• BAC-transgenesis;
  • QTL;
  • Seizure;
  • Epilepsy;
  • Maximal electroshock seizure threshold;
  • Kcnj9;
  • Kcnj10


  1. Top of page
  2. Abstract
  6. Acknowledgments

Summary: Purpose: Previous quantitative trait loci (QTL) mapping studies from our laboratory identified a 6.6 Mb segment of distal chromosome 1 that contains a gene (or genes) having a strong influence on the difference in seizure susceptibility between C57BL/6 (B6) and DBA/2 (D2) mice. A gene transfer strategy involving a bacterial artificial chromosome (BAC) DNA construct that contains several candidate genes from the critical interval was used to test the hypothesis that a strain-specific variation in one (or more) of the genes is responsible for the QTL effect.

Methods: Fertilized oocytes from a seizure-sensitive congenic strain (B6.D2-Mtv7a/Ty-27d) were injected with BAC DNA and three independent founder lines of BAC-transgenic mice were generated. Seizure susceptibility was quantified by measuring maximal electroshock seizure threshold (MEST) in transgenic mice and nontransgenic littermates.

Results: Seizure testing documented significant MEST elevation in all three transgenic lines compared to littermate controls. Allele-specific RT-PCR analysis confirmed gene transcription from genome-integrated BAC DNA and copy-number-dependent phenotypic effects were observed.

Conclusions: Results of this study suggest that the gene(s) responsible for the major chromosome 1 seizure QTL is found on BAC RPCI23-157J4 and demonstrate the utility of in vivo gene transfer for studying quantitative trait genes in mice. Further characterization of this transgenic model will provide new insight into mechanisms of seizure susceptibility.

C57BL/6 (B6) mice are highly resistant to seizures induced experimentally whereas DBA/2 (D2) mice are highly susceptible (Engstrom and Woodbury, 1988; Kosobud and Crabbe, 1990; Ferraro et al., 2002). This observation has served as the basis for much research aimed at gaining insight into the pathogenesis of human epilepsy; however, the biological factors that mediate this strain difference are still unknown. Previous work has demonstrated that gene variation plays a large role in B6 and D2 strain-specific seizure responses and several chromosomal regions have been implicated. In particular, a segment of distal chromosome 1 has been shown to exert a strong influence on the difference in seizure susceptibility between B6 and D2 mice in a variety of diverse paradigms (Buck et al., 1997, 1999; Ferraro et al., 1997, 1999, 2001). Recently, the gene(s) responsible for distal chromosome 1 seizure susceptibility was mapped to a 6.6 Mb interval between Pbx1 and D1Mit150 (Ferraro et al., 2004). The prior discovery of mutations in single ion-channel genes that are sufficient individually to cause seizures both in animals and humans (Noebels 2003) facilitates the identification of several genes in the critical interval as high-priority candidates for analysis. The most compelling of these are the potassium ion channel genes Kcnj9 and Kcnj10.

Evidence exists to support a role for both Kcnj9 and Kcnj10 in the biology of seizures. Kcnj10 encodes a pH-sensitive, inward rectifier potassium channel (KIR4.1) and is expressed throughout the brain in both neurons and glia (Bredt et al., 1995; Neusch et al., 2001; Li et al., 2001, Wu et al., 2004). Targeted knockout of Kcnj10 in mice produces a severe neurological phenotype involving motor abnormalities and postnatal lethality (∼3 weeks; Kofuji et al., 2000). Among common strains of inbred mice, Kcnj10 exhibits a polymorphism that predicts a threonine (B6) to serine (D2 and most other common inbred strains) amino acid substitution at position 261 of the parent protein (Ferraro et al., 2004). Furthermore, the gene maps to a haplotype block within the critical interval on chromosome 1 that discriminates the strains as a function of their threshold for electrically induced seizures (Ferraro et al., 2004). Kcnj9 is a member of the G-protein-activated inward rectifier potassium channel (GIRK) family that is expressed ubiquitously in the brain (Lesage et al., 1994; Dissman et al., 1996). It has been shown to be involved indirectly in seizure mechanisms in that the spontaneous seizure phenotype observed in mice with a targeted knockout of the Kcnj6 gene is exacerbated when this knockout is combined with a Kcnj9 knockout (Torrecilla et al., 2002). Kcnj9 maps within the D2 seizure susceptibility haplotype block, although no coding SNPs have been identified (Ferraro et al., 2004). Nonetheless, Kcnj9 is both a positional and biological candidate for the previously mapped seizure susceptibility locus on distal chromosome 1.

Quantitative trait loci (QTL) mapping studies have provided the chromosomal localization of genes for many complex traits but it is clear that there are limitations with respect to identifying the responsible genes and the causative sequence variation within those genes (Flint et al., 2005). As in other areas of QTL research, mapped loci for seizure-related traits are plentiful; however, identification of specific genetic variants associated with seizure susceptibility or resistance has not been readily forthcoming. Various approaches have been utilized to translate QTL into quantitative trait genes (QTGs) but there is no general formula for success and no single criterion that must be satisfied in order to document a genotype–phenotype correlation. Rather, there is support for the idea that each QTL should be assessed in ways that are most relevant to its specific biology (Abiola et al., 2003). Despite the fact that not all QTL may be amenable to a transgenic approach, the use of gene transfer to analyze the phenotypic effects of a QTL is a potentially powerful method for QTG identification (Ikeda et al., 2002; Henkel et al., 2005). Furthermore, gene transfer is useful for translational studies of complex traits where human gene variations can be isolated and studied in experimental animals (Jiang et al., 2005). Transgenic animals carrying large genomic fragments provide more comprehensive information with regard to genetic effects than animals harboring shorter constructs, and this is especially important for studying clustered genes (Gao et al., 2005). Thus, in order to test more formally the potential influence of Kcnj9 and Kcnj10 on seizure susceptibility in the B6-D2 model, we have created transgenic lines of mice in which the effect of a Kcnj9- and Kcnj10-containing BAC could be studied in relation to the distal chromosome 1 seizure susceptibility QTL Szs1 (Ferraro et al., 1997). Results of the study document BAC-mediated increase of seizure threshold and narrow the region of DNA harboring the causative gene variation to a 186-Kb interval.


  1. Top of page
  2. Abstract
  6. Acknowledgments


Studies were conducted using mice bred in the animal facility at the Department for Veteran's Affairs Medical Center in Coatesville, Pennsylvania, U.S.A. Colonies of B6 and D2 mice have been maintained at the facility since 1995 with annual supplementation of breeders from The Jackson Laboratory (Bar Harbor, ME, U.S.A.). In general, mice are kept on a 14 h light/10 h dark cycle with food and water available freely at all times. The mating scheme for breeding involves housing one male with two females. Pregnant female mice are given their own cage until litters are delivered and weaned. Litters are weaned at 3–4 weeks and pups are group-housed by gender. In addition to B6 and D2 strains, mice from the congenic strain B6.D2-Mtv7a/Ty-27d were also used in these studies. This strain has a B6-derived genetic background with a 6.6 Mb segment of distal chromosome 1 introgressed from D2 (Ferraro et al., 2004). All studies were approved by Institutional Animal Care and Use Committees.

