Human Molecular Genetics

This Article Abstract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (16) Request Permissions Google Scholar Articles by Yang, Y. Articles by Frankel, W. N. Search for Related Content PubMed PubMed Citation Articles by Yang, Y. Articles by Frankel, W. N. Social Bookmarking          What's this?

Human Molecular Genetics, 2003, Vol. 12, No. 9 975-984
DOI: 10.1093/hmg/ddg118
© 2003 Oxford University Press

Spontaneous deletion of epilepsy gene orthologs in a mutant mouse with a low electroconvulsive threshold

Yan Yang¹, Barbara J. Beyer¹, James F. Otto², Timothy P. O'Brien¹, Verity A. Letts¹, H. Steve White² and Wayne N. Frankel^1,*

¹The Jackson Laboratory, 600 Main Street, Bar Harbor, ME 04609, USA and ²Anticonvulsant Drug Development Program, Department of Pharmacology and Toxicology, University of Utah, Salt Lake City, Utah, USA

Received December 11, 2002; Revised February 23, 2003; Accepted March 2, 2003

	ABSTRACT

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
The electroconvulsive threshold (ECT) test has been used extensively to determine the protection conferred by antiepileptic drug candidates against induced seizures in rodents. Despite its clinical relevance, the potential of ECT to identify mouse epilepsy models in genetic studies has not been thoroughly assessed. We adopted the ECT test to screen the progeny of ethylnitrosourea treated male C57BL/6J mice. In a small-scale screen, several mutant lines conferring a low threshold to ECT minimal clonic seizures were mapped to the telomeric region of mouse chromosome 2 in independent founder families. This high incidence was suggestive of a single spontaneous event that pre-existed in the founders of mutagenized stock. Genetic and physical mapping led to the discovery that several lines shared a single mutation, Szt1 (seizure threshold-1), consisting of a 300 kb deletion of genomic DNA involving three known genes. Two of these genes, Kcnq2 and Chrna4, are known to be mutated in human epilepsy families. Szt1 homozygotes and heterozygotes display similar phenotypes to those found in the respective Kcnq2 knockout mutant mice, suggesting that Kcnq2 haploinsufficiency is at the root of the Szt1 seizure sensitivity. Our results provide a novel genetic model for epilepsy research and demonstrate that the approach of using ECT to study seizures in mice has the potential to lead to the identification of human epilepsy susceptibility genes.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
Epilepsy, characterized by recurrent seizures resulting from abnormal, synchronized discharges of neurons in the brain, affects up to 1% of the population worldwide (1). A focus shift in epilepsy research from targeting symptoms to strategies for prevention and disease modification has recently been proposed (2). To move epilepsy research forward, better animal models are needed in that they offer valuable resources for investigators to characterize seizures in the whole organism and to test the efficacy of new therapies.

Electroshock-induced convulsions in mice provide a reliable model for the study of human seizures (3). Electroconvulsive responses and many features of generalized tonic–clonic seizures are conserved in mammals, indicating similar neural substrates or pathways (4). Many of the currently available antiepileptic drugs (AEDs) that prevent specific human seizures were discovered on the basis of their ability to block electroshock-induced seizures in mice and rats. For these reasons, the electroconvulsive threshold (ECT) test is an integral part of the ongoing development program in search of new AEDs in the USA and abroad (5,6). Typical ECT procedures involve passing an electric current of short duration through the brain. The most common form of testing seizure susceptibility in mice by ECT is to use high-frequency pulses (e.g. 60 Hz) through transcorneal electrodes. The endpoints are minimal seizures involving repetitive movement of the jaws and multiple brief contractions of the forelimb muscles, and maximal seizures involving a rigid, violent muscular contraction fixing the hindlimbs in an extended position (tonic hindlimb extension, THE).

As a prelude to using the ECT procedure in genetic research, we previously described ECT baselines for several common inbred mouse strains (7), including C57BL/6J. In the present study we explore the use of ECT in forward genetic studies by screening the progeny of ethylnitrosourea (ENU) treated male C57BL/6J (B6) mice for lowered seizure threshold. We describe the origin, phenotypic characteristics and molecular lesion of a new mutation identified from this screen, seizure threshold 1 (Szt1), which confers a low seizure threshold in a dominant fashion. In the context of these results, we also discuss the relevance of the experimental approach for discovering new mouse models for epilepsy research.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
Origin of Szt1 mutant mice
Szt1 was identified in a mutation screen where B6 mice were mutagenized by ENU. Both first-generation (G₁) and third generation (G₃) mice obtained through a three-generation breeding scheme were tested to identify dominant and recessive mutations, respectively. Based on electroconvulsive thresholds obtained previously from wild-type B6 mice (7), 139 G₁ mice (from 22 independent G₀ founders) and 333 G₃ mice (from 19 independent G₁ families) were tested for their sensitivity to electrically-evoked minimal clonic seizures. Szt1 was one of five phenotypic deviants from the G₁ screen that displayed a low threshold for minimal clonic seizure. Further matings of affected mice to untreated B6 mice confirmed that the phenotype was heritable and that the mode of inheritance was dominant for each of these five deviants. For Szt1, eight out of the 24 mice derived from matings between an affected mouse and a normal B6 mouse were susceptible to minimal clonic seizures at the CC₃ (the estimated current setting for 3% of the mice to reach a seizure response) for B6, suggesting a penetrance of 70% under our ECT testing conditions. Affected mice were viable, fertile and did not show spontaneous or handling provoked seizures. Gross histological examination did not reveal abnormalities in heart, liver, spleen, lung, kidney, gastric-intestine tract, muscle and gonads (data not shown).

Szt1 maps to distal chromosome 2 and homozygotes exhibit perinatal lethality
In order to map the dominant seizure phenotype, we outcrossed B6-Szt1/+ mice to BALB/cByJ (BALB) mice, which have a similar seizure threshold as B6 (7), and backcrossed affected F₁ hybrids to normal BALB mice. Initially, about 20 N₂ affected mice (i.e. those which exhibited a minimal clonic seizure on either of two tests, see testing procedures under Materials and Methods) and 20 unaffected mice (no seizures on either test), were used in a genome scan. We detected linkage with microsatellite markers on telomeric chromosome 2 (data not shown). Ultimately we tested more than 300 N₂ mice for ECT-induced minimal clonic seizures and found strong evidence for linkage (Fig. 1A). Interestingly, two additional dominant mutations from independent G₀ mice were also mapped to distal chromosome 2 (subsequent characterization confirmed that these two additional families carried the same genetic defect as Szt1, see below). Given the small size of the pilot screen (only 139 G₁ mice screened) and the mutagenizing rate of ENU of approximately 1/1000 mutation per locus per generation (8,9), we hypothesized that the Szt1 mutation had arisen spontaneously in the B6 stock prior to ENU mutagenesis experiments.

