GABA withdrawal

Neuroscience

Volume 313, 28 January 2016, Pages 57–72

Cover image

Review

GABA withdrawal syndrome: GABAA receptor, synapse, neurobiological implications and analogies with other abstinences

E. Calixto,

Show more

doi:10.1016/j.neuroscience.2015.11.021

Get rights and content

Highlights

•

GABA withdrawal (GW) increases neuronal excitability in vivo and in vitro.

•

The hyperexcitability induced by GW is a GABAA receptor-dependent phenomenon.

•

GW is a heuristic model of abstinence syndromes induced by GABAergic drugs.

•

GW in vitro in the hippocampus decreases by previous exposure of GABA agonists.

•

GABA withdrawal syndrome in the brain cortex is blocked by the administration of HEPP.

Abstract

The sudden interruption of the increase of the concentration of the gamma-aminobutyric acid (GABA), determines an increase in neuronal activity. GABA withdrawal (GW) is a heuristic analogy, with withdrawal symptoms developed by other GABA receptor-agonists such as alcohol, benzodiazepines, and neurosteroids. GW comprises a model of neuronal excitability validated by electroencephalogram (EEG) in which high-frequency and high-amplitude spike–wave complexes appear. In brain slices, GW was identified by increased firing synchronization of pyramidal neurons and by changes in the active properties of the neuronal membrane. GW induces pre- and postsynaptic changes: a decrease in GABA synthesis/release, and the decrease in the expression and composition of GABAA receptors associated with increased calcium entry into the cell.

GW is an excellent bioassay for studying partial epilepsy, epilepsy refractory to drug treatment, and a model to reverse or prevent the generation of abstinences from different drugs.

Abbreviations

AW, alcohol withdrawal; aloP, allopregnanolone; BLA, basolateral nucleus of the amygdala; DHEAS, dehydroepiandrosterone sulfate; DWS, Diazepam withdrawal syndrome; EEG, electroencephalogram; EPSP, excitatory post synaptic potentials; GABA, gamma-aminobutyric acid; GHB, β-hydroxybutyric acid; GAD, glutamic acid decarboxylase; GW, GABA withdrawal; GWS, GABA withdrawal syndrome; IP, intraperitoneal; LTP, long-term potentiation; NAcc, nucleus accumbens; NMDA, N-methyl-d-aspartate; PDS, paroxysmal depolarization shift; PS, pregnenolone sulfate

Key words

GABA; neuronal hyperexcitability; withdrawal

GABA withdrawal syndrome (GWS), what and what for?

Identification of the relationship between gamma-aminobutyric acid (GABA) and epilepsy allowed the quantification of a cortical hyperexcitability phenomenon caused by chronic treatment of GABA and the subsequent interruption of its administration in the brain cortex of monkeys. In an attempt to identify strategies to manage seizures caused by a genetic model of generalized epilepsy in the Papio papio mandrill (this epilepsy type is inducible because intermittent light stimulation gives rise to epileptiform discharges in the brain cortex, as well as generalized myoclonus; Brailowsky et al., 1989, Brailowsky, 1991b and Brailowsky, 1991a), it was observed that direct intracortical instillation of GABA was able to stop the appearance of seizures; in other words, GABA instillation exerted anticonvulsant effects that lasted until the end of the instillation ( Brailowsky et al., 1988 and Brailowsky et al., 1989). However, the following day, after intracortical instillation of GABA, all of the monkeys exhibited paroxysmal electroencephalographic (EEG) activity at the sites where the instillation was performed.

Independent of the instillation site and the type of instillation cannula used, the EEG showed the presence of polyspike and spike–wave activity that, in the case of the motor cortex, was related with the appearance of myoclonus in the distal portion of the lower limbs; in other words, a different epileptiform activity appeared independent of the activity that the mandrill already demonstrated prior to treatment initiation (Brailowsky, 1991a, Brailowsky et al., 1990 and Brailowsky, 1991b). Years later, hyperexcitability induction in neurons caused by GABA withdrawal (GW) was identified in epileptic rats subjected to amygdala kindling model and in non-epileptic rats. In these small mammals, it was possible to describe the behavioral and electroencephalographic characteristics of the phenomenon (Fukuda et al., 1987, Brailowsky et al., 1988 and Brailowsky, 1991b). This encouraged the use of rats in the model, from which the majority of the data have been obtained and which has permitted studying the phenomenology of the synaptic changes (Fig. 1).

Electrophysiological recordings of the hyperexcitability induced by ...

Fig. 1.

Electrophysiological recordings of the hyperexcitability induced by gamma-aminobutyric acid (GABA) withdrawal syndrome (GWS) in two different models. GWS in vivo; upper panel, (A) recordings of the basal activity of the right and left somatomotor cortex of rats (upper and lower recordings, respectively). Wavelet analyses show the electroencephalographic (EEG) power of the right cortex. (B) Interruption of the intracortical instillation of GABA in the right cortex (5 mM/2 h) produces neuronal hyperexcitability; spike–wave activity is propagated to the contralateral hemisphere. An increase of power is identified in red colors in the Wavelet analysis. © The duration of this activity is about 7–8 days. At day 9 asymmetries in the EGG activity continue to be present; however, values are very similar to basal activity. In vitro; lower panel. Interruption of superfusion of GABA (5 mM/2 h) in brain slices of the hippocampus induces hyperexcitability (black circles). Temporal evolution and representative traces of the fEPSP are depicted in three different stages: before; during, and after GABA application (black bar). The increase of the amplitude of fEPSP, obtained from area CA1 of the hippocampus, is statistically significant at 30 min of GABA interruption. The effect is not reversible in the following 3 h. In contrast, the same stimulation frequency does not induce changes in the synaptic activity (control group; white circles; each symbol in the graphics represents the average ± standard error of the mean, SEM, of each group). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Figure options

Anatomic and enzymatic changes of GABAergic neurotransmission on GW

Radioactive staining studies have permitted the identification of metabolic changes at the GABA instillation site and in subcortical structures connected through distance to the site where GW caused the neuronal hyperexcitability. Through the use of the 2-deoxyglucose (2-DG) radioactive capture/metabolism technique, it was possible to quantify a significant increase of local glucose consumption (from 3 to 5 times higher compared with the control) in the cortical area involved in the appearance of paroxysmal activity and in the ipsilateral thalamic zone of that cortical area (posterior, ventralis-lateral-posterior, ventralis-intermedius, ventralis-lateral, and reticularis oralis nuclei). These brain regions with oxidative metabolic increase correspond to reactive gliosis areas identified in the brains of animals obtained 10 days after the interruption of paroxysmal activity (Menini et al., 1991). These aspects are similar in temporal lobe epilepsy and in abstinence induced by Flunitrazepam or alcohol (Morrow et al., 2001, Olmedo and Hoffman, 2000 and O’Brien, 2005). Subsequent investigations allowed identification of the effects that chronic intracortical instillation of the amino acid exerted on the activity of the enzyme that synthesizes GABA, that is, glutamic acid decarboxylase (GAD). The results demonstrated a 40% reduction of GAD activity at the site where the amino acid was instilled (Salazar et al., 1994). These data are similar to those of the effects produced by acute ingestion of alcohol and by Clonazepam abstinences (Olmedo and Hoffman, 2000 and Kumar et al., 2009).

