EAT-18 is an essential auxiliary protein interacting with the non-alpha nAChR subunit EAT-2 to form a functional receptor.

Nematode parasites infect approximately 1.5 billion people globally and are a significant public health concern. There is an accepted need for new, more effective anthelmintic drugs. Nicotinic acetylcholine receptors on parasite nerve and somatic muscle are targets of the cholinomimetic anthelmintics, while glutamate-gated chloride channels in the pharynx of the nematode are affected by the avermectins. Here we describe a novel nicotinic acetylcholine receptor on the nematode pharynx that is a potential new drug target. This homomeric receptor is comprised of five non-α EAT-2 subunits and is not sensitive to existing cholinomimetic anthelmintics. We found that EAT-18, a novel auxiliary subunit protein, is essential for functional expression of the receptor. EAT-18 directly interacts with the mature receptor, and different homologs alter the pharmacological properties. Thus we have described not only a novel potential drug target but also a new type of obligate auxiliary protein for nAChRs.


Author summary
Soil-transmitted helminths affect about a quarter of the worlds' population. Chemical anthelmintics not only alleviate the threat to human and animal health but also improve agricultural economics and food security. Here we have identified a "druggable" nicotinic acetylcholine receptor (nAChR) subunit, EAT-2, that constitutes the pharyngeal cholinergic receptor in nematodes. The receptor is required for feeding and possibly for reproductive behavior in worms. A selective therapeutic compound targeting this nAChR should either starve the worms or make them sluggish, helping with faster expulsion from the host. The EAT-2 pharyngeal nAChR is a unique receptor formed by five non-α subunits that lack vicinal cysteines in the ligand binding loop-C. To date, all cation selective nAChRs contain at least two α subunits. It is possible that EAT-2 subunits have retained functionality without the vicinal cysteines due to evolutionary modifications and expresses as a new nAChR subtype which doesn't fit the established dogma based on the

Introduction
Nematodes are multicellular organisms that exhibit diverse and complex physiological behaviors. These functions are controlled by a neuromuscular system that employs a large repertoire of highly regulated transporters, neurotransmitters, peptides and ion channels, which all contribute to homeostatic cell-cell communication [1,2]. Nicotinic acetylcholine receptors (nAChRs) are pore-forming membrane proteins belonging to the cys-loop ligand-gated ion channel superfamily. They are conserved throughout metazoan evolution and characterized by a pentameric subunit organization. nAChRs facilitate rapid ionotropic neurotransmission, thereby controlling various physiological behaviors, including reproduction, navigation, feeding, and locomotion [3]. Nematode nAChRs, especially those found on somatic muscle, are targeted by the cholinergic anthelmintic drugs [3][4][5][6][7]. These drugs help alleviate the nematode parasite burden of the 1.5 billion people affected globally as well as mitigate the threat to global food security caused by nematode parasites of livestock [8,9]. Nematodes have a greater number (�29) of nAChR subunits than vertebrates [10,11]; variation in stoichiometry and subunit composition leads to diverse pharmacological sensitivities which makes them attractive anthelmintic targets [6,[10][11][12]. The introduction of recently discovered drugs such as amino-acetonitrile compounds (monepental) and spiroindoles (derquantel), which target nematode nAChRs, further highlight their importance in drug discovery [13][14][15].
All nAChR ion-channels are composed of five subunits forming a central ion-conducting pore and can be either homomeric (one α subunit) or heteromeric (multiple subunits with at least 2 α subunits). Nicotinic acetylcholine receptors from numerous organisms, including Caenorhabditis elegans, have been shown to interact with various chaperone or ancillary proteins such as RIC-3 (resistance to inhibitors of cholinesterase),  and . Ancillary proteins are required for correct folding, assembly of individual subunits into pentamers and trafficking of the mature nAChRs in a subtype dependent manner [12,16,17]. In addition to ancillary proteins, several structurally unrelated auxiliary subunit proteins have been identified for various ionotropic receptors [18][19][20][21]. Auxiliary proteins are essential for functional regulation of ion channels. They are non-pore forming and interact directly with the receptor subunits to modulate channel properties. They do not exhibit any channel activity on their own and are required for certain aspects of in vivo channel function [20]. Boulin et al. [22] identified the first auxiliary subunit for nAChRs, MOLO-1 (modulator of levamisole receptor-1), that regulates biological and biophysical properties of the levamisole-sensitive (L-type) nAChRs in C. elegans. This demonstrates that nAChRs are tractable to regulation by auxiliary proteins contributing to the biological and pharmacological diversity of nAChR subtypes.
In nematodes, the pharynx is a neuromuscular organ that undergoes rhythmic peristalsis to ingest food and is thus crucial for survival [23][24][25]. Pharyngeal peristalsis is under the control of rhythmic activation by excitatory (cholinergic) and inhibitory (glutamatergic) motor neurons innervating the pharyngeal muscle [26,27]. Glutamate-gated chloride channels (GluCls) in the nematode pharynx are one of the primary targets for the avermectins [28][29][30], however little is known about the cholinergic receptors in this tissue. Through genetic screening Raizen et al. [31] identified eat-2 (encoding a non-α nAChR subunit; eating-2) and eat-18 (encoding a single-pass transmembrane domain protein; eating-18) as essential components of pharyngeal cholinergic transmission in C. elegans. Here, we have cloned and functionally expressed EAT-2 and EAT-18 from free-living (C. elegans) and parasitic (Ascaris suum) nematodes. For the first time, we find that a non-α nicotinic subunit (EAT-2) can form a homomeric ligand-gated cation selective ion-channel. The functional expression of this noncanonical receptor is dependent on co-expression with EAT-18. All previously characterized cation selective nAChRs have at least two α subunits with the ligand binding sites located at the interface between each α and its adjacent subunit [10,32]. We have used electrophysiological, biochemical, and molecular techniques to demonstrate that EAT-18 forms part of the mature receptor and functions as an obligate auxiliary protein.

