[1] 13w Status Epilepticus in Mice Deficient for Succinate Semialdehyde Dehydrogenase: GABA A Receptor-Mediated Mechanisms

[1] 280w Succinic semialdehyde dehydrogenase (SSADH) deficiency, or ␥-hydroxybutyric (GHB) aciduria, is a rare autosomal recessive disorder that is characterized by a defect in the degradation of GABA. In the absence of SSADH, both GABA and GHB accumulate. As a result, SSADH deficiency in human is characterized by markedly increased levels of both GHB and GABA in brain, blood, and urine. Clinically, SSADH deficiency may present with a wide spectrum of neurological dysfunction, including language delay, ataxia, hypotonia, and mental retardation. Epilepsy occurs commonly in SSADH deficiency and is characterized by absence, myoclonic, and convulsive seizures, as well as convulsive status epilepticus. [1][2][3] One of our investigators (K.M.G) generated and characterized SSADH-deficient (SSADH Ϫ/Ϫ ) mice that show markedly increased levels of both GABA and GHB in urine and homogenates of liver and brain. 4 In addition to ataxia and poor weight gain, 5 SSADH Ϫ/Ϫ mice display absence, myoclonic, and convulsive seizures and convulsive status epilepticus, a seizure phenotype similar to that in human SSADH deficiency. 3,5,6 Moreover, in the mutant animals, the occurrence of generalized convulsive seizures and status epilepticus is a developmental phenomenon that emerges late in the third postnatal week of life. 6 To understand the cellular events that underlie the developmental appearance of generalized tonic-clonic seizures and status epilepticus in the SSADH Ϫ/Ϫ mouse, we examined putative mechanisms of hyperexcitability in the SSADH Ϫ/Ϫ mouse during the critical developmental window when absence seizures evolve into generalized convulsive seizures and status epilepticus. Because the mutant animals have markedly increased levels of GABA, we focused on putative GABAergic mechanisms in the evolution from absence to convulsive seizures with a particular focus on GABA A receptor (GABA A R)-mediated function.

[1] 162w These data show a significant and progressive selective decrease in [ 35 S]TBPS binding in SSADH Ϫ/Ϫ mice that is age dependent. The decrease in [ 35 S]TBPS binding progressed from P7 to P19 where a nadir was reached that was coincident with the onset of generalized convulsive seizures in the SSADH Ϫ/Ϫ mice in which spontaneous, recurrent, generalized absence sei-zures emerged reliably at P14. 6 The decrease in [ 35 S]T-BPS binding was associated with a smaller decrease in the expression of GABA B R ␤ 2 , as well as decreased GABA A R-mediated inhibitory postsynaptic potentials and increased excitability in hippocampal of the mutant mice. These data are significant, but only correlative, because they do not address directly the issue of a cause-effect relation between the decrease in GABA A Rmediated transmission during the first 3 weeks of life in the SSADH Ϫ/Ϫ mice and the progressive evolution of absence seizures into severe and ultimately fatal generalized convulsive seizures.

[2] 246w The mechanism by which absence seizures in juvenile absence epilepsy transition to generalized convulsive seizures in children is completely unknown, because until now, with the advent of the SSADH Ϫ/Ϫ Fig 5 . Comparison of intrinsic membrane properties in hippocampal neurons of succinic semialdehyde dehydrogenase null (SSADH Ϫ/Ϫ ) and SSADH ϩ/ϩ mice at postnatal day 8 (P8) and P14. Although there was no difference in input resistance and resting membrane potential recorded from P8 animals (A), a significant reduction in input resistance and resting membrane potential was observed in SSADH Ϫ/Ϫ mice after P14 (B). Whole-cell recordings show the voltage deflections, in two pyramidal cells, for the same hyperpolarizing current pulses (C). These responses were used to estimate the input resistances of the cells. mouse model, 6 there has been no animal model of generalized absence seizures that reflect this transition. In this regard, the SSADH Ϫ/Ϫ mouse model is quite different from pharmacological and genetic models of absence seizures, where absence seizures do not progress to generalized convulsive seizures. 17 The SSADH Ϫ/Ϫ mouse model differs also in the involvement of structures beyond the thalamocortical circuitry, because the CA1 region of the hippocampus, a region not involved in rodent models of typical absence seizures, 17 showed decreased GABA A R-mediated function in these experiments. In fact, the involvement of limbic, as well as thalamocortical, circuitry in SSADH Ϫ/Ϫ mice may help to explain why seizure course in these animals progresses to generalized convulsive seizures.