Pharmacological changes in GWS

Experiments focused on determining the role of the two main subtypes of receptors that GABA, GABAA, and GABAB possess indicated that GW is a phenomenon dependent on inotropic receptors. Pharmacologic tests corroborate this affirmation: first, demonstration of the induction of a GWS employing cortical instillations of Isoguvacine, a specific GABAA agonist (Brailowsky et al., 1990 and Brailowsky, 1991b), and second, the corroboration that the specific agonists of the GABAB receptor (Baclofen) induce an increase of neuronal excitability. Additionally, GABAB antagonists (Faclofen) have no effect on GW.

Surprising results have been obtained from pharmacologic studies of GW that utilized anticonvulsant, sedative and hypnotic agents. It is necessary to emphasize that the pharmacological responsiveness of the neuronal hyperexcitability induced by GW changes according to the temporal evolution of the paroxysmal activity. The GWS, during its first 24 h, is resistant to clinical anticonvulsants (Phenytoin, barbiturates, Ethosuccimid, valproic acid, Carbamazepine), and even to the most popular drug choice in cases of epileptic status: Diazepam. Intraperitoneal (IP) doses of up to 15 mg/kg of this benzodiazepine do not affect the discharge frequency of paroxysmal activity, even if the animal is deeply sedated. Even the use of Pentobarbital at anesthetic doses (35 mg/kg) does not significantly modify the frequency of cortical discharges.

In addition to such a dependence on protein synthesis with long-term potentiation (LTP) in its early and maintenance phases in the hippocampus (Frey et al., 1988, Barea-Rodríguez et al., 2000 and Calixto et al., 2003), GWS also depends on the formation of new proteins: anisomycin and cycloheximide, inhibitory agents of protein synthesis that also block the appearance of hyperexcitability.

From day 2 of GWS, changes in pharmacological responsiveness can be observed: antagonists of the N-methyl-d-aspartate (NMDA) receptor, ketamine, amino-phosphate–heptanoate (APH), and MK-801, or benzodiazepines such as Clonazepam exert an anticonvulsive effect on GW. This indicates that other neurotransmitters are involved in neuronal hyperexcitability. Within this context, it is noteworthy that the focal paroxysmal activity produced by GW can be inhibited if local instillation of GABA is re-started, even if this is carried out during the first stages of the hyperexcitability process (Brailowsky et al., 1988).

Related studies identified that taurine, another amino acid with inhibitory effects on cortical activity, is not capable of inducing the abstinence phenomenon. These same negative results were obtained with chronic instillations of glycine at equivalent GABA times and concentrations. Glycine and taurine, both amino acids, exert no inhibitory effects on GW.

This cortical hyperexcitability model caused by GW is a sui generis model that aids in the study of abstinences in the clinic. The use of this model can have an impact in two different ways: (1) on the study of the etiology, evolution, and consequences of abstinences of drugs that increase GABAergic activity, and (2) on the development of strategies that decrease the effects and consequences of pharmacologic abstinences produced by these same drugs.

GW in vitro

The first recordings were obtained from brains of animals that had developed GWS, and the results of these experiments showed changes in the excitability of the neuronal membrane (Silva-Barrat et al., 1989 and Silva-Barrat and Champagnat, 1995). The second type of experiment was designed to characterize neuronal hyperexcitability in brain slices of the brain cortex and the hippocampus without previous manipulation; in other words, it is possible to induce neuronal hyperexcitability by GW in the brain tissue; through these experiments, it was possible to identify changes in synaptic activity and in the GABAA receptor (García-Ugalde et al., 1992, Calixto et al., 2000 and Casasola et al., 2001).

Experiments on the brain cortex obtained from animals that presented GWS with common clinical and electrical characteristics (unilateral myoclonus, plentiful salivation, rhythmic movement of whiskers, immobility, and spikes related with the instillation site shown on the EEG) allowed the identification of biophysical and synaptic changes at the GABA instillation site. Electrical stimulation of the white matter located in the same plane of the recording site induced, in all of the cells analyzed, paroxysmal depolarization with high-frequency spike trains. This activity is denominated paroxysmal depolarization shift (PDS) (Wirsen et al., 1992). In these brain slices, a neuronal population was identified that presented, in addition to synaptically induced PDS, high-frequency spike trains induced by injection of intracellular current. These neurons, with the intrinsic ability to generate PDS, are different from others (i.e., neurons that only could generate PDS by synaptic stimulation) due to the following: (a) they have voltage-dependent potentials (excitatory post synaptic potentials, EPSPs); (b) they present calcium-dependent PDS: The substitution of calcium for cobalt produced the disappearance of current-induced, high-frequency spike trains, this causing the appearance of single potentials, and © they exhibit higher tolerance to the hyperpolarizing effects of GABA when applied to the incubation bath. This pharmacological tolerance was quantified as an increase ranging from 50 to 100 times on the EC50 (half maximal Effective Concentration) of Isoguvacine (specific agonist of GABAA receptors) compared with neurons presenting synaptic PDS (Silva-Barrat et al., 1989 and Montiel et al., 2000).

Intracellular recordings of pyramidal neurons (localized at the GABA installation site, layer III/IV of the brain cortex) obtained from brain slices of animals that presented GWS, allowed the identification of changes in neuronal membrane excitability and in the cell’s discharge properties. An important modification was observed at the threshold of the neuronal firing rate, which was associated with the reduction of the action potentials’ firing adaptability; in addition, a significant reduction of the hyperpolarizing post-potential was also quantified. These results are similar to data obtained under different conditions such as the following: neurons of epileptic foci (Behr et al., 2000a, Behr et al., 2000b, Martin et al., 2001a and Martin et al., 2001b); another withdrawal type (Brown et al., 2005, Krystal et al., 2006, Liang et al., 2007 and Licata and Rowlett, 2008), and under conditions where the brain cortex presents metabolic changes (Kelly and Church, 2004) with effects mediated by benzodiazepines (Capogna et al., 1994, Valverde et al., 1995 and Valverde et al., 1992). Changes in excitable cell properties have been previously identified in withdrawal induced by some GABAergic drugs. Abrupt interruption of chronic ethanol administration induces the manifestation of spontaneous activity, generates increased synaptic activity identified by growth in EPSP amplitude, and the occurrence of synaptic facilitation quantified by the paired-pulse protocol in cells proximal to the brainstem’s cerebral aqueduct (Yang et al., 2002), similar to that generated by GW. The effect of neurosteroids and Clonazepam abstinence on the excitable properties of pyramidal neurons has yet to be explored. However, Diazepam abstinence, which is widely documented, indicates that abstinence generates neuronal hyperexcitability and induces anxiety.

In hippocampal and brain cortex slices incubated with GABA for 120 min, it was possible to confirm the appearance of neuronal hyperexcitability, characterized by the extraordinary increase of pyramidal-neuron field response from hippocampal area CA1. This occurred 1 h after the removal of the amino acid from the perfusion (Fig. 1). Additionally, it was also possible to observe the abolition of recurrent inhibition produced by GABA, this evaluated by means of the paired-pulse stimulation paradigm (García-Ugalde et al., 1992). Subsequently, it was possible to characterize GABA-related neuronal hyperexcitability in brain cortex slices, specifically in the somatomotor cortex (Calixto et al., 2000). In this brain region, GW produces neuronal discharges that grow progressively in amplitude; this is similar to paroxysmal activity quantified in epilepsy models and in benzodiazepine abstinences (flunitrazepam and clonazepam).

