MOUSE MODEL

CROSS-REFERENCE TO RELATED APPLICATION

[0001] The present application is related to and claims priority under 35 U.S. C. § 119(e) to U.S. provisional patent application Serial No. 60/658,569, filed 7 March 2005, incorporated herein by reference.

[0002] This application was made in part with Government support under Grant No. ROl- NS-32666 funded by the National Institutes of Health, Bethesda, Maryland. The federal government may have certain rights in this invention.

BACKGROUND OF THE INVENTION

[0003] The present invention relates to transgenic mice, particularly to knock-in mice, having mutations in the KCNQ2 gene or KCNQS gene. The present invention also relates to targeting vectors intended for the generation of such mice. Furthermore, the present invention relates to murine embryonic stem cells comprising said targeting vectors as well as to a screening method for the identification of compounds for the treatment of a human physiological condition, such as benign familial neonatal convulsions (BFNC), partial epilepsies, therapeutically resistant epilepsies, migraine, neuropathic pain, stroke, dementia and anxiety. [0004] The publications and other materials used herein to illuminate the background of the invention, and in particular, cases to provide additional details respecting the practice, are incorporated by reference, and for convenience are referenced in the following text by author and date and are listed alphabetically by author in the appended bibliography. [0005] Antoni described a hereditary form of neonatal seizures with good outcome for neurologic development (Antoni 1941). Rett and Teubel reported the second case of benign neonatal familial seizures (Rett and Teubel 1964). The first case reported in the English literature appeared in 1968 (Bjerre and Corelius 1968). Since 1989 benign-familial neonatal seizures have appeared in the classification of epilepsy and epilepsy syndromes. They are classified in the chapter of Idiopathic Generalized Epilepsies. [0006] In the majority (54%) of reported neonates with this disorder, seizures begin by 2 to 8 days of life and remit by 16 months (Zonana et al 1984; Miles and Holmes 1990). In several cases (31%), seizures initially appeared not after the first week of life but during the neonatal period. The remaining patients had onset of seizures before 3.5 months (Miles and Holmes 1990).

[0007] Seizures occur in full-term neonates without any known precipitating factors after a normal pregnancy and delivery (Plouin 1990). Both the physical examination and laboratory tests are normal prior to, between, and after the seizures. Psychomotor development is normal in most children, but risk of seizure disorders in their later life is increased (Zonana et al 1984). A family history of similar seizures can be traced in most children affected with benign-familial neonatal seizures. The seizures may be either focal clonic or with a generalized tonic-clonic component and are often accompanied by episodes of apnea (Commission on Classification and Terminology of the International League Against Epilepsy 1989). The seizures are brief, usually 1 to 2 minutes (Miles and Holmes 1990). The frequency of seizures may be as high as 20 to 30 episodes per day (Zonana et al 1984).

[0008] Video tape recordings with EEG have been reported in at least 6 babies (Hirsch et al 1993; Bye 1994; Plouin 1997). In most cases the seizures start with a tonic component, followed by various autonomic and motor changes, which can be unilateral or bilateral and symmetric or asymmetric. Generalized seizures have not been reported. [0009] Benign-familial neonatal seizures is rare, dominantly inherited epileptic syndrome with a penetrance as high as 85%. The disease was first mapped to chromosome 2Oq (Leppert et al 1989). Today it is known that the syndrome exhibits genetic heterogeneity and is caused by mutations in the voltage-gated potassium channel subunit genes KCNQ2 located at 20ql 3.3 (Biervert et al 1998; Singh et al 1998; Lerche et al 1999; Miraglia del Giudice et al 2000; Singh et al., 2003) and KCNQ3 on chromosome 8q24 (Charlier et al., 1998; Hirose et al., 2000; Singh et al., 2003). More recent studies have demonstrated that de novo KCNQ2 mutations occur in newborn children with benign neonatal seizures (Claes et al., 2004).

[0010] Both KCNQ2 and KCNQ3 can form homomeric potassium channels when expressed alone but also can combine to form heteromeric channels (KCNQ2/KCNQ3), leading to the expression of larger currents. It is theorized that the association of both channels is the molecular equivalent of the so-called "M current," a neuronal potassium current highly expressed in the cortex and hippocampus (Brown and Adams, 1980). This current is slowly activated by depolarization, and its slow activation is important in regulating neuron stability, functioning as a brake for repetitive action potential firing. Only a slight reduction (approximately 25%) in KCNQ2/KCNQ3 caused by loss-of-function mutations is enough to increase neuronal excitability that can produce seizures. Recently, it has been postulated that slight reduction of KCNQ channels alone cannot produce seizure activity but can facilitate it under conditions of unbalanced neurotransmission, either by increase in excitation or decrease in

inhibition (Okada et al., 2002). Thus, this unbalance could be one possible explanation as to why the neonatal period is a vulnerable time for the seizures to occur. Another possibility is the differential expression of potassium channels during different stages of maturation. [0011] The combination of molecular biology and electrophysiology has played a seminal role in understanding the gene products that comprise characterized biophysical currents. Technologies in electrophysiology and ion channel screening are useful in the identification of M-channel modulators that may be useful for treating BFNC and other more therapeutically resistant seizure disorders. Such modulators include retigabine and flupirtine that are potassium channel openers (Wu and Dworetzky, 2005; Gribkoff, 2003). In addition, KCNQ modulation has been suggested as a treatment for neuropathic pain (Wu and Dworetzky, 2005; Gribkoff, 2003 Surti & Jan, 2005). M-channel openers have shown efficacy in the cortical spreading depression model of migraine and rat models of anxiety and stroke (Surti & Jan, 2005). hi rodent behavioral tests, M-channel blockers such as linopiridine enhance cognitive performance, suggesting a possible role for such drugs in treating dementia and Alzheimer's disease (Surti & Jan, 2005).

[0012] The Sztl mouse (Otto et al., 2004) provided the first in vivo model system in which to examine the consequences of genetic M channel modification. Studies in the Sztl mouse showed for the first time that KCNQ2 mutation does in fact attenuate IK _(M ) amplitude and increase neuronal excitability in CAl neurons of the hippocampus. However, the Sztl mouse is not an ideal model of BFNC since additional gene products are affected because this mutation not only deletes the genomic DNA encoding the C-terminus of KCNQ2, but also all of Chrna4 and Arfgap-1. The contribution of the latter two genes to the decrease in seizure threshold observed in the Sztl mouse is not known at present. Nonetheless, mutations in Chrna4 have been linked to another human epilepsy, Autosomal Dominant Frontal Lobe Epilepsy (ADFLE); however, Chrna4 hemizygosity (such as that observed in the Sztl mouse) has not been well characterized. Heterozygous knockout Chrna^ ^1' mice display no observable seizure phenotype (McCoIl et al., 2003) and display modestly increased sensitivity to the proconvulsant pentylenetetrazole (PTZ), kainic acid (KA), and bicuculline (BIC), and decreased sensitivity to 4-aminoρyridine (4-AP) (Ross et al., 2000; Wong et al., 2002). ARFGAP-1 is an ADP ribosylating factor that regulates the formation of coat protein 1 (COP-I) vesicles (Yang et al., 2000), but the effects of altered ARFGAP-1 function on neuroexcitability have not been examined. While the Sztl mouse provides a much-needed model of KCNQ2 gene mutation, for the reasons stated above, it is clearly not the ideal model of BFNC.

[0013] There clearly exists a need to develop better animal models of human epilepsy syndromes, including BFNC. There is also a need in the art for improved therapeutic agents for the treatment of BFNC and other partial epilepsies. Thus, there is a need in the art for mice models that can be used for screening potential drug candidates for the treatment of therapeutically resistant epilepsies.

SUMMARY OF THE INVENTION

[0014] The present invention relates to transgenic mice, particularly to knock-in mice, having mutations in the KCNQ2 gene or KCNQ3 gene. The present invention also relates to targeting vectors intended for the generation of such mice. Furthermore, the present invention relates to murine embryonic stem cells comprising said targeting vectors as well as to a screening method for the identification of compounds for the treatment of a human physiological condition, such as benign familial neonatal convulsions (BFNC), partial epilepsies, therapeutically resistant epilepsies, migraine, neuropathic pain, stroke, dementia and anxiety. [0015] Thus, in one aspect, the present invention provides a mouse with a mutation in one or both alleles of the endogenous KCNQ2 gene. The mutation in one or both of the endogenous KCNQ2 genes results in a mouse with a seizure-susceptible phenotype. The mutation is a mutation found in a human BFNC family that has been introduced into the mouse's genome by homologous recombination. The invention also provides cells derived from the mouse. [0016] In a second aspect, the present invention provides a mouse with a mutation in one or both alleles of the endogenous KCNQ3 gene. The mutation in one or both of the endogenous KCNQ3 genes results in a mouse with a seizure-susceptible phenotype. The mutation is a mutation found in a human BFNC family that has been introduced into the mouse's genome by homologous recombination. The invention also provides cells derived from the mouse. [0017] In a third aspect, the present invention provides methods for screening candidate compounds to identify drugs useful in treating therapeutically resistant epilepsies. In one embodiment, the mice of the present invention are utilized as a model to screen for drugs that elevate seizure threshold and prevent seizures. In a second embodiment drugs are screened based on their effect on the electrophysiology of CAl pyramidal neurons isolated from the mice of the present invention. Candidate substances so identified are useful for treating BFNC and are also be useful for treating seizures in individuals with other forms of generalized epilepsy and/or partial epilepsy. The candidate substances are also useful for treating neuropathic pain, stroke, dementia, migraine and anxiety.

BRIEF DESCRIPTION OF THE FIGURES

[0018] Figure IA shows wild-type construct of murine KCNQ2. B, BamHl restriction sites; 5'probe, position of probe used in Southern blots; green bars, locations of exons 2 through 7 of KCNQ2.

[0019] Figure IB shows targeting vector introduced into embryonic stem (ES) cells. The ACN cassette is cloned into intron 5 of KCNQ2 and thymidine kinase (TK) vector at 3' end of targeting vector. Within the ACN cassette the neomycin (neo) gene driven by the mouse RNA polymerase II promoter (polll) confers positive selection and TK gene confers negative selection of ES cells. The 5' arm is 5.4kB and the 3'arm is 4.6kB. *A306T mutation introduced into exon 6 of KCNQ2. During chimeric male spermatogenesis, Cre recombinase driven by the murine angiotensin-converting enzyme promoter, tACE, confers loxP-mediated excision of the ACN cassette to yield a single remaining loxP site within intron 5 and preservation of the mutation in exon 6 of KCNQ2. [0020] Figure 1C shows endogenous mutant construct following germline excision.

[0021] Figure ID shows southern blots of ES cell DNA to identify homologous recombination of ES cells. ES cell DNA restricted by BamHI is electrophoresed on .055-1% agarose gel, transferred to nylon membrane and probed. A 264bp probe located 5' of the targeting vector shows that 7.5kB is the wild-type allele and 6.9kB is the targeted allele verifying homologous recombination of the targeting vector. The same southern blot is later probed with a 632bp neomycin probe to give the expected 2.85kB band demonstrating single copy insertion of the targeting vector. 28 ES cells out of a total 71 ES cells tested were correctly targeted and 12 were positive for the point mutation in exon 6. A single clone was aggregated to blastocysts for chimeric production. [0022] Figure IE shows detection of the presence or absence of the A306T mutation in each experimental animal. Amplification of exon 6 using primers located in the flanking intronic sequence by PCR on mouse genomic DNA followed by single strand conformational polymorphism analysis (SSCP) of the PCR products on a 20% acrylamide gel electrophoresed at 4°C. +/+, homozygous wild-type, +/- heterozygous, -/-, homozygous mutant. [0023] Figure IF shows presence or absence of a single loxP within intron 5 in each experimental animal. Amplification of a single loxP site in intron 5 of KCNQ2 by PCR performed on mouse genomic DNA isolated from each experimental animal. PCR products are

electrophoresed on a 2% agarose gel. +/+, homozygous wild-type, +/- heterozygous, -/-, homozygous mutant.

[0024] Figure 2A shows wild-type construct of murine KCNQ3. B, BsaBI restriction sites; 5'probe, position of probe used in Southern blots; green bars, locations of exons 3 through 8 of KCNQ3.

[0025] Figure 2B shows targeting vector introduced into embryonic stem (ES) cells. The ACN cassette is cloned into intron 6 of KCNQ3 and thymidine kinase (TK) vector at 3' end of targeting vector. Within the ACN cassette the neomycin (neo) gene driven by the mouse RNA polymerase II promoter (polll) confers positive selection and TK gene confers negative selection of ES cells. The 5 ^r arm is 8.2kB and the 3'arm is 1.9kB. *G310V mutation introduced into exon 5 of KCNQ3. During chimeric male spermatogenesis, Cre recombinase driven by the murine angiotensin-converting enzyme promoter , tACE confers loxP-mediated excision of the ACN cassette to yield a single remaining loxP site within intron 6 and preservation of the mutation in exon 5 of KCNQ3. [0026] Figure 2C shows endogenous mutant construct following germline excision.

[0027] Figure 2D shows southern blots of ES cell DNA to identify homologous recombination of ES cells. ES cell DNA restricted by BsaBI is electrophoresed on 0.55-1% agarose gel, transferred to nylon membrane and probed. A 600bp probe located 5' of the targeting vector shows that 10.3kB is the wild-type allele and 14.OkB is the targeted allele verifying homologous recombination of the targeting vector. The same southern blot is later probed with a 632bp neomycin probe to give the expected 14.OkB band demonstrating single copy insertion of the targeting vector. 27 ES cells out of a total 82 ES cells tested were correctly targeted and 14 were positive for the point mutation in exon 5. A single clone was aggregated to blastocysts for chimeric production. [0028] Figure 2E shows Detection of the presence or absence of the G31 OV mutation in each experimental animal. Amplification of exon 5 using primers located in the flanking intronic sequence by PCR on mouse genomic DNA followed by single strand conformational polymorphism analysis (SSCP) of the PCR products on a 20% acrylamide gel electrophoresed at 4°C. +/+, homozygous wild-type, +/- heterozygous, -/-, homozygous mutant. [0029] Figure 2F shows presence or absence of a single loxP within intron 6 in each experimental animal. Amplification of a single loxP site in intron 6 of KCNQ3 by PCR performed on mouse genomic DNA isolated from each experimental animal. PCR products are

electrophoresed on a 2% agarose gel. +/+, homozygous wild-type, +/- heterozygous, -/-, homozygous mutant.

[0030] Figures 3A- 3E show that I _K(M ) amplitude and density are decreased, and deactivation is accelerated in C57/B16.129 CAl neurons carrying the Kcnq2 A306T mutation. Fig. 3A: Sample traces recorded from wild-type B6.129+/+ (WT; black), heterozygous mutant B6.129- Q2+/- (Het; darker grey), and homozygous mutant B6.129-Q2-/- (Horn; lighter grey) CAl neurons in response to the -20 to -60 mV step. I _K(M) amplitude is measured from the deactivation hump (10-20 msec after the hyperpolarizing step) to the steady-state level at the end of the trace. Fig. 3B: I _K(M) amplitude is decreased relative to wild-type in Horn CAl neurons (*, P < 0.001; all return steps). Fig. 3C: I _K(M > density is also decreased in Horn CAl neurons (*, P < 0.005; all return steps). Fig. 3D: Peak amplitudes of sample traces are normalized to illustrate differences in deactivation kinetics (τ values). Fig. 3E: I _K(M) deactication kinetics are faster in Het CAl neurons than WT at the -60 and -50 mV return step (|, P < 0.05). In Horn CAl neurons, I _K ( _M ) deactivation is faster than WT at the -60 mV step (*, P < 0.05). [0031] Figures 4A- 4D show that Spike Frequency Adaptation (SFA) is inhibited in C57/B16.129 CAl neurons carrying the Kcnq2 A306T mutation. Fig. 4A: Sample traces from WT, Het, and Horn CAl neurons illustrate the difference in spike firing patterns induced by the Kcnq2 A306T mutation. A depolarizing current of +100 pA was injected for 800 msec to elicit spike trains. From the beginning to the end of the response, the frequency of the spike response decreases (adapts). Note the difference in the degree of adaptation in the Horn trace. Fig. 4B: Merspike interval number is plotted against the normalized frequency to depict population SFA in CAl neurons of WT, Het, and Horn mice. Het CAl neurons exhibit SFA similar to that of WT CAl neurons (n.s.). The SFA of Horn CAl neurons is significantly inhibited relative to WT (*, P < 0.0001). Fig. 4C: Sample traces in reponse to a stronger +140 pA depolarizing current injectionillustrate increased spike activity and in WT, Het, and Horn CAl neurons. Fig. 4D: In response to a stronger depolarizing current, Het CAl neurons now exhibit significantly decreased SFA relative to WT (*, P < 0.05). SFA is still significantly inhibited in Horn CAl neurons, and is significantly different from Het (f, P < 0.005). [0032] Figures 5A-5G show that processes involving repolarization following a single action potential are facilitated in C57/B16.129 CAl neurons carrying the Kcnq2 A306T mutation. Fig. 5A: Sample traces from WT (black), Het (darker grey), and Horn (lighter grey) CAl neurons illustrate several properties of single action potentials in response to a brief (5 msec) depolarizing current injection. Fig. 5B: Action potential 10-90% rise times are no different

between WT, Het, and Horn. Fig. 5C: 90-10% decay time was not significantly altered in Het or Horn CAl neurons. Fig. 5D: The integrated area of the fast afterdepolarization (fADP) phase of the response was decreased in Horn CAl neurons only (*, P < 0.05). Fig. 5E: fADP deactivation kinetics were accelerated in both Het (f, P < 0.05) and Horn (*, P < 0.05) CAl neurons. Fig. 5F: A slow depolarizing current ramp was applied to determine the threshold for action potential generation (illustrated by the dotted line). Fig. 5G: Action potential threshold was determined to no different in Het or Horn CAl neurons.