Cel-EAT-2 is a non-α nAChR subunit most similar to vertebrate α-7 subunits
C. elegans EAT-2 has the typical functional domains of a pentameric ligand-gated ion channel subunit: a large extracellular N-terminal domain of~200 amino acids required for correct nAChR assembly and agonist binding; a cys-loop separated by 13 intervening amino acids; four transmembrane (TM) domains that form the ion-conducting pore; a cytoplasmic domain between TM3 and TM4 that is involved in modulation of channel activity and ion conductance; and a short extracellular C-terminus. EAT-2 is a non-α subunit as it lacks the pair of adjacent cysteine residues in loop-C required for agonist binding, still overall its sequence is most comparable to the human α-7 subunit with 55% similarity in amino acid residues (S1 Fig). Ligand binding occurs in a cleft formed by three loops (A, B, C) of the principal face of one α subunit and a series of beta strands from loops (D, E, F) of the complimentary interface of the adjacent subunit. All α subunits have either a YXCC or YXXCC motif in loop-C, and this motif was considered essential for ligand binding and modulating the affinity of the receptor binding site [33][34][35]. This lack of vicinal cysteines in the EAT-2 protein subunit suggests that the receptor channel will have different contact residues in the ligand binding pocket and a different pharmacology from other nAChRs.

Cel-eat-18 overlaps with the gene encoding for a CUB/LDL transmembrane protein LEV-10
Cel-EAT-18 is a small, single-pass transmembrane protein expressed in pharyngeal muscle and neurons with no vertebrate homologs [31,36]. There are two splice variants of the gene in C. elegans, encoding EAT-18c (71 aa) and EAT-18d (78 aa), which differ mainly in their C-terminal regions (S2A Fig). The coding sequence of EAT-18 is contained within lev-10, a gene encoding CUB/LDL transmembrane protein LEV-10, localized at cholinergic neuromuscular junctions. LEV-10 functions as an ancillary protein for levamisole sensitive nAChRs in C. elegans and is required for postsynaptic clustering of nAChRs in the body wall muscles. Mutants of the gene display weak levamisole resistance [37]. Interestingly, the first exon of both isoforms of eat-18 is located in the first intron of lev-10 (S2B Fig). Gally et al. [37] confirmed that eat-18 was distinct from lev-10 and was not involved in conferring levamisole resistance. Another distinguishing feature of EAT-18 is the intracellular N-terminal and extracellular Cterminal, which are reversed in lev-10 ( S2C Fig).

Cel-EAT-18
Initiation of the pharyngeal muscle action potential and the frequency of excitatory pharyngeal pumping are under the control of a pair of MC neurons that synapse on marginal cells in C. elegans. MC neurons release acetylcholine producing a fast depolarization of postsynaptic muscle membranes triggering an action potential. The MC neurons behave as a neurogenic pacemaker for rapid pharyngeal pumping. MC neurotransmission requires acetylcholine (ACh) and the nAChR subunit Cel-EAT-2, which is expressed in pharyngeal muscle [31,38]. In order to reconstitute the post-synaptic pharyngeal nAChR, we expressed Cel-EAT-2 in Xenopus laevis oocytes but failed to observe electrophysiological evidence for the formation of a functional nAChR. This could be attributed to the lack of vicinal cysteines in the ligand binding site required for agonist binding and pointed to a possible requirement for an additional protein or subunit. Raizen et al. [31] have shown that similar to Cel-eat-2, mutations in Celeat-18 rendered worms incapable of MC neurotransmission and rapid pharyngeal pumping suggesting the protein may act as either an ancillary or auxiliary protein for the assembly of a functional nAChR. We co-expressed Cel-EAT-2 with Cel-EAT-18c or Cel-EAT-18d cRNA and recorded robust responses to 100 μM ACh in both cases (Fig 1A). The resulting nAChRs produced larger current amplitudes in response to ACh application when the Cel-EAT-18c isoform was used. All of the subsequent recordings were done using the Cel-EAT-18c isoform. The ability of the non-α EAT-2 subunit to express functionally as a homomeric cation selective channel when co-injected with a non-subunit protein makes this cation selective nAChR unique to date.
To further investigate the receptor pharmacology, we examined the concentration-response relationships of selected agonists (Figs 1C, 1D and S3A). 100 μM ACh was used as the internal standard for normalization. Nicotine (pEC 50 = 4.2 ± 0.1) was the most potent agonist after ACh (pEC 50 = 4.8 ± 0.0), whereas carbachol was least potent with a pEC 50 = 3.4 ± 0.0. The concentration-response curves for all the agonists had Hill coefficient values greater than 1, indicating positive cooperativity, with methacholine having the steepest Hill slope (n H = 3.5 ± 1.3). This suggests that the Cel-EAT-2 ion channel has multiple ligand binding sites consistent with other nAChRs.