Hyperexcitability produced by GW can be also induced in cultured cells. Studies in isolated cells demonstrated that the brief exposure of GABA in dissociated neurons from hippocampal area CA1 (that were cultured for 2 weeks and that presented regular intervals in synaptic current generation) gives rise to total suppression of the spontaneous activity of the network. When GABA is washed off, activity increases and currents grow stronger and increase their potency compared with the control. These conditions lasted from 1 to 2 days after the GABA wash-off (Golan et al., 2000).

Through these experiments, it was possible to identify two important aspects of neuronal activity during GW in vitro: (1) the decrease of the release of radioactive GABA produced by the tissue ( Calixto et al., 2000), and (2) a decrease in the number of active GABAA receptors (Casasola et al., 2001).

LTP represents an electrophysiological substrate of memory processes. LTP comprises an increase of synaptic efficiency whose principal characteristic is the increase of synaptic activity after high-frequency stimulation in the recording site’s afferent pathway. It was logical to question whether the hyperexcitability produced by GW in vitro could facilitate the induction or appearance of another phenomenon of neuronal hyperexcitability. In conjunction with this, a set of experiments was executed in which GW facilitates the LTP induction, expression, and maintenance in the brain cortex and hippocampus (central nervous system [CNS] structures that are directly involved in learning and memory processes in mammals) ( Montiel et al., 2000 and Casasola et al., 2004).

The study of the hyperexcitability produced by GW can open new opportunities for developing a new generation of anticonvulsant substances and growing the number of drugs that offer neuronal protection and sedative effects with high impact on the treatment of abstinences produced by commonly used drugs such as benzodiazepines and alcohol (Calixto et al., 1995, Strzelec and Czarnecka, 2001 and O’Brien, 2005).

Different withdrawals and their relation with changes in the GABAA receptor

Postsynaptic changes in the GABAergic system comprise a common mechanism of abstinences that are induced by alcohol, benzodiazepines, neurosteroids, and barbiturates (Brailowsky, 1991a, Brailowsky, 1991b, Calixto et al., 2000, Biggio et al., 2003, Nelson et al., 2005, O’Brien, 2005, Smith et al., 2007, Wulff et al., 2007 and Brust, 2014). In the fields of Psychiatry and Neuroscience, it is known that changes in the GABAA receptor are a pharmacologic consequence of the abstinence induced by its ligand. Two processes of changes in the receptors have been identified at the beginning of these abstinences: (1) a decrease in the density of GABAA receptors (down regulation), associated with (2) a gradual manifestation of new receptors that pharmacologically exhibit a change in sensitivity to its agonists. It is widely documented that new GABAA receptors are expressed under abstinence conditions. These new receptors are less functional or demonstrate lower sensitivity to their agonists. This event is produced by the increase or decrease of the expression of GABAA receptor subunit types (Olmedo and Hoffman, 2000, O’Brien, 2005 and Smith and Gong, 2005); for example, Diazepam decreases expression of subunits α1, β2, and γ2 and increases the expression of subunit α5; this in turn reduces the pharmacologic sensitivity of the receptor (Follesa et al., 2001, Izzo et al., 2001, Cagetti et al., 2002 and Follesa et al., 2004). In addition, something similar occurs with neurosteroids such as progesterone and its reduced metabolite, allopregnanolone (aloP); the abstinence produced by these neurosteroids facilitates the expression of subunits α4, β1 and δ (Griffiths and Lovick, 2005) and, at the same time, decreases the expression of subunit α1 (Gulinello et al., 2001, Griffiths and Lovick, 2005, Smith and Gong, 2005 and Smith et al., 2007). Chronic administration of Pentobarbital increases the expression of subunit β2 (Oh and Ho, 1999). Ethanol withdrawal (EW) generates decreased expression of subunits α1, α4, α5, α6, δ, γ2L, and γ2S, and increases the expression of subunits α2, α3, β1, β2, and β3 (Mhatre and Ticku, 1994, Smith and Gong, 2005, Follesa and Ticku, 1996, Follesa et al., 2001 and Crews et al., 2005; Follesa et al., 2005). Other studies report significant increases of subunits α4 and α2 under alcohol withdrawal (AW) conditions (Cagetti et al., 2003). These changes to modifications in subunit expression are identified hours after interrupting administration of the drugs (latency to onset of abstinence ranges between 4 and 24 h).

Partial studies of the changes in the expression of receptor subunits during GW demonstrate that pyramidal neurons of the brain cortex exhibit a decrease in the expression of GABAA receptor subunit α4. Similar effects are produced by progesterone, aloP, and alcohol (Hui et al., 2010). The findings of our research indicate that, during GWS, there is a significant reduction of GABAA-receptor total expression that is present in the cerebral cortex and hippocampus in early withdrawal stages (Calixto et al., 2000, Casasola et al., 2001 and McBain et al., 2015). With this evidence, it is logical to ask which changes arise in the GABAA receptor during GABA-induced hyperexcitability. Studies have partially explored modifications in the expression of the subunits comprising receptor subunits. Preliminary data from our laboratory indicate that, in pyramidal neurons of the cerebral cortex, the cerebral cortex exhibits decreased expression of the α4 subunit, in the same way that progesterone, aloP, and ethanol do) and of the α2 subunit (contrary to inducing EW).

Benzodiazepines

Benzodiazepines are drugs that increase GABAA receptor function, exerting positive allosteric modulation. The junction between benzodiazepines and the GABAA receptor results in an increment of Cl− conductance. This becomes a hyperpolarization of neurons, allowing a decrease in synaptic transmission. GABAA receptors containing subunits α1, α2, α3, or α5, in combination with subunits β2 and γ2, are receptors with a higher affinity for GABA and benzodiazepines (Leppa et al., 2011). The GABAA receptor subunits comprising subunits α4 and α6 respond weakly to Diazepam, Flunitrazepam, Clonazepam, or Zolpidem (Pirker et al., 2000 and Möhler, 2006). The anticonvulsant, sedative, anxiolytic, or hypnotic effects of benzodiazepines are mediated by GABAA receptors that possess a conformation that is mainly based on subunits α1 and α2 (Möhler, 2006 and Möhler, 2009).

The decreased brain activity phenomenon is dependent on the benzodiazepine dosage. In vivo studies indicate that benzodiazepines diminish activity recorded during an EEG in different brain structures (hippocampus, cortex, among others) in various animal models, including cats, mice, monkeys, rabbits, rats, and dogs ( Depoortere et al., 1983).

The most common signs and symptoms of benzodiazepine withdrawal (BW) include anxiety and agitation. Several studies indicate that prolonged treatment with benzodiazepines significantly decreases the expression of isoforms α1 and γ2 of the GABAA receptor. When treatments are prolonged (for >1 month), changes occur in subunit α5, reducing its expression, while there is increased expression of isoforms α3 and α6 (Doble and Martin, 1996 and Tan et al., 2011).

