at Eli Lilly have developed Steglich’s synthesis of arcyriarubin A l16l. such that this pigment is now available in 72% yield in a single step from indolylmagnesium bromide and 2,3-di~hloromaleirnide.~~ When this is coupled with the Roche method for oxidative C02Me 0-pi ii H 124 1 iii H H H 119 Reagents i NaOMe MeOH; ii I (0.5 equiv.); iii NH,OH reflux 42 Yo Scheme 8 H H OYNF0 Q)--+-n H OCI R - H.. 116 ii-iv H Q)-%-JJH H 122 Reagents i indolylmagnesium bromide (5 equiv.j toluene-Et,O-THF (5 1 lj 72%; ii Pd(OAcj,; iii LiAlH,; iv 10% Pd-C H, 47% Scheme 9 cyclisation of 116 and removal of one of the maleimide carbonyl~,~'*the resulting route to staurosporinone 122 is 68 currently the shortest known affording 122 in 34% overall yield (Scheme 9). Arcyriaflavin 117 is also available albeit in only 10O/O yield from 2,3-bis(3'-indolylmercapto)maleimide on treatment with palladium di~hloride.~~ Approaches to bisindolylmaleimides involving [4 +21 cycloaddition of electron-deficient dienophiles to 2,2'-biindole are less efficient still.'O 5.2 Miscellaneous N-Heterocyclic Pigments Alkaloids containing the canthin-6-one skeleton 125 are produced by some toadstools. The parent 125 itself occurs in the bitter-tasting Cortinarius infractus in which are also found the P-carboline derivatives infractin A 126 and B 127.2-14 Canthin-6-one 125 and several sulfur containing derivatives are present in the North American bolete Boletus curtisii a mushroom that is made remarkable by the protective layer of bright yellow slime that covers the fruit bodies.The pigments responsible for this colour are curtisin A 128 and deoxycurtisin A 129 which are accompanied by the colourless methylthio canthin-6-ones 130-133 as well as by 125.14 The 'Peppery Bolete' Chalciporus piperatus owes its dis- tinctive flavour to the 2H-azepines chalciporone 134 and chalciporonyl propionate 135 which co-occur with the mild 3H-azepines isochalciporone 136 and dehydroisochalciporone 137.l4<'lSimple 2H-azepines were not known as natural products prior to this discovery and very little was known NATURAL PRODUCT REPORTS 1996 1 C02Me 125 126 R=H 127 R =OH SMe 128 R =OH 130 129 R=H SMe 131 132 133 U 134 -L( H-Me o,//,Me 135 0 Me Me 0 137 about their chemistry.Consequently a general synthesis of enantiomerically pure 2,7-dialkyl-2H-azepines was developed (Scheme 10). It was found that like their natural counterparts synthetic 2H-azepines such as 139 (R1= Me; R2= Bu) are both extremely pungent and relatively labile rearranging to their thermodynamically more stable 3H isomers on standing in organic sol~ents.'~ '1" + ~ Li+>R2 R2 R' YH THPO I Boc 138 Boc I i-iii Boc Boc Reagents i Ac,O DMAP; ii PPTS; iii H, Lindlar catalyst; iv PCC; v TFA Scheme 10 NATURAL PRODUCT REPORTS 1996-M.GILL 527 23 M. S. Buchanan T. Hashimoto S. Takaoka Y. Kan and Y. Asakawa Phytochemistry 1996 42 173. 24 M. S. Buchanan T. Hashimoto and Y. Asakawa Phytochemistry 1996 41 821. 25 M. S. Buchanan T. Hashimoto and Y. Asakawa Phytochemistry 1995 40 135. BOC BOC BOC 26 E. Dagne A. A. L. Gunatilaka S. Asmellash D. Abate D. G. I. Kingston G. A. Hofmann and R. K. Johnson Tetrahedron 1994 AcO 50 5615. \ 27 R. L. Edwards D. J. Maitland and A. J. S. Whalley J. Chem. SOC. Perkin Trans. I 1991 1411. 28 R. L. Edwards D. J. Maitland and A. J. S. Whalley J. Chem. SOC. Perkin Trans. I 1989 57. 29 J.B. Hudson J. Zhou L. Harris L. Yip and G. H. N. Towers Photochem. Photobiol. 1994 60,253. 30 H. I. Pass J. Natl. Cancer Inst. 1993 85 443. Reagents i 138 (R = H) THF -78 -+ -50 "C; ii NaBH, MeOH 31 Z. Diwu Photochem. Photobiol. 1995 61 529. 0 "C; iii Ac,O pyridine DMAP CH,Cl, 25 "C; iv PPTS DME 32 A. Mathey W. Van Roy L. Van Vseck G. Eckhardt and glycol 70 "C; v H, Lindlar catalyst poisoned with quinoline 25 "C; W. Steglich Rapid Commun. Mass Spectrom. 1994 8 46. vi Dess-Martin periodinane CH,Cl, 25 "C; vii TFA CH,Cl, 33 P. M. Morgan PhD Thesis University of Melbourne in -10 "C,4 h then DMAP or DABCO 1 YO preparation. 34 W. A. Ayer and L. S. Trifonov J. Nat. Prod. 1994 57, Scheme 11 317. 35 M. Gill and A. Gimenez J. Chem. SOC. Perkin Trans. 1 1995 645.Stimulated by the isolation and characterisation of the 36 A. S. Cotterill and M. Gill Aust. J. Chem. 1994 47 1363. natural products 134 and 135 the first synthesis (Scheme 11) of 37 A. S. Cotterill M. Gill A. Gimenez and N. M. Milanovic the hitherto unknown parent heterocycle 2H-azepine 140 has J. Chem. SOC. Perkin Trans. I 1994 3269. been rep~rted.'~ 38 A. S. Cotterill R. W. Gable and M. Gill Acta Crystallogr. Sect. C Cr-vst. Struct. Commun. 1995 51 500. Acknowledgements. Gratitude is extended to Professor W. 39 N. M. Milanovic PhD Thesis University of Melbourne in Steglich Miinchen for information prior to publication; to the preparation. 