Amygdala Kindling Potentiates Seizure-Stimulated Immediate-Early Gene Expression in Rat Cerebral Cortex

Kindling induces long-term adaptations in neuronal function that lead to a decreased threshold for induction of seizures. In the present study, the influence of amygdala kindling on levels of mRNA for the immediate-early genes (IEGs) c-fos, c-jun, and NGF1-A were examined both before and after an acute electroconvulsive seizure (ECS). Although amygdala kindling did not significantly influence resting levels of c-fos mRNA in cerebral cortex, ECS-stimulated levels of c-fos mRNA (examined 45 min after ECS) were approximately twofold greater in the cerebral cortex of kindled rats relative to sham-treated controls. The influence of kindling on IEG expression was dependent on the time course of kindling, as ECS-stimulated levels of c-fos mRNA were not significantly increased in stage 2 kindled animals. ECS-stimulated levels of c-jun and NGF1-A mRNA were also significantly increased in cerebral cortex of kindled rats relative to sham-treated controls. The influence of kindling on IEG expression was long-lasting because an acute ECS stimulus significantly elevated levels of c-fos and c-jun mRNA in the cerebral cortex of animals that were kindled 5 months previously. In contrast to these effects in cerebral cortex, kindling did not influence ECS-stimulated levels of c-fos mRNA in hippocampus. Finally, immunohistochemical studies revealed lamina-specific changes in the cerebral cortex.(ABSTRACT TRUNCATED AT 250 WORDS)     

Amygdala kindling differentially regulates the expression of the elements involved in TRH transmission

Subthreshold electrical stimulation of the amygdala (kindling) activates neuronal pathways increasing the expression of several neuropeptides including thyrotropin releasing-hormone (TRH). Partial kindling enhances TRH expression and the activity or its inactivating ectoenzyme; once kindling is established (stage V), TRH and its mRNA levels are further increased but TRH-binding and pyroglutamyl aminopeptidase II (PPII) activity decreased in epileptogenic areas. To determine whether variations in TRH receptor binding or PPII activity are due to regulation of their synthesis, mRNA levels of TRH receptors (R1, R2) and PPII were semi-quantified by RT-PCR in amygdala, frontal cortex and hippocampus of kindled rats sacrificed at stage II or V. Increased mRNA levels of PPII were found at stage II in amygdala and frontal cortex, and of pro-TRH and TRH-R2, in amygdala and hippocampus. At stage V, pro-TRH mRNA levels increased and those of PPII, decreased in the three regions; TRH-R2 mRNA levels diminished in amygdala and frontal cortex and of TRH-R1 only in amygdala. In situ hybridization analyses revealed, at stage II, enhanced TRH-R1 mRNA levels in dentate gyrus and amygdala while decreased in piriform cortex; those of TRH-R2 increased in amygdala, CA2, dentate gyrus, piriform cortex, thalamus and subiculum and of PPII, in CAs and piriform cortex. In contrast, at stage V decreased expression of TRH-R1 occurred in amygdala, CA2/3, dentate gyrus and piriform cortex; of TRH-R2 in CA2, thalamus and piriform cortex, and of PPII in CA2, and amygdala. The magnitude of changes differed between ipsi and contralateral side. These results support a trans-synaptic modulation of all elements involved in TRH transmission in conditions that stimulate the activity of TRHergic neurons. They show that reported changes in PPII activity or TRH-binding caused by kindling relate to regulation of the expression of TRH receptors and degrading enzyme.     

Down-Regulation of AMPA Receptor Subunit GluR2 in Amygdaloid Kindling

Abstract: Alterations in glutamatergic transmission are postulated to be important in kindling and epilepsy. The levels of α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) receptor subunits (GluR1, 2, and 4) were compared in amygdalakindled and sham-operated animals using subunit-specific antibodies and quantitative western blotting. Four limbic regions were examined: limbic forebrain, piriform cortex/amygdala, hippocampus, and entorhinal cortex. When subunit levels were examined 24 h after the last stage 5 seizure, levels of GluR2 were found to be selectively reduced in limbic forebrain (30%) and piriform cortex/amygdala (25%), with no changes in other regions examined. In addition, no changes in the other subunits were observed in any region. The decrease in GluR2 that was observed in kindled animals at 24 h was no longer present at 1 week and 1 month after the last stage 5 seizure. Because the GluR2 subunit uniquely determines the calcium permeability of these receptors and because the piriform cortex has been implicated as a source of excitatory drive for limbic seizures, reduced GluR2 expression may be important in increasing neuronal excitability in kindling-induced epilepsy, or may reflect a compensatory mechanism resulting from kindling.     