PLOS PATHOGENS
EAT-2 interacts with auxiliary transmembrane protein EAT-18 to form a functional receptor

PLOS PATHOGENS
EAT-2 interacts with auxiliary transmembrane protein EAT-18 to form a functional receptor erythroidine (30 μM, DHβE). Fig 1E and 1F (and S3B Fig) illustrate the effect of various antagonists on the ACh concentration-response relationship for Cel-EAT-2. d-Tubocurarine produced the most potent inhibition and almost completely blocked the response mediated by ACh (�98% inhibition). Unlike many mammalian nAChRs, the sensitivity and efficacy of the receptor for ACh were not altered by either α-bungarotoxin or DHβE. The antagonist functional profile based on mean current (%) decrease of the control 100 μM ACh current response was: d-tubocurarine > paraherquamide > derquantel >>> α-bungarotoxin � DhβE. In conclusion, the pharmacology of the Cel-EAT-2 receptor is distinct from previously characterized nematode and vertebrate nAChRs [6,[39][40][41][42][43].

Characterization of the acetylcholine response in the A. suum pharynx
Although C. elegans is a powerful model, it is not a parasitic nematode of medical importance. In order to validate pharyngeal nicotinic acetylcholine ion channels as potential anthelmintic drug targets, it is crucial to identify and establish the presence and, in turn the pharmacology of such nAChRs in the pharynx of parasitic worms. We therefore characterized the pharmacology of the A. suum pharynx for comparison with the Cel-EAT-2 receptor. We employed the current-clamp technique to understand the pharmacology of the postsynaptic nAChR response. Application of 100 μM ACh on the pharyngeal preparation produced a large depolarization accompanied by an increase in membrane conductance. The ACh response was inhibited by mecamylamine, and the preparation showed negligible responses to several muscarinic agonists (S1 Data). This confirmed the presence of a nicotinic acetylcholine receptor in the pharynx of the parasite.
We next quantified the effects of selected nicotinic agonists to determine whether the pharyngeal nAChRs are pharmacologically distinct from those of somatic muscle. Our pharyngeal preparations in this group had a mean resting membrane potential of -21.3 ± 1.3 mV and a mean resting conductance (G) of 136.4 ± 14.9 μS (n = 17). The change in conductance (δG) responses to test applications of selected nicotinic agonists were normalized to the ACh δG. Nicotine was the most potent agonist after ACh with mean δG of 92.0 ± 6.2%. Cytisine also produced a large conductance change in the A. suum pharynx (mean δG = 71.2 ± 5.0). The rank order series for vertebrate nicotinic agonists on the A. suum pharynx was: ACh > nicotine > cytisine > epibatidine > DMPP >> choline (Fig 2A and S4A Fig). The rank order series of selected vertebrate nicotinic agonists on the pharynx is different from that of somatic muscle nAChRs and vertebrate host nAChRs (S1 Table). We also tested nine cholinergic anthelmintics on the pharynx to study their effect. Our pharyngeal preparations in these experiments had a mean resting membrane potential of -19.3 ± 1.1 mV and a mean resting conductance of 150.5 ± 11.9 μS (n = 21). The δG responses to test applications of selected cholinergic anthelmintic agents were normalized to the ACh δG.

PLOS PATHOGENS
EAT-2 interacts with auxiliary transmembrane protein EAT-18 to form a functional receptor rank order series on the A. suum pharynx: ACh >> bephenium > thenium > levamisole � morantel � pyrantel � oxantel � tribendimidine. In contrast to somatic muscle nAChRs, none of the cholinergic anthelmintics tested on the pharynx produced >7% of the ACh response (S1 Table).
To further investigate the receptor, we used selected nicotinic antagonists (30μM) to study their inhibitory effects on 100μM ACh responses. Our pharyngeal preparations in this group had a mean resting membrane potential of -20.2 ± 1.1 mV and a mean resting conductance of 129.2 ± 6.8 μS (n = 34). The δG produced by a control application of ACh was set as 100%. We calculated the % inhibition of the δG response to ACh by nicotinic antagonists to determine a rank order series (mean ± SEM, S5 Fig and S6 Fig) The functional spectrum of nicotinic receptor antagonists on the pharynx is distinct from that of vertebrate nAChRs (S1 Table).
We also determined concentration-response curves by plotting the concentration of agonists (1-1000μM, applied for 10s) against the response normalized to 100 μM ACh (applied for 10s) δG within each experiment. Fig 2B (S4B Fig and S4C Fig) shows the concentration-response curves for ACh and nicotine. The pEC 50 of ACh and nicotine were 5.0 ± 0.0 and 5.0 ± 0.1, respectively. The maximal response of ACh was 105.3 ± 2.5 μS, and nicotine was 79.3 ± 4.3 μS. We determined the ACh concentration-response relationships in the presence of nicotinic receptor antagonists: d-tubocurarine (10μM), methyllycaconitine (10μM), paraherquamide (10μM) and dihydro-β-erythroidine (30μM) (Fig 2C). The pEC 50 of the ACh concentrationresponse curve did not significantly differ in the presence of methyllycaconitine, d-tubocurarine or paraherquamide but the maximal response for ACh was inhibited. This suggests that these compounds act as non-competitive antagonists of the pharyngeal ACh response.

