Rational design of insecticides

SESSION IX PERSISTENT AND BIOLOGICALLY DEGRADABLE ANALOGUES OF DDT Bull. Org. mond. Sant 1971, 44, 355-362 Bull. Wld Hith Org. Rational Design of Insecticides G. HOLAN 1 A steric model of diaryl insecticides, which includes the structure of DDT and its analogles, was usedfor the synthesis ofnew highly active insecticides. The new compounds show low toxicity to mammals and their insecticidal activity can be potentiated by micro- somal oxidase inhibitors. There is evidence that they readily undergo biological degrada- tion. However, a spontaneous controlled chemical degradation was also established for the diaryl oxetanes, which form one group of a new series of insecticides. The theoretical model was further investigated using X-ray crystallographty to establish one accurate structure. Biologically, the model is supported by studies of the correlation between insect mortalities and temperature and by measurements of nerve impulses at fly chemoreceptors after treatment with selected compounds. For the past several years we have been engaged in the rational synthesis of broad-spectrum insecticides having low toxicity for mammals. This paper briefly outlines several facets of our work in both the chemical and the biological fields. There are good reasons for using a theoretical model rather than the more common approach of the random synthesis of biologically active com- pounds. If the model is valid, it becomes possible to correct past omissions in the syntheses of related analogues; it also becomes feasible to develop new structures that could not be obtained by intuition. THE MODEL Our model for the synthesis of aryl insecticides was developed from a series of diaryl halocyclopropane analogues of DDT (Holan, 1969). The three- dimensional model, which also accommodates previ- ously known DDT analogues, was obtained by com- paring projections of steric atomic models with accurate insect mortality data for active compounds. The shape that was obtained in this way resembled a "molecular wedge" with two distinct features. First, the influence of the apex of the "wedge" (e.g., the part containing the -CCI3 group of DDT) on the insecticidal activity was found to be indepen- dent of the chemical composition of the group form- ing it, provided one kept to strict dimensional requirements. Second, the two phenyl rings compris- ing the "base" had to have electron-donating substi- I Division of Applied Chemistry, CSIRO, Box 4331 G.P.O. Melbourne, Australia. tuents. The size limitation for this part of the molecule was found to be less restrictive than that for the apex. An hypothesis that assigned definite roles to the two parts of the theoretical model was developed from the known biological effects of this type of insecticide in whole organisms and from more detailed studies of their mode of action in nerve axons (Narahashi & Haas, 1968). It was postulated that (1) lipid-soluble insecticides distribute themselves at the protein-lipid interface ofa nerve membrane and the base containing the phenyl rings locks itself by complex-formation with the aromatic rings into the overlying protein; (2) the apex keeps open a molecu- lar spring, site, or channel for sodium ions, thus in- ducing a leakage of these ions into the nerve axon; and (3) this action causes the delay in the falling phase of the action potential and results in the characteristic multiple spikes of generated nerve impulses. Since this model was formulated we have obtained additional evidence for the validity of its proposed biological role. Complex-formation between the insecticide and nerve protein is consistent with the well-known phenomenon of negative correlation between insect mortality and temperature. This effect has been observed in the hydrocarbon series only for insecticides containing an aryl ring (Barker, 1957; Hoffman & Lindquist, 1949; Guthrie, 1950). The equilibrium constants of charge-transfer com- plexes are similarly