Enduring Effects of Early Life Stress on Firing Patterns of Hippocampal and Thalamocortical Neurons in Rats: Implications for Limbic Epilepsy

Early life stress results in an enduring vulnerability to kindling-induced epileptogenesis in rats, but the underlying mechanisms are not well understood. Recent studies indicate the involvement of thalamocortical neuronal circuits in the progression of kindling epileptogenesis. Therefore, we sought to determine in vivo the effects of early life stress and amygdala kindling on the firing pattern of hippocampus as well as thalamic and cortical neurons. Eight week old male Wistar rats, previously exposed to maternal separation (MS) early life stress or early handling (EH), underwent amygdala kindling (or sham kindling). Once fully kindled, in vivo juxtacellular recordings in hippocampal, thalamic and cortical regions were performed under neuroleptic analgesia. In the thalamic reticular nucleus cells both kindling and MS independently lowered firing frequency and enhanced burst firing. Further, burst firing in the thalamic reticular nucleus was significantly increased in kindled MS rats compared to kindled EH rats (p<0.05). In addition, MS enhanced burst firing of hippocampal pyramidal neurons. Following a stimulation-induced seizure, somatosensory cortical neurons exhibited a more pronounced increase in burst firing in MS rats than in EH rats. These data demonstrate changes in firing patterns in thalamocortical and hippocampal regions resulting from both MS and amygdala kindling, which may reflect cellular changes underlying the enhanced vulnerability to kindling in rats that have been exposed to early life stress.     

Amygdala adenosine A1 receptors have no anticonvulsant effect on piriform cortex-kindled seizures in rat

Adenosine is an endogenous anticonvulsant that exerts its effects through A1 receptors. As the piriform/amygdala is a critical circuit for limbic seizure propagation, in this study, the role of basolateral amygdala A1 receptors on piriform cortex (PC)-kindled seizures was investigated. Rats were kindled by daily electrical stimulation of PC. In fully kindled animals, bilateral intra-amygdala N6-cyclohexyladenosine (CHA; 10-500 micromol/L, a selective A1 receptor agonist) had no effect on kindled-seizure parameters. However, bilateral intra-amygdala 2% lidocaine (reversal neuronal inhibitor) reduced the kindled seizure severity. There was significant increase in stage 4 latency and decrease in stage 5 duration. Bilateral lesion of basolateral amygdala of kindled animals (by electrical DC current) reduced the kindled seizure severity more dramatically. Our results showed afterdischarge duration, stage 5 duration, and seizure duration were decreased and stage 4 latency increased significantly. In addition, daily intra-amygdala CHA had no significant effect on PC kindling acquisition. Therefore, it may be concluded that although the basolateral amygdala neuronal activity has a critical role in the propagation of epileptic seizures from PC, the amygdala A1 receptors have no role in this regard. On the other hand, amygdala A1 receptors have no anticonvulsant or antiepileptogenic effect on PC-kindled seizures.     

Changes of immunoreactive somatostatin and beta-endorphin content in rat brain after amygdaloid kindling

A possible contribution of brain beta-endorphin and somatostatin to the epileptogenicity established by amygdaloid kindling was investigated in rats. Fourteen male rats were chronically implanted with electrodes placed bilaterally into the amygdala. The rats received 1 sec of electrical stimulation to the left amygdala each day. Generalized seizures were observed on average 10 days after initiation of kindling and the electrical stimulation was continued up to twenty-one days. Two months after the completion of the kindling procedure, each kindled and control rat was killed by microwave irradiation and the brains were dissected on ice into thirteen subregions. Each region was homogenized and centrifuged twice in 0.1 N acetic acid. The supernatant extracts were decanted and stored at - 20 degrees C until assay. Immunoreactive beta-endorphin and somatostatin were measured by radioimmunoassays. There were no significant differences in brain beta-endorphin contents between the two groups. In kindled rats, immunoreactive somatostatin was increased significantly in amygdala, sensorimotor, piriform, and entorhinal cortex. The results suggest that changes in somatostatin may be associated with epileptic susceptibility induced by the electrical kindling procedure.     