Functional expression of Asu-EAT-2 requires Asu-EAT-18 and Asu-RIC-3
The pharmacological characterization of A. suum pharyngeal nAChRs revealed significant differences from the Cel-EAT-2 ion channel. In particular, cytisine which produced large depolarization in the A. suum pharynx, failed to activate Cel-EAT-2 ion channel. These pharmacological differences encouraged us to identify the subunits which constituted pharyngeal nAChRs in A. suum. We used Cel-EAT-2 and Cel-EAT-18 sequences as queries in BLASTP homology searches and identified homologs for EAT-2 and EAT-18 in the pig parasite. Comparison of Asu-EAT-2 with Cel-EAT-2 sequences revealed 80% similarity in amino acid composition, with differences among some of the ligand binding residues from various loops (S1

Pharmacological profile of the Asu-EAT-2 nAChR
We were interested in determining the comparative pharmacological profile of A. suum pharyngeal nAChRs and Asu-EAT-2 ion channel in order to establish the contribution of the non-α subunit in pharyngeal pharmacology. We used similar cholinergic agonists, anthelmintic agents, and antagonists as in vivo A. suum pharyngeal experiments on the expressed Asu-EAT-2 receptor. The rank order series of cholinergic agonists and anthelmintic agents based on maximum current response (Fig 3A and S7B Fig) for the receptor was: nicotine > ACh > cytisine > epibatidine > DMPP > oxantel. As with the A. suum pharynx, cholinomimetic anthelmintics such as bephenium, tribendimidine, levamisole, and pyrantel failed to activate the receptor. We also constructed a concentration-response curve for ACh and found it to be � 9 times more potent on the Asu-EAT-2 nAChR compared to Cel-EAT-2 with a pEC 50 = 5.7 ± 0.0 (Fig 3B and S7C Fig).
We tested the antagonistic effects of derquantel, mecamylamine, d-tubocurarine, DHβE, hexamethonium, and methyllycaconitine on the Asu-EAT-2 receptor. The mean % inhibition of the 100 μM ACh current response was used to determine the effect of the antagonists. Mecamylamine and d-tubocurarine produced almost 100% inhibition of the ACh currents, and DhβE was the least potent antagonist (inhibition, 66 ± 8.4%). The functional profile for the antagonists (Fig 3C and  S7D Fig) was: d-tubocurarine~mecamylamine > hexamethonium > methyllycaconitine > deraquantel > DhβE. The rank order series of cholinomimetic anthelmintics, nicotinic agonists, and antagonists on the Asu-EAT-2 receptor differs from that of the A. suum somatic muscle nAChRs as well as the vertebrate nAChRs (S1 Table). In conclusion, the Asu-EAT-2 receptor has a distinct pharmacology and is, therefore, likely suitable to be exploited as a therapeutic target.

Tissue expression of eat-2 and eat-18 in A. suum
In C. elegans, EAT-2 expression is restricted to pharyngeal muscle, while EAT-18 is found in both pharyngeal muscle and some neurons [36]. We used RT-PCR to examine the distribution of Asu-eat-2 and Asu-eat-18 mRNA in various dissected adult A. suum tissues and single somatic muscle cells (n� 5; Fig 3D). We determined that Asu-eat-2 was transcribed in the pharynx, sections of the reproductive tract, and the head region. RT-PCR results revealed the presence of Asu-eat-18 message in the same tissues, as well as gut tissue. We found no evidence of expression of Asu-eat-2 or Asu-eat-18 in somatic muscle cells (S8 Fig). The widespread expression of both the proteins in body tissues other than pharynx was unexpected. It is plausible that Asu-EAT-2 not only assists in feeding but also plays a role in other physiological processes such as reproduction. It also raises the possibility that EAT-18 is interacting with other nAChR subunits in different tissues.

PLOS PATHOGENS
EAT-2 interacts with auxiliary transmembrane protein EAT-18 to form a functional receptor nicotine and cytisine were highly efficacious in in vivo and in vitro recordings in A. suum, while DMPP acted as a weak agonist. In comparison, the Cel-EAT-2 channel failed to respond to cytisine and DMPP application but was activated by oxantel (I max = 11.0 ± 1.3% of ACh response) albeit weakly. Importantly, the comparable pharmacological profile observed for Ascaris in vivo and in vitro recordings suggests it is likely that EAT-2 and EAT-18 constitute the pharyngeal nicotinic response in the parasitic nematode.

