Mapping, cloning and characterization of the mass1 gene in Frings audiogenic seizure-susceptible mice

Frings audiogenic seizure-susceptible mice are a model for sensory-evoked reflex seizures. Their seizure phenotype is characterized by wild running, loss of righting reflex, tonic flexion, and tonic extension in response to high intensity sound stimulation. The present study determined that the mode of inheritance for audiogenic seizure-susceptibility in Frings mice is due to the autosomal recessive inheritance of a single gene. This gene locus was named mass1 for monogenic audiogenic seizure susceptible and was mapped by linkage analysis to the middle region of chromosome 13. further genetic mapping refined the nonrecombinant interval to a distant of approximately 80 kilobases, and a dense physical map spanning the region was developed. Analysis of genomic sequence from the region identified a single novel gene. Within this gene, a single base pair deletion was detected exclusively in Frings mouse DNA, which produces a truncated protein in vitro. The conclusion from these data is that the novel gene is mass1 and the protein truncation is associated with Frings mouse audiogenic seizures. Characterization of mass1 show that is a larger gene multiple exons that encodes three putative alternate transcripts. Expression of mass1 was detected in brain, lung, and kidney. To predict the function of mass1 protein (MASS1), we analyzed the amino acid sequence for homology to known proteins and found significant matches with a cytosolic loop region of the sodium+/calcium2+ exchanger protein family. This homology is due to a repetitive domain within MADD1. MASS1 also contains a multicopper oxidase consensus sequence that is located near the COOH terminus of the protein and would be missing in Frings mutant MASS1 protein. This motif is an important putative domain that will be characterized to determine if it is functional in MASS1. Further study of the MASS1 protein will elucidate its exact function in cells, particularly neurons, which will ultimately determine the role of the wild type protein, and how the Frings mutant MASS1 leads to audiogenic seizure-susceptibility. Understanding the function of this protein and its role in neuronal excitability may provide novel targets for anticonvulsant agent leading to new therapies for the treatment of epilepsy.

MAPPING, CLONING AND CHARACTERIZATION OF THE MASSI GENE IN FRINGS AUDIOGENIC SEIZURE-SUSCEPTffiLE MICE by Shana Skradski A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Pharmacology and Toxicology The University of Utah August 2000 Copyright © Shana Skradski 2000 All Rights Reserved THE UNIVERSITY OF UTAH GRADUATE SCHOOL SUPERVISORY COMMITTEE APPROVAL of a dissertation submitted by Shana Skradski This dissertation has been read by each member of the following supervisory committee and by majority vote has been found to be satisfactory. , Cliair: H. Steve White ~~ A~~_ Kristen A. Keefe Mark F. Leppert THE UNIVERSITY OF UTAH GRADUATE SCHOOL FINAL READING APPROVAL To the Graduate Council of The University of Utah: I have read the dissertation of Shana Skradski in its final form and have found that (1) its format, citations, and bibliographic style are consistent and acceptable; (2) its illustrative material including figures, tables, and charts are in place; and (3) the final manuscript is satisfactory to the Supervisory Committee and is ready for submission to The Graduate School. 6 /, ,t tJtJt5' Date , . H. Steve WhIte Member, Supervisory Committee Approved for the Major Department William R.Crowley I ChairlDean \ Approved for the Graduate Council ~~ oJ'S. cQ~--- David S. Chapman Dean of The Graduate School ABSTRACT Frings audiogenic seizure-susceptible mice are a model for sensory-evoked reflex seizures. Their seizure phenotype is characterized by wild running, loss of righting reflex, tonic flexion, and tonic extension in response to high intensity sound stimulation. The present study determined that the mode of inheritance for audiogenic seizure-susceptibility in Frings mice is due to the autosomal recessive inheritance of a single gene. This gene locus was named mass} for monogenic audiogenic seizure susceptible and was mapped by linkage analysis to the middle region of chromosome 13. Further genetic mapping refined the nonrecombinant interval to a distance of approximately 80 kilobases, and a dense physical map spanning the region was developed. Analysis of genomic sequence from the region identified a single novel gene. Within this gene, a single base pair deletion was detected exclusively in Frings mouse DNA, which produces a truncated protein in vitro. The conclusion from these data is that this novel gene is massl and the protein truncation is associated with Frings mouse audiogenic seizures. Characterization of mass} shows that it is a large gene with multiple exons that encodes three putative alternate transcripts. Expression of mass} was detected in brain, lung, and kidney. To predict the function of the mass} protein (MASS 1), we analyzed the amino acid sequence for homology to known proteins and found significant matches with a cytosolic loop region of the sodium+/ca1cium2+ exchanger protein family. This homology is due to a repetitive domain within MASS 1. MASS 1 also contains a multicopper oxidase consensus sequence that is located near the eOOH terminus of the protein and would be missing in Frings mutant MASS 1 protein. This motif is an important putative domain that will be characterized to determine if it is functional in MASS 1. Further study of the MASS 1 protein will elucidate, its exact function in cells, particularly neurons, which will ultimately determine the role of the wild type protein, and how the Frings mutant MASS 1 leads to audiogenic seizure-susceptibility. Understanding the function of this protein and its role in neuronal excitability may provide novel targets for anticonvulsant agents leading to new therapies for the treatment of epilepsy. v TABLE OF CONTENTS ABSTRACT .......................... : ............................................................................... iv LIST OF FIGURES ........................................................................................... viii ACKNOWLEDGEMENTS ................................................................................ ix Chapter 1. INTRODUCTION AND OVERVIEW OF DISSERTATION ............... 1 Background and Significance .................................................................. : ... 1 History of the Frings Audiogenic Seizure-Susceptible Mice ....................... 3 Comparison of Frings Mice to Other Audiogenic Seizure-Susceptible ....... .. Rodents ........................................................................................................ 7 Overview of the Frings Genetic Project ....................................................... 9 References .................................................................................................. 1 0 2. GENETIC MAPPING OF MASS1 ................................................ 15 Summary .................................................................................................... 15 Introduction ................................................................................................ 16 Materials and Methods ............................................................................... 18 Results ........................................................................................................ 20 Discussion .................................................................................................. 21 References .................................................................................................. 25 3. PHYSICAL MAPPING OF THE MASS1 LOCUS .......................... . 27 Summary .................................................................................................... 27 Introduction ................................................................................................ 28 Materials and Methods ............................................................................... 30 Results ........................................................................................................ 34 Discussion .................................................................................................. 37 References .................................................................................................. 39 4. IDENTIFICATION, CLONING, AND CHARACTERIZATION OF THE MASSI GENE AND ANALYSIS OF THE TRANSLATED PROTEIN SEQUENCE ........................................................... . 41 Sumnlary .................................................................................................... 41 Introduction ................................................................................................ 42 Materials and Method~ ............................................................................... 45 Results ........................................................................................................ 49 Discussion .................................................................................................. 62 References .................................................................................................. 67 5. DISCUSSION AND SUMMARY ................................................ 70 The Importance of the Frings Audiogenic Seizure-Susceptible Mouse Model of Epilepsy .......................................................................... 70 Dissertation Sumnlary ................................................................................ 72 Future Directions and Conclusion ............................................................. 81 References .................................................................................................. 84 vii LIST OF FIGURES Figure Page 1.1 Maximal audiogenic seizure in a Frings mouse .............................................. 5 2.1 Segregation pattern of mass 1 linked markers in the 137 seizure-susceptible and 120 seizure-resistant N2 mice genotyped ............................ 22 2.2 Map representing the interval containing the massllocus .......................... 24 3.1 Linkage map representing the interval containing the massllocus .............. 35 3.2 Large-scale physical map of the massl interval. ........................................... 35 3.3 Fine-scale physical map of the mass 1 interval defined by BACs and cosmids ........................................................................................ 38 4.1 Detailed map of the final massl interval spanning two cosmids .................. 50 4.2 Diagram of the location and genomic structure of massl ............................. 52 4.3 Northern blot analysis of massl transcripts .................................................. 53 4.4 Expression analysis of the massl gene (all transcripts) by RT-PCR in different tissues and cell RNA samples ........................................................ 55 4.5 Analysis of the exon 27 single base pair deletion in Frings mice ................. 58 4.6 In vitro transcription and translation of massl.3 clones constructed from CFI wild type or Frings n10use cDNA ......................................................... 59 4.7 Amino acid sequence alignment of the MASSI repetitive domain .............. 61 4.8 Diagram of the massl cDNA showing the location of the MASS 1 protein repetitive domains and protein motifs ........................................................... 63 ACKNOWLEDGElVIENTS The advice and support of numerous people have helped me throughout graduate school and in finishing this thesis. First and foremost, I came to the University of Utah and stayed in the program because of Dr. Steve White who has helped with every aspect of this project, and has been instrumental in making me the optimist that I am today. I thank Dr. Louis Ptacek for allowing me to work in his lab and for all his enthusiasm, help, and encouragement as well as the rest of my supervisory committee, Dr. Mark Leppert, Dr. Kristen Keefe, and Dr. Louis Barrows, for their insights and suggestions. I especially want to acknowledge the kindness, concern, and wisdom I received from Dr. Harold Wolf, one of the truly great professors at the University of Utah. I also thank all of the White and Ptacek lab members for help with my problems, lab work, and otherwise, every single day as well as each of the pharmacology and toxicology graduate students, past and present, for all the fun extracurricular activities. In addition, I thank Dr. Ying-Hui Fu for sharing her gene­hunting expertise, along with Philip Clair and Anna Clark for their contributions to the Frings mass} project that helped me complete this thesis. I finally thank my family and my husband, David Schmitz, for putting up with these 6 long years of insanity. CHAPTER 1 INTRODUCTION AND OVERVIEW OF DISSERTATION Background and Significance Epilepsy is defmed as the occurrence of recurrent seizures that are typically unprovoked and unpredictable. Epileptic seizures and syndromes are a heterogeneous group of disorders that are estimated to affect 1-3% of the population (Anderson et al., 1999; Hauser et a!., 1996). Seizures are often classified as either partial or generalized in their onset and result from hyper-synchronized frring of neurons. Partial seizures are focal, originating in a discrete brain region, and mayor may not become secondarily generalized to other brain structures. Generalized seizures are diffuse, involving both hemispheres from their onset and are not associated with an identifiable seizure focus. The Frings audiogenic seizure-susceptible mouse represents a model of generalized reflex epilepsy. Human reflex epilepsy syndromes include those seizures that can be induced by sensory stimulation including visual (stimulation by light flicker or patterns) and auditory. Reading, eating, and proprioception have also been identified as sensory seizure triggers (Zitkin and Andermann, 1993). , # At least seven of the human epilepsy syndromes are thought to result from single gene inheritance (Berkovic and Scheffer, 1997; Plaster et a!., 1999). At the present time, the molecular defect has been identified in four of these human epilepsy 2 syndromes. Evidence obtained from two families suggests that nocturnal