Hippocampal Kindling in the Rat Is Associated with Time-Dependent Increases in the Concentration of Glial Fibrillary Acidic Protein

The effect of hippocampal kindling on the regional brain concentration of total glial fibrillary acidic protein (GFAP), a marker of reactive astrocytes, was studied in partially kindled rats, and in fully kindled rats after a post-kindling period of 24 h, 1 week, and 2 months. GFAP concentration was measured in arbitrary units by dot-blots. In the hippocampus, dentate gyrus, basolateral amygdala, pyriform cortex, and entorhinal cortex, limbic structures which are known to be involved in the kindling process, there was an increase in GFAP concentration which was maximal in the fully kindled animals studied after 24 h. In most brain areas, GFAP concentration was still elevated 1 week post-kindling, but had declined to control level 2 months post-kindling. A significant increase in GFAP was also found in septum, ventral pallidum/accumbens nucleus, and primary motor cortex of kindled rats with a post-kindling period of 24 h, whereas in several other brain regions GFAP was unchanged. These results suggest that astrocyte activation, indicative of degenerative changes in nearby neurons, is a transient and regional phenomenon in kindling occurring only during the development of the kindled state.     

Involvement of Putrescine in the Development of Kindled Seizure in Rats

During the development of kindling by daily electrical stimulations applied to the left amygdala of rats, concentrations of the polyamines putrescine, spermidine, and spermine were measured in the left amygdala and the remainder of the cerebrum. A significant increase of putrescine concentration appeared first at the left amygdala in prekindled rats and then propagated to the remainder of the cerebrum with the development of kindling. This increase in putrescine concentration in the left amygdala was higher in prekindled rats than in fully kindled rats and lasted for at least 24 h after the final kindling stimulation. The concentrations of spermidine and spermine were slightly increased in a fully kindled state. To clarify the role of putrescine in kindling, the development of amygdaloid kindling was examined in rats after microinjections of alpha-difluoromethylornithine, a specific inhibitor of polyamine synthesis, and putrescine into the ipsilateral amygdala. Pretreatment with alpha-difluoromethylornithine (50 nmol) for 10 days accelerated both the development of behavioral kindling and the propagation of the afterdischarge from the left amygdala to the frontal cortex. In contrast, pretreatment with putrescine (200 nmol) for 10 days retarded the development of kindling. These results suggest that the increase in putrescine concentration in the kindled brain has an inhibitory effect on the development of kindling.     

Alterations in Glutamate Transporter Protein Levels in Kindling-Induced Epilepsy

Abstract: There is increasing evidence that levels of glutamate are elevated in certain brain regions immediately prior to and during induction and propagation of seizures. Modulation of high-affinity glutamate uptake is a potential mechanism responsible for the elevated levels observed with seizures. To date, three distinct Na⁺-dependent glutamate transporters have been cloned from rat and rabbit: GLT-1, GLAST, and EAAC-1. We performed a series of experiments to determine whether levels of these transporters are altered in amygdala-kindled rats. Levels of GLT-1, GLAST, and EAAC-1 were examined in three brain regions (hippocampus, piriform cortex/amygdala, and limbic forebrain) by quantitative immunoblotting using subtype-specific antibodies. GLAST protein was down-regulated in the piriform cortex/amygdala region of kindled rats as early as 24 h after one stage 3 seizure and persisting through multiple stage 5 seizures. In contrast, kindling induced an increase in EAAC-1 levels in piriform cortex/amygdala and hippocampus once the animals had reached the stage 5 level. No changes in GLT-1 were observed in any region examined. Changes in transporter levels could contribute to the changes in glutamate levels seen with kindling.     

Time- and Region-Specific Variations in Somatostatin Release Following Amygdala Kindling in the Rat

Abstract: Somatostatin biosynthesis is activated during and following kindling epileptogenesis. The aim of this study was to investigate whether this phenomenon translates into enhanced release of the peptide and whether it is involved in kindling maintenance. A marked increase in somatostatin-like immunoreactivity (somatostatin-LI) was observed in hilar interneurons of the hippocampus and in their presumed projections to the outer molecular layer 1 week, but not 1 month, after the last kindled seizure. No overt changes were observed in the striatum or in the cortex. Compared with sham-stimulated controls, (a) in the hippocampus, high-K⁺-evoked somatostatin-LI release was unchanged in synaptosomes taken from rats killed 7 days after the last kindled seizure but was bilaterally reduced after 30 days; (b) in the striatum, it was increased (mainly ipsilaterally to stimulation) 7, but not 30, days after the last seizure; and (c) in the cortex, somatostatin-LI release was bilaterally increased in synaptosomes taken from kindled rats 30, but not 7, days after the last seizure. This study shows that distinct changes occur in synaptosomal somatostatin-LI release after kindling acquisition, depending on the brain area analyzed and on the time elapsed from the last generalized seizure.     