temperature-related (Briegleb, 1961) and their stability is directly dependent on the polarization of the aromatic rings. Fig. 1, which is based on previously synthesized halocyclopropanes and on several of the new insecticides, shows that the 2648 -355- G. HOLAN Fig. 1 Temperature-mortality curves for five different diaryl insecticides * F' r- __j _ _ _ _L L_ _ __ i- _ __ __- 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Teniperiture °C CH3 R - cIi - c-CH3 R CH3 NO - ---- CH -C R CL2-CH3 3 R 1X CI R R \ ci CI Cl - CH-c-Cl ]~~ ~ ~ ~ o2i RZcl R = OC Hs RI = Cl * The fact that the lines are almost parallel (i.e., that their slopes are almost identical) indicates an effect of suitable substitution in the phenyl rings of otherwise dissimilar structures. This negative correlation is related to similar temperature effects of 7r-charge-transfer-complex equilibrium constants and is believed to involve the formation of a complex between the phenyl rings of the insecticides and the overlaying protein of nerve membranes. 356 RATIONAL DESIGN OF INSECTICIDES 357 Fig. 2 Projection of an atomic model of 1,1-bis(p-ethoxyphenyl)-2,2-dimethylpropane, obtained from X-ray crystallographic data illot ops O hydrogen atoms O carbon atoms * oxygen atoms carbon atoms in the neopentane skeleton negative correlation between temperature and mor- tality is the same for all the phenyl-ring-substituted insecticides, provided they contain an electron-donor group in thepara-position. This effect is independent of the magnitude of the insecticidal activity and of the chemical nature of the molecule. From an organic chemist's point of view there is no relation- ship between the dichlorocyclopropane moiety and a tert-butyl group or between this and a nitroalkane, except for their dimensional similarity. Therefore the complex-formation must reside in the suitably substituted aromatic rings of the insecticides. The biological role of the apex of the molecule, which was claimed to have a size related to a hydra- ted sodium ion, was more difficult to prove and the detailed work will be reported elsewhere. Briefly, however, it was found that the insecticide-induced multiplicity of single spikes in the train of impulses measured at a salt-stimulated labellar taste receptor of a susceptible housefly (Barton-Browne & Kerr, 1967) is independent of concentration. In struc- turally similar size-graded halocyclopropane insec- ticides this effect can be related both to the diametre of the apex ofthe molecule and to the insect mortality. DESIGN OF COMPOUNDS The validity of the model was further confirmed by the predictive synthesis of active structures. In the study of numerous compounds in the fluorohydro- carbon, phosphorus, heterocyclic, alicyclic, and hydrocarbon series we did not find a single excep- tion to the defined dimension of a projected van der Waals diameter of the apex of the molecules. This must be within the limits of 0.6 0.05 nm, and it has been found that the LD50 of compounds that con- form to this limitation for a DDT-susceptible strain of the housefly (WHO/IN/Musca domestica/l) is less than 5.0 ,ug/insect. A few selected new structures that illustrate this design are listed in the accom- panying table. In the nitroalkane series the p-ethoxyphenyl ana- logues of the insecticides 1,1-bis(p-chlorophenyl)-2- nitropropane (Prolan) and 1 ,1-bis(p-chlorophenyl)-2- nitrobutane (Bulan, compound XI 1) were not known. This is understandable since, in the absence of a predictive model, the lack of activity of the easily 1 Compounds referred to by roman numerals are those listed in the accompanying table. G. HOLAN Toxicity of new insecticides to the mouse and to a susceptible strain of the housefly* GENERAL FORMULA Rl-4R3 -CR-2 R3 substituent ci CH-C-Cl/I ci CH3 CH-C-CH3/ CH3 CH3 CH-CH2-c-CH3/ CH3 CH3 /-CH, CH-CH-CH3 N02 CH3 CH-C-NO2 CH3 CH-CH-CH2-CH3 /I N02 Compound No. 11 III IV iv V Vi vii Vill ix x Xi Xii Xiii Xiv Other substituents R1 = R2 = Cl (DDT) R' = R2 = C2H50 RI = R2 = CH30 RI = R2 = C2H50 R1 = R2 = Cl RI = R2 = CH30 R' = R2 = C2H50 R' = C2H5S R2= C2HsO RI = 3,4-(-O-CH2-O-) R2= C2H50 R' = R2 = C2HsO R' = R2 = Cl (Bulan) R' = R2 = C2HsO R' = C2H5S R2 = C2H50 R' = 3,4-(-O-CH2-O-) R2= C2H50 LDso (,ug/insect) for Musca domestica a LUnsyn- Syn - ergized ergized b 0.24 1.92 3.2 2.0 12.0 >20 0.48 0.16 0.14 1.01 1.69 0.55 0.11 0.12 0.25 0.32 1.7 0.36 0.065 0.015 0.019 0.053 1.73 0.061 0.04 0.028 358 LDso (mg/kg) for the mouse c 570 5 200 1 150 1 040 360 d -2 000 -2 000 d 1 160 980 d Synergistic ratio 0 5.5 1.9 5.5 7.4 10.7 7.4 10.6 0 9.0 2.8 4.3 RATIONAL DESIGN OF INSECTICIDES Toxicity of new insecticides to the mouse and to a susceptible strain of the housefly (concluded) v GENERAL FORMULA LD50 (,jg/insect) R3 substituent Compound Other substituents for Musca domestica a Synergistic LDso (mg/kg)No. Usn Sy- ratio for the mouse c ergized ergized b XV R1 = R2 = Cl 1.27 0.66 1.9 3 500 R4 = R5 = CH3 R6 = H XVI RI = R2 = Cl >100 >100 _ _ R4 = R6 = CH3 R5= H XVII RI = R2 = C2HsO 0.52 0.01 52 1 200 R4= R5 = CH3 380 d Re = H XVIII R1 = R2 = C2H50 >100 >100 - _ R = Rs= R6 = CH3 XIX R' = C2HsO 1.43 0.086 17.9 - R4 = CH30 R4= R5 = CH3 Re = H * All tests and calculations of LD50 values were carried out according to a previously described method (Holan, 1969). a Using a DDT-susceptible strain of the housefly (WHO/IN/Musca domestica/1). b In potentiation experiments, 0.5 microlitre of a 1 % solution of sesamex in acetone was applied immediately after the application of the dose of insecticide. c The compounds, in olive oil, were injected intraperitoneally into female albino Swiss mice. The average weight of the mice was 20 g and 5 mice were used for each dosage level. The median lethal dose was calculated from the mortalities obtained in 5 days. d Toxicity when synergized with sesamex (1: 1 insecticide: sesamex ratio). prepared 1,1-bis(p-methoxyphenyl) analogue (VI) and the 1-p-ethoxyphenyl-1-o-ethoxyphenyl ana- logue (Jacob et al., 1951) would discourage workers from undertaking the extensive chemical investiga- tions that we found were necessary to solve the difficult synthesis of the bis(p-ethoxyphenyl) deriva- tives VII & XII. Moreover, the activity of the new tertiary nitroalkane insecticide X could not have been predicted without the use of the model. In the hydrocarbon series, the 3,3-dimethyl- butane derivative IV and the 2,2-dimethylcyclo- propane derivative V were prepared after the examination of their projected atomic models. The latter compound is of particular interest because of its relationship to the dimethylcyclopropane struc- tural part of pyrethrin insecticides. The synthesis of 1,1-bis(p-ethoxyphenyl)-2,2-dime- thylpropane (II) gave us an opportunity to obtain from X-ray crystallographic data the first exact structure of a DDT-type insecticide (Fig. 2). This will shortly be followed by the solution of the structure of DDT itself. In the rigidly structured surrounding of the nerve membrane (Hechter, 1965) where the insecticides are thought to act, the compounds are better represented by structures obtained in a solid crystal lattice than by their conformation in solution. 359 G. HOLAN Fig. 3 Projected maximum van der Waals diameters of p-chlorophenyl-substituted 3,3-dimethyl oxetane (left) and trans-3,4-dimethyl oxetane (right) rings* Cl Cl ICH33 \ I \ / 01 I .C- I I H CH31 ;,0.7 nm _-. * The LD5o values (iug/insect) for a susceptible strain of the housefly were 1.27 for the 3,3-dimethyl compound and >200 for the 3,4-dimethyl compound. However, the best illustration of the design is provided by the formation of the dimethyl oxetanes. In a search for a non-persistent insecticide we looked for a semistable heterocyclic ring that would conform to the stated dimensions. From projections of atomic models it was deduced that the correct size would be obtained by the minor change of one methyl group between the 3,3-dimethyl and the 3,4-dimethyl