Recent Advances in Chemical Ecology J. B. Harborne Plant Science Laboratories University of Reading Reading RG6 2AS Reviewing the literature published between July 1985 and December 1987 (Continuing the coverage of literature in Natural Product Reports 1986 Vol. 3 p. 323) 1 Introduction compounds and several types of alkaloid now fall in this 2 Animal-Animal Interactions category. Uniquely cyanogenic glycosides can be both synthe- 2.1 Animal Toxins of Plant Origin sized de novo within the insect and sequestered from dietary 2.2 Chemical Defence in Animals sources. 2.2.1 Beetles and Termites Oviposition stimulants and deterrents in host plants have 2.2.2 Marine Organisms received more attention and a number of flavonoids and 2.3 Pheromones organic bases have been characterized in plant-Lepidopteran 3 Plant-Animal Interactions interactions.The lack of volatility in these particular signals is 3.1 Constitutive Chemical Defence somewhat surprising since it is not entirely clear how they are 3.1.1 Nitrogen Compounds perceived by the female insect. Feeding signals by contrast 3.1.2 Terpenoids usually include volatile components. This is particularly true of 3.1.3 Phenolics and Tannins nectar or pollen feeding and a considerable complexity of 3.1.4 Defence Systems chemical substances has recently been detected in the volatile 3.2 Induced Chemical Defence emanations of orchid blossoms. 3.3 Hormonal Interactions Plant-parasite interactions often involve chemical signals 3.4 Insect Antifeedants and considerable progress has been made in determining the 3.5 Oviposition Stimulants nature of these signals in legume-Rhizobiurn and in Agalinis- 3.6 Pollination Interactions Striga associations.The chemistry and biochemistry of in- 4 Plant-Plant Interactions fection of plants by microbes has been widely studied and 4.1 Allelopathy many new phytotoxins have been described during the period 4.2 Host-Parasite Interactions under review. Most notably the victorins which are the toxins 5 Plant-Microbe Interactions of blight disease (Cochliobolus victoriae) of oats have at last 5.1 Mycotoxins been fully identified. 5.2 Signal Molecules The third edition of a general introduction to ecological 5.3 Phytotoxins biochemistry has just been published2 and a German textbook 5.4 Antimicrobial Agents on the subject has also a~peared.~ A variety of books on 6 References plant-insect interactions are now available*-' as is the third volume in a series 'Handbook of Natural Toxins'.8 Other new books and review articles will be mentioned under specific topics.1 Introduction While there have been no major developments in chemical ecology since the last review,' work has continued unabated on 2 Ani ma I-An imaI Interact ions the identification of the chemical constituents that are involved in a multitude of plant and animal interactions that occur in the 2.1 Animal Toxins of Plant Origin ecosystem.For example an increasing number of classes of Some thirteen classes of plant toxin have now been recorded as plant substance have been identified as being sequestered and being sequestered by insects from food plants and then stored stored by insects for defensive purposes. Iridoids nitro-for defence (see Table 1 ;for earlier literature see refs. 9 and 10). Table 1 Classes of plant toxin that are sequestered by insects and stored for defence" Class of chemical Typical structure Plant source Insect storing it Bianthraquinones Cardiac glycosides Cucurbitacins Cyanogenic glycosides Glucosinolates Iridoids Meth ylazoxymethanols Nitrophenanthrenes Phenols Py razines Pyrrolidine alkaloids Pyrrolizidine alkaloids Quinolizidine alkaloids Hypericin Calotropin Cucurbitacin D Linamarin Sinigrin Aucubin Cycasin Aristolochic acid Salicin 3-isopropyl-DMDP Retronecine Cytisine 2-methox yp yrazine Hypericum hirsutum Asclepias spp.Cucurbita spp. Lotus cornicuiatus Brassica oleracea Plantago lanceolata Zamia JIoridana Aristolochia spp. Salix spp. Asciepias curassavica Omphalea spp. Senecio spp. Cytisus scoparius Beetle Chrysolina brunsvicensis Butterfly Danaus plexippus Beetle Diabrotica balteata Moth Zygaena trifoliib ButteAy Pieris brassicae ButteAy Euphydryas Cynthia Butterfly Eumaeus atala Butterfly Battus archidamas Beetle Chrysomela aenicoliis Butterfly Danaus piexippus Moth Urania fulgens Moth Arctia caja Aphid Aphis cytisorum For references to earlier literature see refs. 9 and 10; other references are in the text. This insect both sequesters linamarin from its food plant and synthesizes it de novo (see text).85 86 HOH~G / 1 RO (1) R = H HO (3) R = Glc (2) 0 (7) Most notably iridoid glucosides have recently been added to the list.''. l2Larvae and pupae of the alpine butterfly Euphydryas Cynthia contain the three iridoids aucubin (I) catalpol(2) and 6-O-glucosylaucubin (3) in concentrations of 0.53 0.31 and 1.48% of the dry weight." The first two compounds are commonly present in the larval food plants (such as Pfantago Zunceofuta),and are directly sequestered from the diet whereas the third compound is probably formed from (1) by conjugation within the insect. Preliminary feeding experiments with the insectivorous stonechat (Suxicofa torquata) indicate that the larvae of E.Cynthia are unpalatable due to their iridoid content and this correlates with the fact that catalpol (2) is one of the bitterest of all the iridoids.12 A more extensive study of ingestion of iridoids from food plants by four North American members of the Lepidoptera has indicated that the fate of the iridoid within the insect is related to predator-avoidance strategy. Thus the aposematic (i.e. having warning colours) butterfly Euphydryas phaeton is unpalatable at all stages sequestering the iridoids (I) and (2) as larvae and retaining them into adulthood. By contrast the butterfly Junonia coenix contains the same two iridoids in the larval and pupal stage but lacks them in the adult which has cryptic coloration. A third situation exists in the catalpa sphinx moth (Ceratomia catafpae) which in the larval stage selectively sequesters catalpol from a range of five iridoids detectable in its food plant ;the adult again is cryptic and lacks iridoids.These three insect species are all specialist feeders on plants of the Plantaginaceae and Scrophulariaceae which are rich in iridoids. The gypsy moth (Lymantria dispar) which is a generalist feeder when raised on an artificial diet containing catalposide as 3.6 % of the dry weight not surprisingly excretes the iridoid unchanged in the faeces. There is little doubt that storage of iridoids by the specialist feeders makes them distinctly unpalatable and bitter to avian predators. l2 A fourth iridoid paederoside (4),has been identified in the aphid Acyrsiphon nipponicus which feeds on members of the Rubiaceae ;the iridoid is present to the extent of 2 YOof the dry weight mainly in the secretion from the c~rnicles.'~ It is again NATURAL PRODUCT REPORTS 1989 derived from the aphid's food plant Paederia scandens which contains paederoside and other iridoids (e.g.asperuloside) which are not sequestered. The reason for selective sequestration may lie in the fact that (4)contains a methyl thiocarbonate group and that methyl mercaptan may be released from it by enzymic hydrolysis in the aphid's defensive secretion. Paedero- side is used by the aphid in defending itself from attack by the ladybird beetle Harmonia axyridis. Another plant toxin which has been confirmed as being stored in an insect is the methylazoxymethanol glucoside cycasin (5).This compound has been identified in the larvae (0.02 YO),pupae (0.3 YO), and adults (1.4 Yof the dry weight) of the rare hairstreak butterfly Eumaeus at& which is endemic to the state of Florida. The insect has brilliant warning coloration; the larvae are red and yellow the adults red and black. It is a specialist feeder on the cycad Zamiafloridana the young leaf of which contains 0.21 %of its dry weight of (5) and there is little doubt that the protective toxin within the insect is of dietary origin.l4 The best-known group of nitrogenous plant toxins that are sequestered by insects is the pyrrolizidine alkaloids (PAS) (see Table 1). A masterly review of the utilization of PAS by insects for defence and for the production of pheromones has been p~b1ished.l~ A book on the chemistry and toxicology of these alkaloids has also appeared.16 According to Boppre,15 the pharmacophagous gathering of these alkaloids from suitable plants by adult Lepidopterans is probably a widespread natural phenomenon.In addition other insects such as male flea beetles grasshoppers and grass flies have been shown to be attracted to PA baits in the field. The two main Orders of butterfly that use PAS for defence and for the production of pheromones are the Danainae of Nohh America and the Ithomiinae of South America. In the latter Order there is good evidence both of a reproductive requirement for these alkaloids and of their use in defence against spiders.A brief report published in 1984 (cf. ref. l) on the PAS of these butterflies has been followed by very complete details.17 What may be a special feature of Ithomiine males is their transference to females during copulation of significant amounts (up to 1.7 mg per individual) of PAS. The alkaloids are then utilized by the females for protecting themselves and their offspring which explains the female interest in pheromone- rich males i.e. those which have been particularly active in gathering PAS pharmacophagously. The adult males collect the PAS from withered borage leaves but the nectars of members of the Tribe Eupatorieae are also an important source. The nectaries of unopened flowers can contain up to 2-4 YOof their dry weight of alkaloids which are provided to the butterfly as a reward for p~llination.'~ The pyrrolizidine alkaloids that were suspected of being present in overwintering danaid butterflies in Mexico have now been fully characterized.They are senecionine (6) integer- rimine seneciphylline intermedine lycopsamine and echina- tine present as such and as their N-oxides.ls The alkaloids are gathered by Monarch butterflies (Danaus pfexippus) from species of Senecio and from other plants that grow in the Mid- west States of the USA during the summer and are presumably mainly stored for defence. The presence of an incompletely characterized dihydropyrrolizine in the alkaloid mixture of the Monarch butterfly confirms that the PAS are also required as precursors of their pheromones.While danaid males mainly use danaidone and hydroxydanaidal as pheromones for attracting females ithomiine males (e.g.of Prittwitzia hymenaea) employ related structures such as methyl hydroxydanaidoate (7). l9 A second group of plant bases that have been acquired by insects for defensive purposes is the quinolizidine alkaloids of members of the Leguminosae. These are mainly sequestered and stored by aphids and a detailed study has been published on the interaction of the lupin aphid (Macrosiphon afbifrons) with its food plants.20 When offered a range of poisonous species of legume to feed on it selectively fed only on Lupinus species. When offered a choice of alkaloid-rich plants of NATURAL PRODUCT REPORTS 1989-5.B. HARBORNE HOH2C Me R ' \EN OH (9) R = Me (8) (10) R = Et OH bH (11) L. albus and alkaloid-low varieties it only heavily infested the former. Such specialization is clearly due to the evolutionary advantages of sequestering and storing the lupin alkaloids. Indeed when analysed M. albifrons was found to contain lupanine 13-hydroxylupanine multiflorine angustifoline and 13-tigloyloxylupanine. Tissue content of alkaloid ranged from 0.6 to 1.8 mg per gram of fresh weight and some of the ingested alkaloid appeared in the honeydew. The alkaloids provide defence to the aphid against ladybirds and the carnivorous carabid beetle Carabus problematicus. Avoidance by the latter of alkaloid-containing aphids was due to the fact that if it ate bitter aphids it became narcotized for up to 48 hours.The acquisition by members of the Lepidoptera of a third group of plant alkaloids -the pyrazines -as warning odours was described in the last Report.' A fourth group of alkaloids -the pyrrolidones -can now be added to the list of toxins that are acquired by insects for defence (Table 1). Thus the 3,4- di h ydroxy -2,5-di( hydroxyme thy1)pyrrolidine (8) (DM DP) has been isolated in high concentration from adults of the day- flying moth Urania fulgens which is yet another warningly coloured insect. In the larval stage it feeds on various Central American species of Omphalea (Euphorbiaceae) from which DMDP has been characterized.21.22 Alkaloids such as (8) are sugar analogues in which the ether oxygen is replaced by nitrogen.DMDP is an analogue of fructose and a potent inhibitor of animal glycosidases. 23 Such alkaloids probably function in plant defence by disrupting the enzyme systems of herbivores and it is likely that sequestration of DMDP by this moth likewise protects it from its avian predators. Cyanogenic compounds are well known to be protective agents in aposematic insects such as the burnet moths and heliconid b~tterflies,~~ but unlike the alkaloids described above have been assumed to be of insect origin rather than derived from their diet. Thus experiments based on feeding labelled valine and isoleucine to the larvae have shown that the five-spot burnet moth (Zygaena trifolii)is able to synthesize its own cyanogenic glycosides particularly linamarin (9) and lotaustralin This is in spite of the fact that its food plant Lotus corniculatus contains both of these cyanogens albeit in varying amounts due to the polymorphism that exists among natural populations.It is now clear from further experiments with the 14C-labelled cyanogens (9) and (lo) that larvae are able to retain between 20 and 45 YOof the glucosides they have consumed.26 Hence the first evidence has been produced that an insect that depends on protective toxins for its survival from predation can both sequester them from the diet and synthesize them de novo in its own tissues. The ability of 2. trifolii to do this explains the relatively high cyanoglucoside content of this particular insect.Other cyanogenic Lepidopterans (such as species of Heliconius) which feed on plants that lack these glucosides have far smaller amounts of linamarin and lotaust- ralin in their tissues. Further biosynthetic experiments in species of Zygaena and Heliconius have confirmed that the (9) and (10) that are synthesized de novo in these two insect groups are formed via 2-methylpropanenitrile and 2-methylbutanenitrile respectively as intermediates. 27 Radioactively labelled precursors were incor- porated into the cyanoglucosides at a high level (15-72 YO)at all stages of the insects' life cycle. The site of biosynthesis and the transport of glycosides (9) and (10) in the larvae of Zygaena trifolii have also been determined.28 Chemical analysis of the six-spot burnet moth (Zygaena jilipendulae) has shown that cyanogenic glucosides are accomp- anied by histamine in the defensive secretions.Linamarin and histamine contents are greater in the wing than in the body. Feeding tests on the common quail (Coturnix coturnix) which is a typical predator of these moths have shown that histamine synergizes with linamarin as a deterrent. A solution of 0.001 YO histamine and 0.028% linamarin was aversive to these birds. 29 Detoxification of hydrogen cyanide in insects may occur by one of two routes conversion into thiocyanate by the enzyme rhodanese [thiosulphate sulphurtransferase] and conversion into 3-cyanoalanine via L-3-cyanoalanine synthase. Wittholm and Naumann30 consider that the presence of the latter route in insects is more closely indicative that an insect is protected by cyanogens than the presence of rhodanese.They surveyed a wide range of Lepidoptera and recorded cyanogenesis for the first time in some fourteen Families. 3-Cyanoalanine was detected in both cryptic and aposematic species including a danaid Danaus chrysippus which is primarily defended by cardenolides and alkaloids (see Table 1) and the cabbage white (Pieris brassicae) which is protected by storage of ingested glucosinolate. In neither case however was linamarin or lotaustralin detected so it is not yet clear whether cyanogenesis is truly defensive in these insects. One of the first plant-derived chemical defences to be recognized in butterflies was the use of milkweed cardenolides by the Monarch butterfly and this interaction continues to be of interest.As reported in 1986,' it is possible to identify which species of Asclepias was fed upon in the larval state from the t.1.c. profile of the cardenolides that are stored in the adult. This fingerprinting technique has now been extended to Monarchs that had been reared on an Eastern American species Asclepias viridi~.~' This plant is utilized as a food plant by migrant populations as they return from the site of their winter hibernation in Mexico. The thin-layer-chromatographic profile of cardenolides is distinct from that of the four other species of Asclepias which have been similarly c~rnpared.~' While some of the plant cardenolides are metabolized during the process of sequestration and storage in the butterfly others are absorbed unchanged.This seems to be true of aspecioside which occurs in two food plants (A. speciosa and A. syriaca) and has also been obtained from the b~tterfly.~~ Its structure has been established as 12fl-hydroxy-5a-tanghinigenin3-0-(6-deoxy-/?-D-alloside) (1 1 ). Among insects other than the Monarch butterfly which depend on dietary cardenolides for defence is the bug Caenocoris nerii. This normally feeds exclusively on cardeno- lide-containing seeds of oleander (Nerium oleander) but it can be successfully reared on sunflower seeds which are cardeno- lide-free. Bugs were fed on oleander or on sunflower seeds and both groups of bug were offered as food to the common quail (Coturnix coturnix);85YOof the sample was uneaten.However the bugs that had been fed oleander seeds were significantly less likely to be killed as compared with those that had fed on sunflower seeds. Thus while other means of protection are probably present in these bugs the presence of cardenolides in the tissues would seem to improve their chances of surviving predation.33 Cardenolides are also sequestered by the brightly (12) R = H (13) R = Me OH OMe h AOGlc OGlc 0 W (14) (15) coloured aphid Aphis nerii from its food plant Asclepias curassavica. In this case the aphid is protected from predation by the orb-web spider Zygiella x-notata. Spiders that eat these toxic aphids build severely disrupted webs due to the psycho- active effects of the ingested cardenolide.Similar results occur if spiders are fed solutions of digito~in.