Diazepam is the benzodiazepine that is most frequently used in the clinical setting (Manchikanti, 2007). It is employed in treatments for the pharmacological management of anxiety, psychosis, and insomnia; also, it is utilized as a tranquilizer, an anesthetic, and a hypnotic, and produces marked muscle relaxation (Baldwin et al., 2013).

After application of Diazepam, the EEG shows an increase in beta activity and decreased alpha activity or slow waves. These changes in beta activity have been correlated with the drug’s anticonvulsant effect (Mandema and Danhof, 1992). Administration of Diazepam increases the levels of steroid hormones, such as testosterone and corticosterone; also, in bovine adrenal glomerulosa cells, it inhibits the production of aldosterone (Papadopoulos, 1993). In the brain, Diazepam increases the activity of the cytochrome p450 multi-enzyme complex, which favors the biosynthesis of aloP, progesterone, and dehydroepiandrosterone, the three so-called neurosteroids (Swinnen et al., 1998). Another benzodiazepine, Lorazepam, generates important pharmacological tolerance: subunit α1 changes in the expression of the GABAA receptor (Fahey et al., 2006). One-week administration of Flurazepam and abrupt interruption induces anxiety associated with increased expression of AMPA-type glutamate receptors and an increase in the activity of Ca2+-channel-dependent voltage in hippocampal neurons, reflecting neuronal hyperexcitability (Shen and Tietz, 2011, Earl et al., 2012 and Cloos and Bocquet, 2013).

Diazepam withdrawal syndrome (DWS)

Between 1971 and 1978, several cases were reported in the U.S. that were related with prolonged treatment with Diazepam and its abstinence: anxiety; insomnia; profuse sweating; dysphoria, agitation; emotional instability; tremor, headache; dizziness; lack of coordination; sensory hypersensitivity; fatigue; lethargy; blurred vision; facial burning sensation; muscle pain; tachycardia; hallucinations; psychosis; delirium and, sometimes, convulsions (Greenblatt and Shader, 1981, Doble and Martin, 1996 and Tan et al., 2011; Arnot et al., 2001 and Kovačević et al., 2014).

In rats, interruption of prolonged treatment with Diazepam creates the appearance of irritability, fearful behavior, and seizures. Spike–wave activity, similar to what happens in the GWS, appears on the EEG (Hu, 2011), in addition to molecular changes in the GABAA receptor, such as the decrease in the density and expression of subunits aside from α1 and α2 and increased ARNm expression of subunits α4, α5, γ2L, γ2S, β2, and β3 in the neurons of the cerebral cortex and hippocampus (Impagnatiello et al., 1996, Pesold et al., 1997 and Ramerstorfer et al., 2011). These changes modify receptor function, causing the reduction of the Diazepam effect (pharmacodynamic tolerance; Follesa et al., 2001, Follesa et al., 2004 and Ator et al., 2010). Recently, it was shown that the glutamatergic system is also modified. During Diazepam withdrawal, there are changes in glutamate receptor conformation; these changes are associated with an increase in the number, as well as in the function, of AMPA receptors and a reduction (down regulation) of NMDA receptors. Increased AMPA receptor function is correlated with the overexpression of Glu1 (essential subunit for neuronal plasticity; Izzo et al., 2001 and Allison et al., 2005). These same changes have been found in glutamatergic neurotransmission during Flurazepam withdrawal (a benzodiazepine 10 times less potent than Diazepam (Van Sickle et al., 2004, Song et al., 2007 and Xiang and Tietz, 2007). The presynaptic component also changes during benzodiazepine withdrawal; recent findings indicate a decrease in the expression of the receptors of group II mGluR, which can contribute to reducing the desensitization mechanisms of neuronal excitation, the latter one of the mechanisms of initiation of benzodiazepine dependence (Okamoto et al., 2013). Diazepam abstinence is associated with a change in the activity of the enzyme protein kinase M zeta (PKMζ), generating an increase in neuronal excitability (Monti et al., 2012).

However, it has been quantified that, depending on the Diazepam application, changes in withdrawal are paradoxically different because the drug’s administration IP gives rise to a reduced binding rate in the AMPA receptor. The opposite occurs with subcutaneous (SC) administration: in this model, the rate is increased of the AMPA receptor binding the same receptor; this effect may be a neuro-adaptive response to an increase in excitatory activity (Steppuhn and Turski, 1993 and Lader, 2011).

The evidence is that it can also increase the activity of other second messengers in benzodiazepine withdrawal identified as Diazepam during withdrawal activity, that is, the arachidonic acid cascade, which is associated with increased susceptibility to the development of increased PTZ-induced seizures (Mori et al., 2012).

It is necessary to note that, similar to what happened with the GWS in the DWS, evolutionarily, changes present first in GABAergic neurotransmission, and then modifications are gradually installed in other neurotransmitters (Caputo and Bernardi, 2010 and Uzun et al., 2010). Experimental evidence in animals indicates that the evolution of Diazepam abstinence is as follows: (1) decrease of GABAergic activity; (2) increased glutamatergic activity; (3) increased serotonergic activity; (4) the appearance of decreased cholinergic neurotransmission, and (5) up regulation of L-type calcium channels in cortical neurons (Katsura et al., 2007).

Abstinence to Diazepam, as well as to GABA, shares more electrophysiological features than differences. However, an exogenous drug induces the DWS, and GWS is a consequence of the interruption of instilling an endogenous neurotransmitter.

Neurosteroids

Neurosteroids are steroids synthesized de novo in the brain by glia (oligodendrocyte and microglia) and neurons. Their concentration is independent on systemic steroids, which are synthesized in gonads and placenta (sex steroids) and suprarenal glands (mineral- and glucocorticoids). Neurosteroid activity in the brain is a consequence of these molecules recognizing specific genomic intracellular receptors that modify neuronal activity ( Murashima and Yoshi, 2010). However, neurosteroids can allosterically modulate the membrane receptors of different GABA-, glutamat-, cholin-, and dopamine-ergic neurotransmissions. In this context, advances in the field of Neuroscience allow the identification of the neurosteroids of 21 carbons that derive directly from progesterone and that are primarily molecules that increase GABAergic tone, similar to what benzodiazepines do. This event is rapid, reversible, and does not involve genomic neuron activity; thus, it is denominated neurosteroid modulating activity or non-genomic neurosteroid activity. It is important to emphasize the following: (1) neurosteroids are synthesized in the brain and have a positive modulatory function in GABAergic neurotransmission, and (2) there is a relationship between the structure and function of these molecules: steroids derived from progesterone are mainly positive allosteric modulators of the GABAA receptor (Veleiro and Burton, 2009).

The concentration of pregnenolone in the brain of mammals is higher in the postnatal stage (40 ng/g) and decreases later. Its concentration increases under stress situations, gestation, and during the postpartum period (Wieland et al., 1981), which contributes to relaxation, sedation, and analgesia in these states (Wilson and Biscardi, 1992, Grobin et al., 1992 and Kabbadj et al., 1993). The mammalian brain also produces neurosteroidal agonists of the GABAA receptor; in addition, in this category are found sulfated derivatives, including pregnenolone sulfate (PS) and dehydroepiandrosterone sulfate (DHEAS). These neurosteroids have been employed as pro-convulsive agents in diverse experimental models (Kokate et al., 1999). According to the heterogeneity of GABAA receptor subunits α1, α3, and α4, these appear to provide greatest sensitivity of the GABAA receptor to the neurosteroids, while the presence of subunits β1 and γ2 is necessary. However, the possible role of other subunits has not yet been identified (Maguire and Mody, 2009).