BAC isolation

BAC isolation was initiated based on the hypothesis that strain variation in Kcnj10 was responsible for the chromosome 1 seizure susceptibility QTL. Thus, a full-length cDNA clone of Kcnj10 was isolated from a B6 brain cDNA pool by PCR using the following set of primers: forward: 5′CTT CGA GAA TTC ATG ACG TCG GTC GCT 3′; reverse: 5′ACC CTG CAG TCA GAC GTT ACT AAT GCG 3′. The PCR amplicon (1160 bp including the entire coding region), was subcloned into the pTOPOII vector as per manufacturer's protocols and then sequenced. We commissioned Research Genetics (now Invitrogen, Carlsbad, CA, U.S.A.) to screen a BAC library constructed from B6 genomic DNA (RPCI-23) using the gel purified full-length clone as a template for a random-primed hybridization probe. Seven BAC clones were identified and subsequently analyzed in our lab by PCR and Southern blot. Full-length (and smaller) fragments of Kcnj10 were amplified, subcloned, and sequenced from clone RPCI-23 157J4. The clone was submitted to the NIH/NHGRI Mouse BAC Sequencing Program and was selected for sequencing. The complete sequence for clone 157J4 is now in Genbank (AC074311). It contains six genes in their entirety: Atp1a4, Atp1a2, Igsf8, Kcnj9, Kcnj10, and Pigm. A schematic diagram of BAC 157J4 is shown in Fig. 1.


Figure 1. Schematic diagram of BAC RPCI23-157J4 (AC 074311). Numbering of gene position is given in relation to overall size of BAC (185.8 Kb). BLAST was used to compare the sequence of individual genes with BAC sequence. The following cDNA sequences were accessed: Atp1a4 (XM355283), Atp1a2 (BC036127), Igsf8 (BC048387), Kcnj9 (NM008429), Kcnj10 (NM020269/AY374423), and Pigm (NM026234).

Download figure to PowerPoint

BAC transgenesis

Female B6.D2-Mtv7a/Ty-27d mice were superovulated and housed overnight with male mice from the same strain. The following morning, reproductive tissues containing fertilized oocytes were transferred to the Transgenic and Chimeric Mouse Facility of the University of Pennsylvania. BAC 157J4 DNA was purified using an affinity column (Princeton Separation Inc., Adelphia, NJ, U.S.A.) and diluted to 5 ng/μl in filtered modified TE buffer (10 mM Tris/0.1 mM EDTA pH 7.5). Picoliter volumes of BAC DNA in circular form were injected into the male pronucleus of fertilized oocytes and approximately 25 injected eggs were implanted into each of seven primed (mated overnight with vasectomized male) female CD-1 mice. Litters were screened for BAC integration into somatic cells by analyzing DNA isolated from tail clips taken at the time of weaning. Germ line transmission was established by mating tail-positive progeny (i.e., BAC DNA detected in tail tissue) with “wildtype” congenic mice and analyzing genomic DNA from mice. As depicted in Fig. 2, BAC-positive transgenic progeny were identified based on the presence of a B6 allele for the previously described SNP in Kcnj10 (Ferraro et al., 2004).


Figure 2. Digitized image of PCR-amplified Kcnj10 alleles following restriction analysis with Fnu4h1 and agarose gel electrophoresis as described in Methods. The B6 allele is represented by a 220 bp band whereas the D2 allele is represented by 2 bands, 120 and 100 bp. Sources of DNA: Lane 1: PhiX HaeIII size marker (New England Biolabs), lane 2: B6 (positive control), lane 3: D2 (positive control), lane 4: BxD F1 (positive control), lane 5: water (negative control), lane 6: transgenic mouse, lane 7: nontransgenic littermate, lane 8: nontransgenic littermate.

Download figure to PowerPoint

Transgene copy number

Copy number was estimated using quantitative RT-PCR analysis of genomic DNA in conjunction with a commercially available Pigm gene expression assay (ABI product number: Mm00452712_s1). Since Pigm has no introns, the expression assay yields amplified product from genomic DNA. We constructed a standard curve by analyzing serial dilutions of genomic DNA from nontransgenic mice (two copies of Pigm) and used this to validate copy number estimates for each transgenic line. For each of the three lines, four nontransgenic and four transgenic mice were analyzed. A difference of 1 cycle-threshold value indicates doubling of copy number.

Brain dissection

Mice were euthanized by cervical dislocation under CO2 anesthesia. Brains were removed from the skull quickly and dissected by hand on an ice-cold glass stage. The brain was oriented with its ventral surface facing up and two coronal cuts were made resulting in three discrete sections. The first cut was made at the level of the optic chiasm and anterior commissure to yield a “forebrain” section that also included olfactory bulbs (but did not include hippocampus). The second cut was made at the junction of the brainstem and diencephalon yielding a “midbrain” section that also included parietal and temporal cortex, the hippocampus and basal ganglia, and a “hindbrain” section that included cerebellum and occipital cortex. Sections were frozen immediately on dry ice and stored at −80°C until analyzed.

RNA extraction

RNA was isolated from brain sections taken from the three transgenic lines as well as from parental mice. Tissues were homogenized in Trizol solution using a Tissuemizer grinder. RNA was isolated following the manufacturer's instructions. Optical density readings were taken at 260 nm and 280 nm using a μQuant microplate spectrophotometer to determine both the concentration of the RNA and its relative purity. Purity is determined by calculating the ratio of the readings at 260 and 280; ratios between 1.9 and 2.0 were taken to indicate that the nucleic acid was of appropriate quality for subsequent experiments. RNA quality was also assessed by electrophoresis on formaldehyde-agarose gels, which provides an estimate of degradation. The influence of contaminating DNA on RNA expression analysis was circumvented by using assays designed to amplify products across exon/intron boundaries and also by DNase treatment.

Real-time PCR

Prior to first strand synthesis, DNase (Sigma) was used to degrade DNA in the RNA sample according to the manufacturer's instructions. We produced cDNA from 2 μg of RNA using the Superscript First-Strand synthesis kit (Invitrogen). The procedure was scaled by a factor of 4 in order to create a pool of cDNA large enough for all experiments.