[0033] Figures 6A-6E show that I _K(M) amplitude and density are decreased, and deactivation is accelerated in C57/B16.129 CAl neurons carrying the Kcnq3 G310V mutation. Fig. 6A: Sample traces recorded from wild-type B6.129+/+ (WT; black), heterozygous mutant B6.129- Q3+/- (Het; darker grey), and homozygous mutant B6.129-Q3-/- (Horn; lighter grey) CAl neurons in response to the -20 to -60 mV step. I _K(M > amplitude is measured from the deactivation hump (10-20 msec after the hyperpolarizing step) to the steady-state level at the end of the trace. Fig 6B: I _K(M > amplitude is decreased relative to wild-type in Horn CAl neurons (*P < 0.005; all return steps). Fig. 6C: I _K(M > density is also decreased in Horn CAl neurons (*P < 0.005; all return steps). At the -50 mV return step, I _R ( _M) density is decreased in Het CAl neurons relative to WT (f , P < 0.05). Fig. 6D: Peak amplitudes of sample traces are normalized to illustrate differences in deactivation kinetics (τ values). Fig. 6E: I _R(M) deactivation kinetics are faster in Het CAl neurons than WT only at the -60 mV return step (j, P < 0.01). In Horn CAl neurons, I _K(M) deactivation is faster than WT only at the -70 mV step (*, P < 0.01).

[0034] Figures 7A-7D show that Spike Frequency Adaptation (SFA) is inhibited in C57/B16.129 CAl neurons carrying the Kcnq3 G310V mutation. Fig. 7A: Sample traces from WT, Het, and Horn CAl neurons illustrate the difference in spike firing patterns induced by the Kcnq3 G310V mutation. A depolarizing current of +100 pA was injected for 800 msec to elicit spike trains. From the beginning to the end of the response, the frequency of the spike response decreases (adapts). Note the decreased adaptation in both the Het and Horn traces. Fig. 7B: Het CAl neurons exhibit less SFA than WT CAl neurons (j, P < 0.0001). The SFA of Horn CAl neurons is also significantly inhibited relative to WT (*, P < 0.0001). Fig. 7C: Sample traces in reponse to a stronger +140 pA depolarizing current injection illustrate increased spike activity and in WT, Het, and Horn CAl neurons. Fig. 7D: Het and Horn CAl neurons exhibit decreased SFA relative to WT (f*, P < 0.0001). At this higher depolarizing current injection, SFA in Horn CAl neurons is significantly decreased relative to Het as well (J, P < 0.05).

[0035] Figures 8A-8G show that processes involving repolarization following a single action potential are facilitated in C57/B16.129 CAl neurons carrying the Kcnq3 G310V mutation. Fig. 8A: Sample traces from WT (black), Het (darker grey), and Horn (lighter grey) CAl neurons illustrate several properties of single action potentials in response to a brief (5 msec) depolarizing current injection. Fig. 8B: Action potential 10-90% rise times are no different between WT, Het, and Horn. Fig. 8C: 90-10% decay time was decreased relative to WT in both Het (t, P < 0.05) and Horn (*, P < 0.05) CAl neurons. Fig. 8D: The integrated area of the fast afterdepolarization (fADP) phase of the response was decreased in Horn CAl neurons only (*, P < 0.05). Fig. 8E: fADP deactivation kinetics were accelerated in both Het (f , P < 0.05) and Horn (*, P < 0.05) CAl neurons. Fig. 8F: A slow depolarizing current ramp was applied to determine the threshold for action potential generation (illustrated by the dotted line). Fig. 8 G: Action potential threshold was determined to no different in Het or Horn CAl neurons. [0036] Figures 9A-9E show that Iκ _(M) amplitude and density are decreased, and deactivation is accelerated in FVB/N.129 CAl neurons carrying the Kcnq3 G310V mutation. Fig. 9 A: Sample traces recorded from wild-type FVBN.129+/+ (WT; black), heterozygous mutant FVBN.129-Q2+/- (Het; darker grey), and homozygous mutant FVBN.129-Q2-/- (Horn; lighter grey) CAl neurons in response to the -20 to -60 mV step. Iκ(M) amplitude is measured from the deactivation hump (10-20 msec after the hyperpolarizing step) to the steady-state level at the end of the trace. Fig. 9B: At the -60, -50, and -40 mV return steps, I _K(M) amplitude is decreased relative to wild-type in Horn CAl neurons (*P < 0.005). Fig. 9C: I _K(M) density is decreased in Horn CAl neurons at all return steps (*P < 0.05 for -70 mV step; P < 0.01 for -60, -50, -40 mV steps). Fig. 9D: Peak amplitudes of sample traces are normalized to illustrate differences in deactivation kinetics (τ values). Fig. 9E: I _K(M) deactication kinetics are faster in Horn CAl neurons than WT at the -60 and -50 mV return step (*, P < 0.05). In Het CAl neurons, I _K(M) deactivation is similar to WT.

[0037] Figures 10A-10D show that Spike Frequency Adaptation (SFA) is inhibited in FVB/N.129 CAl neurons carrying the Kcnq3 G310V mutation. Fig. 1OA: Sample traces from WT, Het, and Horn CAl neurons illustrate the difference in spike firing patterns induced by the Kcnq3 G310V mutation. A depolarizing current of +100 pA was injected for 800 msec to elicit spike trains. From the beginning to the end of the response, the frequency of the spike response decreases (adapts). Note the difference in the degree of adaptation in the Horn trace. Fig. 1OB: Interspike interval number is plotted against the normalized frequency to depict population SFA in CAl neurons of WT, Het, and Horn mice. Het CAl neurons exhibit SFA similar to that of

WT CAl neurons (n.s.). The SFA of Horn CAl neurons is significantly inhibited relative to WT (*, P < 0.05). Fig. 1OC: Sample traces in reponse to a stronger +140 pA depolarizing current injection illustrate increased spike activity and in WT, Het, and Hom CAl neurons. Fig. 10D: In response to a stronger depolarizing current, Het CAl neurons still exhibit similar SFA to WT (n.s.). SFA is still significantly inhibited in Hom CAl neurons (*, P < 0.001).

[0038] Figures 1 IA-I ID show that processes involving repolarization following a single action potential are facilitated in FVB/N.129 CAl neurons carrying the Kcnq3 G310V mutation. Fig. HA: Sample traces from WT (black), Het (darker grey), and Hom (lighter grey) CAl neurons illustrate several properties of single action potentials in response to a brief (5 msec) depolarizing current injection. Fig. HB: Action potential 10-90% rise times are no different between WT, Het, and Hom. Fig. 11C: 90-10% decay time is not significantly affected in either Het or Hom CAl neurons. Fig. 11D: The integrated area of the fast afterdepolarization (fADP) phase of the response was not significantly affected in Het or Hom CAl neurons. Fig. 1 IE: fADP deactivation kinetics show a trend toward significant slowing (P < 0.053) in Het CAl neurons. Fig. HF: A slow depolarizing current ramp was applied to determine the threshold for action potential generation (illustrated by the dotted line). Fig. HG: Action potential threshold was determined to no different in Het or Hom CAl neurons.

[0039] Figure 12 shows video monitoring of mice. Legend to figure 12: For various periods of time, 17 N5F2 Q3-FVB mice were video-monitored to assess the seizure severity and frequency. The X axis goes from 20 to 81 days of age. In each animal, monitoring was done 1-4 days before the first seizure and after the last seizure. Black circles represent generalized tonic clonic seizures, gray circles respresent forelimb and hindlimb clonic seizures and white circles represent forelimb clonic seizures. [0040] Figure 13 show upregulation of NPY in the mossy fibers of the dentate granule cells following seizures in N5F2 Q3-FVB homozygous mutant (-/-) mice compared to wild-type (+/+) mice.

[0041] Figure 14 shows reactive gliosis measured with an antibody to GFAP in the hippocampus of Q3-FVB homozygous mutant mice following multiple spontaneous recurrent seizures.

DETAILED DESCRIPTION OF THE INVENTION

[0042] The present invention relates to transgenic mice, particularly to knock-in mice, having mutations in the KCNQ2 gene or KCNQ3 gene. The present invention also relates to targeting

vectors intended for the generation of such mice. Furthermore, the present invention relates to murine embryonic stem cells and cryopreserved sperm comprising said targeting vectors as well as to a screening method for the identification of compounds for the treatment of a human physiological condition, such as benign familial neonatal convulsions (BFNC), partial epilepsies, therapeutically resistant epilepsies, migraine, neuropathic pain, stroke, dementia and anxiety. [0043] Thus, in one aspect, the present invention provides a mouse with a mutation in one or both alleles of the endogenous KCNQ2 gene. The mutation in one or both of the endogenous KCNQ2 genes results in a mouse with a seizure-susceptible phenotype. The mutation is a mutation found in human BFNC families that has been introduced into the mouse's genome by homologous recombination. The invention also provides cells, including sperm derived from the mouse.

[0044] In one embodiment, the present invention provides the knock-in mouse strain that carries a BFNC-causing missense point mutation - C57/B16λ29-Kcnq2 ^A306τ (B6-Q2). Wild- type, hemizygous, and homozygous N1F2 mice are viable. Spontaneous seizures have been observed that resemble the epileptic seizures of human BFNC. To test the effects of this mutation on seizure susceptibility, electroconvulsive threshold testing was performed in heterozygous mutant and wild-type mice (Otto et al., 2004). The experiments presented herein were designed to test the hypothesis that mice expressing the exact Kcnq2 mutations that cause BFNC in humans will have reduced electroconvulsive seizure thresholds for forebrain, limbic, and hindbrain seizures.

[0045] As shown herein, mice heterozygous for the Kcnq2 A306T amino acid exchange mutation exhibits significantly decreased seizure thresholds in several electroconvulsive threshold (ECT) stimulus paradigms. These differences are dependent on the specific mutation expressed, seizure type elicited, background strain, as well as gender. Studies conducted in Xenopus oocytes suggest that BFNC causing mutations decrease I _K(M) amplitude, and some shift the voltage dependence of M channel activation toward the depolarized range (Schroeder et al., 1998; Singh et al., 2003). The question remains: do these mutations actually alter the function of the native neuronal I _K ( _M ) ₉ thereby precipitating any discernable increase in neuroexcitability, and possibly suggesting a mechanism by which these mutations cause human BFNC? Since the hippocampus is heavily implicated in epilepsy, and Kcnq2, Kcnq3, and Kcnq5 mRNA are highly expressed in the CAl pyramidal cell layer (Shah et al., 2002), we sought to characterize I _K(M) function in CAl hippocampal neurons of mice expressing wild-type Kcnq2, as well as those heterozygous and homozygous for the point mutations mentioned above. Perforated patch

electrophysiology was performed in CAl hippocampal neurons in the acute brain slice preparation.

[0046] The additional experiments presented herein were designed to test the hypothesis that mice expressing Kcnq2 point mutation that causes BFNC in humans will exhibit altered Iκ(M) function and single-cell neuroexcitability. We conclude that CAl neurons in mice homozygous for the Kcnq2 A306T mutation has significantly decreased I _K(M) amplitude and spike frequency adaptation. These results are the first to show that the exact mutations that cause BFNC attenuate the amplitude of the native neuronal I _K(M) , and consequently increase neuronal excitability. We also detail differences in single action potential characteristics that may serve to counteract the effects of Kcnqlβ mutation. These results shed significant light on the consequences of M-channel point mutation as it relates to neuroexcitability and seizure generation, and thus accentuate I _K(M ) as a therapeutic target for the treatment of epilepsy. [0047] In a second aspect, the present invention provides a mouse with a mutation in one or both alleles of the endogenous KCNQ3 gene. The mutation in one or both of the endogenous KCNQ3 genes results in a mouse with a seizure-susceptible phenotype. The mutation is a mutation found in human BFNC families that has been introduced into the mouse's genome by homologous recombination. The invention also provides cells derived from the mouse. [0048] In one embodiment, C57/B\6.l29-Kcnq3 ^G3WV (B6-Q3). Wild-type, hemizygous, and homozygous N2 mice are viable and born within the expected Mendelian proportions in all mutations and strains. Spontaneous seizures have been observed that resemble the epileptic seizures of human BFNC. The effects of this mutation on seizure susceptibility, electroconvulsive threshold testing was performed in heterozygous mutant and wild-type mice (Otto et al., 2004). The experiments presented herein were designed to test the hypothesis that mice expressing the exact Kcnq3 mutations that cause BFNC in humans will have reduced electroconvulsive seizure thresholds for forebrain, limbic, and hindbrain seizures.

[0049] In a second embodiment, FVB/N.129-iTcng3 ^G310V (FN-Q3). Wild-type, hemizygous, and homozygous N1F2 mice are viable. Spontaneous seizures have been observed that resemble the epileptic seizures of human BFNC. The effects of this mutation on seizure susceptibility, electroconvulsive threshold testing was performed in heterozygous mutant and wild-type mice (Otto et al., 2004). The experiments presented herein were designed to test the hypothesis that mice expressing the exact Kcnq3 mutations that cause BFNC in humans will have reduced electroconvulsive seizure thresholds for forebrain, limbic, and hindbrain seizures.

[0050] As shown herein, mice heterozygous for the KcnqS G310V amino acid exchange mutation exhibits significantly decreased seizure thresholds in several electroconvulsive threshold (ECT) stimulus paradigms. These differences are dependent on the specific mutation expressed, seizure type elicited, background strain, as well as gender. Studies conducted in Xenopus oocytes suggest that BFNC causing mutations decrease I _K(M) amplitude, and some shift the voltage dependence of M channel activation toward the depolarized range (Schroeder et al., 1998; Singh et al., 2003). The question remains: do these mutations actually alter the function of the native neuronal I _K(M > thereby precipitating any discernable increase in neuroexcitability, and possibly suggesting a mechanism by which these mutations cause human BFNC? Since the hippocampus is heavily implicated in epilepsy, and Kcnq2, Kcnq3, and KcnqS mRNA are highly expressed in the CAl pyramidal cell layer (Shah et al., 2002), we sought to characterize I _K(M) function in CAl hippocampal neurons of mice expressing wild-type Kcnq3, as well as those heterozygous and homozygous for the point mutations mentioned above. Perforated patch electrophysiology was performed in CAl hippocampal neurons in the acute brain slice preparation.

[0051] The additional experiments presented herein were designed to test the hypothesis that mice expressing Kcnq3 point mutation that causes BFNC in humans will exhibit altered Iκ(M) function and single-cell neuroexcitability. We conclude that CAl neurons in mice homozygous for the Kcnq3 G310V mutation has significantly decreased I _K < _M) amplitude and spike frequency adaptation. These results are the first to show that the exact mutations that cause BFNC attenuate the amplitude of the native neuronal I _K(M) , and consequently increase neuronal excitability. We also detail differences in single action potential characteristics that may serve to counteract the effects of Kcnq2/3 mutation. These results shed significant light on the consequences of M-channel point mutation as it relates to neuroexcitability and seizure generation, and thus accentuate I _K(M) as a therapeutic target for the treatment of epilepsy.

