Curiouser and Curiouser: The Macrocyclic Lactone, Abamectin, Is also a Potent Inhibitor of Pyrantel/Tribendimidine Nicotinic Acetylcholine Receptors of Gastro-Intestinal Worms

Nematode parasites may be controlled with drugs, but their regular application has given rise to concerns about the development of resistance. Drug combinations may be more effective than single drugs and delay the onset of resistance. A combination of the nicotinic antagonist, derquantel, and the macrocyclic lactone, abamectin, has been found to have synergistic anthelmintic effects against gastro-intestinal nematode parasites. We have observed in previous contraction and electrophysiological experiments that derquantel is a potent selective antagonist of nematode parasite muscle nicotinic receptors; and that abamectin is an inhibitor of the same nicotinic receptors. To explore these inhibitory effects further, we expressed muscle nicotinic receptors of the nodular worm, Oesophagostomum dentatum (Ode-UNC-29:Ode-UNC-63:Ode-UNC-38), in Xenopus oocytes under voltage-clamp and tested effects of abamectin on pyrantel and acetylcholine responses. The receptors were antagonized by 0.03 μM abamectin in a non-competitive manner (reduced Rmax, no change in EC50). This antagonism increased when abamectin was increased to 0.1 μM. However, when we increased the concentration of abamectin further to 0.3 μM, 1 μM or 10 μM, we found that the antagonism decreased and was less than with 0.1 μM abamectin. The bi-phasic effects of abamectin suggest that abamectin acts at two allosteric sites: one high affinity negative allosteric (NAM) site causing antagonism, and another lower affinity positive allosteric (PAM) site causing a reduction in antagonism. We also tested the effects of 0.1 μM derquantel alone and in combination with 0.3 μM abamectin. We found that derquantel on these receptors, like abamectin, acted as a non-competitive antagonist, and that the combination of derquantel and abamectin produced greater inhibition. These observations confirm the antagonistic effects of abamectin on nematode nicotinic receptors in addition to GluCl effects, and illustrate more complex effects of macrocyclic lactones that may be exploited in combinations with other anthelmintics.

The increased use of anthelmintics has led to the development of resistance in both animal and human nematode parasites [18]. Resistance to anthelmintics presents a serious drawback in the control of nematode parasite infections and is an increasing medical concern [19]. Combination therapy involving anthelmintic drugs of different classes is an important approach that may slow the development of anthelmintic drug resistance. The simultaneous development of resistance to two drugs is less likely than the development of resistance to one drug alone; hence combinations are predicted to slow the development of resistance [20]. Startect 1 (derquantel and abamectin combination, Fig 1) is a more recently introduced anthelmintic combination that is marketed for use in sheep, but it has been shown to have effects on nematode parasites of other host species such as Ascaris suum of pigs [21]. In this study, we investigated the effects of derquantel alone, abamectin alone, and a combination of derquantel and abamectin on a nAChR subtype from the nematode parasite Oesophagostomum dentatum. This subtype is comprised of Ode-UNC-29:Ode-UNC-63:Ode-UNC-38 (29-63-38) nAChR subunits. Other pharmacologically characterized nAChR subtypes of O. dentatum are Ode-UNC-29:Ode-UNC-63 (pyrantel-sensitive nAChR), Ode-UNC-29:Ode-UNC-63:Ode-ACR-8 (acetylcholine-sensitive nAChR), and Ode-UNC-29:Ode-UNC-63:Ode-UNC-38:Ode-ACR-8 (levamisole-sensitive nAChR) [5]. We chose O. dentatum parasite for further investigation because the worm is easily maintained and passaged, and it is a Clade V nematode, just like the model free-living nematode, Caenorhabditis elegans [22]. O. dentatum is a common nodule worm in pigs, very similar to other Oesophagostomum species which infect humans, most notably in northern Togo and Ghana [23]. The O. dentatum nAChR subtype, Ode-UNC-29:Ode-UNC-63:Ode-UNC-38 used in this study is preferentially activated by pyrantel and tribendimidine, and it is antagonized in a non-competitive manner by derquantel [5]. We show here that abamectin is a potent antagonist of these nAChRs, and that a combination of derquantel and abamectin has greater effects than derquantel or abamectin used alone. These additive effects support the continued use of the derquantel and abamectin combination for the control of nematode parasites. IACUC Log #3-2-5134-s specifically approved this study and granted permission for the culture of the Oesophagostomum parasites. There was no animal suffering or surgery required.