Allele-specific gene expression analysis

In order to determine if genes were transcribed from the BAC construct in vivo, we developed PCR/RFLP assays that exploited sequence differences between endogenous D2-derived genes and B6-derived genes introduced on the BAC. For Kcnj9, a 686 bp fragment from bases 201 to 887 of the Kcnj9 cDNA (NCBI accession #U11860) was amplified using the following primer pair: forward 5′ CGA CTC TGG CCA TCC ATC; reverse: 5′ CGC CCA CGC GAA ACA TGA. An SNP at position 685 results in an MspI restriction site in B6 (“G” allele) and produces fragments of 484 and 202 bp following digestion. The enzyme does not cut the D2 (“C”) allele. For Atp1a2, a 260 bp fragment from bases 276 to 537 of the Atp1a2 cDNA (NCBI accession #NM_178405) was amplified using the following primer pair: forward: 5′CCT CAC CAA TCA GCG AGC; reverse: 5′GTA GTA GGA GAA GCA GCC. A SNP at position 453 results in a NcoI site in D2 (“C” allele) and yields fragments of 84 and 176 bp following digestion. The enzyme does not cut the B6 (“T”) allele. For Kcnj10, a 220 bp fragment, representing bases 685 to 905 from translation start site (NCBI accession #AF322631), was amplified using the following primer pair: forward: 5’ATT CGG CTC AAC CAG GTC AA; reverse: 5’GGT AGG TAG GAA GTG CGA AC. A SNP at position 785 results in an Fnu4H1 restriction site in D2 (“G” allele) that gives fragments of 120 and 100 bp following digestion. Fnu4H1 does not cut the B6 (“C”) allele. As with genotyping assays (Fig. 2), B6 and D2 DNA samples were always analyzed together with putative transgenic DNA as a control for incomplete digestion.

mRNA expression

Applied Biosystems, Inc. (ABI, Foster City, CA, U.S.A.) gene expression assays were used for relative quantification of Atp1a2, Kcnj9, Kcnj10, and Pigm mRNA (ABI product numbers Mm00617899_m1, Mm0434622_m1, Mm0445028_m1 and Mm00452712_s1, respectively). Expression was analyzed in transgenic mice, nontransgenic littermates and mice from the B6 and D2 parental strains. Experiments were carried out using the ABI 7300 Real-Time PCR System (Applied Biosystems Inc.). The comparative CT method for relative quantification was utilized and the “common housekeeping gene”Gapdh (ABI product number Mm99999915_g1) served as a normalization control. Quantitative PCR reactions were set-up following the manufacturer's instructions and standard curves over a 5-log range of template concentration were generated to establish amplification equivalency between Gapdh and the genes of interest. All samples were analyzed in triplicate and averages showed less than 5% coefficient of variation. Relative levels of gene expression for each sample were determined by using the formula 2(-ΔΔCT), which gives the fold change compared to Gapdh.

Western blot analysis

Protein for Western analysis was isolated from frozen brain tissue by homogenization in Trizol or 1× RIPA lysis buffer (1% Nonidet P-40, 1% sodium deoxycholate, 0.1% SDS, 0.15 M NaCl, 0.01 M sodium phosphate, pH 7.2, 2 mM EDTA, and protease inhibitors) followed by centrifugation at 13,000 ×g. Protein concentrations were determined using the BCA protein assay kit (Pierce, Rockford, IL, U.S.A.) according to the manufacturer's instructions. For Kir4.1 (KCNJ10) western blots, 50 μg of total protein from each brain region was separated on a 3–8% Tris-acetate Nu-PAGE gradient gel (Invitrogen). For Kir3.3 (KCNJ9) western blots, 50 μg of total protein from each brain region was separated on a 4–12% Bis-Tris Nu-PAGE gradient gel (Invitrogen). Fractionated protein samples were transferred subsequently to nitrocellulose membranes that were blocked for 1 hr in TBS-T (0.1% Tween-20 dissolved in TBS) containing 10% dry milk. Membranes were incubated in either a 1:400 dilution of an anti-Kir4.1 rabbit polyclonal antibody (Alomone Labs, Jerusalem, Israel) or a 1:200 dilution of an anti-Kir3.3 rabbit polyclonal antibody (Alomone Labs) for 2 h at room temperature. Blots were washed for 30 min with TBS-T and then incubated with a 1:10,000 dilution of an anti-rabbit peroxidase-conjugated secondary antibody (Jackson ImmunoResearch, West Grove, PA, U.S.A.). Blots were then washed for 1 h at room temperature with TBS-T and visualized using enhanced chemiluminescence reagents according to the manufacturer's instructions (Amersham Biosciences, Piscataway, NJ, U.S.A). Blots were imaged and analyzed semi-quantitatively using Image J software ( A time series of film exposures was used to generate a standard curve by which linearity of Western blot signals could be established to insure valid quantification.

Seizure testing

Maximal electroshock seizure threshold (MEST) was determined using a constant current electroshock unit (model #7801, Ugo Basile, Varese, Italy) as described previously (Ferraro et al., 1998, 2001, 2004). Mice were tested with a single shock once per day beginning at age 8–9 weeks. The current level was set to 20 mA for the first trial. It was increased by 2 mA with each successive daily trial until a generalized seizure was elicited and then by 1 mA until a maximal seizure was elicited. All stimuli were delivered at 60 Hz with a 0.4 ms pulse width and 0.2 s duration. A generalized seizure was defined in part by loss of postural control. A maximal seizure was defined by tonic hindlimb extension. Seizures were observed to be induced at all current intensities utilized: lower intensities produced running, stun or facial, and/or forelimb clonus whereas higher intensities produced generalized seizures. The sequence of responses that characterized a trial in which a maximal seizure was observed is as follows: loss of postural control, tonic forelimb flexion, tonic hindlimb flexion, and tonic hindlimb extension. On occasion, a terminal phase involving hindlimb clonus is also observed. Mice were euthanaized by cervical dislocation under CO2 anesthesia immediately after a trial in which a maximal seizure was elicited.

Data analysis

MEST was quantified as described previously with the arithmetic mean used for group comparisons (Ferraro et al., 2004). Multiway ANOVA was conducted with MEST as the dependent variable and gender, genotype (transgenic or nontransgenic littermate) and line (TG-4, TG-12 or TG-18) as independent variables. Newman–Keuls posthoc test was used to compare transgenic and nontransgenic littermate values within each line. Western blots were scanned and analyzed semi-quantitatively using the ImageJ software package ( Multiway ANOVA was conducted with blot density as the dependent variable; line, genotype, and brain region were used as independent variables. Newman–Keuls posthoc test was used to compare transgenic with nontransgenic littermate values for each line. Parental B6 and D2 mice were also compared to one another. Gene expression analysis was conducted using the comparative CT method for relative quantification according to Sequence Detection Systems Chemistry Guide supplied with the ABI 7300 Instrument (ABI, Foster City, CA, U.S.A). Statistical analysis involved comparison of mean ΔCT values between transgenic and nontransgenic mice using the Mann–Whitney U test with Bonferroni correction. Parental B6 and D2 mice were also compared to one another. The Truepistat Software package (Richardson, TX, U.S.A.) was used for all statistical analyses.


  1. Top of page
  2. Abstract
  6. Acknowledgments

We obtained litters from all seven foster dams and in total they comprised 36 progeny. From these, we identified four tail-positive pups, three of which (TG-4, TG-12, and TG-18) were independent germ line founders of new transgenic lines. When mated with wildtype B6.D2-Mtv7a-27d congenic mice, founder mice TG-4 and TG-12 passed on BAC DNA to approximately 50% of offspring (equal numbers of males and females) whereas inheritance from founder TG-18 was much lower at approximately 15%. Transgenic progeny derived from founder TG-18 pass the BAC to approximately 50% of offspring, similar to TG-4 and TG-12 mice. This result suggests delayed integration of BAC DNA (after the first cleavage of the fertilized egg) in founder TG-18 resulting in somatic and germ cell mosaicism. BAC transgenic mice are observed to be generally indistinguishable from nontransgenic littermates. They exhibit the expected B6 coat color and their body weight is similar to same-gender littermates. They are viable into adulthood, exhibit grossly normal locomotor activity, and demonstrate typical breeding patterns.

