In order to facilitate genetic understanding of the neuroactive properties of alcohol we have devised a new approach for studying ethanol withdrawal in C. elegans
. An important aspect of this study is that we used the observation of withdrawal relief as a criterion to discriminate between adaptations directly caused by alcohol, and adaptations resulting indirectly from the pleiotropic effects of alcohol on behaviour. The rationale for this is based on the hypothesis that a low dose of alcohol relieves withdrawal by resetting the balance of signalling in neural circuits that have undergone ethanol-induced allostatic and, or, homeostatic adaptation 
. Thus only those behaviours which are a genuine expression of alcohol-induced neuroadaptation would be predicted to be subject to withdrawal relief.
The assay for withdrawal and withdrawal relief developed here was based on a food race assay. In this assay, C. elegans that were either acutely exposed to ethanol, or chronically exposed and then withdrawn were shown to have a reduced ability to navigate towards food. More detailed analysis of this impairment indicated that for acutely exposed worms this primarily resulted from a reduction in the speed and efficiency of movement while the pattern of locomotory behaviour in terms of the frequency of reversals was relatively unaffected. In marked contrast, worms in ethanol withdrawal appeared severely uncoordinated and exhibited reduced reversal frequency and an increased frequency of omega turns in the first 5 min in a food race task.
We next considered the different components of the withdrawal behaviour i.e. the effect on reversals and on omega turns, to determine whether or not either of these was likely to reflect direct ethanol-induced neuroadaptation. The decrease in reversal frequency was not relieved by a low dose of ethanol and may, at least in part, be explained by an adaptation to food deprivation brought about by ethanol inhibition of pharyngeal pumping 
. This is supported by the rather surprising observation that food deprivation also impaired performance in the food race. However, previous reports indicate that food deprivation inhibits reversal frequency 
and this provides an explanation for poor navigational performance.
The second component of withdrawal behaviour we describe here was the expression of exaggerated body postures. This consisted primarily of a behaviour termed omega turns. In wild-type worms these high angled bends are tightly coupled to reversals in movements called ‘pirouettes’ during exploratory behaviour 
. In ethanol withdrawal, omega turns occurred independently of reversals, leading to unaccompanied omega turns. Previous studies have indicated that unaccompanied omega turns are a rare behaviour in wild-type animals 
and, furthermore, we are not aware of reports of mutants that exhibit this behaviour. The occurrence of unaccompanied omega turns is not simply due to an uncoupling of reversals from omega turns because the absolute frequency of omega turns and the overall posture measure, which indicates bendiness, were both increased. Furthermore, this behaviour was specific to ethanol withdrawal and was not seen during food deprivation. Most importantly, it was ameliorated by a low dose of ethanol. From this we deduce that it is an adaptive behaviour that is expressed when ethanol is withdrawn from worms that have been chronically exposed to ethanol for 6 h or longer. Worms which had developed withdrawal also showed tolerance to a high dose of ethanol, consistent with observations in other animals including humans that these phenomena are inter-related.
The occurrence of unaccompanied omega turns in withdrawal, and the relief of this by a low dose of ethanol, suggests that chronic ethanol exposure has altered the balance of signalling in the command circuits that regulate locomotory pattern during navigational behaviour. Some insight into the neural pathways that might be affected is provided by a study detailing the circuitry that drives reversals and omega turns during exploration 
. These behaviours are differentially regulated by distinct motor circuits. The head and neck motorneurones, SMD and RIV, direct omega turns whilst the forward and backward command interneurones control reversals 
. Intriguingly, laser ablation of the glutamatergic reverse command interneurone AVA resulted in worms that exhibited omega turns in the near complete absence of reversals i.e. unaccompanied omega turns 
and thus superficially would appear to phenocopy this aspect of ethanol withdrawal. Further neurones of interest are the head motorneurones, SMB, SMD and RIV. Laser ablation of SMB increases the amplitude of dorsal-ventral head turns leading to very loopy movement whilst laser ablation of SMD and RIV has the opposite effect leading to a decrease in omega turns 
. This suggests that these motorneurones play a key role in regulating the amplitude of head flexion required to make turns. Thus in ethanol withdrawal the output from SMB, SMD and RIV may be altered. These observations highlight the opportunity for an integrative analysis of adaptive responses to ethanol provided by defining withdrawal in a genetically tractable animal in which the circuits driving sub-behaviours are relatively simple and delineated.
As a first step to interrogating the genetic basis of withdrawal behaviour in C. elegans
we investigated whether mutants previously reported to affect ethanol dependent behaviours altered responses in the food race paradigm. However, neither the NPY-like peptide receptor, NPR-1, previously reported to phenocopy the behaviour of C. elegans
in ethanol withdrawal 
nor the BK potassium channel SLO-1, previously reported as ethanol resistant 
, showed a markedly different profile of ethanol responses compared to wild-type.