[0052] In a further aspect, the mice of the present invention can be used in screening assays to determine whether a test agent (e.g., a drug, a chemical or a biologic) elevates seizure threshold or modulates the M-channel and thus can be used for treating or preventing seizures associated with BFNC or other neurological disorders where the M-channel function is perturbed. In addition, compounds that modulate the M-channel can be used for treating pain, anxiety. The term "modulate," "modulates" or "modulation" refers to enhancement (e.g., an increase) or inhibition (e.g., a reduction) in the specified activity. In one embodiment, the ability of a test compound or agent to modulate the activity of the M-channel is tested in the

mice of the present invention. For example, in in vivo assays such as whole animal, mice of the present invention are exposed to the test compound and the effect on the seizure phenotype or the electroconvulsion threshold is determined. In a second embodiment, the ability of a test compound or agent to modulate the activity of the M-channel is tested in neurons isolated from the mice of the present invention. For example, in in vitro assays, CAl pyramidal neurons (or brain slices) are isolated from the mice of the present invention and are exposed to the test compound. The effect on the seizure pattern or the electrophysiology of the CAl pyramidal neurons is determined. In vitro systems may also be utilized to screen for compounds that disrupt normal regulatory interactions. [0053] A variety of test compounds can be evaluated in accordance with the present invention. In certain embodiments, the compounds to be tested can be derived from libraries (i.e., are members of a library of compounds). While the use of libraries of peptides is well established in the art, new techniques have been developed which have allowed the production of mixtures of other compounds, such as benzodiazepines (Bunin and Ellman, 1992; De Witt et al., 1993), peptoids (Zuckermann, 1994), oligocarbamates (Cho et al., 1993), and hydantoins (DeWitt et al., 1993). An approach for the synthesis of molecular libraries of small organic molecules with a diversity of 104-105 as been described (Carell et al., 1994a; Carell et al., 1994b). [0054] The compounds of the present invention can be obtained using any of the numerous approaches in combinatorial library methods known in the art, including: biological libraries; spatially addressable parallel solid phase or solution phase libraries, synthetic library methods requiring deconvolution, the ^λ one-bead one-compound' library method, and synthetic library methods using affinity chromatography selection. The biological library approach is limited to peptide libraries, while the other four approaches are applicable to peptide, non-peptide oligomer or small molecule libraries of compounds (Lam, 1997). Other exemplary methods for the synthesis of molecular libraries can be found in the art, for example in Erb et al. (1994), Horwell et al. (1996) and Gallop et al. (1994).

[0055] Libraries of compounds may be presented in solution (e.g., Houghten et al., 1992), or on beads (Lam et al., 1991), chips (Fodor et al., 1993), bacteria (U.S. Pat. No. 5,223,409), spores (U.S. Pat. No. 5,223,409), plasmids (Cull et al., 1992) or on phage (Scott and Smith, 1990; Devlin et al., 1990; Cwirla et al., 1990; Felici et al., 1991). In still another embodiment, the combinatorial polypeptides are produced from a cDNA library.

[0056] Exemplary compounds which can be screened for activity include, but are not limited to, peptides, nucleic acids, carbohydrates, small organic molecules, and natural product extract libraries.

[0057] The goal of rational drug design is to produce structural analogs of biologically active polypeptides of interest or of small molecules with which they interact (e.g., agonists, antagonists, inhibitors) in order to fashion drugs which are, for example, more active or stable forms of the polypeptide, or which, e.g., enhance or interfere with the function of a polypeptide in vivo. Several approaches for use in rational drug design include analysis of three-dimensional structure, alanine scans, molecular modeling and use of anti-id antibodies. These techniques are well known to those skilled in the art. Such techniques may include providing atomic coordinates defining a three-dimensional structure of a protein complex formed by said first polypeptide and said second polypeptide, and designing or selecting compounds capable of interfering with the interaction between a first polypeptide and a second polypeptide based on said atomic coordinates. [0058] Following identification of a substance which modulates or affects polypeptide activity, the substance may be further investigated. Furthermore, it may be manufactured and/or used in preparation, i.e., manufacture or formulation, or a composition such as a medicament, pharmaceutical composition or drug. These may be administered to individuals. [0059] A substance identified as a modulator of polypeptide function may be peptide or non- peptide in nature. Non-peptide "small molecules" are often preferred for many in vivo pharmaceutical uses. Accordingly, a mimetic or mimic of the substance (particularly if a peptide) may be designed for pharmaceutical use.

[0060] The designing of mimetics to a known pharmaceutically active compound is a known approach to the development of pharmaceuticals based on a "lead" compound. This approach might be desirable where the active compound is difficult or expensive to synthesize or where it is unsuitable for a particular method of administration, e.g., pure peptides are unsuitable active agents for oral compositions as they tend to be quickly degraded by proteases in the alimentary canal. Mimetic design, synthesis and testing is generally used to avoid randomly screening large numbers of molecules for a target property. [0061] Once the pharmacophore has been found, its structure is modeled according to its physical properties, e.g., stereochemistry, bonding, size and/or charge, using data from a range of sources, e.g., spectroscopic techniques, x-ray diffraction data and NMR. Computational analysis, similarity mapping (which models the charge and/or volume of a pharmacophore,

rather than the bonding between atoms) and other techniques can be used in this modeling process. Such techniques include those disclosed in U.S. Patent No. 6,080,576. [0062] A template molecule is then selected, onto which chemical groups that mimic the pharmacophore can be grafted. The template molecule and the chemical groups grafted thereon can be conveniently selected so that the mimetic is easy to synthesize, is likely to be pharmacologically acceptable, and does not degrade in vivo, while retaining the biological activity of the lead compound. Alternatively, where the mimetic is peptide-based, further stability can be achieved by cyclizing the peptide, increasing its rigidity. The mimetic or mimetics found by this approach can then be screened to see whether they have the target property, or to what extent it is exhibited. Further optimization or modification can then be carried out to arrive at one or more final mimetics for in vivo or clinical testing. [0063] With regard to intervention, any compounds which reverse any aspect of a given phenotype or expression of any gene in vivo and which modulates protein activity or binding with binding partner in vitro should be considered as candidates for further development or potential use in humans. Dosages of test agents may be determined by deriving dose-response curves using methods well known in the art.

[0064] This invention further pertains to agents identified by the above-described screening assays and uses thereof for treating the conditions described herein. Pharmaceutical compositions containing an identified agent as the active ingredient can be prepared according to conventional pharmaceutical compounding techniques. See, for example, Remington's Pharmaceutical Sciences, 18th Ed. (1990, Mack Publishing Co., Easton, Pa.). Typically, a therapeutically effective amount of an active ingredient is admixed with a pharmaceutically acceptable carrier. By a "therapeutically effective amount" or simply "effective amount" of an active compound is meant a sufficient amount of the compound to treat the desired condition at a reasonable benefit/risk ratio applicable to any medical treatment. The actual amount administered, and the rate and time-course of administration, will depend on the nature and severity of the condition being treated. Prescription of treatment, e.g. decisions on dosage, timing, etc., is within the responsibility of general practitioners or specialists, and typically takes account of the disorder to be treated, the condition of the individual patient, the site of delivery, the method of administration and other factors known to practitioners. Examples of techniques and protocols can be found in Remington 's Pharmaceutical Sciences.

[0065] The carrier may take a wide variety of forms depending on the form of preparation desired for administration, e.g., intravenous, oral, parenteral, intramuscular, subcutaneous or

intrathecal. Some examples of the materials that can serve as pharmaceutically acceptable carriers are sugars, such as lactose, glucose and sucrose, starches such as com starch and potato starch, cellulose and its derivatives such as sodium carboxymethyl cellulose, ethyl cellulose and cellulose acetate; powdered tragacanth; malt, gelatin, talc; excipients such as cocoa butter and suppository waxes; oils such as peanut oil, cottonseed oil, safflower oil, sesame oil, olive oil, corn oil and soybean oil; glycols, such as propylene glycol, polyols such as glycerin, sorbitol, mannitol and polyethylene glycol; esters such as ethyl oleate and ethyl laurate, agar; buffering agents such as magnesium hydroxide and aluminum hydroxide; alginic acid; pyrogen-free water; isotonic saline, Ringer's solution; ethyl alcohol and phosphate buffer solutions, as well as other non-toxic compatible substances used in pharmaceutical formulations. For examples of delivery methods see U.S. Pat. No. 5,844,077, incorporated herein by reference.

[0066] Wetting agents, emulsiflers and lubricants such as sodium lauryl sulfate and magnesium stearate, as well as coloring agents, releasing agents, coating agents, sweetening, flavoring and perfuming agents, preservatives and antioxidants can also be present in the composition, according to the judgment of the formulator. Examples of pharmaceutically acceptable antioxidants include, but are not limited to, water soluble antioxidants such as ascorbic acid, cysteine hydrochloride, sodium bisulfite, sodium metabisulfite, sodium sulfite, and the like; oil soluble antioxidants, such as ascorbyl palmitate, butylated hydroxyanisole (BHA), butylated hydroxytoluene (BHT), lecithin, propyl gallate, aloha-tocopherol and the like; and the metal chelating agents such as citric acid, ethylenediamine tetraacetic acid (EDTA), sorbitol, tartaric acid, phosphoric acid and the like.

[0067] Exemplary methods for administering compounds (e.g., so as to achieve sterile or aseptic conditions) will be apparent to the skilled artisan. Certain methods suitable for administering compounds useful according to the present invention are set forth in Goodman and Gilman's The Pharmacological Basis of Therapeutics, 7th Ed. (1985). The administration to the patient can be intermittent; or at a gradual, continuous, constant or controlled rate. Administration can be to a warm-blooded animal (e.g. a mammal, such as a mouse, rat, cat, rabbit, dog, pig, cow or monkey); but advantageously is administered to a human being. Administration occurs after general anesthesia is administered. The frequency of administration normally is determined by an anesthesiologist, and typically varies from patient to patient.

[0068] The pharmaceutical compositions will generally contain from about 0.0001 to 99 wt. %, preferably about 0.001 to 50 wt. %, more preferably about 0.01 to 10 wt. % of the active ingredient by weight of the total composition. In addition to the active agent, the pharmaceutical

compositions and medicaments can also contain other pharmaceutically active compounds. Examples of other pharmaceutically active compounds include, but are not limited to, analgesic agents, cytokines and therapeutic agents in all of the major areas of clinical medicine. When used with other pharmaceutically active compounds, the agents may be delivered in the form of drug cocktails. A cocktail is a mixture of any one of the compounds useful with this invention with another drug or agent. In this embodiment, a common administration vehicle (e.g., pill, tablet, implant, pump, injectable solution, etc.) would contain both the instant composition in combination supplementary potentiating agent. The individual drugs of the cocktail are each administered in therapeutically effective amounts. A therapeutically effective amount will be determined by the parameters described above; but, in any event, is that amount which establishes a level of the drugs in the area of body where the drugs are required for a period of time which is effective in attaining the desired effects.

[0069] The practice of the present invention employs, unless otherwise indicated, conventional techniques of chemistry, molecular biology, microbiology, recombinant DNA, genetics, immunology, cell biology, cell culture and transgenic biology, which are within the skill of the art. See, e.g., Maniatis et al, 1982, Molecular Cloning (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York); Sambrook et al., 1989, Molecular Cloning, 2nd Ed. (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York); Sambrook and Russell, 2001, Molecular Cloning, 3rd Ed. (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York); Ausubel et ah, 1992), Current Protocols in Molecular Biology (John Wiley & Sons, including periodic updates); Glover, 1985, DNA Cloning (IRL Press, Oxford); Anand, 1992; Guthrie and Fink, 1991; Harlow and Lane, 1988, Antibodies, (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York); Jakoby and Pastan, 1979; Nucleic Acid Hybridization (B. D. Hames & S. J. Higgins eds. 1984); Transcription And Translation (B. D. Hames & S. J. Higgins eds. 1984); Culture Of Animal Cells (R. I. Freshney, Alan R. Liss, Inc., 1987); Immobilized Cells And Enzymes (IRL Press, 1986); B. Perbal, A Practical Guide To Molecular Cloning (1984); the treatise, Methods In Enzymology (Academic Press, Inc., N.Y.); Gene Transfer Vectors For Mammalian Cells (J. H. Miller and M. P. Calos eds., 1987, Cold Spring Harbor Laboratory); Methods In Enzymology, VoIs. 154 and 155 (Wu et al. eds.), Immunochemical Methods In Cell And Molecular Biology (Mayer and Walker, eds., Academic Press, London, 1987); Handbook Of Experimental Immunology, Volumes I-IV (D. M. Weir and C. C. Blackwell, eds., 1986); Riott, Essential Immunology, 6th Edition, Blackwell Scientific

Publications, Oxford, 1988; Hogan et al., Manipulating the Mouse Embryo, (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N. Y., 1986); Westerfield, M., The zebrafish book. A guide for the laboratory use of zebrafish (Danio rerio), (4th Ed., Univ. of Oregon Press, Eugene, 2000).

EXAMPLES

[0070] The present invention is described by reference to the following Examples, which are offered by way of illustration and are not intended to limit the invention in any manner. Standard techniques well known in the art or the techniques specifically described below were utilized.

EXAMPLE l

Materials and Methods: Transgenic Mice [0071] Vector Construction [0072] Wild-type clones of mouse KCNQ2 and KCNQ3 were isolated from a λ phage library (Stratagene) and subcloned into pBSk. Figure IA shows wild-type construct of murine KCNQ2. Figure 2 A shows wild-type construct of murine KCNQ3. Point mutations, A306T mutation in exon 6 of KCNQ2 and G310V mutation in exon 5 of KCNQ3, were introduced into the respective constructs using PCR mutagenesis techniques with mutations contained in the primer. The ACN (Bunting et al., 1999) cassette comprises the tACE/Cre/PolII/Neo gene was flanked with two loxP sites (Fig. IB and Fig. 2B). The ACN cassette is cloned into an Xhol site in intron 5 of KCNQ2 and this construct is cloned into a thymidine kinase (TK) vector (Thomas and Capecchi, 1987). The ACN cassette is cloned into a SnaBI site in intron 6 of KCNQ3 and this construct is cloned into TK. Within the ACN cassette the neomycin (neo) gene driven by the mouse RNA polymerase II promoter (polll) confers positive selection and TK gene confers negative selection of ES cells. Each targeting vector was linearized with Notl. During chimeric male spermatogenesis, Cre recombinase driven by the murine angiotensin-converting enzyme promoter, tACE, confers loxP-mediated excision of the ACN cassette to yield a single remaining loxP site (Fig. 1C and Fig. 2C)..

[0073] ES Cells: Transformation, Screening and Blastocyst Aggregation

[0074] Each targeting vector was introduced by electroporation into RI ES cells (Nagy et al.,

1993) and selected for resistance to G418 and FIAU. Approximately 2xlO ⁷ cells were

electroporated and 71 KCNQ2 and 82 KCNQ3 colonies were isolated. DNA from each clone was isolated and analyzed by standard Southern blotting techniques. Homologous recombination was verified using multiple restriction enzymes and probes to demonstrate the expected band shifts on Southern blot (Fig. ID and Fig. 2D). For each mouse strain (KCNQ2 ^A306TUU and KCNQ3 ^G310VUU ), ES cells from a single targeted clone were aggregated with C57Bl/6-derived morulae, and implanted into a pseudopregnant C57B1/6 female. Chimeric progeny were identified by coat color and a single male of each strain was mated to C57B1/6 or FVB/N females for the generation of Fl offspring. Fl offspring were intercrossed to generate F2 experimental animals. An additional 5 backcrosses were performed on the C57B1/6 or FVB/N backgrounds and these offspring were intercrossed to yield N5F2 experimental animals.

[0075] Mouse Genotyping

[0076] Genomic DNA obtained from tail biopsies of Fl and F2 animals were analyzed for

Cre-mediated self excision of the ACN cassette and presence or absence of the respective mutation. Exon 6 of KCNQ2 and exon 5 of KCNQ3 were amplified by PCR primers located in the respective flanking intronic sequence and PCR products were analyzed by single strand conformational polymorphism analysis (SSCP) on a 20% acrylamide gel electrophoresed at 4°C (Fig. IE and Fig. 2E). Under these conditions, the mutation produces a band shift with respect to wild-type. To verify self-excision, primers surrounding the respective loxP site of each mouse strain were used to amplify PCR products that were electrophoresed on a 2% agarose gel. The presence of a single loxP site verifies self-excision (Fig. IF and Fig. 2F).

[0077] Experimental Mice

[0078] Chimeric mice were backcrossed to C57B1/6 and FVB/N strains for the generation of Fl mice. For electrophysiology and seizure threshold experiments, Fl mice were intercrossed to produce Nl F2 progeny that were wild type (+/+), heterozygote (+/-) and homozygote (-/-) for the respective mutations. Electrophysiology experiments were performed on wild type (+/+ ^■ ), heterozygote (+/-) and homozygote (-/-) mice for the following strains: B6.129 F2 Kcnq2 ^A306T , B6.129 F2 Kcnq3 ^G3l0Y and FVB.129 F2 Kcnq3 ^G310V . Seizure threshold experiments were performed on wild-type (+/+) and heterozygote(+/-) mice for the following strains: B6.129 F2 Kcnq2 ^A306τ , B6.129 F2 Kcnq3 ^G3l0Y and FVB.129 F2 Kcnq3 ^G310V . Additional backcrossing and intercrossing were performed to produce incipient congenics, N5F2, on both C57B1/6 and

FVB/N backgrounds for ongoing histological characterization. Five additional backcrosses are performed to produce NlO mice on both C57B1/6 and FVB/N backgrounds.