familial epilepsy is produced by a mutation in the (l4-subunit of the nicotinic acetylcholine receptor (Steinlein et aI., 1997~ Steinlein et aI., 1995), Progressive myoclonic epilepsy has been associated with a mutation in the gene encoding a cysteine protease inhibitor, cystatin B (Pennacchio et aI., 1998). Benign familiaI neonatal convulsions have been linked to mutations in one of two novel related potassium channel genes, KCNQ2 and KCNQ3 (Biervert et aI., 1998; Charlier et aI., 1998; Singh et aI., 1998). Finally, febrile seizures plus (FS+) have been associated with mutations in the ~1 voltage-gated sodium channel regulatory subunit (Wallace et aI., 1998). Therefore, studying relatively simple genetic animaI models of seizures is important to understanding the influence of a single gene or simple polygenic trait on seizure expression. Investigations of transgenic, knockout, and spontaneously occurring mutations in mouse models have identified many seizure genes (Noebels, 1999). Recently, ion channel mutations have been identified in a number of epilepsy mouse models. For example, mutations in (lIM p, and y subunits of a voltage-sensitive calcium channel gene lead to the seizure phenotypes in the tottering (tg), lethargic (lh), and stargazer (stz) mice, respectively (Burgess et aI., 1997; Fletcher et al., 1996; Letts et aI., 1998). The slow-wave epilepsy (swe) mutant mouse phenotype was shown to result from mutations in the sodium/hydrogen exchanger gene nhel (Cox et al., 1997). These mutations result in divergent seizure phenotypes including spike-wave and generalized tonic clonic seizures. Like the human disorders, seizures exhibited by mouse models 3 can result from developmental or metabolic abnormalities or can be idiopathic in their origin. Furthermore, the origin of the mouse seizure phenotypes may be either polygenic or monogenic. Poly.genic defects may better model complex human epilepsy syndromes; however, the simple monogenic models will provide useful information as to how single genes alter the level of CNS excitability and are usually more readily characterized. History of the Frings Audiogenic Seizure-Susceptible Mice The Frings audiogenic seizure-susceptible mouse is an inbred strain derived from random albino stock which were frrst reported in 1951 by Frings, Frings, and Kivert (Frings et aL, 1951). In the early studies, these researchers selected breeding pairs to produce multiple strains of mice with variable degrees of seizure severity in response to sound stimulation (Frings and Frings, 1953). Today the only subset of those strains that remains is one that is maximally susceptible to audiogenic seizures. This strain was subsequently named Frings. Continuous selective inbreeding of audiogenic seizure-susceptible mice has produced the current Frings strain in which greater than 99% of mice tested between 21 and 24 days of age exhibit a maximal seizure in response to a standard high intensity sound stimulation of 110 decibles and 11 kilohertz. The audiogenic seizure phenotype of the Frings mouse strain has been well characterized. It consists of a latency period that extends from the time of sound stimulation until the onset of the wild running phase. The wild running ends when the 4 mouse loses its righting reflex and falls on its side. The mouse then exhibits forelimb clonus followed by tonic flexion, then tonic extension. Respiration is compromised during the tonic extension phase and the increase in CO2 is thought to lead to the . termination of tonic extension. Occasionally, the mouse dies due to respiratory depression. Mice can exhibit all or part of these seizure stages with maximal seizures being defmed as those progressing to tonic extension. Figure 1.1 shows a maximal tonic extension seizure in a Frings mouse. The expression of audiogenic seizures in the Frings mouse strain is age-dependent. This aspect of Frings audiogenic seizures was flrst described in a 1965 study that subjected mice of various ages to a single sound stimulation. The results showed that most mice younger than 2 weeks of age did not respond to sound stimulation. Maximal seizures were not observed until 16 days of age at which point the responses were evenly divided between clonic and tonic extension seizure endpoints. By 20 days of age almost all of the mice displayed maximal seizures, and the high penetrance of seizures persisted until approximately 30 days of age. The expression of maximal seizures gradually declined to 62.5% at 60 days of age. Mice not having maximal seizures progressed to the clonic seizure phase (Castellion et aI., 1965). Thirty-five years after this study, the age-dependent development of audiogenic seizures is virtually unchanged in the Frings mouse strain (unpublished observations). Age-dependent development of the audiogenic response is thought to parallel CNS development. Historical studies have shown that the mouse brain approaches maturity between 15 and 17 days of age. Measurements of neuronal density, brain A B c Figure 1.1. Maximal audiogenic seizure in a Frings mouse. A) The Frings mouse prior to the audiogenic stimulus. B) After the 110 decibel, 11 kilohertz sound stimulation the Frings mouse displays a maximal tonic extension seizure. C) The mouse recovering after the termination of the tonic extension phase of the seizure. 5 6 weight versus body weight, and cortical thickness are indistinguishable between the 15 to 17-day-old and adult mice (Haddara, 1956; Kobayashi, 1963; Sugita, 1918). Likewise, the auditory pathways do not fully develop until approximately 2 weeks of age which correlates with the onset of audiogenic seizure-susceptibility (Alford and Ruben, 1963; Schmidt and Fernandez, 1963). The physiological basis for the loss of full penetrance of maximal seizures in the Frings mice is not known but may be related to the random occurrence of hearing loss or neurochemical changes associated with aging. Audiogenic seizures are nondiscriminatory with respect to clinical categories of anticonvulsant drugs (Chapman et aI., 1984). This point is best exemplified by results obtained with the established anticonvulsants phenytoin, trimethadione (Swinyard et aI., 1963) and ethosuximide (H.S.White, unpublished data). These drugs display widely divergent clinical spectra in humans but are all active against sound-induced seizures in Frings mice. For example, phenytoin is effective against generalized tonic clonic seizures in humans, but trimethadione, (like ethosuximide) is effective against generalized absence seizures. Other antiepileptic drugs (AEDs) active in the Frings mouse include the established AEDs phenytoin, carbamazapine, ethosuximide, clonazepam, and valproate, and the newer AEDs lamotrigine, felbamate, gabapentin, and topiramate (H.S. White, unpublished data). In most cases the effective doses (ED50) of drugs for blocking the tonic extension (maximal) component of the audiogenic seizures is significantly lower than the effective doses for blocking the running phase. Likewise, the tonic extension phase of an audiogenic 7 seizure is more susceptible to modification by drugs than is generalized tonic extension induced by maxnnal electroshock (MES) (Swinyard et at, 1963). No anatomical brain abnormalities or developmental defects have been . identified in naive Frings mice. The strain displays a normal lifespan, breeds well, and produces normal size litters. This leads to the conclusion that the seizure phenotype is an isolated condition, setting this mouse apart from others in which the seizure phenotype is part of multiple components of a complex neurological condition. This allows us to study a seizure disorder without other neurological complications in a simple mouse model of epilepsy and to compare this phenotype to the complex, often multifactorial epileptic disorders. Comparison of Frings Mice to Other Audiogenic Seizure-Susceptible Rodents Audiogenic seizures have been observed in polygenic rodent models, such as the DBN2 mouse (Collins, 1970; Neumann and Collins, 1991; Neumann and Seyfried, 1990; Seyfried and Glaser, 1981; Seyfried et a!., 1980) and GEPR-9 rat (Ribak et at, 1988), as well as in the monogenic Frings strain (Skradski et at, 1998). Although no genes associated with audiogenic seizures in spontaneous mutant models have been cloned, three loci associated with seizure-susceptibility in the DBAl2 mouse (asp 1 , asp2, and asp3) have been mapped to chromosomes 12, 4, and 7, respectively (Neumann and Collins, 1991; Neumann and Seyfried, 1990). Audiogenic seizure-susceptibility has been characterized in two different knockout mouse models. The ftrst model is null mutant mice lacking the X-linked serotonin 5-HT2c receptor which have an increased susceptibility to spontaneous seizures, chemoconvulsants, and audiogenic seizures (Brennan et al., 1997; Tecott et aI., 1995). The second model is knockout mice lacking the fragile X-associated protein, FMRI. The hemizygous and homozygous null mutant mice have an increased susceptibility to audiogenic seizures over normallittermates. However, the penetrance and development of audiogenic seizure-susceptibility of these knockout mice are confounded by the inherent seizure­susceptibility observed in the background strain (Musumeci et aI., 2000). Audiogenic seizures can also be induced in seizure-resistant mice such as C57BL/6 by repetitive sound stimulation (Henry, 1967), suggesting that seizure-susceptibility can be influenced by lnultiple genetic and environmental factors. 8 Although the aUdiogenic seizure phenotype is essentially the same between DBAl2, 5-HT2C knockout, and Frings mice, there is a significantly different age­dependent expression of the seizure-susceptibility. The DBN2 mouse displays maximal seizure-susceptibility between 20-24 days which gradually declines with age (Seyfried, 1982; Seyfried, 1983; Seyfried and Glaser, 1981). This decline in seizure­susceptibility has been attributed to an age-dependent hearing loss in these mice (Erway et al., 1993; Ralls, 1967; Willott et al., 1995). Mice lacking the 5-HT2c receptor display audiogenic seizures by approximately 75 days of age, and complete penetrance is observed by 120 days (Brennan et aI., 1997). In contrast, complete penetrance is observed in the Frings mouse strain between 20-24 days of age and persists well into adulthood. This particular trait affords several distinct advantages. First, these mice are ideally suited for chronic pharmacological studies designed to assess tolerance to antiepileptic drugs. Second, they are well suited for assessing the pathological consequences associated with chronic seizure activity. Overvi~w of the Frings Genetic Project The overall aim of this project was to clone and characterize the gene responsible for audiogenic seizures in the Frings mouse model of epilepsy. To achieve this goal we began by characterizing the inheritance of the seizure phenotype after crossbreeding with a seizure-resistance mouse strain, the C57BU6J. As detailed in Chapter 2 these studies confirmed that the Frings audiogenic seizure phenotype was inherited by the recessive transmission of a single gene. We went on to use that panel of mice to map the Frings audiogenic seizure-susceptibility gene to a 3.6 centimorgan (cM) region in the middle of chromosome 13. The gene was named mass1 for monogenic audiogenic seizure susceptible. Chapter 3 continues with the mapping 9 study where we analyzed a large number of crossbred mouse DNA samples with MIT micro satellite markers across the initial region and developed a dense physical map of yeast and bacterial artificial chromosomes (Y ACs and BACs) for the entire region. The goal of the additional linkage analysis and physical mapping was to limit the [mal region of interest to the minimal possible distance. The conclusion of the physical mapping established the final mass1 interval to be contained within two cosmid clones with a maximum distance of 80 kilobases (Kb). To identify the mass1 gene, as described in Chapter 4, we analyzed the complete sequence from across the interval which identified a single novel gene spanning the entire interval. Within this gene, we identified a single base pair deletion 10 exclusively in Frings mice which was found to produce a truncated protein in vitro. From these data we concluded that this novel gene is mass} and the protein truncation is associated with the Frings mouse audiogenic seizures. Characterization of mass} , shows that it is a large gene with multiple exons that encodes three alternate transcripts with a distinct pattern of tissue specific expression. Analysis of the mass} translated amino acid sequence (MASS 1) identified functional domains that may help determine the function of the protein. The protein contains a repetitive domain that shares homology to a region of the sodium+/ca1cium2+ exchanger protein family as well as a multi copper oxidase motif. Further study of the MASS 1 protein will elucidate its exact function in cells, particularly neurons, which will ultimately determine the role of the wild type protein, and how the Frings mutant protein leads to audiogenic seizure-susceptibility. References Alford, B. R., and Ruben, R. J. (1963). Physiological, behavioral and anatomical correlates of the development of hearing in the mouse. Ann. Oto1. Rhino. Laryngol. 72,237-47. Anderson, V. E., Hauser, W. A., and Rich, S. S. (1999). Genetic heterogeneity and epidemiology of the epilepsies. Adv. Neurol. 79, 59-73. Berkovic, S. F., and Scheffer, I. E. (1997). Epilepsies with single gene inheritance. Brain Dev. }9, 13-8. Biervert, C., Schroeder, B. C., Kubisch, C., Berkovic, S. F., Propping, P., Jentsch, T. J., and Steinlein, O. K. (1998). A potassium channel mutation in neonatal human epilepsy. Science 279, 403-6. Brennan, T. J., Seeley, W. W., Kilgard, M., Schreiner, C. E., and Tecott, L. H. (1997). Sound-induced seizures in serotonin 5-HT2c receptor mutant mice. Nat. Genet. }6, 387-90. 11 Burgess, D. L., Jones, 1. M., Meisler, M. H., and Noebels, J. L. (1997). Mutation of the Ca2+ channel beta subunit gene Cchb4 is associated with ataxia and seizures in the lethargic (lh) mouse. Cell 88, 385-92. Castellion, A. W., Swinyard, E. A., and Goodman, L. S. (1965). Effect of maturation on the development and reproducibility of audiogenic and electroshock seizures in mice. Exp. Neurology 13, 206~ 17. Chapman, A. G., Croucher, M. J., and Meldrum, B. S. (1984). Evaluation of anticonvulsant drugs in DBA/2 mice with sound-induced seizures. Arzneimittelforschung 34, 1261-4. Charlier, C., Singh, N. A., Ryan, S. G., Lewis, T. B., Reus, B. E., Leach, R. J., and Leppert, M. (1998). A pore mutation in a novel KQT-like potassium channel gene in an idiopathic epilepsy family [see comments]. Nat. Genet. 