[3] 180w A unitary hypothesis that would take into account the currently available data to explain the absence and generalized convulsive seizures in the SSADH Ϫ/Ϫ mouse is that the absence seizures in the SSADH Ϫ/Ϫ mouse that appear in the second week of life result from markedly increased levels of GHB in the brains of the mutant animals. GHB is well known to induce this kind of seizure in experimental animals, 21 and the developmental profile fits that of GHB-induced absence seizures. 22 However, GHB does not induce generalized convulsive seizures in rodents 21 ; therefore, the generalized convulsive seizures and status epilepticus in the SSADH Ϫ/Ϫ mouse could arise from decreased GABA A R-mediated inhibition induced by usedependent downregulation of GABA receptors secondary to the markedly increased levels of GABA in the brains of SSADH Ϫ/Ϫ mice. The resultant progressive decrease in GABA A R-mediated inhibition in SSADH brain would initially be heralded by an increase in excitability, as shown in the hippocampal slice data, and would culminate in the onset of generalized convulsive seizures 23 later in postnatal life.

[4] 479w Prolonged occupancy of GABA A Rs by ligands, including GABA, sets in motion a series of mechanisms that can be termed use-dependent regulation. 24 The SSADH Ϫ/Ϫ mice have inordinately increased levels of GABA that could lead to prolonged occupancy of GABA A R, and hence use-dependent downregulation. Downregulation of GABA A R has been reported in cell culture over a period of hours, 25 days, 26 -28 and weeks, 29 as well as in response to systemic administration of progesterone, 30 a neurosteroid allosteric modulator of GABA A R, and alcohol, 31,32 a well-known GABA A R receptor agonist. 33 Chronically increased GABA levels also can downregulate phasic GABA release and reduce presynaptic signaling via GABA B R, another putative mechanism in the SSADH Ϫ/Ϫ mouse. 34 Binding to the GABA A R ion channel site complex is subunit dependent. 19 TBPS is more specific for the channel itself 35,36 than is muscimol or flunitrazepam. Therefore, in view of the observed dynamic changes in TBPS binding, our data may reflect a gradual developmental reduction in GABA A R-gated Cl Ϫ function in the SSADH Ϫ/Ϫ mouse. There is also ample precedent for an isolated change in TBPS binding with alterations of ␤ 2/3 , 37,38 and ␤ subunits of the GABA A R appear to be involved in [ 35 S]TBPS binding. 18,38 -41 Isolated changes in the steroid modulation of TBPS binding also have been demonstrated to be associated with GHB-induced absence seizures. 42 Conversely, it is possible to get selected alterations in muscimol and flunitrazepam binding with no alterations in TBPS binding with other pharmacological manipulations. 43 These [ 35 S]TBPS binding data raised the possibility of a corresponding decrease in expression of ␤ subunits of the GABA A R in SSADH Ϫ/Ϫ brains, and that is what was observed. Indeed, the significant decrease of ␤ 2 GABA A R subunit expression in SSADH Ϫ/Ϫ mice suggests that the progressive decrease in [ 35 S]TBPS binding in SSADH Ϫ/Ϫ mice observed from early in development until the onset of generalized convulsive seizures in the third postnatal week of life may be a consequence of decreased GABA A R ␤ 2 protein expression. The absence of a change in GABA A R ␥ 2 would suggest that in this mutant animal, the selective decrease of ␤ 2 may have triggered the dysfunction of GABA A R-mediated inhibition, as demonstrated electrophysiologically. However, the discrepancy between the magnitude of the decrease of TBPS binding and that of the downregulation of GABA A R ␤ 2 suggests that other subunit(s) of GABA A R also may be involved in the decrease of TBPS binding; therefore, thorough screening of the expression of all subunits of GABA A R in SSADH Ϫ/Ϫ mice may be helpful in understanding the mechanism of the decrease of TBPS binding in the mutant animal with development.