EXAMPLE 2 Materials and Methods: Electroconvulsive Studies

[0079] Mice

[0080] Male and female mice, varying in age between 180 and 320 days were used for all electroconvulsive threshold experiments. There were no significant differences in age between any of the groups. F2 generation mice of the following genotypes were used: wild-type C57/B16.129 (B6) and FVB/N.129 (FN); and heterozygous knock-in mutant C57/B16.129- Zc/jg2 ^A306T/+ (B6-Q2+/-), C57/Bl6λ29-Kcnq2 ^Gmvμ' (B6-Q3+/-), and FVB/N. U9-Kcnq3 ^G3im+ (FN-Q3+/-). Animals were allowed free access to food and water and were housed in a temperature- and light-controlled (12 hrs on/12 hrs off) environment. All animal care and experimental manipulations were approved by the Institutional Animal Care and Use Committee (IUCAC) of the University of Utah, and are in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

[0081] Seizure Thresholds

[0082] Partial psychomotor seizure testing was conducted with a Grass S48 stimulator (6 Hz, 0.2 msec rectangular pulse width, 3 sec duration, varying current intensities). The phenotype for these seizures is rhythmic face movements, forelimb clonus, dorsal neck flexion, rearing and falling, and/or transient gait wobbliness/ataxia (Barton et al., 2001). Minimal clonic and minimal tonic hindlimb extension (THE) seizure testing (60 Hz, 0.2 msec sinusoidal current pulse, varying current intensities) was conducted with a stimulator previously described (Woodbury and Davenport, 1952). Minimal clonic seizures are characterized by rhythmic face and forelimb clonus, rearing and falling, and ventral neck flexion. Minimal THE seizures are characterized by a tonic-clonic flexion-extension sequence that starts with tonic forelimb extension, followed by hindlimb flexion, and terminates in full tonic hindlimb extension (180 degrees to the torso) (Toman, 1951; Peterson, 1998; White et al., 2002). AU ECT testing was conducted between the hours of 1:00 and 5:00 p.m. to avoid confounding results that might arise from differences in circadian rhythm.

[0083] Data Analysis

[0084] Seizure thresholds were tested using the staircase estimation procedure (Otto et al., 2004). Convulsive current (CC) curves were constructed and the CC value that produced a seizure in 50% of the mice tested (CC50) and 95% confidence intervals (CI ₉ 5) were calculated. Mice were separated according to gender, knock-in mutation, and background strain, and CC curves were constructed for three different seizure phenotypes: partial psychomotor, minimal clonic, and minimal tonic hindlimb extension (THE). Statistical comparisons between WT and mutant mice were made by Probit analysis (Minitab 13). Significance was determined at P < 0.05.

EXAMPLE 3

Effects of Kcnq2 and Kcnq3 Point Mutations on Seizure Threshold

[0085] To examine the effects of the Kcnql gene A306T mutation on the B6 mouse background, and the Kcnq3 gene G310V mutation on both B6 and FN mouse backgrounds, electroconvulsive threshold (ECT) testing was performed in mutant and littermate control mice (Otto et al., 2004). The partial psychomotor, minimal clonic, and minimal THE seizure CC curves were prepared for the mutant and wildtype mice as an example comparison. In the comparison of the partial psychomotor, minimal clonic, and minimal THE seizure CC curves for female mutant B6-Q2+/- and WT B6, the shift in seizure threshold was determined to be significant in all three seizure testing paradigms. CC values at which 50% of mice were predicted to seize (CC ₅ o) and corresponding 95% confidence intervals (CI ₉₅ ) were also calculated. Values for male and female B6-Q2+/-, B6-Q3+/-, and FN-Q3+/- heterozygous mice and their littermate controls are summarized in Tables 1-3.

TABLE l

Effects of Kcnq2 A306T and Kcnq3 G310V Mutations on Partial Psychomotor (6 Hz) Seizure Threshold.

Strain

Genotype gender CC ₅ 0 (CI ₉₅ ) comparison

for was threshold was significantly increased compared to WT.

[0086] Table 1 summarizes the CC ₅₀ and CI ₉₅ values in response to partial psychomotor (6 Hz) ECT testing. Several strain- and gender-dependent differences in the seizure threshold altering effects of KCNQ mutation were observed. The Kcnql A306T mutation decreased the seizure threshold of both male and female B6 mice (rows 1 and 2). Interestingly, the Kcnq3 G310V mutation did not effect seizure threshold of male B6 mice, but actually increased the seizure threshold of female B6 mice (rows 3 and 4). The KcnqS G310V mutation decreased the

seizure threshold of male FN mice, and again increased the seizure threshold of female FN mice (rows 5 and 6).

TABLE 2

Effects oiKcnq2 A306T and Kcnq3 G310V Mutations on Minimal Clonic Seizure Threshold.

full CC curve comparisons. *, Heterozygous mutant mouse seizure threshold significantly reduced compared to WT.

[0087] Table 2 shows the CC ₅₀ and CI ₉₅ values in response to the minimal clonic seizure testing paradigm. Note that for this seizure type, the Kcnq2 A306T mutation in B6 mice, and Kcnq3 G310V mutation in B6 and FN mice, all resulted in a significant reduction in the seizure threshold of female mice. Conversely, none of the mutations expressed on any of the genetic background strains significantly affected the seizure threshold of male mice. The minimal

clonic seizure was the only one in which this obvious gender-dependent susceptibility to Kcnql and Kcnq3 mutation was observed.

TABLE 3

Effects of Kcnql A306T and Kcnq3 G310V Mutations on Minimal Tonic Hindlimb Extension (THE) Seizure Threshold.

Strain

Genotype gender CC50 (CI ₉₅ ) comparison

[0088] Table 3 summarizes the calculated CC ₅₀ and CI95 values for the minimal THE seizure. For this seizure type, the seizure thresholds were decreased for all mutations, backgrounds, and genders. This is the only seizure phenotype in which heterozygous mutants all displayed decreased seizure susceptibility relative to littermate control WT mice.

[0089] It is also useful to group the data for each mutation and background strain, then consider how the significance of the effect on seizure threshold changes with seizure severity and spread from partial psychomotor, to minimal clonic, to minimal THE seizure phenotypes. The results across seizure types were grouped one mutation-, gender-, and background-at-a-time. It was found that the only group in which the seizure threshold was significantly decreased for all three seizure phenotypes was the Kcnq2 A306T mutation expressed in female B6 mice. The only seizure model in which mutant mice of all sexes and background strains displayed significantly decreased seizure thresholds was the most severe seizure type; i.e., minimal THE. Also, it is particularly fascinating that the KcnqS G310V mutation actually increased the seizure threshold of female mice of both the B6 and FN background. Finally, none of the mutant male strains displayed an altered seizure threshold in response to minimal clonic testing.

[0090] It has previously been shown that the Sztl mutation, which deletes the genomic DNA encoding the C-terminus of KCNQ2 and at least two other genes, Chrna4 and Arfgap-1, significantly decreases seizure threshold in the partial psychomotor, minimal clonic, and minimal tonic hindlimb extension seizure models (Otto et al., 2004). Although we subsequently showed that this mutation decreases I _K(M) amplitude, it could by no means be assumed that the Kcnq2 component of Sztl was solely responsible for the decreased seizure threshold observed in these mice. The ultimate goal of the present study was to determine whether Kcnq2 A306T and Kcnq3 G310V point mutations that are identical to those that cause BFNC in two human families (Singh et al., 2003) autonomously decrease seizure threshold in the ECT model. The data presented here establish that Kcnq2 and Kcnq3 mutation alone can in fact significantly alter seizure threshold. In addition, and somewhat to our surprise, we found these changes in seizure susceptibility to be dependent on many factors - the type of seizure elicited, the specific amino acid mutation, the background strain of the mice, and their gender.

[0091] The results from the present study show that the Kcnq2 A306T and KcnqS G310V mutations have similar effects on seizure thresholds in the minimal clonic and minimal THE seizure models. In the partial psychomotor model, however, it appears that the Kcnq3 G310V mutation displays a less robust effect. Seizure thresholds of male B6 mice are not affected, and those of female mice are peculiarly increased. These results suggest that either the Kcnq3 G310V mutation has a less robust effect on I _K(M) function, or that compensatory mechanisms designed to counteract the effects of Kcnq mutation may be more effective in Kcnq3 G310V mutant mice.

[0092] Note that only in the minimal THE model is the seizure threshold decreased in all heterozygous mutant mice. This suggests that, for complete manifestation of a lowered seizure threshold phenotype, the cumulative effect of activating many brain regions might be required. As previously described by our group and others (Barton et al., 2001), different combinations of brain regions are activated by the three different stimulus paradigms to induce each seizure type. In brief, partial psychomotor seizures result from activation of forebrain structures; minimal clonic seizures result from activation of forebrain structures; minimal THE seizures result from activation of hindbrain structures. [0093] Thus, the differential effects of the Kcnql and Kcnq3 mutations might be due to the activation of different brain regions. Differences in the expression of KCNQ2 and KCNQ3 subunit proteins, between brain regions and in discrete regions of the cell, might also contribute to these differences. For instance, while KCNQ2 and KCNQ3 are largely coexpressed throughout the brain, on a smaller scale, KCNQ2 is often expressed by itself in more discrete areas of the cell, specifically the axon initial segment and nodes of Ranvier (Devaux et al., 2004).

[0094] The effects of Kcnq 3 G310V mutation expression in female mice are perhaps the most fascinating of this study, hi the partial psychomotor seizure model, female mice carrying this mutation (expressed on both B6 and FN backgrounds) actually exhibited increased seizure thresholds. However, the thresholds of female Kcnq3 G310V mutants from both backgrounds were decreased in the minimal clonic and minimal THE seizure models. Why this apparent reversal of effect in the partial psychomotor model? One potential explanation is that the Kcnq3 G310V mutation in females might lower seizure threshold to the point of inducing interictal burst activity indiscernible to the naked eye. If this were in fact the case, seizure thresholds might actually increase since interictal activity in the more vulnerable brain structures can offer short-term resistance to external stimulus-induced seizures, perhaps providing a preconditioning effect (Bortolotto et al., 1991; Rejdak et al., 2001). In this regard, surface electroencephalogram (EEG) analysis in female B6-Q3+/- and FN-Q3+/- mice to identify possible ictal activity in the forebrain structures, coupled with immunofluorescence studies to examine possible region- specific changes in KCNQ3 expression induced by the G310V mutation, could prove invaluable. [0095] Regarding gender, it appears that female mice are, for the most part, more significantly affected by these mutations than their male counterparts. This trend is especially evident for the minimal clonic seizure testing paradigm, where none of the seizure thresholds of male mice is affected by any mutation on any background strain, but the thresholds of all female

mice are significantly decreased by all mutations on all background strains. The minimal clonic seizure results observed in this study in particular illustrate a central tenet of epilepsy research: exposure to an identical challenge (e.g., genetic mutation) often does not result in the same pathological outcome (e.g., altered seizure threshold or epilepsy) in all subjects. Since female mice display lower seizure thresholds than males in general (Vernadakis and Woodbury, 1972; Frankel et al., 2001), and thus are considered to have increased seizure susceptibility, it is feasible that these mutations increase neuronal excitability such that a lower stimulus intensity is required to reach seizure- threshold and provoke seizure activity. [0096] It has been previously established that genetic background strain can considerably affect the seizure thresholds of mice (Ferraro et al., 1998; Frankel et al., 2001). In the present study, male mice carrying the Kcnq3 G310V mutation on the B6 background exhibited a threshold no different from their WT littermates in partial psychomotor seizure testing, but this mutation did significantly reduce the threshold of male mice carrying the same mutation on the FN background. Thus, the effect of the Kcnq3 G310V mutation appears to be dependent on the genetic background of the mouse. Previous studies support this trend in background-specific changes in seizure activity. For instance, kainic acid treatment induces severe seizures in C57/B16 mice, but only slight seizure activity in 129/SvJ mice (McKhann et al., 2003). [0097] It should be noted that these studies were conducted in mice considerably older than what are typically used for ECT studies. These mice ranged from 180 to 320 days old, whereas most studies of this nature use 8-to-12 week old (56-to-84 day old) mice (Ferraro et al., 1998; Frankel et al., 2001). It should be noted, however, that the CC _5O values of these mice were retested for each seizure type as a final experiment, and were found to be consistent with those established initially. This indicates that seizure thresholds did not shift significantly throughout the study, even after repeated stimulations; nevertheless, these studies should be confirmed in 8- to- 12 week old mice to facilitate comparisons with the previous literature.

[0098] In the present study, we offer evidence that single amino acid mutations that have been identified in human families with BFNC, when expressed in mice, reduce seizure threshold in a variety of, but not all, seizure phenotypes. These results provide evidence of increased excitability in neural networks contained within large brain regions. Similar results were obtained previously in another model of Kcnq mutation, the Sztl mouse (Otto et al., 2004).

EXAMPLE 4

Materials and Methods: CAl Hippocampal Neurons Studies [0099] Mice

[0100] Male and female mice, 8-to-15 weeks of age were used for all electrophysiology experiments. F2 generation mice of the following genotypes were used: wild-type C57/B16.129 (B6+/+), heterozygous C57/B16.129-Zαz?2 ^A306T/+ (B6-Q2+/-) and homozygous C57/B16.129- Zc«#2 ^A306T/A306T (B6-Q2-/-) knock-in; wild-type C57/B16.129 (B6+/+), heterozygous C57/B16.129-Zc^3 ^G310v/+ (B6-Q3+/-) and homozygous B6λ29-Kcnq3 ^GnwiGmv (B6-Q3-λ) knock-in; and wild-type FVB/N.129 (FN+/+), heterozygous FVB/N.129-i_cng3 ^G310V/+ (FN- Q3+/-), and homozygous FVB/N.129-^cn^3 ^G310V/G310V (FN-Q3-/-) knock-in. Animals were allowed free access to food and water and were housed in a temperature- and light-controlled (12 hrs on/12 hrs off) environment. All animal care and experimental manipulations were approved by the Institutional Animal Care and Use Committee (IUCAC) of the University of Utah, and are in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals .

[0101] Acute Brain Slice Preparation

[0102] Brain slices were prepared from wild-type, heterozygous and homozygous Kcnq2

A306T and Kcnq3 G310V mice in a manner similar to previously described methods (Barton et al., 2004). Briefly, mice were anesthetized with Nembutal (25 mg/kg), decapitated, and their brains quickly removed and placed in oxygenated Ringer solution containing (in mM): 200 sucrose, 26 NaHCO ₃ , 10 glucose, 3 KCl, 2 MgSO ₄ , 2 CaCl ₂ , 1.4 NaH ₂ PO ₄ . Brains were trimmed, mounted on a chuck, and 350 μm thick coronal slices were cut using a Vibratome sheer (Vibratome). Slices were then transferred to a holding chamber and allowed to incubate for >1 hr in oxygenated Ringer similar to the above solution, but with 126 mM NaCl in place of sucrose. The NaCl Ringer pH was maintained at 7.37-7.40 with NaOH and continuous bubbling with 95% O ₂ /5% CO ₂ , and mOsm was 305-310.

[0103] Electrophysiological Measurements [0104] Whole-cell perforated patch recordings were obtained from CAl pyramidal neurons in the acute brain slice preparation using a MultiClamp 700A amplifier (Axon Instruments). Signals in voltage-clamp and current-clamp modes were acquired at 10 KHz and 20 KHz, and filtered at 2 KHz and 10 KHz, respectively, for offline analysis using Clampfit 9. Glass

capillaries (World Precision Instruments, Inc.) were pulled to 2.0-3.2 Mω resistances using a micropipette electrode puller (Sutter Instrument Co.). Input and series resistance values of 80- 120 Mω and <15 Mω, respectively, were used as selection criteria for accepting recordings. Capacitance compensation and bridge balance functions were used for voltage-clamp and current-clamp experiments, respectively. Amphoterecin B (0.45-0.5 mg/mL) was dissolved in the intracellular solution containing (in mM): 140 potassium gluconate, 10 HEPES, 10 KCl, and 0.2 MgCl ₂ (pH adjusted to 7.28 with KOH; mOsm = 290). The external NaCl Ringer solution was supplemented with picrotoxin (50 μM) and NBQX (10 μM) to block GABA _A receptor- and non-NMDA receptor-mediated responses, respectively.

[0105] Data Analysis

[0106] In voltage-clamp mode, I _K(M > amplitude was measured as the relaxation current in response to a voltage step protocol from -20 mV to return potentials of -40 mV, -50 mV, -60 mV, and - 7OmV. I _K(M) density was calculated as I _K(M) amplitude (pA) / whole-cell capacitance (pF) for each cell. Kinetic analysis was performed by fitting the deactivation phase of the I _K(M ) trace with the standard single-component exponential fit equation: _, /(t) = A,- • e ^" ^ + C. [0107] In current-clamp mode, action potential trains were recorded using a series of depolarizing current steps ranging from +20 pA to +160 pA. The ability of CAl neurons to accommodate action potential frequency was determined by plotting interspike interval number vs. normalized spike frequency. Each cell served as its own control, and the frequency of each subsequent interspike interval was normalized to the frequency of the first interspike interval. Significant differences in SFA between strains were then determined by non-linear regression analysis (Prism 4.0). Decreased steepness of the fit is interpreted as decreased SFA. Depolarizing current ramps from 0 pA to +30-100 pA were applied to determine the threshold for action potential generation. Brief depolarizing current steps were applied to examine the following single action potential characteristics: 10-90% rise time, 90-10% decay time, afterdepolarization (ADP) area, and ADP decay kinetics. Passive membrane properties were reported by the MultiClamp Commander function. Membrane potential was maintained at -65 mV by direct current injection as needed. [0108] First generation chimeric 129 mice expressing the Kcnq2 A306T and Kcnq3 G310V mutations were engineered separately. Chimeric males carrying each mutation were then bred with wild-type C57/B16 females. Although they come from different progenitor mice, the wild-

type F2 generation mice that arise from both breeding schemes should be genetically comparable. For all cellular physiology parameters measured, data from wild-type B6.129+/+ (B6+/+) progeny of both the Kcnql mutant and Kcnq3 mutant chimeric males were first analyzed separately, then compared for significant differences. Since all ANOVA comparisons revealed non-significant differences (P > 0.50 for all parameter comparisons), wild-type data was pooled. Accordingly, all wild-type data reported for the B6 background strain are pooled.