18, 53-5. Collins, R. L. (1970). A new genetic locus mapped from behavioral variation in mice: audiogenic seizure prone (ASP). Behav. Genet. 1, 99-109. Cox, G. A., Lutz, C. M., Yang, C. L., Biemesderfer, D., Bronson, R. T., Fu, A., Aronson, P. S., Noebels, J. L., and Frankel, W. N. (1997). Sodium/hydrogen exchanger gene defect in slow-wave epilepsy mutant mice [published erratum appears in Cell 1997 Dec 12;91(6):861]. Cell 91, 139-48. Erway, L. C., Willott, J. F., Archer, J. R., and Harrison, D. E. (1993). Genetics of age-related hearing loss in mice: I. Inbred and Fl hybrid strains. Hear. Res. 65, 125- 32. Fletcher, C. F., Lutz, C. M., TN, O. S., Shaughnessy, J. D., Jr., Hawkes, R., Frankel, W. N., Copeland, N. G., and Jenkins, N. A. (1996). Absence epilepsy in tottering mutant mice is associated with calcium channel defects. Cell 87, 607-17. Frings, H., and Frings, M. (1953). The production of stocks of albino mice with predictable susceptibilites to audiogenic seizures. Behavior 5, 305-19. Frings, H., Frings, M., and Kivert, A. (1951). Behavior patterns of the laboratory mouse under auditory stress. J. Mammal. 32, 60-76. Haddara, M. (1956). A quantitative study of the postnatal changes in the packing density of the neurons in the visual cortex of the mouse. 1. Anat. 90, 494-501. Hauser, W. A., Annegers, J. F., and Rocca, W. A. (1996). Descriptive epidemiology of epilepsy: contributions of population-based studies from Rochester, Minnesota. Mayo Clin. Proc. 71, 576-86. 12 Henry, K. R. (1967). Audiogenic seizure susceptibility induced in C57BL-6J mice by prior auditory exposure. Science 158, 938-40. Kobayashi, T. (1963). Brain-to-body ratios and time of maturation of the mouse brain. Am. J. Physiol. 204, 343-6. Letts, V. A., Felix, R., Biddlecome, G. H., Arikkath, J., Mahaffey, C. L., Valenzuela, A., Bartlett, F. S., 2nd, Mori, Y., Campbell, K. P., and Frankel, W. N. (1998). The mouse stargazer gene encodes a neuronal Ca2+-channel gamma subunit [see comments]. Nat. Genet. 19, 340-7. Musumeci, S. A., Bosco, P., Calabrese, G., Bakker, C., De Sarro, G. B., Elia, M., Ferri, R., and Oostra, B. A. (2000). Audiogenic seizures susceptibility in transgenic mice with fragile X syndrome. Epilepsia 41, 19-23. Neumann, P. E., and Collins, R. L. (1991). Genetic dissection of susceptibility to audiogenic seizures in inbred mice. Proc. Natl. Acad. Sci. USA 88, 5408-12. Neumann, P. E., and Seyfried, T. N. (1990). Mapping of two genes that influence susceptibility to audiogenic seizures in crosses of C57BL/6J and DBAJ2J mice. Behav. Genet. 20, 307-23. Noebels, J. L. (1999). Single-gene models of epilepsy. Adv. Neurol. 79, 227-38. Pennacchio, L. A., Bouley, D. M., Higgins, K. M., Scott, M. P., Noebels, J. L., and Myers, R. M. (1998). Progressive ataxia, myoclonic epilepsy and cerebellar apoptosis in cystatin B-deficient mice. Nat. Genet. 20, 251-8. Plaster, N. M., Uyama, E., Uchino, M., Ikeda, T., Flanigan, K. M., Kondo, I., and Ptacek, L. J. (1999). Genetic localization of the familial adult myoclonic epilepsy (FAME) gene to chromosome 8q24. Neurology 53, 1180-3. Ralls, K. (1967). Auditory sensitivity in mice, Peromyscus and Mus musculus. Animal Behavior 15,123-8. Ribak, C. E., Roberts, R. C., Byun, M. Y., and Kim, H. L. (1988). Anatomical and behavioral analyses of the inheritance of audiogenic seizures in the progeny of genetically epilepsy-prone and Sprague-Dawley rats [published erratum appears in Epilepsy Res 1989 May-Jun;3(3):262]. Epilepsy Res. 2, 345-55. Schmidt, R. S., and Fernandez, C. (1963). Development of mammalian endocochlear potential. J. Exptl. Zool. 153, 227-35. 13 Seyfried, T. N. (1982). Developmental genetics of audiogenic seizure susceptibility in mice. In Genetic Basis of the Epilepsies, V. E. Anderson, W. A. Hauser, J. K. Penry and C. F. Sing, eds. (New York: Raven), pp. 199-210. Seyfried, T. N. (1983). Genetic heterogeneity for the development of audiogenic seizures in mice. Brain Res. 27.1 ,325-9 . Seyfried, T. N., and Glaser, G. H. (1981). Genetic linkage between the AH locus and a major gene that inhibits susceptibility to audiogenic seizures in mice. Genetics 99, 117-26. Seyfried, T. N., Yu, R. K., and Glaser, G. H. (1980). Genetic analysis of audiogenic seizure susceptibility in C57BL/6J X DBAl2J recombinant inbred strains of mice. Genetics 94,701-18. Singh, N. A., Charlier, C., Stauffer, D., DuPont, B. R., Leach, R. J., Melis, R., Ronen, G. M., Bjerre, I., Quattlebaum, T., Murphy, 1. V., McHarg, M. L., Gagnon, D., Rosales, T. 0., Peiffer, A., Anderson, V. E., and Leppert, M. (1998). A novel potassium channel gene, KCNQ2, is mutated in an inherited epilepsy of newborns [see comments]. Nat. Genet. 18,25-9. Skradski, S. L., White, H. S., and Ptacek, L. 1. (1998). Genetic mapping of a locus (mass1) causing audiogenic seizures in mice. Genomics 49, 188-92. Steinlein, O. K., Magnusson, A., Stoodt, J., Bertrand, S., Weiland, S., Berkovic, S. F., Nakken, K. 0., Propping, P., and Bertrand, D. (1997). An insertion mutation of the CHRNA4 gene in a family with autosomal dominant nocturnal frontal lobe epilepsy. Hum. Mol. Genet. 6, 943-7. 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, 201-3. Sugita, H. (1918). Comparative studies on the growth of the cerebral cortex. VIII. General review of data for the thickness of the cerebral cortex and the size of the cortical cells in several mammals, together with some postnatal growth changes in these structures. J. Compo Neurol. 29, 241-78. Swinyard, E. A., Castelli on, A. W., Fink, G. B., and Goodman, L. S. (1963). Some neurophysiological and neuropharmacological characteristics of audiogenic-seizure­susceptible mice. J. Pharm. Exp. Ther. 140, 375-84. Tecott, L. H., Sun, L. M., Akana, S. F., Strack, A. M., Lowenstein, D. H., Dallman, M. F., and Julius, D. (1995). Eating disorder and epilepsy in mice lacking 5-HT2c serotonin receptors [see comments]. Nature 374, 542-6. Wallace, R. H., Wang, D. W., Singh, R., Scheffer, 1. E., George, A. L., Jr., Phillips, H. A., Saar, K., Reis, A., Johnson, E. W., Sutherland, G. R., Berkovic, S. F., and Mulley,1. C. (1998). Febrile seizures and generalized epilepsy associated with a mutation in the Na+-channel beta1 subunit gene SCN1B. Nat. Genet. 19, 366-70. 14 Willott,1. F., Erway, L. C., Archer, J. R., and Harrison, D. E. (1995). Genetics of age-related hearing loss in mice. II. Strain differences and effects of caloric restriction on cochlear pathology and evoked response thresholds. Hear. Res. 88, 143-55. Zifkin, B. G., and Andermann, F. (1993). Epilepsy with reflex seizures. In The Treatment of Epilepsy: Principles and Practices, E. Wyllie, ed. (Philadelphia: Lea & Feviger), pp. 614-23. CHAPTER 2 , GENETIC MAPPPING OF MASSI Summary Frings audio genic seizure-susceptible mice are a model for sensory-evoked reflex seizures. Their seizure phenotype is characterized by wild running, loss of righting reflex, tonic flexion, and tonic extension in response to high intensity sound stimulation. The Frings mice are from an inbred colony that has not been genetically characterized. This investigation studied the mode of inheritance for audio genic seizures by crossing the Frings mouse with the seizure-resistant C57BL/6J mouse. Among the backcross progeny generated by crossing (Frings x C57BU6J)Fl mice with the Frings strain, 391 of the 836 N2 progeny were audiogenic seizure-susceptible, a finding consistent with mono genic inheritance. Genetic mapping and linkage analysis of hybrid mice using MIT micro satellite marker sequences localized the seizure gene named mass] for monogenic audiogenic seizure-susceptible to an interval of approximately 3.6 cM located in the middle of mouse chromosome 13. Linkage of the mass] to chromosome 13 is an important step in identifying the gene associated with a monogenic seizure disorder in mice which may ultimately lead to a better understanding of the pathophysiology of human seizure disorders. 16 Introduction To initiate the study of the genetic basis of audiogenic seizures in the Frings mouse, breeding pairs of Frings and the seizure-resistant C57BU6J mouse strains were . mated. The (Frings x C57BU6J)F1 progeny were backcrossed to the Frings parent strain to produce (Frings x C57BU6J)F1 x Frings backcross mice which were used for genetic mapping and studies of seizure thresholds between audiogenic seizure-susceptible and seizure-resistant mice. The Frings mouse audiogenic seizure-susceptibility gene mapping project started with the confrrmation of unpublished observations suggesting that the phenotype was a recessive monogenic trait. In contrast to the polygenic DBAl2J aUdiogenic seizure-susceptible mouse (Collins and Fuller, 1968; Neumann and Collins, 1991; Neumann and Seyfried, 1990; Seyfried and Glaser, 1981; Seyfried et at, 1980), the possibility of the Frings mouse phenotype being a single gene disorder makes it more amenable to mapping and cloning. Although polygenic defects may better model complex human epilepsy syndromes, this simple monogenic model will provide useful information as to how a single gene alters the level of CNS excitability. The Frings seizure gene can be quickly localized using what are termed simple sequence length polymorphisms (SSLPs, also called microsatellite markers). Genotyping a limited number of seizure-susceptible and seizure-resistant mouse DNA samples with SSLP markers across the entire genome will determine where there is an increased association between seizure-susceptible mice and homozygous genotypes above the 50% expected by chance alone (Copeland et at, 1993). Conversely, there should also be an increased association of seizure-resistance with heterozygous 17 genotypes. Once possible candidate regions are identified, more SSLP markers can be analyzed with a full panel of mouse DNA samples to confmn the linkage to that chromosome. Analysis of mice which harbor chromosomal recombinations in the region will determine the bounaaries of the critical interval. Assuming the Frings gene is 100% penetrant, all mice that were phenotyped as seizure-susceptible must be homozygous for the mutation in the gene. They are also homozygous for DNA surrounding the gene until a recombination of the chromosome is encountered (Lander and Botstein, 1987). The nonrecombinant gene interval is defined by all the SSLP markers where no recombinations are detected, bordered by SSLP markers that contain recombinations in DNA samples. In this study we have shown that the mode of inheritance of audiogenic seizure­susceptibility in Frings mice is monogenic. Furthermore, we have mapped the gene to an interval of approximately 3.6 cM on chromosome 13. As a simple monogenic model of isolated sensory-evoked generalized tonic extension seizures, the Frings Inouse represents an important tool that will lead to the identification of a discrete genetic seizure gene. Identification of the mass} encoded protein will provide a greater understanding of the molecular basis for this genetically controlled audio genic seizure­susceptible phenotype. Study of this simple monogenic model will provide information about seizures that will ultimately advance our understanding of epilepsy as a multi­factorial, complex disorder. 18 Materials and Methods Mouse Breeding Frings lruce were crossed to the seizure-resistant strain C57BL/6J to produce (Frings X C57BL/6J)F1 Inice. From the fITst 16 matings of Frings Inales and females to C57BU6J mice, 11 Inouse litters were produced with 103 offspring. The seizure­resistant FI progeny were crossed to the parental Frings strain to produce (Frings X C57BU6J)F1 X Frings backcross Inice. A cumulative total of 67 litters and 836 N2 progeny were produced that were used in either pharmacological (data not shown) or genetic studies. Seizure Testing All mice were phenotyped at postnatal day 21 as seizure-susceptible or seizure­resistant. Mice were placed individually into a round, plastic sound chamber and exposed to 110 decibles (dB), 11 kilohertz (KHz) sound until a full tonic-extension seizure was elicited, or for 20 seconds if no seizure activity was observed (White et aI., 1992). Directly following seizure phenotyping, mice were euthanized by CO2 and bilateral thoracotomy and spleens harvested for DNA preparation. DNA Isolation, Genotyping, and Microsatellite Analysis High molecular weight DNA was isolated from mouse spleens by phenol/chloroform extraction and ethanol precipitation. Primer sequences and marker information were obtained from the Mouse Genome Database (MGD) at the Jackson Laboratories [http://www.informatics.jax.org/mgd.html].Primers were synthesized by standard techniques by the oligonucleotide core facility in the Department of Human 19 Genetics at the University of Utah. One primer was end labeled with p32yATP by T4 polynucleotide kinase using standard techniques (Fouad et al., 1996). The PCR conditions used were 10 mM tris-HCl, 40 mM NaCI, 1.5 mM MgCl, 250 flM spermidine, 200 flM dNTPs, d.5 flM of cold forward and reverse primers, and 9 nM p32yATP end-labeled primer, using 50 ng of mouse genomic DNA in a 25 flL reaction volume. PCR conditions were 94°C for 20 seconds, 58°C for 20 seconds, 72°C for 20 seconds for 5 cycles, and 94°C for 20 seconds, 55°C for 20 seconds, 72°C for 20 seconds with 25 cycles. Samples were electrophoresed on 6% bis-acrylamide gels and visualized by autoradiography. Mapping and Linkage Analysis A genome wide search was undertaken with micro satellite markers derived from the MIT mouse map. Mapping with groups of eight phenotypically affected and unaffected N2 mice was performed to fmd markers that were polymorphic between the Frings and C57BU6J parent strains. Polymorphic markers spaced approximately 50 cM apart were evaluated in affected and unaffected N2 mice for regions with higher than probable homozygosity and heterozygosity, respectively. Once the mapping suggested an association between mouse chromosome 13 and the Frings seizure locus, the genotypes for 76 N2 progeny and 16 parental mice from eight litters were established across the chromosome. Fine mapping in the region of the initially liIlked markers was then performed on 257 N2 progeny (including the fIrst 76) from 29 litters. 