~~ The dietary sequestration of secondary compounds by insects is not confined to toxins and a number of other classes of plant substance can be found especially in butterflies and moths. Carotenoids are widely present the concentrations varying considerably from species to species. Certain aposematic Lepidopterans can accumulate large quantities and Rothschild et al.35 have suggested that in such cases the carotenoids might have a protective function e.g. by preventing free-radical oxidation of phenolic materials. Such protection appears to occur in the Aristolochia-feeding butterflies which store nitrophenanthrenes;for example Battus philenor has 726 pg of carotenoid per gram of dry weight.Evidence in support of this hypothesis is that related butterflies which mimic them in colouring but do not store toxins have significantly lower concentrations of carotenoid in their tissues. On the other hand the Monarch butterfly has low concentrations of carotenoid although it stores cardenolides ; perhaps in this case the butterfly has no need of photo-protection. The coumarins umbelliferone (1 2) and herniarin (1 3) have recently been detected36 in larvae of the small ermine moth Yponomeuta mahalebellus that had been reared on leaves of Prunus mahaleb. While the concentration of coumarins in the plant is 0.54% of the dry weight the moth only contains 0.0034.004YO.Although these moths are distasteful to birds it is doubtful whether the low levels of coumarins in their bodies have a protective function.It is interesting however that the carotenoid content of this moth (0.1% of its dry weight) is relatively high (see above). Dietary flavonoids are also sequestered and stored by about 10% of butterflies although the purpose of this is not yet clear.37 Flavonoids may have a function in wing coloration but they are unlikely to be distasteful to avian predators although this has not yet been tested in detail. Recent studies on flavonoid-ingesting butterflies and their food plants indicate that metabolism or a change in conjugation occurs in most cases. This is true of the marbled white butterfly (Melanargia galathea) where the dietary flavone tricin 7-0-glucoside is recovered in the insect as the 4’-O-glucoside (14)and possibly It as the 4’-0-~ulphate.~* is also true of the swallowtail Eurytides marcellus the body and wings of which contain the flavonol quercetin 3-0-glucoside (1 5); the larval food plant Asimina triloba by contrast contains the 3-0-rutinoside 7-0- glucoside the 3-0-rutinoside and the 3-0-glucoside of quer- ~etin.~~ Likewise the chalkhill blue butterfly (Lysandra coridon) contains a simpler mixture of kaempferol quercetin and isorhamnetin glycosides than its leguminous food plants so some hydrolysis of glycosidic links appears to occur in viv~.~O NATURAL PRODUCT REPORTS 1989 The effect of breadth of diet on autogenic chemical defence has been studied for the generalist grasshopper Romalea guttata.The defensive secretion contains hydroquinone cate- chol p-benzoquinone phenol guaiacol and 4-methoxy-benzaldehyde. There was a dramatic decrease in the number and amount of phenols secreted when it was reduced to feeding on plants of a single species Allium canadense instead of its normal diet of some 26 plant species. Thus diet does have a significant effect on the secretion of toxins even when they are synthesized by the animal and are not of dietary origin. There are two possible explanations which still need testing; the grasshopper may depend on phenolic precursors in the food plants or the physiological stress due to restriction of its diet may reduce the resources that are allocated to 2.2 Chemical Defence in Animals 2.2.1 Beetles and Termites Beetles synthesize an astonishing range of secondary products from simple phenols to complex steroids as part of their defence against predators.The literature up to 1980 is admirably reviewed by Bl~m,‘~ and a more recent account of the chemistry and chemosystematics of the defences of beetles has been provided by Dettner.43 Here it would seem appropriate to mention some recent findings which illustrate the diversity in chemical structures encountered and the protective value of such secretions. That most notorious pest of potato crops the Colorado beetle (Leptinotarsa decemlineata) is decidedly unusual in synthesizing the glutamyl derivative (16) of a new non-protein amino acid in its defence glands.The concentration of this acid in the beetle’s secretion (1.8 x 10-1 mol dm-3) is more than sufficient to deter predation by ants since Myrmica rubra was affected by the toxin at a concentration of 1 x mol dm-3.44 In haemolymph secretions of ladybirds alkaloids such as coccinelline have been reported earlier.2 More recently a novel long-chain diamine (17) has been discovered in five species of ladybird including Harmonia leis con for mi^.^^ By contrast jewel beetles of the Order Buprestidae produce bitter principles such as buprestin B (18) which has a glucose moiety that is acylated at positions 1 2 and 6 with p-hydroxybenzoic or pyrrole-2-carboxylic Some other beetles depend on mixtures of simple aromatic compounds in their defensive secretions.For example the flour beetle Tribolium brevicornis produces 2-hydroxy-4-methoxyacetophenone the corres-ponding propiophenone 3,6-dihydroxy-2-methylbenzoic acid methyl ester and 2-ethyl-3,6-dihydroxybenzoicacid methyl ester. These ketones and esters on the face of it would seem to be fairly mild irritants to predators. However they are potent inhibitors of prostaglandin synthase and it is through such action that they are damaging to the beetle’s enemie~.~’ More elaborate chemical defences are formed within the carrion beetles Silpha americana and S. novaboraciensi~.~~. They produce a series of seven steroids not previously recorded in arthropods in the glandular annex of the rectum.Two of them [(19) and (20)] are potent feeding deterrents to jumping spiders at concentrations of as little as 1 pg. These steroids are discharged from the anus together with enteric fluid whenever the beetles are disturbed. Some beetles secrete mixtures of toxins formed by more than one biosynthetic pathway which are presumably highly effective in defence because of synergism. This appears to be true of larvae of the leaf beetle Gonioctena viminalis the defensive secretion of which is based on the three hydrocarbons hex-2-enal 6-methylhept-5-en-2-one and 6-methylhept-5-en-2- 01 the monoterpene linalool and the aromatic alcohol phenylethan~l.~~ The larvae of another leaf beetle Plagiodera versicolora depend on toxic methylcyclopentanoids and hexa- decyl acetate for defence.42 Ecological experiments have now shown that the defensive secretion is multi-purpose. Although primarily deterrents to avian predators these compounds are repugnant to conspecific adults and to the larvae of a competing NATURAL PRODUCT REPORTS 1989-5. B. HARBORNE H &-OH 0 OLC+ II O H (19) H (20) pCHOHICH2] 23 I c=o / CH 3CCH 2]gC(O)CH= CHOH Table 2 Chemical defence systems of termite soldiers Termite genus' I Biting and injecting Macrotermes Cubitermes Amitermes 2 Poison-brushing Prorhinotermes Schedorhino termes Rhinotermes 3 Glue-squirting Nasutitermes Chemical classes present Alkanes and alkenes Diterpene hydrocarbons Macrocyclic lactones Nitroalkene and farnesene Hydrocarbon ketones P-Keto-aldeh ydes Monoterpenes and cyclic diterpene alcohols I' Other termite genera tend to fall within these three classes.For further details see ref. 52. consumer of willow the Camberwell beauty butterfly (Nymph-alis antiopa). Thus a larva of P. versicolora stakes out its claim to a particular leaf of its food plant and other competing herbivores will be deterred from approaching. The defensive secretions provide it with the means of ensuring that it does not Probably the most remarkable set of defensive secretions within the arthropods are those in termites. In these social insects chemical defence is allotted to sterile soldiers which constitute between 10 and 30% of a given colony.Defence is their only function and up to 8% of their fresh weight may consist of defensive chemicals. Prestwich who has been one of the principal investigators of termite soldiers has provided an excellent summary of their defence systems (Table 2).52As he describes it these insects are walking weapons capable of biting snapping hole-plugging squirting oozing daubing defecating and exploding at their prey. (24) (29 One group of termite soldiers including those of the genus Nasutitermes produce a unique series of cyclic diterpenes the trinervitenes [e.g. (2 l)]. These glue-squirting insects eject a viscous secretion of these sticky diterpenes at their enemies. Interspecific aggression can occur among these insects ; some-times soldiers and workers of the same species may fight each other if they come from different A second group of termite soldiers (see Table 2) contain electrophilic lipids which are contact poisons.These include nitro-alkenes [e.g. (22)] vinyl ketones (23) and P-keto-aldehydes (24). The toxins are brushed onto the cuticle of the attacker by the soldier and once absorbed these lipids are capable of reacting rapidly with biologically essential nucleophilic sites with fatal consequences. These materials are liable to be released on conspecific workers as well as on predators. Termite workers fortunately are protected from the poisons since they can detoxify them via substrate-specific reductases in their tissues. The colonies of a third group of termites are protected by 'biting and injecting' soldiers (see Table 2).The contents of their frontal defence gla& are released into a fresh wound of the attacker and the would-be assailant then bleeds to death. In the genus Amitermes defensive secretions consist mainly of macrocyclic lactones [e.g.(25)] many of which are unique to termites. Sometimes the act of biting is suicidal to the soldier since it may become inextricably 'stapled' to the attacker. 2.2.2 Marine Organisms Some fish which lack physical protection produce defensive secretions to repel the attack of the larger predators such as sharks. This is true of the Pacific sole (Pardachirus pavoninus) which as described in the last Report,l synthesizes six steroidal toxins glycosylated with N-acetylhexosamine for this purpose.Further work on these secretions has now revealed the presence of three ichthyotoxic peptides pardaxins P1 to P3.54 The balance of hydrophobic and hydrophilic amino acids in their 33-residue sequence gives these peptides surfactant properties. Pharmacologically they are remarkably similar in activity to NATURAL PRODUCT REPORTS 1989 Me predator. The anemone fish not only develop immunity to the sting of the anemone but also respond to chemical attractants that are released by the anemone. These attractants which are key signals in the symbiotic association between fish and anemone have been identified in two cases.55 Five chemicals are involved in the association between the anemone Radianthus kuekenthali and the fish Amphiprion perideraion two aply- sinopsins two dihydroaplysinopsins and amphikuemin (26).The indole-based aplysinopsins had been isolated earlier from marine sponges but amphikuemin is a new marine chemical active at a concentration as low as lo-’ mol drn-,. In a second I symbiosis between Stoichactis kenti and Amphiprion ocellaris 0 histamine and tyramine have been identified as the active attractant^.^^ Although barnacles are mainly thought of as a nautical problem because of their tenacious hold on ships’ hulls they 0 are a potential threat to other sea creatures that live anchored OH __ to the sea bed for the same reason. The sea pansy Renilla reniformis however protects itself from being covered by the larvae of a barnacle by synthesizing specific anti-fouling 0 deterpenes.Three substances renillifoulins A-C (27)-(29) (27) R = Mt have been characterized in the secretion of this octocoral. They (28) R = Et are highly effective in inhibiting the settlement of larvae of the (29) R = Pr barnacle Balanus amphitrite amphitrite. Such compounds have obvious commercial potential in the marine industry as additives to anti-fouling paints.56 While marine crustacea such as barnacles and many marine molluscs have protective shells they may also employ chemical means for diverting their predators. This is certainly true of the intertidal limpet ColliseIla limatula which has been described as producing the most potent fish feeding inhibitor yet encountered by chemist~.~’ This compound limatulone (30) is one order of magnitude more effective than polygodial (the hot-tasting antifeedant of the dorid nudibranch Dendrodoris limbata).l It also inhibits predation on the limpet by crabs.Chemically limatulone is interesting in being a C, isoprenoid consisting of identical C, units. A multitude of rare and exotic chemicals have been described OMe from marine organisms and new structures continue to be discovered. Such substances may be shown to have biological activity but it is usually difficult to assign them a function unless their ecology has been explored. For example an H H H H antimicrobial blue pigment with a linear tetrapyrrole structure (31) has been identified in the bryozoan Bugula dentata5* but it CI’ (31) is not known whether the antimicrobial activity is protective in vivo.Similarly a modified peptide jaspamide (32) was recently reported from a species of marine sponge of the genus Jaspis; OH it had both insecticidal and antifungal activity in ~itro.~’ Experimental work however is still needed on its ecological role in the marine environment. 2.3 Pheromones Lepidopteran pheromones particularly those of sexual attrac- tion continue to receive considerable attention. Such sex pheromones are generally species-specific and may be important as isolating mechanisms between species. This seems to be so in the fall cankerworm (Alsophila pometaria). A populational study within the United States of America showed a constancy in pattern in spite of genetical variation.The female pheromone always consisted of (32,6Z,9Z)-nonadeca- 3,6,9- triene (3Z,62,92,1 lE)-nonadeca-3,6,9,1l-tetraene,and (3Z,62,9Z 1lZ)-nonadeca-3,6,9,1I-tetraene in the ratio of 25:65 15.60 By contrast in the oriental armyworm (Mythimna separata) there appears to be infraspecific variation in the blend of compounds melittin the peptide of bee venom. Undoubtedly the pardaxins in the female pheromone. Males of this species from mainland synergize the steroidal glycosides in protecting this sole from China are not attracted to the usual (8 1) blend of (1 1Z)-attack by sharks. hexadec- 1 1 -enyl acetate and (1 12)-hexadec- 1 1 -en01 which is Other fish without any obvious protection from attack released by females in Japan.Instead the Chinese females were depend on a symbiotic association with sea anemones for their found to be releasing three pheromones -(1 1 Z)-hexadec- 1 1-survival. These so-called anemone fish swim in close proximity enal hexadecanal and (1 12)-hexadec-1 1-enol -in the ratio of to the anemones which release fatal stings at any unwary 100:10:0.1. Similar differences in the female pheromones that NATURAL PRODUCT REPORTS 1989-5. B. HARBORNE (34) R = H (35) R =OH are released by Japanese and Chinese insects were observed in the purple stem borer (Sesamia inferens).6’ A nice example of overlap in the composition of sex pheromones between two species is illustrated by the cabbage looper moth (Trichopfusia ni) and the soya bean looper moth (T.incfudens).Females of both species use (7a-dodec-7-enyl acetate as a major component and have two minor components in common. However the sex pheromone of T. ni has three additional species-specific components and that of T. includens has two.6z In the European genus of small ermine moths (Yponomeuta) the females of eight species that have been examined release varying amounts of up to eight pheromones but these are not all necessarily attractants in all taxa. The males of all eight species have abundant receptor cells for (1 1 Z)-and (1 1 E)-tetradec- 1 1-enyl acetate which supports the idea that the simplest pattern of two compounds (as occurs in Y. roreffus)evolved from a more complicated one by loss of desaturase ability.63 Moth pheromones such as (9a-tetradec-9-enyl acetate (93-tetradec-9-ena1 and (1 12)-hexadec- 1 1-enal have been found to be synthesized by the bolas spider Mastophora cornigera.The purpose however is a nefarious one. The spider uses this chemical mimicry to attract male moths as food. These spiders are notable in that they only produce a minimal type of web and that they only capture one type of prey male moths. Analysis of the spider’s secretions suggests that individuals have the ability to vary the composition of their pheromone output can mimic the pheromone blends of the females of more than one species of moth and hence can capture a regular supply of male Grant et af.65 have demonstrated that the alkanes n-tricosane n-tetracosane n-pentacosane and n-heptacosane which are present in the scales of the female of the tussock moth Orgyia feucostigma function as a copulation-releaser pheromone.This is the first demonstration in moths of chemicals (other than those originating from the pheromone gland) that are capable of doing this. The copulation-releasing effect is dependent on the prior stimulation of the males with the female sex pheromone. This contrasts with the situation in the azuki bean weevil (Calfosobruchus chinensis) where a copulation-releaser pheromone (erectin) is independent of the normal sex phero- mone and in some Diptera where a non-volatile sex pheromone itself releases copulation. The chemicals that release copulation in these other insects are also cuticular hydrocarbons but of a greater molecular weight.Many reports have appeared in the period under review of the identification of sex pheromones in moths that had not previously been examined but the structures are usually fairly expectable. Slightly more unusual is the finding of the completely saturated hydrocarbon 5,9-dime thy1 hep tadecane (33) as the sex pheromone of the moth Leucoptera sciteffa.66 (36) 0 Again the finding of (E,E)-and (Z,E)-farnesal as the male pheromone blend of the rice moth (Corcyra cephalonica) is exceptional since moths rarely contain terpenoidsg7 Farnesal has only been reported once before in insects in the mandibular gland of the ant Lasius fufigino~us.