EXAMPLE 5 Effects of Kcnq2 and Kcnq3 Point Mutations on CAl Hippocampal Neurons [0109] As shown in the previous examples, characterization of B6.129 and FVBN.129 mice carrying the Kcnql A306T and Kcnq3 G310V knock-in mutations established a wide range of effects on seizure threshold. Although mice expressing these mutations should be genetically comparable to their wild-type littermates with the exception of a single amino acid switch, the changes in whole-animal seizure threshold detailed in the previous chapter do not specifically implicate alterations in I _K ( _M ) function at the single-cell level. Therefore, we recorded Iκ(M) in CAl neurons in brain slices prepared from mice heterozygous and homozygous for the Kcnql A306T and Kcnq3 G310V mutations, as well as their wild-type littermates, in an effort to characterize changes in I _K(M) specifically.

[0110] Effects of Kcnql A306T Mutation on I _K(M) Amplitude, Density, and Deactivation Kinetics in B6 CAl neurons

[0111] In brain slices prepared from wild-type C57/B16.129 (B6+/+), heterozygous C57/B16.129-i:c« _? 2 ^A306T/+ (B6-Q2+/-) and homozygous C57fB16λ29-Kcnql ^{A306τlA306τ} (B6-Q2- /-) knock-in mice, CAl neurons exhibited an Iκ( _M ) across a range of hyperpolarizing return steps (Fig. 3A; n = 5-14 cells per genotype per return step). In hippocampal CAl neurons in brain slices prepared from B6-Q2+/- mice, I _R(M ) amplitudes were no different than those measured in wild-type B6 slices (Fig. 3B). In B6-Q2-/- CAl neurons, however, I _K ( _M ) amplitude was significantly decreased at every return step tested. Since a decrease in current amplitude could be explained by a decrease in CAl neuron size, whole-cell capacitance, a relative approximation of cellular surface area, was used to convert amplitudes to current densities. As was the case with the current amplitude results, I _K ( _M) density was not significantly affected in B6-Q2+/- CAl neurons, but was decreased in B6-Q2-/- CAl neurons at all voltages (Fig. 3C). I _K(M > amplitude and density data at every hyperpolarizing return step are summarized in Table 4.

TABLE 4

I _K(M) Amplitude and Density Values Are Reduced Across a Range of Voltage Steps in B6.129 CAl Neurons Homozygous for the Kcnql A306T Mutation

Step Ik(M) amplitude (pA) IK(M) density (pA/pF) from -2O mV B6+/+ B6-Q2+/- B6-Q2-/- B6+/+ B6-Q2+/- B6-Q2-/-

-40 mV 46.0 ± 1.4 52.7 ± 4.8 28.6 ± 3.5* 0.37 ± 0.04 0.33 ± 0.03 0.22 ± 0.02*

(n = ll) (n = 9) (n= ll)

-50 mV 37.8 ± 1.1 45.4 ± 5.8 16.6 ± 3.0* 0.32 ± 0.02 0.26 ± 0.04 0.12 ± 0.02*

(n = 10) (n = 5) (n = 7)

-60 mV 22.6 ± 1.6 20.2 ± 1.6 6.5 ± 2.6* 0.17 ± 0.01 0.11 ± 0.03 0.08 ± 0.02*

(n = 14) (n = 5) (n= 8)

-70 mV 11.2 ± 0.9 15.3 ± 2.9 4.2 ± 1.5* 0.09 ± 0.01 0.07 ± 0.02 0.03 ± 0.01*

(n = 13) (n = l l) (n = 8)

All values expressed as mean ± S.E.M.

*, B6-Q2-/- value is significantly different from B6+/+ (P < 0.05).

The numbers of cells recorded from (n values) are identical for I _KQλ) amplitude and density values within the same genotype.

[0112] Normalized current traces illustrate differences in I _K(M > deactivation kinetics in response to the -60 mV return step (Fig. 3D). Deactivation is accelerated in B6-Q2+/- CAl neurons at the -60 and -50 mV return steps (Fig. 3E). Deactivation kinetics are also accelerated in B6-Q2-/- CAl neurons at the -60 mV return step, but deactivation of these channels could not be adequately fit with a standard exponential decay equation.

[0113] Passive Membrane Properties of B6 CAl Neurons Expressing the Kcnq2 A306T

Mutation

[0114] Several fundamental membrane properties were monitored in B6+/+, B6-Q2+/-, and

B6-Q2-/- CAl pyramidal neurons. The Kcnq2 A306T mutation does not significantly alter input resistance, series resistance, resting membrane potential, or membrane capacitance of CAl neurons in B6 mice. These data are summarized in Table 5.

TABLE 5

The Kcnq2 A306T Mutation Does Not Affect the Passive Membrane Properties of B6.129 CAl Neurons

Mouse genotype

Membrane property

WT Het Horn

Input Resistance 107 + 4 103 + 5 103 + 4 (Mω) (n= 17) (n = 9) (n = 10)

Series Resistance 14.6 + 0.3 13.9 + 0.3 14.3 + 0.3 (Mω) (n= 15) (n = 9) (n= 8)

Resting Membrane -61.1 + 0.7 -59.9 ± 0.7 -59.97 + 0.8 Potential (mV) (n = 7) (n = 8) (n = 10)

Capacitance 131 + 5 134 + 13 137 + 11 (pF> (n = ll) (n = 9) (n = 10) AU values expressed as mean + S.E.M. Input resistance, series resistance, and resting membrane potential were taken from values reported by SealTest and MultiClamp Commander functions. Spike threshold values were calculated from the first action potential that was generated in response to a slow depolarizing current ramp.

[0115] Functional Consequences of Kcnq2 A306T Mutation on Spike Frequency Adaptation in B6 CAl Neurons

[0116] I _K(M) activation regulates action potential generation and facilitates spike frequency adaptation (SFA); in response to membrane depolarization, I _K(M) contributes to neuron repolarization and decreases input resistance, which attenuates action potential frequency (Goh and Pennefather, 1987; Castaldo et al., 2002; Yue and Yaari, 2004). To test whether the Kcnq2 A306T point mutation, which also reduces I _K(M) function, increases neuroexcitability via a similar decrease in SFA, we examined SFA in B6+/+, B6-Q2+/-, and B6-Q2-/- CAl neurons. [0117] In response to a +100 pA depolarizing current injection of 800 msec duration, action potential trains were elicited in slices prepared from each mouse genotype (Fig, 4A). The frequency of each successive interspike interval (ISI) was normalized to the frequency of the first, and interspike interval number vs. normalized frequency plots were created (Fig. 4B). The pooled SFA data were then fit with a single exponential decay equation. The non-linear fits were then compared by regression analysis to determine significant differences in SFA (Prism 4.0;

GraphPad Software, San Diego, CA). B6-Q2+/- CAl neurons (gray circles; n = 13) exhibited SFA across the course of the response train that was not significantly different from B6+/+ (black circles; n = 15). B6-Q2-/- CAl neurons (open circles; n = 6), however, display significantly reduced SFA compared to B6+/+. [0118] A stronger depolarizing current (+140 pA) was injected to examine SFA characteristics under conditions of more robust I _K(M ) activation. Sample traces from this current step are shown in Fig 4C. Similar comparisons were made between B6+/+, B6-Q2+/-, and B6- Q2-/- SFA (Fig 4D). Note that at the +140 pA step, in contrast with the +100 pA step, there is a significant difference between B6-Q2+/- SFA compared to that of B6+/+. SFA in B6-Q2-/- CAl neurons is still significantly different from B6+/+, and is also significantly different from B6- Q2+/-. These results demonstrate that a naturally hypofunctional Iκ( _M ) compromises the ability of neurons to modulate action potential firing, and further confirm the significant role of I _K ( _M) in regulating neuronal excitability.

[0119] Effects of Kcnq2 A306T Mutation on Single Action Potential Activity in B6 CAl Neurons

[0120] It has previously been shown that chemically blocking I _K(M) extends two facets of the action potential repolarizing phase: 90-10% decay time and afterdepolarization (Yue and Yaari, 2004). These delays in repolarization are consistent with a decreased outward potassium conductance such as that observed under conditions of I _K(M) block. We therefore examined if the Kcnq2 A306T mutation, which decreases I _K(M) function, similarly affects components of a single action potential. Single action potentials were elicited by delivering brief 5 msec depolarizing pulses of varying intensities such that action potentials were triggered after the offset of the stimulus waveform (Fig. 5A). The following parameters were examined: 10-90% spike rise time, 90-10% spike decay time, integrated area of the afterdepolarization (ADP) phase, and ADP decay kinetics (τ). The action potential 10-90% rise time and 90-10% decay time were not significantly different in B6+/+, B6-Q2+/-, or B6-Q2-/- CAl neurons (Figs. 5B and 5C). Surprisingly, ADP area was decreased in B6-Q2-/- CAl neurons (Fig. 5D), and ADP τ was accelerated in both.B6-Q2+/- and B6-Q2-/- CAl neurons (Fig. 5E). [0121] To examine possible changes in the mV threshold for action potential generation, a slow depolarizing current ramp was applied. The magnitude of the target ramp value was varied such that action potentials were generated just before the offset of the waveform (Fig. 5F).

Action potential threshold was determined as the membrane potential at which the first spike was generated. None of the action potential thresholds significantly differed in B6+/+, B6- Q2+/-, and B6-Q2-/- CAl neurons (Fig. 5G). The effects of Kcnq 2 A306T mutation on single action potential properties are summarized in Table 6.

TABLE 6

Active Membrane Properties of B6.129 CAl Neurons Carrying the Kcnq2 A306T mutation

Mouse genotype

Membrane property

WT Het Horn

AP threshold -48.4 ± 0.8 -46.3 + 0.6 -45.6 + 1.2 (mV) (n = 15) (n = 6) (n = 9)

AP 10-90% rise time 1.01 ± 0.02 0.92 ± 0.04 0.96 ± 0.03 (msec) (n = 9) (n = 6) (n = 6)

AP 90-10% decay time 3.59 + 0.17 3.07 + 0.21 3.42 ± 0.14 (msec) (n = 9) (n = 7) (n = 6)

ADP area 583 + 45 582 + 57 438 + 53*

(p A/msec) (n = 9) (n = 7) (n = 5)

ADP deactivation τ 75.6 + 5.5 56.6 ± 5.6* 55.0 ± 6.7* (msec) (n = 9) (n = 7) (n = 5)

All values expressed as mean ± S.E.M. Input resistance, series resistance, and resting membrane potential were taken from values reported by SealTest and

MultiClamp Commander functions. Spike threshold values were calculated from the first action potential that was generated in response to a slow depolarizing current ramp.

[0122] Effects of the Kcnq3 G310V Mutation on I _K(M) Amplitude, Density, and Deactivation Kinetics

[0123] In brain slices prepared from wild-type C57/B16.129 (B6+/+), heterozygous C57/Bl6.129-i&«^ ^G310V/+ (B6-Q3+/-) and homozygous C57/Bl6λ29-Kcng3 ^ωιmGmv (B6-Q3- /-) knock-in mice, CAl neurons exhibited a viable I _R(M) across a range of hyperpolarizing return steps (Fig. 6A; n = 8-21 cells per genotype per return step). In hippocampal CAl neurons in brain slices prepared from B6-Q3+/- mice, I _R(M) amplitudes were no different than those measured in B6+/+ slices (Fig. 6B). In B6-Q3-/- CAl neurons, I _K(M > amplitude was significantly

decreased at every return step tested. I _K(M) amplitude values were divided by membrane capacitance and converted to current density. I _K ( _M) density was only significantly decreased in B6-Q3+/- CAl neurons at the -50 mV step, but was decreased in B6-Q3-/-CA1 neurons at all return steps (Fig. 6C). The effects of Kcnq3 G310V mutation on I _K(M) amplitude and density data at every hyperpolarizing return step are summarized in Table 7.

TABLE 7

I _K(M) Amplitude and Density Values Are Reduced Across a Range of Voltage Steps in B6.129 CAl Neurons Homozygous for the Kcnq3 G310V Mutation

Step from IK(M) amplitude (PA) IK(M) density (pA/pF)

-20 mV B6+/+ B6-Q3+/- B6-Q3-/- B6+/+ B6-Q3+/- B6-Q3-/-

-40 mV 0.22 ±

46.0 ± 1.4 39.1 ± 2.7 25.0 ± 2.9* 0.37 ± 0.04 0.31 ± 0.02 λ U. λ UJ1 *

(n = ll) (n= 14) (n= 13)

-50 mV 37.8 ± 1.1 0.18 ±

34.3 ± 3.0 16.7 ± 2.5* 0.32 ± 0.02 0.23 ± 0.02f 0.03* (n = 10) (n= 19) (n = 17)

-60 mV 0.04 ±

22.6 ± 1.6 18.3 ± 2.6 3.1 ± 1.7* 0.17 ± 0.01 0.13 ± 0.02

(n = 14) OAJZ

(n = 21) (n = 20)

-70 mV -0.53 ± 0.004 ±

11.2 ± 0.9 6.8 ± 3.0 2.0* 0.09 ± 0.01 0.06 ± 0.02 (n = 13) 0.02* (n = 15) (n= 17) All values expressed as mean ± S.E.M. t, B6-Q3+/- value is significantly different from B6+/+ (P < 0.05).

*, B6-Q3-/- value is significantly different from B6+/+ (P < 0.05).

The numbers of cells recorded from (n values) are identical for I _K(M) amplitude and density values within the same genotype.

[0124] Normalized current traces illustrate differences in Iκ _(M) deactivation kinetics in response to the -60 mV return step (Fig. 6D). Deactivation is accelerated in B6-Q3+/- CAl neurons only at the -60 mV return step (Fig. 6E), while B6-Q3-/- deactivation is only accelerated at the -70 mV return step. I _K ( _M ) deactivation in B6-Q3-/- CAl neurons was too linear to be adequately fit with a standard exponential decay equation.

[0125] Passive Membrane Properties of B6 CAl Neurons Expressing the Kcnq3 G310V Mutation

[0126] Several fundamental membrane properties were monitored in B6+/+, B6-Q3+/-, and B6-Q3-/- CAl pyramidal neurons. The Kcnq3 A306T mutation does not significantly alter input resistance, series resistance, resting membrane potential, or membrane capacitance of CAl neurons in B6 mice. These data are summarized in Table 8.

TABLE 8

The Kcnq3 G310V Mutation Does Not Affect the Passive Membrane Properties of B6.129 CAl Neurons

Mouse genotype

Membrane property

WT Het Horn

Input Resistance 107 ± 4 105 + 6 100 + 4 (Mω) (n = 17) (n = 18) (n = 16)

Series Resistance 14.6 + 0.3 14.0 + 0.2 14.3 + 0.3 (Mω) (n = 15) (n = 19) (n = 25)

Resting Membrane -61.1 ± 0.7 -59.3 ± 0.5 -59.3 ± 0.8 Potential (mV) (n = 7) (n = 15) (n= 15)

Capacitance 131 + 5 140 ± 4 132 ± 7 (PF) (n = l l) (n = 21) (n = 21)

AU values expressed as mean ± S.E.M. Input resistance, series resistance, and resting membrane potential were taken from values reported by SealTest and MultiClamp Commander functions. Spike threshold values were calculated from the first action potential that was generated in response to a slow depolarizing current ramp.

10127] Functional Consequences ofKcnq3 G310V Mutation on Spike Frequency Adaptation in B 6 CAl Neurons

[0128] To test whether the Kcnq3 G310V point mutation increases neuroexcitability via a decrease in SFA, we examined SFA in B6+/+, B6-Q3+/- and B6-Q3-/- CAl neurons. In response to a +100 pA depolarizing current injection of 800 msec duration, action potential trains were elicited in slices prepared from each mouse genotype (Fig. 7A). The frequency of each successive interspike interval (ISI) was normalized to the frequency of the first, and interspike interval number vs. normalized frequency plots were created (Fig. 7B). The pooled

SFA data were then fit with a single exponential decay equation. The non-linear fits were then compared by regression analysis to determine significant differences in SFA (Prism 4.0; GraphPad Software, San Diego, CA). B6-Q3+/- CAl neurons (gray circles; n = 16) exhibited decreased SFA across the course of the response train compared to B6+/+ (black circles; n = 15). SFA in B6-Q3-/- CAl neurons (open circles; n = 18) is also decreased relative to both B6+/+ and B6-Q3+/-.

