Agonist actions of neonicotinoids on nicotinic acetylcholine receptors expressed by cockroach neurons

NeuroToxicology 28 (2007) 829–842

Agonist actions of neonicotinoids on nicotinic acetylcholine receptors expressed by cockroach neurons Jianguo Tan a,1, James J. Galligan b, Robert M. Hollingworth a,* b

a Department of Entomology, Michigan State University, East Lansing, MI 48824, United States Department of Pharmacology and Toxicology, Michigan State University, East Lansing, MI 48824, United States

Received 29 December 2006; accepted 12 April 2007 Available online 20 April 2007

Abstract The agonist actions of seven commercial neonicotinoid insecticides and nicotine were studied on nicotinic acetylcholine receptors (nAChRs) expressed by neurons isolated from the three thoracic ganglia of the American cockroach, Periplaneta americana. Single electrode voltage clamp recording was used to measure agonist-induced inward currents. Acetylcholine, nicotine and all neonicotinoids tested, except thiamethoxam, caused inward currents which were blocked reversibly by methyllycaconitine, a nAChR antagonist. Based on maximum inward currents, neonicotinoids could be divided into two subgroups: (1) those with a heterocyclic ring in their electronegative pharmacophore moiety (i.e. nicotine, imidacloprid and thiacloprid) were relatively weak partial agonists causing only 20–25% of the maximum ACh current and (2) open chain compounds (i.e. acetamiprid, dinotefuran, nitenpyram, and clothiandin) which were much more effective agonists producing 60–100% of the maximum ACh current. These compounds also elicited different symptoms of poisoning in American cockroaches with excitatory responses evident for the low efficacy agonists but depressive and paralytic responses predominating for the most efficacious agonists. No correlation was observed between agonist affinity and efficacy on these nAChRs. Thiamethoxam, even at 100 mM, failed to cause an inward current and showed no competitive interaction with other neonicotinoids on nAChRs, indicating that it is not a direct-acting agonist or antagonist. Despite the probable presence of multiple subtypes of nAChRs on cockroach neurons, competition studies between neonicotinoids did not reveal evidence that separate binding sites exist for the tested compounds. The size of inward currents induced by co-application of neonicotinoid pairs at equal concentration (100 mM) were predominantly determined by the one with higher binding affinity as indicated by EC50 values, rather than by the one with higher binding efficacy as indicated by maximal current (Imax). Agonist efficacy, but not affinity, was positively correlated with insecticidal activity. These findings indicate that: (1) agonist affinity and efficacy vary independently with neonicotinoid structure; (2) high agonist efficacy is dependent on the presence of an acyclic electronegative pharmacophore group; (3) agonist efficacy is a significant factor in the insecticidal activity of neonicotinoids to cockroaches; (4) lower efficacy compounds cause excitatory symptoms (Type A), while high efficacy compounds cause depressive/paralytic symptoms (Type B). # 2007 Elsevier Inc. All rights reserved. Keywords: Single electrode voltage clamping recording; Neonicotinoids; Nicotinic acetylcholine receptors (nAChRs); Imidacloprid; Thiacloprid; Acetamiprid; Dinotefuran; Clothianidin; Nitenpyram; Thiamethoxam; Nicotine; American cockroach; German cockroach

1. Introduction Neonicotinoids, such as imidacloprid (IMI), are increasingly deployed for pest management and have become an important class of insecticides (Kagabu, 1997; Tomizawa and Casida, 2003). Up to now, six additional neonicotinoids have been

* Corresponding author. Tel.: +1 517 432 7718; fax: +1 517 353 5598. E-mail address: [email protected] (R.M. Hollingworth). 1 Present address: Department of Entomology, Cornell University, NYSAES, Geneva, NY 14456, United States. 0161-813X/$ – see front matter # 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.neuro.2007.04.002

marketed since the introduction of IMI in 1991 (Elbert et al., 1990; Thyssen and Machemer, 1999). These neonicotinoids are acetamiprid (ACT) (Takahashi et al., 1992), nitenpyram (NIT) (Minamida et al., 1993), thiacloprid (THI) (Elbert et al., 2000), thiamethoxam (TMX) (Senn et al., 1998), clothianidin (CTD) (Ohkawara et al., 2002), and dinotefuran (DTF) (Kodaka et al., 1998; Wakita et al., 2003). Their structures (Fig. 1) are characterized by having an aromatic heterocyclic chloropyridinylmethyl (CPM), chlorothiazolylmethyl (CTM), or tetrahyrdofurylmethyl (TFM) moiety coupled with either a cyclic or an open-chain electronegative pharmacophore group containing an N-nitroimine, 2-nitromethylene or N-cyanoimine

830

J. Tan et al. / NeuroToxicology 28 (2007) 829–842

Fig. 1. Structures of nicotine, acetylcholine and seven neonicotinoids. In the neonicotinoids, a chloropyridinylmethyl (CPM) group is coupled with a cyclic Nnitroimine moiety in IMI and an N-cyanoimine moiety in THI, with an acyclic N-cyanoimine moiety in ACT and with a 2-nitromethylene moiety in NIT. A chlorothiazolylmethyl (CTM) group is coupled with a cyclic N-nitroimine moiety in TMX and an acyclic N-nitroimine moiety in CTD. A tetrahydrofurylmethyl (TFM) group is coupled with an acyclic N-nitroimine moiety in DTF. Nicotine and acetylcholine are two naturally occurring nAChR agonists. The three-letter abbreviations for these compounds are used in all figures and text.

moiety. Electrophysiological studies and radioligand binding assays have established that neonicotinoid insecticides share a common mode of action in selectively targeting insect nicotinic acetylcholine receptors (nAChRs), acting as agonists or antagonists (Buckingham et al., 1997; Matsuda et al., 2005; Nauen et al., 2001a; Tomizawa and Casida, 2005; Zwart et al., 1994). The described agonist actions of neonicotinoids on nAChRs include partial, full and super activities. The application of IMI activates nAChRs and elicits inward currents on cultured Kenyon cells of the honeybee, Apis mellifera, and on Drosophila-chicken SADb2 and ALSb2 hybrid nAChRs expressed in Xenopus laevis oocytes, acting as a partial agonist that elicits 24–55% of the maximum ACh-induced currents (De´glise et al., 2002; Ihara et al., 2003, 2004; Matsuda et al., 1998). However, on honeybee antennal lobe neurons, IMI can act either as a full agonist or as a partial agonist, suggesting that IMI has different actions on different subtypes of insect nAChRs (Nauen et al., 2001b). Other neonicotinoids, such as CTD or DTF, with an acyclic moiety corresponding to the imidazolidine moiety of IMI, induce either a greater maximum response than ACh or the same response as ACh on SADb2 hybrid nAChRs (Ihara et al., 2004; Kagabu et al., 2002). Thus, they are referred to as super or full agonists, respectively. This diversity of agonist activity of neonicotinoids appears to be related both to their structure and to the subtypes of nAChRs expressed by the target cells, but no comparison of all the commercial neonicotinoids has been made on a single insect neuronal preparation. It is still unclear whether neonicotinoids agonist efficacy (Imax) is related to insecticidal activity as efficacy does not necessarily parallel binding affinity as indicated by EC50 value (concentration producing half maximal response) (Ihara et al., 2004). However, Nishiwaki et al. (2003) demonstrated that the EC50 for neonicotinoids on the Da2b2 hybrid nAChRs