PLOS PATHOGENS
EAT-2 interacts with auxiliary transmembrane protein EAT-18 to form a functional receptor

PLOS PATHOGENS
EAT-2 interacts with auxiliary transmembrane protein EAT-18 to form a functional receptor

Different EAT-18 homologs affect the pharmacology of the EAT-2 nAChR
We expected to see differences in nAChR pharmacology between A. suum and C. elegans due to differences in the amino acid residues of the EAT-2 protein sequences. EAT-2 cannot form a functional receptor on its own and requires EAT-18. To determine the pharmacological relevance of EAT-18, we expressed Cel-EAT-2 with Asu-EAT-18. We tested five agonists on the expressed channel: ACh, nicotine, cytisine, levamisole, tribendimidine, and pyrantel (S9A Fig). No significant differences were observed in the rank order series. Interestingly, the sensitivity of nicotine was affected, illustrating a change in the pharmacology (Fig 4B). Substitution of Asu-EAT-18 for Cel-EAT-18c shifted the concentration-response curve to the left and increased the efficacy of nicotine on the receptor. The EC 50 = 18.7 μM (pEC 50 = 4.7 ± 0.0) for nicotine was approximately 3.5 times lower than before 64.2 μM (pEC 50 = 4.2 ± 0.0). We also observed a significant increase in I max (96.5 ± 2.3% from 54.6 ± 2.6%) when replacing Cel-EAT-18c with Asu-EAT-18. The sensitivity of the pharyngeal receptor to ACh was also altered significantly, but there was no effect on the agonist efficacy (S9B Fig and S9C Fig). The significant shift in ACh and nicotine pEC 50 establishes the electrophysiological evidence of modulation of the receptor by EAT-18 and points to an interaction between the proteins.

EAT-18 co-localizes with EAT-2 on the oocyte surface membrane
EAT-18 is required for functional in vitro expression of EAT-2 and modulates its pharmacological properties. McKay et al. [36] have shown that Cel-EAT-18 is not required for the trafficking of Cel-EAT-2 to the oocyte membrane. It is possible that Cel-EAT-18 functions as an auxiliary subunit rather than an ancillary/chaperone protein. To test this hypothesis, we performed confocal imaging experiments on oocytes expressing GFP tagged Cel-EAT-2 and His tagged Cel-EAT-18c constructs. These experiments revealed that both Cel-EAT-2 and Cel-EAT-18c were co-localized on the oocyte surface membrane. Fig 5C and 5D show the images for double immunostained sections of injected and un-injected oocytes. Cel-EAT-18c is not a typical nAChR subunit protein, and its co-localization with Cel-EAT-2 suggests the possibility of an association between both proteins. Cel-EAT-2-GFP, when expressed alone, localizes to the oocyte surface membrane suggesting that Cel-EAT-18c does not function as an ancillary protein (Fig 5A).
We further assessed the expression of the Cel-EAT-2 channel by Western blot analysis of oocyte protein extracts. Using antibodies that recognize GFP and His tags, we detected Cel-EAT-2-GFP as a 62 kDa and Cel-EAT-18c-His as a 10 kDa protein (Fig 5E and S10A Fig). We were able to detect Cel-EAT-2 in membrane extracts prepared from oocytes co-injected with Cel-EAT-2 and Cel-EAT-18c as well as oocytes injected with Cel-EAT-2 alone. In contrast, Cel-EAT-18c was only present in membrane extracts prepared from oocytes co-injected with both Cel-EAT-2 and Cel-EAT-18c. We did detect Cel-EAT-18c in whole oocyte extracts when Cel-EAT-18c was injected alone. It is plausible that EAT-18 requires EAT-2 for trafficking to the surface membrane, and its role is more complicated than a simple ancillary protein; perhaps related to the functionality of the mature receptor.

EAT-18 forms a part of the EAT-2 nAChR complex
Although Cel-EAT-18c was co-localized with Cel-EAT-2 on the surface membrane of the oocytes and modulated the pharmacology of pharyngeal nAChR, it did not prove the Asu-EAT-18 + Asu-RIC-3 mix (green curve). Bar graphs showing a significant effect of using different EAT-18 proteins with Cel-eat-2 on pEC 50 (top graph) and on the maximum response (bottom graph) produced by application of 300 μM nicotine.

PLOS PATHOGENS
EAT-2 interacts with auxiliary transmembrane protein EAT-18 to form a functional receptor

PLOS PATHOGENS
EAT-2 interacts with auxiliary transmembrane protein EAT-18 to form a functional receptor molecular interaction. Therefore, we performed co-immunoprecipitation experiments to explore a possible direct interaction between Cel-EAT-18c and Cel-EAT-2 (Fig 5F and S10B  Fig). We were able to demonstrate that Cel-EAT-18c-His co-immunoprecipitated with Cel-EAT-2-GFP, which shows that EAT-18 directly interacts with EAT-2 and is a part of the mature receptor complex (Fig 5G).