[5] 269w It is interesting to compare the phenotype of the SSADH Ϫ/Ϫ mouse with the perturbations of GABA A R-mediated inhibition reported herein with that of the GABA A R ␤ 3 null mouse. Whereas the latter is characterized by absence-like seizures and hyperactivity, 44 the developmental progression from absence seizures to lethal status epilepticus 6 that occurs in the SSADH Ϫ/Ϫ mouse is not observed in the ␤ 3 null mouse, suggesting a more generalized deficit of GABA A R-mediated inhibition in SSADH Ϫ/Ϫ . At a cellular level, GABA A R function is controlled by receptor synthesis, assembly, clustering, and cell-surface expression. 45 The difference between GABA A R protein expression and GABA A R binding observed in the SSADH Ϫ/Ϫ mouse raises the possibility that the mechanism for the observed decreased GABA A Rmediated activity in SSADH and resultant seizures could be posttranslational. One potential posttranslational mechanism in this regard is decreased cellsurface expression of GABA A R in SSADH Ϫ/Ϫ due to increased receptor endocytosis in the presence of increased ambient levels of GABA in mutant mouse brain. A number of studies support the idea that postsynaptic GABA A Rs cycle between synaptic sites and intracellular compartments, 46 and that heterodimerization protein-protein interactions between GABA A R and other nonionotropic receptors can influence the endocytosis of GABA A R with a resultant profound influence on GABA A R-mediated inhibition. 47 Thus, any mechanism that regulates the rate of endocytosis of GABA A R is predicted to have profound effects on neuronal excitability 48 and could well play a role in epileptogenesis in the SSADH Ϫ/Ϫ mouse.

[1] 57w LIGANDS. [ 35 S]Tert-butylbicyclophosphorothionate ([ 35 S]T-BPS; specific activity of 104Ci/mmol), a GABA A R antagonist, [ 3 H]muscimol (specific activity of 20Ci/mmol), a GABA A R agonist, and [ 3 H]flunitrazepam (specific activity of 85Ci/ mmol), an agonist at the benzodiazepine site of the GABA A R were purchased from Du-Pont, New England Nuclear (Boston, MA).

[2] 190w [ 35 S]TERT-BUTYLBICYCLOPHOSPHOROTHIONATE AUTORA-DIOGRAPHY. For [ 35 S]TBPS autoradiography and all autoradiographic binding experiments described later, SSADH Ϫ/Ϫ and age-matched wild-type mice were killed by decapitation; the brains were removed and immediately immersed in isopentane at Ϫ35°C. Coronal sections were cut from the anterior to posterior boundaries of the cerebral cortex at 20m at Ϫ20°C and thaw-mounted onto gelatin-coated slides that were dried and stored at Ϫ80°C until used. 7 [ 35 S]TBPS binding was performed by a modification of the method of Edgar and Schwartz. 8 The regions analyzed were the frontoparietal cortex, the ventrobasal thalamus, and the CA1 region of the hippocampus. The precise location of these regions was as described previously. 7 [ 35 S]TERT-BUTYLBICYCLOPHOSPHOROTHIONATE BINDING KINETICS. The kinetics of [ 35 S]TBPS binding to synaptic membranes was determined in SSADH Ϫ/Ϫ and age-matched wild-type mice as Cross and colleagues 9 described previously. Nonspecific binding was determined in the presence of 100M picrotoxin and represented less than 10% of the total binding at concentrations near the apparent dissociation constant (K d ) for [ 35 S]TBPS binding. Protein quantification was determined by BCA assay (Pierce, Rockford, IL).