40 S. Saubern PhD Thesis University of Melbourne 1993. Department of Chemistry University of Canterbury Christ- church New Zealand for providing facilities for writing this 41 E.Dagne I. Casser and W. Steglich Phytochemistry 1992 31 1791. Report; and to Peter M. Morgan University of Melbourne 42 C. Elsworth and M. Gill unpublished work. for some help in information retrieval. 43 A. Antonowitz M. Gill P. M. Morgan and J. Yu Phytochemistry 1994 37 1679. 44 For related compounds from the leaf mould Alternaria solani see 6 References G. Lazarovits W. R. Steele and A. Stoessl 2. Naturforsch. C 1 M. Gill Nat. Prod. Rep. 1994 11 67. Biosci. 1988 43 813. 2 M. Gill and W. Steglich Prog. Chem. Org. Nat. Prod. 1987 45 M. Gill A. Gimenez A. G. Jhingran N. M. Milanovic and A. R. 51 1. Palfreyman manuscript in preparation. 3 A. Takahashi R. Kudo G. Kusano and S. Nozoe Chem. Pharm.46 P. M. Millar BSc Honours Thesis University of Melbourne Bull. 1992 40 3194. 1994. 4 M. S. Buchanan T. Hashimoto S. Takaoka and Y. Asakawa 47 M. Gill and S. Phonh-Axa unpublished work. Phytochemistry 1995 40 1251. 48 M. Gill and J. Yu Nat. Prod. Lett. 1994 5 21 1. 5 M. Holzapfel C. Kilpert and W. Steglich Liebigs Ann. Chem. 49 M. S. Buchanan M. Gill and J. Yu manuscripts in preparation. 1989 797. 50 A. S. Cotterill M. Gill and N. M. Milanovic J. Chem. SOC. 6 H. Besl A. Bresinsky G. Geigenmuller R. Herrmann C. Kilpert Perkin Trans. 1 1995 1215. and W. Steglich Liebigs Ann. Chem. 1989 803. 51 M. Gill Aust. J. Chem. 1995 48 1. 7 M. Gill and R. Watling Plant Syst. Evol. 1986 154 225. 52 M. Gill Beih. Sydowia 1995 10 73. 8 N. Arnold W. Steglich and H. Besl 2.Mykol.1996 62 69. 53 Y. Ye I. Josten N. Arnold B. Steffan and W. Steglich 9 R. Marumoto C. Kilpert and W. Steglich 2. Naturforsch. C Tetrahedron 1996 52 5793. Biosci. 1986 41 363. 54 R. Velten I. Josten and W. Steglich Liebigs Ann. 1995 81. 10 F. Kogl H. Becker G de Voss and E. Wirth Justus Liebigs Ann. 55 L. M. Abrell X-C. Cheng and P. Crews Tetrahedron Lett. 1994 Chem. 1928 465 243. 35 9159. 11 W. Steglich W. Furtner and A. Prox 2. Naturforsch. B Anorg. 56 R. Marumoto D. Klostermayer and W. Steglich unpublished Chem. Org. Chem. Biochem. Biophys. Biol. 1968 23 1044. work. 12 M. Gill and M. J. Kiefel Aust. J. Chem. 1994 47 1967. 57 R. Frode M. Brockelmann B. Steffan W. Steglich and 13 W. Steglich B. Steffan T. Eizenhofer B. Fugmann R. Herrmann R. Marumoto Tetrahedron 1995 51 2553.and J.-D. Klamann in Bioactive Compounds from Plants D. J. 58 S. Takekuma H. Takekuma Y. Matsubara K. Inaba and Chadwick and J. Marsh Ciba Foundation Symposium 154 Wiley Z. Yoshida J. Am. Chem. SOC. 1994 116 8849. Chichester 1990 pp. 56-65. 59 C. Baumann M. Brockelmann B. Fugmann B. Steffan 14 W. Steglich Ernst Schering Research Foundation Lecture Series W. Steglich and W. Sheldrick Angew. Chem. Int. Ed. Engl. 1993 1995 24. 6. 32 1087. 15 K. Justus and W. Steglich Tetrahedron Lett. 1991 32 5781. 60 C. Hopmann and W. Steglich manuscript in preparation. 16 N. A. A. Ali R. Jansen H. Pilgrim K. Liberra and U. Lindequist 61 R. Frode C. Hinze I. Josten B. Schmidt B. Steffan and Phytochemistry 1996 41 927. W. Steglich Tetrahedron Lett. 1994 35 1689.17 W. Adam C. R. Saha-Moller M. Veit and B. Welke Synthesis 62 T. Hashimoto A. Yasuda K. Akazawa S. Takaoka M. Tori and 1994 1133. Y. Asakawa Tetrahedron Lett. 1994 35 2559. 18 S. Zapf A. Werle T. Anke D. Klostermeyer B. Steffan and 63 T. Hashimoto K. Akazawa M. Tori Y. Kan T. Kusumi W. Steglich Angew. Chem. Int. Ed. Engl. 1995 34 196. H. Takahashi and Y. Asakawa Chem. Pharm. Bull. 1994 42 19 J. M. Clough Nat. Prod. Rep. 1993 10 565 and references cited 1531. therein. 64 M. S. Buchanan T. Hashimoto and Y. Asakawa Phytochemistry, 20 J. M. Blunt M. H. G. Munro and G. M. Nicholas personal 1996 41 791. communication. 65 A. Terpin K. Polborn and W. Steglich Tetrahedron 1995 51, 21 R. F. C. Brown and A. J. Robinson Ausr. J. Chem. 1995,48,515. 9941. 22 T.Hashimoto S. Tahara S. Takaoka M. Tori and Y. Asakawa 66 M. M. Faul K. A. Sullivan and L. L. Winneroski Synthesis, Chem. Pharm. Bull. 1994. 42 1528. 1995 1511. 67 W. Harris C. H. Hill E. Keech and P. Malsher Tetrahedron Lett. 1993 34 8361. 68 P. D. Davis C. H. Hill G. Lawton J. S. Nixon S. E. Wilkinson S. A. Hurst E. Keech and S. E. Turner J. Med. Chem. 1992,35 177. 69 A. P. Fonseca A. M. Lob0 and S. Prabhakar Tetrahedron Lett. 1995 36 2689. NATURAL PRODUCT REPORTS 1996 70 J. F. Barry T. W. Wallace and N. D. A. Walshe Tetrahedron 1995 51 12797. 71 0. Sterner B. Steffan and W. Steglich Tetrahedron 1987 43 1075. 72 D. Hamprecht K. Polborn and W. Steglich Angew. Chem. Int. Ed. Engl. 1995 34 1469.

