# General discussion {#discussion} ## Environmental levels of neonicotinoids do not impact on the behavior or development of *C. elegans* Increased use of insecticides requires a better understanding of their environmental impact. Therefore, the initial aim of this project was to investigate the effects of neonicotinoid-insecticides on the Nematoda representative *C. elegans*. Neonicotinoids have been introduced to the market in the 1990s and since then have become the most commonly used insecticides worldwide [@jeschke2011]. They have many advantages, including a high potency against a wide range of pest insects (Section \@ref(potentpests)) and low mammalian toxicity. However, neonicotinoids can have significant field effects on non-target species. The adverse effects of environmental neonicotinoids on bees have been studied for decades (Section \@ref(sublethalbees)), whereas the effects on other ecologically important species is less understood. There are several forms in which neonicotinoids are delivered onto fields including spray and granules, but seed-coating is the most common method [@jeschke2011]. This focused application ensures the presence of drug inside the plant at effective concentrations [@stamm2016], nevertheless neonicotinoids have a long half-life and leeching potential, therefore can reside in soil for prolonged time periods, coming in contact with an inhabiting organisms. This leads to potential exposure to soil nematodes and earth worms, two important cultivars of the biomass and nutrient cyclers, that contribute to the soil fertility [@ingham1985; @neher2001]. The lethal dose of neonicotinoids on earths worms and nematodes varies between 2.74 and 62.08 $\mu$M (Table \@ref(tab:toxallanimal)), however neonicotinoids at lower doses can induce sublethal effects (Section \@ref(effectsofneonicsonworms)). The investigations into their effects on the nematode representative and model organism *C. elegans* were carried using thiacloprid, clothianidin and nitenpyram. The effects of these three compounds on locomotion in liquid and solid media, egg-laying and egg-hatching of wild-type worm were investigated (Chapter 3 and 4). This study reports low potency of neonicotinoids against wild-type *C. elegans* [@kudelska2017]. Neonicotinoids were either not effective, or effective at mM concentrations. These data suggests that neonicotinoids have minimal, if any, effect on locomotion, reproduction or feeding of *C. elegans*, however it is possible that they affect more intricate aspects of worm’s physiology. They impact learning and plasticity in honeybees [@williamson2013], therefore efforts into determining their effects on the ability to perceive and process information could be made. For example, decision making could be tested utilising food leaving [@shtonda2006] or chemotaxis assays [@law2004], whereas learning and memory, using olfactory associative assays [@Kauffman2010]. Exposure of *C. elegans* mutant with increased cuticle permeability [@xiong2017] resulted in increased sensitivity of *C. elegans* to tested compounds. Ten-fold increase in the potency of nitenpyram was noted, as indicated by the shift in the EC~50~ in the thrashing assay. Thiacloprid and clothianidin did not affect a wild-type strain, but inhibited locomotion with an EC~50~ of 337.6 $\mu$M and and 3.5 mm, respectively (Section \@ref(effectsofneonicsonthrashing)). The nematode cuticle is the major route of entry for many drugs, therefore the ability of compounds to cross the cuticular barrier often defines their potency [@alvarez2007]. The cuticle limits bioavailability of many compound used in agriculture [@xiong2017] and by doing so it protects the worm against their potentially harmful effects. The cuticle encapsulates the body of nematodes as well as earthworms. Their structures have several similarities, as revealed by the electron microscopy of redworm *L. terrestris*, greenworm *Allolobophora chlorotica* [@reed1948; @coggeshall1966; @knapp1971] and *C. elegans* preparations [@cox1981]. The main components of the cuticle are the collagen fibrils embedded within the matrix. This relatively thick layer is covered by a much thinner epicuticle, which consists mainly of lipids. In earthworms, the epicuticle is perforated by a coat of outward projections, shorter ellipsoidal bodies and longer microvilli. In *C. elegans* the surface coat consists of the glycoprotein-mesh. Based on the architectural and chemical similarities, it is likely that the cuticle of earthworms may also play a role in drug permeability. This is however poorly investigated. Generally, the concentrations at which neonicotinoids impair on *C. elegans* behaviour are several folds higher than those found in the field [@sanchez-bayo2016] and concentrations effective against insect [@goulson2013]. These data suggests *C. elegans* is not impacted by neonicotinoids in the field, however no field-studies are available to confirm this. (ref:generaldisc-phylogeneticseqalign) **Amino acid sequence alignemnt of the nAChR ligand binding domains from *C. elegans* and thiacloprid-susceptible species.