View larger version (27K): [in this window] [in a new window]   Figure 1. Szt1 is a large genomic deletion. (A) Map position of Szt1 on chromosome 2. N2 (300) mice were genotyped with the indicated SSLP markers and phenotyped by ECT. The complete data set was fed into MapManager QTX (vers. b17 obtained from http://mapmgr.roswellpark.org/mmQTX.html) and the following chart was generated as a graphic output, with a conversion of likelihood ratio statistic to LOD score as described in the program manual. (B) Physical map of the deletion in Szt1 mice. This large genomic deletion (300 kb) is located at the distal end of mouse chromosome 2. The centromeric breakpoint is about 120 kb upstream of Predicted Gene 1 (PG1) and 6 kb downstream of the mouse homolog of human BHLHB4 gene, the telomeric breakpoint is within intron 13 of the Kcnq2 gene. The estimated physical position of the breakpoints was obtained from the Ensembl Mouse Genome Server (www.ensembl.org/Mus_musculus). The deletion affects three known genes Arf1-gap, Chrna4 and Kcnq2, encoding adenosine ribosylation factor-1 GTPase activating protein, 4 neuronal nicotinic acetylcholine receptor subunit and an M-current potassium channel subunit, respectively. Database searches revealed that at least four predicted genes matched to the sequence in this region. All these four predicted genes have been identified as putative homologs of human genes by reciprocal BLAST analysis. PG1 was predicted to encode a novel protein similar to the human NY-REN-2 antigen. PG2 was predicted to encode the mouse homolog of human inhibitor-of-apoptosis protein livin. PG3 matched to the mouse RIKEN cDNA C030019F02 gene. PG4 was predicted to encode a member of the collagen alpha 1 chain family and part of PG4 matched to the mouse RIKEN cDNA 1700051I12 gene. (C) Detection of the Szt1 allele by a 3-primer PCR assay. Genotyping PCR reaction using three primers was carried out as described in Materials and Methods. Note Szt1 homozygote embryo only had a 390 nt PCR product (lane 4) and the wild-type embryo only had the 645 nt product (lane 2). Both bands were present in the heterozygote embryo (lane 3) and adults (lanes 6–8 and 10). Lane 1, X174/Hae III markers; lanes 5, 9 and 11, tail DNA from wild type adult mice.

 
Szt1 heterozygotes from the mapping cross were intercrossed to generate Szt1 homozygotes in order to evaluate a possible Szt1 homozygous phenotype. Interestingly, we did not find any of the expected 25% Szt1 homozygotes at weaning age. Further studies revealed that at embryonic days 11.5 and 18.5, Mendelian ratios of homozygous embryos were observed. Szt1 homozygotes were indeed born alive and were able to survive for a few hours. While homozygotes were alive, we did not observe obvious signs of convulsion nor were abnormalities apparent. However, within the first 12 h after birth, they began gasping, turned blue and eventually died on the first day. Histological studies indicated that they died of pulmonary atelectasis (failure of the alveoli's expansion from gas; data not shown).

Szt1 is a large genomic deletion encompassing human epilepsy gene orthologs
The Szt1 region of mouse chromosome 2 corresponds to 20q13 in the human genome, containing the Kcnq2 potassium channel gene and Chrna4, which encodes the ₄ nicotinic acetylcholine receptor subunit. Several mutations in Kcnq2 and Chrna4 have been identified in human benign familial neonatal convulsions (BFNC) patients (10,11) and human autosomal dominant nocturnal frontal lobe epilepsy (ADNFLE) patients (12), respectively. Further, both epilepsy genes' murine orthologs have been inactivated through targeted mutagenesis via homologous recombination in mouse embryonic stem (ES) cells. Chrna4 null mice showed a heightened basal level of anxiety-like behavior (13) and more recent pharmacological studies suggested that Chrna4 knock-out mice were sensitive to pentylenetetrazole (PTZ) and kainic acid (KA) induced seizure (14). Kcnq2 heterozygotes developed normally and were sensitive to PTZ-induced seizure. However Kcnq2 null mice died at birth due to pulmonary atelectasis (15).

Because of the phenotypic similarities, we decided to check the potential involvement of Kcnq2 in Szt1 mice by designing PCR primers to amplify each exon from DNA samples of Szt1 homozygote, Szt1 heterozygote and wild-type embryos. Although the mouse Kcnq2 gene has 17 coding exons (16), we were unable to detect exons 14, 15, 16 and 17 in Szt1 homozygotes by PCR, thus indicating Szt1 was a deletion and the telomeric breakpoint was within intron 13 of the Kcnq2 gene. In order to identify the centromeric breakpoint, we designed PCR primers to amplify a DNA fragment every 25 kb based on the distal end of mouse chromosome 2 sequence (www.ensembl.org/Mus_musculus). This strategy led to the identification of a roughly 300 kb deletion in Szt1 mice. As shown in Figure 1B, this deletion affected three known genes and as many as four predicted genes or ESTs. Of the three known genes, Kcnq2 was the only one that was partially deleted. The other two, Chrna4 and Arf1-Gap, which encode adenosine diphosphate ribosylation factor-1 GTPase-activating protein (17), were completely deleted. The fact that Szt1 homozygote pups developed to full-term and were able to survive a few hours after birth without apparent developmental defects suggests that none of the affected genes plays a critical role in early mouse embryonic development.