~~ In C. cephafonica,it could be dietarily derived from farnesol.The production of danaidal (34) and (-)-hydroxydanaidal (35) in the scent glands of the arctiid moths Phragmatobia fuliginosa and Pyrrharctia isabeffa is almost certainly due to dietary or pharmacophagous intake of pyrrolizidine alkaloids.68 Males of the first species have mainly danaidal while those of the second have mainly the hydroxy-derivative and the respective females respond prefer- entially to one or other pheromone. One other male moth pheromone which is of dietary origin is the isocoumarin (R)-mellein (36). This has been found in the bumble-bee wax moth (Aphomia sociefla). It is probably of fungal origin in fact since Aspergiffus ochraceus which synthesizes (R)-mellein was detected in the intestines of the last-instar larvae of the moth and in the bumble bee’s nest on which the larvae feed.69 The first report of a pyrazine as a sex pheromone in insects involves a fruit fly rather than a moth.2-Methyl-6-vinylpyrazine (37) has been identified as the male sex pheromone of the papaya fruit fly (Toxotrypana curvicauda). 70 Pyrazines have been previously reported in insects as warning odours in Lepidopterans in poison glands of ants and wasps and as trail pheromones in attine ants. Turning to trail pheromones a new type has been found in the ant Tetramorium impurum where methyl 6-methylsalicylate (38) fulfils this The most remarkable discovery in this area however is of the steroid 5P-cholestane-3,24-dione(39) as the trail pheromone in larvae of the eastern tent moth (Malacosoma americanum).It is the first steroidal trail phero- mone to be found and the first report of a larval trail pherom~ne.’~ Like other trail substances it is highly active and the caterpillars have a threshold sensitivity to (39) of g per millimetre of trail. This caterpillar lives in colonies of 50 to 300 individuals and foragers move from the tent to food plants especially young leaves. They not only lay a trail with (39) but also mark the tastiest leaves e.g. of Prunus serotina so that caterpillars can find the best food by following a trail.73 The pheromones of beetles do not receive the same attention as those of moths although their ecological chemistry can be just as interesting. It has just been discovered for example that two species of beetle share the same female pheromone stegobinone (40) the common furniture beetle (Anobium punctatum) and the drugstore beetle (Stegobium paniceum).Although both beetles belong to the same Subfamily Anobiinae they are not closely related and indeed the two genera appear to be separated in evolutionary time by between 40 and 50 million years. In Nature the two species are allopatric so that NATURAL PRODUCT REPORTS 1989 attempt to determine the nature of this pheromone Preti et a1.81 have used gas chromatography-mass spectrometry to determine the steroids in male and female axillary secretions. The results show that the concentration of the androstenol(43) 0 varies in both men and women with the highest concentrations HI (411 (421 H0’.+ H (43) there is no confusion in breeding.However the drugstore beetle has recently become a pest of warehouses and stores so it is now sympatric with the furniture beetle.74 Only one stereoisomer [(l’R,2S,3R)] of (40) is active and this activity decays if it is stored for two weeks due to conversion into the (1’S,3R)-form which is not only inactive itself but is inhibitory of the (l’R)-f~rm.’~ The best known aphid pheromones are those of alarm; one or other of the two sesquiterpenes germacrene A and (E)-p-farnesene is commonly released. The turnip aphid (Lipaphis erysimi) however responds only weakly to (a-p-farnesene . although related species respond strongly. Dawson et ~ 1have ~ now found that the alarm response in this aphid is substantially increased by mixing (E>-p-farnesene with plant-derived iso- thiocyanates which have been identified in the aphid’s volatile secretions.Ally1 and but-3-enyl isothiocyanates turned out to be more active than but-2-yl and pent-4-enyl derivatives while 2-phenylethyl isothiocyanate (which is the major volatile compound of the food plant) was inactive. These results show that at least in certain species of aphids plant volatiles can serve two purposes i.e. as feeding stimulants and for activation of pheromones. The first identification of sex pheromones in aphids has also been reported from the same lab~ratory.~’ Nepetalactone (41) and the related lactol(42) are released from the hindlegs of the female of the vetch aphid (Megoura viciae) to attract the male.Nepetalactone is a well-known monoterpene lactone being present as a feline attractant in catmint (Nepeta cataria) but any connection between this occurrence and the sex life of the aphid is hard to conceive. Moving up the evolutionary scale there is a report of testosterone being released in the urine of male sea lampreys (Petrornyzon marinus) to attract fern ale^.')^ Goldfish (Carassius auratus) also depend on a steroid 17a,20P-dihydroxypregn-4-in women being produced in the mid-follicular phase prior to ovulation. It has not however been clearly implicated as the key pheromone in this interaction. The odour of the axillary secretion is due in part to aliphatic acids which range from C to C, in chain length and it is conceivable that they could also be involved.In quite a different type of experiment the androstenol (43) was examined as a spacing pheromone.82 Rest-room stalls were sprayed with either (43) or androsterone as a control. Men avoided the odour of (43) whereas women’s choice was unaffected by either steroid. While the quantity of odour that was used in this experiment was greater than that in human armpits the results do suggest that this compound may have a pheromonal effect in humans. 3 Plant-Animal Interactions 3.1 Constitutive Chemical Defence Much effort continues to be expended on testing the hypothesis that plants are defended from herbivory by their constitutive secondary chemistry. Not all experiments necessarily give a clear-cut answer one way or the other.Beart et for example examined the ability of different gallotannins to complex with the protein bovine serum albumin expecting the biosynthetically more evolved compounds to bind more tightly than simple pentagalloylglucose as predicted by the co-~ evolution theory. In fact there was no such correlation between the structure of a gallotannin and its capacity to bind to this protein. However this is not really surprising since such a test in vitro is rather far removed from what actually takes place when a herbivore feeds on plant tissues. The interaction between a phytophagous insect and a tannin-containing plant is a highly complex encounter as has been revealed by a number of new experiments described below.Clearly there are constraints on chemical co-evolution as Berenbaum et discovered when they investigated the parsnip webworm (Depressaria pastinacella) which is the only major herbivore on wild parsnip and an insect which has no other host plant. Resistance to feeding is present in parsnip and controlled by the synthesis of four furanocoumarins but this is unfortunately negatively correlated with low production of seeds. Thus the wild parsnipwebworm interaction appears to have reached a ‘stalemate’ in the co-evolutionary arms race. The plant has to allow some limited feeding on its tissues or else it will not be successful in reproducing itself. Various suggestions have been made that low to moderate herbivory may not actually be harmful to plants; conceivably en-3-one to synchronize male-female spawning readine~s.~~ herbivory might increase plant productivity by stimulating growth.85 BelskyE6 has however examined the experimental Steroids such as androsterone may be important as male pheromones in musk deer (Moschus rnoschiferus).80The classic musky odour of these deer has always been considered to be due to muscone (3-methylpentadecanone) which is reportedly present as 22% of the glandular secretion of Nepalese stags.However analysis of the glands of Siberian musk deer failed to show it to be present. Instead eight steroids including androsterone were detected. This has a santal-like odour and could conceivably be replacing muscone as the male pheromone in these particular populations of the deer.The search for human pheromones continues (see last Report1) and there is now some evidence implicating 5a-androst-16-en-3a-01 (43) in this role. One of the established pheromonal interactions in Man is that the length of menstrual cycle and the fertility of women is affected by (i) intimate contact with men or (ii) contact with male axillary secretions or (iii) the axillary secretions of women. One of the results of this is the menstrual synchrony that develops among female undergraduates living together in a hall of residence. In an evidence adduced and finds that in no case is it as yet really satisfactory. He concludes that this is a hypothesis which has yet to be proved. In the present section the different plant toxins that have been investigated from the ecological viewpoint will be considered under three general headings -nitrogen compounds terpenoids and phenols and tannins -and then recent work on latexes and trichome defence systems will be reviewed.3.1.I Nitrogen Compounds Although there is now a large body of evidence to indicate that cyanogenesis protects plants from their herbivores there is still much to be learnt about a defence system which is genetically variable in most of the plants where it is present. There have been two contrasting studies of slugs feeding on Trifofium repens in which the interaction has been considered from both the plant’s and the slug’s point of view. Analysis of the NATURAL PRODUCT REPORTS 1989-5. B.HARBORNE k"y" RO OR n (45) releasable cyanide in young cyanogen-positive clover seedlings showed that there is a 33% increase from five-day-old to 35- day-old plants with especially high levels occurring in stem and c~tyledon.~' When garden slugs (Arion hortensis) were allowed to graze on five-day-old seedlings they killed both cyanogenic and acyanogenic forms. However discrimination occurred when they were offered 16- 23- and 35-day-old seedlings; while there was still some grazing on the cyanogenic forms very few of these seedlings suffered lethal damage. In this case cyanogenesis would seem to be maximally protective when the plant is most vulnerable to destruction and to be concentrated in the more vital organs (stem and cotyledon) rather than in the more expendable leaf tissue.In a second paper on the grazing of the slug Deroceras reticulatum on cyanogenic and acyanogenic clover it was found that slug populations vary in their response to cyanogenesis according to the frequency of cyanogenic plants in a given population.88 Slugs that had been obtained from sites with a low frequency of cyanogenic clover tended to avoid eating cyanogenic forms whereas slugs from sites where there was a high frequency of cyanogenic plants appeared to have become adapted to the toxin and showed much less dis- crimination. Thus the selective advantage that is enjoyed by a cyanogenic form under grazing by this slug will only be substantial where the frequency of cyanogenic plants is low (e.g.between 11 and 24%). There will be less protection when the frequency of the cyanogenic form is high (e.g.65 %). These results help to explain why polymorphism for cyanogenesis continues to remain widespread in clover populations. Non-protein analogues of phenylalanine and tyrosine such as L-dopa are well known to occur in quantity in legume seeds and to provide protection from seed predators such as bruchid beetles. A new analogue 4-aminophenylalanine has now been detected in seeds of five species of Vigna.89Although there are quantitative variations in the amounts in the seeds toxicity tests against two important pest bruchids showed that the levels present were always sufficient to provide a barrier to feeding. One of the bruchids (Zabrotes subfasciatus; lethal dose 0.3 YO) was more susceptible than the other (Callosobruchus maculatus ; lethal dose 0.75 YO),but even at lower concentrations this non- protein amino acid when added to artificial diets markedly increased the development times of both species.Sub-lethal doses thus had significant ecological effects on two predators. Moreover 4-aminophenylalanine is four times as toxic as L-dopa to the same beetles. Hiptagin (44)and three related nitro-derivatives of glucose occur in the root and foliage of the leguminous pasture plant Lotus pedunculatus and they appear to be responsible for the resistance of the root to the grass grub Costelytra zealandica. Ingestion by third instar larvae was significantly reduced when they were offered feeding discs containing between 0.1 and 0.25 YOof hiptagin ;these levels are similar to those found in the plant.Resistance to attack by other insects has been observed suggesting that the nitro-compounds protect this species of Lotus from a range of predator^.^^ Although alkaloids are the most widespread nitrogen- containing substances of secondary metabolism occurring in some 20 YOof species of flowering plants their contribution to plant defence has often been overlooked. This is the view of Barbosa and Kris~hik,~~ who have pointed out that the alkaloid content of eastern American deciduous trees is probably just as important for defence as the much discussed quantitative tannin defence of these trees. Alkaloids are widely present in such a flora and can significantly alter the feeding preferences of a highly polyphagous herbivore such as the gypsy moth (Lymantria dispar).For example the tulip tree (Liriodendron tulipifera) which is never attacked contains fourteen alkaloids including glaucine in the leaves. Avoidance is related at least in part to the alkaloid content. Thus when glaucine was incorporated into the diet of gypsy moth larvae it reduced their survival and upset the growth characteristics of the insect. Alkaloids may also be important factors in the chemical defence of aquatic higher plants; such plants generally lack structural defences and yet are seldom eaten.92 A survey of fifteen representative species showed an alkaloid content of between 0.13 mg per gram of dry weight in Heteranthera dubia and 0.56 mg per gram of dry weight in Potamogeton crispus.These levels are within the range which would have a deterrent effect on potential herbivores. The discovery that alkaloids occur widely in these plants contrasts with earlier in which it was suggested that secondary compounds were poorly represented in aquatic plant species. As mentioned in the last Report,l the pattern of accumulation of alkaloids within a given plant may be closely correlated with a defensive role; the concentration of caffeine in the coffee plant is highest in those tissues most vulnerable to attack. A similar study of the production of indole alkaloids in Catharanthus roseus under conditions of drought and wounding stress has now shown that the synthesis of these alkaloids never falls below the level that is needed for prote~tion.~~ Although drought stress causes a decrease in relative alkaloid content (i.e.YOdry weight) the concentrations in the living tissues differ little from those in non-stressed plants.In growing tissues wounding leads to an increase of 100% in accumulation of alkaloids whereas wounding has little effect on alkaloid levels in non-growing tissue. Overall the results indicate a very economic allocation of energy and nutrients and show that the synthesis of alkaloids is maintained at a sufficient level to provide effective chemical defence against predators at all times. An example of a specific co-evolutionary response to herbivory has come to light in the alkaloids of the genus Nicotiana.Species of Nicotiana Section Repandae (for example N. repanda) are highly toxic to the tobacco hornworm (Manduca sexta) in spite of its well-known ability to avoid nicotine poisoning when feeding on tobacco. Topical application of 1 mg of the crude exudate of leaves of N. repanda to larvae caused 100YOmortality in 24 Analysis of the alkaloids showed that N-acyl analogues of nicotine such as (45) are present and the insect is not able to detoxify them. Significantly (45) is non-toxic to other insect pests such as Heliothis virescens so the synthesis of N-acyl-nicotine derivatives probably represents a special adaptation in Nicotiana to feeding by the tobacco hornw~rm.