Neither the transgenic founders nor any of their progeny have been observed to exhibit spontaneous seizures or any other major behavioral or motor dysfunction; however, assessment of seizure susceptibility documented significantly higher MEST in transgenic mice compared to nontransgenic littermates for all three lines with TG-18 exhibiting the largest magnitude of effect (13, 7, and 6 mAmp for TG-18, TG-12, and TG-4 lines, respectively). The difference in MEST between transgenic mice and nontransgenic littermates was statistically significant for all lines (Fig. 3) and this result was independent of gender (female data not shown). Estimation of copy number using the Pigm expression assay documented an amplification difference of 2 CT for TG-18 and 1 CT for TG-4 and TG-12. These results suggest that eight copies of Pigm are present (six copies of the BAC are inserted) in line TG-18 and four copies of Pigm present (two copies of the BAC inserted) in TG-4 and TG-12. Thus, the magnitude of the effect on seizure phenotype correlated with estimated copy number as TG-18 mice showed a two-fold greater MEST increase versus nontransgenic littermates compared to TG-12 and TG-4 mice. It is possible that incorporation of a longer BAC concatamer into the genome of founder TG-18 is related to the delay in integration proposed above to explain the initial low BAC transmission rate.


Figure 3. Seizure phenotype in parental (B6 and D2), congenic (B6.D2-Mtv7a-27d) and BAC-transgenic mice. Maximal electroshock seizure threshold in 8- to 9-week-old male mice is shown. Strain designations: B = B6; D = D2, C = B6.D2-Mtv7a-27d (Ferraro et al., 2004); the symbols “+” and “−“ refer to BAC-positive transgenic mice and their nontransgenic littermates, respectively. The top of each bar represents the mean MEST (+ SD). Values (given in mAmp): B = 65 ± 7, D = 25 ± 2, C = 43 ± 6, TG-4 (−) = 47 ± 3, TG-4 (+) = 54 ± 3, TG-12 (−) = 44 ± 5, TG-12 (+) = 51 ± 3, TG-18 (−) = 45 ± 4, TG-18 (+) = 58 ± 4; * p < 0.05, ** p < 0.01 vs. nontransgenic littermates; *** p < 0.001 vs. B and D (ANOVA, Newman–Keuls; n = 20–22 per group).

Download figure to PowerPoint

We confirmed in vivo expression of genes from the BAC construct by using allele-specific RT-PCR/RFLP assays. Transgenic mice exhibited both B6 and D2 alleles for Atp1a2, Kcnj9, Kcnj10, and Pigm (data not shown), evidence that these genes are transcribed from the integrated BAC. Results of quantitative RT-PCR analysis show that relative message levels for Atp1a2, Kcnj9, Kcnj10, and Pigm are consistently higher in transgenic mice compared to nontransgenic littermates and although minor regional differences between the three lines are noted, overall gene expression is increased approximately two-fold (Table 1). We also conducted RT-PCR analysis in parental strains and show that, in general, expression levels are similar between the strains; however, it is noted that D2 mice exhibit significantly higher levels of Kcnj10 in hindbrain (1.74-fold) and significantly higher levels of Atp1a2 in all brain regions (1.9 to 2.7 fold) compared to B6 mice (Table 1).

Table 1. Regional brain gene expression1 in transgenic mice, nontransgenic littermates, and parental strains
 TG-4TG-12TG-18Parental strains
  1. Levels of Atp1a2, Kcnj9, Kcnj10, and Pigm mRNA are expressed as fold-increase in transgenic mice compared to nontransgenic littermates; for parental strains, the ratio D2:B6 is shown. Upper limit of range was determined by solving the expression 2−ΔΔCT with ΔΔCT+ s, where s = the standard deviation of the ΔΔCT value. Lower limit was calculated similarly as ΔΔCT− s. Relative quantification involved normalization to Gapdh (see Methods for details). FB, forebrain; MB, midbrain; HB, hindbrain. * p < 0.05; ** p < 0.01 (Mann–Whitney U Test; n = 6 per group). Bold values highlight statistically significant differences.

 MB1.320.77–1.931.201.09–1.271.351.01–1.83  2.70**2.52–2.88
 HB1.390.82––1.19  3.03**2.42–3.781.74*1.55–1.92

We show results of Western blot analyses in Figs. 4 and 5. With respect to Kir3.3 (KCNJ9) immunoreactivity, there is a major band visible at the level of the 45 kD marker in all samples (Fig. 4 left panel). This result is consistent with the predicted size of the Kir3.3 monomer that contains 376 amino acid residues (NCBI accession #NM_008429). Several minor bands are also noted on the blot; however, preincubation of primary antibody serum with peptide antigen (supplied by the antibody manufacturer) for 2 h at 37°C resulted in loss of the major band at 45 kD without altering the other minor bands (data not shown). Relative to actin, we demonstrate that Kir3.3 immunoreactivity is significantly higher in forebrain and midbrain of transgenic mice compared to nontransgenic littermates; differences in hindbrain did not reach levels of statistical significance (Figure 5). Western blot results also show the expected banding pattern for Kir4.1 (KCNJ10) immunoreactivity. Thus, there is a single band that migrates with the 45 kD size marker in all samples (Fig. 4) and this is consistent with the predicted size of the Kir4.1 monomer that contains 379 amino acid residues (NCBI accession #AB039879.1). Blots incubated with serum that was preincubated with the antigen peptide supplied by the primary antibody manufacturer were blank (data not shown). Relative to actin, we show that Kir4.1 immunoreactivity is significantly higher in midbrain and hindbrain of transgenic mice compared to nontransgenic littermates whereas there was no difference in forebrain levels (Fig. 5). The magnitude of the difference in both Kir3.3 and Kir4.1 immunoreactivity between transgenic and nontransgenic mice is consistent with the magnitude of differences in mRNA level. Comparison of Kir3.3 and Kir4.1 immunoreactivity in brain from parental B6 and D2 mice shows no statistically significant differences with the exception of hindbrain where D2 mice have significantly more Kir4.1 immunoreactivity compared to B6 (Fig. 5).


Figure 4. Western blot analysis. Representative experiment showing Kir3.3 (left) and Kir4.1 (right) immunoreactivity in the hindbrain from individual transgenic mice and nontransgenic littermates (n = 3 each).

Download figure to PowerPoint


Figure 5. Quantification of Western blots. The top of each bar represents the mean (+ SD) level of Kir3.3/KCNJ9 or Kir4.1/KCNJ10 immunoreactivity in parental mice, transgenic mice, and nontransgenic littermates normalized to actin. FB, forebrain; MB, midbrain; HB, hindbrain. * p < 0.05; ** p < 0.01 (ANOVA, Neuman–Keuls; n = 3 per group for transgenic and nontransgenic littermates; n = 6 for parental B6 and D2 mice).