Materials and Methods
Cloning of nAChR subunits from O. dentatum and ancillary factors from Haemonchus contortus O. dentatum nAChR subunits and H. contortus ancillary factors used in this research study have been previously cloned and reported [5].

Two-microelectrode voltage-clamp (TEVC) electrophysiology
We followed methods we have described in detail before [5]. Briefly, 100 μM BAPTA-AM was added to the oocytes in incubation media approximately 3 hours prior to electrophysiological recordings from the oocytes. TEVC recordings were carried out by impaling the oocytes with two microelectrodes; a voltage sensing electrode, Vm, and a current injecting electrode, Im, were used to inject the current required to hold the membrane at the set voltage. The microelectrodes were pulled with a Flaming/Brown horizontal electrode puller (Model P-97, Sutter Instruments), filled with 3 M potassium chloride and the microelectrode tips carefully broken with a piece of tissue paper to achieve a low resistance of 2

Drugs
Acetylcholine (ach) and pyrantel (pyr) were purchased from Sigma-Aldrich (St Louis, MO, USA), while derquantel (der) and abamectin (aba) were gifts from Zoetis (Kalamazoo, MI). Acetylcholine and pyrantel were dissolved in recording solution. Derquantel and abamectin were dissolved in DMSO and added to recording solution at a final concentration of 0.01% DMSO.

Drug applications
Acetylcholine which is a natural agonist of many nAChRs served for normalization and as a control in our experiments. Each agonist concentration was applied for 10 s, with sufficient time allowed for wash off between drug applications. For every oocyte recording, 100 μM acetylcholine was applied initially and this current response was measured, subsequent responses were measured and normalized to this acetylcholine current. To generate control agonist concentration-response plots, acetylcholine was applied at different concentrations between 0.1-100 μM or pyrantel was applied at different concentrations between 0.03-10 μM. Higher concentrations of pyrantel were not tested because pyrantel acts as an open channel blocker at high concentrations [24]. For the antagonist experiments, a control 100 μM acetylcholine was also first applied for 10s, and then followed by a 2 min or 10 min challenge with the antagonist (derquantel and/or abamectin), and a final application of the agonist (pyrantel) in the continued presence of the antagonist.