** Figure taken from Hopewell et al. 2017. ```{r generaldisc-phylogeneticseqalign-label, fig.cap="(ref:generaldisc-phylogeneticseqalign)", fig.scap= "Amino acid sequence alignemnt of the nAChR ligand binding domains from \\textit{C. elegans} and thiacloprid-susceptible species.", fig.align='center', echo=FALSE} knitr::include_graphics("fig/gen_discussion/phylogeneticsequence70.png") ``` The sensitivity of *C. elegans* to neonicotinoids differs from that of parasitic nematodes and earth worms. In *C. elegans* neonicotinoids impact on locomotion at low mM concentrations (Figure \@ref(fig:thrashing-tc-comp-label) and \@ref(fig:BB-plot-label)), but have no effects on reproduction (Figure \@ref(fig:EL-plot-capt-label) and \@ref(egg-laying-lbl)). In the plant parasitic nematode, neonicotinoids are effective at $\mu$M concentrations. The LD50 of thiacloprid on the root-knot nematode *M. incognita* is 143.23 $\mu$M [@dong2014], whereas the IC~50~ from the *C. elegans* egg-hatching experiments is 300 $\mu$M [@dong2014; @dong2017]. Earthworms seem to be the most susceptible, with reported toxic lethal doses greater or equal to 2.74 $\mu$M (Table \@ref(tab:toxallanimal)) and doses effective against their reproduction and mobility at 488.85 nM, or higher (Section \@ref(effectsofneonicsonworms)). Thus, there is a differential neonicotinoid-susceptibility between *C. elegans* and parasitic nematodes and between *C. elegans* and earth worm species. The same is true among the insects, where some species are much more susceptible than others (Section \@ref(tab:toxallanimal)). Therefore, the sensitivity of each species to neonicotinoids should be considered separately when evaluating the environmental impact of these insecticides. This highlights the complexity and difficulty of the neonicotinoids risk characterisation and usage management. Can the differential susceptibility to neonicotinoids be predicted based on nAChR sequence analysis and phylogenetic relationships? Comparison of the amino acid sequence of *C. elegans* muscle-type receptor with homologous subunits from neonicotinoid-susceptible species of insects, fish and annelid (segmented) worms revealed amino acids important in thiacloprid binding are highly conserved. However, phylogenetically, insect and *C. elegans* receptors are the most distantly related to insects that segmented worms and fish, which could explain lower susceptibility of *C. elegans* to clothianidin, thiacloprid and nitenpyram in comparison to earthworms. (ref:gendiscussion-celegansphylogenetictree) **Neighbour joining phylogenetic tree analysis of the amino acid sequences of nACRs ligand binding domains from *C. elegans* and thiacloprid-susceptible species.** To generate the tree, the ligand binding domain of *C. elegans* ACR-16 subunits and highly homologous sequences from parasitic nematode and representative species of insects, annelid worms and fish were selected. Bootstrap values are shown as percentages after 1000 replicates. The scale bar indicates the fraction of substitutions per amino acid. Figure taken from Hopewell et al. 2017. ```{r gendiscussion-celegansphylogenetictree-label, fig.cap="(ref:gendiscussion-celegansphylogenetictree)", fig.scap= "Neighbour joining phylogenetic tree analysis of the amino acid sequences of nACRs ligand binding domains from \\textit{C. elegans} and thiacloprid-susceptible species.",fig.align='center', out.width= '40%', echo=FALSE} knitr::include_graphics("fig/gen_discussion/phylogenetic-tree.png") ``` ## *C. elegans* as a model to study the mode of action of neonicotinoids The adverse effect of neonicotinoids on biological pollinators and the emerging resistance (Section \@ref(resgenevidence)) creates a threat to the food safety and highlights the need for the synthesis of novel and more selective insecticides. One of the first steps towards the synthesis of new insecticides is the understanding of their mode of action [@metcalf1971]. Neonicotinoids act by targeting insect nAChRs (Section \@ref(neonicstarget)), however their mode of action and receptor specificity differs. They can have agonistic, antagonistic and super-agonistic action, depending on the animal preparation upon which they are applied and on the neonicotinoids itself (Section \@ref(moaneonicsinsects)). In addition, even neonicotinoids sharing the same pharmacophore, target distinct receptors [@thany2009; @moffat2016]. Difficulties in the heterologous expression of insect's receptors hiders their pharmacological characterisation and description of neonicotinoid-receptor specificity (Section \@ref(ric3insect)). The model organism *C. elegans* is an alternative system in which the mode of action of chemical agents can be studied *in-vivo* [@lewis1987; @lewis1980]. *C. elegans* expresses at least 29 nAChR subunits (Figure \@ref(fig:seqidentityecd-label)) [@jones2007a] which form receptors at the neuromuscular junction and in the nervous system. Muscle-type receptors are involved in the regulation of locomotion, reproduction and feeding (Section \@ref(pharmacelegans) \@ref(nachrutantfeeding)). Neuronal type receptors are expressed in circuits involved in the sensory processing and chemosensation [@yassin2001]. The suitability of the *C. elegans* system for the mode of action studies was investigated by scoring the sensitivity of native pharyngeal nAChRs receptors to neonicotinoids. The pharynx is a neuromuscular system in which feeding is carried out by a musculature under the influence of the pharyngeal nervous system. In the presence of food, 5-HT is released, which stimulates MC neuron. MC releases acetylcholine which acts directly on EAT-2 containing nAChRs to drive fast pumping. Thus, EAT-2 is a molecular determinant of the fast pharyngeal response. Although EAT-2 is the only nAChR subunit with clearly characterised function, ACR-7 is also expressed in the pharyngeal muscle [@saur2013]. To determine the effects of neonicotinoids on the pharyngeal nAChRs, dissected head preparations were exposed to clothianidin, thiacloprid and nitenpyram. Their effects on 5-HT stimulated pharyngeal pumping were investigated. Nitenpyram and clothianidin inhibited 5-HT stimulated pumping at 25 mM and 500 $\mu$M, respectively (Section \@ref(dissectedanimalnicotineandneonics)). The impact of neonicotinoids on unstimulated pharynx was also investigated and revealed no effects of thiacloprid and nitenpyram. In contrast clothianidin stimulated pumping. The lowest dose of clothianidin effective against this *C. elegans* behaviour was 75 $\mu$M. These data suggest nAChR expressed in the pharynx have low affinity to neonicotinoids. In contrast, neonicotinoids are effective at nM concentrations on target insect receptors (section \@ref(electrophysevidence) and \@ref(ligbinding)) suggesting nAChR pharmacophore present in pest and beneficial insects are distinct from those found in the *C. elegans*. (ref:gendiscussion-celegansinsectalagnment) **Sequence alignment of the pharmacophore of insect and *C. elegans* nAChR subunits.** Ligand binding pocket is formed from the loops originating from the principal (a) and complementary (b) receptor subunits. Amino acids important in forming drug-receptor interactions are color-coded as in Figure \@ref(fig:binding-pocket-label). Non-conserved residues are underlined. Numbering is according to the AChBP of Ls sequence. Mp = *M. persicae* (peach aphid), Dm = *D. melanogaster* (fruit fly), Ce = *C. elegans*. Sequence alignment generated by MUSCLE. ```{r gendiscussion-celegansinsectalagnment-label, fig.cap="(ref:gendiscussion-celegansinsectalagnment)", fig.scap= "Sequence alignment of the pharmacophore of insect and \\textit{C. elegans} nAChR subunits", fig.align='center', out.width= '80%', echo=FALSE} knitr::include_graphics("fig/gen_discussion/celegans_and_insect_added_loopG_and_B.png") ``` The nAChR pharmacophore consists of the contributions from the $\alpha$ (principal) subunit and non-$\alpha$ (complementary) subunit (Section \@ref(bindingsite) and \@ref(pharmacophore)). Residues important in binding of agonists have been identified in mollousc AChBP and nAChR bound to agonists, including nicotine and acetylcholine, thiacloprid, imidacloprid and clothianidin [@celie2004; @hansen2005; @ihara2008; @talley2008; @ihara2014; @zouridakis2014; @morales-perez2016]. The residues identified as those important in binding of agonists were compared between worm and insect nAChRs, by aligning their sequences. The chosen insect subunits are those that form receptor chimeras in the recombinant system and confer high binding affinity to neonicotinoids (i.e Kd below 10 nM, Section \@ref(chimerareceptors) and Table \@ref(tab:bindignrecombinant)) and $\beta1$ subunit, identified as a molecular determinant of neonicotinoid-resistance in *M. persicae* [@bass2011]. Amino acid sequences of these subunits were aligned against nAChR subunits forming functional receptors at the body wall muscle, namely ACR-16 [@ballivet1996; @boulin2008; Section \@ref(muscletypenachr)], as well as EAT-2 and ACR-7 which are two subunits identified in the pharyngeal system [@mckay2004; @saur2013 and Section \@ref(eat2atthepharynx)]. A comparison of nAChR binding pocket residues in *C. elegans* and insect receptors identified several differences, in particular in the loops from the complementary site of the binding pocket, loops D, E and G. This might suggest that contributions from the complementary site determine neonicotinoids-selectivity [@marotta2014; @hansen2004; @harpsoe2013; @marotta2014]. However, there are receptor subunits, such as ACR-16, in which almost all residues are conserved, implying that susceptibility to neonicotinoids cannot be predicted based on the sequence of nAChR binding pocket, or that residues important in neonicotinoid binding have not been identified correctly and that crystal stuctures of insect receptors are needed to better understand the interactions between these insecticides and their target. ## *C. elegans* pharynx as a platform for the pharmacological characterisation of nAChRs Low sensitivity of *C. elegans* to nitenpyram, clothianidin and thiacloprid in behavioural and cellular assays precludes its use as a model to study the mode of action of neonicotinoids per se, but highlights its potential use as a suitable background for the heterologous expression of insect nAChRs. New insecticides are needed to prevent the negative environmental impact of neonicotinoids and combat emerging resistance. The first step towards the synthesis of neonicotinoids with improved selective toxicity profile is the heterologous expression of nAchRs from pests and non-target species in a suitable host, but no heterologous or native insect nAChR complexes have been expressed yet. Cell lines and *Xenopus* oocytes are routinely used as biological systems for heterologous nAChR expression (Section \@ref(biologicalsystemfornachrexpression)). Model organism *C. elegans* is an alternative system in which recombinant proteins can be expressed by generation of transgenic worms [@crisford2011; @sloan2015]. In comparison to other systems, *C. elegans* is cheap and easy to maintain, whereas transgenic worms can be preserved for years at - 80 $^\circ$C. The function of recombinant nAChRs can be studied *in-vivo* by evaluating their impact on behaviours underpinned by the cholinergic neurotransmission. Acetylcholine is a major neurotransmitter of the pharynx, released by at least 7 out of 14 neurons (Table \@ref(tab:pharynx-neurons)). It is key and necessary for the induction of fast feeding in response to food (Section \@ref(achpumping)). Its action on the pharynx is mediated by nAChRs, most notably EAT-2 expressed at the NMJ (Section \@ref(eat2atthepharynx)). The *C. elegans* pharynx does not possess an innate susceptibility to low concentrations of neonicotinoids, creating an appropriate genetic background in which effects of these compounds on recombinantly expressed receptor can be studied. In addition, it expresses an array of chaperon proteins, creating a favorable environment for the maturation and function of these ion channels (Section \@ref(cematnachr)). *C. elegans* pharynx has been successfully used for the heterologous expression of non-native proteins [@crisford2011], including nAChRs [@sloan2015]. EAT-2 is a single nAChR subunit that confers pharyngeal 5-HT sensitivity and feeding response in *C. elegans* [@mckay2004]. We show that *eat-2* mutant is a suitable genetic background, in which the expression of heterologous nAChRs could be scored in behavioural assays. Transgenic line in which *eat-2* is heterologously expressed, rescued the blunted pharyngeal response to food and restored 5-HT resistance of the *eat-2* mutant (Section \@ref(behaviourofeat2rescue)). However, the expression of human $\alpha7$ nAChR in the *C. elegans* mutant pharynx had no observable phenotypical consequences (Section \@ref(feedingalpha7celegans) and Section \@ref(htandtransgenicalpha7)). Similarly, expression of this receptor in the wild-type worm revealed no differences in pharyngeal responses to nAChR agonists nicotine or choline (Section \@ref(pharmaalpha7transegnicworms)). Staining with fluorescently labelled human $\alpha7$ nAChR antagonist $\alpha-Bgtx$, reveled increased fluorescence in the pharyngeal muscle of transgenic, when compared to control worms (Section \@ref(bgtxstaining)). $\alpha$-bgtx binds to the extracellular, domain of the receptor [@dellisanti2007]. This suggests that $\alpha7$ receptor is expressed on the cell surface of the pharyngeal muscle, however due to the lack of phenotype its function is unclear. Further pharmacological