ISSN:0265-0568    

DOI:10.1039/NP9961300513

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

The Sesterterpenoids James R. Hanson School of Molecular Sciences University of Sussex Brighton Sussex BN I 9QJ UK Reviewing the literature published between November 1991 and March 1996 (Continuing the coverage of literature in Natural Product Reports 1992 vol. 9 p. 481 ) 1 Introduction 2 Linear Sesterterpenoids 3 Monocarbocyclic Sesterterpenoids 4 Bicarbocyclic Sesterterpenoids 5 Tricarbocyclic Sesterterpenoids 6 Tetracarbocyclic Sesterterpenoids 7 Fungal and Miscellaneous Sesterterpenoids 8 References I Introduction This report follows the pattern of its predecessors1 and is devoted to the occurrence of the sesterterpenoids. The sesterterpenoids have been reviewed2 and listed in the Dictionary of Terpen~ids.~ Many sesterterpenoids are of marine origin and they have been described in reviews devoted to marine natural product^.^ Although the synthesis of the sesterterpenoids is outside the scope of this review it is worth noting that a number of syntheses have been reported including several of man~alide,~ luffariolide,6 ceroplasto17 and other ophiobolanes.* Sesterterpenoids are biomarkers amongst sedimentary iso- prenoids and the structure determination of theseg and their analysis'O has been reported.Not all C, compounds are sesterterpenoids and further biosynthetic studies have been reported1' on mero terpenoid mixed polyketide- terpenoid fungal metabolites such as the citreohybridones which are metabolites of Peniciffium citreoviride and arise from a C15 and a C, unit.2 Linear Sesterterpenoids Marine organisms particularly sponges have continued to provide the source of linear sesterterpenoids. The terminal units often comprise either a furan a y-lactone or a tetronic acid moiety. These sesterterpenoids co-occur with degradation products and in some instances with diterpenoid chains. One of the linear sesterterpenoids 1 obtainedlZ from the Caribbean sponge Thorecta horridus possessed a marked inflammatory activity inducing the release of histamine and causing oedema in the paw of test animals. This compound has also been isolatedz5 from Luflarieffa geometrica and named luffarin Q. Thorectolide mono-acetate 2 obtained l3 from a New Cale- donian marine sponge identified as a Hyrtios species in contrast possessed anti-inflammatory properties.A tetronic acid 3 with antimicrobial activity has been isolated14 from an Australian sponge Psammoczniu species whilst a relative isopalinurin 4 which is a mild protein phosphatase inhibitor was obtained15 from a South Australian Dysidea species. The epoxyfuran carboxylic acid 5 was isolated16 together with some linear diterpenoids from a Western Australian Spongia species. The cometins A-C (e.g. 6)contain two furan rings and are further products obtained17 from Australian sponges. Variabilin 7 is a widespread member of this series and its absolute stereochemistry and that of some related tetronic acids have been established.18-19 (18R)-Variabilin 8 and its 11-methyloctadecanoate ester 9 have been z1 from the H0& 1 2 R=OCOMe ?H 3 U ?H 4 U 0 5 6 QH 7 OR 8 R=H '0 9 R =C(CH*)&H(CH&Me I 8 Me /\ 0 \ * 0 I 0 10 sponge Ircinia fefix.Degraded linear sesterterpenes have been isolated from sponges of the genera Spongia Carteriospongia and Hippospongiu. Untenospongin C 10 is an example which has been obtainedz2 from an Okinawan Hippospongia sp. whilst 529 530 0 11 OH 12 the difuran isonitenin 11was isolated23 from Spongiu oficinulis. The degradation of tetronic acids with alkaline hydrogen peroxide to mimic the formation of these compounds has been rep~rted.,~ A series of bislactones exemplified by luffarin R 12 have been isolated25 from Luflariella geometricu.3 Monocarbocyclic Sesterterpenoids Igernellin 13 has been obtained26 from a Palauan sponge Igernellu sp. A number of antibacterial sesterterpenoids such as 14 have been found2' in an Australian Lufuriellu species from the Great Barrier Reef. The absolute configuration at C-4 of manoalide 15 seco-manoalide 16 and neomanoalide 17 has been established.28 These were isolated from the sponge Hyrtios erectu. Manoalide 15 and its relative luffariellolide 18 are powerful anti-13 OH 14 R = OH OAC 0 16 OH 17 NATURAL PRODUCT REPORTS 1996 inflammatory agents and inhibitors of phospholipase A,. This activity has been associated with the ability of manoalide to form an adduct with lysine residues. Schiff base formation has been dem~nstrated~~ between bee venom PLA and manoalide.The synthesis of radiolabelled manoalide analogues has been reported.30 Marine sponges of the genus Lufluriellu are a rich source of manoalide-related sesterterpenoids. An Okinawan sponge of this genus has ~ielded~'.~~ a series of compounds of this type the luffariolides A-G exemplified by 19 and 20. The fascio- spongides A 21 B 22 and C 23 are manoalide relatives which have been from a sponge Fasciospongia sp. The six- membered ring has been cleaved in the formation of 22 and 23. There is considerable variation in the secondary metabolite content of some sponges. (4E,6E)-Dehydromanoalide 24 has been along with the bicyclic luffalactone 30 from a large scale collection of the sponge Luflariella variubilis.A number of the secondary metabolites of sponges play a protective role. The absolute stereochemistry has been es-for tabli~hed~~ cyclolinteinone 25 which has ichthytoxic properties and has been isolated from Cucospongiu linteiformis. HO. 20 U 22 23 OH 4 OHC 0 24 NATURAL PRODUCT REPORTS 1996-5. R. HANSON 531 pco2H 26 '*. H I 27 """& The Australian marine sponge Luflariella geometrica has to be a rich source of acyclic and cyclic