The effect of mouse spinal cord transection on the antinociceptive response to the gamma-aminobutyric acid agonists THIP (4,5,6,7-tetrahydroisoxazolo [5,4-c]pyridine-3-ol) and baclofen

Pathophysiological power of improper tonic GABA(A) conductances in mature and immature models

High-affinity extrasynaptic gamma-aminobutyric acid A (GABA_A) receptors are tonically activated by low and consistent levels of ambient GABA, mediating chronic inhibition against neuronal excitability (tonic inhibition) and the modulation of neural development. Synaptic (phasic) inhibition is spatially and temporally precise compared with tonic inhibition, which provides blunt yet strong integral inhibitory force by shunting electrical signaling. Although effects of acute modification of tonic inhibition are known, its pathophysiological significance remains unclear because homeostatic regulation of neuronal excitability can compensate for long-term deficit of extrasynaptic GABA_A receptor activation. Nevertheless, tonic inhibition is of great interest for its pathophysiological involvement in central nervous system (CNS) diseases and thus as a therapeutic target. Together with the development of experimental models for various pathological states, recent evidence demonstrates such pathological involvements of tonic inhibition in neuronal dysfunction. This review focuses on the recent progress of tonic activation of GABA_A conductance on the development and pathology of the CNS. Findings indicate that neuronal function in various brain regions are exacerbated with a gain or loss of function of tonic inhibition by GABA spillover. Disturbance of tonic GABA_A conductance mediated by non-synaptic ambient GABA may result in brain mal-development. Therefore, various pathological states (epilepsy, motor dysfunctions, psychiatric disorders, and neurodevelopmental disorders) may be partly attributable to abnormal tonic GABA_A conductances. Thus, the tone of tonic conductance and level of ambient GABA may be precisely tuned to maintain the regular function and development of the CNS. Therefore, receptor expression and factors for regulating the ambient GABA concentration are highlighted to gain a deeper understanding of pathology and therapeutic strategy for CNS diseases.

Pharmacological enhancement of δ-subunit-containing GABA(A) receptors that generate a tonic inhibitory conductance in spinal neurons attenuates acute nociception in mice

The development of new strategies for the treatment of acute pain requires the identification of novel nonopioid receptor targets. This study explored whether δ-subunit-containing GABA(A)Rs (δGABA(A)Rs) in neurons of the spinal cord dorsal horn generate a tonic inhibitory conductance in vitro and whether δGABA(A)R activity regulates acute nociception. Whole-cell recordings revealed that δGABA(A)Rs generate a tonic inhibitory conductance in cultured spinal neurons and lamina II neurons in spinal cord slices. Increasing δGABA(A)R function by applying the δGABA(A)R-preferring agonist 4,5,6,7-tetrahydroisoxazolo [5,4-c]pyridine-3-ol (THIP) increased the tonic current and inhibited neuronal excitability in spinal neurons from wild-type (WT) but not δ subunit null-mutant (Gabrd(-/-)) mice. In behavioral studies, baseline δGABA(A)R activity did not regulate acute nociception; however, THIP administered intraperitoneally or intrathecally attenuated acute nociception in WT but not Gabrd(-/-) mice. In the formalin nociception assay, the phase 1 response was similar for WT and Gabrd(-/-) mice. In contrast, the phase 2 response, which models central sensitization, was greater in Gabrd(-/-) mice than WT. THIP administered intraperitoneally or intrathecally inhibited phase 1 responses of WT but not Gabrd(-/-) mice and had no effect on phase 2 responses of WT mice. Surprisingly, THIP reduced the enhanced phase 2 response in Gabrd(-/-) mice. Together, these results suggest that δGABA(A)Rs in spinal neurons play a major physiological and pharmacological role in the regulation of acute nociception and central sensitization. Spinal δ-subunit-containing GABA(A) receptors were identified with electrophysiological methods and behavioral models as novel targets for the treatment of acute pain.