Endogenous and synthetic neurosteroids with greater activity on the GABAA receptor are those having ring “A” reduced in its C #5 and hydroxylated in its C #3, both in position α, and an electronegative atom (usually oxygen) in C #17 or C #20 (Finn and Gee, 1993). aloP and tetrahydrodeoxycorticosterone possess such capacities. aloP comprises the neurosteroid with the greatest biological activity and pharmacological potency on the GABAA receptor. Its effects are anxiolytic, hypnotic, antiepileptic, and anesthetic (Majewska, 1992 and Lan and Gee, 1994).

Characterized by their affinity, at least two different GABAA receptor sites have been identified as those interacting with neurosteroids. The high-affinity site that has been postulated for PS refers to a hydrophobic portion inside the receptor because it is insensitive to protein digestion. However, for DHEAS, the effects of proteolytic enzymes significantly alter receptor binding. This suggests different receptor-binding sites for these molecules and the participation of structures outside of the receptor (Majewska, 1992, Reddy, 2010 and Turkmen et al., 2011). Interaction at the GABAA receptor level is different between agonists and antagonist neurosteroids (Majewska, 1992 and Reddy, 2010). Both molecules exhibit differences in their polarity: agonists are highly lipo-soluble, whereas antagonists are highly water-soluble (sulfate group); this property gives rise to the different behavior that both structures possess in the plasma membrane. It is possible that electronegative atom C #20 of the neurosteroid agonists is the main polar difference between both neurosteroid groups and, in turn, the fundamental factor binding the protein. 3α Hydroxylation may also be responsible for modification of the receptor. 3α hydroxylation, as well as that of the ceto group of C #7 or C #2, interacts with GABAA protein subunits. Neurosteroids can interact with G proteins coupled with effector systems in the neuronal membrane. One year later, it was identified that some neurosteroids were able to modulate calcium currents through mechanisms of transduction coupled with pertussis toxin-sensitive G proteins (Ffrench-Mullen et al., 1994). There is a relationship between brain excitability and cyclical changes of sex hormones (estrogen and progestin). Clinical experimental evidence support estrogen and progestin as modulators in the frequency of the occurrence of some epilepsy types, in contrast to poor effects in androgens and other steroids (Reddy, 2010). The effects of sex steroids on some epilepsy types of have been studied in animal models; for example, epileptic seizures induced by drugs or electroshock increase with estrogen administration, while intravenous (IV) injections of progesterone reduce these crises, as well as those induced by Penicillin in ovariectomized cats (Veleiro and Burton, 2009 and Reddy, 2010).

Neurosteroid withdrawal syndrome (NWS)

In the clinical ambit, the relationship between sex hormones and neuronal hyperexcitability is evident: generalized seizures in women are more frequent during the menstrual or the premenstrual phase, during which there is a decrease in progesterone concentration and an increased estrogen concentration is initiated, whereas after the luteal phase, there is a significant decrease of the presence of this crisis type (Veleiro and Burton, 2009). Another factor suggesting a close relationship between neuronal hyperexcitability and neurosteroids is the change in the frequency of the presence of generalized and non-generalized epilepsy in pregnancy. Some works (Andréen et al., 2009, Reddy, 2010 and Reddy, 2013) mention an exacerbation of the generalized crisis in the first gestational trimester (Ferando and Mody, 2012). In the immediate postpartum, progesterone increases nearly 10-fold; during this stage, the occurrence of epileptic seizures exhibits a significant decrease. Epilepsy related with steroid levels is denominated “catamenial epilepsy”. It has been reported that epileptic crises are more frequent in women with ovulatory cycles compared with women who do not ovulate. Medroxyprogesterone, a progesterone analog, is employed as an anovulatory and is characterized as a potent antiepileptic. The relationship between neuronal hyperexcitability and neurosteroids is based on experimental and clinical evidence as follows: (a) the menstrual period defines cyclical changes in both progesterone and estrogen levels; the second part of the cycle depends on progesterone; in this case, susceptibility to an epileptic crisis is lower than in any part of the cycle (Andréen et al., 2009 and Reddy, 2010); therefore, (b) the effects of estrogen and progesterone involve favoring and decreasing interictal activity, respectively, or a significant decrease in the presentation of this crisis type and even of the resistance effect of benzodiazepines ( Veleiro and Burton, 2009 and Gangisetty and Reddy, 2010); © these changes in the concentration of sex steroid hormones are evident in certain biological stages, such as puberty and pregnancy, during which the frequency of epileptic episodes changes; (d) administration of progesterone and estrogen analogs and antagonists permit the facilitation or prevention of their effects on cerebral excitability ( Calixto et al., 1995, Veleiro and Burton, 2009 and Reddy, 2010).

Neurosteroids increase GABAA receptor activity, similar to benzodiazepines or barbiturics, but performed at different receptor sites (Gangisetty and Reddy, 2010 and Carver and Reddy, 2013). Pregnans neurosteroids increase GABA binding, potentiate receptor function (increase in Cl− current), and further enhance neuronal hyperpolarization. They are positive allosteric modulators, and their synthesis is increased by stress during functional recovery processes involving brain injuries. Over the last few years, it has been reported that these molecules, such as progesterone and aloP, cause a decrease in neuronal activity, but sharply falling plasma concentrations in women can induce dysphoria, anxiety, and irritability, this providing knowledge of a premenstrual syndrome. However, during pregnancy, which averages 9 months, there is an increase in pregnans neurosteroids, but at the time of the labor, there is an abrupt decrease in neurosteroid concentration, a factor that can induce postpartum depression. This indicates that the process toward hormonal changes throughout the menstrual cycle permits the modification of neurosteroids, which may also be the cause of withdrawal syndromes (Reddy, 2014).

Recent data indicate that high aloP levels modify the expression of GABAA receptor subunits, thereby changing the conditions of spatial learning and plasticity of hippocampal area CA1 according to the estrous cycle of the female rodents (Lovick et al., 2005 and Sabaliauskas et al., 2014), with α4 subunit expression favoring this plasticity. Expression of this subunit in the receptor, favoring more rapid desensitization and decreased hyperpolarization, can exert an effect on the GABAA receptor (Gong and Smith, 2014 and Barth et al., 2014). This can be important in the onset of behavioral change-dependent hormones during the adolescence factor (Smith, 2013). In addition to factors such as the stress alternative, it also increases the expression of the receptor’s α4 subunit (Shen et al., 2013).

This point resembles other withdrawal syndromes as follows: changes in receptor expression by increasing neurosteroids and a sharp drop in their concentrations that induce changes in the expression of subunits α1 and α2, their decrease is in turn associated with the increased expression of α1 subunit, which decreases GABAA receptor activity.