20 Results Seizure Inheritance Frings mice were crossed with the seizure-resistant C57BU6J mouse strain to . establish the mode of transmission of the Frings gene. In the Frings parent strain, penetrance of the audiogenic seizure-susceptible phenotype is observed to be greater than 99% (unpublished observations). Testing of the F1 progeny (n=103) showed 100% of the mice to be seizure-resistant at 21 days of age. The backcross matings of the F1 mice with the Frings parental strain produced 836 N2 progeny of which 391 were audiogenic seizure-susceptible (47%), which does not significantly differ from the expected segregation ratio of 50% (X2= 3.48, df=l, p=0.06). Based on these results, the transmission of the Frings audiogenic-seizure phenotype is consistent with an autosomal recessive trait produced by a single gene. There was no association between audiogenic seizure-susceptibility and coat color or gender, thereby eliminating these chromosomal regions from further testing. Linkage to Chromosome 13 A genome-wide search was undertaken with groups of 8 seizure-susceptible and seizure-resistant N2 progeny. MIT microsatellite mouse marker sequences were used to localize the Frings seizure locus. Because the swiss albino strain was not included in the MIT database, it was necessary to first determine whether the MIT n1arkers employed were polymorphic between the Frings and C57BU6J parental strains. For all the markers tested, no residual heterozygosity was detected in the Frings parent strain, further confirming the Frings mouse as being a true inbred strain. 21 Analysis of 50 polymorphic MIT microsatellite markers spanning the genome localized the audiogenic seizure gene to chromosome 13. Genotyping 257 N2 mice and analyzing linkage and recombinants localized the massl interval to the region of chromosome 13 between D13Mit200 and D13Mit126 (Figure 2.1). Markers D13Mit97,231,9 were not separable in the 257 backcross progeny. Analysis of genotypes showed 3 obligate and 2 nonobligate recombinant mice from different litters with recombinations between D13Mit126 and D13Mit97,23l,9, thereby defining the distal border of the interval. The proximal boundary of the interval was defined by 2 obligate and 2 nonobligate recombinant mice from 4 separate litters with recombinations between Dl3Mit200 and D13Mit97,231,9 (Figure 2.1). The distance between the markers that border the interval was calculated to be 3.6 cM in this backcross and approximately 4 cM as estimated from the 1997 Chromosome 13 Committee Report (Justice and Stephenson, 1997). This locus was named massl for monogenic audiogenic seizure-susceptible (nomenclature approved by the International Committee on Standardized Genetic Nomenclature for Mice). Mouse litter #63 (one of the 29 litters genotyped) contained 10 pups of which 6 mice were audiogenic seizure-susceptible. Of these 6, one male and 2 female seizure­susceptible mice did not carry a homozygous haplotype in the proposed interval containing massl (Figure 2.1, far left). Discussion Results from this study establish the Frings audiogenic seizure-susceptible mouse as a monogenic model of seizures. The massl locus has been mapped to the Dl3Mit66 Dl3Mit200 massl, Dl3Mit97# Dl3Mitl26 Dl3Mit202 II II 011 011 011 011 II 110 110 011 011 011 II 110 110 II 011 011 II 11.0 110 110 110 011 II 11011 II 110110 3* 110 123 4 6 2 2 3 2 0 2 } 3.9.:t 1.2 cM } 1.6 .:to.8 cM } 2.0 .:to.9 cM } 0.8 .:to.6 cM 22 Figure 2.1. Segregation pattern of massllinked markers in the 137 seizure­susceptible and 120 seizure-resistant N2 mice genotyped. Each column represents the chromosomal haplotype with the number of animals observed with that haplotype given below. The black boxes represent the C57BL/6J allele, and the open boxes represent the Frings allele. At the far left are the haplotypes for 3 seizure-susceptible mice from one litter which are heterozygous across the proposed interval for massl (shown by *). The # denotes that similar results were obtained with markers Dl3Mit231 and Dl3Mit9 (not shown). 23 middle region of chromosome 13 and is not associated with any of the loci responsible for audiogenic seizures in DBAl2J mice (aspI-3) (Neumann and Seyfried, 1990, Neumann et aI., 1991) Identification and functional expression of the protein encoded . by massI will provide insight into the etiology of sound-induced seizures and will potentially lead to better understanding of other reflex epilepsies. Of the 137 affected N2 progeny tested, 3 mice from one litter were heterozygous over the interval (Figure 2.1, "*"). It is possible that these 3 samples have been mislabeled. It is also possible that the mice represent a phenocopy. A phenocopy would be mice having an identical seizure phenotype caused by a mutation in a different gene. Because the samples were all from one litter and they were the only anomalous genotypes identified for all of the mice tested the possibility of phenocopy in these mice seems to be a likely explanation. According to the 1997 chromosome committee reports (Justice and Stephenson, 1997), several candidate genes have been mapped to this region of mouse chromosome 13 shown in Figure 2.2. Of these, adenylyl cyclase II (Adcy2) (Edelhoff et aI., 1995) and the dopamine transporter (DatI) (Lossie et aI., 1994) represent two good physiological candidates. Cyclic AMP which is regulated by adenylyl cyclase is proconvulsive when injected into rodents brains (Boulton et aI., 1993; Ferrendelli, 1986), As such, a dysregulation of cAMP could be responsible for audiogenic seizures in Frings mice. Likewise, evidence suggests that modification of dopaminergic tone in the CNS can modify seizure threshold. For example, D2 receptor stimulation is anticonvulsant whereas activation ofDl receptors is proconvulsant (al-Tajir et aI., 1990). Therefore, Adcy2 and DatI are good physiological candidate genes for "'!IIIIi" D 13Mit66 3.9cM ..... D13Mit200 1.6cM ... t- D13Mit97,231,9 mass} 2.0cM ..... D13Mit126 O.8cM ...... D13Mit202 i 13 Figure 2.2. Map representing the interval containing the massllocus. The map was generated by the analysis of 257 N2 backcross progeny and recombinants between the loci. The calculated genetic distances are shown on the left. 24 25 audiogenic seizures in the Frings mouse. However, as detailed in Chapter 4, completion of the physical map determined that neither Dat} , Adcy2, nor any other candidate genes were localized within the final mass} intervaL Mapping of mass} to chromosome 13 localizes the first monogenic mouse locus for audiogenic seizures in a spontaneously occurring mutation modeL No known seizure-related loci have been mapped to this region of mouse chromosome 13. A region of conserved synteny is present between the mass} interval of mouse chromosome 13 and regions of human chromosome 5. Genetic mapping of sensory-evoked human epilepsy syndromes may provide a link to the Frings audiogenic seizure locus. This linkage study of the mass} gene will provide the basis for identifying the mass} protein. Characterization of this protein will lead to a greater understanding of the factors contributing to brain hyperexcitability and the etiology of seizures. Ultimately, these studies may provide novel pharmacological targets for the treatment of epilepsy. References al-Tajir, G., Chandler, C. J., Starr, B. S., and Starr, M. S. (1990). Opposite effects of stimulation of D 1 and D2 dopamine receptors on the expression of motor seizures in mouse and rat. Neuropharmacology 29,657-61. Boulton, C. L., McCrohan, C. R., and O'Shaughnessy, C. T. (1993). Cyclic AMP analogues increase excitability and enhance epileptiform activity in rat neocortex in vitro. Bur. J. PharmacoL 236, 131-6. Collins, R. L., and Fuller, J. L. (1968). Audiogenic seizure prone (asp): a gene affecting behavior in linkage group 8 of the mouse. Science }62, 1137-9. Copeland, N. G., Jenkins, N. A., Gilbert, D. J., Bppig, J. T., Maltais, L. J., Miller, J. C., Dietrich, W. F., Weaver, A., Lincoln, S. B., Steen, R. G., and et al. (1993). A genetic linkage map of the mouse: current applications and future prospects [see comments]. Science 262,57-66. 26 Edelhoff, S., Villacres, E. C., Storm, D. R., and Disteche, C. M. (1995). Mapping of adenylyl cyclase genes type I, II, III, IV, V, and VI in mouse. Mamm. Genome 6, 111- 3. Ferrendelli, J. A. (1986). Roles of biogenic amines and cyclic nucleotides in seizure mechanisms. Adv. Neurol. 44, 393-400. Fouad, G. T., Servidei, S., Durcan, S., Bertini, E., and Ptacek, L. J. (1996). A gene for familial paroxysmal dyskinesia (FPD1) maps to chromosome 2q. Am. J. Hum. Genet. 59, 135-9. Justice, M. J., and Stephenson, D. A. (1997). Mouse chromosome 13. Mamm. Genome 7, S223-37. Lander, E. S., and Botstein, D. (1987). Homozygosity mapping: a way to map human recessive traits with the DNA of inbred children. Science 236, 1567-70. Lossie, A. C., Vandenbergh, D. J., Uhl, G. R., and Camper, S. A. (1994). Localization of the dopamine transporter gene, Dat 1, on mOll:se chromosome 13. Mamm. Genome 5,117-8. Neumann, P. E., and Collins, R. L. (1991). Genetic dissection of susceptibility to audiogenic seizures in inbred mice. Proc. Natl. Acad. Sci. USA 88, 5408-12. Neumann, P. E., and Seyfried, T. N. (1990). Mapping of two genes that influence susceptibility to audiogenic seizures in crosses of C57BU6J and DBN2J mice. Behav. Genet. 20, 307-23. Seyfried, T. N., and Glaser, G. H. (1981). Genetic linkage between the AH locus and a major gene that inhibits susceptibility to audiogenic seizures in mice. Genetics 99, 117- 26. Seyfried, T. N., Yu, R. K., and Glaser, G. H. (1980). Genetic analysis of audiogenic seizure susceptibility in C57BU6J X DBN2J recombinant inbred strains of mice. Genetics 94, 701-18. White, H. S., Patel, S., and Meldrum, B. S. (1992). Anticonvulsant profile of MDL 27,266: an orally active, broad-spectrum anticonvulsant agent. Epilepsy Res. 12, 217- 26. CHAPTER 3 . PHYSICAL MAPPING OF THE MASS1 LOCUS Summary Frings audiogenic seizure-susceptible mice are a model for sensory-evoked reflex seizures. Their seizure phenotype is characterized by wild running, loss of righting reflex, tonic flexion, and tonic extension in response to high intensity sound stimulation. Previous work determined that the Frings seizure phenotype was due to the autosomal recessive inheritance of a single gene and mapped this gene to the middle region of chromosome 13. The gene was named mass} for monogenic audiogenic seizure-susceptible. In this study we refmed the genetic map of the mass} region by genotyping an additional 1200 (Frings X C57BL/6J)F1 intercross mice and developed a dense physical map across the interval. U sing clones from the physical map we identified new micro satellite markers in the region and analyzed these markers with the recombinant mice from the mass) interval. Recombinations from two mice determined that the mass} genetic interval was spanned by two cosmid clones with a maximum physical distance of 80 kilobases (Kb). Refining the genetic map and developing a physical map of the mass} interval are important steps towards identifying and cloning the mass} gene. 28 Introduction The Frings audiogenic seizure-susceptible mouse is a genetic model of generalized seizures evoked by a sensory stimulation. Our frrst study confrrmed that the Frings audio genic seizure phenotype was due to the autosomal recessive transmission of a single gene. This was determined by analyzing the segregation of the seizure-susceptibility phenotype for more than 800 (Frings X C57BU6J)F1 X Frings backcross lnice. Approxitnately 47% of the N2 mice were audio genic seizure­susceptible which did not differ significantly from the expected 500/0. This gene was named mass1 for monogenic audiogenic seizure susceptible. The mass1 gene was mapped to chrolnosolne 13 with a small panel of seizure-susceptible and seizure­resistant lnice. The mouse DNA sample set was expanded to include 257 N2 mice which were genotyped with SSLP markers spanning chromosome 13. Data from these mice determined that the mass1loci was 3.6 cM located near the middle of the chromosome. The mass1 interval defmed by the original set of N2 mice was not small enough to efficiently clone the Frings seizure gene. Therefore, the purpose of this study was to refme the mass1 interval to the minimal possible distance and to complete a physical map of that region. Comparing several published mouse gene cloning projects (King et aI., 1997; Kingsmore et at, 1996; Perou et al., 1996; Segre et aI., 1995) led us to believe that analyzing approximately 2000 meioses would produce a recombinational interval sufficiently small to allow the rapid cloning of the mass1 gene. Approximately 1200 (Frings X C57BU6J)F1 intercross progeny were genotyped with markers across 29 the interval to identify new recombinant mice in the region. These mice were used to fmely map the mass 1 interval. Along with breeding and screening F2 mice, we proceeded to assemble a . physical map of the evolving region. A complete physical map of the region serves many functions. First, it will allow us to conclusively determine jf any of the known candidate genes and SSLP markers currently mapped in the general mass1 vicinity are truly in the fmal interval. Second, having the genomic DNA from the region in cloned segments will allow us to identify new SSLP markers in the region which can be used to narrow the interval. Finally, the genomic clones can be used to identify candidate cDNAs from the region by hybridizing them to cDNA libraries (Copeland et a!., 1993). The genomic clones can also be used as templates for large scale sequencing projects from which the open reading frames of genes can be identified (King et at, 1997). The physical mapping began with yeast artificial chromosomes (Y ACs) with an average insert size of 1 Mb covering large distances of the region. The Y ACs were identified from contigs described on the Genetic and Physical Maps of the Mouse Genome [http://www.genome.wi.mit.edulcgi-binlmouse/index] internet site. The YACs were characterized with the simple sequence length polymorphisms (SSLP) and sequence tagged site (STS) markers listed on the maps. Unlike SSLPs which are used for recombinational analysis, STS markers are PCR primer pairs that amplify small segments of DNA that are mapped to specific locations in the genome making them useful for aligning contigs. 