Identification and functional expression of a glutamate- and avermectin-gated chloride channel from Caligus rogercresseyi, a southern Hemisphere sea louse affecting farmed fish

Abstract

Parasitic sea lice represent a major sanitary threat to marine salmonid aquaculture, an industry accounting for 7% of world fish production. Caligus rogercresseyi is the principal sea louse species infesting farmed salmon and trout in the southern hemisphere. Most effective control of Caligus has been obtained with macrocyclic lactones (MLs) ivermectin and emamectin. These drugs target glutamate-gated chloride channels (GluCl) and act as irreversible non-competitive agonists causing neuronal inhibition, paralysis and death of the parasite. Here we report the cloning of a full-length CrGluClα receptor from Caligus rogercresseyi. Expression in Xenopus oocytes and electrophysiological assays show that CrGluClα is activated by glutamate and mediates chloride currents blocked by the ligand-gated anion channel inhibitor picrotoxin. Both ivermectin and emamectin activate CrGluClα in the absence of glutamate. The effects are irreversible and occur with an EC(50) value of around 200 nM, being cooperative (n(H) = 2) for ivermectin but not for emamectin. Using the three-dimensional structure of a GluClα from Caenorabditis elegans, the only available for any eukaryotic ligand-gated anion channel, we have constructed a homology model for CrGluClα. Docking and molecular dynamics calculations reveal the way in which ivermectin and emamectin interact with CrGluClα. Both drugs intercalate between transmembrane domains M1 and M3 of neighbouring subunits of a pentameric structure. The structure displays three H-bonds involved in this interaction, but despite similarity in structure only of two these are conserved from the C. elegans crystal binding site. Our data strongly suggest that CrGluClα is an important target for avermectins used in the treatment of sea louse infestation in farmed salmonids and open the way for ascertaining a possible mechanism of increasing resistance to MLs in aquaculture industry. Molecular modeling could help in the design of new, more efficient drugs whilst functional expression of the receptor allows a first stage of testing of their efficacy.

Figure 1. Alignment of CrGluClα amino acid sequence with other glutamate-gated chloride channel α subunits.

PROMALS3D program (http://prodata.swmed.edu/promals3d/) was used to generate an alignment with the following glutamate receptor GluClα proteins: L. salmonis (Ls) GI:115361509, D. melanogaster (Dm) GI:1507685 and C. elegans (Ce) GI:559559. The predicted signal peptide of CrGluClα, identified using SignalP 4.1 programme, is shown in bold italic. M1–M4 are transmembrane domains; cysteines involved in C-loops are highlighted in yellow; in grey and in red are residues involved in glutamate and ivermectin binding respectively. Arrows indicate amino acids variations between Caligus obtained from different sources (see Figure 2). The position of the transmembrane domains and binding sites is based on the crystal structure of the C. elegans GluClα .

Figure 2. Variations in deduced CrGluClα primary structure.

The rooted phylogenetic tree was done by using the UPGMA method after aligning twelve representative clones with CLUSTALW software (http://www.genome.jp/tools/clustalw). The positions of amino acid changes (V27I, D73G, R387K and L411Q), or deletions (isoleucine 20 (I or Δ), and alanine 376 plus serine 377 (AS or Δ)) are shown. The clone name is composed by CrGluClα, source and Arabic number. Source: Valdivia (Vald), Darwin (Dw) and Errázuriz (Err). Arabic number is the number assigned during the cloning process.

Figure 3. Glutamate-activated picrotoxin-sensitive currents of CrGluClα RNA-injected Xenopus oocytes.

A. Current traces obtained at the indicated membrane potentials during bath application of glutamate at concentrations from 1 to 100 µM during the times shown in the boxes. Current and time calibration bars shown apply also to recordings shown in panels B and D. B. Picrotoxin (PTX) at 100 µM reversibly abolishes glutamate-evoked current. The trace illustrated was taken at 60 mV. Respective current voltage relations taken from voltage-ramps applied during this experiment are shown in panel C. In panels D and E, it is shown that most of the glutamate evoked outward current is dependent on the presence of extracellular Cl⁻, all but 7.6 mM of which was replaced by gluconate during the time identified as Low [Cl⁻]_o. The current-voltage relations in E have been corrected by subtraction of the current remaining in 100 µM PTX. Panel F shows a graph summarizing results (means±SEM from 7 experiments) of the concentration-response relationship of the glutamate-sensitive currents. The responses were normalised to the maximal effect evoked by glutamate. The line is a Hill equation fitted simultaneously to both −80 and 60 mV data sets giving values for EC₅₀ of 6.89±0.83 µM and n_H of 1.33±0.18.

Figure 4. Irreversible activation of CrGluClα by ivermectin.

A. Current trace obtained at 60 mV during bath application of 50 µM glutamate and then to ivermectin at concentrations going from 1 to 3000 nM during the times shown in the boxes. The effect of picrotoxin addition at 100 µM is shown at the end of the trace. The trace in panel B shows that whilst ivermectin-induced current blockade by picrotoxin is reversible, the current evoked by the ivermectin is persistent even after prolonged washing. As for the glutamate response, ivermectin-induced outward current was greatly decreased upon decreasing [Cl⁻]_o to 7.6 mM. In C and D current voltage relations taken from voltage-ramps applied during experiment in B are shown. Those in D have been corrected by subtracting the current remaining in the presence of 100 µM PTX. E. Dose-response relationship of the Ivermectin-sensitive currents. Data are normalized (mean ± SEM) to the maximal effect of ivermectin and originate in five separate experiments. The line is a Hill equation fitted simultaneously to measurements taken at −80 and 60 mV giving a value for EC₅₀ of 181±10 nM, with corresponding n_H value of 2.1±0.26.