expressed in Xenopus oocytes, have a higher positive correlation with the insecticidal activity against houseflies (r = 0.893) than the Imax values, suggesting that affinity is a more important determinant of insecticidal activity than efficacy. Nevertheless, a contribution of agonist efficacy to insecticidal activity cannot be ruled out, since neonicotinoids with higher efficacy depolarize neurons to a greater degree than those with lower efficacy. Clothianidin and DTF show high insecticidal activity and induce greater maximum responses than IMI on Da2b2 hybrid nAChRs (Ihara et al., 2004; Kagabu et al., 2002), but their EC50 values on Da2b2 hybrid nAChRs measured by two electrode voltage clamp electrophysiology are lower than those of IMI. In order to elucidate in more detail the contribution of agonist action to the insecticidal activity of neonicotinoids, it is essential to measure their agonist actions on nAChRs expressed by native neurons from insects. In this paper, we studied the agonist actions and interactions of the seven commercial neonicotinoids on native nAChRs expressed by thoracic neurons from the American cockroach, Periplaneta americana. Our data provide an overall picture of the structural factors that contribute to the agonist actions on cockroach neuronal nAChRs. We also examined the potential relationship between their agonist actions and insecticidal activities. 2. Materials and methods 2.1. Chemicals All neonicotinoids except nitenpyram were obtained as analytical standards (99% purity) from the USEPA. Nitenpyram (analytical standard), nicotine hydrogen tartrate (98%), acetylcholine chloride (99%), methyllycaconitine citrate, atropine (99%) and piperonyl butoxide (90%) were obtained from Sigma–Aldrich (St. Louis, MO).

J. Tan et al. / NeuroToxicology 28 (2007) 829–842

831

2.2. Neuron isolation

2.4. Drug application

Neurons were isolated from the three thoracic ganglia of male adult American cockroaches, Periplaneta americana, which were purchased from Carolina Biological Supply Company (Burlington, NC) and maintained at room temperature (21–23 8C) with free access to water and dry dog food. The isolation of neurons from the ganglia was performed at room temperature using enzymatic digestion and mechanical dissociation as described by Salgado and Saar (2004). Briefly, cockroaches were immobilized using CO2, and then the head, legs and wings were removed. The insects were fixed dorsal side up with pins on a dissection dish. The dorsal cuticle, fat body, gut and some muscles were removed in order to access the ventral nerve cord. The three thoracic ganglia were carefully dissected and placed in cockroach dissection saline solution containing 185 mM NaCl, 3.0 mM KCl, 4 mM MgCl2, 10 mM D-glucose, and 10 mM HEPES, with pH adjusted to 7.2 with 1N NaOH. The ganglia, after being desheathed with forceps, were incubated for 20 min at room temperature in the cockroach dissection saline containing collagenase (Type IA, 0.5 mg/ml; Sigma–Aldrich). The ganglia were then rinsed three times with dissection saline supplemented with 5 mM CaCl2 and fetal calf serum (10% by volume; Sigma–Aldrich), and mechanically triturated through a series of pipettes of decreasing tip opening. The dissociated neurons, suspended in the supplemented dissection saline, were allowed to settle and adhere for at least 30 min on microscope cover glasses (1.2 cm diameter) coated with polyL-lysine (mol. wt. >30,000; Sigma–Aldrich) in 35 mm polystyrene petri dishes.

Drugs were applied in two ways. When establishing steadystate concentration–response curves, each agonist was applied at a series of concentrations for 20 s from a drug pipette by pressure ejection (8 psi) controlled by a Picospritzer II (General Valve Corporation, Fairfield, NJ). The drug pipette, with an opening of 5 mm, was aimed directly at the neuron from a distance of approximately 200 mm, perpendicular to the flow of external solution. Thus, the drug solution could be washed away from the neuron with fresh external solution soon after the termination of ejection. When testing antagonist actions, each compound was applied at a concentration of 100 mM by addition to the external solution for 10 min in order to reach equilibrium prior to agonist application. Competitive interaction between paired agonists was observed by measuring inward currents induced by pressure application of two compounds in combination at a concentration of 100 mM for each. Three neonicotinoids (IMI, DTF and NIT) of varied structure were selected to pair with the other neonicotinoids. IMI represents a CPM-cyclic N-nitroimine neonicotinoid, DTF represents a TFM-acyclic N-nitroimine neonicotinoid, and NIT represents a CPM-acyclic 2-nitromethylene neonicotinoid. All neonicotinoids were dissolved and diluted in DMSO to 1000 the desired test concentration, which was then directly diluted in cockroach saline, giving a final DMSO concentration <0.1%. Application of DMSO (0.1%) alone in saline had no effect on ACh-evoked currents. In all experiments, 1 mM atropine was included in the external solution in order to block muscarinic receptors (Salgado and Saar, 2004). 2.5. Data analysis

2.3. Single electrode voltage clamp recording Membrane currents were recorded using the single electrode voltage clamp recording method (Ren and Galligan, 2005; Salgado and Saar, 2004) at room temperature. The cover glass with dissociated neurons was transferred to a recording chamber (2 cm  4 cm) and superfused with cockroach dissection saline supplemented with 5 mM CaCl2 at a flow rate of 0.5 ml/min from the opening of a 500 mm internal diameter glass capillary tube (FHC, Bowdoinham, ME) placed 800 mm from the neuron. Individual neurons were visualized at 200 magnification using an inverted microscope (Olympus CK-2). Large neurons, 50–75 mm in diameter, were selected and impaled with microelectrodes fabricated on a PN-30 model puller (Narishige, Japan) and filled with 2 M KCl (tip resistance = 25–35 mV). An amplifier (Axoclamp 2A, Axon Instruments, Foster City, CA) was used to record membrane current in single electrode voltage clamp mode. Neurons were voltage-clamped at zero current potential (40 to 95 mV). The switching rate was adjusted to 5–6 kHz and the gain setting was 8–25. Recording conditions were optimized by adjusting capacitance neutralization. Signals were digitized and acquired using a MiniDigi 1A analog-digital converter (Axon Instruments) and Axoscope 9.0 software (Axon Instruments).