[3] 105w [ 3 H]MUSCIMOL AUTORADIOGRAPHY. [ 3 H]Muscimol binding was performed on SSADH Ϫ/Ϫ and age-matched wild-type control mice as Titulaer and colleagues 10 described, but with modifications. In brief, tissue slices were prewashed in 50mM tris(hydroxymethyl)aminomethane (Tris)-citrate buffer (pH 7.1), 150mM NaCl, for 30 minutes at 4°C and air-dried. Sections were incubated in 10nM [ 3 H]muscimol in 50mM Triscitrate buffer for 40 minutes at 4°C, rinsed 3 times in cold 50mM Tris-citrate buffer for 5 minutes each, dipped once in ice-cold distilled water, and air-dried. Nonspecific binding was determined in the presence of 100M muscimol and represented less than 10% of the total binding.

[4] 217w [ 3 H]FLUNITRAZEPAM AUTORADIOGRAPHY. [ 3 H]Flunitrazepam binding was performed on SSADH Ϫ/Ϫ and agematched wild-type control mice by a modification of the method of Carlson and colleagues. 11 A total of 10M flunitrazepam was used to determine nonspecific binding, which represented less than 10% of the total binding. ANALYSIS OF BINDING. Five mutant and five wild-type control animals were used in the autoradiographic studies, and each experiment was done in triplicate. The autoradiographic data were analyzed as reported elsewhere. 12 Dried tissue sections were opposed to hyperfilm-bmax film (Amersham, Arlington, IL) with [ 3 H] microscale standards (Amersham) for 2 to 3 weeks at room temperature. The films were developed in D-19 (Kodak, Rochester, NY), fixed, and air-dried. Quantitative analysis of the resulting autoradiograms was performed densitometrically using a microcomputer-based densitometer system (MCID; Imaging Research; Ontario, Canada). In brief, a standard curve between the optical density of [ 3 H] standards and tissue radioactivity equivalents (pmol/mg of tissue) was constructed using a nonlinear regression analysis. The average optical density values of the selected brain regions were in the linear portion of this standard curve. The value (measured in pmol/ mg) in each brain region was calculated by interpolation using the image analyzer. 13 Five to eight readings were determined and averaged for each anatomic area analyzed.

[5] 151w IMMUNOBLOTTING. SSADH Ϫ/Ϫ and age-matched SSAD-H ϩ/ϩ mice were killed by decapitation under light halothane anesthesia; brains were immediately excised, and cerebral cortex, hippocampus, and thalamus were dissected on ice-cold glass. The anatomic dissections were performed as described previously. 14,15 A variety of primary antibodies were used to probe the membranes. For GABA A R, anti-GABA A ␣ 1 (Lab Vision, Fremont, CA), anti-GABA A ␤ 2 (1:500; Novus Biologicals, Littleton, CO), and anti-GABA A ␥ 2 (1:1,000; Chemicon International, Temecula, CA) subunits (2-hour incubation at room temperature), rabbit polyclonal antibodies were used. Appropriate rabbit (Vector Laboratories, Burlingame, CA) secondary horseradish conjugated antisera were used in the secondary incubation (1 hour at room temperature), and then developed by enzyme chemiluminescence (Amersham Pharmacia). The antibody is specific against both GABA A ␤ 2 heavy (H) and light chain (L). There is no known reactivity with other GABA subunits according to the manufacturer.

[6] 138w IMMUNOHISTOCHEMISTRY. Brains were excised and chilled in isopentane (Ϫ40°C) for 1 minute and brought to Ϫ20°C for cryosectioning. Sections were fixed in ice-cold phosphate-buffered saline (PBS), 4% paraformaldehyde for 10 minutes, then rinsed in PBS and immersed for 5 minutes at room temperature in 0.3% H 2 O 2 prepared in PBS to block endogenous peroxidase. After a fast wash, sections were preincubated for 2 hours in 5% normal goat serum PBS with 0.25% Triton-X100 (Sigma Labs, St. Louis, MO) to block nonspecific reactions, and then incubated in rabbit anti-GABA A ␣ 1 (1:200; Lab Vision) or ␤ 2 antiserum (1:300; Novus Biologicals) overnight at 4°C with shaking. Secondary incubation, with horseradish peroxidase-conjugated goat anti-rabbit IgG (1:500; Vector Laboratories), was performed at room temperature for 1 hour. The immune complexes were visualized with enhanced metal diaminobenzidine tetrahydrochloride (Pierce).