Download figure to PowerPoint


  1. Top of page
  2. Abstract
  6. Acknowledgments

In the present study, we demonstrate elevation of electrical seizure threshold through in vivo transfer of a BAC that harbors six linked genes, two of which (Kcnj9 and Kcnj10) are primary candidates for major seizure susceptibility QTL on distal chromosome 1 reported previously in B6 (seizure resistant) and D2 (seizure susceptible) mice (Ferraro et al., 1997; 1999; 2001). The experimental strategy involved random insertion transgenesis in which the BAC construct was delivered by microinjection into 1-day old fertilized oocytes that were derived from a seizure-susceptible congenic strain, B6.D2-Mtv7a/Ty-27d, produced during refinement of the chromosome 1 QTL map position (Ferraro et al., 2004). In general, phenotypes that result from random insertion methods must be interpreted with caution since it is possible that genomic integration of exogenous DNA can interrupt the sequence of an endogenous (unknown) gene(s), effectively yielding a functional knockout. In our study, demonstration of decreased seizure susceptibility (increased seizure threshold) in three independent transgenic lines (Fig. 3) minimizes this possibility and suggests that a BAC-derived gene(s) is responsible for the observed phenotype. Further support for this conclusion may be generated by comparing the magnitude of the change in seizure susceptibility in relation to transgene (BAC) copy number. Thus, the MEST increases shown for male mice in Fig. 3 represent fold-increases of 1.3 for TG-18, the line harboring six copies of the BAC, compared to 1.15 and 1.08-fold increases for lines TG-12 and TG-4, respectively, each of which harbor two copies. Overall, female mice had lower MEST values compared to males (data not shown); however, the ability of BAC genes to reverse the seizure susceptible phenotype was independent of gender and, as for the males, the magnitude of phenotypic effect was greater in TG-18 females compared to females from the other two lines again emphasizing the potential effect of copy number variation.

It is not clear which of the BAC transgenes is responsible for the observed effects on seizure threshold. Of the six known genes that are present on BAC 157J4, biological plausibility is highest for Kcnj9 and Kcnj10 with the latter gene supported by the largest body of data. We reported previously that Kcnj10 exhibits a nonsynonymous (coding) SNP that predicts an amino acid variation in the intracellular C-terminus of the channel protein at position 262 wherein the B6 sequence encodes a threonine residue and the D2 sequence encodes serine (Ferraro et al., 2004). Although this change is conservative, it is possible that it could affect the phosphorylation state of the protein since threonine and serine residues are substrates for different kinases (Light et al., 2000; Wischmeyer and Karschin, 1996). The C-terminal region of the protein is also known to be involved in channel gating as well as interaction with other channel subunits and membrane-associated proteins (Nishida and MacKinnon, 2002) and therefore subtle genetic variation may have significant functional effects. Human KCNJ10 exhibits a coding SNP, Arg271Cys, which predicts an amino acid variation at a relative position nine residues removed from the variation discovered in the mouse and which discriminates patients with common forms of epilepsy from control subjects in genetic association studies (Buono et al., 2004). Importantly, this positive result has been replicated by an independent laboratory (Lenzen et al., 2005) making it one of the few confirmed genetic association findings in common epilepsies and strengthening the hypothesis that Kncj10 is involved in differences in seizure susceptibility between B6 and D2 mice. In vitro studies aimed at examining the electrophysiology of these channel polymorphisms were unable to document significant functional differences when the variants were expressed in Xenopus oocytes (Shang et al., 2005) suggesting that the influence of the variation is subtle at the physiological level, that it requires participation of molecular elements not present in oocytes or that it is nonfunctional but is in genetic association with the causative variation. Further study is required to evaluate these possibilities.

With regard to Kcnj9, it is first important to note that there are no known coding SNPs between B6 and D2 mice. Therefore, if this gene influences seizure phenotype in our model, it must involve a strain-specific difference in expression either at the message or protein level. This hypothesis is not supported by data generated in parental B6 and D2 mice that document similar relative levels of both Kcnj9 mRNA and Kir3.3 immunoreactivity throughout the brain in the two strains. Thus, the likelihood that Kcnj9 contributes to the observed differences in seizure phenotype is reduced.

Also located on BAC 157J4 are two genes encoding alpha subunits of Na+/K+-ATPase, Atp1a4 and Atp1a2. Although both of these genes are biologically plausible candidates for influencing seizure susceptibility in terms of the critical role of their enzyme system in membrane physiology, enthusiasm for them is weak for several reasons. The α4 subunit gene exhibits testes-specific expression (Underhill et al., 1999) and is therefore not likely to be involved in a nervous system phenotype. The α2 gene is expressed at high levels in excitable tissues including brain; however, there are no coding SNPs between B6 and D2 mice. Differences in brain mRNA expression level exist (Table 1); however, mice with a targeted knockout of Atp1a2 do not experience spontaneous seizures nor do they exhibit a phenotype involving behavioral hyperexcitability (Moseley et al., 2003). With regard to studies in human epilepsy, analysis of SNPs in ATP1A2 has not uncovered genetic association with common forms of the disease (Buono et al., 2000; Lohoff et al., 2005).

Another gene found on the BAC, Igsf8, encodes a protein involved in immune function. There is a coding SNP in the gene between B6 and D2 mice, His221Arg, and it warrants further consideration since there is evidence of causal links between immune function and seizure susceptibility (McKnight et al., 2005).

The final known gene present on BAC 157J4 is Pigm. Pigm encodes a transferase involved in the transport of carbohydrate moieties used to modify proteins related to G-protein function (Maeda et al., 2001). Thus, Pigm plays a role in signal transduction mechanisms. It is expressed in the brain and contains a coding SNP, Gln206His, between B6 and D2 mice. This potentially functional variation cannot yet be excluded as a putative seizure susceptibility factor particularly in light of a recent report documenting that a Pigm promoter mutation in humans causes a severe clinical phenotype that includes seizures (Almeida et al., 2006). Finally, Ensembl reports a novel, hypothetical gene at the 3′ end of the BAC (ENSMU00000066692). It is estimated to be 99 Kb in size with two exons that contain an open reading frame of 104 bp encoding a 34-amino acid peptide with no known sequence homology.

Nonsynonymous coding SNPs are not the only gene variants of potential functional significance. Promoter and intronic SNPs can also have important effects with respect to levels of gene expression. However, despite the fact that sequence variation exists between B6 and D2 mice in the putative 5′ regulatory regions of all genes present on the BAC, analysis of relative mRNA using RT-PCR showed similar levels between strains for primary candidates Kcnj9 and Kcnj10. An exception to this pattern is observed in hindbrain where D2 mice express higher levels of Kcnj10 compared to B6 (Table 1). It is unlikely that higher expression of Kcnj10 in D2 hindbrain explains the relative seizure susceptibility of this strain since overexpression of Kcnj10 in the BAC transgenic lines results in decreased susceptibility (increased MEST). Although it is possible that coordinated regulation of transcription and turnover could mask potential strain differences in message level, our results suggest that the classic B6 and D2 difference in seizure susceptibility is not related to a difference in the expression of either of these two genes. It may be more likely that the phenotypic effect is related to the expression of a functional protein variant resulting from a coding SNP as opposed to expression of a greater amount of a protein that has an inherent level of activity similar to the endogenous protein. An important caveat relates to the anatomical resolution of the methods employed in that a brain nucleus-specific difference would likely be difficult to detect. In addition to functional effects of variation in promoter and coding sequences, it is also possible that intronic SNPs could have biological significance by altering splicing mechanisms and resulting species of message; however, in the present study, Northern blot analyses using mRNA extracted from brains of B6 and D2 mice demonstrated single bands for Atp1a2, Kcnj9, and Kcnj10 (data not shown).