In all experiments, maximum inward currents (Imax) induced by agonists were normalized to the response caused by 100 mM ACh in the same neuron. The normalized data from each neuron were fitted to the Hill equation using Origin 7 software (OriginLab Corp., Northampton, MA) as follows:  Y ¼ Emax

Xn ðECn50 þ X n Þ



where Emax is the maximum response, EC50 the concentration producing a half-maximum normalized response, n the Hill coefficient, and X is the agonist concentration. Each EC50 value was converted into a pEC50 value (pEC50 = log EC50). The mean value of pEC50 and its standard error (S.E.M.) were calculated from a set of repetitions in different neurons. Data are expressed as the mean  S.E.M. ANOVA (Tukey’s) oneway test was used to establish significant differences among treatments. A statistically significant difference was considered to exist at the P < 0.05 level. 2.6. Insecticidal assay Insecticidal tests against adult male B. germanica and P. americana were conducted as described by Mori et al. (2002)

832

J. Tan et al. / NeuroToxicology 28 (2007) 829–842

and Kiriyama and Nishimura (2002). Adult male B. germanica from a laboratory susceptible strain were topically treated with 1 ml acetone containing 30 mg piperonyl butoxide (PBO) to inhibit insecticide metabolism through the activity of cytochrome P450 monooxygenases. After 1 h, 0.2 ml DMSO solution containing a test compound (4 nmole/g cockroach) was injected into the thorax through the membrane between the second and the third abdominal segment using a Hamilton syringe dispenser. Thirty German cockroaches were tested for each compound. With adult male P. americana, 2 ml of acetone containing 60 mg PBO was placed on the dorsal thorax and the first abdominal segment. After 1 h, 0.4 ml DMSO solution containing different concentration of test compound was injected as for B. germanica. Four to eight American cockroaches were injected for each of five to six doses chosen to cover the range of 10–90% mortality based on initial toxicity tests. The injected cockroaches were maintained at room temperature along with a piece of braided cotton roll (Richmond Dental, Charlotte, NC) soaked in water. Mortality was recorded 48 h after the injection as judged by prostration of the insect with little if any movement of appendages. LD50 values to American cockroaches were estimated using MicroProbit 3.0 software. In order to verify the toxicity, LD50 values of IMI and THI were also estimated using the ‘up-and-down’ method with 13 consecutive doses increased by 0.3 log unit (Dixon and Mood, 1948, Dixon, 1991). No mortality was observed in control group in which only DMSO were injected. 3. Results 3.1. Inward currents caused by ACh and neonicotinoids ACh (100 mM) evoked an inward current in all neurons tested at potentials between 40 and 95 mV (Fig. 2). AChinduced currents remained stable during repeated ACh application for up to 120 min (data not shown), thus allowing the establishment of full concentration–response curves in a single neuron without correction for run-down or desensitization. The inward currents were characterized by a rapid rising phase with a peak amplitude of up to 15 nA followed by partial desensitization. The currents returned to zero steady-state level after washing the neurons with drug-free saline. ACh-induced currents were reversibly blocked by the bath application of methyllycaconitine (MLA, 0.1 mM) (Fig. 3), indicating that the currents were mediated by nAChRs. As shown, the majority of records revealed a relatively slow desensitization and, at most, only a small non-desensitizing component of the type described by Salgado and Saar (2004). Nicotine (NIC) and all neonicotinoids, except TMX, induced inward currents (Fig. 2), confirming the agonistic effects of these compounds on cockroach neuronal nAChRs. Bath application of MLA reversibly blocked all neonicotinoidinduced currents (see Fig. 3D for an example of the inhibition of the IMI-induced current). Although these currents also showed a rapid rising phase similar to those induced by ACh, they were slower in returning to a zero steady-state level after agonist washout. CTD and NIT were the most obvious cases (Fig. 2F

and G). Therefore, the dissociation of nicotine and other neonicotinoids from their binding sites was slower than that of ACh. Based on the maximum current, agonists could be divided into two broad groups. One group, comprised of NIC, IMI and THI (Fig. 2A–C), caused maximum currents that were equivalent to 22  1, 21  1 and 22  1% of the maximum ACh current, respectively (Table 1). The second group, comprised of ACT, DTF, CTD, and NIT (Fig. 2E–H), induced maximum currents that were 61  2, 80  2, 100  3, and 98  3% of the maximum ACh currents, respectively (Table 1). Therefore, NIC, IMI and THI are relatively weak partial agonists, whereas ACT, CTD, DTF and NIT are much more effective or full agonists. Interestingly, TMX failed to induce a significant inward current (Fig. 4D), although ACh, IMI and CTD produced currents in the same neuron (Fig. 4A–C). Further, the bath application of TMX (100 mM) did not affect the currents induced by ACh, IMI and CTD (Fig. 4E–G). In sharp contrast to TMX, the bath application of CTD (100 mM), a compound similar to TMX in aromatic CTM structure, desensitized the receptors and completely inhibited the currents caused by ACh, IMI and CTD (Fig. 4I–K). It is therefore evident that TMX was not a direct-acting nAChR agonist and did not interact with these nAChRs. 3.2. Agonist concentration–response curves All compounds, except TMX, activated cockroach neuronal nAChRs in a concentration-dependent manner (Fig. 5). The EC50 values of ACh (15.0  3.3 mM, n = 4) and DTF (7.8  0.9 mM, n = 4) were significantly higher than those of the other compounds. The EC50 values of NIC, IMI, and ACT were 5.1  1.0 mM (n = 3), 2.5  0.8 mM (n = 3), and 3.7  0.7 mM (n = 5), respectively. There were no significant differences between these values. THI, CTD and NIT composed another group with the lowest EC50 values of 1.4  0.1 mM (n = 4), 1.6  0.1 mM (n = 3), and 1.5  0.5 mM (n = 4), respectively. In contrast to the EC50 values, there were no statistical differences between the Hill coefficients of these agonists, which were indistinguishable from 1.0. 3.3. Interactions between neonicotinoids Interactions between these neonicotinoids were studied by measuring inward currents caused by agonist pairs each at a concentration of 100 mM which, when applied individually, gave a maximum response. Fig. 6 shows the competitive action of IMI on nAChRs when it was paired with other agonists. IMI acted as a partial agonist inducing 21  1% of the maximum ACh current (Table 1). Paired application of IMI with ACh, NIC, THI, and DTF, respectively, evoked similar inward currents as those caused by IMI alone (Fig. 6A–D), whereas, the combinations of IMI with ACT or CTD produced a larger inward current than IMI alone, but smaller than ACT or CTD alone (Fig. 6E and F). These results indicate that their agonistic actions occur primarily at the same site of action and thereby a competitive interaction exists between them. Together with the

J. Tan et al. / NeuroToxicology 28 (2007) 829–842

833

Fig. 2. Representative inward currents caused by ACh, nicotine and neonicotinoids at 100 mM concentration on the same neuron. Each agonist was applied from a drug pipette placed 200 mm from the neuron by pressure ejection (8 psi) for 20 s using a Picospritzer II. The bar above each trace denotes the time of addition of either ACh or neonicotinoid insecticide. At the same time, the external solution was keeping running at 0.5 ml/min from a glass tube placed 800 mm from the neuron. The pressure ejection drug solution flow from pipette was perpendicular to the flow of external solution. Thus, the agonist could be washed off from the neuron by fresh external solution soon after the termination of ejection. The maximum current induced by each agonist was normalized to the maximum currents of ACh on the same neuron. The Imax mean values and their standard errors (S.E.M.) are summarized in Table 1.