30 In conjunction with the Y AC mapping, bacterial artificial chromosomes (BACs) with an average insert size between 80 to 120 Kb were identified with primers in the region from a PCR-based library (Shizuya et aI., 1992). The BACs are more stable than Y ACs which are often chimeric (King et aI., 1997), and the insert DNA is more readily produced, isolated, manipulated, and sequenced. The BACs were used to identify new SSLP markers and to walk across the fmal massi interval. When the massi interval was contained within a single BAC, it was subcloned into a cosmid contig. Fine mapping and physical mapping ultimately produced an interval containing the massi gene on two overlapping cosmids with a maximal distance of 80 Kb. A gene interval of this size is amenable to quickly identifying and cloning massi. The two cosmids can be used to identify one or more candidate cDNAs by hybridizing them to mouse brain cDNA libraries as well as providing a template for sequencing to identify open reading frames. Once candidate cDNAs are identified they can be screened for the mutation responsible for audio genic seizures in Frings mice. Physically mapping the massi region is an important step in identifying the massi gene which will provide valuable information on the genetic basis of this seizure phenotype as well as mechanisms underlying CNS hyperexcitability. Materials and Methods Mouse Breeding, Seizure Testing and DNA Collection Frings mice were crossed to the seizure-resistant strain C57BL/6J to produce 1200 (Frings X C57BU6J)F1 intercross offspring. The Frings mice used in this study 31 were bred in our colony and the C57BU6J mice were supplied by the Jackson Laboratory (The Jackson Laboratory, Bar Harbor, ME). All mice were phenotyped at postnatal day 21 as seizure-susceptible or seizure-resistant as described in Chapter Two. Directly following seizure phenotyping, tail sections were cut for DNA preparation. Potential recombinant mice within the region were tested again to confmn the seizure phenotype, a second tail section was cut, and the mice were euthanized by CO2 and bilateral thoracotomy and spleens harvested for DNA preparation. High molecular weight DNA was isolated from mouse tails or spleens by pheno]Jchlorofonn extraction and ethanol precipitation. Fine Mapping All F2 mice were initially tested with markers D13Mit312, D13Mit97, and D13Mit69 to identify new recOlubinant mice in the mass1 region. Recombinant mice were genotyped with additional markers (D13Mit99, D13Mit9, D13Mit190) across the region to determine the minimal nonrecombinant interval with available SSLPs. All known MIT micro satellite markers were identified from the Chromosome 13 Committee map located at [http://www.informatics.jax.org/ccr/searches/contents .cgi?&year=1999&chr=13]. Primer sequences and information for the markers were obtained from the Whitehead Institute Database site Genetic and Physical Maps of the Mouse Genome [http://www.genome.wi.mit.edulcgi-binimouse/index].Primer synthesis and SSLP analysis were performed as previously described in Chapter 2. 32 Physical Mapping Yeast artificial chromosomes. YAC maps spanning the region were obtained from the Physical Maps of the Mouse Genome [http://www.genome.wi.mit.edulcgi- . bin/mouse/index]. Y ACs that appeared to contain SSLP markers known to be within the region were obtained frOlTI Research Genetics (Haldi et al., 1996) and YAC DNA was prepared by standard techniques (Silverman, 1996). All STSs shown to be associated with each Y AC clone from the map were synthesized and tested to confrrm that the clones were correct and aligned with overlapping YAC clones. Standard PCR conditions for physical mapping analyses were 10 mM tris-HCI, 40 mM NaCl, 1.5 mM MgCI, 30 J-lM dNTPs, 0.5 J-lM of forward and reverse primers, and 50 ng of DNA in a 25 J-lL reaction volume. PCR thermo cycles were 94°C for 2 minutes, followed by 35- 40 cycles of 94°C for 10 seconds, 54°C for 30 seconds, and 72°C for 30 seconds with a 5 minute [mal extension at 72°C. Bacterial artificial chromosomes. BACs were identified and isolated from the PCR-based mouse BAC library available from Research Genetics using all known STSs and SSLPs found in the region on linkage and Y AC maps. BAC DNA was prepared using purification columns by the recommended procedure (Magnum columns, Genome Systems, Inc). BAC end sequence was obtained using T7 and SP6 primers. Individual BAC insert sizes were determined by complete digestion of the BAC DNA with Notl and separating the fragments on a 1.0% agarose gel in 0.5X TBE circulating buffer. The field inversion gel electrophoresis (FIGE) program was 180 volts forward, 120 volts reverse, 0.1 seconds initial switching time linearly ramped to 3.5 seconds switching time for 16 hours. 33 Simple sequence length polymorphism (SSLP) identification. BAC DNA was partially digested with Sa~3A1 into fragments ranging from 1 to 3 kb and subcloned into the BamH1 site of pUC18 with the Re ad y-To-Go cloning kit (Amersham Pharmacia Biotech). New repeats were identified by plating the subclone library, lifting duplicate Hybond-N membranes (Amersham Pharmacia Biotech) and hybridizating with (CAho and (ATho oligonucleotides endlabe1ed with y32p-ATP. Hybridized membranes were exposed to autoradiographic f11m. Clones producing a positive signal were sequenced and primer pairs were designed to amplify new repeat sequences. New SSLP markers were tested with control and recombinant mice to fmely map the interval. Bacterial artificial chromosome (BAC) walking. When the fme mapping showed that the mass] interval was spanned by a single YAC clone and the interval did not include any known STSs or SSLPs, the BAC library was utilized to walk across the final gap. Continuous BAC identification, end sequencing, and STS development produced a final BAC contig of seven clones. Simultaneous SSLP searches of the new BACs produced markers which narrowed the mass] interval as the walking progressed. Cosmid subcloning. BAC 290J21 was partially digested with Sau3A1 into 30- 40 Kb fragments which were subcloned into cosmids as per the instructions for the SuperCos 1 cosmid vector kit (Stratagene) and packaged with Gigapack III Gold 34 Packaging Extract (Stratagene) using XL I-Blue mrf' competent cells. Cosmids were then aligned by amplification with all STSs across the region. Results Fine Mapping The massl interval spanning from D13Mit200 to D13Mit126 was estimated to be 3.6 cM at the limits of the initial set of 257 N2 mice tested as shown in Chapter 2. Approximately 1200 additional (Frings X C57BL/6J )F1 intercross mice were genotyped with microsatellite markers D13Mit312, D13Mit97, and D13Mit69 which spanned the interval as shown in Figure 3.1. Analysis of the recombinations determined that the massl region was distal to the D13Mit97 marker and proximal of D13Mit69. Two additional microsatellite markers, D13Mit9 and D13Mit190, were identified within this interval from the Chromosome 13 Committee map. Genotyping of the border-defining recombinant mice with these markers narrowed the interval to between D13Mit9 and D13Mit190. Of the 1200 F2 mice 3 were recombinant at D13Mit9 and 10 mice were recombinant at D13Mit190. No other known SSLP markers were mapped to within this interval. Physical Mapping This distance between the markers D 13Mit9 and D 13Mit190 was covered by three overlapping YACs 151C12, 87F11, and 187D1 found on the contig WC13.27 shown in Figure 3.2. These YACs contained 4 known STSs, D13SLC106, D13SLCl17, D13SLClll and D13SLC105 which were used to identify BACs from 35 13 I ............................................................................ I ....................................... ~ ...................... . 0.8 cM 2.0cM 1.6cM I D13Mit69/ D13Mit126 D13Mit9 D13Mit190 Dat1 Adcy2 I D13Mit97! I D13Mit312 \ D13Mit200 Nhe3 mass 1 Figure 3.1. Linkage map representing the interval containing the massllocus. The map was generated by the analysis of 257 N2 backcross progeny and recombinants between the loci. The calculated genetic distances are shown on the top. The loci Nhe3, Datl, and Adcy2 represent the sodium hydrogen exchanger 3, dopamine transporter 1, and adenylyl cyclase 2 genes. The boxed loci represent the three SSLP markers used to genotype the F2 progeny. .YAG 18701. YAG 180G8 • • • • YAG 151G12 • • • • YAG 87F11 • • • • • • YAG 464G7 • • • • • Figure 3.2. Large-scale physical map of the massl interval. The loci D13Mit190 and D13Mit9 are MIT microsatellite SSLP markers. D13SLC10 and D 13SLCll are novel SSLP markers, and the others are STS markers. 36 the BAC library. Using small-insert pUC19 subclone libraries of the BACs, we screened for new SSLP markers. On BAC 177N3 we identified a Inarker D13SLC10 where 6 of the 10 recombinant mice from D 13Mit190 were still recombinant which narrowed the distal boundary. 'Finally, we identified a marker, D13SLCll, on BAC 265F23 where 4 mice were still recombinant. This marker, D13SLCll, defmed the distal border and with D13Mit9 defmed the massl interval to the distance spanned by a single YAC, 151C12 (Figure 3.2). No known SSLPs or STSs were contained within the massl interval, therefore, we walked into the region using BAC end sequences to develop new STSs. Simultaneous with the BAC walking, SSLP searches of new BACs continued to narrow the interval. Seven overlapping BACs were required to cover the distance betweenD13SLCll and D13Mit9. Identification of BAC 290J21 both crossed the gap, completing the BAC walk, and recombinations in SSLPs from each end of the BAC showed that it represented the entire massl region. BAC 290J21 contained a proximal SSLP, D13SLC14, with 1 recombinant mouse of the 3 from D13Mit9, Inouse #5a9, and also a distal marker, D13SLC15, which was recombinant for the same 4 mice as D13SLCll, mice #54c8, #2d11, #30c7, and #33c5. This localized the massl region to an area that was less than 150 Kb which was the size estimated for BAC 290J21 by Field Inversion Gel Electrophoresis. BAC 290J21 was made into cosmid library with inserts of approximately 30 to 40 Kb as well as a small-insert pUC 19 subclone library. Sequences from randolnly selected pUC19 clones were used to develop new STSs across the BAC, and these new 37 markers were then used to align cosmids into a complete contig of BAC 290J21. A minimum of 5 overlapping cosnlids were required to cover the massl region as shown in Figure 3.3. SSLP screening of the pUC19 library detected 5 new repeat markers within BAC 290J21. Markers Dl3SLCl6, Dl3SLCl7, and Dl3SLCl8 were alllnapped to cosmid C12Ajust proximal of the flanking marker Dl3SLC]4, and all4lnarkers were recombinant for mouse 5a9. Marker Dl3SLCl9 was mapped to cosmid C5A which placed it within the massl interval. Analysis of this marker showed it was recolnbinant for 2 (54c8 and 2dll) of the 4 mice recolnbinant at Dl3SLCl5 making it the new distal boundary of the region. One fmal marker, Dl3SLC20, mapped to cosmid C13A and was recombinant for only mouse 2d 11 that refmed the interval to a maximum of 2 overlapping coslnids, CIB and CI3A. Discussion Fine mapping and physical mapping are important to rapidly minimize and clone regions containing genes of interest. Problems occur if any of the individual samples used to defme the interval have the incorrect phenotype. In our initial linkage mapping we discovered three seizure-susceptible N2 mice from one litter that were not homozygous for the Frings allele across the mass] interval (see Chapter 2). All the N2 mice were phenotyped and sacrificed, and spleen DNA was prepared for genotyping. This protocol did not allow us to recheck the samples. However, using a different sampling protocol for the 1200 F2 mice, we did not encounter any more problelnatic mice. The F2 mice were tested for seizure-susceptibility and then DNA was extracted h~¢i~"~~®hi"i";l'ill 40a6 & 1 e4 ~:',~}~~:#,:~(:-.~':<:"lii~t;~*{·~~f. ~~)1~i~/~*n~t~b~;;~~{Vii:~i Sa9 I::.:~~:w::::~~'':~~§t~~~*%t;::~~~:i:'i'~?r:i.::tj~~~it~!ii~w~i::·~'t<f~~~~~:::#$-~.;l{;i;¥tWik;Z€<ffj*,~ii=m#ii:%:-~~~@m\'H.m;:;~¥#_~¥~_h:.'tw$i.,'$·:Y#@iwj 2d 11 ·~t'iMt:r~:::;·: «':':::}~~""'. i::m . -';%$~;t*~m¥,;.o/.···· ~~'1" f:-:\tW:~>~ .. ~~'t.:., ;'-,H3}.~~.·~ 5408 ,~ ~ 3Oc7& 3305 It) !:.] ~ ~ Q) ~ ~ ;;; <:> a ~ ~ ~ l\l (;j ~ 5 "'" "- co ~ ~ 0 0 '" rl 0 '" 0 U ~ G [) ~ ~ ii2il iGil ~~ iGil i[i)l iGil G~ ~~ iGil i(iJl ~~ [ii)l iGil [~) ~~ ~~ [~) ~~ (~J ~ ., ., ., ., '" i2 _ Ci Q a Ci Ci Q Q Ci Ci Ci Q a Q Q'" C'"i Q Ci Ci Ci Ci Q Ci BAC29OJ21 • • ••••••• C • •• • • •• C • • • • • BAC 251C17 • <) • .8ACi4919. • i A5 • • • • ••••• 18 ~ CSB • • • ..,gB 0---4 • • • .cWA 0 • • • • • CSA • • •• 091i?A.... • • III mass 1 .. <80 Kb Closest Recombinant Mice Recombinant IEEJl Unknown 0 Nonrecombinant - STSs and SSLPs I BACs Cosmids Figure 3.3. Fine-scale physical map of the massl interval defined by BACs and cosmids. Dl3SLC- numbers between 10 and 100 are novel SSLP markers, and Dl3SLC- numbers 100 to 200 are novel STS markers. The bars above the map represent the genotypes of the nearest recombinant mice. The hatched bars represent regions where the mice are recombinant, black filled bars are regions where the mice are nonrecombinant, and white filled bars are regions where the genotypes are unknown. The final massl interval is spanned by cosmids C13A and CIB with a maximal distance of 80 Kb. VJ 00 39 from tail samples for genotyping. Later, important recombinant mice were tested again for seizure-susceptibility and a second DNA sample genotyped to eliminate errors in sampling. This leads us to believe that the mass} interval we defined is accurate and the mutation responsible for audiogenic seizures in Frings mice must be contained within the 2 cosmids. Inaccurate maps represent a second problem to consider. Markers are often not in the correct order and Y AC inserts are often chimeric. The map of the mass} interval is believed to be reliable as the final interval was covered with BAC clones and using Y AC clones was not necessary. BACs are more stable and all new SSLP and STS markers were tested against the BAC and Y AC maps with no anomalies detected in any case. Fine mapping of the mass} interval decreased the distance containing the gene responsible for causing audiogenic seizures in Frings mice from 3.6 cM (an estimated 7 Mb) to less than 80 Kb. Narrowing the interval to these 2 cosmids will allow us to rapidly screen for candidate cDNA sequences and identify the mass} gene. Identifying mass} will provide more information on the genetic basis of seizures and mechanisms of CNS hyperexcitability, as well as possibly providing novel target for antiepileptic therapies. References Copeland, N. G., Jenkins, N. A., Gilbert, D. J., Eppig, J. T., Maltais, L. J., Miller, J. C., Dietrich, W. F., Weaver, A., Lincoln, S. E., Steen, R. G., et al. (1993). A genetic linkage map of the mouse: current applications and future prospects [see comments]. Science 262,57-66. 