Based on the sequence information around the deletion breakpoints in Szt1 mice, we designed a 3-primer PCR assay to detect the Szt1 allele and wild-type allele. As shown in Figure 1C, a common primer and a deletion-specific primer generated a 390 nt PCR product whereas the same common primer and a third primer amplified a 645 nt wild-type allele product in one genotyping reaction. Using this assay, we screened the 22 DNA samples collected from mice used in the original ENU study. Three out of 22 G₀ mice carried the Szt1 allele. Each of these three G₀ mice gave rise to multiple families that were found to have a low electroconvulsive threshold (data not shown). From these data we concluded that Szt1 indeed arose spontaneously on B6 background prior to ENU mutagenesis. We also screened the B6 production stock at The Jackson Laboratory and various sources of DNA samples from B6 mice, but we did not detect the presence of Szt1 allele, thus suggesting that Szt1 appeared to be an isolated spontaneous event that occurred in a small expansion stock.

Reduced Kcnq2 transcript expression from the Szt1 allele
The telomeric breakpoint of the Szt1 allele is within the 13th intron of Kcnq2, thereby deleting the last four exons which encode the cytoplasmic tail of this protein. We performed mRNA analysis to investigate how the deletion affected the amount of normal Kcnq2 expressed in Szt1/+ heterozygotes. Despite several attempts with different Kcnq2 probes, strong cross-hybridization to the ribosomal RNA was observed. Nevertheless, when poly A+ RNA from E18.5 embryos was used, we could detect a major band around 8.5 kb in wild-type, corresponding to that reported in the literature (16), and a less intense band in Szt1 heterozygote. This band was absent in the Szt1/Szt1 homozygote (Fig. 2). Using ß-actin as a loading control, we determined that the Szt1 heterozygote only had 71% of the Kcnq2 transcript of the wild-type embryo. Further, we performed real-time PCR analysis using a pair of N-terminal primers that covered all the alternatively spliced variants of Kcnq2 (16). Again, Szt1 heterozygote was found to have 63% of the Kcnq2 transcript in the wild-type littermate control (data not shown). Taken together, these data suggest that Szt1 heterozygotes have decreased level of Kcnq2 RNA transcript.

View larger version (53K): [in this window] [in a new window]   Figure 2. The Szt1 deletion interferes with the expression of Kcnq2. Top half of a northern blot of E18.5 brain poly-A+ RNA from Szt1/Szt1, Szt1/+ and +/+. An N-terminal Kcnq2 specific probe and a mouse ß-actin probe were used sequentially. The size markers shown were from Ambion's RNA Millennium size markers.

 
Reduced seizure threshold in Szt1 heterozygous mice
We developed a partial congenic strain (N₆) on the BALB background in order to further evaluate the seizure threshold of Szt1 heterozygotes. Female Szt1 mutants and +/+ controls at 7 weeks of age were used to generate a response curve for minimal clonic and maximal tonic hindlimb extension seizures induced by ECT. As shown in Figure 3, for each testing paradigm, Szt1's response curve was significantly shifted towards the left, indicating a much reduced seizure threshold. For example, the median response level of congenic Szt1/+ mice (6.6 mA; Fig. 3 legend) was much closer to that of the highly sensitive FVB/NJ strain (6.1 mA, 95% CI 5.7–6.4; data from 7) when compared with +/+ mice (8.3 mA; Fig. 3 legend). A similar decrease in seizure threshold was also seen in male Szt1 mice (data not shown).

View larger version (22K): [in this window] [in a new window]   Figure 3. Szt1 heterozygous mice are hypersensitive to ECT. Thirty female Szt1/+ and 32 female littermate +/+ BALB. B6-Szt1 congenic mice were tested for ECT over a range of electric current settings for both minimal and maximal seizure and each ECT response was recorded. As stated in the text, if the rhythmic movement of the jaws and forelimbs was displayed, a minimal seizure was scored while the occurrence of tonic hindlimb extension was recorded as a maximal seizure. The data were analyzed in the computer program MiniTab (Minitab Inc.) and a response curve was generated using the log-Probit procedure. The median response level (CC50) for minimal seizures were as follows +/+=8.3 mA (95% CI 8.0–8.7), Szt1/+=6.6 mA (95% CI 6.4–6.9); for maximal seizures +/+=14.2 mA (95% CI 13.6–15), Szt1/+=12.4 mA (95% CI 11.7–13.0).

 
We also tested the response of Szt1 heterozygotes to PTZ-induced tonic–clonic seizures. PTZ is a non-competitive GABA_A antagonist that lowers neuronal inhibition. Each mouse received a single subcutaneous injection of 50 mg/kg body weight of PTZ and was monitored continuously for 30 min. The number of Szt1 mice experiencing generalized convulsive seizures after the PTZ injection was significantly higher than that observed in wild-type littermates, indicating a lowered seizure threshold or heightened neuroexcitability in the mutant mice (Table 1).

View this table: [in this window] [in a new window]   Table 1. Sensitivity of PTZ-induced seizure. We obtained the F1 hybrid from affected G1 B6 mice outcrossed to BALB mice. Then affected F1 hybrids were backcrossed to normal B6 mice to generate N2 mice that were tested for sensitivity to subcutaneous PTZ injection induced seizure as described in the Results (35). More than 60 mice were tested and the two-tail Fisher exact test results (P=0.0122) indicated the hypersensitivity to PTZ in Szt1 heterozygotes compared with wild-type littermates. This group contained both female and male mice and each genotype had similar number of mice from both sexes (Szt1, 14 females and 12 males; wild type, 21 females and 18 males)

 
Hippocampal defect of Szt1 heterozygotes
Given the seizure susceptibility, we examined hematoxylin and eosin (H–E) stained brain sections of the Szt1 heterozygote. Except for the hippocampus, no obvious neuronal defect was apparent in the brain. Gross examination of the hippocampus revealed a slightly malformed CA3 subfield, in particular the region between the two dentate gyrus blades—in Szt1 heterozygous mice compared with wild-type littermates. The continuous architecture of the pyramidal neurons in CA3c region appeared disrupted in Szt1 heterozygous mice. Although this subtle phenotype appeared in more than 10 Szt1 heterozygotes when compared with wild-type littermates ranging from 6 weeks to 6 months old (data not shown), the result is considered preliminary because of the variability commonly observed in the dentate hilar region.