EC50 measurements (Table 1), these data suggest that the size of the inward currents caused by paired agonist application were mainly determined by the agonist with higher binding affinity, rather than by the agonist with higher agonist efficacy. Figs. 7 and 8 show very similar results when DTF or NIT was paired with the other agonists: the inward currents were predominantly determined by the agonist that had the higher binding affinity (i.e. lower EC50). For instance, when DTF was combined with NIC, IMI, THI and ACT, the inward currents were determined by NIC, IMI, THI and ACT (Fig. 7A–D), rather than by DTF itself, due to its lower binding affinity. When DTF was combined with two full-agonists, ACh or CTD,

the inward currents were still determined by the agonist with higher binding affinity. In the paired-application of NIT, one of the highest affinity agonists, with ACh, ACT, DTF and NIC, respectively, the inward currents were determined by NIT (Fig. 8A–D). However, when NIT was co-applied with NIC, IMI, or THI, inward currents were determined by both of the co-applied drugs (Fig. 8E–G) as their binding affinities were similar. Therefore, the binding affinity, rather than efficacy determined inward currents caused by a pair of agonists acting at the same site(s). In no case did the paired application of agonists cause a current greater than that induced by the most efficacious agonist of the pair, indicating that compound-

834

J. Tan et al. / NeuroToxicology 28 (2007) 829–842

Fig. 3. Representative inward currents caused by ACh and IMI at 100 mM concentration before and after bath application of methyllycaconitine (MLA) at 100 nM concentration. ACh and IMI were applied from a drug pipette by pressure ejection for 20 s as described in Fig. 2. MLA was applied in the external solution 10 min prior to the recording of ACh or IMI current in order to reach an equilibrium status. Inward currents were induced by pressure ejection of ACh (A) and IMI (B) at 100 mM concentration, respectively, before MLA was applied in the external solution. No current was caused by ACh (100 mM) (C) or IMI (100 mM) (D) after MLA (100nM) was applied. An ACh current (E) and IMI current (F) were again recorded after the neuron was washed for 10 min with MLA-free external solution.

specific, independently acting receptor types were probably not present. 3.4. Insecticidal activity The insecticidal activities of six neonicotinoids were determined by injecting them into PBO-pretreated P. americana. In American cockroaches injected with a toxic dose of a test compound, two types of response were seen. In Type A, the onset of intoxication began with strong excitatory symptoms, including uncoordinated quivering, hyper-excitability and rapid spontaneous movements, wing flexing, violent body shaking

and leg tremors. Finally, the cockroaches were unable to right themselves after falling over, became prostrate with weak uncoordinated movements, and eventually died. At lower doses, this knock-down was reversible and recovery occurred. In the Type B response, almost no excitatory effect was seen. The roaches slowly became immobile with some postural changes (lowered head) and then became paralyzed. They eventually died without experiencing knockdown. Interestingly, there was a structural correlation among the compounds causing these two types of response. Neonicotinoids with a heterocyclic electronegative moiety (IMI and THI) always caused Type A symptoms. Nicotine also caused rapid excitatory

Table 1 Mean  S.E.M. values for EC50, Hill coefficient and maximum current Compound Acetylcholine Nicotine Imidacloprid Thiacloprid Acetamiprid Dinotefuran Clothianidin Nitenpyram a

pEC50 (M)a mean  S.E.M. A

4.85  0.09 5.31  0.08BC 5.63  0.13C 5.86  0.01E 5.46  0.09BCDF 5.12  0.04ABD 5.79  0.01EF 5.72  0.12E

Hill coefficient mean  S.E.M. 1.37  0.09 1.09  0.13 1.02  0.16 1.02  0.17 1.06  0.12 1.18  0.15 1.05  0.17 1.12  0.16

N 4 3 3 4 5 4 3 4

Imaxb mean  S.E.M. D

100 21.91 20.69 21.87 61.47 80.33 99.77 98.28

(0.81)A (0.94)A (1.30)A (2.32)B (2.31)C (1.81)D (2.99)D

N 96 17 24 19 14 20 28 11

pEC50 = log EC50. The EC50 value was estimated from the Hill equation fitting of dose–response curve recorded from each neuron, and then the mean value of pEC50 and its standard error (S.E.M.) were calculated from the number of individual neurons tested as indicated by N. The ANOVA (Tukey’s test) one-way test showed there are significant differences between compounds at p < 0.05 level (d.f. = 7, F = 21.59). The significance of differences is indicated by the letters A, B, C, D, E, or F. The pEC50 values with same letter indicate no significant difference statistically. b Maximum current (Imax) was the maximum inward current induced by each agonist at a concentration of 100 mM on each neuron, normalized to the maximum inward current of 100 mM ACh on the same neuron. The ANOVA (Tukey’s test) one-way test showed there are significant differences between compounds at p < 0.05 level (d.f. = 7, F = 466.48). The significant difference is marked by letter A, B, C, or D. The Imax values with same letter indicate no significant difference statistically.

J. Tan et al. / NeuroToxicology 28 (2007) 829–842

835

Fig. 4. Representative inward currents induced by ACh, IMI and CTD at 100 mM concentration before and after application of TMX and CTD at 100 mM concentration. Before application TMX or CTD, pressure ejection of ACh (A), IMI (B) and CTD (C) caused significant currents. TMX (D) did not induce a significant current. These currents were not inhibited by bath application of 100 mM TMX for 10 min (E–H). In contrast to TMX, bath application of CTD with external solution for 10 min could reversibly inhibit the currents caused by ACh (I), IMI (J) and CTD (K).

symptoms of the same kind. By contrast, three compounds with an acyclic electronegative group (CTD, DTF, NIT) caused Type B symptoms. ACT, also with an acyclic group, was the exception to this correlation and caused excitation (Type A) symptoms rather than depression and paralysis. When tested in German cockroaches, IMI and THI caused no mortalities at a dose of 4 nmole/g cockroach. However, DTF, CTD, NIT and ACT produced 90  7, 77  10, 40  7, and 20  7% mortality, respectively (Table 2). For American cockroaches, the LD50 values are 2.00 (THI), 0.70 (IMI), 0.42 (ACT), 0.28 (DTF), 0.18 (CTD), 0.33 (NIT) nmole/g insect (Table 2). THI showed significantly lower toxicity (11-fold) than CTD. When the LD50 values were estimated using the ‘up and down’ method, THI also showed an 11-fold lower toxicity than CTD (Table 2). These LD50 values again showed a positive correlation (R = 0.90) with agonist efficacy, but not with agonist affinity (Fig. 9A and B). 4. Discussion Fig. 5. Concentration–response curves for inward currents induced by eight nAChRs agonists. Data were expressed as a percentage of the current amplitude caused by ACh (100 mM) in each neuron. Points are mean  S.E.M. response amplitude from the number (N) of neurons as shown in Table 1. Curves were fitted to the data points using the Hill equation.