Data analysis
We used Clampfit 9.2 (Molecular Devices, CA, and USA) to measure peak current responses and normalized them to the first 100 μM acetylcholine responses. GraphPad Prism 5.0 software (GraphPad Software Inc., USA) was used to analyze the data, and results were expressed as mean ± S.E.M. Concentration-response data points were fitted with the Hill equation as previously described: where R max is the maximum response % relative to the control 100 μM ach response; the EC 50 is the concentration producing the half-maximum response and n H is the slope factor or Hill coefficient [25]. We used the unpaired two-tailed Student's t-test to test for statistical significance and a p value < 0.05 was considered significant. The Bliss additive effect dose-response relationship for pyrantel current responses was calculated as previously described [26] to predict the linear additive effects of derquantel and abamectin on the Ode-UNC-29:Ode-UNC-63:Ode-UNC-38 nAChR. To do so, we determined: the normalized mean current responses to each control pyrantel concentration (CR[pyr]); the mean current responses to each concentration of pyrantel in the presence of derquantel (derquantel effect) and; the mean current responses to each concentration of pyrantel in the presence of abamectin (abamectin effect). We then used the fractional inhibition (mean reduction in response/CR[pyr]) produced by derquantel alone, Fd, and abamectin alone, Fa, at each pyrantel concentration to determine the additive effect. We used Fa(1-Fd) to denote the fractional inhibition produced by abamectin when derquantel is already present, and Fd+Fa(1-Fd) to denote the fractional inhibition produced by the combination of derquantel and abamectin. Lastly, the normalized additive response was calculated as: The difference between the observed and calculated additive effects for the combination of derquantel and abamectin were tested for statistical significance using the paired two-tailed Student's t-test, taking p < 0.05 as significant.  Table). Our results showed the EC 50 for pyrantel to be significantly smaller (p < 0.001) than the EC 50 for acetylcholine, but showed no significant difference (p > 0.05) in R max between acetylcholine and pyrantel. Based on these EC 50 values, pyrantel is 32.5 times more potent than acetylcholine on the Ode-UNC-29:Ode-UNC-63:Ode-UNC-38 nAChR. Also there was a slightly steeper Hill slope (nH) for pyrantel than acetylcholine (nH = 1.2 ± 0.1, n = 6 for pyrantel; nH = 1.0 ± 0.1, n = 4 for acetylcholine; p < 0.05). We used pyrantel rather than acetylcholine for most subsequent experiments because of its potency and because it is a more selective agonist than acetylcholine for this nAChR subtype [5].  smaller than R max in the presence of 0.3 μM abamectin (98.4 ± 6.5, n = 5). This suggested that the inhibitory effects of lower concentrations of abamectin can be greater than those of higher concentrations. This reversed inhibitory effect was tested and confirmed when we increased the concentration of abamectin from 0.3 μM to 1 μM and 10 μM. While the EC 50 values did not change (0.3 ± 0.1 μM, n = 6 in the presence of 1 μM abamectin, and 0.4 ± 0.1 μM, n = 5 in the presence of 10 μM abamectin), the R max values increased to 108.0 ± 3.5%, n = 6 in the presence of 1 μM abamectin, and 106.0 ± 9.4%, n = 5 in the presence of 10 μM abamectin. We point out that the effects of abamectin were reversible on washing, although the current responses did not return to control responses even after a 10 mins wash. S2 Table summarizes the EC 50 , R max , n H and n numbers for these experiments. Interestingly, when the EC 50s for pyrantel alone are compared with that for pyrantel in the presence of 0.1 μM, 0.3 μM, 1 μM or 10 μM abamectin, there were no significant differences in the EC 50 values. In contrast however, the R max values for pyrantel in the presence of 0.1 μM, 0.3 μM, 1 μM and 10 μM abamectin were significantly less than the R max for pyrantel alone, Fig 4B. We emphasize again that the R max for pyrantel in the presence of abamectin concentrations !0.3 μM were greater than the R max for pyrantel in the presence of 0.1 μM abamectin, with the difference being statistically significant (p < 0.05) for 1 μM. These observations show that abamectin has a bi-phasic effect, suggesting two allosteric sites of action: a high affinity site causing antagonism, and a lower affinity site causing a reduction in the antagonism, Fig 4C and 4D.   following the administration of abamectin at its therapeutic dose of 0.2 μM at 12 hours [27]. 0.1 μM derquantel was selected because the reported concentration is 13 μM at 12 hours in the intestine [28] with an uncertain but lower concentration being present within the worm. We used pyrantel as the preferred agonist to study the effects of abamectin and derquantel on the Ode-UNC-29:Ode-UNC-63:Ode-UNC-38 receptor because it is more selective than acetylcholine for this receptor [5]. However, we also tested the effects of abamectin and derquantel with acetylcholine as the agonist instead of pyrantel. We found, as with pyrantel, that a lower concentration of 0.1 μM abamectin was a more potent non-competitive acetylcholine antagonist than 0.3 μM or 1 μM abamectin, S1 Fig. We also found that 0.3 μM derquantel acts as a non-competitive antagonist of acetylcholine and that the combination of derquantel and abamectin shows additive inhibitory effects without demonstrating synergism, S2 Fig.