To further evaluate the potential defects in other regions of the hippocampus, we took advantage of transgenic mice where certain subsets of neurons, including representatives of CA1, CA2 and CA3 subfields, have been selectively labeled with yellow fluorescent protein (YFP) (18). The expression of YFP in neural cells is driven by a modified version of the regulatory elements from the Thy1 gene and the pattern of expression is heritable (18). Szt1 heterozygotes and B6 Thy1-YFP line H mice were bred to generate Szt1 and YFP double-positive mice. In this transgenic line, a subset of neurons in layer V of the cerebral cortex and a subset of hippocampal pyramidal neurons express YFP (18) and this technique has been shown to offer greater power to detect hippocampal defect than H–E staining (19). In contrast to wild-type littermates, Szt1 heterozygotes appeared to contain fewer YFP-labeled pyramidal neuron cell bodies throughout CA1, CA2 and CA3 subfields (n=4, Fig. 4). Consistent with this paucity of cell bodies, fewer YFP-labeled dendritic structures of the CA1 pyramidal neurons were observed in Szt1 brain—a difference that was consistent throughout the all hippocampal sections (data not shown), ruling out the possibility of an artifact due to oblique sectioning. We determined the number of YFP positive pyramidal neurons from Szt1 heterozygotes and their wild-type littermates from matched brain sections and found that Szt1 hippocampus contained about 40% less YFP-labeled pyramidal cells (P<0.001, n=4; Table 2). We also counted the number of YFP-positive cortex layer V neurons but did not detect significant difference between Szt1 heterozygotes and wild-type controls (Table 2), suggesting the difference was hippocampus-specific. Overall, these results indicate that the Szt1/+ hippocampus is missing a significant fraction of pyramidal neurons and that in some subfields this difference can be visualized qualitatively by using the YFP transgenic mice.

View larger version (96K): [in this window] [in a new window]   Figure 4. Hippocampal defect in Szt1 heterozygotes. (A) Coronal sections of mouse brains expressing YFP. A male Szt1 YFP transgene double heterozygote, 12 weeks of age, is shown in panel A1, wild-type YFP transgene heterozygote control is shown in panel A2. Higher magnification images focusing on CA2 and CA3 hippocampal subfields are shown in panels A5 and A6 for Szt1 and control, respectively. Higher magnification images focusing on CA1 subfield are shown in panels A3 and A4. The cellular nature of the YFP positive neurons was confirmed by DAPI counterstaining (data not shown). Note there were fewer neurons and dendritic structures labeled by YFP in Szt1 brain compared with control brain (arrows, representative of results seen on multiple section planes, n=4).

 

View this table: [in this window] [in a new window]   Table 2. The Szt1/+ hippocampus contains fewer YFP positive pyramidal neuronsa

 
Body size
During the course of our work, we noticed that Szt1 mice always weighed less than their wild-type littermates. To confirm this body weight difference, we monitored the body weight between the age of 4 weeks and 12 weeks of more than 100 mice. Both female and male Szt1 heterozygotes consistently weighed 4–8% less than their wild-type littermates (Fig. 5). Although metabolic monitoring did not reveal significant difference in food and water intake, oxygen consumption, carbon dioxide production, heat generation or rest activity between Szt1 heterozygotes and the wild-type littermates (YY and WNF, unpublished results), body length (nasal–anal length) measurements showed that Szt1 mice were 3–5% shorter than wild-type controls (data not shown), indicating Szt1 heterozygotes are smaller.

View larger version (17K): [in this window] [in a new window]   Figure 5. Szt1/+ mice are smaller than littermate controls. We monitored the body weight of more than 100 mice that came from matings between Szt1/+ B6 isocongenic mice. Single-caged mice were excluded from this study. Each mouse was weighed at 4, 6, 8, 10 and 12 weeks of age. At each time point, both female and male Szt1 heterozygotes weigh less than the wild-type littermates (Student t-test, P<0.01, assuming two-tailed distribution and two-sample equal variance). The average weight of each genotype at a fixed time point was plotted against time to generate a growth curve using MS-Excel program.

 

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
Recent progress in genome-wide ENU mutagenesis in the mouse has provided geneticists with new tools to study mammalian gene function (20–23). ENU induces point mutations in the genome, and therefore can cause a wide range of effects on the mutated gene. However, to be an effective approach towards developing new animal models for medical research, ENU mutagenesis needs to be combined with a robust and relevant phenotyping assay. Here, we report on the identification of a mouse epilepsy model through a unique assay—the ECT. ECT is an industry standard method for assessing the efficacy of antiepileptic drugs but its potential use as a gene discovery tool in large-scale genetic screens has not be explored. The results presented here show that the genetic defect in one such mouse epilepsy model, Szt1, detected by ECT, includes genes implicated in human familial epilepsies, thus validating ECT as a useful screening paradigm. Our aim was to identify seizure disease loci in mice that would advance our understanding of human epilepsy. The identification of candidate genes in mouse epilepsy models has facilitated the discovery of human disease genes, as shown for at least one case of idiopathic generalized epilepsy (24–26). Detailed characterization of the associated mouse models (27) can potentially extend their relevance to the human condition. Further, advances in epilepsy research have been predicted on the availability of model systems that attempt to reproduce the human condition. Acute seizure models such as maximal electroshock (MES) or PTZ-induced seizures have been widely used in drug discovery programs (28,29). However, these models employ normal rodents and do not take into account the influence of genetic heterogeneity on seizure threshold and/or response. To this end, the mutant mice described here are sensitive to electrically and chemically induced seizure, and therefore may represent an alternative model for the screening of new antiepileptic drugs.