Amygdala Kindling Modifies Extracellular Opioid Peptide Content in Rat Hippocampus Measured by Microdialysis

Abstract: Opioid peptide release in the hippocampus was shown to be increased immediately following amygdala kindling stimulation in freely moving rats using microdialysis combined with a universal opioid peptide radioimmunoassay (RIA). Extracellular opioid peptide levels were elevated (55% above basal levels) within the first 10 min after electrical stimulation-induced partial seizures in previously nonkindled animals. Fully kindled rats showed lower extracellular opioid peptide levels (40% reduction) during the interictal period [16 ± 2.1 days (mean ± SEM) after the last stage V seizure], in comparison with values obtained from the sham-kindled group under basal conditions. However, opioid peptide release in fully kindled rats increased above 152% of interictal levels within the first 20 min after onset of fully kindled seizures, attaining peak levels equal to that of the partial kindled group and returning to prestimulation conditions 40–60 min following the ictal events. The majority of the immunoreactive material recovered from the hippocampus within the first 20 min following partial and generalized kindled seizures coeluted with dynorphin-A (1–6), dynorphin-A (1–8), and Leu-enkephalin by HPLC/RIA analysis. It is proposed that the enhanced opioid peptide release in hippocampus induced by amygdala kindling stimulation might be associated with either enhanced excitability or seizure suppression as seizure susceptibility fluctuates. The reduced interictal opioid peptide levels may also underlie some interictal behavioral disturbances.     

Decreased benzodiazepine receptor binding in amygdala-kindled rat brains

3H-Flunitrazepam (3H-FLU) binding was measured in multiple brain regions of amygdala-kindled rats two weeks following the sixth Stage 5 convulsion. As compared to 'yoked' controls, the kindled animals displayed significant reductions in 3H-FLU binding in the ipsilateral cortex (20%) and in the hypothalamus (20%). Scatchard plots revealed that these reductions were due to changes in the maximal number of available binding sites (Bmax) rather than to alterations in receptor affinity (KD). No significant changes were found in the contralateral cortex, or in either the contralateral or ipsilateral amygdala, hippocampus or striatum. These data suggest that kindling is associated with long-lasting changes in the benzodiazepine receptor system and possibly with related changes in GABA-mediated neural inhibition.     

Z944, a Novel Selective T-Type Calcium Channel Antagonist Delays the Progression of Seizures in the Amygdala Kindling Model

Temporal lobe epilepsy (TLE) is the most common form of drug resistant epilepsy. Current treatment is symptomatic, suppressing seizures, but has no disease modifying effect on epileptogenesis. We examined the effects of Z944, a potent T-type calcium channel antagonist, as an anti-seizure agent and against the progression of kindling in the amygdala kindling model of TLE. The anti-seizure efficacy of Z944 (5mg/kg, 10mg/kg, 30mg/kg and 100mg/kg) was assessed in fully kindled rats (5 class V seizures) as compared to vehicle, ethosuximide (ETX, 100mg/kg) and carbamazepine (30mg/kg). Each animal received the seven treatments in a randomised manner. Seizure class and duration elicited by six post-drug stimulations was determined. To investigate for effects in delaying the progression of kindling, naive animals received Z944 (30mg/kg), ETX (100mg/kg) or vehicle 30-minutes prior to each kindling stimulation up to a maximum of 30 stimulations, with seizure class and duration recorded after each stimulation. At the completion of drug treatment, Ca_V3.1, Ca_V3.2 and Ca_V3.3 mRNA expression levels were assessed in the hippocampus and amygdala using qPCR. Z944 was not effective at suppressing seizures in fully kindled rats compared to vehicle. Animals receiving Z944 required significantly more stimulations to evoke a class III (p<0.05), IV (p<0.01) or V (p<0.0001) seizure, and to reach a fully kindled state (p<0.01), than animals receiving vehicle. There was no significant difference in the mRNA expression of the T-type Ca²⁺ channels in the hippocampus or amygdala. Our results show that selectively targeting T-type Ca²⁺ channels with Z944 inhibits the progression of amygdala kindling. This could be a potential for a new therapeutic intervention to mitigate the development and progression of epilepsy.     