A common point between the withdrawal of benzodiazepine and of ethanol is the recipient of interleukin 1, which in their presence increases alcohol consumption, and their lack reflects a decrease in the consumption of alcohol and benzodiazepines in mice (Blednov et al., 2015).

Other neurotransmitters involved in GW

The increase of neuronal excitability, in vivo and in vitro, produced by GABAergic abstinences is associated with other neurotransmitters; for example, Diazepam and Pentobarbital abstinences produce a significant increase in the expression of NMDA receptors ( Jang et al., 1998). AW exerts the same effect on the NMDA receptors of hippocampal area CA1, this is a late characteristic of AW (Nelson et al., 2005). When GABA-produced hyperexcitability is present, neurons of the somatomotor cortex that exhibit an increase in their activity also demonstrate a decrease in the activity of the muscarinic receptors (Silva-Barrat et al., 2001), an increase in the activation of narcotic receptors (Silva-Barrat et al., 2005), and a decrease in the outward current of K+, (Silva-Barrat and Champagnat, 1995), as well as an increase in the expression of the acetyl-cholinesterase enzyme in epileptic discharge foci (Araneda et al., 1994). Withdrawal induced by the interruption of the administration of drugs such as benzodiazepines produces an increase in the expression of opioid receptors and, in addition, pre-treatment with benzodiazepines decreases the symptoms of morphine abstinence (MA) ( Valverde et al., 1992 and Suzuki et al., 1996). These synaptic changes suggest a functional and neurochemical arrangement in the brain cortex and hippocampus as a consequence of the hyperexcitability produced by the abstinences. The implications of the GABAergic changes that impact other neurotransmissions comprise an unexplored field. It is possible that in late GWS stages, glutamat-, cholin-, and opio-ergic neurotransmissions increase their activity.

Recently, the GABAA receptor has been explored under conditions of BW; this can generate plasticity in the ventral tegmental area (VTA), increasing glutamatergic activity that, in turn, modifies dopamine release in this area (Vashchinkina et al., 2012). The nucleus accumbens (NAcc) is also a structure that favors mesolimbic addictive behavior, permitting greater dopamine release in this brain region. Abstinence by a rat model with chronic intermittent ethanol (CIE) treatment is also modulated by GABAA receptors (Liang et al., 2014a). Although the effect of alcohol is different in different brain regions, changes in GABAA receptor activity in the amygdala are responsible for anxiety processes. In particular, the basolateral nucleus of the amygdala (BLA) decreases expression of the α4 and δ subunits of ethanol during treatment; this would comprise the main contribution of drug withdrawal behavior initiated in BLA withdrawal, while the entrance of a new intake of ethanol, increasing α4 and γ2 subunit expression, decreases the expression of α1, α2. and δ, as well as the induction of GABAA receptor plasticity in the NAcc, which decreases the expression of the δ subunit, the increase of the α1 and α4 subunits, as well as those of γ2 and α5 (Liang et al., 2014b). These changes in the composition of the modified biophysical receptor channel generate changes in brain structure activity (Korpi et al., 2007, Lindemeyer et al., 2014 and Capogna, 2014). Additionally, chronic ethanol treatment modifies phosphorylation from tyrosine GABAA receptor kinase, which increases the desensitization process and dependence on alcohol intake (Ravindran and Ticku, 2006 and Marutha Ravindran et al., 2007). In culture, a change in ethanol expression induces cortical neurons as follows: it induces increased expression of subunits α4, decreased the expression of subunits α1 and α2, and increases expression of the NR2B NMDA receptor subunit (Sheela Rani and Ticku, 2006).

Is GWS refractory to treatment?

One of the targets of clinical therapeutics is to decrease the expression of the symptoms produced by abstinence (Heberlein et al., 2009). From the experimental view, this process can entail a new approach: our results suggest that it is possible to block or reduce the neuronal hyperexcitability produced by GW with the sequential administration of at least two pro-GABAergic drugs. This finding is novel in the Neurosciences field. For example, in some abstinences produced by drugs that are agonists of the GABAA receptor, such as the abstinence induced by alcohol, an up regulation process of the receptor’s subunit α2 occurs; however, the combined administration of alcohol and β-hydroxybutyric acid (GHB) reverts the increase in this subunit’s expression ( Follesa et al., 2001 and Follesa et al., 2004). Also, a 48-h exposure of aloP to cultured hippocampal neurons facilitated the increase of subunit α4 expression, this process blocked by short administration of alcohol at low concentrations (these concentrations in themselves do not induce abstinence) (Smith and Gong, 2005). On the other hand, it was identified that administration of Diazepam in early stages of alcohol intake reduces changes in GABAA receptor subunits; these results are not obtainable with the administration of Baclofen (Sanna et al., 2003). Also, Diazepam pretreatment blocks the increase in subunit α4 expression produced by alcohol abstinence (Sanna et al., 2003). Prior administration and co-administration of progesterone, aloP, and PS with Diazepam or Triazolam reduce the tolerance, anxiety, and hyperactivity produced by benzodiazepines ( Reddy and Kulkarni, 1997, Olmedo and Hoffman, 2000 and Liang et al., 2007).

Global analyses of these findings suggest the possibility of decreasing the abstinence produced by GABAergic drugs through the administration of another drug that modulates the same receptor. Twenty years ago, it was thought that GWS was refractory to treatment. Recent discoveries indicate that GW can be diminished with the administration of structures derived from the phenyl alkyl amines, such as HEPP, which possesses a powerful effect, in that this drug decreases the appearance of the latency of spike–wave activity and the subsequent hyperexcitability reported in the EGG (Fig. 2B). In another set of experiments, we found that previous administration of progesterone (5 days/45 mg/kg) modifies the expression of GW in vivo; it delays latency and reduces the power of the EEG ( Fig. 2C). With these results, we can conclude that the electrophysiological expression of GWS in vivo can be diminished. In anticipation of complementing our research, we formulated the following question: “Could the expression of GW also be reduced in vitro?” Our line of work is essential for demonstrating what we stated previously. Data obtained from electrophysiological in vitro tests on hippocampal slices demonstrate that GW can be reduced by 47% by alcohol if this agonist of the GABAA receptor is previously exposed to GABA; in the case of Diazepam, GW is reduced by 67%, and aloP totally blocked the neuronal hyperexcitability produced by GW in vitro ( Fig. 2D). In other words, GW can be decreased if an agonist of the GABAA receptor modifies the protein structure of the receptor before GABA recognizes its receptor. These results open new possibilities in the treatment of abstinences that involve GABAergic neurotransmission.

In vivo, HEPP and chronic progesterone modify gamma-aminobutyric acid (GABA) ...

Fig. 2.