For those engaged in forward genetics/phenotype-driven screen in model organisms, the nature of the genetic defect in Szt1 is of particular interest. Although we set out to pursue an ENU-induced point mutation that would confer seizure susceptibility, we discovered a spontaneous mutation that arose in the parental mouse stock prior to ENU treatment. This mutation causes perinatal lethality in homozygotes, a low threshold to electroshock- and PTZ-induced seizures and an abnormal hippocampus in heterozygotes. If it were not for the ECT screen, none of the potentially important phenotypes of Szt1 mice would have been so easily detected. Spontaneous mutations with multiple, readily observable phenotypes have contributed a great deal to our understanding of molecular mechanisms of neurological disease, but these mutations were discovered because they produce at least one obvious phenotype. Our results remind us that less obvious (but not necessarily less important) spontaneous mutations can arise in inbred mouse strains and can ‘contaminate’ founder stocks used for mutagenesis screens. This phenomenon may be more of a nuisance than is generally appreciated in the genetics community. Although it is routine to determine inbred strain baselines prior to embarking on a mutagenesis screen, such studies are usually done on a modest number of mice. Moreover, it is not common knowledge that many inbred strains, such as B6, are not necessarily maintained from ‘clonal’ stocks but rather from several independent pedigreed lines separated by several generations. It is these pedigreed lines that are expanded to supply groups of mice to researchers for their studies. Thus, an inbred strain that contributes to a baseline study done at one time of year may not have exactly the same genetic composition as that used in a screen done later in the same year because it may come effectively from a different subline. Although the spontaneous mutation rate is very low (approximately 2x10^-6 mutations/locus/generation) (30) and the spontaneous mutation conferring an overt phenotype will be noticed readily and set aside, it is likely that mutations with less obvious phenotype (such as Szt1) could populate a sizable fraction of the colony because of genetic drift between pedigrees comprising an inbred strain. Although it was fortunate that Szt1 was not present in any of The Jackson Laboratory's C57BL/6J foundation stocks, our results nevertheless suggest that it would be a good idea for researchers to check baseline data routinely from their respective inbred strains, and also to become familiar with the breeding schemes used by their animal suppliers. Should genetic drift become more of a problem as researchers assay more subtle phenotypes, it may become necessary for animal suppliers to routinely generate their colonies from a standard set of frozen gametes.

Although the genetic defect in Szt1 affects three known genes and as many as four additional genes, several aspects of the phenotype caused by this large genomic deletion are reminiscent of those displayed by Kcnq2 knock-out mice (perinatal lethality in null and sensitivity to PTZ in heterozygous) (15). The second epilepsy gene, Chrna4, has also been inactivated in the mouse, however, the Chrna4 null mice were viable and fertile and had normal brain histology and home-cage behavior (13). A recent report indicated that Chrna4 knock-out mice were sensitive to chemoconvulsant induced seizures, but testing results from Chrna4 heterozygotes were not available (14), barring a direct comparison with the Szt1 heterozygote. The third known gene in the deleted region, Arf1-Gap, encodes a Golgi complex protein that has been shown to regulate vesicle transport in vitro (17,31). Studies of in vivo function of this gene such as germline inactivation in the mouse have not been reported. On the other hand, Kcnq2 null mice died at birth due to collapsed lung and the heterozygotes were sensitive to PTZ-induced seizures (15). It has been proposed that KCNQ2 protein may regulate the pulsive catecholamine release from adrenal chromaffin cells at birth. Disruption of Kcnq2 gene may therefore abolish the normal initiation and maintenance of respiration in newborn mice, leading to respiratory stress (15). Although these investigators did not report abnormal findings in body weight and hippocampus, subtle phenotypes such as those found in Szt1 could be overlooked easily.

Mutations in the Kcnq2 gene cause BFNC in humans and a unique feature of this syndrome is the age-related onset. Affected infants experience convulsions during the first few weeks of life and then remit (10,11). Up to 85% affected infants remain seizure-free throughout the rest of their life and only 15% will experience a recurrence later in life (10,11). Interestingly, we did not observe this age-dependent seizure phenotype in Szt1 mice. Szt1 homozygotes die at birth due to lung defect; prior to death, they did not show signs of convulsion. Szt1 heterozygote pups are indistinguishable from the wild-type pups. We did not observe the occurrence of spontaneous seizure or handling-provoked seizure during their lifespan. We detected the hippocampal defect in Szt1 heterozygote as early as 6 weeks of age and the abnormalities did not seem to exaggerate as they aged, suggesting a developmental origin.

Kcnq2 and its close homolog, Kcnq3, encode subunits of a unique type of potassium channel called the ‘M-channel’ (32). M-channels are voltage-dependent potassium channels with slow opening and closing properties. Because the ion channels responsible for the propagated action potential operate much faster, M-channels favor the firing of a single action potential but inhibit the repetitive firing of action potentials (33). Strong immunoreactivity has been localized to neurons controlling the excitability and synchronization of hippocampal pyramidal cells (34), providing a connection between dysfunction of M-channels and seizure disorders in human patients. The fact that there was a roughly 30–40% message reduction of Kcnq2 transcript in the heterozygous suggests that the C-terminal genomic deletion in Szt1 destabilized the Kcnq2 RNA transcript and inhibited the expression of the protein. Taken together, we propose that the sensitivity to electroshock and chemoconvulsant-induced seizure in Szt1 heterozygotes may be due to Kcnq2 haploinsufficiency (reduction of functional KCNQ2 protein).

In summary, we sought to identify new mouse mutants with seizure disorders through a combination of ENU mutagenesis and ECT, a clinically relevant seizure model. Surprisingly, in a small-scale, pilot screen we uncovered a spontaneous mutation arising prior to mutagenesis. This mutation was a large genomic deletion at the distal mouse chromosome 2, affecting genes implicated in human hereditary epilepsy syndromes. The fact that the ECT screen provided us with a mutant mouse harboring defects in human seizure susceptibility gene orthologs encourages us to proceed further with ECT as a tool for obtaining animal models for human epilepsy.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
Animals
All mouse strains were raised at The Jackson Laboratory. They were fed standard NIH diet containing 6% fat and acidified water ad libitum. To generate Szt1 BALB congenic mice, Szt1 heterozygotes were outcrossed to BALB mice and the affected F₁ hybrids were repeatedly backcrossed to BALB mice. The segregation of the Szt1 allele was followed by the polymorphism of the SSLP markers D2Mit213, D2Frk2 and D2Frk1 and ECT testing. D2Frk2 forward primer, TGCTAAACCAGACAGCCTGA; D2Frk2 reverse primer, TGGAGTACCAGGGATGTAGTTCA. D2Frk1 forward primer, GGGCTGCAGACCACCTAAA; D2Frk1 reverse primer, TCCAGGAGCCGTGTAGAAA. The mice used in the PTZ study, body weight monitoring and brain histology characterization had not been subjected to ECT test prior to the respective experiment. All animal procedures followed AAALAC guidelines and were approved by institutional ACUC.

Electroconvulsive threshold test
We followed the procedures described previously (7). Briefly, mice were restrained and a drop of anesthetic containing 0.5% tetracaine and 0.9% NaCl was placed onto each eye and a pre-set current was applied via silver transcorneal electrodes using an electroconvulsive stimulator (Ugo Basile model 7801). The stimulator was set to produce rectangular wave pulses with the following parameters: 299 Hz, 0.2 s duration, 1.6 ms pulse width for high frequency shock. Each mouse was tested once a day and twice in total with at least one day rest between the two testing days. We detected the seizure sensitivity of Szt1 mice by using the estimated CC₃ for normal B6 mice (6.5 mA for female and 8.0 mA for male) (7).