[0129] A stronger depolarizing current (+140 pA) was injected to examine SFA characteristics under conditions of more robust I _K(M) activation. Sample traces from this current step are shown in Fig. 7C. Similar comparisons were made between B6+/+, B6-Q3+/-, and B6- Q3-/- SFA (Fig. 7D). At the +140 pA step, both B6-Q3+/- and B6-Q3-/-SFA are reduced relative to B6+/+, and are significantly different from each other as well.

[0130] Effects of Kcnq3 G310V Mutation on Single Action Potential Activity in B6 CAl Neurons [0131] We examined the effects of the Kcnq3 G310V mutation on several components of the single compound action potential. Single action potentials were elicited by delivering brief 5 msec depolarizing pulses of varying intensities such that action potentials were triggered after the offset of the stimulus waveform (Fig. 8A). The following parameters were examined: 10- 90% spike rise time, 90-10% spike decay time, integrated area of the afiterdepolarization (ADP) phase, and ADP decay kinetics (τ). Action potential

[0132] 10-90% rise time was not significantly affected in B6-Q3+/- or B6-Q3-/- CAl neurons (Fig. 8B). Surprisingly, every parameter that involves fast repolarization was actually accelerated in CAl neurons carrying Kcnq3 G310V mutation. Both B6-Q3+/- and B6-Q3-/- CAl neurons had faster 90-10% decay times relative to B6+/+ (Fig. 8C). B6-Q3-/- neurons displayed decreased ADP area (Fig. 8D), and ADP τ was accelerated in both B6-Q3+/- and B6- Q3-/- CAl neurons (Fig. 8E).

[0133] To examine possible changes in the mV threshold for action potential generation, a slow depolarizing current ramp was applied. The magnitude of the target ramp value was varied such that action potentials were generated just before the offset of the waveform (Fig. 8F). Action potential threshold was determined as the membrane potential at which the first spike was generated. None of the action potential thresholds significantly differed in B6+/+, B6-

Q3+λ, and B6-Q3-/- CAl neurons (Fig. 8G). The effects of Kcnq3 G310V mutation on single action potential properties are summarized in Table 9.

TABLE 9

Active Membrane Properties of B6.129 CAl Neurons Carrying the Kcnq3 G310V Mutation

Mouse genotype

Membrane property

WT Het Horn

AP threshold -48.1 + 0.9 -48.6 + 0.9 -49.7 ± 0.7 (mV) (n = 12) (n = 7) (n = 7)

AP 10-90% rise time 1.01 + 0.02 0.93 + 0.03 0.95 + 0.04 (msec) (n = 9) (n = 17) (n = 19)

AP 90-10% decay time 3.59 + 0.17 3.07 + 0.15f 3.06 + 0.15* (msec) (n = 9) (n = 19) (n = 16)

ADP area 583 + 45 576 + 20 455 ± 30*

(p A/msec) (n = 8) (n = 15) (n= 10)

ADP deactivation τ 75.6 + 5.5 59.3 + 4.3t 57.0 + 5.1* (msec) (n = 9) (n = 17) (n = 12)

All values expressed as mean + S.E.M. Input resistance, series resistance, and resting membrane potential were taken from values reported by SealTest and MultiClamp Commander functions. Spike threshold values were calculated from the first action potential that was generated in response to a slow depolarizing current ramp.

[0134] Effects of the Kcnq3 G310V Mutation on I _K(M) Amplitude, Density, and Deactivation Kinetics in FN CAl Neurons [0135] In brain slices prepared from wild-type FVB/N.129 (FN+/+), heterozygous FYB/Nλ29-Kcnq3 ^Gnw/+ (FN-Q3+/-) and homozygous FVB/Nλ29-Kcnq3 ^Gmv/ωi0W (FN-Q3-/- ) knock-in mice, CAl neurons exhibited an I _K < _M) across a range of hyperpolarizing return steps (Fig. 9A; n = 5-9 cells per genotype per return step). In hippocampal CAl neurons in brain slices prepared from FN-Q3+/- mice, I _K(M) amplitudes were no different than those measured in FN+/+ slices (Fig. 9B). In FN-Q3-/- CAl neurons, I _K(M) amplitude was significantly decreased at the -60, -50, and -40 mV return steps. I _K(M) amplitude values were divided by membrane capacitance and converted to current density. I _K(M) density was not affected in FN-Q3+/- CAl

neurons, but was decreased in FN-Q3-/- CAl neurons at all return steps (Fig. 9C). The effects of Kcnq3 G310V mutation on I _K(M) amplitude and density at every hyperpolarizing return step in FN CAl neurons are summarized in Table 10.

TABLE 10

I _K(M) Amplitude and Density Values Are Reduced Across a Range of Voltage Steps in FN.129 CAl Neurons Homozygous for the Kcnq3 G310V Mutation

Step from IK(M) amplitude (pA) IK(M) density (pA/pF) -2O mV FN+/+ FN-Q3+/- FN-Q3-/- FN+/+ FN-Q3+/- FN-Q3-/-

0.32 ±

-4O mV 78.3 ± 10.1 80.9 ± 5.7 43.7 ± 4.7* 0.57 ± 0.10 0.59 ± 0.06 0.03*

(n = 5) (n = 6) (n = 9)

0.21 ±

-5O mV 64.0 ± 7.2 72.1 ± 4.3 29.2 ± 4.2* 0.37 ± 0.05 0.45 ± 0.05 0.03* (n = 6) (n = 5) (n = 9)

-6O mV 41.5 ± 6.6 0.29 ± 0.09 ±

43.0 ± 5.0 13.5 ± 1.1* 0.27 ± 0.04 0.003 (n = 6) (n = 6) (n = 8) 0.01*

-7O mV 29.5 ± 10.2 25.7 ± 4.0 11.8 ± 3.4 0.19 ± 0.06 0.16 ± 0.03 0.007 ± 0.02* (n = 5) (n = 6) (n = 9)

All values expressed as mean ± S.E.M.

*, FN-Q3-/- value is significantly different from FN+/+ (P < 0.05). The numbers of cells recorded from (n values) are identical for I _K(M) amplitude and density values within the same genotype.

[0136] Normalized current traces illustrate differences in I _K(M) deactivation kinetics in response to the -60 mV return step (Fig. 9D). Deactivation in FN-Q3+/- CAl neurons is not significantly affected at any return step (Fig. 9E), while FN-Q3-/-Hom deactivation kinetics are accelerated at every step measurable. At the -40 mV return step in FN-Q3-/- CAl neurons, I _K(M) deactivation was too linear to be adequately fit with a standard exponential decay equation.

[0137] Passive Membrane Properties of FN CAl Neurons Expressing the Kcnq3 G310V Mutation

[0138] Several fundamental membrane properties were monitored in FN+/+, FN-Q3+/-, and FN-Q3-/-Hom CAl pyramidal neurons. The Kcnq3 A306T mutation does not significantly alter input resistance, series resistance, resting membrane potential, or membrane capacitance of CAl neurons in FN mice. These data are summarized in Table 11.

TABLE Il

The Kcnq3 G310V Mutation Does Not Affect the Passive Membrane Properties of FN.129 CAl Neurons

Mouse genotype

Membrane property

WT Het Horn

Input Resistance 93 + 6 93 + 5 95 + 4 (Mω) (n = 6) (n = 7) (n= 14)

Series Resistance 13.5 ± 0.5 13.8 + 0.3 13.8 + 0.3 (Mω) (n = 7) (n = 7) (n= 16)

Resting Membrane -60.1 + 2.4 -61.3 + 1.5 -58.35 ± 1.58 Potential (mV) (n = 5) (n = 6) (n= 10)

Capacitance 160 + 15 162 + 7 155 ± 6 (pF) (n = 6) (n = 6) (n = 15)

All values expressed as mean ± S.E.M. "f , Het is significantly different from WT (P < 0.05). *, Horn is significantly different from WT (P < 0.05). Input resistance, series resistance, and resting membrane potential were taken from values reported by SealTest and MultiClamp Commander functions. Spike threshold values were calculated from the first action potential that was generated in response to a slow depolarizing current ramp.

[0139] Functional Consequences ofKcnq3 G310V Mutation on Spike Frequency Adaptation in FN CAl Neurons [0140] To test whether the Kcnq3 G310V point mutation increases neuroexcitability via a decrease in SFA, we examined SFA in FN+/+, FN-Q3+/- and FN-Q3-/- CAl neurons. In response to a +100 pA depolarizing current injection of 800 msec duration, action potential trains were elicited in slices prepared from each mouse genotype (Fig. 10A). The frequency of each successive interspike interval (ISI) was normalized to the frequency of the first, and interspike interval number vs. normalized frequency plots were created (Fig. 10B). The pooled

SFA data were then fit with a single exponential decay equation. The non-linear fits were then compared by regression analysis to determine significant differences in SFA (Prism 4.0; GraphPad Software, San Diego, CA). FN-Q3+/- CAl neurons (gray circles; n = 6) exhibited similar SFA across the course of the response train compared to FN+/+ (black circles; n = 6). SFA in FN-Q3-/- CAl neurons (open circles; n = 12) is significantly decreased relative to both FN+/+ and FN-Q3+/-.

[0141] A stronger depolarizing current (+140 pA) was injected to examine SFA characteristics under conditions of more robust I _KQA) activation. Sample traces from this current step are shown in Fig. 1OC. Similar comparisons were made between FN+/+, FN-Q3+/-, and FN-Q3-/- SFA (Fig. 10D). At the +140 pA step, FN-Q3+/- SFA is still not significantly different from FN+/+, and FN-Q3-/- SFA is reduced relative to FN+/+.

[0142] Effects of Kcnq3 G310V Mutation on Single Action Potential Activity in FN CAl Neurons [0143] We examined the effects of the Kcnq3 G310V mutation on several components of the single compound action potential. Single action potentials were elicited by delivering brief 5 msec depolarizing pulses of varying intensities such that action potentials were triggered after the offset of the stimulus waveform (Fig. HA). The following parameters were examined: 10- 90% spike rise time, 90-10% spike decay time, integrated area of the afterdepolarization (ADP) phase, and ADP decay kinetics (τ). Action potential 10-90% rise times were not significantly affected in FN-Q3+/- or FN-Q3-/- CAl neurons (Fig. HB). In contrast to the Kcnq3 G310V mutation on the C57/B16.129 background, in FVB/N.129 mice, 90-10% decay times, ADP area, and ADP τ were not significantly affected (Figs. HC, HD and HE). However, in FN-Q3+/- CAl neurons, ADP τ is trending toward being significantly slower than FN+/+ (P = 0.053). [0144] To examine possible changes in the mV threshold for action potential generation, a slow depolarizing current ramp was applied. The magnitude of the target ramp value was varied such that action potentials were generated just before the offset of the waveform (Fig. 1 IF). Action potential threshold was determined as the membrane potential at which the first spike was generated. None of the action potential thresholds significantly differed in FN+/+, FN- Q3+/-, and FN-Q3-/-Hom CAl neurons (Fig. 11G). The effects of Kcnq3 G310V mutation on single action potential properties in FN CAl neurons are summarized in Table 12.

TABLE 12

Active Membrane Properties of FN.129 CAl Neurons Carrying the Kcnq3 G310V Mutation

Mouse genotype

Membrane property

WT Het Horn

AP threshold -51.2 + 1.4 -50.6 + 1.6 -50.7 ± 0.9 (mV) (n = 6) (n = 7) (n = 15)

AP 10-90% rise time 0.81 ± 0.06 0.88 + 0.06 0.83 + 0.03 (msec) (n = 7) (n = 7) (n = 13)

AP 90-10% decay time 2.88 ± 0.21 3.21 ± 0.30 3.02 ± 0.18 (msec) (n = 7) (n = 7) (n = 15)

ADP area 504 ± 37 623 ± 49 486 ± 23 (p A/msec) (n = 5) (n = 7) (n = 14)

ADP deactivation τ 53.9 ± 4.4 72.3 ± 6.2(f ) 56.4 ± 2.9 (msec) (n = 9) (n = 17) (n = 12)

All values expressed as mean ± S.E.M. (|), Het is trending towards significantly different from WT (P = 0.053). Input resistance, series resistance, and resting membrane potential were taken from values reported by SealTest and MultiClamp Commander functions. Spike threshold values were calculated from the first action potential that was generated in response to a slow depolarizing current ramp.

[0145] In an effort to better characterize the physiologic effects of Kcnq2 and Kcnq3 mutations and their involvement in BFNC specifically, we conducted an electrophysiology study in "knock-in" mice carrying Kcnq2 and Kcnq3 point mutations identical to those identified in human patients afflicted with BFNC - Kcnq2 A306T and Kcnq3 G310V. [0146] The results of the present investigation establish that the exact mutations that cause human BFNC precipitate decreased I _K(M) amplitude, density, and accelerated deactivation kinetics in CAl neurons of the hippocampus. Spike frequency adaptation (SFA) is also inhibited in the same neurons, which is consistent with a hypofunctional I _K(M) , and provides evidence of increased neuronal excitability. We also detail several unexpected differences in the characteristics of single action potentials that are caused by these mutations. It is especially noteworthy that the two separate mutations do not precipitate identical changes in I _K(M) function, SFA, or single action potential characteristics; nor does the same mutation (Kcnq3 G310V) expressed on two different background mouse strains have uniform effects on these parameters.

These results are the first to describe changes in native neuronal I _K ( _M ) function that result from BFNC-causing Kcnq2 and Kcnq3 mutations. We have shed significant light on the consequences of Kcnq2 and Kcnq3 mutations in I _K ( _M ) function and their significance in regulating neuroexcitability at the single-cell level. Overall, the effects of these mutations on the single-cell biophysical properties examined appear consistent with the previously established alterations in whole-animal seizure threshold.

[0147] The Kcnq2 A306T mutation expressed in C67/B16.129 (B6) mice and the Kcnq3 G310V mutation expressed in both B6 and FVB/N.129 (FN) mice significantly attenuate I _K ( _M ) function in mice homozygous for these mutations via a combination of decreased current amplitude and density, as well as accelerated deactivation kinetics. In CAl neurons of B6- Q2+/-, B6-Q3+/-, and FN-Q3+/- mice, altered deactivation kinetics and decreased current density likely combine to decrease overall I _K(M) functionality as well. This presumption is supported by the decreased SFA observed, for instance, in B6-Q3+/- CAl neurons. In this population, I _K < _M) density, and deactivation kinetics are only significantly attenuated at the -50 mV and -60 mV steps, respectively. However, SFA is inhibited in B6-Q3+/- neurons relative to B6+/+ at all depolarizing steps that produce action potential trains (data shown for +100 p A and +140 pA steps). These results suggest that a combination of modifications in any of some or all of the components of the I _R(M) response can translate into significant changes in neuronal excitability. [0148] The apparent decrease in I _K(M) amplitude may be due to a depolarizing shift in the voltage of M channel activation or decreased peak channel conductance. Issues of voltage clamp error, space clamp error, and leak channel conductances preclude an accurate determination of the voltage of I _K(M) activation in the acute brain slice recording paradigm. However, it has been established that some BFNC-causing Kcnq2 and Kcnq3 mutations shift the voltage of KCNQ2/KCNQ3 current activation toward depolarized potentials in Xenopus laevis oocytes (Singh et al., 2003). The I/V relationships shown for I _K(M) amplitude and density hint at a depolarizing shift in I _K ( _M) activation, and would be consistent with and supportive of the results shown herein. [0149] The discrepancy between the effects of the Kcnq3 G310V mutation on CAl neuron physiology when expressed on the B6 and FN backgrounds is especially noteworthy. The most obvious divergence is illustrated by the SFA data: this mutation does not significantly affect SFA in FN-Q3+/- CAl neurons, but does significantly inhibit SFA in B6-Q3+/- CAl neurons. In addition, while B6-Q3+/- CAl neurons display a slight decrease in I _K(M) density and

deactivation kinetics, FN-Q3+/- CAl neurons display no change in I _K(M > density or deactivation kinetics. The correlation between decreased I _K(M) function and inhibited SFA observed in B6- Q3+/- CAl neurons, and the lack of effect on both I _K(M) function and SFA in FN-Q3+/- CAl neurons, is especially compelling evidence for the critical role of I _K(M) in regulating neuronal excitability. In CAl pyramidal neurons, KCNQ2 protein is normally concentrated at the Nodes of Ranvier and the axon initial segment (AIS), the anatomical site of action potential initiation, and KCNQ3 is most often co-expressed with KCNQ2 (Devaux et al., 2004). It is therefore not surprising that the Kcnq2 A306T and Kcnq3 G310V mutations examined here produce such considerable and diverse changes in action potential characteristics. [0150] It has not been determined whether the mutant forms of KCNQ2 and KCNQ3 expressed in these knock-in mice are inserted into the plasma membrane. We do know that the KCNQ2 A306T and KCNQ3 G310V proteins are functionally expressed in Xenopus oocytes and effectively reduce M-channel function (Schroeder et al., 1998; Singh et al., 2003), whereas KCNQ2 subunits lacking the C-terminus do not express (Schwake et al., 2000). In fact, interactions between KCNQ2 and KCNQ3 C-terminal segments are crucial for proper channel assembly and insertion into the membrane (Maljevic et al., 2003). Consistent with these studies, BFNC can also be caused by deletion mutations that insert premature stop codons in the C- terminal coding sequence. Computational analysis and oocyte expression data suggest that these mutations, which truncate the KCNQ2 C-terminus, also prevent expression of the truncated protein (Biervert et al., 1998; Pereira et al., 2004). In addition, Kcnq2-I- knock-out and homozygous Sztl-I- mice are perinatal lethal; both die of lung atelectasis (Watanabe et al, 2000; Yang et al., 2003). If the Kcnq2 A306T mutation precluded KCNQ2 protein insertion, mice homozygous for this mutation would likely die as well. To date, the only deaths observed in homozygous mutant knock-in mice have been postnatal as a result of lethal seizure events. We believe that since the Kcnq2 A306T and Kcnq3 G310V mutations involve residues within the pore-lining region and not the C-terminus, the mutant Kcnq2 and Kcnq3 gene products are in fact inserted into the membrane. We also believe that because mice homozygous for these mutations do not exhibit perinatal lethality, and display a quantifiable I _K(M > the mutant KCNQ2 and KCNQ3 subunits are functional. [0151] The accelerated repolarization following the firing of a single action potential observed in the mutant CAl neurons warrants further discussion. The M channel blocker linopirdine actually delays membrane repolarization in a recording paradigm similar to the one employed here (Yue and Yaari, 2004), an effect that is concordant with the role of I _K ( _M >