In vivo, HEPP and chronic progesterone modify gamma-aminobutyric acid (GABA) withdrawal syndrome (GWS). (A) Classic initiation of neuronal hyperexcitability produced by GWS is quantified in Wavelet analyses (upper panels/red arrow); the respective encephalographic (EEG) recordings are also shown (lower panels). (B) HEPP application (50 mg/kg) reduces the maximal stage of hyperexcitability induced by GWS (green arrow/acute effect). © Previous short treatment of Progesterone (dose of 45 mg/kg did not induce any withdrawal syndrome). (D) In vitro, aloP blocks the induction of GW. AloP exposure did not induce hyperexcitability in brain slices (10 μM/1 h; black rhombus). However, the same exposure of aloP before GABA application blocks the hyperexcitability induced in area CA1 of the hippocampus. Figures illustrate temporal evolution and representative traces of the experiments, and each symbol in the graphics represents the average ± standard error of the mean (SEM) of each group. Horizontal bars represent time superfusion. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Figure options

Decrease of withdrawal syndromes in GABAergic drugs

Reducing the signs and symptoms of withdrawal syndromes to the fullest expression is the expected event in therapeutic hospitalizations from the experimental viewpoint; this process can acquire a new approach that aims to achieve the blockage or decrease of neuronal hyperexcitability at the expense of joint or sequential administration of at least two GABAergic drugs (Mhatre et al., 2001). This aspect is recent in the fields of Neuroscience and Clinical Neurology; for example, during ethanol withdrawal, which induced a subunit α2 upregulation process, joint administration of ethanol and GHB reversed the increased expression of this subunit (Follesa et al., 2004). Also, the 48-h exposure of the neurosteroid aloP to cultured hippocampal neurons favors increased α4 subunit expression; this process is blocked by short administration of ethanol at low concentrations that, in themselves, do not induce withdrawal. Data of the same group demonstrated that ethanol could generate reduced subunit α4 expression (Smith and Gong, 2005). Moreover, it was identified that Diazepam administration in early alcohol-intake stages reduces the changes in GABAA receptor subunits induced by ethanol withdrawal, an aspect that cannot be compensated for by Baclofen (Sanna et al., 2003). Likewise, pretreatment with Diazepam and, separately, GHB, blocks the effect of subunit α4 upregulation induced by ethanol withdrawal (Follesa et al., 2003). Joint administration of ethanol and aloP decreases tolerance and pharmacological dependence in rodents (Morrow et al., 2001). Previous administration and co-administration of progesterone, aloP, and PS with Diazepam or Triazolam reduce tolerance, anxiety, and hyperactivity quantified by BW (Reddy and Kulkarni, 1997). The overall reading of this evidence allows the development of a hypothesis related with the possibility of blocking or reducing abstinence induced by a GABAergic drug by applying another that occupies the same receptor. The decrease factor of the second drug is based on (a) that the sequential administration (prior or parallel to the one generating abstinence, but for short periods), or that (b) the concentration is small.

GW: unique hypothesis regarding withdrawal syndromes induced by agonist by GABAA receptors?

GW is a phenomenon of acute induction, is a result of the increase and abrupt decrease of GABA concentrations in the synaptic cleft, and modifies pre- and postsynaptic functions (Brailowsky et al., 1988) (Fig. 3). Many of the molecular mechanisms underlying the induction of hyperexcitability that characterize GABA abstinence possess similarities to those that can produce other drugs that recognize the same GABAA receptor but that bind different sites. Alcohol, benzodiazepines, and neurosteroids modulate the GABAA receptor, and exposure of these drugs by long-term treatments also induces synaptic changes favoring increased neuronal excitability: modification of the kinetics of Cl− channel activation and inactivation (Smith and Gong, 2005, Smith et al., 2007, Cagetti et al., 2003 and Smith, 2013) and changes in receptor subunit expression (Liang et al., 2014a and Liang et al., 2014b). However, there are also significant differences in the expression of other induction processes of pharmacological and clinical expression, among these withdrawal, e.g., AW is strongest when the interruption of alcohol is intermittent (alcohol administration–interruption–re-administration–induction of withdrawal) (Cagetti et al., 2003 and Sanna et al., 2003). Unlike DW and GWS, these are very strongly induced when discontinuance of administration is acute (Cloos and Bocquet, 2013). For the early phase of withdrawal from benzodiazepines, neurosteroids, and alcohol, the modification is very important of other neurotransmission types, such as glutamatergic, adrenergic, dopaminergic, and the opioid system (Shen et al., 2010, Shen and Tietz, 2011 and Liang et al., 2014b); however, in GW, initially there are no fundamental changes in the NMDA receptor and it is not blocked by Naloxone, indicating that, at least at the beginning, the GWS entertains no relationship with the glutamatergic synaptic or other neurotransmissions (Brailowsky, 1991). The EEG pattern in DW is similar to that induced by GABA and progesterone; however, it causes a different pattern in AW (Rangaswamy and Porjesz, 2014). For pharmacological clinical management and treatment of AW, benzodiazepines and barbiturics are required. However, GW is very resistant to Diazepam, barbiturates, and neurosteroids, and only the drug denominated HEPP has acute antiepileptic effects and decreases in GWS hyperexcitability (Fig 2B). Induction of anxiety caused by DW and AW is very strong with regard to the induction of GW. It is known that GABAB receptor agonists, such as Baclofen and Muscimol, can change drug-seeking behavior when injected into limbic structures. However, this process has not been studied for benzodiazepine withdrawal or for GW. Previous results indicate that Baclofen does not change the electrophysiological evolution of GWS, this suggesting that the GABAB receptor is not involved in the same manner in the induction of GW, in that this GABAB receptor does indeed participate in the generation of abstinence to ethanol.

Synaptic mechanisms that increase the neuronal activity in GABA withdrawal (GW). ...

Fig. 3.

Synaptic mechanisms that increase the neuronal activity in GABA withdrawal (GW). Basic sketch of the GABAergic synapsis, pre- and postsynaptic components are shown; GABA production, storage and release occur in presynaptic, while in post-synapsis, the synaptic and extra-synaptic GABAA receptor modulates inhibition. GABA instillation (5 mM/120 min) significantly increases the quantity of the amino acid in the synaptic cleft; this possesses a direct repercussion in both synaptic components. In the presynaptic component, GABA synthesis is reduced by the decrease of GAD activity and expression (1); release of the neurotransmitter is decreased (2); consequently, there is a decrease of GABA concentration in the synaptic cleft (3). In post-synapsis, the immediate changes are observed as a decrease in the density of receptors (4) and gradually, the new GABAA receptors exhibit a decrease in their pharmacological sensitivity due to changes produced by the expression of new subunits (5). Finally, the increase in excitability is facilitated by an increase in calcium conductance; this potentiates the excitable properties of the neurons (6).

Figure options

One advantage of studying GW is the ability to quantify the changes induced in a specific area of the brain where GABA was applied and to quantify the temporal evolution of hyperexcitability in the EEG of an exact site. GABA concentrations > 5 mM induce faster withdrawal latency (Fig 1) in a manner similar to that which Diazepam and Progesterone can induce. On comparing abstinence induced in the cerebral cortex relative to that which can be generated in the hippocampus, the electrophysiology compared demonstrates that the hippocampus exhibits greater epileptiform activity (frequency, amplitude, and short latency) but takes less time (average length is 7 days compared with the hippocampus, which is 4 days) (Calixto et al., 2000 and Casasola et al., 2004).