The Role of GABA in the Mediation and Perception of Pain

A great deal of effort has been expended in attempting to define the role of GABA in mediating the transmission and perception of pain. Pursuit of this question has been stimulated by the fact that GABAergic neurons are widely distributed throughout the central nervous system, including regions of the spinal cord dorsal horn known to be important for transmitting pain impulses to the brain. In addition, GABA neurons and receptors are found in supraspinal sites known to coordinate the perception and response to painful stimuli and this neurotransmitter system has been shown to regulate control of sensory information processing in the spinal cord. The discovery that GABA receptor agonists display antinociceptive properties in a variety of animal models of pain has provided an impetus for developing such agents for this purpose. It has been shown that GABA receptor agonists, as well as inhibitors of GABA uptake or metabolism, are clinically effective in treating this symptom. However, even with an enhanced understanding of the relationship between GABAergic transmission and pain, it has proven difficult to exploit these findings in designing novel analgesics that can be employed for the routine management of pain. Work in this area has revealed a host of reasons why GABAergic drugs have, to date, been of limited utility in the management of pain. Chief among these are the side effects associated with such agents, in particular sedation. These limitations are likely due to the simultaneous activation of GABA receptors throughout the neuraxis, most of which are not involved in the transmission or perception of pain. This makes it difficult to fully exploit the antinociceptive properties of GABAergic drugs before untoward effects intervene. The discovery of molecularly and pharmacologically distinct GABAA receptors may open the way to developing subtype selective agents that target those receptors most intimately involved in the transmission and perception of pain. The more limited repertoire of GABAB receptor subunits makes it more difficult to develop subtype selective agents for this site. Nonetheless, a GABAB agonist, CGP 35024, has been identified that induces antinociceptive responses at doses well below those that cause sedation (Patel et al., 2001). It has also been reported that, unlike baclofen, tolerance to antinociceptive responses is not observed with CGP 44532, a more potent GABAB receptor agonist (Enna et al., 1998). While the reasons for these differences in responses to members of the same class remain unknown, these findings suggest it may be possible to design a GABAB agonist with a superior clinical profile than existing agents. Besides the challenges associated with identifying subtype selective GABAA and GABAB receptor agonists, the development of GABA analgesics has been hindered by the fact that the responsiveness of these receptor systems appear to vary with the type and duration of pain being treated and the mode of drug administration. Further studies are necessary to more precisely define the types of pain most amenable to treatment with GABAergic drugs. Inasmuch as the antinociceptive responses to these agents in laboratory animals are mediated, at least in part, through activation or inhibition of other neurotransmitter and neuromodulator systems, it is conceivable that GABA agonists will be most efficacious as analgesics when administered in combination with other agents. The results of anatomical, biochemical, molecular, and pharmacological studies support the notion that generalized activation of GABA receptor systems dampens the response to painful stimuli. The data leave little doubt that, under certain circumstances, stimulation of neuroanatomically discreet GABA receptor sites could be of benefit in the management of pain. Continued research in this area is warranted given the limited choices, and clinical difficulties, associated with conventional analgesics.

GABA(B) receptor alterations as indicators of physiological and pharmacological function

Given the widespread distribution of GABA(B) receptors throughout the central nervous system, and within certain peripheral organs, it is likely their selective pharmacological manipulation could be of benefit in the treatment of a variety of disorders. Studies aimed at defining the clinical potential of GABA(B) receptor agonists and antagonists have included gene deletion experiments, examination of changes in receptor binding, subunit expression and function in diseased tissue, as well as after the chronic administration of drugs. The results indicate that a functional GABA(B) receptor requires the combination of GABA(B(1)) and GABA(B(2)) subunits, that receptor function does not always correlate with subunit expression and receptor binding, and that GABA(B) receptor modifications may be associated with the clinical response to antidepressants, mood stabilizers, and GABA(B) receptor agonists and antagonists. Moreover, changes in GABA(B) binding or expression suggest this receptor may be involved in mediating symptoms associated with chronic pain, epilepsy and schizophrenia. This, together with results from other types of studies, indicates the potential therapeutic value of developing drugs capable of selectively activating, inhibiting, or modulating GABA(B) receptor function.