40 Haldi, M. L., Strickland, C., Lim, P., VanBerkel, V., Chen, X., Noya, D., Korenberg, J. R., Husain, Z., Miller, J., and Lander, E. S. (1996). A comprehensive large-insert yeast artificial chromosome library for physical mapping of the mouse genome. Mamm Genome 7, 767-9. King, D. P., Zhao, Y., Sangoram, A. M., Wilsbacher, L. D., Tanaka, M., Antoch, M. P., Steeves, T. D., Vitaterna, M. H., Kornhauser, J. M., Lowrey, P. L., Turek, F. W., and Takahashi, J. S. (1997). Positional cloning of the mouse circadian clock gene. Cell 89, 641-53. Kingsmore, S. F., Barbosa, M. D., Tchernev, V. T., Detter, J. C., Lossie, A. C., Seldin, M. F., and Holcombe, R. F. (1996). Positional cloning of the Chediak-Higashi syndrome gene: genetic mapping of the beige locus on mouse chromosome 13. J Investig. Med. 44, 454-61. Perou, C. M., Moore, K. J., Nagle, D. L., Misumi, D. J., Woolf, E. A., McGrail, S. H., Holmgren, L., Brody, T. H., Dussault, B. J., Jr., Monroe, C. A., Duyk, G. M., Pryor, R. J., Li, L., Justice, M. J., and Kaplan, J. (1996). Identification of the murine beige gene by Y AC complementation and positional cloning. Nat. Genet. 13, 303-8. Segre, J. A., Nemhauser, J. L., Taylor, B. A., Nadeau, J. H., and Lander, E. S. (1995). Positional cloning of the nude locus: genetic, physical, and transcription maps of the region and mutations in the mouse and rat. Genomics 28, 549-59. Shizuya, H., Birren, B., Kim, U. J., Mancino, V., Slepak, T., Tachiiri, Y., and Simon, M. (1992). Cloning and stable maintenance of 300-kilobase-pair fragments of human DNA in Escherichia coli using an F-factor-based vector. Proc. Natl. Acad. Sci. USA 89,8794-7. Silverman, G. A. (1996). Purification of YAC-Containing Total Yeast DNA. In Methods in Molecular Biology, Vol. 54, D. Markie, ed. (Totowa, NJ: Humana Press Inc.), pp. 65-8. CHAPTER 4 IDENTIFICATION: CLONING, AND CHARACTERIZATION OF THE MASS1 GENE AND ANALYSIS OF THE TRANS LA TED PROTEIN SEQUENCE Summary Frings audiogenic seizure-susceptible mice are a model for sensory-evoked reflex seizures. Their seizure phenotype is characterized by wild running, loss of righting reflex, tonic flexion, and tonic extension in response to high intensity sound stimulation. Previous work determined that the Frings seizure phenotype was due to the autosomal recessive inheritance of a single gene and mapped this gene to an 80 Kb interval in the middle region of chromosome 13. In this study we identified a single novel gene completely spanning the 80 Kb interval by analyzing the sequence across the entire nonrecombinant region. Characterization of mass1 shows that it is a large gene with multiple exons that encodes for three putative alternate transcripts. Expression of mass1 was detected in brain, lung, and kidney. Within this gene, we detected a single base pair deletion exclusively in Frings mice which produces a truncated protein in vitro. From these data we concluded that this novel gene is mass1 and the protein truncation is associated with the Frings mouse audiogenic seizures. To predict the function and characteristics of the mass] translated protein (MASSI), we analyzed the amino acid sequence for homology to known proteins and found significant matches 42 with the large third cytosolic loop region of the sodium+/calciuln2+ exchanger protein family. This homology is due to a highly repetitive domain within MASSI. MASSI also contains a multicopper oxidase consensus sequence that is located near the eOOH terminus of the protein and would be missing in Frings mutant MASS 1 protein. This motif is an important putative domain that will have to be characterized to determine if it is functional in MASS 1. Further study of the MASS 1 protein will elucidate its exact function in cells, particularly neurons, which will ultimately determine the role of the wild type protein and how the Frings mutant MASSI leads to audiogenic seizure­susceptibility. Introduction The Frings audiogenic seizure-susceptible mouse is a genetic model of generalized seizures evoked by a sensory stimulation. We have shown that the Frings audio genic seizure phenotype is due to the autosomal recessive transmission of a single gene. This gene was naIned mass} for monogenic audiogenic seizure-susceptible. The mass} gene was mapped to chromosome 13, and extensive physical mapping determined that the genetic mutation is contained within two cosmid clones with a maximum distance of 80 Kb. The goal of this study was to identify cDNAs in the mass} interval and to determine which gene contains the genetic mutation responsible for causing seizures in the Frings audiogenic seizure-susceptible mouse. To accomplish this goal we used cosmid clones as probes to screen multiple mouse brain cDNA libraries to identify cDNAs which have homology to genomic DNA from the mass} interval. All cDNAs 43 identified would automatically become candidate genes for massI. This region was sufficiently small that one would expect to fmd one or possibly two candidate genes in the region. However, the library screening was unsuccessful. As an alternative strategy we sequenced the entire nonrecombinant massI interval and screened this genomic sequence for open reading frames to fmd putative exons of genes in the region. Once an open reading frame sequence was identified, it was characterized to confinn that the gene is expressed in mouse brain RNA and the full-length sequence was determined from the 5' untranslated region to the polyA tail. Reverse transcription and PCR CRT-PCR) of overlapping segments from across the entire gene from Frings and seizure-resistant C57BL/6J mouse brain RNA samples were compared to identify differences in nucleotide sequences. All nucleotide changes could possibly represent the genetic mutation responsible for audio genic seizure-susceptibility in Frings mice. Conversely, they could represent polymorphisms between strains of mice that have no effect on protein sequence or function. In this study we report a single novel gene spanning the entire nonrecombinant massI interval. Characterization of massI shows that it is a large gene with 35 exons that encodes three alternate transcripts with a discreet tissue specific pattern of expression. Within the Frings massI gene we identified a single base pair deletion. This deletion has only been detected in Frings mice and is predicted to change the reading frame, leading to a premature stop codon and then to a truncated protein. In vitro transcription and translation of cDNA clones confirmed the truncation of Frings mass} translated protein compared to the wild type protein. From these data we conclude that this novel gene is mass} and the protein truncation causes audiogenic seizures in the Frings mouse. Sequence similarity searches of the mass} cDNA by BLAST at the National Center for Biotechnology Information [http://www.ncbi.nlm.nih.gov/blast/blast.cgiJ yielded no significant homology to any known genes. However, by searching the 44 mass} translated amino acid sequence it was determined that short repetitive regions of the MASS 1 sequence were similar to a domain of the sodium+ -calcium2+ exchanger (Na+/Ca2+ exchanger) protein family. The amino acid sequence was also found to contain a multicopper oxidase domain. To conclude the study, we performed a detailed analysis the amino acid sequence of MASS 1, including the homology between MASSI and the Na+/Ca2+ exchangers, and described functional domains identified within the sequence including the multicopper oxidase consensus sequence. Identification of the mass} gene is a significant step toward understanding the genetic basis of audiogenic seizures in the Frings mouse model of epilepsy. This novel gene shows no homology to known ion channels, making it a unique among seizure­causing genes. While further studies may determine that the function of MASS 1 is to modulate ion gradients across cell membranes or cytosolic ion concentrations, it will most likely be through an indirect mechanism that has not yet been characterized. Analysis of the MASSI protein sequence will direct future research towards identifying and characterizing the function of MASS 1 and how the truncated protein in Frings mice leads to audiogenic seizure-susceptibility. Targeting novel mechanisms or pathways involved in seizures may produce better pharmacological agents that avoid side effects associated with current treatments. Materials and Methods Identifying and Cloning the mass} Gene 45 Cosmid subcIones. BAC 290J21 was partially digested with Sau3Al into 30- 40 Kb fragments which were subcloned into cosmids as per the instructions for the SuperCos 1 cosmid vector kit (Stratagene) and packaged with Gigapack III Gold Packaging Extract (Stratagene) using XLI-Blue mrf competent cells. Cosmids were then aligned by PCR amplification with all STSs across the region. Cosmid sequencing was performed by standard techniques using 1200 ng of cosmid DNA and 3.2 pmole of gene-specific massl oligos ranging from 18 to 24 nucleotides in length. Cloning the eDNA. The massl cDNA was identified by reverse transcription­PCR (RT-PCR) using primers developed .between exons predicted using Genefmder [http://dot.imgen.bcm.tmc.edu:93311gene-fmder/gf.html] and Grail [http://avalon.epm.ornLgov/Grail-binlEmptyGrailForm]. Total RNA was prepared from whole mouse brain of C57BL/6J, Frings and Fl mice with Trizol reagent as per instructions (Molecular Research Center, Inc.). The standard reverse transcription reaction conditions were 1.0 Ilg RNA, 15 ng random hexamers, Ix First Strand Buffer, 10 mM DTT, 1 mM dNTPs, 40 U RNAse Inhibitor, and 200 U Superscript II reverse transcriptase (Gibco BRL Lifetechnology). The mass} gene was cloned in multiple segments by PCR and RACE using C57BU6 mouse brain Marathon-ready cDNA (Clonetech) following the supplier's recommendations. All PCR products were cloned 46 into the pCR2.1-TOPO TA cloning vector (Invitrogen). High fidelity PCR clones were made using pix DNA polymerase (GibcoBRL Life Technologies), and Inultiple sequencing reactions were compared to detennine the fmal mass} sequence. Analysis of the massl Gene Reverse transcription.peR. The RT reactions to determine tissue specificity of mass} expression were performed as described in the previous section on samples from CFl (Charles River, Wilmington, MS), C57BU6J (The Jackson Laboratory, Bar Harbor, MA), or Frings mouse tissues and cells. The tissue panel samples were isolated from a single C57BL/6J mouse. PCR conditions to amplify the cDNAs were 10 mM tris-HCl, 40 mM NaCl, 1.5 mM MgCl, 30 J..lM dNTPs, 0.5 J..lM of forward and reverse primers, and 1 J..lL of the cDNA in a 25 J..lL reaction volume. PCR thermo cycles were 94°C for 2 minutes, followed by 25 (fJ-actin primers) or 40 (mass} primers) cycles of 94 °c for 10 seconds, 54°C for 30 seconds, and 72°C for 30 seconds with a 5 minute [mal extension at 72°C. The mass} primers spanned from exon 22 to exon 23, the forward was 5' CAG AGO ATG GAT ACA GTA C 3' and the reverse was 5' GTA ATC TCC TCC ITG AGT TG 3' and the expected product size was 487 base pairs. The fJ-actin primers also spanned an intron and were forward 5' GeA GTG TGT TGG CAT AGA G 3' and reverse 5' AGA TCC TGA CCG AGe GTG 3' and the expected product size was 327 base pairs. PCR products were separated by gel electrophoresis on 2% agarose gels in IX TAE buffer at 120V, and the bands visualized by staining with ethidium bromide using an ultraviolet(UV) light source. The size ladders are 100 bp (Gibco/BRL Life Technologies) increments with 600 base pairs staining brightest. 47 Mutational analysis. For SSCP, the mouse DNA samples NJ, AKR/J, BALB/cJ, C57BL/6J, C3H/HeJ, CAST/EiJ, LP/J, NON/LtJ, Nod/LtJ, SPRET/EiJ, and DBA2/J were supplied by the Jackson Laboratory (Bar Harbor, MA). The CF1 Inice were supplied by Charles River (Wilmington, MS), and the seizure-susceptible EL, EP, and SAS mice were supplied by Dr. T. Seyfried (Boston University, Boston, MS). PCR reactions were identical to those conditions listed above except 0.3 ~L of a32P-dCTP was included in a 10 ~L total reaction volume. A 30 ~L aliquot of dilution buffer (0.1 % SDS/lOmM EDTA in ddH20) was added to the PCR reactions. A 10 ~L aliquot of the dilute PCR reaction was mixed with 1 0 ~L of loading dye (brolnophenol blue/xylene cyanol) and 2 ~L samples were separated by nondenaturing electrophoresis on an 8% bis-acrylamide, 10% glycerol, nondenaturing gel at 20W for 14 hours at rOOln telnperature with a fan. The PCR forward primer sequence was 5' TTT ATT GTA GAG GAA CCT GAG 3' and the reverse primer sequence was 5' GCC AGT AGC AAA CTG TCC 3' and the expected product size was 126 base pairs. Exon 27 PCR products were sequenced to determine that the aberrant band was due to a single G deletion in the Frings mouse mass1 gene as shown for C57BL/6 and Frings DNA. Analysis of the massl Translated Protein Sequence Amino acid sequence analysis. The amino acid sequence of MASS 1 was deduced from the nucleotide sequence of the cloned mass1 cDNA by DNA Star. The amino acid sequence was compared to known proteins by BLAST sequence similarity searching [http://www.ncbi.nlm.nih.gov/blastlblast.cgi]. Identification of functional domains utilized PSORT II Prediction [http://psort.nibb.ac.jp/form2.html], Sequence 48 Motif Search [http://www.motif.genome.ad.jp/]. Global and Domain Similarity Search [http://www-nbrf.georgetown.edu/pirwww Isearchl dmsim.html], and Pattern Match [http://www-nbrf.georgetown.edulpirwww Isearch/patmatch.html]. In vitro protein expression. The in vitro protein truncation test utilized massl.3 cDNA clones in pBluescript SK. The cDNAs were produced by RT-PCR amplification as described previously using pix Platinum DNA polymerase (GibcoIBRL). The cDNAs were produced from both CFl and Frings RNA. The clones were sequenced to confirm a complete open reading frame as well as confirm the ex on 27 splice variants