A Transient Upregulation of Glutamine Synthetase in the Dentate Gyrus Is Involved in Epileptogenesis Induced by Amygdala Kindling in the Rat

Reduction of glutamine synthetase (GS) function is closely related to established epilepsy, but little is known regarding its role in epileptogenesis. The present study aimed to elucidate the functional changes of GS in the brain and its involvement in epileptogenesis using the amygdala kindling model of epilepsy induced by daily electrical stimulation of basolateral amygdala in rats. Both expression and activity of GS in the ipsilateral dentate gyrus (DG) were upregulated when kindled seizures progressed to stage 4. A single dose of L-methionine sulfoximine (MSO, in 2 µl), a selective GS inhibitor, was administered into the ipsilateral DG on the third day following the first stage 3 seizure (just before GS was upregulated). It was found that low doses of MSO (5 or 10 µg) significantly and dose-dependently reduced the severity of and susceptibility to evoked seizures, whereas MSO at a high dose (20 µg) aggravated kindled seizures. In animals that seizure acquisition had been successfully suppressed with 10 µg MSO, GS upregulation reoccurred when seizures re-progressed to stage 4 and re-administration of 10 µg MSO consistently reduced the seizures. GLN at a dose of 1.5 µg abolished the alleviative effect of 10 µg MSO and deleterious effect of 20 µg MSO on kindled seizures. Moreover, appropriate artificial microRNA interference (1 and 1.5×10⁶ TU/2 µl) of GS expression in the ipsilateral DG also inhibited seizure progression. In addition, a transient increase of GS expression and activity in the cortex was also observed during epileptogenesis evoked by pentylenetetrazole kindling. These results strongly suggest that a transient and region-specific upregulation of GS function occurs when epilepsy develops into a certain stage and eventually promotes the process of epileptogenesis. Inhibition of GS to an adequate degree and at an appropriate timing may be a potential therapeutic approach to interrupting epileptogenesis.     

Histamine H1 receptor binding capacities in the amygdalas of the amygdaloid kindled rat

The histamine H1 receptor binding capacity of the amygdalas of amygdaloid kindled rats was studied. In the kindled nonstimulated amygdala, significant decreases in K(D) and B(max) values compared with those of control amygdala were found 1 week after the last kindled seizure. One month after the last kindled seizure, the decreased K(D) value was sustained in the kindled nonstimulated amygdala. This decreased Bmax value 1 week after the last kindled seizure in nonstimulated amygdala may partly and transiently contribute to kindled seizure susceptibility. The decreased K(D) value in nonstimulated amygdala observed until 1 month after the last kindled seizure indicates the long-lasting increment of binding affinity of the pyrilamine binding site of the histamine H1 receptor in the steady state of kindled seizure susceptibility.     

Autoradiographic imaging of altered synaptic alphabetagamma2 and extrasynaptic alphabeta GABAA receptors in a genetic mouse model of anxiety

To image the possible alterations in brain regional GABAA receptor subtype properties in a genetic animal model of human anxiety, mice heterozygous for the deletion of GABAA receptor gamma2 subunit (gamma2+/-) were studied using ligand autoradiographic assays on brain cryostat sections. The [35S]TBPS binding assay was designed to reveal impaired GABA and channel site coupling shown to be more prominent in recombinant alpha1/6beta3 than in alpha1/2beta3gamma2 or beta2 subunit-containing GABAA receptors expressed in HEK 293 cells. Increased GABA-insensitive [35 S]TBPS binding in the gamma2+/- mouse brains was evident in the cerebral cortex and in subcortical regions, the alterations being regionally similar to the loss of gamma2 subnunit-dependent benzodiazepine (BZ) sites as revealed by [3H]Ro 15-4513 autoradiography. As the gamma2 subunit protein is needed for synaptic clustering of GABAA receptors, these results indicate that the extrasynaptic alphabeta3 receptors can be visualized in vitro as atypical GABA-insensitive [35S]TBPS binding sites. The results suggest that GABAAergic synaptic inhibition is widely decreased in the brains of anxiety-prone gamma2+/- mice, while extrasynaptic GABAA receptors are increased. These autoradiographic imaging findings further demonstrate the need to develop GABAA receptor subtype-selective in vivo ligands to aid in assessing the contributions of various subcellular receptor populations in anxious and other patient groups.     