Figure 5. Irreversible activation by emamectin of CrGluClα receptor expressed in Xenopus oocytes.

A. Current trace obtained at 60 mV during bath application of emamectin at concentrations going from 5 nM to 10 µM during the times shown in the boxes. The effect of picrotoxin addition at 100 µM is also shown immediately after removal of emamectin. Wash out of picrotoxin in the absence of emamectin returns current to the levels attained in the presence of high concentration of emamectin. B: Current-voltage relations taken from voltage-ramps applied during experiment in A. Control denotes current prior to emamectin addition. C. Dose-response relationship of the emamectin-sensitive currents. Data are normalized (mean ± SEM) to the maximal effect of emamectin and originate in seven separate experiments. The line is a Hill equation fitted simultaneously to measurements taken at −80 and 60 mV giving a value for EC₅₀ of 202±21 nM and n_H 1.1±0.11.

Figure 6. Interaction of emamectin and ivermectin with CrGluClα: results of molecular docking and molecular dynamics experiments.

A and B. The final conformation of the transmembrane domains (M1–M4) of CrGluClα for each subunit (S1–S5) after 140 ns MD runs are shown schematically for the receptor after docking with ivermectin (A) or emamectin (B). Emamectin and ivermectin are shown as stick structures. C–F. Higher magnification views of the drug positions in their binding sites. Dotted lines in E and F indicate H-bonds with the residues in stick structures. The view in C and D is from above the channel having removed the extracellular domain. That in E and F is a side-view. Different colours of the transmembrane domains indicate different subunits. Ivermectin and emamectin are displayed with their disaccharide moiety in yellow with the rest of the molecule in orange.

Figure 7. Conservation of residues involved in drug binding in Caligus and C. elegans GluClα.

Residues involved in hydrogen-bonding (green) and van der Waals (grey) interactions of ivermectin or emamectin with CrGluClα (CrGluClα-IVM, CrGluClα-EMA, see text) are compared with those identified in the CeGluClα-ivermectin complex (CeGluClα-IVM) . Only those interactions existing during >70 ns of the 140 ns MD runs are shown. Residues highlighted in red are those apparently involved in creating a hydrophobic seal in the channel pore in the absence of agonists .

Figure 8. A. Channel pore radius along the z-axis in the receptor alone (black) and in the CrGluClα-emamectin (green) and CrGluClα-ivermectin (red) systems.

A length, mainly intramembrane, of the pore is shown. The discontinuous lines show the pore segment lined by M2 α-helices. Values have been taken at the end of MD trajectories. B and C. Lateral views of the pore at time zero (C) and at the end (B) of the MD trajectory for CrGluClα in absence of drugs. Only four M2 helices are shown with the fifth removed for clarity. Residues P288, A292 and L299 are shown in licorice. Water occupancy is shown as licorice and surface.

Figure 9. CrGluCl-T318A expression in Xenopus oocytes.

A. Current trace obtained in a CrGluClα-T318A-expressing oocyte that exhibited high spontaneous current at the indicated membrane potential. Bath application of glutamate at concentrations from 1 to 100 µM took place during the times shown in the boxes. This was followed by exposure to ivermectin, decrease of bath [Cl⁻] to 7.6 mM and addition of 100 µM PTX. B. Average with SEM of spontaneous currents in CrGluClα-T318A-expressing oocytes (n = 7), and increase in current evoked by 100 µM glutamate (n = 7), and the current decrease upon addition of ivermectin (n = 4) or emamectin (n = 5). C. Current voltage relations of spontaneous current exhibited in a CrGluClα-T318A-injected oocyte in normal bath solution and after all but 7.6 mM extracellular Cl⁻ was replaced by gluconate. The current-voltage relations have been corrected by subtraction of the current remaining in 100 µM PTX. D. Concentration-response relationship of the glutamate-sensitive CrGluClα-T318A currents (means±SEM from 7 experiments) normalised to the maximal effect evoked by glutamate. The response of WT CrGluClα is shown for comparison.

Figure 10. Effect of ivermectin or emamectin upon CrGluClα-T318A-dependent currents.

A and B. Effect of increasing concentrations of emamectin and ivermectin on WT CrGluClα-dependent currents. The currents were first activated by addition of 50 µM glutamate and during continuous exposure to glutamate increasing concentrations of emamemctin or ivermectin were added. Reduction of [Cl⁻] to 7.6 mM or addition of 100 mM PTX are also indicated. C and D. Similar experiments performed using the CrGluClα-T318A mutant. E and F. Concentration dependence of the effects of avermectins on WT and T318A mutant CrGluClα receptors. Normalised average responses to emamectin and ivermectin are shown (means±SEM, n = 6 for all experiments shown). Fits of Hill decay functions to the average values are shown by the solid lines.