[7] 140w BRAIN SLICES AND SOLUTIONS. P8-to P14-old mice (8 SSADH Ϫ/Ϫ and 10 SSADH ϩ/ϩ mice) were anesthetized with halothane and decapitated to obtain hippocampal slices. Transverse brain slices (450m) were obtained by a Vibratome (Series 1000; Technical Products International (Ellisville, MD), St. Louis, MO) and maintained in artificial cerebrospinal fluid containing 125mM NaCl, 2.5mM KCl, 1.25mM NaH 2 PO 4 , 2mM MgSO 4 , 2mM CaCl 2 , 25mM NaHCO 3 , and 10mM glucose. To investigate possible epileptiform activity (specifically primary or secondary afterdischarges, or both), we increased KCl concentration to 5mM and decreased the concentration of MgSO 4 to 0.9mM. When needed, drugs were applied by superfusion. D-2amino-5-phosphopentanoic acid (D-AP5; 20M, made of 50mM stock solution in distilled water) and 6-cyano-7nitroquinoxaline-2,3-dione (CNQX; 100M, made of 50mM stock in dimethylsulfoxide) were obtained from Tocris Cookson and diluted daily.

[8] 145w WHOLE-CELL RECORDINGS. Neuronal recordings were obtained with the use of the whole-cell configuration of the patch-clamp technique 16 from the CA1-CA2 hippocampal pyramidal neurons. Electrodes were filled with internal solution of 150mM potassium gluconate, 10mM N-2hydroxyethylpiperazine-NЈ-2-ethanesulfonic acid (HEPES), 2mM Mg-ATP, 5mM KCl, and 0.1mM EGTA; osmolarity was 275 Ϯ 5mOsm, pH 7.2 adjusted with potassium hydroxide. Electrodes had tip resistances between 5 and 6 M⍀. Neuronal responses were recorded with the use of an Axoclamp 2-A amplifier (Axon Instruments, Foster City, CA) in bridge mode. For extracellular stimulation, a bipolar stimulation electrode was placed in the Schaffer collaterals. Initial low-stimulus intensities were gradually increased to elicit stable postsynaptic responses (as measured by either extracellular local field potentials or whole-cell recorded potentials for at least 15-20 minutes) and were maintained at this level for the duration of all the experiments that required this type of stimulation.

[9] 98w EXTRACELLULAR RECORDINGS. For extracellular stimulation, current pulses (single-current stimuli 150 milliseconds at 0.66Hz) were delivered through a bipolar stimulating electrode positioned in the Schaffer collaterals. The extracellular recording electrode was filled with artificial cerebrospinal fluid and placed in the CA1 cell body layer. Electrical signals were recorded using an Axoclamp 2A amplifier (Axon Instruments). PCLAMP software (Axon Instruments) was used for analysis of membrane potential, input resistance (which was measured from the linear part of the current-voltage plot), and amplitude of the postsynaptic responses. The amplitudes of the postsynaptic responses were measured from the peak to the baseline.