oxetanes; similarly, the addition of a methyl group at the 4-position of the 3,3-dimethyl oxetanes would alter unfavourably the down- ward projection of the model (Fig. 3). The 3,3- dimethyl oxetanes have the correct dimensions and, when potentiated, the bis(p-ethoxyphenyl) derivative XVII shows an insecticidal activity 25 times that of DDT. The 3,4-dimethyl and 3,3,4-trimethyl derivatives XVI and XVIII do not conform and are completely inactive. The initial selection of the oxetane ring was also based on previous reports (Margerum et al., 1959) that indicated that the compounds would readily undergo degradation by electrocyclic fission to formaldehyde and the biologically inactive 1,1 -bis(aryl)-2,2-dimethyl- ethylenes (Fig. 4). We have found that this chemical degradation can be controlled by the removal or addition of free-radical initiators. Fig. 4 A-O :9:113 Cl 1CH3 ICH3 H3C I I 1 C HC ~~~~H3C CH3 360 + H CHO RATIONAL DESIGN OF INSECTICIDES 361 POTENTIATION OF INSECTICIDAL ACTIVITY In the alkoxyphenyl or alkylthiophenyl deriva- tives it is possible to assume the formation of quinonic resonance structures.' These enhance the reactivity of the molecules and their ability to form n-charge-transfer complexes. The reson- ance structures would also contribute to a highly increased susceptibility to electrophilic oxidative attack at the benzylic carbon. That this takes place is strongly supported by the effect of microsomal mixed-function oxidase inhibitors (e.g., sesamex) on the activity of the alkoxy and alkylthio insecticides (see the accompanying table). With the p-chloro- phenyl-substituted insecticides, such as Bulan (XI) or DDT (1), no potentiation is observed. Furthermore, comparison of the synergistic ratios of the p-chloro- phenyl-substituted and p-ethoxyphenyl-substituted oxetanes XV and XVII demonstrates the contribu- tion of the ethoxy group to the ease of oxidative degradation of the insecticides. If the assumption is valid that the intrinsic insecticidal activity is observed only when the compounds are potentiated (i.e., when the full dose of the non-degraded compound reaches the active site), then several of the alkoxyphenyl compounds are much more active than their halogen analogues. The potentiation of the alkoxyphenyl compounds does not discriminate to any extent be- tween the synergists reported to be inhibitors of the mixed-function oxidase in insects (Wilkinson, 1967). When synergized with a 1.0% solution of piperonyl butoxide, the nitrobutane XII showed an LD,0 of 0.05 pg/insect for the housefly, whereas it showed an LD50 of 0.12 jg/insect when synergized with a 0.5 ' solution of 4-dimethylamino-5-nitro-1,2- methylenedioxybenzene (Wilkinson, 1967). Potentiation would be an attractive method for the control of insects in particular, since the data in the accompanying table show that, for several compounds (VIII, IX, XII, and XVII), potentiation takes place predominantly in the insect rather than the mammal. I These have been referred to as " quinone-methide" resonance structures. The addition of synergists to overcome resistance in insects or to increase the activity of insecticides is not an established method in public-health and agricultural pest control. Therefore we attempted the synthesis of " self-potentiated" compounds. This approach was based on the observation that a lack of symmetry of substituents in the phenyl rings does not negate the activity of the compounds. Com- pounds IX and XIV contain the 3,4-methylenedioxy- phenyl group, which is a major substituent of several insecticide synergists (Wilkinson, 1967). Keeping within the dimensional limits of our model, we substituted a p-ethoxy group in the other phenyl ring. The non-synergized insecticidal activity of these unsymmetrical compounds reaches nearly the level of the synergized activity of their bis(p-ethoxy- phenyl) analogues VII and XII. A similar increase in activity is found for the thiophenyl derivatives VIII and XIII. We