4.1. Agonist efficacy and structure Although all neonicotinoid insecticides act selectively on insect nAChRs (Matsuda et al., 2001, 2005; Tomizawa and

836

J. Tan et al. / NeuroToxicology 28 (2007) 829–842

Fig. 6. Representative inward currents induced by several agonists at 100 mM concentrations applied alone and paired with IMI. All agonists and their combinations with IMI were applied from a drug pipette by pressure ejection as described in Fig. 2.

Casida, 2003, 2005), their agonist actions vary significantly from partial efficacy to full or super efficacy with different insect-related receptors (Brown et al., 2006; De´glise et al., 2002; Ihara et al., 2003, 2004; Kagabu et al., 2002; Matsuda et al., 1998; Nauen et al., 2001b). Among the chemical structures of the neonicotinoids tested in this study, there are three heterocyclic aromatic moieties, CPM, CTM and TFM, coupled, respectively, to either a cyclic or an acyclic Nnitroimine moiety (e.g. IMI, TMX, CTD and DTF), Ncyanoimine moiety (e.g. THI and ACT), or a 2-nitromethylene moiety (e.g. NIT) (Fig. 1). The structural diversity of the neonicotinoids gives rise to questions of the relationship of their structural elements to receptor affinity and efficacy, and the relationship of these to their toxicity to insects. In the present study, we observed that the compounds tested can be divided into two groups based on their maximum inward currents on cockroach neuronal nAChRs. One group (NIC, IMI and THI) are rather weak partial agonists, while another group (NIT, CTD, DTF and ACT) are full or nearly full agonists. Comparing their structures (Fig. 1), this division in agonist efficacy is related to whether the electronegative pharmacophore moiety is cyclic or acyclic. IMI and THI have the cyclic N-nitroimine, or N-cyanoimine moiety, respectively, and they

are partial agonists. NIC is also a weak partial agonist and has a cyclic pyrrolidine group in its structure. All the high-efficacy agonists have an acyclic electronegative moiety, i.e. Nnitroimine, 2-nitromethylene, or N-cyanoimine. Ihara et al. (2006) studied the effects of a series of imidacloprid (cyclic) and clothianidin (acyclic) analogs on nicotinic receptors using neurons from the terminal abdominal ganglia of P. americana and patch clamp methodology. They similarly found that the acyclic analogs had higher efficacies than their corresponding cylic analogs. It is possible that the open chain structure provides greater flexibility in this part of the molecule allowing it to bind in a way that enhances agonist efficacy. The role of the aromatic moiety (CPM, CTM or TFM) is not as significant as the electronegative pharmacophore structure in determining the agonist efficacy, since IMI, a weak partial agonist, and NIT, a full agonist, share the same CPM moiety. No obvious correlations between molecular structure and receptor affinity (EC50) were evident. 4.2. Agonist affinity and competition There is a potential for a high diversity of insect nAChRs due to the existence of 10 or more receptor subunit genes in several

J. Tan et al. / NeuroToxicology 28 (2007) 829–842

837

Fig. 7. Representative inward currents induced by several agonists at 100 mM concentrations applied alone and paired with DTF. All agonists and their combinations with DTF were applied from a drug pipette by pressure ejection as described in Fig. 2.

insect species (Gundelfinger and Schulz, 2000; Jones et al., 2005; Littleton and Ganetzky, 2000). Further, alternative splicing and pre-mRNA editing provide for additional diversity of insect nAChRs (Grauso et al., 2002; Jones et al., 2005; Lansdell and Millar, 2000; Sattelle et al., 2005). Functionally, several receptor subtypes with differing responses to agonists and antagonists have been reported in cockroaches (e.g. see Buckingham et al. (1997); Salgado and Saar (2004)). However, the competition studies conducted here with different paired neonicotinoids suggested that the agonistic activities of all the neonicotinoids tested, except for TMX, are probably due primarily to their actions on the same site(s) or subunit(s) of the nAChRs present and no additional currents were observed with any pair combinations compared to those induced by the compounds individually. The maximum inward current of two neonicotinoids in combination at equal concentration was primarily determined by their relative affinities. Further, no evidence for the existence of receptors with varied affinities for any of these compounds were evident from the concentration–

activity curves since the Hill coefficients were very close to 1.0 in each case. Two special cases are discussed further below. 4.3. Dinotefuran and clothianidin Dinotefuran (DTF) and clothianidin (CTD) share the same acyclic N-nitroguanidine moiety but they differ considerably in the presence of a TFM moiety in DTF and a CTM moiety in CTD (Fig. 1). The high insecticidal activity of synergized DTF to German cockroaches (Mori et al., 2002; Miyagi et al., 2006) is somewhat unexpected based on the data from ligand binding assays with American cockroach nerve cord membrane preparations using labeled epibatidine (Mori et al., 2002) which showed that DTF was much less effective as a competitor for binding than other neonicotinoids. Electrophysiological studies with the isolated nervous system of P. americana (Kiriyama et al., 2003) and with Drosophila-chicken SADb2 hybrid nAChRs expressed in Xenopus oocytes (Kagabu et al., 2002) also indicate a potency for DTF in stimulating and/or

838

J. Tan et al. / NeuroToxicology 28 (2007) 829–842

Fig. 8. Representative inward currents induced by several agonists at 100 mM concentrations applied alone and paired with NIT. All agonists and their combinations with NIT were applied from a drug pipette by pressure ejection as described in Fig. 2.

blocking activity that is somewhat lower than or equal to that of other neonicotinoids. Recently, Miyagi et al. (2006) studied the binding of labeled dinotefuran itself to American cockroach whole nerve cord membranes and concluded that a distinct high affinity binding site for dinotefuran exists which could explain its unexpectedly high toxicity. A recent study by Honda et al. (2006) also concluded that DTF has a unique mode of

interaction with nAChRs in the leafhopper Homalodisca and concluded that the binding orientation or recognition site for the TFM moiety of DTF may not be identical to that for the CPM moiety in IMI or the CTM moiety in CTD. However, this did not seem to be true with receptors from Drosophila melanogaster in which all neonicotinoids, including DTF, appeared to interact with the same site. This suggests that

J. Tan et al. / NeuroToxicology 28 (2007) 829–842

839

Table 2 Toxicity of neonicotinoids against American and German cockroaches and their symptoms Compound

B. germanicac

P. americana LD50 (nmole/g)a

Types of symptomsb

Mortality mean  S.E.M.

