Phylogenomics of ligand-gated ion channels predicts monepantel effect

Abstract

The recently launched veterinary anthelmintic drench for sheep (Novartis Animal Health Inc., Switzerland) containing the nematocide monepantel represents a new class of anthelmintics: the amino-acetonitrile derivatives (AADs), much needed in view of widespread resistance to the classical drugs. Recently, it was shown that the ACR-23 protein in Caenorhabditis elegans and a homologous protein, MPTL-1 in Haemonchus contortus, are potential targets for AAD action. Both proteins belong to the DEG-3 subfamily of acetylcholine receptors, which are thought to be nematode-specific, and different from those targeted by the imidazothiazoles (e.g. levamisole). Here we provide further evidence that Cel-ACR-23 and Hco-MPTL-1-like subunits are involved in the monepantel-sensitive phenotype. We performed comparative genomics of ligand-gated ion channel genes from several nematodes and subsequently assessed their sensitivity to anthelmintics. The nematode species in the Caenorhabditis genus, equipped with ACR-23/MPTL-1-like receptor subunits, are sensitive to monepantel (EC(50)<1.25 µM), whereas the related nematodes Pristionchus pacificus and Strongyloides ratti, which lack an ACR-23/MPTL-1 homolog, are insensitive (EC(50)>43 µM). Genome sequence information has long been used to identify putative targets for therapeutic intervention. We show how comparative genomics can be applied to predict drug sensitivity when molecular targets of a compound are known or suspected.

Figure 1. Phylogenetic tree based on the LBD region of putative LGIC genes.

NJ tree (clustalw) from the LBD region of conceptually translated putative LGIC genes as detected with Genewise after an initial Blast screen with the 210 seeds (see Materials and Methods) - 1426 sequences in total. Thousand bootstrap iterations were performed and branches below 50% bootstrap support were collapsed. Nematode sequences are shown in shades of green, platyhelminthes yellow, insect purple and vertebrate red. Some C. elegans and human subunits are labelled, and the labels for proteins involved in drug susceptibility are coloured: levamisole - cyan, monepantel - orange and ivermectin – blue. Green asterisks indicate branches that similar to these latter appear broad and nematode specific and could be attractive for further investigation as targets for other compounds.

Figure 2. Detailed view of the DEG-3 sub-family.

Detailed view of the DEG-3 sub-family from the LBD region NJ tree, with branches below 50% support after 1000 bootstrap iterations joined. A few related subfamilies are shown as collapsed branches. CEL, Caenorhabditis elegans, CBN, C. brenneri, CBR, C. briggsae, CRE, C. remanei, CJA, C. japonica, PPA, Pristionchus pacificus, HCO, Haemonchus contortus, HBA, Heterorhabditis bacteriophora, MHA, Meloidogyne hapla, NM3MIC/MIN, M. incognitia, NB3AYC Ancylostoma ceylanicum, SRA, Strongyloides ratti, BMA, Brugia malayi, TSP, Trichinella spiralis, SME, Schmidtea mediterranea. Pristionchus pacificus lacks a close MPTL-1 homolog and was predicted to be less sensitive to AAD-1566 than species such as C. japonica, H. contortus and H. bacteriophora.

Figure 3. Sensitivity to AAD-1566 determined in vitro.

The percentage of the average number of adult worms present after 3 days exposure relative to a control is plotted versus drug concentration for Caenorhabditis elegans, mutated C. elegans strain acr-23 (cb27), C. japonica, C. briggsae, C. remanei, C. brenneri, Pristionchus pacificus and S. ratti. Sigmoid dose-response curve fit was performed in Prism. EC₅₀ values with standard errors, estimated from data points with four replicates, are shown inset.