Discussion
Abamectin as a non-competitive antagonist of Ode-UNC-29:Ode-UNC-63:Ode-UNC-38 Abamectin, a mixture of avermectins B1a and B1b (Fig 1), belongs to the macrocyclic lactone class of anthelmintics which have been shown to act on GluCls [11]. In addition to the effect of abamectin on GluCls, Puttachary et al [21] showed, using muscle contraction and currentclamp electrophysiological techniques, that abamectin acts as a non-competitive antagonist of Ascaris suum (Clade III nematode) somatic muscle nAChRs. In this paper we were interested to see if abamectin had similar effects on the expressed O. dentatum (Clade V nematode) receptor, Ode-UNC-29:Ode-UNC-63:Ode-UNC-38. Interestingly, we found that abamectin was also a potent non-competitive antagonist of this receptor (Fig 3B).
Bi-phasic effects of abamectin on Ode-UNC-29:Ode-UNC-63:Ode-UNC-38 The inverse dose-dependent bi-phasic effects of abamectin on Ode-UNC-29:Ode-UNC-63: Ode-UNC-38 suggest that abamectin has two allosteric sites of action on this receptor: one, a higher affinity negative allosteric modulator (NAM) site causing antagonism; and the other, with lower affinity, a positive allosteric modulator (PAM) site, causing a reduction in the antagonism (Fig 4D). The NAM and PAM sites of action of abamectin on the receptors may be in the outer lipid phase of the membrane [29] between the TM1 transmembrane domain of one receptor subunit and the TM3 domain of the adjacent receptor subunit [30]. Since the receptor channel is composed of heterogeneous subunits, binding between one pair of subunits may stabilize closing (negative allosteric modulator) while binding between another pair of subunits may stabilize opening (positive allosteric modulator). Derquantel and abamectin combination produces greater effects than either derquantel or abamectin used alone The effects of a combination of derquantel and abamectin on a mixture of in situ A. suum somatic nAChR subtypes have been described by Puttachary at al [21]: when derquantel was used alone, A. suum somatic nAChRs were antagonized in a competitive manner (right-shift in EC 50 , but no change in Rmax); when abamectin was used alone, A. suum somatic nAChRs were antagonized in a non-competitive manner (reduction in Rmax, but no change in EC 50 ); when derquantel and abamectin were used in combination, abamectin potentiated the antagonism of derquantel. In our experiments with Ode-UNC-29:Ode-UNC-63:Ode-UNC-38, derquantel showed non-competitive (Fig 5C) rather than competitive antagonism. With the combination of derquantel and abamectin together, the R max was significantly smaller than for derquantel alone and abamectin alone (Fig 5D). Additionally, the effects of the combination were greater than those of 0.1 μM derquantel at pyrantel concentrations ! 0.3 μM suggesting that the use of abamectin and derquantel in combination is likely to be more effective than either one of the two drugs administered alone.

Conclusion
Our observations reveal that abamectin, in addition to its recognized effects on GluCl channels also has effects on an UNC-29: UNC-63: UNC-38 subtype of nematode nAChR and that the antagonism is non-competitive with a bi-phasic inverse dose-dependent effect characteristic of two sites of action. The higher affinity site, a negative allosteric site, produces more closing of the nAChR in contrast to the other lower affinity site, a positive allosteric site that produces more opening. The significant observation is that lower concentrations of abamectin produced greater inhibition of the nAChR than higher concentrations. The combination of derquantel and abamectin as Startect R can produce inhibition of nematode worm movement and feeding. Derquantel and abamectin inhibits the nAChRs; and in addition, abamectin inhibits movement and feeding by activating GluCls. The double action of the combination of derquantel and abamectin has the potential to slow resistance development in nematode parasites by acting at a number of different sites of action. The emergence of resistance in numerous nematode parasite species makes the control of helminth infections with the currently available anthelmintics more difficult but the recent introduction and careful use of appropriate anthelmintic combinations may be used to reduce the rate of development of anthelmintic resistance. It should also be noted that effects of abamectin does not involve a simple mode of action and that there are effects on more than one LGIC in addition to the GluCls.