Nicotine Imidacloprid

Not determined 0.70 (0.45–1.57)

A A

Not determined 0

Thiacloprid

2.00 (1.23–3.78) 3.90 (2.31–6.58) d

A

0

Acetamiprid Dinotefuran

0.42 (0.27–0.57) 0.28 (0.15–0.46)

A B

20  7 90  7

Clothianidin

0.18 (0.12–0.31) 0.34 (0.19–0.60) d

B

77  10

Nitenpyram

0.33 (0.21–0.73)

B

40  7

a b c d

LD50 values were estimated using Micro-Probit 3.0 software. 95% confidence limits in parentheses. A, excitation symptoms; B, depression/paralytic symptoms. The two types are described in the text. Mortality was measured at a dose of 4 nmole/g beetle (see text). LD50 values were estimated using ‘up-and-down’ method.

homopterans may have atypical receptors compared to other insects, but leaves in question to what extent DTF acts differently from other neonicotinoids in other types of insects. In the present study with P. americana, we did not find any clear evidence that DTF interacts with a unique group of nAChRs. The competition pattern of DTF with other neonicotinoids is very similar to that of IMI or NIT. No additional current was obtained when DTF was combined with other neonicotinoids at a maximally stimulating concentration which might have indicated the existence of an additional uniquely DTF-sensitive group of receptors. Nor did the concentration–response curve for DTF show evidence for receptors with multiple binding affinities for DTF. The Hill coefficient for DTF was approximately 1.0 as with the other neonicotinoids. It is possible that the DTF-specific receptors reported in P. americana by Miyagi et al. (2006) are located outside the thoracic ganglia we examined and, since the DTFspecific receptors reported by Miyagi et al. were present at low density, and therefore were probably a minor component of the total nAChR population in the nerve cord, their responses might not have been sufficient to detect in our system. The affinity (EC50) of DTF is significantly lower than that of CTD and most

of the other neonicotinoids, though it is a highly effective agonist. This agrees with the result of (Kagabu et al., 2002) using Drosophila-chicken SADb2 hybrid nAChRs. Our result is also in general agreement with the whole nerve cord recordings from P. americana made by Kiriyama and Nishimura (2002) who reported no effects for DTF that were quantitatively or qualitatively different from those of other neonicotinoids such as IMI and CTD. One further complication in assessing the varying results with DTF in cockroaches is that DTF is notably more toxic to German cockroaches than IMI (Mori et al., 2002 and Table 2). However, the difference is much lower in American cockroaches (Kiriyama and Nishimura, 2002 and Table 2). Correlations between receptor binding studies in P. americana and toxicity studies in B. germanica could therefore be misleading. 4.4. Thiamethoxam and clothianidin TMX (Maienfisch et al., 2001a,b) and CTD (Jeschke et al., 2003) are second-generation commercial neonicotinoids which have an identical CTM moiety in their structures. They exhibit higher insecticidal activity than IMI against both

Fig. 9. Correlations between agonist affinity (pEC50) (A) and agonist efficacy (Imax) (B) and insecticidal activity to American cockroaches (pLD50 = log LD50). The data are shown in Tables 1 and 2. The error bars of the Y-axis represent the 95% confidence limits of pLD50s. The error bars of the X-axis represent the standard errors of pEC50s (A) and Imax (B).

840

J. Tan et al. / NeuroToxicology 28 (2007) 829–842

chewing and sucking insects. TMX differs from CTD only in having a cyclized rather than an acyclic N-nitroguanidine moiety (Fig. 1). Based on non-competitive behavior in binding assays with aphid membrane preparations, it has been claimed that TMX may interact with a different site or group of nAChRs than CTD and IMI (Kayser et al., 2004; Wellmann et al., 2004). In this study, although CTD was highly effective, TMX failed to activate the cockroach neuronal nAChRs. This is in accord with the idea that cyclization of the electronegative moiety in these molecules disfavors efficacy, but in this case it appears that TMX may not even bind to the receptors since it does not antagonize the effects of other neonicotinoids. The effects of TMX on cockroach neurons were also studied by Nauen et al. (2003) on both the desensitizing (nAChD) and nondesensitizing (nAChN) receptors described by Salgado and Saar (2004). Bath-applied TMX (30 mM) had no agonistic activity against the nAChN, but with the nAChD, a relatively weak agonistic action was observed. Our finding is consistent with studies using whole-cell voltage clamp on neuronal cell bodies from Helithis virescens larvae where no current was induced by TMX at concentrations as high as 300 mM (Nauen et al., 2003). The lack of a potent agonist action of TMX on native nAChRs from several insects, coupled with the observation that TMX is readily converted to CTD in insect tissues, support the conclusion that TMX acts as a proinsecticide in many insects (Nauen et al., 2003), although one cannot eliminate the possibility that a subset of receptors in aphids and other homopterans responds to TMX directly. 4.5. Agonist actions and toxicity The relationship between the actions of neonicotinoids on nAChRs and their toxic effects in insects is complex, largely unclear, and may vary between different insect species (Salgado and Saar, 2004). An initial stimulation of many of the receptors may relate to some of the early excitatory effects observed, but desensitization follows rapidly and leads to persistent receptor block. Other nAChRs are more resistant or completely recalcitrant to desensitization and may continue to respond in the presence of agonists until a depolarizing block of transmission could occur. The presence of multiple receptor types with different responses to nicotinic stimulation and the lack of detailed knowledge of the specific roles of these receptors in the insect nervous systems in different insects are barriers to understanding the relationship between receptor responses in vitro and the poisoning process in vivo. Attempts to correlate poisoning symptoms or lethality with neonicotinoid structure and effects on isolated nervous systems may provide useful information but it may also oversimplify a very complex situation. A series of studies with structurally varied neonicotinoids that correlate their potency in stimulating and then blocking nervous activity in isolated nerve cords from P. americana with their lethality have suggested that lethality is more related to the neuroblocking activity of a series of neonicotinoids than to the neuroexcitatory activity (Kiriyama and Nishimura, 2002).