~~ 3.1.2 Terpenoids The effects of plant monoterpenes on feeding of ruminants tend to be indirect because the compounds are not necessarily toxic to the animal but have an inhibitory effect on the micro- organisms in the rumen when present in sufficient concentration in the plant fodder.A study of the effect of the monoterpenes in gymnosperm tissue on food selection by calves of the red deer (Cervus elaphus) suggests that there is at first rejection largely on the basis of odour concentration but that there is some degree of acceptance with time.97 Essential oils can have a more dramatic effect on feeding behaviour in voles and (47) (48) I &I GkOW (49) snowshoe hares and it has been shown that pine oil is an effective feeding repellent since the animals do not become habituated to the odours.Natural resistance in the Douglas fir (Pseudotsuga menziesii) to grazing by both deer and snowshoe hares is probably due to elevated terpenoid levels in the foliage as compared to susceptible strains.s8 Animals such as the greater glider (Petauroides uolans) and the brushtail possum (Trichosurus vulpecula) are adapted to feeding on leaves that contain significant levels of essential oil ;indeed these animals feed almost exclusively on eucalyptus leaves.99 They are able to do this because the microbial population in their hindguts is largely protected from the deleterious effects of Eucalyptus oils by their absorption from the stomach and small intestine and their subsequent detoxification in the liver.Levels of essential oils in plants can vary considerably according to environmental conditions and this can affect the extent of herbivorv on a plant. For example the generalist moth Pseudoplusia includens was inhibited from feeding on leaves of camphorweed (Heterotheca subaxillaris) that had been collected from plants grown under nitrate-limiting conditions due to the elevated levels of essential oil. By contrast larval consumption and larval growth were enhanced on leaves from plants that had grown under conditions of adequate nitrate supply. The content of volatile oils in the leaves dropped from 6.25 to 4.02 mg per gram of dry weight when the nitrate supply was adequate.loO Similar observations in Cleome serrulata showed that the toxin content (in this case a glucosinolate rather than a terpenoid) increased as a result of drought stress to levels which inhibited damage by insects.lol Chemical defence in the tropical green alga Caulerpa ashmeadii to herbivory by fish is provided by several sesquiter- penoids such as (46) and (47) which are unique to this species.lo2Earlier claims that the alkaloid caulerpin which also occurs in the alga was responsible for deterrency were not borne out in practice.It was inactive against the tropical damselfish Pomacentrus coeruleus at a dose of 20 pg CM-~ whereas the sesquiterpenoids caused death at a dose of 2.5 pg cmP3. As a consequence of insect-plant co-evolution plant toxins are often differential in their effects on insect feeders.For example the triterpenoid cucurbitacins of members of the NATURAL PRODUCT REPORTS 1989 Cucurbitaceae are normally repellent to generalist insect feeders while they can be attractive to specialists. Indeed some cucumber beetles actually sequester these toxins and use them for defence (see Section 2.1). However it is always possible as a result of continuing co-evolution for plants to evolve toxins that are capable of repelling specialist feeders. This seems to be true of the cucurbit Mornordica charantia which produces a cucurbitacin glycoside momordicine I1 (48).lo3 This substance at a concentration of 3.2 mg CM-~(comparable to the natural levels in the plant) reduced the feeding of red pumpkin beetles (Aulacophora foveicollis) in bioassays in uitro.The related cucurbitacins B and E were feeding stimulants to the same beetle. Momordicine 11 is unique among known cucurbitacins in that C-19 is an aldehyde group there is a 23-0-glucosyl group and it lacks a keto-group at C-1 1. Although cucurbitacins are characteristically bitter and hence repellent they can occasionally be sweet. In fact the rhizomes of Hemsleya carnosiflora (Cucurbitaceae) contain six of these triterpenoids :one is tasteless two are sweet and three are bitter. The relationship between taste and structure in this series is complex but substitution of the triterpenoid skeleton at both ends with sugar substituents may lead to sweetness as in one of the sweet substances (49).lo4 In concluding this section it should be mentioned that a book on natural plant molluscides has recently been pub- lished,lo5 since the most active agents are saponins.One of the best candidates for controlling the bilharzia snail in Africa is the leguminous tree Swartzia madagascariensis the fruits of which are rich in molluscidal sa~onins.~O~ More recent examination of the leaves of the related Swartzia simplex showed the presence of similar saponins but in general they were less active than the components of S. madagascariensis.'06 One final point of interest is the discovery in plants of toxins which at one time appeared to be strictly of animal origin. Bufadienolides are characteristic toad poisons but they are also present in plants of the Crassulaceae. Three novel bufadienolide glycosides have been reported in Cotyledon orbiculata and are the cause of poisoning of cattle which eat this plant,lo7 while two toxic bufadienolide orthoacetates have been isolated from flowers of Bryophyllum tubgorum.lo8 3.1.3 Phenolics and Tannins Many of the classical methods for measuring phenolic/ tannin levels in plant tissues suffer from empiricism and may fail to distinguish between different groups of natural phenol. This obviously makes it difficult in experiments on animal herbivory to know precisely what factors are responsible for deterrence particularly since only certain polyphenols notably tannins are able to precipitate protein and hence exert an adverse nutritional effect. During the period under review attention has been focussed on methodology and efforts have been made to improve the analysis of tannins.Two important papers have appeared on a critical analysis of techniques for measuring tannins in ecological studies.loS. 110 Bovine serum albumin has been covalently labelled with Remazol Brilliant Blue R to provide a substrate for a convenient spectrophotometric assay for the precipitation of proteins by vegetable tannins.l1' Concentrations of tannins in plant extracts have been deter- mined by allowing them to react with protein and then measuring the precipitated complex by radial diffusion in an agar slab. The limit of detection is 0.025 mg of a condensed tannin and the precision is 6%.l12 The relative affinities of condensed tannins that had been purified from sorghum quebracho wattle and runner bean for six dissimilar proteins were determined by a competitive-binding assay.The results indicated that tannin-protein interactions may be specific for different tannins as well as for different proteins. The highly specific nature of these interactions suggests that the differences in affinity are ecologically significant. 113 Hagerman and rob bin^"^ have pointed out that some protein-tannin inter-actions may produce soluble as well as insoluble complexes NATURAL PRODUCT REPORTS 1989-5. B. HARBORNE H and that the metabolic effects of the former are as yet unknown. Ecological experiments have shown that simple soluble phenolics can sometimes have adverse effects on feeding of animals.Ferulic acid whether present as such or released by enzymic breakdown appears to be a significant factor in deterring feeding of the maize weevil Sitophilus zeamais on maize seed.l15 in vitro ferulic acid (50) is an effective antifeedant at a concentration of 0.05 mg per gram. Similarly chlorogenic acid is a feeding deterrent to the leaf beetle Lochmaea capreae cribrata which feeds on members of the Salicaceae.'" Both phenols and tannins are important deterrents to the molluscs Helix pomatia and Arion ater although this snail and slug are happy to eat plants that contain substances which are toxic to mammals."' Caffeic acid esters with either glucose or rhamnose occur variably in populations of the plant Plantago major and there is evidence that molluscs tend selectively to avoid plants that are rich in the rhamnose ester.'18 These results with terrestrial molluscs tie in with similar studies of marine molluscs grazing on seaweeds where phenolics are again the major anti- herbivore chemicals.119 The mollusc Tegula funebralis will feed on thirteen species of brown algae but the six most preferred species have an average phenolic content of 0.83 YOof their dry weight whereas the seven least preferred average 4.53YOof their dry weight of phenolic.120 The ecological effects of hydrolysable tannins have been little studied so it is interesting to find that geraniin which occurs in leaves of Geranium species is a growth inhibitor to the polyphagous moth Heliothis virescens. However it does not seem to be involved in interactions with proteins.Geraniin is a pro-toxin because it releases the true growth inhibitor ellagic acid (51) upon hydrolytic cleavage. Ellagic acid appears to be an anti-nutritional factor because of its ability to chelate essential rnetals.lZ1 Adaptation of mammals to the adverse effects of dietary tannin through the synthesis of proline-rich (PR) proteins was mentioned in the last Report.' This adaptation which has been demonstrated in rats and humans has now been analysed in some The salivary PR proteins that are formed have a high affinity for tannins which is in part due to their being glycosylated. The oligosaccharide moieties enhance the affinity and selectivity of binding to proteins and increase the solubility of the resulting tannin-glycoprotein complex.The carbohydrate content of PR proteins is up to 40% by weight. The sugar units appear to keep the protein in an open conformation allowing the maximal amount of hydrogen-bonding to occur with the tannins. lZ3 Adaptation to tannin-rich diets may take other forms and there is evidence of an endocrine adaptation in sheep involving an increase in the rate of release of glycerol from adipose The mule deer (Odocoileus hemionus) is one animal which is not able to adapt to high-tannin diets by the synthesis of PR proteins so that its ability to graze on tree tissues is governed to a large extent by the levels of tannin present.lZ5 Tannins effectively protect the leaves and flowers of deciduous trees in summer from grazing by mule deer because of the levels present and the anti-nutritional effects of precipitation of protein.Other soluble phenolics that are present are probably also protective because they are toxic to the deer. Adaptation of insects to dietary tannin clearly takes a different form than that in mammals. In generalist insects such M e2CH C(0) OH HO (52) as Manduca sexta and Schistocerca gregaria there is evidence of adaptation through the production of surfactants which prevent tannin-protein complexes from precipitating out in the gut.126 Even so generalist insects may still limit their intake of leaves that are rich in tannins. This is true of the armyworm Spodoptera latifascia ;when offered leaves of Cecropia peltata that contained high and low levels respectively of tannins it chose the latter.12' The production of leaves with a high tannin content can however be economically costly to the plant.In this species of Cecropia there is clear evidence that seedling trees with high levels of tannins produce fewer leaves per plant than less well protected trees with a low tannin ~0ntent.l~' 3.I .4 Defence Systems The production of a latex which is widespread in the angiosperms,128 has always been assumed to be defensive particularly since the milky secretion often contains additional toxic materials in the way of terpenoids or alkaloids. Experimental evidence for a defensive role however has been limited and has been mainly confined to studies with ants.'" However in a new series of experiments Dussoord and Eisner130 have established that many mandibular insects in order to feed on latex-bearing plants have managed to overcome the latex defence by vein-cutting behaviour.Thus if caterpillars of Danaus plexippus feed on milkweed plants they cut the leaf veins before feeding distal to the cuts. This cutting of the veins blocks the flow of latex to the feeding sites and represents a counter-adaptation by the insect to the plant's defence. In more direct experiments with droplets of the latex from the plant Asclepias syriaca it was shown that they repelled feeding by the larvae of the generalist feeder Spodoptera er idania. Another defence system which has been studied more extensively in recent years is the glandular trichomes that are on the upper surface of the leaves of many plants.A technique has been developed for visualizing epidermal glandular struc- tures on plant leaves by pressing them against filter papers that have been soaked in Tollen's reagent when such structures appear as dark spots on a light ba~kgr0und.l~' Exploration of the glandular trichomes of Solanum berthaultii which are effective against aphids and similar insect pests has continued. One part of the defence involves the immobilization of the insects by an exudate which becomes brown and hard due to the interaction between chlorogenic acid and a phenolase both of which are present in the trichome but in separate compartments. Recent work on the phenolase has shown that it is unique to the trichome tissue and differs from the phenolases in other parts of the leaf.132 The sticky non-volatile material that is released from a second type of trichome on these leaves has now been characterized as a mixture of novel sucrose esters of which 3,4-di-O-isobutyryl-6-O-caprylsucrose (52) is a major component.133 These compounds appear to be multifunctional in defence in that they also provide resistance to infection by potato blight (Phytophthora infestans).Type B trichomes are common not only throughout the genus Solanum but also on leaves of other Solanaceous plants such as species of Datura Lycopersicon Nicotiana and Petunia. Chemical investigations here have shown that acylated glucose and sucrose esters with different organic acids linked variously at positions 1 2 3,4,and 6 are present in these trichomes.Two NATURAL PRODUCT REPORTS 1989 HO \ OH (56) R = H (57) R = 0 of the major components in the exudate of type B trichomes of Datura metel for example are 2,3-di-O-hexanoyl-P-~-glucose and 1 ,2,3-tri-0-134 hexano yl -P-~-glucose. Type B trichomes also contain volatile sesquiterpenes which in some cases are alarm pheromones for aphids. P-Caryo- phyllene and (a-p-farnesene previously reported to be present in trichomes of Solanum berthaultii have now been detected in the glandular trichomes of S. tuberosum. 135 Two unidentified sesquiterpenes have been detected in the trichomes of Lyco-persicon hirsutum where they co-occur with the previously known toxins tridecan-2-one and undecan-2-one.Both sesquiter- penes were acutely toxic to larvae of Spodoptera exigua and Keiferia lycopersicella. 136 Another class of toxins that are usually secreted in leaf trichomes is allergenic chemicals. Falcarinol (53) has been identified as the substance that is responsible for contact dermatitis caused by the leaves of the popular ornamental house-plant Scheflera arb~ricola.'~~ Falcarinol also causes dermatitis in several other plants of the Araliaceae notably the common ivy (Hedera helix).13* It is accompanied by a new allergen didehydrofalcarinol(54) in the latter ~1ant.l~' Allergic activity is quite specific to these structures since the closely related compounds falcarindiol dehydrofalcarinol falcarinol and falcarinone are all inactive.137,13* In the Hydrophyllaceae allergenic chemicals of the trichomes are usually isopentenyl- substituted quinols. Three novel derivatives (55)-(57) have been identified as the contact dermatitis agents of Phacelia campanularia while the known 2-geranyl-4,6-dihydroxyphenyl The acetate was found in P. pedi~ellata.~~~ activities of (59457) were comparable with that of pentadecylcatechol which is one of the allergens of poison ivy. 3.2 Induced Chemical Defence The phenomenon of induced chemical defence is now well established (cf. ref. 1). Insects feeding on the leaves of many but not all plants cause an increase in the synthesis of secondary products which means that the leaves become unpalatable to that insect within a few hours or days.While the trigger is mechanical damage it is not clear whether the signal that passes throughout the plant is a soluble one within the tissues or an airborne volatile. Evidence for an airborne signal has been obtained by Zering~e'~~ in the case of the cotton plant. Cotton leaves were exposed for seven days to volatile chemicals (including myrcene) originating either from mechani- cally damaged cotton leaves or from leaves that were infected with Aspergillusjiavus. In both cases the triggering of a 52 and a 34 YOincrease in phloroglucinol-derived gossypol derivatives occurred in undamaged leaves. Since gossypols are natural insecticidal materials being localized in subepidermal pigment glands on cotton leaves they could be responsible for the reduction in palatability of the leaves although this was not directly tested.lg1 In another characteristic induced response feeding of the mountain pine beetle (Dendroctonus ponderosae) on lodgepole pine (Pinus contorta) causes an increase in defensive mono- terpenes in the phloem surrounding the feeding site.Miller et al.lg2have now shown that equally elevated levels of mono- terpenes can be produced by treating tissue of the lodgepole pine with a blue-staining fungus (Ceratocystis clavigerum) with a pectic fraction (PIIF) from tomato leaf or with a fungal cell- wall fragment (chitosan) that is known to cause the synthesis of phytoalexins in plants. Relative concentrations of mono-terpenes in the tree (in milligrams per gram of dried phloem tissue) were 2.4 (control) 9.2 (beetle infestation) 7.4 (fungal infection) 8.4 (PIIF fraction) and 21.3 (chitosan).These results suggest the existence of a common recognition-defence mechanism in higher plants which is brought into play by either microbial infection or attack by insects. Inducible defence in soya bean (Glycine max) to the Mexican bean beetle (Epilachna varivestis) was found to be correlated with increased phenylalanine ammonia-lyase activity and increased biosynthesis of Again there is a close parallel with the molecular events which take place during the synthesis of phytoalexins in soya bean. However the analogy with antimicrobial defensive systems should not be pressed too far since induced defence against feeding by insects may have a different chemical basis.Further work on the chemistry of herbivory-induced defence in plants is urgently needed. 