Relationship between the antinociceptive response to desipramine and changes in GABAB receptor function and subunit expression in the dorsal horn of the rat spinal cord

Although tricyclic antidepressants are among the drugs of choice for the treatment of neuropathic pain, their mechanism of action in this regard remains unknown. Because previous reports suggest these agents may influence gamma-aminobutyric acid (GABA) neurotransmission, and GABAB receptors are known to participate in the transmission of pain impulses, the present experiments were undertaken to examine whether the administration of desipramine alters GABAB receptor subunit expression and function in the dorsal horn of the rat spinal cord. For the study, rats were injected (i.p.) once daily with desipramine (15 mg/kg) for 7 consecutive days, during which their thermal withdrawal threshold was monitored, and after which GABAB receptor function, and the levels of GABAB receptor subunit mRNA, were quantified in the spinal cord dorsal horn. The results indicate that 4-7 days of continuous administration of desipramine are necessary to observe a significant increase in the thermal pain threshold. Moreover, it was found that 7 days of treatment with desipramine enhances GABAB receptor function, as measured by baclofen-stimulated [35S]GTPgammaS binding, and increases mRNA expression for the GABAB(1a) and GABAB(2), but not GABAB(1b), subunits. These findings suggest the antinociceptive effect of desipramine is accompanied by a change in spinal cord GABAB receptor sensitivity that could be an important component in the analgesic response to this agent.

Role of arginine in immunonutrition

Arginine plays an important role in many physiologic and biologic processes beyond its role as a protein-incorporated amino acid. Dietary supplementation of arginine can enhance wound healing, regulate endocrine activity and potentiate immune activity. Under normal unstressed conditions the arginine requirement of adult humans is fulfilled by endogenous sources, however this is compromised during times of stress, especially in critical illness. These finding have led to use of arginine supplementation as part of an immune-enhancing dietary regimen to help combat the immune suppression seen in such patients. Though the results from studies examining the use of this type of immunonutrition in critically ill patients are far from definitive, they are promising that this mode of therapy may be of some advantage. A better understanding of the in vivo biology of arginine and its metabolism is necessary to truly define a benefit from arginine supplementation.

Antinociception produced by systemic R(+)-baclofen hydrochloride is attenuated by CGP 35348 administered to the spinal cord or ventromedial medulla of rats

This study examined the sites in the central nervous system at which subcutaneously-administered R(+)-baclofen hydrochloride (baclofen), the most active isomer of this prototypic gamma-aminobutyric acid (GABA)B receptor agonist, acts to produce antinociception in the rat. To determine whether baclofen acts in the spinal cord, either saline or the GABAB receptor antagonist CGP 35348 was injected intrathecally in rats pretreated 24 min earlier with 1 or 3 mg/kg s.c. baclofen. Intrathecal (i.t.) injection of 3 or 10 micrograms of CGP 35348 antagonized the increase in tail-flick and hot-plate latency produced by either dose of baclofen. To determine whether baclofen acts at sites in the ventromedial medulla (VMM), either saline or CGP 35348 was microinjected in the nucleus raphe magnus or nucleus reticularis gigantocellularis pars alpha of rats pretreated 24 min earlier with 1 or 3 mg/kg s.c. baclofen. Microinjection of 0.5 or 3 micrograms of CGP 35348 at sites in the VMM produced at best only a very modest attenuation of the antinociceptive effects of baclofen. These data suggest that systemically-administered baclofen acts at sites in both the spinal cord and the VMM, but that its antinociceptive effects are likely to be mediated to a greater extent by a spinal, rather than medullary site of action. However, a definitive comparison of the relative contribution of GABAB receptors in these two regions is precluded by differences in the diffusion and concentrations of the antagonist in the spinal cord and brainstem. Finally, microinjection of 0.5 or 3.0 micrograms of CGP 35348 in the nucleus raphe magnus or nucleus reticularis gigantocellularis pars alpha of saline-pretreated rats did not alter tail-flick or hot-plate latency. This finding suggests that, unlike GABAA receptors, GABAB receptors do not mediate the tonic GABAergic input to neurons in these nuclei.

Comparison of the antinociceptive and antispastic action of (-)-baclofen after systemic and intrathecal administration in intact, acute and chronic spinal rats