Our results, which are limited to a quite small group of compounds with a relatively low variation in potency, showed no correlation between receptor affinity (pEC50) and insecticidal activity against P. Americana or B. germanica (Table 2 and Fig. 9). However, a positive correlation was found between agonist efficacy and insecticidal activity (Fig. 9), indicating that agonist efficacy is likely to be an important factor in the insecticidal activity of neonicotinoids. This is not surprising because an agonist with a higher efficacy will open ion channels and depolarize the membrane to a greater degree than one with a lower efficacy. In their study with P. americana neurons, Ihara et al. (2006) also concluded that the synergized insecticidal potencies of IMI and CTD analogs had no correlation with their agonist affinities (pEC50) and speculated that this could be due to differences in the hydrophobicity, and thus in the bioavailability, of the compounds. In contrast, the pEC50 values for several neonictinoids on SAD/b2 hybrid nAChRs expressed in Xenopus oocytes shows a higher correlation than the Imax value with the insecticidal activity to houseflies (Nishiwaki et al., 2003), suggesting that affinity is more important than the efficacy for nAChRs in killing these insects. In fact, it seems logical that both high affinity and high efficacy should favor toxicity since both would favor an enhanced ability to impact receptors. Therefore, additional studies are needed to fully understand how agonist binding affinity and binding efficacy, agonist activity and prolonged receptor desensitization, and the possible involvement of multiple and functionally varied receptor subtypes are related to the in vivo responses and insecticidal activity of neonicotinoids. Our observations raised one other interesting point. The compounds with a cyclic electronegative group (IMI, THI and nicotine) produced the same Type A poisoning syndrome, i.e. strong excitation symptoms followed by prostration and death as described by Schroeder and Flattum (1984). However, the acyclic compounds, CTD, DTF, NIT, did not induce such excitation symptoms before paralysis and death (Type B). Kiriyama and Nishimura (2002) also briefly noted such a difference of symptoms between IMI (Type A) and CTD (Type B). The obvious difference in the poisoning syndromes among these neonicotinoids matches well with their agonist efficacies (Imax) on cockroach neuronal nAChRs. The only exception is ACT, an acyclic analog that causes Type A symptoms. However, ACT is the least efficacious analog in the acyclic group, so it may be that Type B symptoms are only caused by compounds which are essentially full agonists. Compounds with very high efficacy (CTD, DTF, NIT) may cause such a short period of receptor excitation before neuroblocking actions predominate that the symptoms observed are primarily depressive and paralytic. Acknowledgements We would like to thank Dr. Vincent L. Salgado (BASF, Research Triangle Park, NC) for his helpful advice and discussions, and colleagues in Dr. James J. Galligan’s laboratory for their help in the course of this project. This

J. Tan et al. / NeuroToxicology 28 (2007) 829–842

project was supported by NRI grant no. 2003-35302-13569 from USDA-CSREES. References Brown LA, Ihara M, Buckingham SD, Matsuda K, Sattelle DB. Neonicotinoid insecticides display partial and super agonist actions on native insect nicotinic acetylcholine receptors. J Neurochem 2006;99:608–15. Buckingham S, Lapied B, LeCorronc H, Grolleau F, Sattelle DB. Imidacloprid actions on insect neuronal acetylcholine receptors. J Exp Biol 1997;200:2685–92. De´glise P, Gru¨newald B, Gauthier M. The insecticide imidacloprid is a partial agonist of the nicotinic receptor of honeybee Kenyon cells. Neuroscience Lett 2002;321:13–6. Dixon WJ. Staircase bioassay: the up-and-down method. Neurosci Biobehav Rev 1991;15:47–50. Dixon WJ, Mood AM. A method for obtaining and analyzing sensitivity data. J Am Stat Assoc 1948;48:109–26. Elbert A, Overbeck H, Iwaya K, Tsuboi S. Imidacloprid: a novel systemic nitromethylene analog insecticide for crop protection. Proc Brighton Crop Prot Conf—Pests Dis 1990;1:21–8. Elbert A, Erdelen C, Kuhnhold J, Nauen R, Schmidt HW. Thiacloprid: a novel neonicotinoid insecticide for foliar application. Proc Brighton Crop Prot Conf—Pests Dis 2000;1:21–7. Grauso M, Reenan RA, Culetto E, Sattelle DB. Novel putative nicotinic acetylcholine receptor subunit genes, Da5, Da6 and Da7, in Drosophila melanogaster identify a new and highly conserved target of adenosine deaminase acting on RNA-mediated A-to-I Pre-mRNA editing. Genetics 2002;160:1519–33. Gundelfinger ED, Schulz R. Insect nicotinic acetylcholine receptors: genes, structure, physiological and pharmacological properties. In: Clementi F, Fornasari D, Gotti C, editors. Handbook of Experimental Pharmacology Neuronal Nicotinic Receptor, 144. Berlin: Springer; 2000. p. 497–521. Honda H, Tomizawa M, Casida JE. Insect nicotinic acetylcholine receptors: neonicotinoid binding site specificity is usually but not always conserved with varied substituents and species. J Agric Food Chem 2006;54:3365–71. Ihara M, Brown LA, Ishida C, Okuda H, Sattele DB, Matsuda K. Actions of imidacloprid, clothianidin and related neonicotinoids on nicotinic acetylcholine receptors of American cockroach neurons and their relationships with insecticidal potency. J Pestic Sci 2006;31:35–40. Ihara M, Matsuda K, Otake M, Kuwamura M, Shimomura M, Komai K, et al. Diverse actions of neonicotinoids on chicken a7, a4b2 and Drosophilachicken SADb2 and ALSb2 hybrid nicotinic acetylcholine receptors expressed in Xenopus laevis oocytes. Neuropharmacology 2003;45:133–44. Ihara M, Matsuda K, Shimomura M, Sattelle DB, Komai K. Super agonist actions of clothianidin and related compounds on the SADb2 nicotinic acetylcholine receptor expressed in Xenopus laevis oocytes. Biosci Biotechnol Biochem 2004;68:761–3. Jeschke P, Uneme H, Benet-Buchholz J, Sto¨lting J, Sirges W, Beck ME, et al. Clothianidin (TI-435)—the third member of the chloronicotinyl insecticide (CNITM) family. Pflanzenschutz-Nachrichten Bayer 2003;56:5–25. Jones AK, Grauso M, Sattele DB. The nicotinic acetylcholine receptor gene family of malaria mosquito, Anopheles gambiae. Genomics 2005;85:176– 87. Kagabu S. Chloronicotinyl insecticides-discovery, application and future perspective. Rev Toxicol 1997;1:75–129. Kagabu S, Matsuda K, Komai K. Preparation of dinotefuran related compounds and agonistic action on SADb2 hybrid nicotinic acetylcholine receptors expressed Xenopus laevis oocytes. J Pestic Sci 2002;27:374–7. Kayser H, Lee C, Decock A, Baur M, Haettenschwiler J, Maienfisch P. Comparative analysis of neonicotinoid binding to insect membranes. I. A structure–activity study of the mode of [3H]imidacloprid displacement in Myzus persicae and Aphis craccivora. Pest Manag Sci 2004;60: 945–58. Kiriyama K, Nishimura K. Structural effects of dinotefuran and analogues in insecticidal and neural activities. Pest Manag Sci 2002;58:669–76.