[10] 522w Statistical analysis of [ 35 S]TBPS binding at different ages of SSADH Ϫ/Ϫ mice were analyzed by one-way analysis of variance followed by Newman-Keuls multiple-comparisons test. Two individual group means were compared using a twotailed, independent Student's t test. Quantification of immunoblotting was obtained by dosimeter. Each lane was loaded with 15l of samples, and the loading amount normalization was obtained by comparison with the antiactin band on the same blot. Two individual group means were compared using a two-tailed, independent Student's t test. The unpaired Student's t test was used in all of electrophysiological experiments. Significance was considered at p Ͻ 0.05; numerical values are expressed as means Ϯ standard error. The seizure phenotype of the SSADH Ϫ/Ϫ mouse has been described. 6 In brief, spontaneous, recurrent, absence-like seizures appeared during the second week of life at P14. These seizures meet all of the behavioral, electrophysiological, and electrographic criteria for absence seizues. 17 The absence seizures persisted until around P18 to P20, when they were superseded by the occurrence of myoclonic seizures that evolved rapidly into generalized convulsive seizures. Status epilepticus then rapidly emerged and often was lethal by P23. Three specific radiolabeled ligands, [ 35 S]TBPS, [ 3 H]muscimol, and [ 3 H]flunitrazepam, were first used to assess the binding properties of GABA A R in freshfrozen brain sections at P7, before the onset of absence seizures, and from P14 to P19, during the evolution of absence to generalized convulsive seizures. We found that [ 35 S]TBPS binding in SSADH Ϫ/Ϫ mice was significantly reduced at P7 and P14 when compared with wild-type mice and became progressively more diminished until the third postnatal week of life just before the onset of generalized convulsive seizures (Fig 1 ). A significant ( p Ͻ 0.01; n ϭ 5) reduction in [ 35 S]TBPS binding was observed in SSADH Ϫ/Ϫ mice before the developmental onset of generalized convulsive seizures (see Figs 1A, B). The decrease in [ 35 S]TBPS was observed throughout all brain regions examined (see Fig 1C ) and was progressive over the first 3 weeks of life (see Fig 1D). One-way analysis of variance showed there was a significant difference of [ 35 S]TBPS binding when mice of different ages were compared, with the older mice having significantly lower [ 35 S]TBPS binding. P14 showed significantly lower [ 35 S]TBPS binding than P7 ( p Ͻ 0.001; n ϭ 5), and P19 showed significantly lower binding than P14 ( p Ͻ 0.001; n ϭ 5). Linear regression analysis also showed that the decrease in [ 35 S]TBPS binding in SSADH Ϫ/Ϫ mice was significantly age dependent in all three regions measured ( p Ͻ 0.001; n ϭ 5). Scatchard analysis of [ 35 S]TBPS binding at P18, before onset of seizures, showed a significant ( p Ͻ 0.01; n ϭ 3) decrease in B max in SSADH Ϫ/Ϫ (720.3 Ϯ 127.2fmol/mg) versus SSADH ϩ/ϩ mice (2,051.0 Ϯ 170.5fmol/mg). There was no significant difference between SSADH Ϫ/Ϫ and SSADH ϩ/ϩ mice in K d of Wu et al: Receptors and Seizures 45 during the developmental time period under examination (Figs 2A-D).

[11] 107w mice when compared with SSADH ϩ/ϩ mice at postnatal day 17 (P17) to P19 (D). Fifteen-microliter protein from brain were resolved by sodium dodecyl sulfate polyacrylamide gel electrophoresis and transferred to nitrocellulose as described. Values represent percentages (means Ϯ standard deviation; n ϭ 3) of the band intensity of GABA A R ␤ 2 in SSADH ϩ/ϩ mice. Student's unpaired t test showed that significant decrease of GABA A R ␤ 2 in SSADH Ϫ/Ϫ in comparison with control value (*p Ͻ 0.05; **p Ͻ 0.01). There was no significant change of expression of GABA A R ␣ 1 and GABA A R ␥ 2 (B, C).

[12] 61w whole-cell patch-clamp recordings of pyramidal hippocampal neurons in SSADH Ϫ/Ϫ and wild-type control mice. As shown in Figure 5, there was no difference in input resistance and resting membrane potential recorded at P8 (see Fig 5A ); however, by P14, a significant reduction in input resistance and depolar-ization of resting membrane potential had emerged in SSADH Ϫ/Ϫ mice (see Fig 5B).

[13] 69w Inhibitory neurotransmission was estimated from intracellular postsynaptic responses in the presence of the glutamate receptor blockers 6-cyano-7-nitroquinoxaline-2,3-dione and d-2-amino-5-phosphopentanoic acid (to block excitatory postsynaptic potentials). Extracellular stimulation of the Schaffer collaterals evoked GABA Amediated inhibitory postsynaptic potentials that were blocked by bicuculline. GABA A responses from the neurons of SSADH Ϫ/Ϫ mice were significantly lower than inhibitory responses elicited from control mice at P8 and P14 (Figs 6A, B).