3.3 Hormonal Interactions Juvenile hormone I11 (58) has been isolated in large amounts (I 51 ,ug per gram of wet weight) in leaves of the sedge Cyperus iria accompanied by methyl (2E,6E)-farnesoate. 144 This is remarkable since although juvenile hormone analogues and anti-juvenile hormones (e.g.the precocenes) have been obtained from plants this is the first report of the actual insect hormone the commonest of five closely related structures occurring in plant tissues. It parallels the much earlier discovery of the insect moulting hormones a-and P-ecdysone from both plant and animal sources.* As with the occurrence of excessive amounts of ecdysteroids in plants as compared with the trace amounts that are found in insects the concentration of juvenile hormone I11 in C.iria is at least 150 times the largest amount that ever accumulates in insect tissues. Nymphs of the grasshopper Melanoplus sanguinipes that fed on Cyperus iria grew at a similar rate to controls that fed on wheat seedlings but they showed pronounced morphogenetic effects when they moulted to adults. If the ovaries were dissected from females that had fed on C. iria they were found to be markedly underdeveloped compared to those of normal females and the treated females laid no egg pods. These findings show that the accumulation of juvenile hormone I11 in this plant represents a novel defensive mechanism of plants different from the synthesis of juvenile hormone analogues which may well be effective against a wide range of insect species.lg4 The effects of anti-juvenile hormone agents both natural and synthetic on insects have recently been reviewed.lg5Precocene 11 when fed to the grasshopper Melanoplus sanguinipes was converted into the 2-hydroxymethyl derivative (59) and a demethylated product was also obtained. Related acetyl-chromenes were similarly oxidized to the more polar 2-hydroxymethyl analogues which were significantly less toxic than the parent compounds and more readily excreted.'16 NATURAL PRODUCT REPORTS 1989-5. B. HARBORNE OH 0 (611 (62) Table 3 Oviposition stimulants from plants for female butterflies Insect and host plant Black swallowtail (Papilio polyxenes) on carrot (Daucus carota) Citrus swallowtail (Papilio xuthus) on Citrus unshiu The swallowtail Papilio protenor on Citrus natsudaidai Pipevine swallowtail (Atrophaneura alcinous) on Aristolochia species Chemicals involved Reference Luteolin 7-U-(6-O-malonylglucoside) trans-chlorogenic acid organic bases 159 Vicenin-2 narirutin hesperidin rutin adenosine 5-hydroxy-N-methyltryptamine 157 158 bufotenine synephrine stachydrine (+)-1D-chiro-inositol Naringin hesperidin 160 Four aristolochic acids and sequoyitol 161 Insect moulting hormones continue to be reported from new plant sources.Thus ecdysterone was found in roots (0.62 % of the dry weight) and leaves (0.92%) of Pfafia iresinoides (Amaranthaceae) while polypodine B and pterosterone were also detected.14' 2-Deoxyecdysone 3-0-glucoside (blechnoside A) and 2-deoxyecdysone 25-0-glucoside (blechnoside B) were isolated from the fronds of the fern Blechnum minus.148 Four 14a-hydroxypinnasterols such as (60) were discovered in the red alga Laurencia pinnata; this is a rare example of moulting hormone activity from a marine alga.lg9 The metabolism of 20-hydroxyecdysone was investigated in Heliothis virescens an insect which is relatively impervious to phytoecdysones in its food.Four major non-polar metabolites were identified in the frass the 22-linoleate 22-palmitate 22- oleate and 22-stearate.While phosphates sulphates and glucosides have been variously recognized as conjugates in insects this is the first report of fatty acid esters being formed as a detoxification procedure.150 3.4 Insect Antifeedants Many plant terpenoids are antifeedant and the structures and activities of these natural pesticide agents have been re-viewed.151 The best known terpenoid antifeedant is azadi- rachtin from the Indian neem tree (Azadirachta indica) and its structure has finally been established as (61),152 after two incorrect formulations had been proposed earlier. X-Ray crystallography confirmed the correctness of formula (61).153 The mode of action of the sesquiterpene lactone tenulin (62) which is a major constituent of bitterweed (Helenium amarum) against phytophagous insects has been investigated.At 3 pmol g-' in artificial diets it reduced the growth and delayed the development of the larvae of the European corn borer (Ostrinia nubilalis) as well as being antifeedant. A pronounced carry-over effect was observed in the fecundity of adult moths emerging from treated larvae. The LD, of (62) in the migratory grasshopper Melanoplus sanguinipes was 0.88 pmol per insect. The toxicity of (62) was antagonized by its co-administration with cysteine suggesting that the cyclopentenone group of (62) undergoes Michael addition of biological nucleophiles in vivo. After testing a series of related structures it was found that only tenulin analogues that are capable of acting as electrophilic acceptors had significant antifeedant a~tivity.'~' From studying the effects of three alkaloids on the feeding of the red turnip beetle (Entomoscelis americana) on leaves it has been suggested that sensory inhibition is one mechanism of feeding deterrence.155 Sparteine nicotine and quinine stimulate a cell in the galeal sensilla of larvae and adults of the beetle which is also stimulated by glucosinolates. The alkaloids are deterrents although glucosinolates are stimulants to feeding. The alkaloids also inhibit the response of the sugar-sensitive cell the latter effect probably being responsible for feeding deterrence. While there may be special receptor cells which are adapted to deterrent compounds in insects the possibility exists that secondary compounds may also be deterrent generally because they interfere with the sucrose-stimulating cell.Much other recent work on the physiological basis of insect feeding behaviour is reviewed in a book dedicated to Vincent Dethier a pioneer in this field.156 3.5 Oviposition Stimulants As a result of recent investigations in both Japanese and American laboratories we have a reasonably complete picture of the chemical attractants which stimulate a female swallowtail butterfly specifically to lay its eggs on the leaf of a chosen host plant (see Table 3). In all four cases that were studied the attractants are non-volatile polar compounds located in the (63) OH HO 0 (641 Me OH (65) leaf vacuole but it is not entirely clear as yet how they are perceived by the butterfly.The most complex mixture of chemicals that is needed for total oviposition response is those in the leaf of Citrus unshiu where four flavonoids four organic bases a quaternary ammonium compound (63) and an inositol are inv01ved.~~~~ 158 The response is entirely a synergistic one since no single component in whatever amount stimulated oviposition. While some of the constituents are universally present in leaves (e.g. adenosine) others (e.g. the flavonoids) are more specific to Citrus species and certainly the combination of compounds each present at a particular concentration must be entirely specific to the host plant concerned. In the case of the black swallowtail (Papilio polyxene~),’~~ the response is mediated by only certain of the components of the leaf flavonoids.Thus the luteolin 7-0-(6-0-malonylglucoside) (64) from carrot leaf is stimulatory whereas the closely related deacylated compound luteolin 7-0-glucoside (also present in the leaf) is inactive. Similarly with the swallowtail Papilio protenor the female responds to the presence of flavanone structures but not to any of a number of other related flavonoid molecules.160 Swallowtails besides ovipositing (and feeding) on plants of the Umbelliferae and Rutaceae also oviposit in some instances on members of the Aristolochiaceae. Such insects as the pipevine swallowtail (Atrophaneura alcinous) not surpris- ingly respond to the characteristic nitro-compounds the aristolochic acids of these plants.16’ In the case of feeding on Aristolochia debilis the female is responsive to a mixture of four of the seven aristolochic acids in the leaf and in addition responds to sequoyitol.The same leaf also contains myo-inositol but this is completely inactive. Besides the presence of oviposition cues in plant leaves there may be oviposition deterrents. Thus female butterflies will rapidly learn to avoid plants which are likely to be deleterious to their offspring when the eggs hatch out and the larvae begin to feed. Some of the best examples of such deliberate avoidance are plants of the genera Cheiranthus and Erysimum of the Cruciferae which because of their rich content of glucosinolate attractants should be chosen for egglaying by the cabbage white butterfly (Pieris brassicae).Earlier work with this butterfly has clearly established that sinigrin is such a powerful oviposition stimulant that females will even lay their eggs on green blotting paper if it is previously soaked in a solution of this glucosinolate.16’ Both Cheiranthus and Erysimum are unusual among cruciferous genera in that their species contain NATURAL PRODUCT REPORTS 1989 additional leaf toxins in the form of cardiac glycosides. Rothschild et al. 163 have now demonstrated for the first time that such compounds can be natural contact deterrents for Lepidopterans. These have found that in the case of the Siberian wallflower (Cheiranthus x allionii),a strophanthidin glycoside is specifically responsible for oviposition deterrence to the cabbage white butterfly.This compound which is one of fourteen cardiac glycosides in the ‘deterrent’ fraction of the leaf not only occurs within the leaf but also on the leaf surface and it is clear from neural-response experiments that female butterflies can detect its presence. The toxic effects of the cardenolides on the larvae were demonstrated by transferring eggs onto the wallflower. The eggs hatched normally and the larvae fed on the leaf because of the mustard oil feeding stimulant that was present ; however all died before metamorphosis beyond the second instar. The chemical treatment of plants can also less specifically lead to deterrence in egglaying on otherwise acceptable leaves.Thus artificial spraying of cabbage plants with solutions of coumarin or rutin caused some deterrence in egglaying by Pieris r~pae.’~~ In this case though the chemicals deterred the females from alighting on the leaf and they are therefore not contact cues (as in the case of the cardiac glycosides). Some insects such as the European cherry fruit fly (Rhagoletis cerasi) deliberately mark fruit after egglaying in order to avoid accidentally laying two eggs in the same fruit. This ensures that the emerging larvae have sufficient food to go through to adulthood. The nature of the oviposition deterrent which is present in the faeces has been identified as a hydroxy- fatty acid (65) conjugated with glucose and taurine. It would thus seem to be a normal insect metabolite which has been adapted for practical use in this way as a chemical signal for preventing oviposition.165 A similar deterring pheromone is produced by the apple maggot fly (R. pomonella) and the compound remains active for three weeks under normal weather conditions.166 The last Report’ described an oviposition attractant of the mosquito Culex pipiens that had been identified as 6-acetoxy-5- hexadecanolide. Apical droplets of egg rafts contain this active attractant which induces further oviposition of this and other species of Culex. More recent work has shown that only the (-)-(SR,6S)-isomer is active (at dosages of OSpg per 100 ~m-~ of water) to Culex quinquefasciatus. This isomer was also active towards C. tarsalis at higher doses but it had no effect on Aedes aegypti or Anopheles quadrimaculatus and hence appears to be genus-~pecific.~~~ 3.6 Pollination Interactions Floral fragrances are important in attracting insects especially beetles bees and moths for pollination.These fragrances have been little studied because of the analytical difficulties. Pellmyr .~~~ et ~ 1have demonstrated a new method of determining floral volatiles by measuring both floral and vegetative fragrances and then subtracting the latter from the former to produce a ‘true’ picture of the floral essence. When they applied this method to four species of Actaea (Ranunculaceae) these authors showed that all four contain from ten to fifteen geraniol- and nerol-based monoterpenes. Species differences in odour are primarily due to quantitative variations and pollinators modify their behaviour according to the particular terpenes that dominate the scent.In Cimicijiuga simplex (of the same plant Family) two of the floral substances -methyl anthranilate and isoeugenol -synergize to attract pollinating butterflies. Other forms of this plant species which are pollinated by bumble bees lack these two attractants and have mainly terpenoid volatiles. In the genus Ophrys of the Orchidaceae chemical mimetism occurs in the volatiles of flowers since identical substances can be found in the cephalic secretions of their pollinat~rs.’’~ For NATURAL PRODUCT REPORTS 1989-5. B. HARBORNE (66) example linalool citronellol citronellal geraniol geranial and (E,E)-farnesol occur both in the scents of Ophrys species and in the secretions of female bees of the genus Andrena.Hence male bees are attracted to pseudocopulate with and hence pollinate the flowers of these orchids. Analyses of the volatiles of Ophrys lutea and 0.fusca by gas chromato- graphy-mass spectrometry have led to the identification of no less than 184 components 138 being aliphatic thirteen aromatic and the remainder mono-or sesqui-terpene~.'~' Among the compounds that are most attractive to male bees are octan- 1-01 nonan- 1-01 geraniol citral and farnesol. The wild orchid 0. lutea is able to attract the males of more species of Andrena than most other species of Ophrys; this is due to the wide range of volatiles released from the Pollination by Andrena fuscipes is due to the fact that aliphatic hydrocarbons together with 6-methylhept-5-en-2-one, gera-niol nerol citral and (E,E)-farnesol are present in the flower volatiles and the cephalic secretions of the bees.Unlike the situation in Ophrys the pollinators of Cypripedium species include female bees which are attracted to the flowers by deception through visual morphological and chemical cues. Chemically there are also differences from Ophrys since each species of Cypripedium appears to have its own character- istic volatiles. Thus C. calceolus contains straight-chain acetates (e.g. octyl and decyl acetates) the volatiles of C. parvzjlorum include terpenes (e.g. cis-and trans-p-ocimene) and C.pu-bescens releases aromatics (e.g. 1,3,5-trimethoxybenzene).173 Pollination by bees occurs with members of the Orchidaceae from both the Old World (e.g. Ophrys and Cypripedium) and the New World but with the American plants the male bees are attracted to the flower volatiles because they collect them and use them in lek formation to attract females for mating.' An estimated 650 species of Neotropical orchids are pollinated solely by male Euglossine bees and almost every species has its own distinctive fragrance based on either terpenoids or Examination of the volatiles of nine species of Catuseturn showed up to fourteen compounds to be present but only one component carvone oxide (66) which was a major constituent (up to 74% in C. maculatum) was attractive to the Eulaema bee pollinators and collected by them.175 Growing in the same habitat is a plant of the Euphorbiaceae Dalechampsia spathulata which proved to be pollinated by the same bees; on analysis of its scent it was found to contain 62 % of the total volatiles as carvone oxide.This appears to be a very rare example of convergent evolution in which a euphor-biaceous plant has mimicked the chemical lure of an orchid in order to ensure its pollination by the same bee. Orchids are less commonly pollinated by moths and in Platanthera chzorantha it has been found that the odours that are produced nocturnally when the moths are active differ from those emitted during the day.176 Methyl benzoate and some monoterpenes dominate the nocturnal secretions and these seem to be necessary to elicit the search behaviour of the appropriate species of moth.The floral fragrances of species of the moth-pollinated genera Zypgynum and Exospermum (Winteraceae) are based on isoprenoids and short-chain a1iphati~s.l~~ Ethyl acetate is present in large amounts and contributes to the more effective pollination of the flowers by making moths drowsy. By contrast pollination by moths in Datura innoxia appears to be mediated by tropane alkaloids in the floral nectar. These narcotize the hawkmoth Manduca quinquemaculata which gets hooked on this species and returns again and again to obtain another ' fix I.178 Pollens from different plants have characteristic odours that can be distinct from those of the whole flowers as perceived by both human noses and insect antennae.The oily coating of the pollen grains is the main source of volatiles and this may be an important attractant to pollen-foraging honeybees which can discriminate between pollen odours. A comparison between Rosa rugosa and R. canina has now shown that pollen odours are indeed chemically different from those of whole fl0~ers.l~~ Of 31 compounds that are present in R. rugosa nine are restricted to the flowers and ten to the pollen. Specific flower compounds include benzyl alcohol pentan- 1-01 and citronellol while the pollen-specific volatiles are geranylacetone geranial undecan-2-one and tetradecanal. Whereas the pollen of R. rugosa contains 22 volatile components that of R.canina has only five. Flower colour (and hence floral pigments) have long been recognized as being important in the attraction of pollinators. Recent experiments with the blue-flowered larkspur (Del-phinium nelsonii) and its white mutant forms have emphasized the role of anthocyanin pigmentation in attracting an optimally foraging insect such as a bee.laO The ability of plants to respond rapidly to changes in availability of their pollinator has recently been illustrated by some observations on the scarlet gilia (Ipomopsis aggregata) which is pigmented by pelargonidin glycoside.ls' In populations growing near Flagstaff Arizona it has been observed that in a minority of plants there is a shift in flower colour during the season from red through pink to white.'