Changes in benzodiazepine receptors in the limbic system following kindling of the olfactory bulb in rats

Repeated electrical stimulations of the olfactory bulb led to the progressive development of a generalized epilepsy (kindling effect). One week after the last stimulation eliciting a stage 5 seizure, diazepam-(3H) binding was studied in olfactory bulb-kindled rats. Numbers of benzodiazepine receptors were increased in kindled olfactory bulb and amygdala. No significant change was observed in hippocampus. This modification could be considered as a response of the inhibitory mechanisms to repeated seizures which is insufficient to counteract the installation of the kindling effect.     

Monophosphoryl Lipid A and Pam3Cys Prevent the Increase in Seizure Susceptibility and Epileptogenesis in Rats Undergoing Traumatic Brain Injury

Five percent of all epilepsy cases are attributed to traumatic brain injury (TBI), which are known as post-traumatic epilepsy (PTE). Finding preventive strategies for PTE is valuable. Remarkable feature of TBI is activation of microglia and subsequent neuroinflammation, which provokes epileptogenesis. The toll-like receptor agonists monophosphoryl lipid A (MPL) and tri-palmitoyl-S-glyceryl-cysteine (Pam3Cys) are safe, well-tolerated and effective adjuvants existing in prophylactic human vaccines. We examined the impact of early injection of MPL and Pam3Cys to rats, on the rate of kindled seizures acquisition following TBI. Rats received a single dose (1 µg/rat) of MPL or Pam3Cys through intracerebroventricular injection. 5 days later, trauma was exerted to temporo-parietal cortex of rats by controlled cortical impact device. After 24 h, traumatic rats underwent amygdala kindling. Brain level of the inflammatory cytokine tumor necrosis factor-alpha (TNF-α) was also measured in traumatic rats by immunoblotting. Compared to non-traumatic (sham-operated) rats, traumatic rats showed three times lower seizure threshold (133?±?5 µA vs. 416.3?±?16 µA, p?<?0.001); about three times less number of stimuli to become kindled (5?±?1 vs. 14?±?2, p?<?0.01); longer duration of kindled seizure parameters including entire seizure behavior, generalized seizures, and afterdischarges (p?<?0.001); and a two times increase in the TNF-α level. MPL and Pam3Cys did not change kindling rate and the seizure parameters in sham-operated rats. The MPL- and Pam3Cys-pretreated traumatic rats displayed seizure threshold, speed of kindling, and duration of kindled seizure parameters, similar to the non-traumatic rats. Pretreatment by MPL and Pam3Cys prevented the increase in TNF-α level by trauma. Given that MPL and Pam3Cys currently have clinical use as well-tolerated vaccines with reliable safety, they have the potential to be used in prevention of PTE.     

Immunoaffinity purification of GABAA receptor alpha-subunit iso-oligomers. Demonstration of receptor populations containing alpha 1 alpha 2, alpha 1 alpha 3, and alpha 2 alpha 3 subunit pairs.

Novel methods for the isolation of gamma-aminobutyric acidA (GABAA) receptor alpha subunit iso-oligomers have been developed. Thus, populations of GABAA receptors containing the GABAA receptor alpha 1 subunit, the alpha 2 subunit, and the alpha 3 subunit have been purified from sodium deoxycholate extracts of bovine cerebral cortex with the retention of specific [3H]flunitrazepam-binding activity by anti-alpha 1 324-341, anti-Cys alpha 2 414-424, or anti-Cys alpha 3 454-467 antibody affinity chromatography, respectively. The relative abundance of the different specificity alpha subunits in these preparations was compared with benzodiazepine affinity chromatography-purified GABAA receptors by immunoblotting. In each case, it was found that although the immunoreactivity with the specific alpha subunit antibody that was used for purification was enriched in immunoaffinity-purified receptors, reactivity with the other alpha subunit specificity antibodies, together with anti-gamma 2 1-14 Cys immunoreactivity was found. Immunoprecipitation of GABAA receptors purified by anti-alpha 1 324-341 antibody affinity chromatography by all three anti-alpha subunit antibodies employed, together with the use of anti-alpha 1 324-341 and anti-Cys alpha 2 414-424 antibody affinity columns in series, further substantiated the partial co-purification of the different polypeptides. These results demonstrate the copurification of the gamma 2 subunit with each population of alpha 1, alpha 2, alpha 3 subunit-enriched GABAA receptors. They also show the existence of minor populations of GABAA receptors that contain alpha 1 alpha 2, alpha 1 alpha 3, and alpha 2 alpha 3 subunit pairs within single oligomers.