Discussion
nAChRs are vital components of the metazoan neuromuscular junction and essential targets for anti-parasitic interventions. They are typically composed of 5 subunits, including at least 2 α subunits. Here we describe for the first time a non-α nAChR subunit that can form a functional homomeric cation selective receptor when coexpressed with Cel-EAT-18. Even though EAT-2 lacks the essential vicinal cysteines in loop-C, the pharyngeal subunit contains most of the residues that form the "aromatic box" of the α-7 ligand binding site [44]. These include a W149 (loop-B; when mature peptide numbering of α7 subunits is used) and Y197 (loop-C) contributed by the "principle face" of one subunit and W54 (loop-D) provided by the "complimentary" face of the adjacent subunit (S1 Fig). This could explain how a non-α subunit, EAT-2, was able to constitute a functional cation selective receptor. Acetylcholine-gated chloride channels are another example of cys-loop ligand-gated subunits that lack the vicinal cysteines but still form functional channels, albeit anion selective [45]. It is possible that through evolution, nematodes express additional nicotinic acetylcholine receptors subtypes in addition to the relatively small but widely studied group of receptors from vertebrates. Variable residues neighboring the conserved core of the aromatic residues and the non-conserved loop-E and Loop-F account for the differences in pharmacological properties of nicotinic acetylcholine receptors [32]. The differences in the pharmacological profile of vertebrate receptors and somatic nematode nAChRs could be attributed to these variations in amino acids.
Cel-EAT-18 functions as an obligate auxiliary protein and modifies the pharmacological properties of this cys-loop ion channel. Several previously characterized nAChRs require ancillary proteins, either RIC-3 alone or in combination with UNC-50 and UNC-74, for successful in vitro expression [6,12,16,17,39]. In vitro expression of Cel-EAT-2 receptors instead only required co-expression with EAT-18, which has little similarity to ancillary proteins. Both proteins are localized in the pharyngeal muscles, and mutations in Cel-eat-18 caused pharyngeal pumping defects similar to Cel-eat-2 mutations [31,36]. Based on our confocal imaging studies, Cel-EAT-18c is localized on the oocyte surface only when injected with Cel-EAT-2. This supports our hypothesis that EAT-18 interacts with EAT-2 and likely requires the pore-forming subunit protein for trafficking to the surface membrane. Other evidence of interaction is provided by the pharmacological modulation of the ACh and nicotine responses by using EAT-18 proteins from different nematode species with Cel-EAT-2. The co-immunoprecipitation EAT-18-His (n = 6) on the surface membrane. The overlay image (yellow fluorescence) shows the co-localization of both the proteins. (D) Double immunostained sections of un-injected (negative control; n = 6) Xenopus laevis oocytes. (E) Western blot analysis of Xenopus oocyte extracts. Un-injected oocytes served as a negative control. Cel-EAT-2-GFP was immunostained with anti-GFP antibodies and was present in the extracts prepared from oocytes co-injected with Cel-EAT-18-His as well as oocytes injected with Cel-EAT-2-GFP alone. Cel-EAT-18-His was immunostained with anti-His antibodies and was present in the extracts prepared from oocytes co-injected with Cel-EAT-2-GFP. , ✣: non-specific interacting protein bands labeled by anti-GFP and anti-his antibodies, respectively, served as a positive control. (F) Co-immunoprecipitation experiments revealed Cel-EAT-18 directly interacts with EAT-2 and constitute part of the receptor complex. Cel-EAT-2-GFP was immunoprecipitated using anti-GFP, followed by Western blot analysis of Cel-EAT-18-His using anti-His antibodies. Un-injected oocytes and oocytes injected with Cel-EAT-2-GFP alone served as negative controls for co-immunoprecipitation experiments. : non-specific interacting protein bands labeled by anti-GFP and anti-his antibodies, respectively, served as a positive control. (G) Schematic representation for physical interaction between EAT-2 and EAT-18. https://doi.org/10.1371/journal.ppat.1008396.g005

PLOS PATHOGENS
EAT-2 interacts with auxiliary transmembrane protein EAT-18 to form a functional receptor experiments have provided more concrete evidence of direct physical interaction between Cel-EAT-2 and Cel-EAT-18. EAT-18 modifies the pharmacological properties of the EAT-2 nAChR and is necessary for in vivo cholinergic MC neurotransmission, which regulates pharyngeal pumping. Therefore EAT-18 meets the criteria of an auxiliary protein. However, unlike other auxiliary proteins, successful expression of a functioning channel requires the presence of EAT-18. Various auxiliary subunits have been discovered for the ionotropic glutamate-gated (kainate and AMPA) and GABA B receptors that modulate their properties or functional expression levels [18][19][20][21]. The interaction of these auxiliary subunits with the receptor complex holds a physiological relevance. MOLO-1 (modulator of levamisole receptor-1) is the first example of an auxiliary subunit for levamisole sensitive acetylcholine receptors in C. elegans [22]. Like MOLO-1, null mutants of EAT-18 result in significant physiological defects, and many of the nematode species express highly conserved orthologues of EAT-18 (S12 Fig) suggesting evolutionary conservation of function. In contrast to EAT-18, MOLO-1 is not required for the functional expression of the somatic levamisole nAChR but only regulates the trafficking, localization, or gating kinetics; instead, EAT-18 is essential for EAT-2 to form a functional receptor in vitro. This implies that EAT-18 not only meets the criteria for an auxiliary protein but may belong to a novel class of proteins not previously described, adding to the types of auxiliary subunits identified for the cys-loop cation channels.
Identification of a suitable target and its validation is one of the most important steps in developing a new drug. An ideal anthelmintic target should meet certain criteria in order to be considered relevant for pharmacological intervention; important physiological function, conservation across parasite species and pharmacological divergence from host receptors. Parasite nAChRs are regarded as popular targets because they contribute to vital physiological functions. Additionally, their diversity, conserved structure among various species of nematodes, and distinct pharmacology from mammalian orthologues make them "druggable". The pharynx is a muscular organ required for feeding in nematodes. While the nematode pharynx has been exploited as a target tissue for the avermectins (GluCls) [26,27,30,46], less is known about the nAChRs in this tissue. In C. elegans two genes, eat-2 and eat-18, were required for MC neurotransmission. Cel-EAT-2 and Cel-EAT-18 are both localized in the pharyngeal muscles, and mutations in these genes caused defects in feeding behavior in the worms [31,36]. We hypothesize that activation of the pharyngeal nAChR formed by the EAT-2 subunit and EAT-18 auxiliary protein will lead to an effect similar to levamisole in somatic muscle and cause pharyngeal paralysis in nematodes. We were able to successfully co-express EAT-2 and EAT-18 from C. elegans, a model nematode, and A. suum, a parasitic species, in Xenopus oocytes and characterize the pharmacology of this conserved receptor.