Baclofen is particularly effective in treating spasticity of spinal origin in humans. However, most investigations of this drug in animals have only assessed its antinociceptive effect, presumably because of the difficulty in developing animal models of spasticity. This study attempted to evaluate both, the antinociceptive and antispastic action of (-)-baclofen (the more active enantiomer) by incorporating the chronic spinal preparation, in which spasticity gradually develops following spinal transection. Separate groups of intact, acute (1 day) or chronic (20-25 days) spinal rats were pretested on the nociceptive tail-flick (TF) assay prior to either subcutaneous (SC; 1-30 mg/kg) or intrathecal (IT; 0.1-12 micrograms) injection of (-)-baclofen and retested at specific post-injection intervals. Hindlimb spasticity was elicited in chronic spinal rats by mechanical stimulation to the abdomen. Because the clinical use of baclofen generally involves chronic administration, both responses were tested for 3 successive days to assess tolerance. Results confirmed the analgesic effect of SC and IT (-)-baclofen in intact rats. As previously reported, the antinociceptive effect of IT (-)-baclofen was increased in acute spinal rats. However, three weeks after spinalization there was a profound decrease in this response. In contrast, antinociception produced by SC (-)-baclofen was reduced in acute and chronic spinal rats compared to intact animals; but there was no difference between the acute and chronic conditions. In spite of this differential decrease in antinociception after IT, relative to SC, administration, both routes of administration produced an antispastic effect in chronic spinal rats. There was no antinociceptive tolerance to SC administration and only minimal tolerance to IT (-)-baclofen (in intact rats); the antispastic effect did not become tolerant. A peripheral action might explain the dichotomy between SC and IT (-)-baclofen in regard to antinociception. However, further research is needed to determine why both routes of administration were effective against spasticity while only SC (-)-baclofen retained an antinociceptive action in chronic spinal rats.

Baclofen as an adjuvant analgesic

Baclofen is a gamma-aminobutyric acid (GABA) agonist approved for the treatment of spasticity and commonly used in the management of many types of neuropathic pain. Controlled studies have demonstrated the efficacy of this drug in trigeminal neuralgia. Although its precise mechanism of analgesic action is unknown, it is likely that a drug-induced increase in inhibitory activity is sufficient to interrupt the cascade of neural events that culminates in aberrant activity of wide dynamic range neurons, or more rostral neurons in nociceptive pathways, that is the substrate for some types of neuropathic pain. The optimal use of baclofen as an adjuvant analgesic requires an understanding of its pharmacology, side effect spectrum, and dosing guidelines that have proven useful in clinical practice. Failure of baclofen therapy following a prolonged trial requires dose tapering prior to discontinuation due to the potential for a withdrawal syndrome.

Phaclofen antagonizes the antinociceptive but not the sedative effects of (-)-baclofen

1. Intraperitoneal (ip) injection of (-)-baclofen induced long-lasting antinociceptive and sedative effects in rats. 2. Phaclofen, the phosphonic derivative of baclofen, fully antagonized the antinociceptive effect of (-)-baclofen. When injected intracerebroventricularly (icv), but not ip, phaclofen antagonized in a dose-dependent fashion (50-200 micrograms) the delays in behavioral response induced by (-)-baclofen (2.5-10 mg/kg ip) in both hot plate and tail flick tests. 3. In addition phaclofen (100 micrograms icv) counteracted the loss of the righting reflex induced by (-)-baclofen (7.5-15 mg/kg ip). 4. In contrast, phaclofen (100-200 micrograms icv) counteracted only in part the sedative effect of (-)-baclofen. In rats pretreated with the antagonist (200 micrograms icv), the electrocorticographic hypersynchrony due to (-)-baclofen (5 mg/kg ip) is replaced by a synchronized pattern associated with behavioral sedation. 5. These data are consistent with the reported antagonism by phaclofen on the effects of (-)-baclofen. They also seem to indicate that in rats phaclofen-sensitive GABA-B receptors play an important role in the analgesic effects of baclofen, but only a minor role in the sedative effects of this drug.

Antinociceptive effect of intrathecal morphine in tolerant and nontolerant spinal rats

The antinociceptive effect of intrathecal morphine on the tail-flick (TF) reflex of rats was significantly enhanced within one day after spinal transection (ED50 = 0.125 microgram) relative to the effect obtained in intact rats (ED50 = 5.9 micrograms). By 20-30 days after spinalization the potency of intrathecally administered morphine had substantially declined. Intact rats, made tolerant to the antinociceptive effect of systemic morphine (3.0 mg/kg, SC on each of seven consecutive days), were not tolerant to intrathecal morphine (ED50 = 6.5 micrograms). In contrast, rats that were pretreated with either morphine alone, repeated TF tests alone, or both of these treatments, were tolerant to intrathecal morphine when tested one day after spinal transection. The results suggest first, that the antinociceptive effect of intrathecal morphine in intact rats is tonically inhibited by descending supraspinal input and that removal of this input is responsible for the enhanced antinociceptive effect of intrathecal morphine in spinal rats. Second, the data suggest that tolerance to the antinociceptive effect of intrathecal morphine in intact rats may also be tonically inhibited by supraspinal input, because spinal opiate tolerance is expressed after spinal transection.