It must also be considered that the mechanism responsible for seizure protective effects in the BAC-transgenic mice may be distinct from that involved in differences in seizure phenotype between the parental strains. Thus, although Kcnj10 subunits may assemble to form homomeric channel complexes, in some parts of the brain they are coexpressed and coassembled with Kcnj16 (Wu et al., 2004; Butt and Kalsi 2006) such that overexpression of Kcnj10 alone could alter local channel formation and/or properties. Kcnj9 also contributes to the formation of heteromeric protein complexes (Jelacic et al., 2000); however, this subunit may be detrimental to the assembly of functional channels (Schoots et al., 1999) possibly via a mechanism whereby channel traffic is directed to lysosomes rather than the cell membrane (Ma et al., 2002). Thus, although we did not observe Kcnj9 or Kcnj10 expression differences between B6 and D2 mice, altered expression levels of BAC-derived genes might alter the stoichiometry that defines the function of inward rectifier potassium channels.

B6 and D2 mice differ with respect to numerous phenotypic traits, and as a result, they have been the subject of many studies since being established as inbred strains in the middle of the 20th Century. Remarkable B6 and D2 strain differences with respect to seizure susceptibility were noted shortly after their introduction into biomedical research by comparing their responses to convulsive auditory stimuli (Hall 1947). Whereas the greater relative susceptibility of D2 mice to audiogenic seizures is evident only when the animals are very young (about 3 weeks of age or less), these mice retain relative seizure susceptibility through adulthood when other experimental seizure paradigms are employed including chemoconvulsant and electroshock tests (Engstrom and Woodbury, 1988). Another paradigm that exploits this strain difference involves susceptibility to ethanol withdrawal seizures (Goldstein and Kakihana, 1975). Studies designed to map the chromosomal location of genes that mediate differences in seizure susceptibility between B6 and D2 mice have documented a complex polygenic determinism with QTL detected on many different chromosomes; however, two loci in particular have been detected repeatedly. Thus, QTL on distal chromosome 1 (∼90 cM) and mid-chromosome 4 (∼40 cM) have been reported in screens using kainic acid (Ferraro et al., 1997), alcohol withdrawal (Buck et al., 1997), pentylenetetrazol (Ferraro et al., 1999), and pentobarbital withdrawal (Buck et al., 1999). It is likely that any gene influencing such diverse seizure phenotypes will be of fundamental importance in controlling neuronal excitability. Moreover, identification of such genes could be relevant to many different types of human epilepsy, particularly common, multifactorial forms that comprise the vast majority of all cases.

Over the past several years, considerable evidence has accumulated to suggest that the gene on chromosome 4 that mediates differences between B6 and D2 mice in susceptibility to depressant withdrawal seizures is Mpdz, a PDZ domain-containing protein (Fehr et al., 2002; Shirley et al., 2004). PDZ proteins are adapter proteins that serve as central organizers of protein complexes at the plasma membrane (Sarkar et al., 2004). Interestingly, Kir channels have functional PDZ binding domains (Yoo et al., 2004) and analysis of Kir4.1 and Kir3.3 sequences with the Eukaryotic Linear Motif (ELM) resource for functional sites in proteins ( reveals consensus PDZ1 motifs in the extreme C-terminal region of both proteins. Whereas functional analysis of PDZ domains on Kir proteins suggests that Kir3.3 does not bind PDZ sequences (Nehring et al., 2000), a number of studies have shown that Kir4.1 does (Horio et al., 1997; Kurschner et al., 1998; Tanemoto et al., 2004; Tanemoto et al., 2005). Thus, given that previous results from analysis of locus–locus interactions detected epistasis between the distal chromosome 1 kainate seizure susceptibility QTL and the kainate susceptibility QTL on mid-chromosome 4 (Ferraro et al., 1997), it is possible that a potential functional relationship exists between Mpdz and the product of the chromosome 1 QTL.

In summary, we have used an in vivo BAC transgenesis strategy involving B6 genes from the critical interval of the chromosome 1 seizure QTL to protect against classical D2 seizure susceptibility. Given the modest increases in expression noted between transgenic and nontransgenic littermates for the BAC genes studied, and the 100% identity of amino acid sequence between B6 and D2 mice for Kcnj9 and Atp1a2, the nonsynonymous (Thr262Ser) coding SNP in Kcnj10 remains the variation that is most likely to define the seizure-related QTL on distal chromosome 1. Phenotypic effects caused by functional changes in populations of inward rectifier potassium channels resulting from an alteration of the balance of subunit expression cannot be ruled out nor can the effects of closely associated gene variants. These issues will be addressed by BAC recombineering (knock-in) experiments ongoing currently in our lab in which the B6 SNP in Kcnj10 will be changed to D2. We predict that full characterization of these transgenic models will provide new insight into mechanisms of seizure susceptibility.


  1. Top of page
  2. Abstract
  6. Acknowledgments

Acknowledgments:  We thank Glenn A. Doyle, Ph.D. for helpful discussion and critical reading of the manuscript and James Martin for excellent technical assistance. This work was supported by grant NS040554 (TNF) and is dedicated to the memory of Dr. Gregory T. Golden.