and Frings 7009~G mutation. The proteins were produced with the TnT Quick Coupled Transcription/Translation System (Promega) with the T7 RNA polymerase as recommended by the supplier. Redivue 35S-Methionine (Amersham Pharmacia Biotech) was incorporated into the proteins. The standard reaction contained 40 ilL TnT Quick Master Mix, 2 ilL 35S-Methionine, and 1 mg template DNA in a 50 ilL reaction volume. The samples were incubated at 30°C for 90 minutes. A 10 ilL aliquot of the sample was diluted with loading dye, boiled for 5 minutes and loaded onto a 10% SDS-Polyacrylamide Gel Electrophoresis (SDS-PAGE) gel. The electrophoresis conditions were 15 rnA for 30 minutes and 30 rnA for 3 hours in IX running buffer (25 mM Tris-HCI, 192 mM glycine, and 0.1 % SDS). Rainbow colored protein molecular weight marker (RPN 756, Amersham Life Science) was run with the protein samples to determine the molecular weight of the samples. The ladder bands were marked on the autoradiography film to determine the size of the samples. Gels were fixed in 25% isopropanol, 10% acetic acid in ddH20 for 30 minutes at room temperature, and the signal amplified for 30 minutes in Amplify Reagent (Ambion). Gels were dried and exposed to auto radio graphy fIlm for 12 hours. Results Candidate Gene Identification 49 We developed intragenic STS markers to known candidate genes (DatI, Adcy2, and Nhe3) which were approximately mapped to the region containing massl. PCR analysis of the STSs showed that none of the Y ACs, BACs, or cosmids comprising the physical map contained these genes. To directly identify candidate genes from the two cosmids (C1B and C13A) shown in Figure 4.1, we screened mouse brain cDNA libraries using the cosmid DNA as hybridization probes. The library screening experiments were unsuccessful at identifying any candidate cDNAs from the region; therefore, we employed an alternate strategy of sequence walking across the cosmids using the mapped markers as anchor points. The cosmid walking sequences were edited and compiled to produce the complete genomic sequence from marker Dl3SLCl4 to DI3SLC20. The complete nonrecombinant massl interval was approximately 36 kB (Figure 4.2A). Analysis of the sequence by two exon-fmding pro grams, Genefmder and Grail, predicted one multiple exon gene spanning the massl interval oriented from the distal to proxinlal end. Reverse transcription-peR (RT-PCR) with primers spanning a putative intron amplified products of the appropriate size from Frings and C57BU6J total brain RNA. Sequence analysis of these bands confrrmed that they matched the genomic sequence within the exons and identified the frrst intron-exon boundries. Figure 4.1. Detailed map of the final mass 1 interval spanning two cosmids. The D13SLC- markers from 10 to 100 are novel SSLPs identified in the region. D13SLC- marker numbers 100 to 200 are novel STSs identified in the region. The gray bar represents the region that was completely sequenced to identify cDNAs in the mass 1 interval. 51 Cloning and Analysis of massl eDNA RT -PCR experiments identified exons 24 to 28 spanning 1 Kb of open reading frame sequence that could be amplified from mouse brain RNA; however, further attempts to amplify putative adjacent exons failed. Rapid amplification of cDNA ends (RACE) completed the 3' end of the gene which contained 330 base pairs of untranslated sequence from the first stop codon to the polyA tail. Multiple 5'RACE reactions produced the complete cDNA sequence of mass} and identified three alternate transcripts each containing a unique 5' untranslated sequence. Between exon 21 and the 3'UTR, the cDNA sequences were aligned with 36 Kb of complete genomic sequence to determine the intron-exon boundaries; however, exon 1 to exon 20 were not contained in that interval (Figure 4.2A). Intron-exon boundaries for these exons were determined by sequencing cosmid genomic DNA using primers designed from the cDNA sequence. The mass} gene encodes three putative alternate transcript lengths. The longest transcript is approximately 9.4 Kb, the second 7.1 Kb, and the shortest 3.7 Kb. Analysis of the genomic structure determined that there were 35 exons in the longest transcript. As diagrammed in Figure 4.2B, each transcript contains a unique 5' untranslated region leading into the rest of the gene sequence. All three transcripts contain a possible splice variant in exon 27 where 83 base pairs of sequence are either included or removed from the transcript. Northern blot analyses of mouse RNA failed to produce conclusive data to confirm these transcript sizes. However, several autoradiograms with very long exposure times suggested that the 9.4 and 7.1 Kb 52 A ~ g (;; ~ ~ ~ ~r(oj~ ~ C\i uIS) § (.~.,.). ("'\-j U(Q t(o\j u u u u .,J u u u u CI) u u.,J ~ .,J (;j .,J ~ STSs and SSlPs ....J ....J ~ ~~~ ~ ~ CI) ~ ~ ~ CI) ~ CO) CO) C') Q C') C') c Q CO) C') CO) C') c C tS C C,... 'c,... Q.... c c Q tS Q c I Chromosome 13 C1B C20B II • I I I o-e I I I I o--e C12A Cosmids C13A I II I 0 I I I I I I 0 I II I I I ¥ I Genomic Sequence Complete _ Partial 0 .::. I mass1 Genomic Structure 8 3.7 Kb (14 exons) mass1.2 transcript mass1.1 transcript 9.4 Kb (35 exons) Figure 4.2. Diagram of the location and genomic structure of massl. A) Map of the region detailing the genomic structure of the massl gene and the alignment of the exons with the physical map. The gray bar represents the region that was completely sequenced. The exons aligned below it were located in that sequence. The white bar represents the region that was partially sequenced to identify intron-exon boundaries and the exons located in that region are shown below the bar. B) The enlarged diagram of the massl genomic structure shows the three putative transcripts and the exons that are included in each transcript. transcripts are expressed in mouse brain. Figure 4.3 shows an example northern blot where the 7.1 Kb band is visible, but the larger and smaller transcripts are not clear. 53 Analysis of the expression of mass} in mouse tissues by RT -PCR of brain, heart, kidney, liver, lung, muscle, intestine, and spleen RNA shows that the gene is predominantly found in the brain, lung, and kidney (Figure 4.4A). Further analysis of the adult mouse brain determined that the gene is expressed at similar levels in all brain regions examined including hippocampus, brain stem, cerebellum, midbrain, and cortex (Figure 4.4B). The mass} gene has also been shown to be expressed in both pure astrocyte cultures, as well as singly isolated, pooled mouse cultured cortical neuron RNA (Figure 4.4C). Mutational Analysis of massl As shown in Table 1, 17 single nucleotide polymorphisms (SNPs) were identified between Prings and C57BU6J mice within the nonrecombinant coding region, exons 21 to 35. Seven of these SNPs altered the amino acid sequence of the protein and could, theoretically, result in the Frings audiogenic seizure-susceptibility. However, listed as SNP number six in Table 1, a single base pair deletion was detected in the Frings mouse mass} gene by sequence analysis of PCR products. This deletion is located in ex on 27 before the long and short splice variants. In Figure 4.5A the sequence chromatogram shows a single G deletion at position 7009 in the Frings mouse DNA sample compared to the seizure-resistant control C57BU6J. This deletion results in a frame shift of the open reading frame changing the valine to a stop codon which is expected to produce a truncated MASS 1 9.5 Kb .... 7.5 Kb .... 4.4 Kb .... 2.4Kb .... 1.4 Kb .... 54 Figure 4.3. Northern blot analysis of massl transcripts. Total RNA from adult CF1 brain, kidney, lung and spleen, as well as embryonic day 15 CF1 brain tissue was hybridized with a cDNA probe spanning exons 30 to 35. The faint band marked by the left arrow appears to be of the size expected for massl.2. The right arrows indicate the approximate locations of the RNA size markers. A mass 1 p-actin B ~ mass1 55 ~ mass1 ~ p-actin Figure 4.4. Expression analysis of the mass} gene (all transcripts) by RT -PCR in different tissue and cell RNA samples. A) Analysis of mass} in multiple tissue RNA samples of a CFl mouse shows expression is primarily in the brain, kidney, and lung. B) Further analysis of brain RNA detected mass} expression in all regions tested. C) Pooled cultured cortical neuron RNA and cultured astrocyte RNA show mass} expression by RT-PCR. The mass 1 specific primers spanned between exon 22 to exon 23 and the expected product size is 487 base pairs. The ~-actin primers also spanned two exons and the expected product size is 327 base pairs. The ladder is in 100 base pair increments with the 600 base pair band as the darkest. Table 1 All Base Pair Changes Identified Between C57BL/6 and Frings mass1 Coding Sequence Within the Nonrecombinant Interval from Exon 21 to Exon 35 Nucleotide Polymorphism 2 3 4 5 6* Nucleotide Change C57BU6J####Frings C5904T G5941C T6186C A6847G A6973G 7009ilG G7104T C7233T A7335G G7450A G7701A A8176G C8223T T8226A G8326A A8688G Amino Acid Change C57BU6J -7 Frings No Change Alanine -7 Proline No Change Threonine -7 Alanine No Change Valine -7 Stop No Change No Change No Change Valine -7 Isoleucine No Change Isoleucine -7 Valine No Change No Change Valine -7 Isoleucine No Change 56 7 8 9 10 11 12 13 14 15 16 17 G9201C Methionine -7 Isoleucine * Denotes the deletion in Frings mouse DNA proposed to be responsible for the audiogenic seizure-susceptible phenotype. 57 protein in Frings mice. Further analysis of the deletion in other Inouse strains by gel electrophoresis showed that the deletion is only detected in Frings mouse DNA and not in any of the other seizure-resistant or seizure-susceptible mouse strains tested (Figure 4.5B). Analysis of the mass} Translated Protein Sequence The massl gene produces three putative transcripts named massl.l (9.4 Kb), mass1.2 (7.1 Kb), and massl.3 (3.7 Kb) which could produce three protein isoforms. The long transcript open reading frame contains 9327 nuc1eotides and is expected to produce an approximately 337 kilodalton (kD) protein. Likewise, the mediuln transcript open reading frame contains 6714 nucleotides and the predicted protein size is 244 kD. The short transcript open reading frame is 2865 nucleotides and the predicted protein size is approximately 103 kD. Overall the MASSI protein is strongly acidic and has a -192 charge at pH 7.0. The hydropathy plot indicated numerous putative membrane spanning domains. However, these regions did not share homology with any known membrane spanning domain sequences. The massl.3 cDNA clones containing Frings and eFI mouse sequences were used to produce MASS 1.3 protein in vitro. This was to conftrm that the 7009ilG mutation identifted in the Frings mouse massl gene sequence produced a truncated protein 709 amino acids shorter than the wild type protein. As shown in Figure 4.6 the MASSl.3 wild type translated protein is the expected size, 103 kD, and the Frings MASSl.3 is 26 kD. The positive control is luciferase protein which has the expected size of 61 kD. A B C~~'mA'TTO C57BU6 Seizure Resistant Control Mice 180 .. ~ 200 Frings Seizure F1 Susceptible [FRING X Control Mice C57BUBJ] ~""''M"CI 110 180 19<0 Frings AGS-Susceptible Mice F2 Nearest Recombinant Mice / --- ---,~~/ "/ " ~~~%~~~~~~~~~ ~<? ~~~~~""'v~C)~"Z- (f) v~~<1t:. >V ~t::. . ~V ~V ~t::.~ ...,.O'~~...,.~~O' ~c .~ c~.0'> ~...c,. g~~~ c g(J \~ CP "=' "=' 3. ~ 3.~ <6 ~ c Wild type Allele Flings Allele Figure 4.5. Analysis of the exon 27 single base pair deletion in Frings mice. A) Sequence chromatogram of the ex on 27 segment from C57BU6 and Frings DNA. The Frings mouse DNA contains a single G deletion at nucleotide. B) High resolution gel electrophoresis of PCR products from a 150 base pair segment of exon 27 spanning 7009L\G. Shown are seizure-resistant control mice (lanes 1-11), seizure-susceptible control mice (lanes 12-15), heterozygous [Frings x C57BL/6]F1 mice (lanes 16-20), and multiple Frings audiogenic seizure-susceptible mouse samples (lanes 21-32). Lanes 33 to 40 show the genotypes of the F2 nearest recombinant mice (Un represents unaffected and Aff represents affected mice). The upper band corresponds to the wild type allele and the lower band is the Frings 7009L\G allele. C) The 7009L\G is located in exon 27 of the mass} gene which is common to all putative transcripts. LIl 00 +- 220kD +- 97.5kD +- 66kD +-46kD +-30kD +- 21.5 kD 59 Figure 4.6. In vitro transcription and translation of massl.3 clones constructed fronl CFl wild type or Frings mouse cDNA. Lane 1 is the luceriferase positive control cDNA clone that produces a 61 kD protein. Lane 2 is the CF 1 wild type massl.3 protein product. The expected size for the wild type MASS 1.3 protein is 103 kD. Lane 3 shows the truncation of the Frings massl.3 protein product, and its expected size is 26 kD. The arrows indicate the approximate locations of the rainbow molecular weight bands. 