have not as yet determined whether the compounds actively inhibit the oxidase system or only present a substrate that is more resistant to the enzymatic oxidation. This we hope to resolve by further synthesis and the testing of combinations of the insecticides. The high activity and rapid degradation of some of the new compounds, together with their low acute toxicity for mammals (see the table) could make them a valuable alternative to some existing insecticides. However, the lack of persistence of the new insecti- cides is unsatisfactory under present practice of control of insect pests. Consequently, we are con- tinuing our research and hope to obtain in the future a structure with a correct balance of activity and persistence. The results of studies of the chemistry of the com- pounds reported herein, of X-ray determinations of their structure, and of investigations of their degrada- tion and of their activity against other insects will be published elsewhere. The new insecticides are covered by patents assigned to the Commonwealth Scientific and Industrial Research Organization. ACKNOWLEDGEMENTS The author thanks Dr E. Shipp, Entomology Depart- ment, School of Life Sciences, University of New South Wales, for supervising the entomological work; Dr C. L. Kennard and Mr T. Delacy for X-ray structure analysis; Mrs J. Johnston, Mrs T. Ejmont, Mr F. Romer, and Miss N. Ali for assistance in the biological studies; and Mr D. O'Keefe, Mr R. Eibl, and Mr R. Walser for assistance in the chemical investigations. Studies of toxicities for mammals were carried out in part at the Department of Pharmacology, University of Melbourne, and at the Commonwealth Serum Laboratories, Mel- bourne. 362 G. HOLAN REFERENCES Barker, R. J. (1957) J. econ. Ent., 50, 446-450 Barton-Browne, L. & Kerr, R. W. (1967) Ent. exp. appl., 10, 337-346 Briegleb, G. (1961). In: Electron-Donator-Acceptor Kom- plexe, Berlin, Springer, pp. 140-146 Guthrie, F. E. (1950) J. econ. Ent., 43, 559-560 Hechter, 0. (1965) Ann. N. Y. Acad. Sci., 125, 625-646 Hoffman, R. A. & Lindquist, A. W. (1949) J. econ. Ent., 42, 891-893 Holan, G. (1969) Nature (Lond.), 221, 1025-1029 Jacob, T. A., Bachman, G. B. & Haas, H. B. (1951) J. org. Chem., 16, 1572-1574 Margerum, J. D., Pitts, J. N., Rutgers, J. G. & Searle, S. S. (1959) J. Amer. chem. Soc., 81, 1549-1551 Narahashi, T. & Haas, H. G. (1968) J. gen. Physiol., 51, 177-198 Wilkinson, C. F. (1967) J. agric. Food Chem., 15, 139- 147 DISCUSSION WRIGHT: How stable are these compounds in relation to the existing chlorinated hydrocarbons? HOLAN: They are less stable. HAYES: The term " persistent " applied to pesticides may be misleading. All organic compounds undergo degrada- tion at a slower or a faster rate. The serious fault of DDT is not merely that it is persistent but that it accumul- ates to a high degree in all steps of at least a few food chains. HOLAN: We are attempting to alter the stability of our compounds to make them more persistent. Several, such as the oxetanes, would be expected to be stable on dry surfaces but possibly to hydrolyse in an aqueous environment. DAUTERMAN: A number of chlorinated-hydrocarbon insecticides induce microsomal enzymes, which has implications for health. Do you have any information on whether your compounds induce microsomal enzymes? HOLAN: No. The inducement of oxidative enzymes follows from the model I have proposed. A " molecular wedge " should affect membranes generally and should release bound enzymes from mitochondrial surfaces. Our compounds are less stable in biological environ- ments than is DDT, and therefore such action is less likely. DONNINGER: I do not fully understand your reply to Dr Dauterman's question. Have any long-term mam- malian feeding studies been carried out and if so is any liver enlargement observed? HOLAN: We have not carried out any chronic toxicity tests. In a short study of a few of the new compounds, Dr Barnes demonstrated no cumulative toxicity.