  1. Top of page
  2. Abstract
  6. Acknowledgments
  • Abiola O, Angel JM, Avner P, Bachmanov AA, Belknap JK, Bennett B, Blankenhorn EP, Blizard DA, Bolivar V, Brockmann GA, Buck KJ, Bureau JF, Casley WL, Chesler EJ, Cheverud JM, Churchill GA, Cook M, Crabbe JC, Crusio WE, Darvasi A, De Haan G, Dermant P, Doerge RW, Elliot RW, Farber CR, Flaherty L, Flint J, Gershenfeld H, Gibson JP, Gu J, Gu W, Himmelbauer H, Hitzemann R, Hsu HC, Hunter K, Iraqi FF, Jansen RC, Johnson TE, Jones BC, Kempermann G, Lammert F, Lu L, Manly KF, Matthews DB, Medrano JF, Mehrabian M, Mittlemann G, Mock BA, Mogil JS, Montagutelli X, Morahan G, Mountz JD, Nagase H, Nowakowski RS, O'Hara BF, Osadchuk AV, Paigen B, Palmer AA, Peirce JL, Pomp D, Rosemann M, Rosen GD, Schalkwyk LC, Seltzer Z, Settle S, Shimomura K, Shou S, Sikela JM, Siracusa LD, Spearow JL, Teuscher C, Threadgill DW, Toth LA, Toye AA, Vadasz C, Van Zant G, Wakeland E, Williams RW, Zhang HG, Zou F. (Complex Trait Consortium). (2003) The nature and identification of quantitative trait loci: a community's view. Nature Reviews: Genetics 4:911916.
  • Almeida AM, Murakami Y, Layton DM, Hillmen P, Sellick GS, Maeda Y, Richards S, Patterson S, Kotsianidis I, Mollica L, Crawford DH, Baker A, Ferguson M, Roberts I, Houlston R, Kinoshita T, Karadimitris A. (2006) A. Hypomorphic promoter mutation in PIGM causes inherited glycosylphosphatidylinositol deficiency. Nature Medicine 12:846851.
  • Bredt DS, Wang TL, Cohen NA, Guggino WB, Snyder SH. (1995) Cloning and expression of two brain-specific inwardly rectifying potassium channels. Proceedings of National Academy of the Sciences USA. 92:67536757.
  • Buck KJ, Metten P, Belknap JK, Crabbe JC. (1997) Quantitative trait loci involved in genetic predisposition to acute alcohol withdrawal in mice. Journal of Neuroscience 17:39463955.
  • Buck K, Metten P, Belknap J, Crabbe J. (1999) Quantitative trait loci affecting risk for pentobarbital withdrawal map near alcohol withdrawal loci on mouse chromosomes 1, 4, and 11. Mammalian Genome 10:431437.
  • Buono RJ, Ferraro TN, O’Connor MJ, Sperling MR, Abbey M, Finanger E, Lohoff F, Mulholland N, Berrettini WH. (2000) Lack of association between temporal lobe epilepsy and a novel polymorphism in the alpha 2 subunit gene (ATP1A2) of the sodium potassium transporting ATPase. American Journal of Medical Genetics 96:7983.
  • Buono RJ, Lohoff FW, Sander T, Sperling MR, O’Connor MJ, Dlugos DJ, Ryan SG, Golden GT, Zhao H, Scattergood TM, Berrettini WH, Ferraro TN. (2004) Association between variation in the human KCNJ10 potassium ion channel gene and seizure susceptibility. Epilepsy Research 58:175183.
  • Butt AM, Kalsi A. (2006) Inwardly rectifying potassium channels (Kir) in central nervous system glia: a special role for Kir4.1 in glial functions. Journal of Cellular and Molecular Medicine 10:3344.
  • Dissmann E, Wischmeyer E, Spauschus A, Pfeil DV, Karschin C, Karschin A. (1996) Functional expression and cellular mRNA localization of a G protein-activated K+ inward rectifier isolated from rat brain. Biochemical and Biophysical Research Communications 223:474479.
  • Engstrom FL, Woodbury DM. (1988) Seizure susceptibility in DBA and C57 mice: the effects of various convulsants. Epilepsia 29:389395.
  • Fehr C, Shirley RL, Belknap JK, Crabbe JC, Buck KJ. (2002) Congenic mapping of alcohol and pentobarbital withdrawal liability loci to a <1 centimorgan interval of murine chromosome 4: identification of Mpdz as a candidate gene. Journal of Neuroscience 22:37303738.
  • Ferraro TN, Golden GT, Smith GG, Schork NJ, St Jean P, Ballas C, Choi H, Berrettini WH. (1997) Mapping murine loci for seizure response to kainic acid. Mammalian Genome 8:200208.
  • Ferraro, TN, Golden GT, Snyder R, Laibinis M, Smith GG, Buono RJ, Berrettini WH. (1998) Genetic influences on electrical seizure threshold. Brain Research 813:207210.
  • Ferraro TN, Golden GT, Smith GG, St Jean P, Schork NJ, Mulholland N, Ballas C, Schill J, Buono RJ, Berrettini WH. (1999) Mapping loci for pentylenetetrazol-induced seizure susceptibility in mice. Journal of Neuroscience 19:67336739.
  • Ferraro TN, Golden GT, Smith GG, Longman RL, Snyder RL, DeMuth D, Szpilzak I, Mulholland N, Eng E, Lohoff FW, Buono RJ, Berrettini WH. (2001) Quantitative genetic study of maximal electroshock seizure threshold in mice: evidence for a major seizure susceptibility locus on distal chromosome 1. Genomics 75:3542.
  • Ferraro TN, Golden GT, Smith GG, DeMuth D, Buono RJ, Berrettini WH. (2002) Mouse strain variation in maximal electroshock seizure threshold. Brain Research 936:8286.
  • Ferraro TN, Golden GT, Smith GG, Martin JF, Lohoff FW, Gieringer TA, Zamboni D, Schwebel CL, Press DM, Kratzer SO, Zhao H, Berrettini WH, Buono RJ. (2004) Fine mapping of a seizure susceptibility locus on mouse Chromosome 1: nomination of Kcnj10 as a causative gene. Mammalian Genome 15:239251.
  • Flint J, Valdar W, Shifman S, Mott R. (2005) Strategies for mapping and cloning quantitative trait genes in rodents. Nature Reviews Genetics 6:271286.
  • Gao J, Wei Y, Huang Y, Liu D, Liu G, Wu M, Wu L, Zhang Q, Zhang Z, Zhang R, Liang C. (2005) The expression of intact and mutant human apoAI/CIII/AIV/AV gene cluster in transgenic mice. Journal of Biological Chemistry 280:1255912566.
  • Goldstein DB, Kakihana R. (1975) Alcohol withdrawal convulsions in genetically different populations of mice. Advances in Experimental Medicine and Biology 59:343352.
  • Hall CS.. (1947) Genetic differences in fatal audiogenic seizures between two inbred strains of house mice. Journal of Heredity 38:26.
  • Henkel A, Wei Z, Cohen DE, Green RM. (2005) Mice overexpressing hepatic Abcb11 rapidly develop cholesterol gallstones. Mammalian Genome 16:903908.
  • Horio Y, Hibino H, Inanobe A, Yamada M, Ishii M, Tada Y, Satoh E, Hata Y, Takai Y, Kurachi Y. (1997) Clustering and enhanced activity of an inwardly rectifying potassium channel, Kir4.1, by an anchoring protein, PSD-95/SAP90. Journal of Biological Chemistry 272:1288512888.
  • Ikeda A, Zheng QY, Zuberi AR, Johnson KR, Naggert JK, Nishina PM. (2002) Microtubule-associated protein 1A is a modifier of tubby hearing (moth1). Nature Genetics 30:401405.
  • Jelacic TM, Kennedy ME, Wickman K, Clapham DE. (2000) Functional and biochemical evidence for G-protein-gated inwardly rectifying K+ (GIRK) channels composed of GIRK2 and GIRK3. Journal of Biological Chemistry 275:3621136216.