GABAergic modulation of nociceptive threshold: effects of THIP and bicuculline microinjected in the ventral medulla of the rat

Neurons of the nucleus raphe magnus (NRM) and nucleus reticularis gigantocellularis pars alpha (NGCp alpha) have been implicated in the regulation of nociceptive threshold and production of antinociception. Previous studies have shown that the activity of these neurons is modulated by noradrenergic, cholinergic and serotonergic afferents. The present study examined whether these neurons are additionally subject to regulation by a GABAergic input. Microinjection of the GABAA receptor agonist 4,5,6,7-tetrahydroisoxazolo[5,4-c]pyridin-3-ol (THIP; 0.3 or 1.0 microgram) in the NRM or NGCp alpha significantly decreased tail flick latency (TFL) and increased responsiveness to noxious pinch. Hot plate latency (HPL) was not affected by microinjection of 0.3 microgram THIP. Although HPL was increased after microinjection of 1.0 microgram THIP, this effect may reflect motoric disturbances. In contrast to the hyperalgesia produced by THIP, microinjection of the GABAA receptor antagonist bicuculline methiodide (0.04 or 0.1 microgram) produced a small, but significant increase in TFL. Responsiveness to noxious pinch and HPL were not affected by either dose. These findings indicate that neurons of the NRM or NGCp alpha involved in the regulation of nociceptive threshold are subject to an inhibitory GABAergic input mediated by a GABAA receptor. However, in contrast to previously described inhibitory inputs, the GABAergic influence does not appear to be tonically active to a substantial extent in the unanesthetized rat.

The GABA agonist THIP, attenuates antinociception in the mouse by modifying central cholinergic transmission

The effect of THIP, a direct-acting gamma-aminobutyric acid (GABA) receptor agonist, on the antinociceptive response to a variety of agents was examined using the mouse tail-immersion assay. Alone, THIP produced an antinociceptive response in smaller doses (5 mg/kg) but was ineffective at doses exceeding 10 mg/kg. Treatment with THIP (15 mg/kg) was found to block the antinociceptive response to an inhibitor of the uptake of GABA, an inhibitor of GABA-transaminase, a direct-acting GABA receptor agonist and to a cholinesterase inhibitor. In contrast, THIP had no effect on the antinociceptive responses to morphine, clonidine or oxotremorine. The results indicate that large doses of THIP reduce cholinergic activity in a pathway important for mediating the antinociceptive action of GABAergic drugs and physostigmine.

GABAergic mechanisms of analgesia: an update

Both directly acting (GABAA and GABAB agonists) and indirectly acting GABAergic agents (GABA uptake inhibitors and GABA-transaminase inhibitors) produce analgesia in a variety of animal test systems. Analgesia produced by GABAA agonists is probably due to a supraspinal action, although spinal sites may also play a role. GABAA agonist analgesia is insensitive to naloxone, bicuculline, picrotoxin and haloperidol, but is blocked by atropine, scopolamine and yohimbine suggesting a critical role for central cholinergic and noradrenergic pathways in this action. The lack of blockade by the GABAA antagonist bicuculline is difficult to explain. Both bicuculline and picrotoxin have intrinsic analgesia actions which may not necessarily be mediated by GABA receptors. The GABAB agonist baclofen produces analgesia by actions at both spinal and supraspinal sites. Baclofen analgesia is insensitive to naloxone, bicuculline and picrotoxin, and blockade by cholinergic antagonists occurs only under limited conditions. Catecholamines are important mediators of baclofen analgesia because analgesia is potentiated by reserpine, alpha-methyl-p-tyrosine, phentolamine, ergotamine, haloperidol and chlorpromazine. A role for serotonergic mechanisms is less well defined. Methylxanthines, which produce a clonidine-sensitive increase in noradrenaline (NA) turnover, increase baclofen analgesia by a clonidine-sensitive mechanism. Both ascending and descending NA pathways are implicated in the action of baclofen because dorsal bundle lesions, intrathecal 6-hydroxydopamine and medullary A1 lesions markedly decrease baclofen analgesia. However, simultaneous depletion of NA in ascending and descending pathways by locus coeruleus lesions potentiates baclofen analgesia suggesting a functionally important interaction between the two aspects. Baclofen analgesia within the spinal cord may be mediated by a distinct baclofen receptor because GABA does not mimic the effect of baclofen and the rank order of potency both of close structural analogs of baclofen as well as antagonists differs for analgesia and GABAB systems. The spinal mechanism may involve an interaction with substance P (SP) because SP blocks baclofen analgesia, and desensitization to SP alters the spinal analgesic effect of baclofen. GABA uptake inhibitors produce analgesia which is similar to that produced by GABAA agonists because it is blocked by atropine, scopolamine and yohimbine. Analgesia produced by GABA-transaminase inhibitors is similar to that produced by GABAA agonists because it can be blocked by atropine, but it is potentiated by haloperidol while THIP analgesia is not.(ABSTRACT TRUNCATED AT 400 WORDS)