ENU mutagenesis
We injected B6 males at 8–10 weeks of age with 2x100, 2x110, 2x120, 3x80, 3x85 mg/kg ENU intraperitoneally. Following an 10 week sterility period each ENU-treated B6 male (G₀) was mated to B6 females to generate G₁ males. The G₁ males were used in a three-generation breeding scheme, where G₁ males were mated with B6 females to generate G₂ daughters that were backcrossed with their G₁ fathers in order to produce G₃ offsprings recessive for ENU-induced mutations.

Brain histology
For YFP imaging, we followed the procedures described by Feng et al. (18). The genotype of each mouse obtained from matings between Szt1 heterozygote and thy1-YFP line H heterozygote was confirmed by PCR using YFP specific primers (18). Four percent paraformaldehyde in PBS was used to fix the brain and the samples were post-fixed in 4% paraformaldehyde in PBS for 1 h before being immersed in 30% sucrose in PBS. After the brain sank to the bottom, the sample was further immersed in 15% sucrose-PBS and 50% OCT embedding medium (EMS, Ft Washington, PA, USA) for 12 h. Brain frozen sections of 8 µm were cut coronally and were stained with nuclear DNA dye DAPI (25 ng/ml in 1xPBS) for 5 min. Coverslips were then mounted onto mounting medium (Gel/Mount, Biomeda, Foster City, CA, USA). Sections were examined through a Nikon Eclipse 600 epi-fluorescent microscope and images were captured by a SPOT 2.2.1 camera (Diagnostic Instruments Inc.).

Szt1 genotyping
The genotypes of all mice were confirmed by PCR. Mouse tail DNA was amplified 35 cycles (15 s, 95°C; 30 s, 62°C; 1 min, 72°C) on a thermal cycler. Three primers were used in the genotyping PCR reaction. K14D (part of the 14th intron of KCNQ2) CTGGCCATTCACAGACAGG; BC3D (344 nucleotides upstream of the centromeric breakpoint), CCACCCTACGGTTACCAGTG; and BP1U (24 nucleotides downstream of the telomeric breakpoint), CTCAAGCTGGAGCCTCTCAC. The reaction mix was separated by 2% agarose gel. K14D and BP1U generated a 645 nt PCR product from wild-type allele, whereas BC3D and BP1U amplified a 390 nt fragment from the Szt1 allele due to the deletion.

Northern blot analysis and quantification
Total RNA was collected from E18.5 embryo brain using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). Poly-A selection was performed using Oligotex mRNA kit following the manufacturer's recommendation (Qiagen, Valencia, CA, USA). Three micrograms of poly-A+ RNA were separated on a denaturing agarose gel and transferred to a Nytran Plus membrane (Scheicher and Schuell, Keene, NH, USA). We generated an N-terminal specific probe to the Kcnq2 transcript following the procedure described in Nakamura et al. (16). The probe was gel-purified and the hybridization was carried out as described (16). The blot was washed and exposed to an X-ray film for 6 days at -80°C. The same blot was stripped and reprobed with a mouse ß-actin probe to obtain another film. Both films were imaged by Fuji Luminescent Image Analyzer LAS-1000 plus and subsequently quantified by Fuji Image Gauge Version 3.4.

	ACKNOWLEDGEMENTS

 
We thank Drs Sue Ackerman, Robert Burgess, Greg Cox, Tom Ferraro, Elyse Schauwecker, Nanda Singh and Karen Wilcox for comments, advice, or sharing of reagents and unpublished results. We also thank Dr Natalie Blades, Carolyne Dunbar and Jeff Forthofer for technical assistance. This work was supported by grants from the National Institutes of Health (NS40246 to H.S.W. and W.N.F., NS31348 to W.N.F.) and a TJL Postdoctoral Fellowship to Y.Y.

	FOOTNOTES

 
^* To whom correspondence should be addressed. Tel: +1 2072886354; Fax: +1 2072886077; Email: wnf{at}jax.org

	REFERENCES

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 

1. Hauser, W.A. and Hesdorffer, D.C. (1993) Incidence and prevalence. In Hauser, W.A. and Hedorffer, D.C. (ed.), Epilepsy: Frequency, Causes and Consequences. Demos Press, New York, pp. 1–51.

2. Jacobs, M.P., Fischbach, G.D., Davis, M.R., Dichter, M.A., Dingledine, R., Lowenstein, D.H., Morrell, M.J., Noebels, J.L., Rogawski, M.A., Spencer, S.S. and Theodore, W.H. (2001) Future directions for epilepsy research. Neurology, 57,1536–1542.[Abstract/Free Full Text]

3. Peterson, S.L. (1998) Electroshock. In Peterson, S.L. and Albertson, T.E. (ed.), Neuropharmacology Methods in Epilepsy Research. CRC Press, Boca Raton, FL, pp. 1–26.

4. Krall, R.L., Penry, J.K., Kupferberg, H.J. and Swinyard, E.A. (1978) Antiepileptic drug development: I. History and a program for progress. Epilepsia, 4, 393–408.

5. Loscher, W. (1998) New visions in the pharmacology of anticonvulsion. Eur. J. Pharmac., 342, 1–13.[CrossRef][Web of Science][Medline]

6. White, H.S., Wolf, H.H., Woodhead, J.H. and Kupferberg, H.J. (1998) The National Institutes of Health anticonvulsant drug development program: Screening for Efficacy. In French, J., Leppik, I. and Dichter, M.A. (ed.), Antiepileptic Drug Development: Advances in Neurology. Lippincott-Raven, Philadephia, PA, Vol. 76.

7. Frankel, W.N., Taylor, L., Beyer, B., Tempel, B.L. and White, H.S. (2001) Electroconvulsive thresholds of inbred mouse strains. Genomics, 74, 306–312.[CrossRef][Web of Science][Medline]

8. Justice, M.J., Noveroske, J.K., Weber, J.S., Zheng, B. and Bradley, A. (1999) Mouse ENU mutagenesis. Hum. Mol. Genet., 10, 1955–1963.