sesterterpenes. Luffarellin P 26 is an example of a monocarbocyclic sester- 36 terpenoid obtained from this sponge. A novel furano-sesterterpene untenic acid 27 which activates the sarcoplasmic OH reticulum Ca'+-ATPase involved in muscle action has been found in an extract of an Okinawan marine sponge.36 4 Bicarbocycl ic Sesterterpenoids The studies on the Australian sponge Luflariella geometrica referred to previo~sly,~~ have revealed the presence of a large number of related bicarbocyclic sesterterpenoids the luffarins A-N exemplified by luffarin A 28 and luffarin I 29.These The number of known norsesterterpene cyclic peroxides has possess the same bicarbocyclic unit and differ in the nature of continued to increase. Two further examples which have been the side chain. Luffalactone 30 from Luflariella variabilis is a isolated40 from a Thai sponge of the genus Mycale are similar metabolite. 34 mycaperoxide A 34and B 35,both of which exhibit significant OH cytotoxic and antiviral activity. Trunculin F 36 is a further I HOJP0degradation product which has been iso1ated4l from an OH lo Australian sponge Latrunculia convulosa.A sponge from the Great Barrier Reef a Coscinoderma sp. has been to produce a cytotoxic and antibacterial sesterterpenoid quinol coscinoquinol 37. Not all sesterterpenoids are of marine origin and a number of Salvia (Labiatae) species have been shown to contain them. '*., H Three recent examples are salvileucolide 38 from S. sahendica a plant found in Northern and the norsesterterpenes 29 yosgadensonol 39 and its 13-epimer 40 which were isolated 28 from a Turlush species S. yosgadensis. Diterpenoids are A number of bicarbocyclic sesterterpenoids have a carbon widespread amongst Salvia species. skeleton reminiscent of the clerodane diterpenoids.Caco-spongionolide B 31 and its 25-deoxyderivative 32 have been isolated3'.38 from the Adriatic sponge Fasciospongia cavernosa. They possess ichthytoxic activity whilst cacospongionolide B also shows cytotoxic activity. Palaulol 33 is an anti-inflam- matory sesterterpene which was from a Palauan *; Fascaplysinopsis species of sponge. 0 '*.,H lo 30 31 R=OH 32 R=H 0 39 13P-Me 41 40 1%-Me 5 Tricarbocyclic Sesterterpenoids A number of cytotoxic a-hydroxybutenolides the spongiano- lides A-F have been from a Spongia species collected off the coast of Florida. They are exemplified by spongianolide A 41 and C 42. These metabolites inhibited protein kinase C. The spongianolides C and D are identical to 33 the lintenolides A and B which were isolated46 from the 532 NATURAL PRODUCT REPORTS 1996 r! Po I I 'a, H 52 42 R=Ac 1 43 R = OCOCH2CH(OH)Me 0 OH 54 a-R 44 45 Caribbean sponge Cacospongia linteiformis.The lintenolides C-E e.g. 43 and 44 exist as 16-epimers and act as anti-feedants against predatory fish.47 The vulgarosides e.g. 45 are the ring opened form of these compounds which have been from an Italian plant Cydonia vulgaris (Rosaceae) which is used in folk medicine to treat skin disorders. The conulosins A 46 and B 47 have been i~olated~~.~~ from the Australian sponge Latrunculia convulosa. Apart from chlorinated homo-diterpenes the South African nudibranch Chromodoris hamiltoni has been shown50 to contain as a minor metabolite a sesterterpene hamiltonin E 48.Some related ketals have been isolated from a sponge originally assigneds1 to the genus Dactylospongia but now known52 to be Petrosaspongia nigra. The structure of 49 was established by X-ray crys- tallograp hy . Hyrtiosal 51 was from the Okinawan marine sponge Hyrtios erecta. The metabolite which possesses a rearranged carbon skeleton was shown to inhibit the pro- liferation of KB cells. A plausible scheme for the origin of the hyrtiosane skeleton involves the rearrangement of an epoxide with the extrusion of an aldehyde 50+51. A Caribbean collection of the sponge Cucospongia cf. linteiformis has afforded some novel mono- and tri-carbocyclic metabolites for which a plausible unifying biogenetic scheme has been written.The structure of lintenone 5354 and its stereoisomer 5455may be derived by cyclization and rearrange- 0 &H '. H 46 47 55 ment of an epoxide 52. The monocyclic compounds including cyclolinteinone 5556may arise by simple rearrangment of the epoxide 52 although the recent isolation55 of the 3-deoxy analogue of 55 makes this part of the biogenetic scheme less attractive. 6 Tet raca r boc yc Iic Sest ert er pen0 ids The suberitenones A 56 and B 57 possess a novel carbon skeleton. They were obtained5' from an Antarctic sponge of the genus Suberites. Suberitenone B inhibited the cholesteryl ester transfer protein. The majority of tetracarbocyclic sesterterpenoids belong to the scalarane series.The superacid cyclization of bicyclo- geranylfarnesic acid and of geranylfarnesic acid has been to lead to compounds of the scalarane type e.g. 58. 12-0-Desacetylfuroscalarol59 and 12-0-desacetylscalarin 60 have been from an Okinawan sponge of the genus 0 0 fi0 '8.. H 40 49 56 57 f;;> OH mcty-, I I I\-0 '*.,H 0 '. A 50 51 59 NATURAL PRODUCT REPORTS 1996-5. R. HANSON 533 R2 MeCH&HCH&&OH -.- W '*. =A 72 O H 60 R' = a-OH; R2 = OH 61 R=@OH 73 66 R' = b-OH; R2 = H 62 R=a-OH XCHOAcO. AcO CHO6.XOMe 63 R=H 65 64 R=OAc Hyrtios. These compounds enhance nerve growth factor synthesis and might provide leads for compounds for the treatment of disorders of the central nervous system such as Alzheimer's disease.'O Isocalarafuran A 61 and B 62 have been isolated from a South Australian sponge Spongia hispida whilst 12-desacetoxyscalaradial 63 was obtaineds2 from Caco-spongia mollior.Details of the X-ray crystal structure of scalaradial 64 have been reported.63 The predator-prey re-lationship between the nudibranch Hypselodoris orsini and the sponge Cacospongia mollior has been examineds4 in terms of the biotransformation of the dietary sesterterpenoid by the mollusc. 