*' The shift is correlated precisely with the coincident southern migration of humming birds which are the primary pollinators in mid-July and with the need to be attractive to the remaining pollinator a hawkmoth (Hyles lineata).Colour shifting involves decreasing the amount of anthocyanin that is formed in the petal and eventually completely turning off the synthesis of anthocyanins. Another plant species which can attract more than one pollinator through its floral colours is Pedicularis densflora (Scrophulariaceae).Ia3 Here the flower has scarlet corollas to attract hummingbirds and magenta calyces and bracts with highly ultraviolet-reflective hairs to attract bumble bees. While younger flowers are pollinated by bees the older flowers are visited by birds.Instead of a shift in colour per se there is an increase in nectar sugar (from 15 to 25 %) as the flower ages which correlates with the nutritional requirements of the changing pollinator. 4 Plant-Plant Interactions 4.1 Allelopathy A book on the science of allelopathy has been published based on the proceedings of a There is a useful section on techniques but it is clear from this book (and from much else that is written on allelopathy) that there are unsolved problems in the bioassays that are frequently employed for monitoring the inhibitory effects of allelochemicals on seed germination. Weidenhamer et alla5 have shown that the volume of a solution of allelochemical and the number of seeds to which it is applied can considerably alter the results obtained.Lower phytotoxic concentrations of juglone or ferulic acid for example can produce equivalent or greater inhibitory effects than higher concentrations when the amount available per seed for uptake is greater. This means that in the field allelopathic effects may be more pronounced in situations where overall plant densities are low because of environmental or other constraints. Decreased growth of plant when their density decreases is the opposite of what usually occurs and the demonstration of such effects strongly implies that allelopathic substances are present in the soil. NATURAL PRODUCT REPORTS 1989 CH2 II ’0 (67) OH H (69) R = H (70) (71) R = Glc I I ( 72) (73) The chemistry of allelopathy has been reviewed186 and a number of new allelochemicals have been described.The sesquiterpene lactone cnicin (67) has been identified as an allelopathic agent in the knapweed Centaurea maculosa and is responsible for the inhibition of growth in competitor species caused by this plant.ls7 Similarly tricin (68) has been reported as the allelopathic chemical of quackgrass (Agvopyron vepens) being released from the herbage and having an adverse effect on root growth of its competitors.18* Likewise heptan-2-one and ( +)-heptan-2-01 are the allelopathic volatiles of Amaranthus palmeri. Vapours of these two compounds at concentrations of 1 p.p.m. inhibit the germination of onion and carrot seeds and are autotoxic to amaranth seeds.ls9 Finally the benzoxazinones (69) and (70) have been recognized as the allelopathic agents of winter rye (Secafe cereale).They are released when tissue is damaged from a bound glycosidic form (71). Compounds (69) and (70) have no effect on the germination of seeds of other species but inhibit their root growth at concentrations of 0.37 and 1.05 mmol dm-3 respectively. lg03 lgl A concerted effort has been made to study the allelopathic effects of three shrubs that grow in the Florida scrub community on competing sandhill grasses.192 Aqueous washes of the leaves of Ceratiola ericoides (Empetraceae) contained the triterpenoid ursolic acid (72) and the dihydrochalcone ceratiolin (73); the latter is not itself phytotoxic but it spontaneously decomposes in water to the phytotoxic compound dihydrocinnamic acid.In a second shrub of these communities Conradina canescens (Labiatae) phytotoxic effects were due to four monoterpenoids camphor myrtenol borneol and carvone. In a third plant Calamintha ashei (also Labiatae) the most active ingredients were again monoterpenes (menthofuran epievodone and OH OH (74) 0 11 0 (75) calaminthone) but these were again accompanied by ursolic acid. This triterpenoid is a natural surfactant and seems to play a supporting role in allelopathy enhancing the solubilization of lower terpenoids via micellization and also assisting in their transport to target seeds or seedlings. A new autotoxic agent that has been characterized is caffeine the purine alkaloid of coffee.It has been detected in soil and leaf litter in coffee plantations and exerts a deleterious effect on new plantings probably because it is so slowly turned over in the Caffeine besides being autotoxic in coffee will inhibit germination of the seeds of many weedy species such as Amaranthus speciosus. On the other hand it has no effect on the crop plant Vigna mungo so could be useful as a selective herbicide in such a crop.1s4 4.2 Host-Parasite Interactions There are more than 3000 species of higher plants which have the ability to form specialized intrusive organs called haustoria by which they are able to attach themselves to other higher plants and feed from them.lg5 They are thus parasitic on their host plant depending on the host either completely or partly for their nutritional needs.For successful parasitization two distinct chemical messages need to be passed out from the roots of the host plant one to stimulate the germination of the seeds of the parasite and the other to induce the formation of haustoria in the newly germinated seed so that it can attach itself to the host. Chemical studies of these germination stimulants and host- recognition substances are still at an early stage but as a result of recent work we know that typical secondary substances in root exudates are responsible for these signals. The first germination stimulant to be characterized was the sesquiterpene strigol but this was artificial in the sense that it was obtained from root exudates of cotton which is not a natural host of Striga asiatica.The first natural stimulant for seeds of Striga species has been isolated from the roots of Sorghum bi~olorl~~ and it turns out to be quite different from strigol in fact a simple p-diphenol (74) with an unsaturated aliphatic side-chain. It is easily oxidized to the corresponding p-quinone (79 which is inactive. This compound is an ideal messenger from the point of view of Striga species. It will be in the active quinol form as it emerges from the sorghum root but as it moves further and further away it will gradually be oxidized by the oxygen in the soil and become inactive. Hence only seed that is near enough to the sorghum root for a root of the parasite to be able to latch onto it will be triggered into germination.The second message of the Striga-Sorghum interaction -the haustorial inducer -has also been identified and turns out to be a simple quinone namely 2,6-dimethoxy-p-benzoquinone. lg7 This substance appears to be released from the surface of the root of the host by enzymic digestion and then triggers off the formation of haustoria in the parasite. NATURAL PRODUCT REPORTS 1989-5. B. HARBORNE W M eO ( 79) (80) Other haustoria-inducing compounds have been described from two leguminous plants. Two were isolated from the root exudate of tragacanth (Astragalus species) which stimulated the formation of haustoria in the parasite Agalinis purpurea.They are a simple dihydrostilbene xenognisin A (76) and a related isoflavone xenognisin B (77). A third was isolated from the root of Lespedeza sericea which is another leguminous host of Agalinis species and identified as a triterpene soyasapogenol A (78).lg7 Hence several different classes of chemical apparently have the ability to trigger recognition of the host in this parasite. An unusual host-parasite interaction has been encountered where the grass Imperata cylindrica partially parasitizes the bulbs of Pancratium bzjlorum (Amaryllidaceae). A red necrotic zone develops in the bulb around the point of attachment and chemical analysislg8 indicates that the host produces new phenolic compounds -a flavylium derivative (79) and a chalcone (80)-in response to parasitic attachment.There are also some alterations in alkaloid metabolism since lycorine occurs bound as the 1-glucoside and 1-palmitylglucoside in healthy bulb tissue whereas free lycorine occurs in the necrotic tissue. 5 Plant-Microbe Interactions 5.1 Mycotoxins Ryegrass staggers is a nervous disease of sheep and cattle; it is prevalent anywhere that Lolium perenne is a dominant pasture species. The tremorgenic agents involved are the mycotoxins lolitrem B (81) and related structures. These potent neurotoxins are produced in L.perenne if an endophytic fungus Acremonium loliae infects the grass. Eradication of this fungus from ryegrass would therefore seem to be agriculturally desirable since it would remove a potential threat to browsing livestock.Such an operation however is not without problems because the presence of the fungus not only increases the vigour of the grass but also protects it against insect pests such as the Argentinian stem weevil (Listronotus bonariensis). Rowan et al. have now demonstrated that the resistance of infected ryegrass to attack by insects is due to the synthesis by the endophyte of a novel alkaloid peramine (82) which is an antifeedant. lg9 Peramine has also been detected in the mycelium of A. loliae CHMe2 CHMeEt 0 -CH-C4O)NMeCHC(O) ( 84) H2N \ HOH2C 0 0 (85) so it is fairly clearly of fungal origin. Biosynthetically it could be formed by the cyclization of proline with arginine but this has yet to be established.This alkaloid is clearly biosynthetically distinctive from the neurotoxin of ryegrass i.e. lolitrem B the origin of which has recently been explored by Weedon and Mantle. 2oo These authors isolated a related indole isoprenoid paxilline (83),from submerged cultures of A. loliae and from infected ryegrass and concluded that it was a likely precursor of (81). Another example of a mycotoxin-producing fungus con- ferring resistance to insects on a host plant appears to be that of Fusarium avenaceum on the foliage of balsam fir (Abies balsamea). At least this would explain the unexpected collapse of infestations of the spruce budworm (Choristoneura fumiferana) on this tree.201 Thus the cyclohexadepsipeptide enniatin A (84)from I;. avenaceum has been found in foliage of balsam fir and the pure compound is toxic to the budworm larvae at a concentration of 0.04%.The spread of a cereal pathogen to a gymnosperm is another remarkable feature of this complex interaction. The Fusarium species that infect cereals are known to produce a range of mycotoxins besides the cyclodepsipeptide (84).One further example has come to light in Fusarium roseum namely fusarochromanone (85) which has been obtained from infected overwintering oats in Alaska.202 It is toxic to poultry and reduces the hatchability of fertile hens' eggs. The ergot alkaloids produced by the ergot fungus Claviceps purpurea infecting rye are one of the oldest known groups of NATURAL PRODUCT REPORTS 1989 (89) R = H (90) R = OMe Table 4 Variation in the nodulating signal in different legume-Rhizobium symbioses Symbiosis Inducing flavonoids Reference R.trifolii on clover 4,7-dihydroxyflavone 21 1 4’-hydroxy-7-methoxyflavone 4’,7-dihydroxy-3’-methoxyflavone(geraldone) R. meliloti on lucerne 3’,4’,5,7-tetrahydroxyflavone(luteolin) 210 R. leguminosarum on pea 3’,4’,5,7-tetrahydroxyflavanone(eriodictyol) 209 4’,5,7-trihydroxyflavone7-glucoside mycotoxins and have been found for the first time in tall fescue grass (Festuca ar~ndinacea).~~~ The fungal endophyte is Acre- monium coenophialum and its infection on tall fescue causes several different toxic symptoms in grazing cattle one being ‘fescue foot’ (a gangrene of the cattle’s hooves). Ergovaline was the principal ergot alkaloid in infected fescue the total alkaloid content of which varied between 1.5 and 14 mg per kilogram of plant tissue.The fungal origin of the ergot alkaloids in tall fescue is apparent since the same compounds are produced by A. coenophialum in pure culture. Finally it is worth reporting that the mycotoxin that is responsible for the disease lupinosis (a liver disease of sheep and cattle) has been characterized as (86). This substance phomopsin A (a product of the fungus Phomopsis lepto- strorniformis which infects lupin fodder species) is an unusual partly cyclic he~apeptide.~~~ 5.2 Signal Molecules In many plant-microbe interactions there is a stage at which the micro-organism detects a susceptible host plant through the release from that host of a compound (or compounds) which acts as a chemical signal.Such a signal produces an immediate response in the microbe which usually involves the expression of the genes that are necessary for establishing an infection. The chemical signals of both the crown gall infection of dicotyledo- nous plants and of the even more familiar legume-Rhizobium symbiosis have now been characterized. Stachel et aL205 were the first to report that the acetosyringone (87) and a-hydroxyacetosyringone (88) that are released from wounded cells of tobacco root are capable of activating the transfer of T-DNA in Agrobacterium tumefaciens which then leads to establishing the symptoms of crown gall disease in the host. The synthesis of these two signal molecules is stimulated by the deliberate wounding of plant tissue and this agrees with earlier observations that wounded cells are more susceptible than undamaged cells to this bacterial infection.Compounds (87) and (88) have not been reported before as natural products and they would seem to be specific to this interaction although some lignin precursors (such as sinapic acid) are also moderately active. In a more extensive study of structure-activity relationships Spencer and Towers206 have shown that coniferyl alcohol (89) and sinapyl alcohol (90) are as active as acetosyringone in triggering the transfer of T-DNA and that a range of related substances containing either guaiacyl or syringyl moieties have some activity. They propose that Agrobacterium tume- faciens may be capable of detecting plant cells which are undergoing synthesis of lignins or repair of cell walls and that several types of lignin precursor may provide the signal for attack.Acetosyringone has only so far been positively identified in the A. tumefaciens-tobacco interaction but it has also been shown to be an efficient signal for the transformation of Arabidopsis thaliana by Agrobacterium tumefaciens.207 Unfortunately chemical studies of the trigger have not yet shown why it is that crown gall disease is restricted to dicotyledons and fails in monocotyledons. Roots of monocots are known to be as capable as those of dicots in releasing Iignin precursors that are based at least on guaiacyl residues.20s An even more intriguing signal system has been uncovered in the legume-Rhizobium symbiosis since it has been shown that secondary compounds of the flavonoid class can both stimulate and repress the expression of the nodulating (nod) genes of Rhizobium specie^.^^^-^" Most attention has been given to the switching-on process and several common flavones [e.g.luteolin (91)] and a flavanone have been identified (see Table 4) in the root exudates of leguminous plants which are capable of inducing the expression of nod genes at low (10 nmol dm-3) concentrations. Different structures are involved in the three symbioses that have been investigated (see Table 4) and this ties in precisely with the species-specific nature of most legume-Rhizobium interactions.There is an earlier stage of the interaction concerned in recognition which is probably mediated by plant lectins. The signal compounds are with one exception flavone or flavanone aglycons and this seems to be important for stimulatory activity since most of the flavonoid glycosides that were tested in the R. leguminosarum-pea system were in-active.209 The concentration of flavone that is released from the root might even determine the ability of that plant to fix nitrogen most efficiently. Indeed Kapulnik et aL212 have found a strain of alfalfa designated HP32 which fixed more nitrogen than the parental strain HP and which had 77% more luteolin in the roots. Equally interesting is the finding that in R. leguminosarum on pea other flavonoids notably the isoflavones daidzein (92) and genistein (93) are capable of inhibiting the expression of nodulating genes.The concentration of these isoflavones that is NATURAL PRODUCT REPORTS 1989-5. B. HARBORNE 103 R H HO-$-OH I (91) R = OH (92) R = H (94) R = H (93) R = OH required for inhibition is however two orders of magnitude higher than that required for gene induction. In the symbiosis of R. trifolii on clover the two inhibitors umbelliferone (7-hydroxycoumarin) and formononetin (4'-methyldaidzein) have been found in root exudates. A ten-fold excess of umbelliferone to 4,7-dihydroxyflavone (94) which is the principal root stimulant resulted in complete inhibition of expression of nod genes. Competition for a common binding site at the bacterial surface would appear to be taking place.213 These results indicate that flavonoids and related phenolics can regulate the expression of these genes in Rhizobium species by switching them off as well as on.This may be important to the host plant since initiation of too many nodules on the root could be an excessive drain on the nutritional resources of the legume and just as damaging as insufficient nodulation. 