9. Russell, W.L., Kelly, E.M., Hunsicker, P.R., Bangham, J.W., Maddux, S.C. and Phipps, E.L. (1979) Specific-locus test show ethylnitrosourea to be the most potent mutagen in the mouse. Proc. Natl Acad. Sci. USA, 76, 5818–5819.[Abstract/Free Full Text]

10. Singh, N.A., Charlier, C., Stauffer, D., DuPont, B.R., Leach, R.J., Melis, R., Ronen, G.M., Bjerre, I., Quattlebaum, T., Murphy, J.V. et al. (1998) A novel potassium channel gene, KCNQ2, is mutated in an inherited epilepsy of newborns. Nat. Genet., 11, 25–29.

11. Biervert, C., Schroeder, B.C., Kubisch, C., Berkovic, S.F., Propping, P.J. and Jentsch, T.J. (1998) A potassium channel mutation in neonatal human epilepsy. Science, 279, 403–406.[Abstract/Free Full Text]

12. Steinlein, O.K., Mulley, J.C., Propping, P., Wallace, R.H., Phillips, H.A., Sutherland, G.R., Scheffer, I.E. and Berkovic, S.F. (1995) A missense mutation in the neuronal nicotinic acetylcholine receptor alpha 4 subunit is associated with autosomal dominant nocturnal frontal lobe epilepsy. Nat. Genet., 11, 201–203.[CrossRef][Web of Science][Medline]

13. Ross, S.A., Wong, J.Y.F., Clifford, J.J., Kinsella, A., Massalas, J.S., Horne, M.K., Scheffer, I.E., Kola, I., Waddington, J.L., Berkovic, S.F. and Drago, J. (2000) Phenotypic characterization of an alpha 4 neuronal nicotinic acetylcholine receptor subunit knock-out mouse. J. Neurosci., 20, 6431–6441.[Abstract/Free Full Text]

14. Wong, J.Y.F., Ross, S.A., McColl, C., Massalas, J.S., Powney, E., Finkelstein, D.I., Clark, M., Horne, M.K., Berkovic, S.F. and Drago, J. (2002) Proconvulsant-induced seizures in 4 nicotinic acetylcholine receptor subunit knockout mice. Neuropharmacology, 43, 55–64.[CrossRef][Web of Science][Medline]

15. Watanabe, H., Nagata, E., Kosakai, A., Nakamura, M., Yokoyama, M., Tanaka, K. and Sasai, H. (2000) Disruption of the epilepsy KCNQ2 gene results in neural hypersensitivity. J. Neurochem., 75, 28–33.[CrossRef][Web of Science][Medline]

16. Nakamura, M., Watanabe, H., Kubo, Y., Yokoyama, M., Matsumoto, T., Sasai, H. and Nishi, Y. (1998). KQT2, a new putative potassium channel family produced by alternative splicing. Isolation, genomic structure, and alternative splicing of the putative potassium channels. Receptors Channels, 5, 255–271.[Web of Science][Medline]

17. Cukierman, E., Huber, I., Rotman, M. and Cassel, D. (1995). The ARF1 GTPase-activating protein: zinc finger motif and Golgi complex localization. Science, 270, 1999–2002.[Abstract/Free Full Text]

18. Feng, G., Mellor, R.H., Bernstein, M., Keller-Peck, C., Nguyen, Q.T., Wallace, M., Nerbonne, J.M., Lichtman, J.W. and Sanes, J.R. (2000) Imaging neuronal subsets in transgenic mice expressing multiple spectral variants of GFP. Neuron, 28, 41–51.[CrossRef][Web of Science][Medline]

19. Corbo, J.C., Deuel, T.A., Long, J.M., LaPorte, P., Tsai, E., Wynshaw-Boris, A. and Walsh, C.A. (2002) Doublecortin is required in mice for lamination of the hippocampus but not the neocortex. J. Neurosci., 22, 7548–7557.[Abstract/Free Full Text]

20. Nolan, P.M., Peters, J., Strivens, M., Rogers, D., Hagan, J., Spurr, N., Gray, I.C., Vizor, L., Brooker, D., Whitehill, E., Washbourne, R. et al. (2000) A systematic, genome-wide, phenotype-driven mutagenesis programme for gene function studies in the mouse. Nat. Genet., 25, 440–443.[CrossRef][Web of Science][Medline]

21. Hrabe de Angelis, M., Flaswinkel, H., Fuchs, H., Rathkolb, B., Soewarto, D., Marschall, S., Heffner, S., Pargent, W., Wuensch, K., Jung, M. et al. (2000) Genome-wide, large-scale production of mutant mice by ENU mutagenesis. Nat. Genet., 25, 444–447.[CrossRef][Web of Science][Medline]

22. Herron, B.J., Lu, W., Rao, C., Liu, S., Peters, H., Bronson, R.T., Justice, M.J., McDonald, J.D. and Beier, D.R. (2002) Efficient generation and mapping of recessive developmental mutations using ENU mutagenesis. Nat. Genet., 30, 185–189.[CrossRef][Web of Science][Medline]

23. Bucan, M. and Abel, T. (2002) The mouse: genetics meets behavior. Nat. Rev. Genet., 3, 114–123.[CrossRef][Web of Science][Medline]

24. Burgess, D.L., Jones, J.M., Meisler, M.H. and Noebels, J.L. (1997) Mutation of the Ca2+ channel ß subunit gene Cchb4 is associated with ataxia and seizures in the lethargic (lh) mouse. Cell, 88, 385–392.[CrossRef][Web of Science][Medline]

25. Escayg, A., De Waard, M., Lee, D.D., Bichet, D., Wolf, P., Mayer, T., Johnston, J., Baloh, R., Sander, T. and Meisler, M.H. (2000) Coding and noncoding variation of the human calcium-channel beta4-subunit gene CACNB4 in patients with idiopathic generalized epilepsy and episodic ataxia. Am. J. Hum. Genet., 66, 1531–1539.[CrossRef][Web of Science][Medline]

26. Meisler, M.H., Kearney, J., Ottman, R. and Escayg, A. (2001) Identification of epilepsy genes in human and mouse. A. Rev. Genet., 35, 567–588.[CrossRef][Web of Science][Medline]