  • Jiang Z, Dalton TP, Jin L, Wang B, Tsuneoka Y, Shertzer HG.  (2005) Toward the evaluation of function in genetic variability: characterizing human SNP frequencies and establishing BAC-transgenic mice carrying the human CYP1A1_CYP1A2 locus. Human Mutation 25:196206.
  • Kofuji P, Ceelen P, Zahs KR, Surbeck LW, Lester HA, Newman EA. (2000) Genetic inactivation of an inwardly rectifying potassium channel (Kir4.1 subunit) in mice: phenotypic impact in retina. Journal of Neuroscience 20:57335740.
  • Kosobud AE, Crabbe JC. (1990) Genetic correlations among inbred strain sensitivities to convulsions induced by 9 convulsant drugs. Brain Research 526:816.
  • Kurschner C, Mermelstein PG, Holden WT, Surmeier DJ. (1998) CIPP, a novel multivalent PDZ domain protein, selectively interacts with Kir4.0 family members, NMDA receptor subunits, neurexins, and neuroligins. Molecular and Cellular Neuroscience 11:161172.
  • Lenzen KP, Heils A, Lorenz S, Hempelmann A, Hofels S, Lohoff FW, Schmitz B, Sander T. (2005) Supportive evidence for an allelic association of the human KCNJ10 potassium channel gene with idiopathic generalized epilepsy. Epilepsy Research 63:113118.
  • Lesage F, Duprat F, Fink M, Guillemare E, Coppola T, Lazdunski M, Hugnot JP. (1994) Cloning provides evidence for a family of inward rectifier and G-protein coupled K+ channels in the brain. FEBS Letters 353:3742.
  • Li L, Head V, Timpe LC. (2001) Identification of an inward rectifier potassium channel gene expressed in mouse cortical astrocytes. Glia 33:5771.
  • Light PE, Bladen C, Winkfein RJ, Walsh MP, French RJ. (2000) Molecular basis of protein kinase C-induced activation of ATP-sensitive potassium channels. Proceedings of National Academy of the Sciences USA 97:90589063.
  • Lohoff FW, Ferraro TN, Sander T, Zhao H, Dahl JP, Berrettini WH, Buono RJ. (2005) No association between common variations in the human alpha 2 subunit gene (ATP1A2) of the sodium-potassium-transporting ATPase and idiopathic generalized epilepsy. Neuroscience Letters 382:3338.
  • Ma D, Zerangue N, Raab-Graham K, Fried SR, Jan YN, Jan LY. (2002) Diverse trafficking patterns due to multiple traffic motifs in G protein-activated inwardly rectifying potassium channels from brain and heart. Neuron 33:715729.
  • Maeda Y, Watanabe R, Harris CL, Hong Y, Ohishi K, Kinoshita K, Kinoshita T. (2001) PIG-M transfers the first mannose to glycosylphosphatidylinositol on the lumenal side of the ER. The EMBO Journal 20:250261.
  • McKnight K, Jiang Y, Hart Y, Cavey A, Wroe S, Blank M, Shoenfeld Y, Vincent A, Palace J, Lang B. (2005) Serum antibodies in epilepsy and seizure-associated disorders. Neurology 65:17301736.
  • Moseley AE, Lieske SP, Wetzel RK, James PF, He S, Shelly DA, Paul RJ, Boivin GP, Witte DP, Ramirez JM, Sweadner KJ, Lingrel JB. (2003) The Na, K-ATPase alpha 2 isoform is expressed in neurons, and its absence disrupts neuronal activity in newborn mice. Journal of Biological Chemistry 278:53175324.
  • Nehring RB, Wischmeyer E, Doring F, Veh RW, Sheng M, Karschin A. (2000) Neuronal inwardly rectifying K(+) channels differentially couple to PDZ proteins of the PSD-95/SAP90 family. Journal of Neuroscience 20:156162.
  • Neusch C, Rozengurt N, Jacobs RE, Lester HA, Kofuji P. (2001) Kir4.1 potassium channel subunit is crucial for oligodendrocyte development and in vivo myelination. Journal of Neuroscience 21:54295438.
  • Nishida M, MacKinnon R. (2002) Structural basis of inward rectification: cytoplasmic pore of the G protein-gated inward rectifier GIRK1 at 1.8 A resolution. Cell 111:957965.
  • Noebels JL. (2003) Exploring new gene discoveries in idiopathic generalized epilepsy. Epilepsia 44 (Suppl 2):1621.
  • Sarkar D, Boukerche H, Su ZZ, Fisher PB. (2004) mda-9/syntenin: recent insights into a novel cell signaling and metastasis-associated gene. Pharmacology and Therapeutics 104:101115.
  • Schoots O, Wilson JM, Ethier N, Bigras E, Hebert TE, Van Tol HH. (1999) Co-expression of human Kir3 subunits can yield channels with different functional properties. Cell Signal 11:871883.
  • Shang L, Lucchese CJ, Haider S, Tucker SJ. (2005) Functional characterisation of missense variations in the Kir4.1 potassium channel (KCNJ10) associated with seizure susceptibility. Brain Research and Molecular Brain Research 139:178183.
  • Shirley RL, Walter NA, Reilly MT, Fehr C, Buck KJ. (2004) Mpdz is a quantitative trait gene for drug withdrawal seizures. Nature Neuroscience 7:699700.
  • Tanemoto M, Abe T, Ito S. (2005) PDZ-binding and di-hydrophobic motifs regulate distribution of Kir4.1 channels in renal cells. Journal of American Society of Nephrology 16:26082614.
  • Tanemoto M, Abe T, Onogawa T, Ito S. (2004) PDZ binding motif-dependent localization of K+ channel on the basolateral side in distal tubules. American Journal of Physiology: Renal Physiology 287:F1148F1153.
  • Torrecilla M, Marker CL, Cintora SC, Stoffel M, Williams JT, Wickman K. (2002) G-protein-gated potassium channels containing Kir3.2 and Kir3.3 subunits mediate the acute inhibitory effects of opioids on locus ceruleus neurons. Journal of Neuroscience 22:43284334.
  • Underhill DA, Canfield VA, Dahl JP, Gros P, Levenson R. (1999) The Na,K-ATPase alpha4 gene (Atp1a4) encodes a ouabain-resistant alpha subunit and is tightly linked to the alpha2 gene (Atp1a2) on mouse chromosome 1. Biochemistry 38:1474614751.
  • Wischmeyer E, Karschin A. (1996) Receptor stimulation causes slow inhibition of IRK1 inwardly rectifying K+ channels by direct protein kinase A-mediated phosphorylation. Proceedings of National Academy of Sciences USA 93:58195823.
  • Wolford JK, Hanson RL, Kobes S, Bogardus C, Prochazka M. (2001) Analysis of linkage disequilibrium between polymorphisms in the KCNJ9 gene with type 2 diabetes mellitus in Pima Indians. Molecular Genetics and Metabolism 73:97103.
  • Wu J, Xu H, Shen W, Jiang C. (2004) Expression and coexpression of CO2-sensitive Kir channels in brainstem neurons of rats. Journal of Membrane Biology 197:179191.
  • Yoo D, Flagg TP, Olsen O, Raghuram V, Foskett JK, Welling PA. (2004) Assembly and trafficking of a multiprotein ROMK (Kir 1.1) channel complex by PDZ interactions. Journal of Biological Chemistry 279.