However, note that ADP area and τ values are decreased in both B6-Q2+/- and B6-Q3+/- CAl neurons, and 90-10% action potential decay times are decreased in B6-Q3+/- CAl neurons. Interestingly, only the FN-Q3+/- CAl neurons exhibit a trend toward prolonged decay times, increased ADP area, ADP τ, which would actually be the predicted result of a decreased IK(M> The results presented here suggest that one or more gene products (such as ionic conductances, phosphorylating or dephosphorylating enzymes) may be upregulated or downregulated to compensate the decreased I _K(M) function observed in B6-Q2+/- and B6-Q3+/-, but not FN-Q3+/-, CAl neurons. This is evidence of yet another background strain-dependent difference in the effects of Kcnq3 mutation. [0152] The results presented here provide a direct link between genetic alterations in Kcnq2 and Kcnq3, decreased I _K(M > function, and increased single-cell excitability. We suggest that decreased I _K(M) amplitude and subsequently inhibited SFA are the primary source of the lowered seizure threshold previously observed in the B6-Q2+/-, B6-Q3+/-, and FN-Q3+/- mice, and may contribute to seizure generation in human BFNC patients expressing these mutations.

EXAMPLE 6

Seizure Characteristics of N5F2 Homozygous Mutant Mice

[0153] 24-hour video monitoring has documented that spontaneous seizures occur in all 4 strains (Q3-FVBN, Q3-B6, Q2-FVBN Q2-B6 ) of N5F2 homozygous mutant mice. For example, Q3-FVBN mice can have over 600 spontaneous recurrent seizures (SRS) before 82 days of age that begin with generalized tonic clonic seizures (grade 5, black circles) until approximately 40 days (?) which are followed by forelimb and hindlimb clonic seizures (grade 3, gray circles) until approximately 60 days, followed by forelimb clonic seizures (grade 2, white circles) (Figure 12). This suggests that the seizure severity is reduced with increasing age. During the first approx 50 days, animals will have periods of no seizures that can last for up to 5(?) days which are interrupted by at least one generalized tonic clonic seizure. All homozygous Q3-FVBN mice exhibit seizures, suggesting that the SRS phenotype is fully penetrant. Q3 mice with the same mutation on the B6 background exhibit spontaneous seizures in less than half of the mice that were video monitored, suggesting incomplete penetrance of seizures. [0154] KCNQ2 homozygous mutant mice on either B6 or FVBN genetic background exhibit at least one generalized tonic clonic seizure documented with video monitoring. KCNQ2-FVBN mice die as a result of seizures while KCNQ2-B6 mice survive at least one SRS.

EXAMPLE 7

Histochemical Characteristics of N5F2 Homozygous Mutant Mice [0155] Upregulation of NPY [0156] The upregulation of neuropeptide Y (NPY) in mossy fiber axons of dentate granule cells is a hallmark of seizure activity although the significance of this increased expression is unclear (reference: Scharfman & Gray, EXS. 2006;(95):193-211Plasticity of neuropeptide Y in the dentate gyrus after seizures, and its relevance to seizure-induced neurogenesis). NPY upregulation was determined using a commercial antibody in N5F2 Q3-B6, Q3-FVB and Q2-B6 mouse brain slices following either single or multiple seizures. Following seizures, animals were sacrificed, brains were removed and fixed in formalin. Brains were then embedded in paraffin, sliced at 3um, mounted on charged slides and hybridized with an antibody to NPY using standard automated techniques. NPY is upregulated in the mossy fibers of dentate granule cells of Q3-FVB (figure 13), Q3-B6 and Q2-B6 (data not shown) homozygous mutant mice following seizures. This upregulation is absent from wild-type mice. Of note is the absence of NPY staining in the inner molecular layer of the dentate gyrus indicating that no mossy fiber sprouting has occurred in these animals.

[0157] Reactive Gliosis [0158] Astrocytes play a significant role in glutamate uptake, potassium buffering and sequestration of reactive oxygen species during normal brain function. In CNS injury such as seizures, astrocytes in the hippocampus become activated, also termed reactive gliosis, but it is presently unclear whether this process contributes to CNS pathology or is a result of it. Glial fibrillary acidic protein (GFAP) is the major intermediate filament protein in mature astrocytes and can be used as a marker of reactive gliosis. We examined our mutant N5F2 Q3-FVB animals for the presence of reactive gliosis and found that indeed, multiple spontaneous recurrent seizures results in an increase in GFAP expression (figure 14). Following seizures, animals were sacrificed, brains were removed and fixed in formalin. Brains were then embedded in paraffin, sliced at 3um, mounted on charged slides and hybridized with an antibody to GFAP using standard automated techniques.

[0159] The results presented above were obtained from mice that were genetically engineered to express the exact point mutations known to cause BFNC in humans. Specifically,

the Kcnql A306T mutation was expressed on the C57/B16.129 (B6) background, and the Kcnq3 G310V mutation was expressed on both the C57/B16.129 and FVB/N.129 (FN) backgrounds. Examples 2 and 3 summarized the alterations in seizure thresholds that result from knock-in point mutations in Kcnql and Kcnq3. Overall, mice heterozygous for these mutations (expressed on both backgrounds) largely exhibited decreased seizure thresholds. This is the first report to show that mice carrying BFNC-causing mutations display reduced seizure thresholds. Interestingly, we uncovered a number of unexpected differences in seizure thresholds that were dependent on individual mutation, mouse gender, and background strain. [0160] Examples 4 and 5 are an electrophysiological characterization of CAl neurons in C57/B16.129-iføzg2 ^A305T , C57/B\6λ29-Kcnq3 ^Gmy , and FVB/N.129-J-cw#3 ^G310V mice. In CAl neurons of mice homozygous for each of these mutations, I _K(M > amplitude and density were attenuated, deactivation kinetics were accelerated, and SFA was significantly inhibited. In heterozygous mice, however, a number of quite diverse effects were observed. The results of this chapter suggest a gene dosage effect for some, but not all, of the mutations examined. In addition, expressing the same mutation (Kcnq3 G310V) on different genetic backgrounds (i.e., B6 and FN) produced quite disparate effects. Surprisingly, both mutations expressed on the B6.129 background resulted in accelerated repolarization following a single action potential. This is indirect evidence of a cellular compensatory mechanism that may in fact neutralize the excitability that results from these Kcnq mutations. These data are the first to directly show that BFNC-causing mutations attenuate native neuronal I _K(M) and SFA.

[0161] Table 13 summarizes the most salient observed differences in seizure threshold and electrophysiology parameters in the mouse models oϊKcnq2 and Kcnq3 mutations examined in the present study.

TABLE 13

Summary of Differences in Seizure Threshold and Electrophysiology Parameters Affected by the Kcnql and Kcnq3 Mutations Studied

Seizure Phenotypes Electrophysiology Parameters

Minimal Minimal I _K (M) amp/ IK(M) AP fADP fADP

6 Hz clonic THE density kinetics SFA decay area kinetics

B6-Q2+A c? 4 : ? 4 4" r 4 ns ns 4

B6-Q3+/- 3 ns : $ t <? ns : $ J r ι° 4 4 ns 4

FN-Q3+/- cJ j, : $ t cj ns : $ | ns ns ns ns ns 1(0.053)

B6-Q2-/- nd nd nd r 4" 4 ns 4 4

B6-Q3-λ πd nd nd r r 4 4 4 4

FN-Q3-/- nd nd nd 4" 4° 4 ns ns ns

B6-Sztl+/- 4 4 4 r ns 4 nd nd nd

I, significant decrease relative to WT littermate controls (P < 0.05; Probit analysis for seizure phenotypes, ANOVA for electrophysiology parameters).

T, significant increase relative to WT littermate controls (P < 0.05; Probit analysis for seizure phenotypes, ANOVA for electrophysiology parameters). ^a , I _K ( _M ) amplitude, density, and deactivation kinetics significantly affected at one or more return steps. ns, no significant difference relative to WT littermate controls, nd, no data obtained.

[0162] The ECT results obtained in the Kcnql A306T and Kcnq3 G310V knock-in mice suggest that these mutations precipitate several gender-dependent differences in seizure thresholds. In minimal clonic ECT testing, seizure thresholds of all heterozygous mutant female mice (B6-Q2+/-, B6-Q3+/-, and FN-Q3+/-) were significantly decreased, while none of the male mice were affected. This is consistent with the idea that female mice, which often exhibit decreased seizure threshold relative to male mice, can therefore be more susceptible to an additional preconvulsive challenge. However, the results obtained in the partial psychomotor seizure testing paradigm imply a change in the opposite direction. In this seizure phenotype, female mice expressing the Kcnq3 G310V mutation exhibited significantly increased seizure thresholds, a trend that was evident regardless of genetic background. We posit that the Kcnq3

G310V mutation may have actually lowered the seizure threshold of female mice drastically enough to induce spontaneous interictal bursting and/or undetected seizure activity, possibly resulting in a seizure-refractory period. If this were the case, and female mice were tested during the refractory period, it could explain the apparent increase in seizure threshold in these mice (Revilla et al., 1999). In addition, in the knock-in mice, we show electrophysiological evidence of accelerated membrane repolarization following single spike activity (Examples 4 and 5). This result is quite the opposite of what might be expected under conditions of decreased I _K (M), and therefore inhibited membrane repolarization. It is possible that such an anticonvulsant compensatory mechanism might be robust enough to protect against seizure activity in local networks such as the forebrain structures activated by partial seizure testing. The more severe seizure phenotypes (minimal clonic and minimal THE) that recruit additional brain structures might then overcome these protective mechanisms, and a reduced seizure threshold would be observed. [0163] Being cognizant of the widespread gender-dependent differences in seizure thresholds in the knock-in mutant mice, all electrophysiology data initially were maintained and analyzed separately according to gender. Scattergrams were constructed for mice of each genotype, and between-gender statistical comparisons were made for each electrophysiological parameter Although separating the data according to gender often lowered the power coefficient of the analysis (StatMate 2, GraphPad Software), none of these comparisons was determined to be significant (P > 0.50). We therefore pooled the data acquired from CAl neurons in male and female mice according to each genotype for further statistical comparisons. [0164] It is curious that whole-animal behavioral measures in male and female mice uncover such drastic differences in sensitivity to Kcnq mutations, but none of the single-cell parameters measured exhibited a significant gender bias. There are countless explanations for this apparent discrepancy. For instance, the effects of Kcnq mutation on I _K(M) function were only examined in one cell type. Other neuron populations might be affected by these mutations differently than the CAl neurons examined here. The sex hormones estrogen and progesterone modify neuronal excitability in CAl neurons, facilitating and preventing seizure activity, respectively (Woolley and McEwen, 1993). In addition, GABA systems are particularly susceptible to hormonal regulation, specifically the neurosteroids (Wilson, 1992), even demonstrating regional specificity (Wilson and Biscardi, 1997). Since our brain slice recording solutions were not supplemented with any sex hormones and slices were constantly perfused, hormone levels were likely depleted over time. Normal sex hormone-dependent cellular mechanisms might therefore

be rendered ineffective in our in vitro system, and therefore explain the absence of gender effect in the electrophysiology studies.

[0165] Due to limited availability of age-matched WT and heterozygous mutant mice, the ECT studies were conducted in mice that were much older (20-50 weeks) than those used for the electrophysiology studies (8-15 weeks). In addition, the electrophysiology studies were also only designed to test I _K(M) function and other biophysical mechanisms that are affected by I _K ( _M > There are many other mechanisms that contribute to seizure threshold that were not examined in these studies. For instance, the tonic activity level of many neurotransmitter systems (i.e., glutamate, GABA, acetylcholine) might also be up- or down-regulated in response to the Kcnq mutations examined here. Along these lines, studies where CA3 neurons are stimulated and synaptic transmission is recorded in CAl neurons would be useful in examining alterations in such neurotransmitter systems. Likewise, a characterization of other voltage-gated ion channels (i.e., sodium, potassium, and calcium channels) might uncover additional compensatory mechanisms. Further investigation of any of these matters might provide insight into the gender- related differences in seizure susceptibility, but is well beyond the scope of this dissertation.

[0166] The electrophysiology studies conducted in knock-in mice uncovered several interesting alterations in I _K(M > function, neuronal excitability, and single spike characteristics (Examples 4 and 5). Between our Sztl electrophysiology study and other Xenopus oocyte studies characterizing the effects of BFNC-causing mutations, it was not surprising that both knock-in mutations reduced IK( _M ) amplitude. However, the accelerated I _K ( _M) deactivation kinetics we observed were somewhat surprising. Faster KCNQ2/KCNQ3 current deactivation kinetics have been reported in Xenopus oocytes expressing KCNQ2 R214W mutant subunits (Castaldo et al., 2002). However, KCNQ2 residue 214 is contained within a different region of the M channel (the S4 voltage-sensing region). In addition, results obtained in Xenopus oocytes often do not closely parallel those obtained in the native system (Dorr, 1993; Lewis et al., 1997; Sivilotti et al., 1997).

[0167] The apparent gene dosage effect observed in some, but not all, of the knock-in mutations with respect to IK( _M ) function and spike frequency adaptation (SFA) is particularly interesting. A gene dosage effect was previously described in KCNQ2/KCNQ3 currents recorded in Xenopus oocytes expressing the Kcnql A306T mutation, but not the Kcng3 G310V mutation (Schroeder et al., 1998). We observed a gene dosage effect in SFA in both B6-Q2 and B6-Q3, but not FN-Q3, mutant mice.

[0168] The effects of the Kcnq3 G31 OV mutation on I _K ( _M ) function and SFA when expressed on different backgrounds warrants further discussion. I _K(M ) density and deactivation kinetics were significantly reduced in B6-Q3 ^+/" mice, producing a decrease in SFA. In FN-Q3 ^{+ "} mice, Iκ _(M) density was not affected, but deactivation kinetics were significantly reduced; CAl neurons of these mice did not display altered SFA. Interestingly, BFNC-causing mutations only display about 80% penetrance in human families ~ 20% of family members carrying the same mutations are not afflicted with BFNC and do not display a higher incidence of adult onset epilepsy (Leppert et al., 1993). In mice, it appears that differences in B6.129 and FN.129 mouse genetic backgrounds may account for the differential sensitivity to M channel mutation. It could be that such differences are analogous to subtle differences in combinatorial genetics within human families that account for incomplete penetrance of the BFNC phenotype. [0169] It is particularly fascinating that the degree to which SFA is affected closely parallels the degree to which I _K(M) function is affected in CAl neurons heterozygous for these knock-in mutations, hi addition, and perhaps most importantly, spontaneous seizures have been seen in both B6-Q2 ^"/- and B6-Q3 ^7" mice. This congruous relationship lends further support to the considerable role that the ∑ _K(M) plays in regulating neuroexcitability.

[0170] It would be interesting to examine the pharmacosensitivity of the knock-in mice to not only linopirdine and retigabine, which target the M channel, but also other proconvulsants and anticonvulsants. Such an examination conducted in the knock-in mice of the present invention would suggest better treatment strategies for BFNC patients who develop adult onset epilepsy. For instance, if an I _K(M) enhancing anticonvulsant (i.e., retigabine) proved to be less potent in a knock-in mouse, BFNC patients with similar point mutations might also be less responsive to the drug. [0171] Some BFNC patients develop epilepsy in adulthood after a considerable latent period. The latent period between an initial insult and increased seizure susceptibility is a characteristic of many forms of epilepsy, specifically temporal lobe epilepsy (White, 2002). A mutation in the Kcnq2IKcnq3 genes and an increased likelihood of developing epilepsy could be considered an initial insult and evidence of increased seizure susceptibility, respectively. Along these lines, the Kcnq mutant knock-in mice could be used to conduct studies designed to observe the long- term consequences of genetic mutations on seizure susceptibility, or so-called "second hit" studies (Walker et al., 2002).