27. Zhang, Y., Mori, M., Burgess, D.L. and Noebels, J.L. (2002) Mutations in high-voltage-activated calcium channel genes stimulate low-voltage-activated currents in mouse thalamic relay neurons. J. Neurosci., 22, 6362–6371.[Abstract/Free Full Text]

28. Walker, M.C., White, H.S. and Sander, J.W. (2002) Disease modification in partial epilepsy. Brain, 125, 1937–1950.[Abstract/Free Full Text]

29. White, H.S. (2002) Animal models of epileptogenesis. Neurology, 59 (Suppl. 5), S7–S14.[Abstract/Free Full Text]

30. Russell, L.B. and Russell, W.L. (1992) Frequency and nature of specific-locus mutations induced in female mice by radiations and chemicals: a review. Mutat. Res., 296, 107–127.[Web of Science][Medline]

31. Yang, J.S., Lee, S.Y., Gao, M., Bourgoin, S., Randazzo, P.A., Premont, R.T. and Hsu, V.W. (2002) ARFGAP1 promotes the formation of COPI vesicles, suggesting function as a component of the coat. J. Cell Biol., 159, 69–78.[Abstract/Free Full Text]

32. Wang, H.S., Pan, Z., Shi, W., Brown, B.S., Wymore, R.S., Cohen, I.S., Dixon, J.E. and McKinnon, D. (1998) KCNQ2 and KCNQ3, potassium channel subunits: molecular correlates of the M-channel. Science, 282, 1890–1893.[Abstract/Free Full Text]

33. Cooper, E.C. (2001) Potassium channels: how genetic studies of epileptic syndromes open paths to new therapeutic targets and drugs. Epilepsia, 42 (Suppl. 5), 49–54.

34. Cooper, E.C., Harrington, E., Jan, Y.N. and Jan, L.Y. (2001) M channel KCNQ2 subunits are localized to key sites for control of neuronal network oscillations and synchronization in mouse brain. J. Neurosci., 21, 9529–9540.[Abstract/Free Full Text]

35. Ferraro, T.N., Golden, G.T., Smith, G.G., St Jean, P., Schork, N.J., Mulholland, N., Ballas, C., Schill, J., Buono, R.J. and Berrettini, W.H. (1999) Mapping loci for pentylenetetrazol-induced seizure susceptibility in mice. J. Neurosci., 19, 6733–6739.[Abstract/Free Full Text]

CiteULike    Connotea    Del.icio.us    What's this?

This article has been cited by other articles:

				R. Roeloffs, A. D. Wickenden, C. Crean, S. Werness, G. McNaughton-Smith, J. Stables, J. O. McNamara, N. Ghodadra, and G. C. Rigdon In Vivo Profile of ICA-27243 [N-(6-Chloro-pyridin-3-yl)-3,4-difluoro-benzamide], a Potent and Selective KCNQ2/Q3 (Kv7.2/Kv7.3) Activator in Rodent Anticonvulsant Models J. Pharmacol. Exp. Ther., September 1, 2008; 326(3): 818 - 828. [Abstract] [Full Text] [PDF]

				N. A. Singh, J. F. Otto, E. Jill Dahle, C. Pappas, J. D. Leslie, A. Vilaythong, J. L. Noebels, H. Steve White, K. S. Wilcox, and M. F. Leppert Mouse models of human KCNQ2 and KCNQ3 mutations for benign familial neonatal convulsions show seizures and neuronal plasticity without synaptic reorganization J. Physiol., July 15, 2008; 586(14): 3405 - 3423. [Abstract] [Full Text] [PDF]

				J. A. Kearney, Y. Yang, B. Beyer, S. K. Bergren, L. Claes, P. DeJonghe, and W. N. Frankel Severe epilepsy resulting from genetic interaction between Scn2a and Kcnq2 Hum. Mol. Genet., March 15, 2006; 15(6): 1043 - 1048. [Abstract] [Full Text] [PDF]

				D. N. Cook, G. S. Whitehead, L. H. Burch, K. G. Berman, Z. Kapadia, C. Wohlford-Lenane, and D. A. Schwartz Spontaneous Mutations in Recombinant Inbred Mice: Mutant Toll-like Receptor 4 (Tlr4) in BXD29 Mice Genetics, March 1, 2006; 172(3): 1751 - 1755. [Abstract] [Full Text] [PDF]

				R. B. Sher, C. Aoyama, K. A. Huebsch, S. Ji, J. Kerner, Y. Yang, W. N. Frankel, C. L. Hoppel, P. A. Wood, D. E. Vance, et al. A Rostrocaudal Muscular Dystrophy Caused by a Defect in Choline Kinase Beta, the First Enzyme in Phosphatidylcholine Biosynthesis J. Biol. Chem., February 24, 2006; 281(8): 4938 - 4948. [Abstract] [Full Text] [PDF]

				J. F. Otto, Y. Yang, W. N. Frankel, H. S. White, and K. S. Wilcox A Spontaneous Mutation Involving Kcnq2 (Kv7.2) Reduces M-Current Density and Spike Frequency Adaptation in Mouse CA1 Neurons J. Neurosci., February 15, 2006; 26(7): 2053 - 2059. [Abstract] [Full Text] [PDF]

				T. P. O'Brien and W. N. Frankel Moving forward with chemical mutagenesis in the mouse J. Physiol., January 1, 2004; 554(1): 13 - 21. [Abstract] [Full Text] [PDF]

This Article Abstract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (16) Request Permissions Google Scholar Articles by Yang, Y. Articles by Frankel, W. N. Search for Related Content PubMed PubMed Citation Articles by Yang, Y. Articles by Frankel, W. N. Social Bookmarking          What's this?

Online ISSN 1460-2083 - Print ISSN 0964-6906

Copyright © 2009 Oxford University Press

Oxford Journals Oxford University Press

• Site Map
• Privacy Policy
• Frequently Asked Questions

Other Oxford University Press sites: Oxford University Press Oxford Journals China Oxford Journals Japan American National Biography Booksellers' Information Service Children's Fiction and Poetry Children's Reference Corporate & Special Sales Dictionaries Dictionary of National Biography Digital Reference English Language Teaching Higher Education Textbooks Humanities International Education Unit Law Medicine Music Online Products Oxford English Dictionary Reference Rights and Permissions Science School Books Social Sciences Very Short Introductions World's Classics