experiments of transgenic strains (as described in the Discussion of Chapter 6) should be carried out to determine whether $\alpha7$ retains its function upon expression in the pharynx of *C. elegans*. The expression of $\alpha7$ was driven by a myo-2 promoter, which should drive expression in all cells of the pharyngeal musculature [@altun2009a]. However, $\alpha-Bgtx$ staining of transgenic worms was concentrated in pm7 and pm8 muscle cells, suggesting $\alpha7$ is expressed in the terminal bulb, which does not overlap with the localisation of native EAT-2, native protein. Thus, native EAT-2 promoter should be used to ensure correct localisation of the expressed protein. ## *C. elegans* as a model for mammalian toxicity studies To better understand the pharmacological profile of *C. elegans* nAChRs, the effects of endogenous neurotransmitter acetylcholine as well as agonist nicotine and cytisine were tested on the pharynx. Acetylcholine, nicotine and cytisine were applied on cut-head and the effects on EPG was assayed. All three induced potent and transient stimulation of pumping leading to muscle tetanus (Figure \@ref(fig:epg-nicotine-2-label)). Concentrations effective were in the low $\mu$M range. The order of potency as measured by the EC~50~ was: nicotine > cytisine > acetylcholine. In comparison to human $\alpha7$, effective concentrations were in a similar range [@papke2002]. In addition, like pharyngeal nAChRs, human $\alpha7$ receptors insensitive to low doses of neonicotinoids with the EC~50~ values of 0.74 mM and 0.73 mM, respectively for clothianidin and imidacloprid, on the heterologously expressed channel [@cartereau2018]. This suggests the pharmacophore of human $\alpha7$ and pharyngeal nAChRs is conserved. This was confirmed by aligning the sequences of key amino acids forming the nAChR binding pocket in $\alpha7$ and two of the pharyngeal nAChRs (Figure \@ref(fig:pharmacophoreceleganspharynxandhumanalpha7-label)). (ref:pharmacophoreceleganspharynxandhumanalpha7) **Amino acid sequence alignment of human $\alpha7$ and two of the pharyngeal *C. elegans* nAChRs ligand binding pockets.** Amino acid sequences forming nAChR binding pocket were aligned. Amino acids important in agonist binding are highlighted, as in Figure \@ref(fig:binding-pocket-label). Ce = *C. elegans*, Hs = human. Sequence alignment generated with MUSCLE. ```{r pharmacophoreceleganspharynxandhumanalpha7-label, fig.cap="(ref:pharmacophoreceleganspharynxandhumanalpha7)", fig.scap="Amino acid sequence alignemnt of human $\\alpha7$ and pharyngeal \\textit{C. elegans} nAChRs ligand binding pockets.", fig.align='center', out.width= '150%', echo=FALSE} knitr::include_graphics("fig/gen_discussion/celeganseat2andacr7andhumanalpha7.png") ``` Comparison of human $\alpha7$ and pharyngeal nAChR subunits EAT-2 and ACR-7 revealed high sequence similarity. Almost all residues forming ligand binding pocket are conserved between these subunits, suggesting pharyngeal nAChR subunits are homologous to human $\alpha7$. Besides pharyngeal nAChR subunits, homologs of over two thirds of human proteins can be found in *C. elegans* [@sonnhammer1997; @lai2000]. The similarities between mammals and *C. elegans* extends beyond genetics. *C. elegans* has conserved synaptic function, due to the presence of almost all vertebrate neurotransmitters and conserved neuronal signalling pathways [@bargmann1998; @kaletta2006]. There are however some differences. *C. elegans* expresses inhibitory glutamate-gated chloride channel which is absent in mammals [@cully1994], but lacks voltage-gated sodium channels which is present in humans [@bargmann1998]. Despite these limitations, *C. elegans* emerges as an attractive model for toxicity studies [@hunt2017]. Methods to use *C. elegans* as a model to study acute toxicity as well as developmental and reproductive toxicology have been developed [@boyd2010; @xiong2017]. A good correlation between the rank order of toxicity of many compounds on *C. elegans* and mammals for acute toxicity, growth and reproduction endpoints have been found [@williams1988b; @boyd2010]. This includes organophosphates, which act by disrupting cholinergic neurotransmission at the synapse [@chadwick1947]. The rank order of acute toxicity, as measured by LC~50~, correlated well with the LD~50~ ranking of these agents in rats and mice [@cole2004]. This highlight the potential suitability of *C. elegans* as a model for toxicity testing of cholinergic and other neurotoxins. Introduction of *C. elegans* into toxicity testing has a potential to reduce the use of conventional mammalian models, resulting in the reduction of cost and duration of such studies [@williams1988].