12-Epiheteronemin 65 has been founds5 in a New Caledonian collection of Hyrtios erecta. It was the major metabolite in this collection whereas heteronemin predominated in collections from both Australia and the Red Sea. The scalarin derivative 66 has been founds6 in an Indian collection of Heteronemu (= Hyrtios) erecta.A number of mono- and di-alkylated scalaranes are known. The formation of these is interesting both from a biosynthetic point of view and as chemotaxonomic markers within the sponges of the order Dictyoceratida and their associated nudibranch molluscs. A collection of Lendenfeldia frondosa from the Solomon Islands afforded6' the homoscalaranes 67-70. The acetate of 67 possessed anti-inflammatory activity. Another homoscalarane of this type 71 was obtaineds8 from Phyllospongiu dendyi. 67 R=a-OH 69 68 R = D-OH MeO. OH 70 71 OCCH2CHMe 6 AH 74 75 The sponge Phyllospongia (syn. Carteriospongiu) foliascens has proved to be a rich source of bishomoscalaranes. A family of lactones the phyllactones A-G exemplified by 7269and 73,'O have been isolated from a collection from the Nansha Islands in the South China Sea.The sesterterpene metabolites of P. foliascens collected in two localities differed in the relative stereochemistry of the substituents at C-12 and C-16. The lactone 74 was obtained'l from an Indian collection of P. foliascens. A series of cytotoxic bishomoscalaranes have been ~btained'~ from a Dictyoceratid sponge Strepsichordaia lenden- feldi including 75 and 76. The inorolides e.g. 77 and 78 have been dete~ted'~ as metabolites of the nudibranch Chromodoris inornata and may arise from dietary sesterterpenoids. 7 Fungal and Miscellaneous Sesterterpenoids The stereochemistry and biosynthesis of terpestacin 79 a fungal metabolite from an Arthrinium species has been e~tablished.'~ The absolute configuration at the ring junction was found to be opposite to that of the known retigeranic acid and variculanol at their corresponding centres.The compound inhibits syncytium formation an event involved in the interaction between the human immunodeficiency virus and human T-cells. 75 The structure of fusaproliferin 80 originally YAc 8 & '8.. I H -. 76 77 vc '-.,H 78 15 79 80 534 NATURAL PRODUCT REPORTS 1996 The biotransformation of ophiobolin A 83has been st~died.’~ Metabolites such as 84 and 85 were obtained using Polyangium cellulosum which showed significant biological activity against Trichophy t on men tagroph y tes.The structure 86 has been proposeds2 for raoulic acid which I v is the major bioactive anti-leukaemic constituent of the New Zealand plant Raoulia australis (Asteraceae). Whilst it is difficult to propose a biogenesis of this compound from geranylfarnesyl pyrophosphate the carbon skeleton may be generated by diprenylation of a sesquiterpene and hence raoulic acid may not be a genuine sesterterpenoid. 8 References 1 2 3 4 J. R. Hanson Natural Product Reports 1992 9 48 1. J. R. Hanson in Rodd’s Chemistry of Carbon Compounds ed. M. Sainsbury 2nd edn 2nd supplement vol. 2B Elsevier Amster- dam 1994 p. 441. Dictionary of Terpenoids ed. J. D. Connolly and R. A. Hill Chapman and Hall London 1991 vol. 2 p. 1097. D. J. Faulkner Natural Product Reports 1996,13,75and previous Scheme 1 5 reports.C. W. Jefford J.-C. Rossier J. Boukouvalas and P. Huang Helv. Chim. Acta 1994,77 661 ;P. Bury G. Hareau P. Kocienski and D. Dhanak Tetrahedron 1994,50,8793; E. Lattmann J. Coombs kntalthe retigeran own as plography. roliferin has been e~tablished’~ by This metabolite is related to terpesic acid stereochemistry at C-14 an X-ray crys- tacin but has d C-15. The 6 7 and H. M. R. Hoffman Synthesis 1996 171. G. Hareau and P. Kocienski Synlett. 1995 893. L. A. Paquette T. Z. Wang and Vo Nha Huu J.Am. Chem. SOC. 1993 115 1676; L.A. Paquette and M.A. Poupart J. Org. Chem. 1993 58 4245. compound eratum whlif was isolated” froich is a serious pa m the fungus Fthogen on maize. usarium pro- 8 J. H. Rigby T. McGuire C. Senanayake and K. Khemani J. Chem.Soc. Perkin Trans. I 1994 3449. The struct ure of the fungal m etabolite astellat 0181 which is 9 S. T. Belt D. A. Cooke S. J. Hird and S. Rowland J. Chem. Soc. pr~duced’~ by Aspergillus var iecolor (syn. A. s tellatus) was Chem. Commun. 1994 2077. the catablished bpyrophosphaown in Scessh rbon skeleton is y NMR studies. Bte by a series of cheme 1. The initial formed from geiosynthetic studies yclizations and step in this cycliza ranylfarnesyl have shown” rearrangements tion may lead 10 11 A. Alajbeg Nafta (Zagreb) 1994 45 383 (Chem. Abstr. 1995 122 243597). S. Kosemura H. Miyata K. Matsunaga and S. Yamamura Tetrahedron Lett. 1992 33 3883. implicated in either teraleurodiscal. id have shmponent to acco the formation pestacin or fusaprFurther studies on ownso that the origA with the S conf of retigeranic oliferin whilst latthe lichen product inal samples contaiguration at C-18 acid 82 and retigeranic ined a minor and a major er stages are 12 13 14 E.Fattorusso V. Lanzotti S. Magno L. Mayol M. DiRosa and A. Ialenti Bioorg. Med. Chem. Lett. 1991 1 639. M.-L. Bourguet-Kondracki C. Debitus and M. Guyot .I. Chem. Res. (S) 1996 192. L. Murray H. Hamit J. N. A. Hooper L. Hobbs and R. J. Capon Aust. J. Chem. 1995 48 1899. this centre. mponent,co retigeranic acid B 82 with the R-co nfiguration at 15 L. Murray A. T. R. Sim J. A. P. Rostas and R. J. Capon Aust. J. Chem. 1993 46 1291. 16 P. A. Searle and T. F. Molinski Tetrahedron 1994 50 9893. 