Nociceptive responses to altered GABAergic activity at the spinal cord

GABA agonists and antagonists were injected intrathecally at the spinal cord, to determine their effect on nociceptive thresholds. Tactile stimulation, applied against the flank by a medium diameter von Frey fiber (5.5 g force), elicited distress vocalizations after, but not before injection of the GABA antagonists, bicuculline MI or picrotoxin (0.25 and 1 microgram dosages). Vocalization threshold to tail shock was significantly reduced by bicuculline MI or picrotoxin. Tail flick withdrawal latency from radiant heat was not altered by GABA antagonists. The GABA agonist, muscimol, significantly elevated vocalization threshold to tail shock at a 5 micrograms dose. At a lower dose level (1 microgram), muscimol significantly reduced vocalization threshold to tail shock. Tail flick latency was significantly prolonged by the 5 micrograms dose of muscimol; however, flaccid paralysis of the hind limbs was also evident. Nociceptive thresholds were not altered by GABA or saline injection. These findings indicate that GABAergic activity contributes to the tonic modulation of nociception at the spinal cord.

The parafasciculus thalami as a site for mediating the antinociceptive response to GABAergic drugs

Electrophysiological (single cell) experiments were undertaken to examine whether neurons in the rat parafasciculus thalami (PF) are involved in mediating the antinociceptive response to GABAergic drugs. The results indicated that: noxious stimuli excite most PF neurons; microiontophoretic application of morphine, GABA or the GABA agonist, THIP, attenuated the spontaneous firing rate of PF neurons; morphine, THIP and GABA reduced the neuronal excitation induced by noxious stimuli; application of the GABA receptor antagonist, bicuculline, prevented the effects of THIP and GABA on PF activity; while naloxone blocked the response to morphine on PF neurons, it failed to influence the actions of GABA and THIP; and the injection of THIP or GABA into the PF produced an antinociceptive response as assessed by the rat tail-immersion assay, whereas pentobarbital was inactive. The findings suggest that GABA receptors located in the PF may mediate the antinociceptive response to GABAergic drugs, and that the action of these agents is unrelated to opiate receptors.

GABA uptake inhibitors produce a greater antinociceptive response in the mouse tail-immersion assay than other types of GABAergic drugs

Antinociception produced by the GABA uptake inhibitors d,l- SKF-89976A and SKF-100330A was characterized and compared to that produced by other types of GABAergic drugs. Using the mouse tail-immersion assay it was found that the antinociception produced by the uptake inhibitors was antagonized by scopolamine, a cholinergic muscarinic receptor antagonist. However, neither SKF compound demonstrated any significant affinity for muscarinic receptor binding sites suggesting that they are not direct-acting cholinomimetics. In vitro uptake experiments revealed that the SKF compounds selectively inhibit GABA transport, having no effect on the accumulation of aspartic acid, glutamic acid, beta-alanine or glycine. Moreover, antinociception and GABA uptake inhibition were stereoselective for SKF-89976A, with the d-isomer being more active in both tests. When comparing antinociceptive responses at maximally effective doses it was also found that the SKF compounds were substantially more efficacious than direct-acting GABA receptor agonists or a GABA transaminase inhibitor. These data suggest that uptake inhibitors may be facilitating GABA transmission in a system that is less affected by other types of GABAergic compounds.