[1] 99w The SSADH mouse model was generated by standard gene targeting and characterized by Hogema and colleagues. 4 Mice were genotyped by two-allele three-primer polymerase chain reaction using tail genomic DNA, as described previously. 4 Experiments were conducted comparing the knockout (SSADH Ϫ/Ϫ ) versus their littermate wild-type (SSADH ϩ/ϩ ) mice used as control at age postnatal day 7 (P7) to P19, before the onset of generalized convulsive seizures. 6 Developing SSADH Ϫ/Ϫ and SSADH ϩ/ϩ mice were used for the experiments described in the next section at the times specified during the first three postnatal weeks of life.

[2] 223w [ 35 S]TBPS binding is GABA A R subunit dependent. 18,19 Therefore, the significant decrease in [ 35 S]TBPS binding and decrease in B max for [ 35 S]TBPS binding observed in SSADH Ϫ/Ϫ mice early in life raised the possibility of altered subunit expression of GABA A R. Immunohistochemical staining showed no difference between SSADH ϩ/ϩ and SSADH Ϫ/Ϫ mice in the expression pattern of GABA A ␤ 2 , a pattern of expression that was similar to that reported previously in normal rat brain. 20 However, the immunohistochemical data suggest that there was a decrease in GABA A ␤ 2 expression in SSADH Ϫ/Ϫ mice that was present throughout all brain areas (Fig 3A). Quantitative immunoblotting analysis showed a significant decrease of GABA A ␤ 2 subunits (see Figs 3C, D), but not GABA A R ␣ 1 , GABA A R ␥ 2 (Figs 3B, C), and GABA A ␤ 3 subunits (data not shown), in the neocortex, hippocampus, and cerebellum of SSADH Ϫ/Ϫ mice. The decrease of GABA A ␤ 2 in the thalamus of SSADH Ϫ/Ϫ mice did not reach significance (data not shown). The magnitude of the observed decrease in GABA A ␤ 2 in SSADH Ϫ/Ϫ mice did not appear to be as great as that observed for [ 35 S]TBPS binding in the mutant animals.

[3] 258w Electrophysiological recordings from wild-type and SSADH Ϫ/Ϫ hippocampal slices were used to examine the epileptiform activity observed in SSADH Ϫ/Ϫ mice. There were two groups of animals: at P7 to P8 and at P14 to P16. Hyperexcitability was observed with extracellular recordings only in the P14 to P16 group. Extra- cellular recordings of somatic field potentials in CA1 pyramidal layers showed spontaneous local field potentials synchronous events in 4 of 6 (67%) hippocampal slices from SSADH Ϫ/Ϫ (n ϭ 6 mice) and 0 of 6 slices from SSADH ϩ/ϩ mice at P13 to P14 days of age (Fig 4A). In addition, primary afterdischarges were observed in response to 15 to 100Hz train in 5 of 6 (83%) hippocampal slices from SSADH Ϫ/Ϫ mice (n ϭ 7 mice), with a mean afterdischarge duration of 16.1 Ϯ 5.5 seconds, but no afterdischarges were found in wild-type mice hippocampal slices (0/6). Input-output curves, calculated from the evoked postsynaptic responses of CA1 pyramidal layers (the amplitudes of the peak evoked responses were measured as detailed earlier in Materials and Methods), further illustrate the increased excitability in the slices from knockout animals (see Figs 4B, C). The ability of the mutant neurons to respond to electrical stimulation of Schaffer collaterals was significantly greater in SSADH Ϫ/Ϫ mice, as judged by the higher amplitudes and greater number of population spikes evoked by a single pulse stimulation to the Schaffer collaterals recorded in the SSADH Ϫ/Ϫ slices. These data indicate that hippocampal slices from SSADH Ϫ/Ϫ mice are hyperexcitable in the second week of life.

[4] 10w ROTRANSMISSION. Passive membrane properties and inhibitory neurotransmission were estimated using