841

Kiriyama K, Nishiwaki H, Nakagawa Y, Nishimura K. Insecticidal activity and nicotinic acetylcholine receptor binding of dinotefuran and its analogues in the housefly, Musca domestica. Pest Manage Sci 2003;59:1093–100. Kodaka K, Kinoshita K, Wakita T, Kawahara N, Yasui N. MIT-446: a novel systemic insect control compound. Proc Brighton Crop Prot Conf—Pests Dis 1998;1:21–6. Lansdell SJ, Millar NS. The influence of nicotinic receptor subunit composition upon agonist, a-bungarotoxin and insecticide (imidacloprid) binding affinity. Neuropharmacology 2000;39:671–9. Littleton JT, Ganetzky B. Ion channels and synaptic organization: analysis of the Drosophila genome. Neuron 2000;26:335–43. Maienfisch P, Angst M, Brandl F, Fischer W, Hofer D, Kayser H, et al. Chemistry and biology of thiamethoxam: a second generation of neonicotinoid. Pest Manage Sci 2001a;57:906–13. Maienfisch P, Huerlimann H, Rindlisbacher A, Gsell L, Dettwiler H, Haettenschwiler J, et al. The discovery of thiamethoxam: a second-generation neonicotinoid. Pest Manage Sci 2001b;57:65–76. Matsuda K, Buckingham SD, Freeman JC, Squire MD, Baylis HA, Sattelle DB. Effects of the a subunit on imidacloprid sensitivity of recombinant nicotinic acetylcholine receptors. Br J Pharmacol 1998;123:518–24. Matsuda K, Buckingham SD, Kleier D, Rauh JJ, Grauso M, Sattelle DB. Neonicotinoid insecticides acting on insect nicotinic acetylcholine receptors. Trends Pharmacol Sci 2001;22:573–80. Matsuda K, Shimomura M, Ihara M, Akamatsu M, Sattelle DB. Neonicotinoids show selective and diverse actions on their nicotinic receptor targets: electrophysiology, molecular biology, and receptor modeling studies. Biosci Biotechnol Biochem 2005;69:1442–52. Minamida I, Iwanaga K, Tabuchi T, Aoki I, Fusuka T, Ishizuka H, et al. Studies on acyclic nitroethene compounds. Part 2. Synthesis and insecticideal activity of acyclic nitroethene compounds containing a heteroarylmethylamino group. J Pestic Sci 1993;18:41–8. Miyagi S, Komaki I, Ozoe Y. Identification of a high-affinity binding site for dinotefuran in the nerve cord of the American cockroach. Pest Manage Sci 2006;62:293–8. Mori K, Okumoto T, Kawahara N, Ozoe Y. Interaction of dinotefuran and its analogues with nicotinic acetylcholine receptors of cockroach nerve cords. Pest Manage Sci 2002;58:190–6. Nauen R, Ebbinghaus-Kintscher U, Elbert A, Jeschke P, Tjeten K. Acetylcholine receptors as sites for developing neonicotinoid insecticides. In: Ishaaya I, editor. Biochemical sites of insecticide action and resistance. Berlin: Springer-Verlag; 2001. p. 77–105. Nauen R, Ebbinghaus-Kintscher U, Schmuck R. Toxicity and nicotinic acetylcholine receptor interaction of imidacloprid and its metabolites in Apis mellifera (Hymenoptera: Apidae). Pest Manage Sci 2001b;57:577–86. Nauen R, Ebbinghaus-Kintscher U, Salgado VL, Kassmann M. Thiamethoxam is a neonicotinoid precursor converted to clothianidin in insect and plants. Pestic Biochem Physiol 2003;76:55–69. Nishiwaki H, Nakagawa Y, Kuwamura M, Sato K, Akamatsu M, Matsuda K, et al. Correlations of the electrophysiological activity of neonicotinoids with their binding and insecticidal activities. Pest Manage Sci 2003;59:1023–30. Ohkawara Y, Akayama A, Matsuda K, Andersch W. Clothianidin, a novel broad spectrum neonicotinoid insecticide. Proc Brighton Crop Prot Conf—Pests Dis 2002;1:51–8. Ren J, Galligan JJ. Dynamics of fast synaptic excitation during trains of stimulation in myenteric neurons of guinea-pig ileum. Auton Neurosci 2005;117:67–78. Salgado VL, Saar R. Desensitizing and non-desensitizing subtypes of alphabungarotoxin-sensitive nicotinic acetylcholine receptors in cockroach neurons. J Insect Physiol 2004;50:867–79. Sattelle DB, Jones AK, Sattelle BM, Matsuda K, Reenan R, Biggin PC. Edit, cut and paste in the nicotinic acetylcholine receptor gene family of Drosophila melanogaster. BioEssays 2005;27:366–76. Senn R, Hefer D, Hoppe T, Angst M, Wyss P, Brandl F, et al. CGA 293343: a novel broad-spectrum insecticide supporting sustainable agriculture worldwide. Proc Brighton Crop Prot Conf—Pests Dis 1998;1:27–36. Schroeder ME, Flattum RF. The mode of action and neurotoxic properties of the nitromethylene heterocycle insecticides. Pestic Biochem Physiol 1984;22:148–60.

842

J. Tan et al. / NeuroToxicology 28 (2007) 829–842

Takahashi H, Mitsui J, Takakusa N, Matsuda M, Yoneda H, Suzuki J, et al. NI25, a new type of systemic and broad spectrum insecticide. Proc Brighton Crop Prot Conf—Pests Dis 1992;1:89–96. Thyssen J, Machemer L. In: Yamamoto I, Casida JE, editors. Imidacloprid: toxicology and metabolism. Nicotinoid insecticides and the nicotinic acetylcholine receptor. Tokyo: Springer-Verlag; 1999. p. 213–22. Tomizawa M, Casida JE. Selective toxicity of neonicotinoids attributable to specificity of insect and mammalian nicotinic receptors. Annu Rev Entomol 2003;48:339–64. Tomizawa M, Casida JE. Neonicotinoid insecticide toxicology: mechanisms of selective action. Annu Rev Pharmacol Toxicol 2005;45:247–68.

Wakita T, Kinoshita K, Yamada E, Yasui N, Kawahara N, Naoi A, et al. The discovert of dinotefuran: a novel neonicotinoid. Pest Manage Sci 2003;59:1016–22. Wellmann H, Gomes M, Lee C, Kayser H. Comparative analysis of neonicotinoid binding to insect membranes. II. An unusual high affinity site for [3H]thiamethoxam in Myzus persicae and Aphis craccivora. Pest Manage Sci 2004;60:959–70. Zwart R, Oortgiesen M, Vijverberg PM. Nitromethylene heterocycles: selective agonists of nicotinic receptors in locust neurons compared to mouse N1E-115 and BC3H1 cells.. Pestic Biochem Physiol 1994;48: 202–13.