[0172] Finally, due to robust KCNQ2, KCNQ3, and KCNQ5 expression, and large I _K(M) amplitudes observed in dorsal root ganglion and superior cervical ganglion neurons, the I _R(M) has

recently garnered attention in the field of neuropathic pain research. Retigabine has analgesic effects in the tail flick model of acute neuropathic pain, and the sciatic ligation model of chronic neuropathic pain. Linopridine coinjection diminishes the efficacy of retigabine (Dost et al., 2004), suggesting that at least part of retigabine 's anti-hyperalgesic properties are mediated by I _K(M) enhancement. The knock-in mice therefore can also employed in models of neuropathic pain to determine the long-term effects of an innate decrease in I _K ( _M > Sciatic ligation or formalin injection studies might well uncover differences in the susceptibility of these mice to insults that produce neuropathic pain. [0173] The results presented herein are the first to show reductions in native neuronal I _K ( _M ) function, as well as increased single-cell neuroexcitability, that result from the expression of BFNC-causing mutations. Our approach of combining whole-animal behavior with single-cell biophysics has helped solidify the link between attenuated I _K(M) function and increased seizure susceptibility that result from Kcnq2 and KcnqS mutation. Thus, the studies contained herein have significantly contributed to our understanding of how Kcnq mutations might precipitate human epilepsy, and further elucidate the neurophysiologic role of I _K(M > We believe that because proper I _K(M) function is so critically involved in regulating neuronal excitability, it will prove useful in the management of not only BFNC, but also other forms of epilepsy. Moreover, drugs designed to increase I _K(M > activity could in fact lead to improved treatment of virtually any pathology characterized by aberrant neuronal excitability. These include, but are by no means limited to, neuropathic pain, migraine, stroke, dementia and anxiety.

[0174] It will be appreciated that the methods and compositions of the instant invention can be incorporated in the form of a variety of embodiments, only a few of which are disclosed herein. It will be apparent to the artisan that other embodiments exist and do not depart from the spirit of the invention. Thus, the described embodiments are illustrative and should not be construed as restrictive. It will also be appreciated that in this specification and the appended claims, the singular forms of "a," "an" and "the" include plural reference unless the context clearly dictates otherwise. It will further be appreciated that in this specification and the appended claims, the term "comprising" or "comprises" is intended to be open-ended, including not only the cited elements or steps, but further encompassing any additional elements or steps.

BIBLIOGRAPHY

[0175] Antoni, N. (1941). Aketenskapsforbudet for epileptiker bor upphavas atminstone tills vidard. Svensk Laktartidningen 38:233-43.

[0176] Barton, M.E. et al. (2001). Pharmacological characterization of the 6 Hz psychomotor seizure model of partial epilepsy. Epilepsy Res 47:217-227.

[0177] Barton, M.E. et al. (2004). The effect of CGX-1007 and CI-1041, novel NMDA receptor antagonists, on NMDA receptor-mediated EPSCs. Epilepsy Res 59:13-24.

[0178] Biervert, C. et al. (1998). A potassium channel mutation in neonatal human epilepsy. Science 279:403-406. [0179] Bjerre, I. and E. Corelius (1968). Benign neonatal familial convulsions. Acta Pediatr Scand 57:557-61.

[0180] Bortolotto, Z.A. (1991). Effects of bilateral clamping of carotid arteries on hippocampal kindling in rats. Physiol Behav 49:667-671.

[0181] Brown, D.A. and P.R. Adams (1980). Muscarinic suppression of a novel voltage- sensitive K+ current in a vertebrate neurone. Nature 283:673-676.

[0182] Bunin, B. A. and Ellman, J.A. et al. (1992). "A general and expedient method for the solid-phase synthesis of 1,4-benzodiazepine derivatives." J Am Chem Soc 114:10997-10998.

[0183] Bunting, M. et al. (1999). Targeting genes for self-excision in the germ line. Genes & Dev 13:1524-1528. [0184] Bye, A.M. (1994). Neonate with benign familial neonatal convulsions: recorded generalized and focal seizures. Pediatr Neurol 10:164-165.

[0185] Carell, E. et al. (1994a). "A Novel Procedure for the Synthesis of Libraries Containing Small Organic Molecules." Angew Chem MEd Engl 33:2059-2061.

[0186] Carell, E. et al. (1994b). "A Solution-Phase Screening Procedure for the Isolation of Active Compounds from a Library of Molecules." Angew Chem Int Ed Engl 33 :2061 -2064.

[0187] Castaldo, P. et al. (2002). Benign familial neonatal convulsions caused by altered gating of KCNQ2/KCNQ3 potassium channels. JNeurosci 22: 1-6.

[0188] Charlier, C. et al. (1998). A pore mutation in a novel KQT-like channel gene in an idiopathic epilepsy family. Nat Genet 18:53-55. [0189] Cho, CY. et al. (1993). "An Unnatural biopolymer." Science. 261:1303-1305.

[0190] Claes, et al. (2004). "De novo KCNQ2 mutations in patients with benign neonatal seizures." Neurology (55:2155-8).

[0191] Commission on Classification and Terminology of the International League Against Epilepsy (1989). Proposal for revised classification of epilepsies and epileptic syndromes. Epilepsia 30:389-99.

[0192] Cull, M.G. et al. (1992). "Screening for Receptor Ligands Using Large Libraries of Peptides Linked to the C Terminus of the lac Repressor." Proc Natl Acad Sd USA 89:1865- 1869.

[0193] CwMa, S.E. et al. (1990). "Peptides on Phage: A Vast Library of Peptides for Identifying Ligands." Proc Natl Acad Sd USA 87:6378-6382.

[0194] Devlin, J.L. et al. (1990). "Random Peptide Libraries: A Source of Specific Protein Binding Molecules." Science 249:404-406.

[0195] Devaux, JJ. et al. (2004). KCNQ2 is a nodal K+ channel. JNeurosci 24:1236-1244.

[0196] DeWitt, S.H. et al. (1993). "Diversomers": An Approach to Nonpeptide, Nonoligomeric Chemical Diversity." Proc Natl Acad Sd USA 90:6909-6913.

[0197] Dorr, R.T. (1993). Clinical properties of yeast-derived versus Escherichia coli- derived granulocyte-macrophage colony-stimulating factor. Clin Ther 15:19-29; discussion 18.

[0198] Dost, R. et al. (2004). The anti-hyperalgesic activity of retigabine is mediated by KCNQ potassium channel activation. Naunyn Schmiedebergs Arch Pharmacol 369:382-390. [0199] Erb, E. et al. (1994). "Recursive Deconvolution of Combinatorial Chemical Libraries." Proc Natl Acad Sd USA 91:11422-11426.

[0200] Felici, F. et al. (1991). "Selection of antibody ligands from a large library of oligopeptides expressed on a multivalent exposition vector." JM?/ Biol 222:301-310.

[0201] Ferraro, T.N. et al. (1998). Genetic influences on electrical seizure threshold. Brain Res 813:207-210.

[0202] Fodor, S. et al. (1993). "Multiplexed biochemical assays with biological chips." Nature 364:555-556.

[0203] Frankel, W.N. et al. (2001). Electroconvulsive thresholds of inbred mouse strains. Genomics 74:306-312. [0204] Gallop, M.A. et al. (1994). "Applications of combinatorial technologies to drug discovery. 1. Background and peptide combinatorial libraries." J Med Chem 37:1233-1251.

[0205] Goh, J. W. and P. S. Pennefather (1987). Pharmacological and physiological properties of the after-hyperpolarization current of bullfrog ganglion neurones. J Physiol 394:315-330.

[0206] Gribkoff, V.K. (2003). The therapeutic potential of neuronal KCNQ channel modulators. Expert Opin Ther Targets 7:737-748.

[0207] Heron, S.E. et al. (2002). Sodium-channel defects in benign familial neonatal- infantile seizures. Lancet 14:851-852.

[0208] Hirose, S. et al. (2000). A novel mutation of KCNQ3 (c.925TaC) in a Japanese family with benign neonatal convulsions. Ann. Neurol 47:822-826. [0209] Hirsch, E. et al. (1993). Electroclinical signs of benign neonatal familial convulsions. Ann Neurol 34:835-41.

[0210] Horwell, D. et al (1996). "'Targeted' molecular diversity: design and development of non-peptide antagonists for cholecystokinin and tachykinin receptors." Immunopharmacology 33:68-72. [0211] Houghten, R.A. et al. (1992). "The use of synthetic peptide combinatorial libraries for the identification of bioactive peptides." Biotechniques 13:412-421.

[0212] Lam, K.S. (1997). "Application of combinatorial library methods in cancer research and drug discovery." Anticancer Drug Des 12:145-167.

[0213] Leppert, M. et al. (1989). Benign familial neonatal convulsions linked to genetic markers on chromosome 20. Nature 337:647-648.

[0214] Leppert, M. et al. (1993). Searching for human epilepsy genes: a progress report. Brain Pathol 3:357-369.

[0215] Lerche, H. et al. (1999). A reduced K+ current due to a novel mutation in KCNQ2 causes neonatal convulsions. Ann Neurol 46:305-312. [0216] Lewis, T.M. et al. (1997). The ion channel properties of a rat recombinant neuronal nicotinic receptor are dependent on the host cell type. J Physiol 505(Pt 2):299-306.

[0217] Maljevic, S. et al. (2003). C-terminal interaction of KCNQ2 and KCNQ3 K+ channels. J Physiol 548:353-360.

[0218] McCoIl, CD. et al. (2003). Electroencephalographic characterisation of pentylenetetrazole-induced seizures in mice lacking the alpha 4 subunit of the neuronal nicotinic receptor. Neuropharmacology 44:234-243.

[0219] McKhann, G.M., 2 ^nd et al. (2003). Mouse strain differences in kainic acid sensitivity, seizure behavior, mortality, and hippocampal pathology. Neuroscience 122:551-561.

[0220] Miles, D.K. and G.L. Holmes (1990). Benign neonatal seizures. J Clin Neurophysiol 3:369-379.

[0221] Miraglia del Giudice, E. et al. (2000). Benign familial neonatal convulsions (BFNC) resulting from mutation of the KCNQ2 voltage sensor. Eur J Med Genet 8:994-997.

[0222] Nagy, A. et al. (1993). Derivation of completely cell culture-derived mice from early- passage embryonic stem cells. Proc Natl Acad Sci USA 90:8424-8428 [0223] Okada, M. et al. (2002). Impaired M-current and neuronal excitability. Epilepsia 43:36-38.

[0224] Otto, J.F. et al. (2004). Mice carrying the sztl mutation exhibit increased seizure susceptibility and altered sensitivity to compounds acting at the m-channel. Epilepsia 45:1009- 1016. [0225] Peterson, S. (1998). Electroshock. In Neuropharmacology Methods in Epilepsy Research, CRC Press, Boca Raton, FL.

[0226] Plouin, P. (1990). Benign neonatal convulsions. In Neonatal seizures, Wasterlain, CG. and P. Pert, eds., Raven, New York, pp.:51-59.

[0227] Plouin, P. (1997). Benign neonatal convulsions. In Epilepsy: a comprehensive textbook, Engel, J. and T.A. Pedley, eds., Lippincott-Raven, Philadelphia, pp. 2247-2455.

[0228] Rejdak, R. et al. (2001). Brain tolerance and preconditioning. Pol J Pharmacol 53:73-79.

[0229] Rett, A. and R. Teubel (1964). Neugeborenen-krampfe im kahmen einer epileptisch belasteten familie. Wiener Klinische Wochenschrift 76:609-613. [0230] Revilla, V. et al. (1999). Seizure-refractory period after a single stimulation and inhibition of seizures after repetitive stimulation in the gerbil: effects on blood Cortisol levels. Epilepsia 40:1-4.

[0231] Ronen, G.M. et al. (1993). Seizure characteristics in chromosome 20 benign familial neonatal convulsions. Neurology 43:1355-1360. [0232] Ross, S. A. et al. (2000). Phenotypic characterization of an alpha 4 neuronal nicotinic acetylcholine receptor subunit knock-out mouse. JNeurosci 20:6431-6441.

[0233] Schroeder, B.C. et al. (1998). Moderate loss of function of cyclic-AMP-modulated KCNQ2/KCNQ3 K+ channels causes epilepsy. Nature 396:687-690.

[0234] Schwake, M. et al. (2000). Surface expression and single channel properties of KCNQ2/KCNQ3, M-type K+ channels involved in epilepsy. J Biol Chem 275:13343-13348. [0235] Scott, J.K. and J.P. Smith (1990). "Searching for Peptide Ligands with an Epitope Library." Science 249:386-390.

[0236] Shah, M.M. et al. (2002). Molecular correlates of the M-current in cultured rat hippocampal neurons. J Physiol 544:29-37.

[0237] Singh, N.A. et al. (1998). A novel potassium channel gene, KCNQ2, is mutated in an inherited epilepsy of newborns. Nat Genet 18:25-29.

[0238] Singh, N.A. et al. (2003). KCNQ2 and KCNQ3 potassium channel genes in benign familial neonatal convulsions: expansion of the functional and mutation spectrum. Brain 126:2726-2737.

[0239] Sivilotti, L.G. et al. (1997). Recombinant nicotinic receptors, expressed in Xenopus oocytes, do not resemble native rat sympathetic ganglion receptors in single-channel behaviour. J Physiol 500(Pt 1):123-138.

[0240] Surti, T.S. and Jan, L. Y. (2005). "A potassium channel, the M-channel, as a therapeutic target." Curr Opin Investig Drugs 6:704-11.

[0241] Thomas, K.R. and M.R. Capecchi (1987). Site-directed mutagenesis by gene targeting in mouse embryo-derived stem cells. Cell 51:503-512.

[0242] Toman, J. (1951). Neuropharmacologic considerations in psychic seizures. Neurology 1:444-460.

[0243] VernadaMs, A. and D.M. Woodbury (1972). Influence of Cortisol on brain and spinal excitability in developing rats. UCLA Forum Med Sd 15:35-42. [0244] Walker, M.C. et al. (2002). Disease modification in partial epilepsy. Brain 125:1937-1950.

[0245] Watanabe, H. et al. (2000). Disruption of the epilepsy K.CNQ2 gene results in neural hyperexcitability. JNeurochem 75:28-33.

[0246] White, H.S. (2002). Animal models of epileptogenesis. Neurø/og3/ 59:S7-S14. [0247] White, H.S. et al. (2002). General principles: discovery and preclinical development of antiepileptic drugs. In: Antiepileptic Drugs, 5 Edition, Levy, R.H. et al., eds, Lippincott Williams and Wilkins, Philadelphia.

[0248] Wilson, M. A. (1992). Influences of gender, gonadectomy, and estrous cycle on GABA/BZ receptors and benzodiazepine responses in rats. Brain Res Bull 29: 165-172. [0249] Wilson, M.A. and R. Biscardi (1997). Influence of gender and brain region on neurosteroid modulation of GABA responses in rats. Life Sci 60:1679-1691.

[0250] Wong, J.Y. et al. (2002). Proconvulsant-induced seizures in alpha(4) nicotinic acetylcholine receptor subunit knockout mice. Neuropharmacology 43:55-64.

[0251] Woodbury, L. and V. Davenport (1952). Design and use of a new electroshock seizure apparatus, and analysis of factors altering seizure threshold and pattern. Arch Int Pharmacodynam Ther 92:97 -107.

[0252] Woolley, CS. and B.S. McEwen (1993). Roles of estradiol and progesterone in regulation of hippocampal dendritic spine density during the estrous cycle in the rat. J Comp Neurol 336:293-306.

[0253] Wu, Y-J. and S.I. Dworetzky (2005). Recent developments on KCNQ potassium channel openers. Cur Med Chem 12:453-460.

[0254] Yang, J.S. et al. (2002). ARFGAPl promotes the formation of COPI vesicles, suggesting function as a component of the coat. J Cell Biol 159:69-78.

[0255] Yang, Y. et al. (2003). Spontaneous deletion of epilepsy gene orthologs in a mutant mouse with a low electroconvulsive threshold. Hum MoI Genet 12:975-984. [0256] Yue, C. and Y. Yaari (2004). KCNQ/M channels control spike afterdepolarization and burst generation in hippocampal neurons. JNeurosci 24:4614-4624.

[0257] Zonana, J. et al. (1984). Familial neonatal and infantile seizures: an autosomal- dominant disorder. Am J Med Genet 18:455-459.

[0258] Zuckermann, R.N. (1994). "Discovery of nanomolar ligands for 7-transmembrane G- protein-coupled receptors from a diverse N-(substituted)glycine peptoid library." J Med Chem 37:2678-2685.