5.3 Phytotoxins The chemistry and biological function of host-specific phyto- toxins have been reviewed by Uen~;~l~ there is particularly good coverage in this review of the complex mixtures of the so-called AK and AM toxins that are produced by Alternaria kikuchiuna and A. mali. The highlight of recent years is the elucidation of the structure of the elusive victorins (the major host-selective toxins of Cochliobolus victoriae) that were first described some thirty years ago.Victorin C (95) is a cyclic peptide based on the six subunits glyoxylic acid 3-hydroxy- leucine 5,5-dichloroleucine 3-hydroxylysine dehydrochloro- alanine and victalanine (which is a novel pentacyclic analogue of phenylalanine).215 Three other victorins (B D and E) have also been characterized from this fungus and they all have very closely similar The phytotoxins of Phyllosticta maydis which causes yellow leaf blight disease of Zea mays are also host-selective in their action. Each of the four main components that have been characterized as long-carbon-chain polyketols shows toxic effects at to mol dm-3 in susceptible varieties of maize which have Texas-male-sterile cytoplasm but none is toxic to resistant varieties even at concentrations of mol dmP3.In order to explore structure-activity relationships twelve mimics of PM-toxin A (96) have been synthesized.216 The mimics have four P-ketol groups spaced by varying lengths of methylene chains or by a 1,3-diene chain. Mimics with shorter methylene side-spacers or with diene side-spacers are 30- to 300-fold less toxic than (96) but the remaining analogues are equally toxic. It appears therefore that intramolecular associations at the P-ke$ol groups may yield a circular cage structure with distinguishable hydrophilic and hydrophobic domains. Such formation of channels may be involved in the mode of action of these toxins.Turning to non-host-specific toxins it appears that 3-(methy1thio)propionic acid (97) must be the simplest of all known phytotoxins. This compound is produced by the cassava pathogen Xanthomonas campestris pv. manihotis and is effective at levels of 6,ug per gram of fresh weight in producing characteristic symptoms of bacterial blight in artificially infected leave^.^" The phytotoxin hymatoxin A from Hypoxylon mammatum which is a parasite of aspen (Populus tremuloides) is notable for having a sulphate substituent at one of the hydroxyl groups of its diterpene structure (98).218 The two phytotoxic butenolides (99) and (100) of Seiridium cardinale (a (95) (96) MeSCH2CH 2 CO2H (97) Ho++rcHo (101) OH (102) (103) pathogen of Cupressus sempervirens) are unusual in their 3,4- disubstitution; the only other report of such butenolides in fungi is from Hypoxylon serpenx219 A fruitful new source of interesting phytotoxins has been the fungal pathogens of wild grass species.These substances have obvious potential as herbicides for controlling these grasses where they become serious weeds. Bipolaroxin (101) has been identified from Bipolaris cynodontis which is a fungal pathogen of the notorious weed bermuda grass (Cynodon ductylon).220 Again exserohilone (102) is a novel toxin of Exserohilum holmii this being a fungal pathogen of crowfoot grass (Dactyloctenium aegyptium). This substance produces necrotic lesions with a reddish-brown border on the leaves of several plant species when applied at concentrations of between lo-" and mol dm-3.221 More unexpectedly the simple indole derivative tryptophol (103) is a phytotoxin of Drechslera (104) OH COz H (105) COZH OH (106) OH COzH (107) OH (108) COzH OH (1091 OH COZH (1 10) nodulosum which is a pathogen of goose grass (Eleusine indica).Tryptophol occurs in infected leaves of this grass at a concentration of 3.5pg per 0.45 gram of dry weight and causes necrotic lesions when injected at a concentration of 6.2 x mol dm-3.222 The mode of action of most phytotoxins in damaging the host plants is still not clear but it appears that Stemphylium botryosum f.sp. lycopersici probably produces its virulent effects on tomato plants by scavenging the iron from living tissue via the production of siderophores. Evidence that this is so has been obtained by showing that S. botryosum is capable of producing siderophores under both iron-sufficient and iron- deficient conditions. 2239224 In iron-sufficient conditions two novel phytotoxic chelates of iron(m) are formed stemphyl- oxins I and 11. Under conditions of iron deficiency the cultures synthesize three different siderophores :these are dimerum acid coprogen B and a monohydroxamate derivative. Stemphylium botryosum var. lactucum [a related pathogen on lettuce (Lactuca sativa)] wreaks its damage in this host plant via more typical phytotoxins which do not apparently chelate metals.One stemphylin which was characterized in 1975 as a chromone glucoside has now been shown to be identical to altersolanol A (104).225 The two known anthraquinones dactylariol and macrosporin have been obtained from this fungus as well as the two further phytotoxins stemphyltoxin and stemphyperylenol. 226 The possibility that phytotoxins of pathogenic fungi have other effects than damaging the host plant was hinted at in the last Report.' Thus they may protect an infected plant from further fungal infection. This happens in the case of Epichloe NATURAL PRODUCT REPORTS 1989 typhina [a pathogen of timothy grass (Phleum pratense)]. It has already been shown that E. typhina synthesizes three fungitoxic sesquiterpenes that prevent further infection by Cladosporium phlei which is another potential pathogen of timothy grass.' Four further fungitoxic metabolites the unsaturated hydroxy- fatty acids (1 05)+ 108) have now been characterized.227 Toxins that are produced by one fungus may well be directly deleterious to a second and it has been known for some time that Pythium ultimum which is a major root pathogen of many temperate crops can be controlled by introducing Laetisaria arvalis into the soil. This latter basidiomycete secretes an allelopathic agent that induces rapid lysis of hyphae in Pythium species and in similar pathogens. The active compound has been identified by Bowers et as laetisaric acid (8-hydroxylinoleic acid) (109) ;the term 'thallophytic allelopathy ' has been coined for this type of microbial interaction.It is interesting incidentally that Pythium ultimum itself actually produces an unsaturated fatty acid (1 10) as a phytotoxin when causing the black root disease of sugar beet.229 5.4 Antimicrobial Agents As a consequence of the co-evolution of plants with their microbial parasites plants have elaborated a complex series of defensive barriers that is capable of providing them with resistance to diseases. Besides the many classes of low-molecular-weight antimicrobial agent that may be present on plant surfaces or in their epidermal layers there are also macromolecular barriers. The latest type of antifungal sub- stance to be recognized in plants is lectins that have chitinase Defensive systems may also be induced by microbial invasion the best known of these being the phytoalexin response.However increased lignification synthesis of glyco- proteins or production of extensin may also occur in response to infection. In this Report constitutive antimicrobial com- pounds will be discussed before a consideration of some of the more recent studies of induced defence. The chemistry of antifungal compounds and of phytoalexins has recently been reviewed 231,232 and much has been written in the symposium proceedings edited by Bailey233 and by L~gtenberg,~~~ on elicitation of phytoalexins. A mixture of 5-substituted resorcinols the two principal components being (1 11) and (1 12) occur in the peel of unripe fruits of the mango (Mangifera indica).There is good circumstantial evidence that they are responsible for the latency of black spot disease (Alternaria alternata) in this fruit. Thus the disease only develops after the fruit has ripened and precisely when these antifungal agents have disappeared from the The resorcinols are only present in trace amounts in the flesh but when unripe fruit is peeled there is a massive build-up of antifungal activity and the peeled fruit becomes resistant. At the same time there is an increase in quantity of the resorcinols in the fruit from 30 ,ug per gram of fresh weight to a value of 160,ug per gram of fresh weight after four Bound toxins within the plant provide a second line of defence to surface toxins and many secondary compounds have been implicated in this role.' Further examples continue to be reported.Two antifungal saponins camellidins I and 11 from Camellia japonica have been characterized as the tetraglycosides (113) and (114). At concentrations of 30 to 100 p.p.m. they interfere with the germination of fungal conidia by causing them to swell. 237 Two new antifungal naphthoxirenes (1 15) and (1 16) and their monoglucosides have been found in the root bark of Sesamum angolense. Fungitoxicity in this series is related to the presence of free hydroxyl groups in a peri-position to a carboxyl Antimicrobial activity has rarely been identified in floral tissues of higher plants so it is notable that Nagai and Tada239 have found the antibacterial agents chinesin I (1 17) and chinesin I1 (1 18) in flowers of Hypericum chinense.A structural resemblance to the bitter hop principles humulone and lupulone is apparent here. 105 NATURAL PRODUCT REPORTS 1989-5. B. HARBORNE Ho (111) Ho (112) Glc Gal (113) R = AC (114) R = H R (115) R = 0 1116) R = H,OH (118) R z Me (119) (120) R’ = OMe R2= S (121) R’ = H R2 =S (122) R’ = OMe R2=O H (123) (124) OH as Phytophthora infestans. Likewise Wippich and Wink241 have shown that quinolizidine alkaloids and gramine are harmful to the development of powdery mildew (Erysiphe graminis) on barley because they inhibit the germination of conidia of the fungus at concentrations of between 1 and 5mmol dm-3. The presence of tannins in plants has also long been associated with the idea of resistance to disease since phenolic compounds are not only antimicrobial in their own right but become even more toxic as a result of oxidation to quinonoids.The pathological evidence for such a role has been limited. However Nicholson et al.242have implicated tannins as resistant factors through the discovery that some pathogenic fungi may have developed a protective response to their presence in plant tissues. Thus the fungus (Colletotrichum graminicola) that causes anthracnose on sorghum produces its spores in a water-soluble mucilage. A glycoprotein fraction within this mucilage was shown to have an exceptionally high affinity for condensed tannins thus protecting the spores from inhibition of germination.What percentage of fungi have such a protective device has yet to be determined but it is probably absent from the unspecialized moulds such as Fusarium species which attack sorghum grains. Varieties of sorghum that are resistant to grain mould have been shown to have elevated levels of flavan-4-01s and proanthocyanidins when compared to susceptible cultivar~.~~~ Interest in the phytoalexins (anti fungal agents formed de novo following infection) continues unabated. A series of five novel sulphur-containing indole derivatives (1 19)-(123) have been characterized as phytoalexins in Brassica campestri~~~~ and Raphanus sativus. 245 These substances are clearly related to the constitutive mustard oil glycoside glucobrassicin of the Cruciferae but their biosynthetic origin has yet to be established.Anthraquinones have been reported as phytoalexins for the first time in Cinchona species,246caffeic acid esters for the first time in Rehmanniu glutinosu (Scr~phulariaceae),~~’ and bi- phenyls for the first time in Cercidiphyllum japonicum (Cercidiphylla~eae).~~~ Furanocoumarins which are charac- teristic phytoalexins of members of the Umbelliferae have been detected in Citrus species a an thy let in)^^^ and in Helianthus species (xanthoto~in).~~~ The spiroketal enol ether (124) has been obtained from the infected leaves and stems of Coleo-stephus my~onis.~~~ dihydroflavonols including (1 25) Four have been reported as phytoalexins of the legume Shuteria vestita.252 Whether or not plant alkaloids provide disease resistance in the angiosperms has been somewhat debatable.Roddi~k~~O considers the evidence in the case of the steroidal alkaloids of the Solanaceae to be more in favour than against. He indicates that because solanine and tomatine can destabilize membranes they are likely to be highly damaging to potato pathogens such Benzoic acid accumulates in leaf tissues of Pinus radiata in response to damage by the phytotoxin dothistromin which is the product of Dothistroma At 10 p.p.m. benzoic acid causes 90% inhibition of the fungus so it would appear to be a true phytoalexin. Interestingly this response in the needles differs from that in the sapwood where the stilbenes pinosylvin and its monomethyl ether are formed following fungal attack.254 Four linear furanocoumarins have been detected in healthy tissues of parsley (Petroselinum crispurn) so that although there is massive synthesis of these compounds in cultured cells following fungal elicitation there is some question as to whether they should be termed phytoalexins.Other putative defensive agents such as the hydrolytic enzymes 1,3-p-glucanase [glucan 1,3-P-glucosidase] and chitinase were also present in considerable amount in the cotyledons of healthy parsley.255 Pathogenesis-related (PR) proteins are produced in plants when they react hypersensitively to infection by viruses and other pathogens. They could be involved in resistance to diseases but their precise role is still unclear.Usually the proteins are present in small amounts in uninfected leaves and there is a dramatic increase after suitable induction. If tobacco leaves are treated with tobacco mosaic virus they produce some thirteen separable proteins; in the case of one of them there is a 20000-fold increase in c~ncentration.~~~ One protein of M 33000 has 1,3-p-glucanase activity; another of M 27000 has chitinase activity; while a third of M 13000 has an unknown 106 function but an amino-acid sequence similar to that of the sweet protein tha~matin.~~’-~~~ The discovery of chitinase activity ties in with the recent finding that plant chitinases are potent inhibitors of fungal growth.260 The synthesis of PR proteins in response to microbial infection has an analogy in the production of heat-shock (HS) proteins when plants are subjected to high temperatures.That there is an association between the two is apparent in the recent work of Stermer and Hammerschmidt.261 They were able to produce disease resistance in cucumber (Cucumis sativa) to CZadosporium cucumerinum simply by subjecting seedlings to heat shock at 50 “C for 40 seconds. Resistance develops within 15 to 21 hours of the heat shock and appears to be due to changes in the cell wall and particularly to an increased synthesis of the cell-wall protein extensin. Finally it is worth noting that it is possible to inhibit fire blight (Erwinia amylovora) in susceptible apple plants by an equally simple method by injecting the leaves with D-galactose or L-fucose.These sugars somehow inhibit the bacteria from forming normal cell walls at the interface between host and pathogen and this is sufficient to stop the disease.262 6 References 1 J. B. Harborne Nut. Prod. Rep. 1986 3 323. 2 J. B. Harborne ‘Introduction to Ecological Biochemistry’ Aca- demic Press London 3rd edn. 1988. 3 D. Schlee ‘Okologische Biochemie ’ Springer Berlin 1986. 4 ‘Insects and the Plant Surface’ ed. B. Juniper and R. Southwood Edward Arnold London 1986. 5 ‘Natural Resistance of Plants to Pests’ (A.C.S. Symposium Series No. 296) ed. M. B. Green and P. A. Hedin American Chemical Society Washington D.C. 1986. 6 ‘Insect-Plant Interactions’ ed. J. R. Miller and T. A. Miller Springer Berlin 1986. 7 ‘Insects-Plants’ ed.V. Labeyrie G. Fabres and D. Lachaise Dr W. Junk Dordrecht 1987. 8 ‘Marine Toxins and Venoms’ ed. A. T. Tu Marcel Dekker New York 1988. 9 M. Rothschild in ‘Insect/Plant Relationships’ ed. H. F. van Emden John Wiley Chichester 1973 pp. 59-83. 10 S. S. Duffey Annu. Rev. Entomol. 1980 25 447. 11 A. Franke H. Rimpler and D. Schneider Phytochemistry 1987 26 103. 12 M. D. Bowers and G. M. Puttick J. Chem. Ecol. 1986 12 169. 13 R. Nishida and H. Fukami J. Chem. Ecol. 1988 in press. 14 M. Rothschild R. J. Nash and E. A. Bell Phytochemistry 1986 25 1853. 15 M. BopprC Naturwissenschaften 1986 73 17. 16 A. R. Mattocks ‘Chemistry and Toxicology of Pyrrolizidine Alkaloids’ Academic Press London 1986. 17 K. A. Brown Ann. Mo. Bot.Gard. 1987 74 359. 18 R. B. Kelley J. N. Seiber A. D. Jones H. J. Segall and L. P. Brower Experientia 1987 43 943. I9 S. Schulz W. Francke J. Edgar and D. Schneider Z. Natur-forsch. Sect. C 1988 43 99. 20 W. Wink and P. Romer Naturwissenschaften 1986 73 210. 21 J. M. Horn D. C. Lees N. G. Smith R. J. Nash L. E. Fellows and E. A. Bell in ‘Insects-Plants’ ed. V. Labeyrie G. Fabres and D. Lachaise Dr W. Junk Dordrecht 1987 p. 394. 22 G. Kite personal communication. 23 A. D. Elbein and R. J. Molyneux in ‘Alkaloids Chemical and Biological Perspectives ’ ed. S. W. Pelletier John Wiley New York 1987 Vol. 5 pp. 1-54. 24 A. Nahrstedt Plant Syst. Evol. 1985 150 35. 25 V. Wray R. H. Davis and A. Nahrstedt 2.Naturforsch. Sect. C 1983 38 583. 26 A.Nahrstedt and R. H. Davis Phytochemistry 1986 25 2299. 27 R. H. Davis and A. Nahrstedt Insect Biochem. 1987 17 689. 28 S. Franzl A. Nahrstedt and C. M. Naumann J. Insect Physiol. 1986 32 705. 29 H. Muhtasib and D. L. Evans J. Chem. Ecol. 1987 13 133. 30 K. Wittholm and C. M. Naumann J. Chem. Ecol. 1987 13 1789. 31 S. P. Lynch and R. A. Martin J. Chem. Ecol. 1987 13 47. 32 H. T. A. Cheung T. R. Watson S. M. Lee M. M. McChesney and J. N. Seiber J. Chem. SOC. Perkin Trans. 1 1986 61. NATURAL PRODUCT REPORTS 1989 33 D. L. Evans N. Castoriades and H. Badruddine Oikos,1986,46 325. 34 S. B. Malcolm Proc. 4th Annual Meeting of the International Society for Chemical Ecology Hull England 1987 p. 30. 35 M. Rothschild R. Mummery and C. Farrell Biol.J. Linn. SOC. 1986 28 359. 36 S. Y. Fung and W. M. Herrebout J. Chem. Ecol. 1987 13 2041. 37 J. B. Harborne Annu. Proc. Phytochem. SOC. Eur. 1985 25; 393. 