17 18 S. Urban and R. J. Capon Ausr. J. Chem. 1992 45 1255. M. Ishibashi M. Kurosaki Y. Mikami and J. Kobayashi Nut.Prod. Lett. 1993 3 189. 19 R. J. Capon T. R. Dargaville and R. Davis Nut. Prod. Lett. 1994 4 5 1. 81 82 20 A. Martinez C. Duque N. Sato R. Tanaka and Y. Fujimoto Nut. Prod. Lett. 1995 6 1. 21 A. Martinez C. Dunque N. Hara and Y. Fujimoto Nut. Prod. 22 Lett. 1995 6 281. J. Kobayashi H. Shinonega H. Shigemori and T. Sasaki Chem. Pharm. Bull. 1993 41 381. 23 L. A. Lenis L. Nunez C. Jimenez and R. Riguera Nut. Prod. Lett. 1996 8 15. 24 G. Guella I. Mancini I. N’Diaye and F. Pietra Tetrahedron Lett. 1991 32 6415. 25 26 M. S. Butler and R. J. Capon Aust. J. Chem. 1992 45 1705. G. Liu J. Pika and D. J. Faulkner Nut. Prod. Lett. 1995,7 297. Hi) 27 G. M. Koenig A. D. Wright and 0. Sticher J. Nut. Prod. 1992 83 @- 84 28 29 55 174. M. Kobayashi T. Okamoto K.Hayashi N. Yokoyama T. Sasaki and 1. Kitagawa Chem. Pharm. Bull. 1994 42 265. B. C. M. Potts D. J. Faulkner M. S. de Carvalho and R. S. 11 Jacobs J. Am. Chem. Soc. 1992 114 5093. OH 30 31 S. G. Senderoff A Y. L. Shu K. Lawrie and J. Heys J. Labelled Compd. Radiopharm. 1992 31 519. M. Tsuda H. Shigemori M. Ishibashi T. Sasaki and J. Kobayashi J. Org. Chem. 1992 57 3503. H I HO 85 86 32 33 J. Kobayashi C.-M. Zeng M. Ishibashi and T. Sasaki J. Nut. Prod. 1993 56 436. A. Montagnac M. Pals and C. Debitus J. Nut. Prod. 1994 57 186. that NATURAL PRODUCT REPORTS 1996J. R. HANSON 34 B. C. M. Potts R. J. Capon and D. J. Faulkner J. Org. Chem. 1992 57 2965. 35 M. R. Conte E. Fattorusso V. Lanzotti S. Magno and L. Mayol Tetrahedron 1994 50 13469. 36 N.Shoji A. Umeyama K. Kishi S. Arihara Y. Ohizumi and J. Kobayashi Aust. J. Chem. 1992 45 793. 37 S. De Rosa A. Crispino A. De Giulio C. Iodice R. Pronzato and N. Zavadnik J. Nat. Prod. 1995 58 1776. 38 S. De Rosa R. Puliti A. Crispino A. De Giulio C. De Sina C. Iodice and C. A. Matia Tetrahedron 1995 51 10731. 39 E. W. Schmidt and D. J. Faulkner Tetrahedron Lett. 1996 37 395 1 40 J. Tanaka T. Haga K. Suwanborirux U. Kokpol G. Bernard- inelli and C. W. Jefford J. Org. Chem. 1993 58 2999. 41 M. S. Butler and R. J. Capon Aust. J Chem. 1993 46 1363. 42 G. V. Alea A. R. Carroll and B. F. Bowden Aust. J. Chem. 1994 47 191. 43 F. M. Moghaddam B. Zaynizadeh and P. Ruedi Phytochemistry 1995 39 715. 44 G. Topcu A. Ulubelen T. C. M. Tam and C.-T.Che J. Nat. Prod. 1996 59 113. 45 H. He P. Kulanthaivel and B. J. Baker Tetrahedron Lett. 1994 35 7189. 46 M. R. Conte E. Fattorusso V. Lanzotti S. Magno and L. Mayol Tetrahedron 1994 50 849. 47 A. Carotenuto E. Fattorusso V. Lanzotti S. Magno and L. Mayol Liebigs Ann. Chem. 1996 77. 48 N. De Tommasi F. De Simone C. Pizza and N. Mahmood J. Nat. Prod. 1996 59 267. 49 M. S. Butler and R. J. Capon Nat. Prod. Lett. 1992 1 171. 50 J. Pika and D. J. Faulkner Tetrahedron 1995 51 8189. 51 A. R. Lal R. C. Cambie C. E. F. Rickard and P. R. Bergquist Tetrahedron Lett. 1994 35 2603. 52 R. C. Cambie A. R. La1 and C. E. F. Rickard Acta Crystallogr. Sect. C, 1996 52 709. 53 K. Iguchi Y. Shimada and Y. Yamada J. Org. Chem. 1992,57 522.54 E. Fattorusso V. Lanzotti S. Magno L. Mayol and M. Pansini J. Org. Chem. 1992 57 6921. 55 A. Carotenuto M. R. Conte E. Fattorusso V. Lanzotti and S. Magno Tetrahedron 1995 51 1075 1. 56 M. C. Conte E. Fattorusso V. Lanzotti S. Magno and L. Mayol Tetrahedron 1994 50 13 469. 57 J. Shin Y. Seo J.-R. Rho E. Baek B.-M. Kwon T.-S. Jeong and S.-H. Bok J. Org. Chem. 1995 60 7582. 58 P. F. Vlad N. D. Ungur and Nguyen Van Hung Mendeleev Commun. 1992 61. 59 Y. Doi H. Shigemori M. Ishibashi F. Mizobe A. Kawashima S. Nakaike and J. Kobayashi Chem. Pharm. Bull. 1993 41 2190. 60 J. Kobayashi and F. Mizobe Japan Patent 06239741 (Chem. Abstr. 1995 122 1097). 61 R. Davis and R. J. Capon Aust. J. Chem. 1993 46 1295. 62 S. De Rosa R.Puliti A. Crispino A. De Giulio C. A. Mattia and L. Mazzarella J. Nar. Proc. 1994 57 256. 63 R. Puliti C. A. Mattia and L. Mazzarella Acta Cr,.stallogr. Scct. C 1995 51 1703. 64 G. Cimino A. Fontana F. Gimenez A. Marin E. Mollo E. Trivellone and E. Zubia Experientia 1993 49 582. 65 M. L. Bourguet-Kondracki M. T. Martin C. Debitus and M. Guyot Tetrahedron Lett. 1994 35 109. 66 Y. Venkateswarlu M. A. Farooq Biabani and T. P. Rao Indian J. Chem. Sect. B. 1995 34 563. 67 K. A. Alvi and P. Crews J. Nat. Prod. 1992 55. 859. 68 B. C. Rao R. S. H. S. N. Kalidindi V. D. Rao and S. Sreedhara Indian J. Chem. Sect. B. 1993 32 288. 69 X. Fu L. Zeng J. Su M. Pais and P. Potier J. Nat. Prod. 1992 55 1607. 70 X. Fu L. Zeng J. Su M. Pais and P. Potier J.Nat. Prod. 1993 56 1985. 71 M. Reddy Y. Venkateswarlu and J. V. Rao Indian J. Chem. Sect. B 1993 32 1196. 72 B. F. Bowden J. C.Coll J. Li R. C. Cambie M. R. Kernan and P. R. Bergquist J. Nat. Prod. 1992 55 1234. 73 T. Miyamoto K. Sakamoto H. Amano R. Higuchi T. Komori and T. Sasaki Tetrahedron Lett. 1992 33 58 11. 74 M. Oka S. Iimura Y. Narita T. Furumai M. Konishi. T. Oki Q. Gao and H. Kakisawa J. Org. Chem. 1993 58 1875. 75 M. Oka M. Konishi S. Iimura Y. Narita. T. Furumai T. Oki Q. Gao and H. Kakisawa Tennen Yuki Kagobutsu Toronkai Koen Yoshishu 1992 34 558; (Chem. Abstr. 1994 120 158 264). 76 A. Santini A. Ritieni V. Fogliano G. Randazzo L. Mannina A. Logrieco and E. Benedetti J. Nut. Prod. 1996 59. 109. 77 G. Randazzo V.Fogliano A. Ritieni L. Mannina E. Rossi A. Scarallo and A. L. Segre Tetrahedron 1993 49 10883. 78 I. H. Sadler and T. J. Simpson Magn. Reson. Chem. 1992 30 518. 79 T. J. Simpson J. Chem. Soc. Perkin Tram. 1 1994. 3055. 80 H. Sugawara A. Kasuya Y. Iitaka and S. Shibata Chem. Pharm. Bull. 1991 39 3051. 81 E. Li A. M. Clark D. P. Rotella and C. D. Hufford J. Nal. Prod. 1995 58 74. 82 S. J. Bloor Tetrahedron Lett. 1993 34 5617.