The pharyngeal nAChR composed of EAT-2 and EAT-18 as a novel drug target
The pharyngeal cys-loop ligand-gated ion channel formed by EAT-2 meets the criteria for a suitable anthelmintic drug target [47]: 1) it performs a neuromuscular function essential for parasite biology; 2) this receptor is druggable, it has distinct pharmacology from somatic muscle receptors and is insensitive to many of the currently used cholinergic anthelmintics including morantel, tribendimidine, and pyrantel; 3) EAT-2 and EAT-18 are present in multiple relevant parasitic nematode species (S11 Fig and S12 Fig) and the protein sequences are highly conserved; 4) Cel-EAT-2 is only 36% identical to human α-7 nAChR subunit, and there are no mammalian homologs for EAT-18 providing potential for selectivity; it is also pharmacologically distinct from vertebrate nAChRs. In order to identify novel drug targets for anthelmintic agents, it is crucial to understand the properties and function of the target proteins. We have

PLOS PATHOGENS
EAT-2 interacts with auxiliary transmembrane protein EAT-18 to form a functional receptor successfully elucidated the components and pharmacological profile of the pharyngeal nAChR by employing various molecular, biochemical, and electrophysiology techniques.

Molecular biology
Plasmid constructs (Life Technologies Inc., USA) containing C. elegans EAT-2 (Accession number: Y48B6.4) & EAT-18 (Accession number: isoform-c-Y105E8A.7c.1 and isoform-d -Y105E8A.7d.1) were cloned into XhoI and ApaI restriction sites of the pTB-207 expression vector using In-Fusion cloning kit (Takara Bio USA, Inc.; EAT-2: 5' end-TGGCGGCCGctcg agATGACCTTGAAAATCGCATTTTTCA and 3' end-ATCAAGCTCgggcccTTATTCAATA TCAACAATCGGACTAT; EAT-18: 5' end-TGGCGGCCGctcgagATGCGAAGCCTGGAGC GAAT and 3' end-ATCAAGCTCgggcccTCAAAGTGTTGATCGCATTTCCTCA). For biochemistry and immunofluorescence assays, Cel-EAT-2 was tagged with GFP in between the transmembrane regions 3 and 4 between leucine 377 and 378; EAT-18 was tagged with the 6xHis tag at the C-terminal. Full-length sequences of A. suum EAT-2 (Accession number: GS_09411) and EAT-18 were amplified from total RNA extracted from the dissected whole pharynx of A. suum. Briefly, TRIzol Reagent (Life Technologies Inc., USA) was used to extract total RNA from A. suum adult worms. cDNA was synthesized by using SuperScript VILO Master Mix (Life Technologies Inc., USA) and served as a template for the amplification. Fulllength product was sub-cloned into pTB207 expression vector by adding XhoI and ApaI restriction enzyme sites respectively to the forward primer (5' end: TGGCGGCCGctcgagATG CAAATATTTTCTATGGTAATT) and reverse primer (3' end: ATCAAGCTCgggcccTTAAT TCCATACGTTTGGGG) using In-Fusion cloning. Z-competent E. coli JM109 cells (Zymo Research, USA) were used for the transformation of the ligated product. The final cloned constructs of all the plasmids were sequenced with pTB207 vector primers (forward, T7 and reverse, SP6). Only positive clones were used for cRNA synthesis using in vitro transcription with the mMessage mMachine T7 transcription kit (Life Technologies Inc., USA), and the cRNA was aliquoted and stored at -80˚C.