38 A. Wilson J. Chem. Ecol. 1985 11 1161; ibid. 1986 12 49. 39 A. Wilson Phytochemistry 1986 25 1309. 40 A. Wilson J. Chem. Ecol. 1987 13 473. 41 C. G. Jones T. A. Hess D. W. Whitman P. J. Silk and M. S. Blum J. Chem. Ecol. 1987 13 283. 42 M. S. Blum ‘Chemical Defense of Arthropods’ Academic Press New York 1981. 43 K. Dettner Annu. Rev. Entomol. 1987 32 17. 44 D. Daloze J. C. Braekman and J. M. Pasteels Science 1986 233 221. 45 M. F. Braconnier J. C. Braekman D. Daloze and J. M. Pasteels Experientia 1985 41 519. 46 W. V. Brown A. J. Jones M. J. Lacey and B. P. Moore Aust. J. Chem.1985 38 197. 47 R. W. Howard and D. D. Mueller J. Chem. Ecol. 1987 13 1707. 48 J. Meinwald B. Roach K. Hicks D. Alsop and T. Eisner Experientia 1985 41 516. 49 J. Meinwald B. Roach and T. Eisner J. Chem. Ecol. 1987 13 35. 50 K. Dettner and G. Schwinger Experientia 1987 43 458. 51 M. J. Raupp F. R. Milan P. Barbosa and B. A. Leonhardt Science 1986 232 1408. I 52 G. D. Prestwich in ‘Natural Product Chemistry’ ed. Atta-ur- Rahman Springer Berlin 1986 pp. 3 18-329. 53 Y. Roisin J. M. Pasteels and J. C. Braekman Biochem. Syst. Ecol. 1987 15 253. q 54 S. A. Thompson K. Tachibana K. Nakanishi and I. Kubota Science 1986 233 341. 55 M. Murata K. Miyagawa-Kohshima K. Nakanishi and Y. Naya Science 1986 234 585. 56 P. A. Keifer K.L. Rinehart Jr. and I. R. Hooper J. Org. Chem. 1986 51,4450. 57 K. F. Albizati J. R. Pawlik and D. J. Faulkner J. Org. Chem. 1985 50 3428. 58 S. Matsunaga N. Fusetani and K. Hashimoto Experientia 1986 42 84. 59 T. M. Zabriskie J. A. Klocke C. M. Ireland A. H. Marcus T. F. Molinski D. J. Faulkner C.-F. Xu and J. C. Clardy J. Am. Chem. Soc. 1986 108 3123. 60 C. Mitter and J. A. Klun J. Chem. Ecol. 1987 13 1823. 61 P.-C. Zhu F.-L. Kong and Y.-Q. Yu J. Chem. Ecol. 1987 13 977. 62 C. E. Linn J.-W. Du A. Hammond and W. L. Roelofs J. Chem. Ecol. 1987 13 1351. 63 C Lofstedt W. M. Herrebout and J.-W. Du Nature (London) 1986 323 621. 64 M. K. Stowe J. H. Tumlinson and R. R. Heath Science 1987 236 964. 65 G. G. Grant D. Frech L. Macdonald K.N. Slessor and G. G. S. King. J. Chem. Ecol. 1987 13 345. 66 W. Francke S. Francke M. Toth G. Szacs P. Guerin and H. Arn Naturwissenschaften 1987 74 143. 67 P. Zagatti G. Kunesch F. Ramiandrasao C. Malosse D. R. Hall R. Lester and B. F. Nesbitt J. Chem. Ecol. 1987 13 1561. 68 S. B. Krasnoff L. B. Bjostad and W. L. Roelofs J. Chem. Ecol. 1987 13 807. 69 3. Kunesch P. Zagatti A. Pourreau and R. Cussini Z. Natur-forsch. Sect. C 1987 42 657. 70 T. Chuman P. J. Landolt R. R. Heath and J. H. Tumlinson J. Chem. Ecol. 1987 13 1979. 71 E. D. Morgan and D. F. Ollett Naturwissenschuften 1987 74 596. 72 D. Crump R. M. Silverstein H. J. Williams and T. D. Fitz- gerald J. Chem. Ecol. 1987 13 397. 73 S. Peterson Ecol. Entomol. 1987 12 253. 74 P.R. White and M. C. Birch J. Chem. Ecol. 1987 13 1695. 75 H. Kodama K. Mochizuki M. Kohno A. Ohnishi and Y. Kuwahara J. Chem. Ecol. 1987 13 1859. NATURAL PRODUCT REPORTS 1989-J. B. HARBORNE 76 G. W. Dawson D. C. Griffiths J. A. Pickett L. J. Wadhams and C. M. Woodcock J. Chem. Ecol. 1987 13 1663. 77 G. W. Dawson D. C. Griffiths N. F. James A. Mudd J. A. Pickett L J. Wadhams and C. M. Woodcock Nature (London) 1987 325 614. 78 M. A. Adams J. H. Teetu Y. Katz and P.B. Johnsen J. Chem. Ecol. 1987 13 387. 79 J. G. Dulka N. E. Stacey P. W. Sorensen and G. J. van der Kraak Nature (London) 1987 325 251. 80 V. E. Sokolov M. Z. Kagan V. S. Vasilieva V. I. Prihodko and E. P. Zinkevich J. Chem. Ecol. 1987 13 71. 81 G. Preti W. B. Cutler G.M. Christensen H. Lawley G. R. Huggins and C. R. Garcia J. Chem. Ecol. 1987 13 717. 82 A. R. Gustavson M. E. Dawson and D. G. Bonett J. Comp. Physiol. Psychol. 1987 101 210. 83 J. E. Beart T. H. Lilley and E. Haslam Phytochemistry 1986 24 33. 84 M. R. Berenbaum A. R. Zangerl and J. K. Nitao Evolution 1986 40,1215. 85 S. J. McNaughton Am. Nat. 1986 128 765. 86 A. J. Belsky Am. Nut. 1987 129 777. 87 J. C. Horrill and A. J. Richards Heredity 1986 56 277. 88 R. S. L. Burgess and R. A. Ennos Oecologia 1987 73 432. 89 A. N. E. Birch L. E. Fellows S. V. Evans and K. Doherty Phytochemistry 1986 25 2745. 90 C. Gnanasunderam and 0.R. W. Sutherland Phytochemistry 1986 25 409. 91 P. Barbosa and V. A. Krischik Am. Nat. 1987 130 53. 92 M. L.Ostrofsky and E. R. Zettler J. Ecol. 1986 74 279. 93 J. W. McLure in ‘Phytochemical Phylogeny’ ed. J. B. Harborne Academic Press London 1970 pp. 233-268. 94 P. M. Frischknecht M. Battig and T. W. Baumann Phyto-chemistry 1987 26 707. 95 J. E. Huesing and D. Jones Phytochemistry 1987 26 1381. 96 E. Zador and D. Jones Plant Physiol. 1986 82 479. 97 S. Elliott and A. London J. Chem. Ecol. 1987 13 1343. 98 C. M. Bell and A. S. Harestad J. Chem. Ecol. 1987 13 1409. 99 W. J. Foley E. V. Lassak and J. Brophy J. Chem. Ecol. 1987 13 2115. 100 C. A. Mihaliak D. Couvet and D. E. Lincoln J. Chem. Ecol. 1987 13 2059. 101 S. M. Louda M. A. Farris and M. J. Blaa J. Chem. Ecol. 1987 13 569. 102 V. J. Paul M. M. Littler D. S. Littler and W. Fenical J. Chem.Ecol. 1987 13 1171. 103 M. V. Chandravadana J. Chem. Ecol. 1987 13 1689. 104 R. Kasai K. Matsumoto R. L. Nie T. Morita A. Awazu J. Zhou and 0.Tanaka Phytochemistry 1987 26 1371. 105 ‘Plant Molluscicides’ ed. K. E. Mott Wiley Chichester 1987. 106 C. Borel M. P. Gupta and K. Hostettmann Phytochemistry 1987 26 2685. 107 P. S. Steyn F. R. van Heerden R. Vleggaar and L. A. P. Ander- son J. Chem. Soc. Perkin Trans. I 1986 1633. 108 R. J. Capon J. K. Macleod and P. B. Oelrichs Aust. J. Chem. 1986 39 1711. 109 S. Mole and P. G. Waterman Oecologia 1987 72 137. 110 S. Mole and P. G. Waterman Oecologia 1987 72 148. 111 T. N. Asquith and L. G. Butler J. Chem. Ecol. 1985 11 1535. 112 A. E. Hagerman J. Chem. Ecol. 1987 13 437. 113 T. N. Asquith and L.G. Butler Phytochemistry 1986 25 1591. 114 A. E. Hagerman and C. T. Robbins J. Chem. Ecol. 1987 13 1243. 115 A. Serrato J. T. Arnason C. Nozzolillo J. D. H. Lambert B. J. R. Philogene C. Fulcher K. Davidson L. Peacock J. Atkinson and P. Morand J. Chem. Ecol. 1987 13 751. 116 K. Matsudo and S. Senbo Appl. Entomol. Zool. 1986 21 41 1. 117 P. Molgaard Biochem. Syst. Ecol. 1986 14 113. 118 P. Molgaard J. Ecol. 1986 74 1127. 119 G. R. Johnson and K. H. Mann J. Exp. Mar. Biol. Ecol. 1986 97 231. 120 P. D. Steinberg Ecol. Monogr. 1985 55 333. 121 J. A. Klocke B. van Wagenen and M. F. Balandrin Phyto-chemistry 1986 25 85. 122 H. Mehansho L. G. Butler and D. M. Carlson Annu. Rev. Nutr. 1987 7 423. 123 T. N. Asquith J. Uhlig H. Mehansho L. Putman D.M. Carl- son and L. Butler J. Agric. Food Chem. 1987 35 331. 124 T. N. Barry T. F. Allsop and C. Redekopp Br. J. Nutr. 1986 56 607. 125 C. T. Robbins T. A. Hanley A. E. Hagerman 0.Hjelford D. L. Baker C. C. Schwarte and W. W. Munte Ecology 1987 68 98. 126 J. S. Martin M. M. Martin and E. A. Bernays J. Chem. Ecol. 1987 13 606. 127 P. D. Coley Oecologia 1986 70 238. 128 C. R. Metcalfe and I. Chalk ‘Anatomy of the Dicotyledons’ Clarendon Press Oxford 2nd edn. 1983 Vol. 11 pp. 7CL81. 129 F. M. Jones Trans. R. Entomol. SOC. 1932 80 345. 130 D. E. Dussoord and T. Eisner Science 1987 237 898. 131 T. Eisner M. Eisner and J. Meinwald J. Chem. Ecol. 1987 13 943. 132 P. Y. Bouthyette N. Eannetta K. J. Hannigan and P. Gregory Phytochemistry 1987 26 2949.133 R. R. King Y. Pelletier R. P. Singh and L. A. Calhoun J. Chem. SOC. Chem. Commun. 1986 1078. 134 R. R. King and L. A. Calhoun Phytochemistry 1988 27 3761. 135 D. A. Ave P. Gregory and W. M. Tingey Entomol. Exp. Appl. 1987 44 131. 136 S. Y. Lin J. T. Trumble and J. Kumamoto J. Chem. Ecol. 1987 13 837. 137 L. Hansen and P. M. Boll Phytochemistry 1986 25 529. 138 P. M. Boll and L. Hansen Phytochemistry 1987 26 2955. 139 G. Bruhn H. Faasch H. Huhn B. M. Hansen J. Brohan and W. A. Konig Z. Naturforsch. Teil B 1987 42 1328. 140 G. W. Reynolds and E. Rodriguez Phytochemistry 1986 25 1617. 141 H. J. Zeringue Phytochemistry 1987 26 1357. 142 R. H. Miller A. A. Berryman and C. A. Ryan Phytochemistry 1986 25 611. 143 H.S. Chiang D. M. Norris A. Ciepiela P. Shapiro and A. Oosterwyk J. Chem. Ecol. 1987 13 741. 144 Y.C. Toong D. A. Schooley and F. C. Baker Nature (London) 1988 333 170. 145 G. B. Staal Annu. Rev. Entomol. 1986 31 391. 146 P. Proksch M. B. Isman L. Witte and T. Hartmann Phyto-chemistry 1987 26 2227. 147 N. Nichimoto Y. Shiobara M. Fujino S. Inoue T. Takemoto F. de Oliveira G. Akisue M. K. Akisue G. Hashimoto 0.Tanaka R. Kasai and H. Matsuura Phytochemistry 1987,26 2505. 148 A. Suksamrarn J. S. Wilkie and D. H. S. Horn Phytochemistry 1986 25 1301. 149 A. Fukuzawa M. Miyamoto Y. Kumagai and T. Masamune Phytochemistry 1986 25 1305. 150 I. Kubo S. Komatsu Y.Asaka and G. de Boer J. Chem. Ecol. 1987 13 785. 151 T. A. van Beek and A. E. de Groot Recl.Trav. Chim. Pays-Bas 1986 105 513. 152 W. Kraus M. Bokel A. Klenk and H. Pohnl Tetrahedron Lett. 1985 26 6435. 153 H. B. Broughton S. V. Ley A. M. Z. Slawin D. J. Williams and E. D. Morgan J. Chem. SOC. Chem. Commun. 1986 46. 154 J. T. Arnason M. B. Isman B. J. R. Philogene and T. G. Wad- dell J. Nat. Prod. 1987 50 690. 155 B. K. Mitchell and J. F. Sutcliffe Physiol. Entomol. 1984 9 57. 156 ‘Perspectives in Chemoreception and Behaviour’ ed. R. F. Chapman E. A. Bernays and J. G. Stoffolano Springer Berlin 1987. 157 R. Nishida T. Ohsugi S. Kokubo and H. Fukami Experientia 1987 43 342. 158 R. Nishida personal communication. 159 P. Feeny Phytochemistry 1988 27 3439. 160 K. Honda J. Chem. Ecol. 1986 12 1999. 161 R. Nishida and H.Fukami J. Chem. Ecol. 1988 14 in press. 162 W. A. L. David and B. 0.C. Gardiner Bull. Entomol. Res. 1962 53 91. 163 M. Rothschild H. Alborn G. Stenhagen and L. M. Schoon- hoven Phytochemistry 1988 27 101. 164 B. E. Tabashnik J. Chem. Ecol. 1987 13 309. 165 J. Hurter E. F. Boller E. Stadler B. Blattmann H.-R. Buser N. U. Bosshard L. Damm M. W. Kozlowski R. Schoni F. Raschdorf R. Dahinden E. Schlumpf H. Fritz W. J. Richter and J. Schreiber Experientia 1987 43 157. 166 A. L. Averill and R. J. Prokopy J. Chem. Ecol. 1987 13 167. 167 Y. S. Hwang M. S. Mulla J. D. Chaney G. G. Liu and H. J. Lu J. Chem. Ecol. 1987 13 245. 168 0.Pellmyr G. Bergstrom and I. Groth Phytochemistry 1987 26 1603. 169 I. Groth G. Bergstrom and 0.Pellmyr Biochem. Syst.Ecol. 1987 15 441. 170 A. K. Borg-Karlson and I. Groth Phytochemistry 1986 25 1297. 171 A. K. Borg-Karlson G. Bergstrom and I. Groth Chem. Scr. 1985 25 283. 172 A. K. Borg-Karlson and J. Tengo J. Chem. Ecol. 1986 12 1928. 173 G. Bergstrom G. Birgersson I. Groth and L. A. Nilsson J. Chem. Ecol. 1988 14 in press. 174 N. H. Williams and W. M. Whitten Biol. Bull. (Woods Hole Mass.) 1983 164 355. 175 W. M. Whitten N. H. Williams W. S. Armbruster M. A. Bat- tiste L. Strekowski and N. Lindquist Syst. Bot. 1986 11 222. 176 L. A. Nilsson Bot. Notiser. 1978 131 35. 177 L. Thien P. Bernardt G. W. Gibbs 0.Pellmyr G. Bergstrom I. Groth and G. McPherson Science 1985 227 540. 178 V. Grant and K. A. Grant Bot. Gaz. (Chicago) 1983 144 280.179 H. E. M. Dobson J. Bergstrom G. Bergstrom and I. Roth Phytochemistry 1987 26 3 17 I. 180 N. M. Waser and M. V. Price Nature (London) 1983 302 422. 181 J. B. Harborne and D. M. Smith Biochem. Syst. Ecol. 1978 6 127. 182 K. N. Paige and T. G. Whitham Science 1985 227 315. 183 L. W. Macier Bull. Torrey Bot. Club 1986 113 101. 184 ‘The Science of Allelopathy’ ed. A. R. Putnam and C. S. Tang John Wiley Chichester 1986. 185 J. D. Weidenhamer T. C. Marton and J. T. Romeo J. Chem. Ecol. 1987 13 1481. 186 ‘The Chemistry of Allelopathy ’ (A.C.S. Symposium Series No. 268) ed. A. C. Thompson American Chemical Society Washing- ton D.C. 1985. 187 R. G. Kelsey and L. J. Locken J. Chem. Ecol. 1987 13 19. 188 L. A. Weston B. A. Burke and A. R. Putnam J.Chem. Ecol. 1987 13 403. 189 W. J. Connick J. M. Bradons M. Legendre S. L. Vail and R. M. Menges J. Chem. Ecol. 1987 13 463. 190 J. P. Barnes and A. R. Putnam J. Chem. Ecol. 1987 13 889. 191 J. P. Barnes A. R. Putnam B. A. Burke and A. J. Aasen Phyto-chemistry 1987 26 1385. 192 N. H. Fischer N. Tanrisever and G. B. Williamson in ‘Natural Products Potential in Agriculture’ ed. H. G. Cutler American Chemical Society Washington D.C. 1988. 193 G. R. Waller D. Kumari J. Friedman N. Friedman and C. H. Chou in ‘The Science of Allelopathy’ ed. A. R. Putnam and C. S. Tang John Wiley Chichester 1986 pp. 243-270. 194 D. E. Hathway Biol. Rev. Cambridge Philos. Soc. 1986 61 435. 195 P. R. Atsatt in ‘Encyclopedia of Plant Physiology New Series’ ed.0. L. Lange P. S. Nobel C. B. Osmond and H. Ziegler Springer Berlin 1983 Vol. 12C pp. 519-535. 196 M. Chang D. H. Netzly L. G. Butler and D. G. Lynn J. Am. Chem. Soc. 1986 108 7858. 197 M. Chang and D. G. Lynn J. Chem. Ecol. 1986 12 561. 198 S. Ghosal Y. Kumar D. K. Chakrabarti J. Lal and S. K. Singh Phytochemistry 1986 25 1097. 199 D. D. Rowan M. B. Hunt and D. L. Gaynor J. Chem. Soc. Chem. Commun. 1986 935. 200 C. M. Weedon and P. G. Mantle Phytochemistry 1987 26 969. 201 D. B. Strongman G. M. Strunz P. Giguere C. M. Yu and L. Calhoun J. Chem. Ecol. 1988 14 753. 202 S. V. Pathra W. B. Gleason Y. W. Lee and C. J. Mirocha Can. J. Chem. 1986 64 1308. 203 P. C. Lyons R. D. Plattner and C. W. Bacon Science 1986,232 485. 204 M.F. Mackay A. van Donkelaar and C. C. J. Culvenor J. Chem. Soc. Chem. Commun. 1986 1219. 205 S. E. Stachel E. Messens M. van Montagu and P. Zambryski Nature (London) 1985 318 624. 206 P. A. Spencer and G. H. N. Towers Phytochemistry 1988 27 278 1. 207 S. N. Sheikholleslam and D. P. Weeks Plant Mol. Biol. 1987 8 291. 208 C. Tang and C. Young Plant Physiol. 1982 69 155. 209 J. L. Firmin K. E. Wilson L. Rossen and A. W. B. Johnston Nature (London) 1986 324 90. 210 N. K. Peters J. W. Frost and S. R. Long Science 1986 233 977. 211 J. W. Redmond M. Batley M. A. Djordjevic R. W. Innes P. L. Kuempfel and B. G. Rolfe Nature (London) 1986 323 632. NATURAL PRODUCT REPORTS 1989 212 Y. Kapulnik C. M. Joseph and D. A. Phillips Plant Physiol. 1987 84 1193.213 M. A. Djordjevic J. W. Redmond M. Batley and B. G. Rolfe EMBO J. 1987 6 1173. 214 T. Ueno in ‘Pesticide Science and Biotechnology’ ed. R. Greenhalgh and T. R. Roberts Blackwells 1987 pp. 89-96. 215 V. Macko T. J. Wolpert W. Acklin B. Jaun J. Seibl J. Meili and D. Arigoni Experientia 1985 41 1366. 216 Y. Suzuki L. W. Coleman J. M. Daly,Y. Kono H. W. Knoche and S. Kakeuchi Phytochemistry 1987 26 687. 217 D. Perreaux H. Maraite and J. A. Meyer Physiol. Mol. Plant Pathol. 1986 28 323. 218 B. Bodo D. Davoust D. Lecommandeur S. Rebuffat I. Genetet and J. Pinon Tetrahedron Lett. 1987 28 2355. 219 L. Sparapano A. Evidente A. Ballio A. Graniti and G. Rand- azzo Experientia 1986 42 627. 220 F. Sugawara G. Strobel L. E. Fisher G. D. van Duyne and J.Clardy Proc. Natl. Acad. Sci. USA 1985 82 8291. 221 K. Sugawara F. Sugawara and G. Strobel J. Org. Chem. 1985 50 5631. 222 F. Sugawara and G. Strobel Phytochemistry 1987 26 1349. 223 S. Manulis I. Barash Y. Kashman and D. Netzer Phyto-chemistry 1984 23 2193. 224 S. Manulis Y. Kashman and I. Barash Phytochemistry 1987 26 1312. 225 G. Assante and G. Nasini Phytochemistrv 1987 26 703. 226 A. Arnone G. Assante L. Merlini and G. Nasini J. Chem. Soc. Perkin Trans. 1 1986 525. 227 H. Koshino S. Togiya T. Yoshihara S. Sakamura and T. Shimanuki Tetrahedron Lett. 1987 28 73. 228 W. S. Bowers H. C. Hoch P. H. Evans and M. Katayama Science 1986 232 105. 229 A. Ichihara M. Hashimoto and S. Sakamura Agric. Biol. Chem. 1985 49 2207. 230 R.A. Dixon Nature (London) 1986 324 303. 231 J. B. Harborne Annu. Proc. Phytochem. Soc. Eur. 1987 27 195. 232 J. B. Harborne in ‘Natural Resistance of Plants to Pests’ (A.C.S. Symposium Series No. 296) ed. M. B. Green and P. A. Hedin American Chemical Society Washington D.C. 1986 pp. 22-35. 233 ‘Biology and Molecular Biology of Plant-Pathogen Interactions ’ ed. J. A. Bailey Springer Berlin 1986. 234 Recognition in Microbe-Plant Symbiotic and Pathogenic Inter- actions’ ed. B. Lugtenberg Springer Berlin 1986. 235 M. Cojocaru S. Droby E. Glotter A. Goldman H. E. Gottlieb B. Jacoby and D. Prusky Phytochemistry 1986 25 1093. 236 S. Droby D. Prusky B. Jacoby and A. Goldman Physiol. Mol. Plant Pathol. 1986 29 173. 237 C. Nishino S. Manabe N.Enoki T. Nagata T. Tsushida and E. Hamaya J. Chem. Soc. Chem. Commun. 1986 720. 238 0.Potterat H. S. Evans J. D. Msonthi and K. Hostettmann Helv. Chim. Acta 1987 70 1551. 239 M. Nagai and M. Tada Chem. Lett. 1987 1337. 240 J. G. Roddick in ‘Solanaceae Biology and Systematics’ ed. W. G. D’Arcy Columbia University Press New York 1986 pp. 201-222. 241 C Wippich and M. Wink Experientia 1985 41 1477. 242 R. L. Nicholson L. G. Butler and T. N. Asquith Phytopath-ology 1986 76 1315. 243 R. Jambunathan L. G. Butler R. Bandyopadhyay and L. K. Mughogho J. Agric. Food Chem. 1986 34 425. 244 M. Takasugi N. Katsui and A. Shirata J. Chem. Soc. Chem. Commun. 1986 1077. 245 M. Takasugi K. I. Monde N. Katsui and A. Shirata Chem. Lett. 1987 1631. 246 R.Wijnsma J. T. Go I. N. van Weerden P. A. A. Harkes R. Verpoorte and A. Baerheim Svendsen Plant Cell Rep. 1985 4 241. 247 Y. Shoyama M. Matsumoto and 1. Nishioka Phytochemistry 1987 26 983. 248 M. Takasugi and N. Katsui Phytochemistry 1986 25 2751. 249 A. J. Khan G. Kunesch S. Chuilon and A. Ravise Fruits 1985 40,807. 250 P. S. Marshall Ph.D. thesis Reading University 1988. 251 P. S. Marshall J. B. Harborne and G. S. King Phytochemistry 1987 26 2493. 252 J. L. Ingham S. Tahara and S. Z. Dziedzic J. Nut. Prod. 1986 49 631. NATURAL PRODUCT REPORTS 1989-5. B. HARBORNE 253 R. A. Franich M. J. Carson and S. D. Carson Physiol. Mol. Plant Pathol. 1986 28 267. 254 W. E. Hillis and T. Inoue Phytochemistry 1966 5 483. 255 W. Knogge E.Kombrink E. Schmelzer and K. Hahlbrock Planta 1987 171 279. 256 W. S. Pierpoint Phytochemistry 1986 25 1595. 257 S. Kauffmann M. Legrand P. Geoffroy and B. Fritig EMBO J. 1987 6 3209. 258 M. Legrand S. Kauffmann P. Geoffroy and B. Fritig Proc. Natl. Acad. Sci. USA 1987 84 6750. 259 J. Rigden and R. Coutts Trends Genet. 1988 4 87. 260 A. Schlumbaum F. Mauch U. Vogeli and T. Boller Nature (London) 1986 324 365. 261 B. A. Stermer and R. Hammerschmidt Physiol. Mol. Plant Pathol. 1987 31 453. 262 R. C. Hignett Physiol. Mol. Plant. Pathol. 1987 30 131.