ISSN:0265-0568    

DOI:10.1039/NP9961300529

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Hot off the Press Robert A. Hill' and Andrew R. Pitt' Department of Chemistry Glasgow University Glasgow UK G 12 800. E-mail bobh! chem.gla.ac.uk Department of Pure and Applied Chemistry Strathclyde University Thomas Graham Building 295 Cathedral Street Glasgow UK G 1 IXL. E-mail a.r.pitt(u strath.ac.uk Reviewing the recent literature on natural products and bioorganic chemistry .4C2:< terpenoid 1 with a new hassane skeleton has been isolated from Stil\licr Crpiurici (J. G. Luis ct cd. Tetraheclron 1996 52. 12309).The hassane 1 is related to the apianes recently isolated from the same species [see 'Hot off the Press' in Nat. Prod. Kcp.. 1996 /3(5) iii] and the authors propose a common biogenctic origin from an abietane. Neovibsanine A 2 from I ?hurrii~ni ciit'cihuki is an unusual cleaved diterpenoid (Y.Fukuyama et ul. Tetrahrciron Lctt. 1996 37 6767). The co- occurring vibsanine B 3 has been converted photochemically into the neovibsanine skeleton. Caryose 4 is the first example of a carbocycI ic m on osaccha ride from Pseudonzonas carpphj./li (M. Adinolfi. Ccirhoiijdr. Res.. 1996 284 11 1). A new class of carbon skeleton 5 has been identified from the poison gland of the African Myrmicaria ant (F. Scroder ct a/.,Chem. Commurz. 1996. IH,2 139):consisting of two unbranched C15 chains it has been called myrmicarin 430A. Dermatolactone 6 from the Basidiomycete 1rpc.y 1rrctezr.v has a new sesquiterpenoid carbon skeleton that is probably related to the illudane group (.4. Mayer et (I/. Ph?.tot.l?c.nti.vtr?..1996 43 375). An unusual spiroether. cleroindicin A 7 has been isolated from C/ero-1 I I H2FiH HO 'OH 4 elendrum iiitlicwm (J. Tian et til, Chiri. Chcrii. Lctt.. 1996. 7. 279). Nostocine A 8 is a violet pigment produccd as an extracellular metabolite from the freshwater cyanobacterium IVostoc spotr-giaqfhrnw (K. Hirata ct [I/.. Hc.tr.roc:~.e./es. 1996. 43. 1 513). The structure of this nitrogen-rich heterocycle was established by X-ray analysis. Nostocine A 8 shows growth inhibitory activity against a range of organisms and it is suggested that nostocine A 8 plays an allelopathic role. Fischerellin A 9. from the cyanobacterium Fischerc.//ri inuscicdcr also shows interesting allelopathic activity as a fungicide and as a photosystem 11 inhibitor (L.Hagmann and F. Juttner. Tctrtilrctlrori Lctt.. 1996. 37. 6539). S. Sepulveda-Boza and B. K. Cassels have reviewed the plant metabolites that are active against trypanosomiasis due to Trj.pcrno.voriici cnci (Plaritcr Md.. 1996. 62. 98). A. I>. Wright and co-workers have reviewed their search for marine- derived natural products with selective antimalarial activity. (J. Nut. Prod. 1996 5Y 710). There have been a number of syntheses reported recently that have resulted in the review of the structure of the targct natural product. The structure of bruceoside C. a quassinoid from BI-ucecr,jci\miccr. has been shown to be incorrect. P. A. Grieco and co-worker synthesized the alleged aglycone of bruceosidc C and found that it was not identical to the aglycone derived from the natural product (J.Org. Chcriz..1996. fi I 53 16). The correct structure for bruceoside C has not yet been established. The structure assigned to a clerodadienoic acid from Epriui purpurcri has also been shown to be incorrect. T.-H. Lee and C.-C. Liao have synthesized the clerodadienoic acid 10 and 8 9 COOH 6 7 10 ... 111 NATURAL PRODUC‘T REPOKTS. I1Nh The biogenesis of sapidolide A 14 from Brrcarirrctr strpicltr is intriguing (N. C. Barua and co-workers. Tctrcrlic~tlrotiLc’tt.. 1996,37 6791). The structure of sapidolide A 14 was confirmed by X-ray analysis. H. G. Floss and co-workers have studied the HO‘& biosynthesis of 3-amino-5-hydroxybenzoic acid 15 the prc-11 cursor of the mC,N units in the ansamycin antibiotics (J..4m. & 0 HO@ 12 13 confirmed its structure by X-ray analysis but the spectral data did not agree with those of the natural product (Tetruhedroti Lctt. 1996. 37 6869). T. Koike rt LII. claim to have synthesised dictamnol with the cis-fused ring junction 11 (Cheni. Ptinrni. Birll 1996 44 646). However A. de Groot and co-workers have also synthesized cis-dictamnol 11 and report that it is not identical to natural dictamnol (Cheni. Pliarni. Bull.. 1996 44 1400).De Groot’s group point out that the synthetic route used by Koike‘s group involved the ketone intermediate 12 which will readily undergo epimerization to the truns-fused junction and that the structure of natural dictamnol should be revised to trans-dictamnol 13.14 C/ieni. Soc. 1996 118 7486). Using cell-free extracts from the rifamycin B producer ,4nij.c~o/irrop.~i.s and tiic~~litf~rt.~riif~ithe ot~ij~~~s ansatrieni n A producer Strc~pt collitius. t hey es t;i b1ished that 3-amino-5-hydroxybenzoic acid 15 is produced by a new variant of the shikimate pathway (Scheme 1 ) in\,olving 4-amino- 3.4-dideoxy-~-crrnb ino-he p u1osonic acid 7-phos pI1a t e ( ami ti o DAHP) 16 rather than the normal shikiniatc piithwa) intermediate 3-deoxy-D-crrcihitin-hepulosoiiic acid 7-phosphate (DAHP). Deuterium kinetic isotope effects have been used to dem- onstrate that two different (S)-adcnosyl-L-iiicthioninc A”-sterol methyl transferase enzymes are involved in the bio- synthesis of 24(28)-rnethylenecyclo~~rtaiiol and cyclosadol (24-methylcycloart-23-en-3~-ol) (W.D. Nes and co-workers. Tcft.+ hedroii Lett. 1996 37 6823). Sarcoglane 17 from Strrcophj~toti glnucuni probably arises by cyclisation of a cembranoid (Y. Kashman and co-workers. 7i’truhc~lrori Lett.. 1996 37 6909). Sinulariadiolide 18 from a Sinulurici species also appears to be derived from a cembranoid precursor (Y. Yamada and co-workers J. Org. Cheni. 1995 60 5998). Labelling studies have established that the hydroxy group at C-2’ of bicycloinycin 19 from Strcptotiijws supporoiic.iisi.s is introduced with inversion of configuration (E.L. Bradley ct ul.. T~~rrtrhc~tlroti Lett.. 1996. 37 6935). The opening of an epoxide intermediate is proposed to explain this inversion.Further evidence for the compart- mentalisation of biosynthetic pathways has been presented by C. E. Domenech c’t (21. (Eirr.J. Bioclrcwi. 