There is a consensus that peptidergic pathways come in to play particularly when an animal is subject to environmental stressors 
and neuropeptides are strong candidates as ethanol effectors in addiction 
therefore we further investigated the role of neuropeptides using a more inclusive approach i.e. by employing egl-3(ok979)
, a putative null for a neuropeptide precursor convertase enzyme 
and severely depleted in neuropeptides 
. Interestingly, this mutant did not exhibit withdrawal but was affected by acute exposure to ethanol. The phenotypes we observed for egl-3
were otherwise relatively mild, similar to earlier studies 
, and serve to illustrate the neuromodulatory role of neuropeptides.
A further consideration is the dose-dependence of the effects of ethanol. In this study the lowest concentration that gave rise to withdrawal signs was around 136 mM, and for relief was 50 mM. These concentrations were applied externally and it has previously been reported that the internal concentration is ten fold lower than the external 
. However, observations on the fast kinetics of ethanol effects in intact worms 
and the similar sensitivity of exposed and intact tissue to ethanol 
conflict with this and suggest that equilibration of ethanol across the worm cuticle is likely to be rapid. Therefore we predict that the internal concentration of ethanol approximates to the external concentration. Importantly, the rapid kinetics of the ethanol effects makes it impossible to accurately determine the internal concentration using a biochemical assay and so this issue remains to be resolved.
In terms of the withdrawal relief experiments, we observed an effect at 50 mM ethanol. On the basis of the argument outlined in the paragraph above, we predict that the tissue concentration in the worm in these experiments is very close to 50 mM. It could be argued that the relieving concentration of ethanol, 50 mM, is additive to ethanol remaining inside the worm following the ethanol conditioning procedure. However, evidence from our own studies and the literature indicate this is unlikely to be the case. Thus whilst there is some dispute concerning the internal concentration of ethanol in worms following immersion in a known external concentration 
an aspect that all the publications agree on is that if you take a worm that has been placed in a high concentration of ethanol (300 to 400 mM) and then rapidly collect, cool, wash and homogenise the worms then the concentration of ethanol measured in the worms is approximately 20 mM 
. Therefore, as we included an extended wash procedure, with no cooling, we can predict that the concentration remaining inside the worm after ethanol conditioning and washing is less than 20 mM. Thus, in the withdrawal relief experiments conducted at 50 mM ethanol it is unlikely that the tissue concentration is much in excess of this value. On this basis, the ability to induce a withdrawal phenomenon that was subject to ethanol relief with 50 mM ethanol indicates that C. elegans
can exhibit responses to ethanol within a range that is sub-lethal in higher animals, including human.
In conclusion, this study demonstrates the phenomenon of ethanol withdrawal and withdrawal relief in C. elegans
. Whilst this simple animal system cannot model complex aspects of human alcohol addiction such as motivation, craving and cue-dependent relapse 
it can provide a reductionist correlate of ethanol-induced neural plasticity which underpins negative reinforcement and therefore contributes to alcohol addiction. By defining, and investigating the genetic basis for, distinct but inter-related ethanol-induced behavioural states in C. elegans
() we have shown that neuropeptide signalling has a pivotal role in ethanol withdrawal. Further studies are required to identify the precise neural substrate of this adaptive response in C. elegans
. In particular, it will be very informative to identify which C. elegans
neuropeptides and neuropeptide receptors are required for the egl-3
dependent effect of alcohol withdrawal. Despite the fact that the C. elegans
neuropeptides are a large and diverse family of approximately 250 molecules 
this is nonetheless a tractable problem given the molecular techniques that can be applied in this model system. As far as the receptors are concerned, there are 60 rhodopsin-like G protein-coupled receptors in C. elegans
predicted to bind small molecule neurotransmitters or neuropeptides 
and these are therefore candidate effectors for alcohol withdrawal. Within this group there are homologues of many of the mammalian neuropeptide receptor families 
and further studies will be aimed at defining which of these are involved in C. elegans
withdrawal. For the time-being we can note a key role for neuropeptides in this phenomenon. This provides a tantalizing parallel with mammalian studies in which neuropeptides have also been implicated in chronic responses to ethanol, including NPY 
, CRF 
, the opioid peptides 
and tachykinins 
. This reinforces the importance of conserved pathways consistent with current investment in the development of drugs targeting peptidergic systems for the treatment of alcoholism 
. A recent focus of interest comes from data from mammalian studies which implicate hyperactivity in the extra-hypothalamic neuropeptide corticotrophin-releasing factor (CRF) signalling pathway in behavioural neuroadaptation to ethanol and stress-induced reinstatement of alcohol consumption 
. A seductive model is one in which chronic ethanol acts as a stressor to trigger neuropeptidergic signalling and neural plasticity in an evolutionarily conserved manner.
A model to describe the effects of acute and chronic ethanol on the perfomance of C. elegans in the food race task.