Abstinence-induced neurosteroids, benzodiazepines, or alcohol possess a systemic pharmacological character, i.e., these molecules are structurally modified in the liver, and their clearance by the kidneys is immediate. Chronic alcohol or benzodiazepine treatment sets into operation mechanisms of hepatic enzyme induction, which decrease the availability of drugs in the nervous system from the first manifestations, i.e., tolerance to these drugs does not depend solely on the effects on the brain. These drugs exert an effect on several brain structures in parallel; these effects depend on the change in the composition of the receptor and brain structure. The effect of alcohol on the brain takes place in various structures: at the levels of the brainstem, basal ganglia, substantia nigra, cerebellum, cerebral cortex, and hippocampus. However, alcohol also possesses dopaminergic and glutamatergic effects, indicating that its effect on the brain involves more activation on inducing dependency, withdrawal, and repetition of intake. Analogously, benzodiazepines do not induce direct and acute changes in withdrawal induction in other neurotransmission types, but may increase neurosteroid synthesis in mitochondria, promoting cholesterol entry for metabolic synthesis of steroid hormones from glial cells and neurons. These pro-inhibitor, anxiolytic, antiepileptic, and hypnotic effects include short latency and, depending on the molecular structure of each benzodiazepine, these effects can be long or short.

The GWS is a phenomenon induced at a specific anatomical site, with direct installation in the receptors. The spread of the epileptiform phenomenon is possible in other structures, but herein lies a great difference in the process of hyperexcitability in relation with other models of abstinence: in GWS, the synaptic modification site of abstinence to GABA is accurate and localizable, in comparison with abstinences to other drugs, whose anatomical sites of origin are diverse; various of these sites can be activated in series or in parallel, and the propagation of hyperexcitable activity is more diverse.

Abstinence from various drugs whose origin is the GABAA receptor expressed more differences than similarities among them. We propose the hypothesis that each withdrawal induced by different synaptic GABAergic drugs induced the various changes, modifications, and specific biophysical changes that characterize each of them, that the differences are marked, and that each should be studied with the restrictions, limitations, and validation of their scope in order to know whether they are comparable among themselves. Therefore, GW comprises a model study that demonstrates excellent conditions for studying synaptic changes with analogies and differences in the pharmacological modulation of the GABAA receptor, which we plan to continue studying.

Conclusions

Briefly, here are 10 important points regarding GW:

(1)

GW produces pre- and postsynaptic changes related with changes in GABA concentration in the synaptic cleft. Consequently, neuronal excitability increases (decrease of GABA synthesis and release. decrease of GABAA receptors and changes in their subunits; Fig. 3).

(2)

The GWS resembles the focal status epilepticus associated with neuronal hyperexcitability in humans.

(3)

GW is related with syndromes that present modifications in inhibitory transmission and resembles other models of epileptogenesis induced by GABA blockers, such as Bicuculline or Picrotoxin. Both hyperexcitability models present neuronal hyperexcitability, glucidic, hypermetabolic, and excitotoxicity-related neuropathological changes.

(4)

Pharmacologically, the hyperexcitability produced by GW comprises a phenomenon dependent on the GABAA receptors and is followed by the decrease of the activity and expression of the enzyme GAD; this is associated with tolerance to specific agonists of this receptor (isoguvacine). On its genesis, both the decrease in GABA release and the decrease of the GABAA receptor are present. On the other hand, the pharmacological reactivity of GWS changes according to temporal evolution: GWS is resistant to the majority of common anticonvulsant drugs during the first 24 h and is sensitive to benzodiazepines after day 2; however, GWS can be modified by the administration of HEPP or chronic progesterone (Fig 2).

(5)

GW not only presents synaptic changes, but also changes in the properties of the neuronal firing rate: the cells become more excitable in early stages, and the firing threshold decreases. This is associated with changes in action potentials and in addition, the hyperpolarizing postsynaptic potential reduces its amplitude, which is followed by a significant increase of intracellular calcium.

(6)

GW can induce neuronal hyperexcitability in brain slices (Fig. 1) and dissociated neurons.

(7)

GWS can be related with abstinence syndromes that occur after the abrupt interruption of a sustained administration of drugs, such as barbiturates, benzodiazepines, neurosteroids, and alcohol. All these drugs facilitate GABAergic neurotransmission.

(

The early mechanisms of this abstinence are dependent on the GABAA receptor, while the maintenance mechanisms of abstinence involve another type of neurotransmission that facilitates the hyperexcitability process.

(9)

The increase in neuronal excitability produced by GABA deprivation opens new possibilities to investigate the plastic changes that occur in the neuroglia as a consequence of the changes in the levels of GABA (the predominant inhibitory neurotransmitter of the central nervous system, CNS) (i.e., it facilitates LTP). Also, neurochemical and -pharmacological analyses are easy to perform because tissue sample limitations are avoided.

(10)

Finally, the increase of neuronal excitability produced by GW in brain slices can be decreased if an agonist of the receptor is previously exposed to GABA. The neurosteroid, aloP, produces this effect with high efficacy (Fig. 2D).

Acknowledgments

The author would like to thank Javier Gálvez and Sonia Santos for technical services provided. This work is supported by CONACYT-México Grant No. 166823.

References

Allison et al., 2005

C. Allison, J.A. Pratt, T.L. Ripley, D.N. Stephens

AMPA receptor autoradiography in mouse brain following single and repeated withdrawal from Diazepam

Eur J Neurosci, 21 (2005), pp. 1045–1056

View Record in Scopus | Full Text via CrossRef | Citing articles (6)

Andréen et al., 2009

L. Andréen, S. Nyberg, S. Turkmen, G. Van Wingen, G. Fernández, T. Bäckström

Sex steroid induced negative mood may be explained by the paradoxical effect mediated by GABAA modulators

Psychoneuroendocrinology, 34 ( (2009), pp. 1121–1132

Article |  PDF (489 K) | View Record in Scopus | Citing articles (66)

Araneda et al., 1994

S. Araneda, C. Silva-Barrat, C. Menini, R. Naquet

High expression of noradrenaline, choline acetyltransferase and glial fibrillary acid protein in the epileptic focus consecutive to GABA withdrawal. An immunocytochemical study

Brain Res, 655 (1994), pp. 135–146

Article |  PDF (2678 K) | View Record in Scopus | Citing articles (

Arnot et al., 2001

M.I. Arnot, M. Davies, I.L. Martin, A.N. Bateson

GABA(A) receptor gene expression in rat cortex: differential effects of two chronic diazepam treatment regimes

J Neurosci Res, 64 (6) (2001), pp. 617–625

View Record in Scopus | Full Text via CrossRef | Citing articles (21)

Ator et al., 2010

N.A. Ator, J.R. Atack, R.J. Hargreaves, H.D. Burns, G.R. Dawson

Reducing abuse liability of GABAA/benzodiazepine ligands via selective partial agonist efficacy at alpha1 and alpha2/3 subtypes

J Pharmacol Exp Ther, 332 (1) (2010), pp. 4–16

View Record in Scopus | Full Text via CrossRef | Citing articles (32)

Baldwin et al., 2013

D.S. Baldwin, K. Aitchison, A. Bateson, H.V. Curran, S. D

Hi Kris, now I am not sure either  , I had access to the full paper and then I thought it was a good one. So I copied it. 

You can read only the abstract and a small paragraph entitled Benzos.

Long and complicated.

GWS=gaba withdrawal syndrome, gaba as in gaba or gaba-ergic drugs like benzos ?

I will read again later.