1996 23Y. 720). They have shown different incorporation of labelled acetate meva- lonate and leucine into sterols carotenoids and gibberellins in Gihhcrclltr ,f i! jikitroi. suggesting that the biosq,n t hesis is i n physi ca 11y sepa rated conipa r tni e ti t s wi t h di f rere11 t s u hst ra t e pools. The Battersby group have published the synthetic route to the put at ive s pi roc ycI ic py r ro1enon e i n tcrni ed i;i t c ;i I 1;i 1ogiic 20 17 18 HO C02HHx2H HN HO‘ 0b N H6H 2 HO’ 19 P A CO2H I OH 15 Scheme I 20 NATL'IIAI. PRODUCT REPORTS.1996-HOT OFF THE PRESS for the reaction of L'roporphyrinogen 111 synthase the enzyme that catalyses the formation of the cyclic tetrapyrrole nucleus for the tetrapyrrole pigments (J. Clieni. Soc. Perkin Tram. I 1996 2079). The subsequent paper describes how they were able to use a combination of X-ray crystallography on a model s>.stem and CD correlation to determine the absolute stereo- chemistry of the active enantiomer of the spirocycle (J. Chem. Soc.. Perkin Trim. I 1996. 2091). A whole issue of C'hcn?.Rev. edited by B. E. Smart has been given over to fluorine chemistry including extensive coverage of synthetic methodology (Cheni. Re\'.. 1996. 96 1555-1 757). A5 3 tribute to Vladmir Prelog on his 90th birthday issue 5 ( 1996) of tic/\*.Cizki.Actcr contains several dedicated articles including a discussion of the study of the origin of life by A. Eschenmoser and M. V. Kisakurek (Hdv.Cliim. Actn 1996 79. 1249) and a review of how methods of evolution biology inay be used to solve problems of synthetic chemistry by G. Quinkert and co-workers (Helv. Chiui. Actu 1996 79 1260). A unique mode of covalent catalysis by the B dependent 1-aii.linocyclopropane-I -carboxylate (ACC) deaminase has re-cently been uncovered (K. Li clt ul. J. Am. Clienz. Soc. 1996 118. 8763). The o.\-o-methylene analogue 21 is a time dependent inhibitor of the enzyme. with 22 being the only product isolated. consistent with nucleophilic attack by an enzyme based nuclcophile on the cyclopropane ring of the imine fi)rmcd between B and AAC.The deuteriated e.uo-methylene analogue 23 shows that nucleophilic attack must take place at C.3 (Scheme 2). 21 22 23 Scheme 2 The catalytic activities and the structure-function relation- ships of the enzymes of the cytochrome P450 group have been rekieued (B. A. Halkier PIi?,toc.h~.mi.str~.,1996 43 1). J. B. Jones and co-workers have published a review discussing the need to expand the knowledge of 'enzyme-substrate inter- actions that regulate and control enzyme specificity' as a first step to allowing the reliable selection of the enzyme best suited to ;I particular chiral transformation (Acru Cheni. Sctrnti. 1996 30. 697). This will allow accurate predictions of substrate suita bi I i t y and rational mu tagenesi s for tailoring specificity .Thc group 01' Mutter have reported the use of template aswnbled synthetic proteins (TASP) as functional protein mimics (Arige\t-. Clicwi.,tnt. Ed. Eng. 1996 35 1482). The idea ot' peptide loops grafted onto a semi-rigid template (in this case 3 qclic peptide) as a mimic of a protein active site or antibody combining site appears to be a realistic goal. Breslow and Still havc used a combinatorial approach to probe the specificity of binding of peptides to P-cyclodextrin (Aizgeit.. Cheni.. Int. Ed. Engl.. 1996,35 1490). A high selectivity for L-Phe-D-Pro and D-Phe-L-Pro was observed which was rationalized by NMR and molecular modelling studies. The generally accepted hydrogen bond accepting ability of fluorine has been called into question by O'Hagan's group (7etrnhedron 1996.52 I26 13). An evaluation of F-. .H contacts in the Cambridge Structural Database System showed very few C F. -H -X short distance contacts with C-F- -H-C van der Waals contacts being more common. Ah initio calculations suggested that C( sp")-F forms stronger H-bonds than C(sp')-F which is supported by the data in the structural database. A greater understanding of the nature of the noncovalent interactions between proteins and Ligands in solution and in the electrospray mass spectro- metric gas phase is emerging. C. V. Robinson c't trl. (J. .4t?i. Chew. Sor. 1996 IIH. 8646) have studied the protein-CoA ligand binding and assembly by both altering the length of the alkyl chain of the CoA and altering key residues on the protein.This shows that in the gas phase the requirements are diferent to the solution. with H-bond and n-stacking interactions being more iniport a n t than h yd r ophobic i nt e r ac t i on s . H yd r oge n ~ deuterium exchange experiments are also providing key evidence for the understanding of these interactions. The paper also includes some discussion of possi ble desolvation processes. The concept of reactive immunisation for the generation of' catalytic antibodies is discussed in a review by Learner and Barbas 111 (Actu Cjiem. Sccrntf. 1996. 50. 672.). Immunization with a chemically reactive species that will undergo a reaction with the antibody results in that reaction becoming part of the mechanism of the catalytic event.This was then used to generate an antibody to catalyse an aldol condensation ~iu an iminc formed with the antibody. They describe the procedure a> immunising with a 'reaction' rather than a 'molecule'. The subsequent paper discusses several issues related to the applicability of antibody-catalysed reactions to asymmetric synthesis (E. Keinan ct al. Actcr Cliern. Scarid. 1996. 50. 679). R 0 LU Scheme 3 An extensive review of the chemistry of the DNA alkylator CC1065 and an elegant elucidation of the mechanism of the sequence selectivity of the alkylation has been published by D. L. Boger and D. S. Johnson (Angcw,.Cheni. Itit. Ed. Eiigl. 1996 35 1438) clearly showing the relationship between structure functional reactivity and biological properties. The trigger for the alkylation appears to be a twist induced in the linking amide bond leading to a more activated cyclopropyl ring (Scheme 3). A review of the various strategies for targeting RNA with conformationally restricted antisense agents has recently been published (P. Herdewijn Leibigs .4nn.. 1996 1337). The generally accepted mode of action of actinomycin by binding to double stranded DNA has been called into question by R. M. Wadkins ct (11. (J. Mol. Biol.. 1996. 262. 53). They present compelling evidence both experimental and modelling of actinomycin binding to single stranded DNA. which may be an important mode of binding for the inhibition of RNApol.

ISSN:0265-0568    

DOI:10.1039/NP996130iiif

出版商:RSC    

年代:1996

数据来源: RSC