Electrophysiology
Two-electrode voltage-clamp in Xenopus oocytes. Oocyte injections and two-electrode electrophysiology recordings were performed as previously described [48].
A. suum current-clamp recordings from the pharynx. A. suum pharyngeal dissections and electrophysiology recordings were adapted from Martin [49]. Briefly, pharynx was dissected out from the head region of the worm and mounted on Sylgard (Sigma-Aldrich Inc., USA)-lined in a double jacketed bath chamber maintained at 28˚C. The muscle layer surrounding the anterior 3 rd of the pharynx was preserved for anchoring. The intestine attached to the posterior end of the pharynx was used for stretching and pinning down. The preparation was continuously perfused with calcium-free Ascaris Perienteric Fluid-Ringer (calcium-free APF-Ringer) composition (mM): NaCl 23, Na-acetate 110, KCl 24, MgCl 2 11, glucose 11, and HEPES 5; NaOH or acetic acid was added to adjust the pH to 7.6. The experimental compounds were dissolved in calcium-free APF-Ringer and applied as described in the results. The rate of localized perfusion was 3.5-4 ml min −1 through a 20-gauge needle, which was placed directly above the recording region of the pharynx. The pharyngeal preparations with resting membrane potentials less than -15 mV and the resting conductances less than 250 μS were selected for analysis. We used 3 M potassium acetate in our micropipettes to get the final resistances of 4-7 MO for the voltage sensing and 0.5-1 MO for the current injecting electrode for current-clamp recordings. The current-injecting electrode injected hyperpolarizing step

PLOS PATHOGENS
EAT-2 interacts with auxiliary transmembrane protein EAT-18 to form a functional receptor currents of −1000 nA for 500 ms at 0.3 Hz. Pharyngeal preparations with constant resting membrane potentials more negative than −15 mV for 20 min and a stable input conductance of <250 μS were selected for the recordings. Data analysis. GraphPad Prism 8.0 software (GraphPad Software Inc., USA) was used to analyze the data. In two-electrode voltage-clamp recordings, the peak currents were measured and normalized to 100 μM ACh response and expressed as mean ± S.E.M. The data for sigmoid concentration-response curves were fitted to the Hill equation [17]. We used One-way analysis of variance (ANOVA) and Extra sum of squares F-test to test statistical differences (statistically different if p<0.05). Tukeys' multiple comparison was used as a post-hoc test.
In current-clamp recordings, the peak changes in membrane conductance (δG max ) in response to drug applications were normalized to δG response to ACh application (100 μM ACh, applied for 10s) within each preparation. We constructed the sigmoidal concentrationresponse plots by fitting the data by nonlinear regression to determine the pEC 50 and the maximal response (R max ). Extra sum of squares F-test was used to test statistical differences between pEC 50 , slope, and maximal response. The significance levels were set to P<0.05.

Biochemistry
Immunostaining for confocal microscopy. Oocytes were prepared for confocal imaging following the previously published protocol [50]. In brief, oocytes were fixed in 4% paraformaldehyde at 4˚C overnight. Fixed oocytes were embedded in 3% low-melting point agarose, and 50-μm thick slices were cut using a vibratome. The oocyte slices were blocked with 0.2% Bovine Serum Albumin (BSA) plus 0.1% Tween 20 in PBS overnight at 4˚C. This was followed by incubation with the primary antibody (ABfinity Histidine tag recombinant rabbit oligoclonal antibody at 1:500 dilution for detecting Cel-EAT-18c, ThermoFisher#A-710286, and goat anti-GFP rabbit IgG antibody for detecting Cel-EAT-2, Abcam # ab6673, 1:1000) overnight at 4˚C, and then incubated with a fluorescent secondary antibody (Alexa-Fluor 488 goat anti-rabbit IgG antibody, 1:1000, ThermoFisher#A-11008; Alexa-Fluor Plus 680 donkey anti-goat IgG secondary antibody, ThermoFisher#A-32860, 1:15000) for 1 hour at room temperature. The slices were mounted on glass slides using a Fluoromount-G mounting medium (Thermo Fisher Scientific, USA) followed by confocal imaging (Leica SP5 X MP confocal/multiphoton microscope system).
Western blot analysis using Xenopus oocytes. Oocyte protein extraction and Western blot analysis protocol were adapted from Lin-Moshier & Merchant [51] with the following modifications. Ten oocytes with currents � 500 nA in response to 100 μM ACh were pooled and suspended in 100 μL of homogenization buffer (50 mM HEPES, pH = 7.5, 150 mM NaCl, 1 mM PMSF, 1 mM EDTA, pH = 8.0 and Protease Inhibitor cocktail, Sigma Aldrich, MO, USA). The homogenized sample was centrifuged at 800g for 5 min at 4˚C, and the supernatant was transferred into a clean tube. The centrifugation step was repeated twice to ensure complete removal of the yolk particles. Samples were boiled with Laemmli buffer for 5 min and then subjected to electrophoresis (4-12% Bis-Tris gel for Cel-EAT-2-GFP and 16% Tricine gels for Cel-EAT-18-His). The gels were blotted onto PVDF membranes and blocked with BlockOut blocking buffer (Rockland Immunochemicals Inc., PA, USA). The blots were probed with a 1:10000 dilution of primary antibody (mouse monoclonal anti-GFP antibody to detect GFP tagged Cel-EAT-2, Proteintech # 66002-1-Ig; HRP-conjugated anti-His antibody to detect His tagged Cel-EAT-18, Proteintech # HRP-66005) at 4˚C overnight. HRP conjugated antimouse antibody at 1:10000 dilution was used as the secondary antibody (SA00001-1) for GFP tagged protein. Immunoreactivity was visualized by enhanced chemiluminescence (GE healthcare, IL, USA).