60 Detailed comparisons of MASS 1 to the Na+/Ca2+ exchangers identified a small homologous domain. This homology was to the ~ 1 and ~2 repetitive elements in the third cytosolic loop of the exchanger that contains the Ca2+ regulatory binding dOlnain (Nicoll et al., 1996). Further analysis of MASS 1 detennined that this domain is repetitive and occurs 18 different times with the sequence. Alignlnent of these sequences shows several highly conserved amino acids within this domain (Figure 4.7) including a Proline-Glutamic Acid-X-X-Glutrunic Acid (PEXXE) domain that is preceded by one to three acidic residues (D or E). Three aspartic acid residues (DDD) found in the Na+/Ca2+ exchanger ~1 segment and in the segment of the very large G­protein coupled receptor-l (Accession AAD55586 by Nikkila et at, 1998) have been shown to be Ca2+ regulatory binding sites (Levitsky et at, 1994; Matsuoka et aI., 1995; Matsuoka et at, 1993), In the MASSI repeat, however, this DDD domain is not well conserved with only repeat number 3 containing the exact DDD motif, and repeats 1,9, and 18 which contain conservative substitutions of glutamic acid residues. The location of the repeat elements in shown in Figure 4.8. Analysis of the MASS 1 sequence by Pattern Match identified a multicopper oxidase I consensus sequence site was also identified in COOH terminal region of MASS 1. As shown in Figure 4.8 the multicopper oxidase I signature is complex~ spanning 21 amino acids, making this motif particularly interesting. The multicopper oxidase I site is located in the region of the MASS 1 protein that would be truncated by the Frings 7009~G mutation. Frings mice would therefore be lacking this potentially important domain. Biochemical analysis of this putative dOlnain will determine if this 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 G G TEA EGET' >GET I Ai I Consensus GTLVFLEGETEANITVTVLDDDIPELDESFLVVLL Na/Ca Ex ~1 Seg GT ..... GET Lg G Protein Seg TLFL GE\·'.E\'Y V. Na/Ca Ex ~2 Seg G E1'1 V; Possible 19 . T Possible 20 Nl rAN, 'ELDEr.:fFLV: L: (;EE:F\?V.E Lie, . DPE. PE:i~ 61 Figure 4.7. Amino acid sequence alignment of the MASS 1 repetitive domain. The first 18 lines represent the well conserved amino acid repeat domain found in MASS 1. Positions of highly conserved amino acids are boxed. The next line shows the consensus sequence for the MASS 1 repeat, and below it are the sequences of the Na+/Ca2+ ~1 and ~2 segments that share homology with the MASSI repeat. Also shown is a homologous region of the very large G-protein coupled receptor-l (Accession 55586). The boxed DDD segment has been shown to be the regulatory Ca2+ binding site. The last two lines represent two additional putative repeat domains in MASS 1 that only contain the PEXXE domain and none of the other highly conserved amino acids. Mass1 Sequence Multicopper Oxidase I Signature ~~SPNTSED~C~ ~~xxxxxxxx~x~ Y L C M I WVT M F Y W L M F Y W 62 Figure 4.8. Diagram of the massl cDNA showing the location of the MASS 1 protein repetitive domains and protein motifs. The bar represents the massl.l cDNA containing exons 1-35. The numbers above the cDNA represent the locations of the 18 repetitive domains found in amino acid sequence. Below the bar is the MASS 1 amino acid sequence followed by the multicopper oxidase I consensus sequence. The boxed residues show the MASS 1 amino acid matches to the consensus sequence. Residues 3,5,7, 17 and 21 have multiple alternatives that are listed below. X represents any amino acid. The "*,, represents the locations of the three tyrosine kinase phosphorylation motifs, the "#"identifies the locations of the two cAMP/cGMP-dependent phosphorylation motifs, and the "I\" shows the location of the glycosoaminoglycan attachment motif. 63 is a functional multicopper oxidase I donlain. Other less common motifs found within MASS 1 include three tyrosine kinase phophorylation motifs, two cAMP/cGMP­dependent phosphorylation motifs, and one glycosaminoglycan attachment motif (Figure 4.8). Finally, numerous common putative protein modification sites were identified including 52 casein kinase IT phosphorylation sites, 32 protein kinase C phosphorylation sites, 41 N-myristylation sites, and 61 N-glycosylation sites. Further analysis of the MASS 1 protein will be required to determine if any of these consensus sites are functional domains. Discussion Sequencing the cosmid DNA from the nonrecombinant mass} interval identified a single gene with multiple exons. This gene encodes three different putative transcripts and is expressed in mouse brain, lung and kidney. Computer-based BLAST nucleotide sequence similarity have shown that the mass} sequence does not share significant similarity to any sequences in the databases. Not finding any mass} cDNA sequence in the databases supports our hypothesis that it is expressed in low abundance in the brain. This hypothesis is based on the fact that screening two independent cDNA libraries for the mass} cDNA did not produce any positive clones, and low expression was detected by Northern blots, RT-PCR, and in situ hybridization. The low abundance could be due to low expression of the mass} mRNA or the message being unstable and quickly degraded. Without a cDNA clone of the mass} gene isolated from a cDNA library, the entire gene sequence had to be determined by PCR-based 5' and 3' RACE. PCR 64 approaches have been used to clone all or parts of other genes (Reppert et al., 1994), but the results may be questionable because of artifacts inherent with PCR-based assays. Problems :include producing inaccurate sequence due to Taq DNA polymerase errors and errors due to amplifying parts of homologous genes. To avoid these problenls, the mass1 fmal sequence was compiled from segments amplified with a high fidelity Pfx DNA polymerase to produce accurate sequence from multiple templates. The mass1 cDNA intronlexon boundaries were determined by sequencing using three adjacent cosmid genomic clones as templates (Cosmids C1B, C13A, and C20B). Because the cDNA sequence could be exactly matched to cosmid genomic sequence from the mass} interval it is unlikely that any parts of the proposed mass] transcripts are runplified from homologous genes. The expression of the mass1 gene has been shown to be primarily in brain, lung, and kidney tissue. This pattern of expression suggests that MASS 1 could be involved with ion transport or perhaps a secretory function. Whether the protein directly regulates membrane excitability is unknown. This would contrast with most of the previously identified seizure-related proteins which have been associated with ion channels like the voltage-gated potassium channels, KCNQ2 and KCNQ3, in hUl11an benign familial neonatal convulsions (Biervert et al., 1998; Charlier et al., 1998; Singh et al., 1998) and the voltage-gated sodium channel 131 regulatory subunit that is associated with febrile seizures plus (FS+ ) (Wallace et al., 1998). Likewise, the tottering (tg), lethargic (lh), and stargazer (stg) mouse models of absence seizures have been associated with the 0." ~, and y subunits of the voltage-gated calcium channel, respectively (Burgess et aI., 1997; Fletcher et aI., 1996; Letts et al., 1998). However, other proteins have been associated with seizures like the cystein protease inhibitor, cystatin B (Pennacchio et al., 1998), where it is not entirely clear by what luechallislu mutations in these proteins result in seizures. 65 The homology of the MASS 1 protein sequence repetitive dOluains to the sodium+-calcium2+ exchanger (Na+/Ca2+ exchanger) ~1 and ~2 repeat domains may provide an important link into determining the function of this novel protein. Although the homology between these proteins is limited to a short segment of the cytosolic loop of the exchanger, it is most likely to be functionally significant in MASS 1 because that domain is repeated 18 times within the protein sequence. The Na+/Ca2+ exchanger is a plasma membrane associated protein that cotransports three sodiulu ions into a cell and one calcium ion out of the cell using the sodium electrochemical gradient (Nicoll et aI., 1996). The Na+/Ca2+ exchanger can be regulated by intracellular calcium at a Ca2+ binding site on the third cytosolic loop that is distinct from the Ca2+ transport site. This binding site is composed of three aspartic acid residues (DDD). When Ca2+ is bound at the regulatory site the transporter is activated (Levitskyet aI., 1994; Matsuoka et a!., 1995; Matsuoka et aI., 1993). One of the MASSI repetitive domains contains the DDD domain, and three others have conservative D to E substitutions. This could mean that the function of these particular repetitive domains is involved with Ca2+ binding. Although none of the Na+/Ca2+ exchangers have been directly linked to seizures in mammals, as mentioned previously, the voltage-gated Ca channels have been associated with three absence mouse seizure models (Burgess et al., 1997~ Fletcher et 66 aI., 1996; Letts et al., 1998). Therefore, Ca2+ dysregulation is directly implicated in neuronal hyperexcitability. Loss ofNHEl, a member of the sodiuln+/hydrogen+ exchanger family, has also been associated with absence seizures in the slow-wave epilepsy (swe) Inouse model of epilepsy (Cox et aI., 1997). Together these Inodels show that the loss of regulation of both sodium and calcium ions is important in seizure susceptibility, Inaking the MASSI homology to the Na+/Ca2+ exchangers of particular interest. The Inulticopper oxidase I consensus sequence identified within the MASS 1 alnino acid sequence is also an interesting putative functional dOlnain. The multicopper oxidases are a slnall family of proteins that oxidize substrates while reducing Inolecular O2 to H20. The oxidation of substrate occurs in a series of steps during which the multicopper oxidase stores electrons to prevent the fonnation of free radicals. Two of the multicopper oxidases, Fet3p in yeast (Askwith et aI., 1994) and ceruloplasmin in humans (Harris et al., 1995), have been shown to oxidize and transport iron. Other known multicopper oxidase substrates include Mn2+ (Brouwers et aI., 1999), biogenic amines like serotonin, epinephrine, and dopamine, as well as (+ )-lysergic acid diethylamide(LSD) (Zaitsev et aI., 1999). Therefore, loss of this MASSI putative functional domain could possibly result in problems with the metabolism of iron or other metals, copper sequestering, neurotransmitter processing, and/or oxidative stress. Furthermore, the tyrosine kinase and cAMP/cGMP-dependent phosphorylation sites may be functionally significant. However, with a large protein such as MASS 1 putative protein motif sites commonly occur by chance, and detailed biochemical analysis of the 67 protein will be required to determine which, if any, of the motifs produce functional domains within the protein. Further research to determine the function of MASS 1 should help us understand how a defect in this protein results in seizures in the Frings AGS-susceptible mouse. From the mouse mass 1 cDNA, a putative human massl homolog has been identified (A. Clark, unpublished results). Through mapping and characterization of the human homolog it may be possible to find an association of mass 1 with a human epilepsy disorder. Together, the studies of the mouse and human MASS 1 will provide insight into the function of this novel protein and the mechanism through which it leads to seizure-susceptibili ty. References Askwith, C., Eide, D., Van Ho, A., Bernard, P. S., Li, L., Davis-Kaplan, S., Sipe, D. M., and Kaplan, J. (1994). The FET3 gene of S. cerevisiae encodes a multicopper oxidase required for ferrous iron uptake. Cell 76, 403-10. Biervert, C., Schroeder, B. C., Kubisch, C., Berkovic, S. F., Propping, P., Jentsch, T. J., and Steinlein, O. K. (1998). A potassium channel mutation in neonatal human epilepsy. Science 279,403-6. Brouwers, G. J., de Vrind, J. P., Corstjens, P. L., Cornelis, P., Baysse, C., and deVrind de Jong, E. W. (1999). cumA, a gene encoding a multicopper oxidase, is involved in Mn2+ oxidation in Pseudomonas putida Gb-1. Appl. Env. Microbiol. 65, 1762-8. Burgess, D. L., Jones, J. M., Meisler, M. H., and Noebels, J. L. (1997). Mutation of the Ca2+ channel beta subunit gene Cchb4 is associated with ataxia and seizures in the lethargic (lh) mouse. Cell 88, 385-92. Charlier, C., Singh, N. A., Ryan, S. G., Lewis, T. B., Reus, B. E., Leach, R. J., and Leppert, M. (1998). A pore mutation in a novel KQT -like potassium channel gene in an idiopathic epilepsy family [see comments]. Nat. Genet. 18, 53-5. 68 Cox, G. A., Lutz, C. M., Yang, C. L., Biemesderfer, D., Bronson, R. T., Fu, A., Aronson, P. S., Noebels, J. L., and Frankel, W. N. (1997). Sodiurn/hydrogen exchanger gene defect in slow-wave epilepsy mutant mice [published erratum appears in Cell 1997 Dec 12;91(6):861]. Cell 91, 139-48. Fletcher, C. F., Lutz, C. M., TN, O. S., Shaughnessy, J. D., Jr., Hawkes, R., Frankel, W. N., Copeland, N. G., and Jenkins, N. A. (1996). Absence epilepsy in tottering mutant mice is associated with calcium channel defects. Cell 87, 607-17. Harris, Z. L., Takahashi, Y., Miyajima, H., Serizawa, M., MacGillivray, R. T., and Gitlin, J. D. (1995). Aceruloplasminemia: molecular characterization of this disorder of iron metabolism. Proc. NatI. Acad. Sci. USA 92, 2539-43. Letts, V. A., Felix, R., Biddlecome, G. H., Arikkath, J., Mahaffey, C. L., Valenzuela, A., Bartlett, F. S., 2nd, Mori, Y., Campbell, K. P., and Frankel, W. N. (1998). The mouse stargazer gene encodes a neuronal Ca2+-channel gamma subunit [see comments]. Nat. Genet. 19,340-7. Levitsky, D.O., Nicoll, D. A., and Philipson, K. D. (1994). Identification of the high affinity Ca(2+)-binding domain of the cardiac Na(+)-Ca2+ exchanger. J. BioI. Chern. 269, 22847-52. Matsuoka, S., Nicoll, D. A., Hryshko, L. V., Levitsky, D.O., Weiss, J. N., and Philipson, K. D. (1995). Regulation of the cardiac Na( + )-Ca2+ exchanger by Ca2+. Mutational analysis of the Ca(2+)-binding domain. J. Gen. PhysioI. 105,403-20. Matsuoka, S., Nicoll, D. A., Reilly, R. F., Hilgemann, D. W., and Philipson, K. D. (1993). Initial localization of regulatory regions of the cardiac sarcolemmal Na( + )-Ca2+ exchanger. Proc. NatI. Acad. Sci. USA 90, 3870-4. Nicoll, D. A., Hryshko, L. V., Matsuoka, S., Frank, J. S., and Philipson, K. D. (1996). Mutagenesis studies of the cardiac N a( + )-Ca2+ exchanger. Ann. N. Y. Acad. Sci. 779, 86-92. Nikkila, H., Nunez, B. S., Pascoe, L., Curnow, K. M., and White, P. C. (1998). Sequence similarities between a novel putative G-protein coupled receptor and Na+/Ca2+ exchangers define a novel cation binding domain: UT Southwestern Medical Center. Pennacchio, L. A., Bouley, D. M., Higgins, K. M., Scott, M. P., Noebels, J. L., and Myers, R. M. (1998). Progressive ataxia, myoclonic epilepsy and cerebellar apoptosis in cystatin B-deficient mice. Nat. Genet. 20, 251-8. 69 Reppert, S. M., Weaver, D. R., and Ebisawa, T. (1994). Cloning and characterization of a mammalian melatonin receptor that mediates reproductive and circadian responses. Neuron 13,1177-1185. 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., McHarg, M. L., Gagnon, D., Rosales, T. 0., Peiffer, A., Anderson, V. E., and Leppert, M. (1998). A novel potassium channel gene, KCNQ2, is mutated in an inherited epilepsy of newborns [see comments]. Nat. Genet. 18, 25-9. Wallace, R. H., Wang, D. W., Singh, R., Scheffer, I. E., George, A. L., Jr., Phillips, H. A., Saar, K., Reis, A., Johnson, E. W., Sutherland, G. R., Berkovic, S. F., and Mulley, J. C. (1998). Febrile seizures and generalized epilepsy associated with a mutation in the Na+-channel beta1 subunit gene SCN1B. Nat. Genet. 19,366-70. Zaitsev, V. N., Zaitseva, I., Papiz, M., and Lindley, P. F. (1999). An X-ray crystallographic study of the binding sites of the azide inhibitor and organic substrates to ceruloplasmin, a multi-copper oxidase in the plasma. J. BioI. Inorg. Chern. 4, 579-87. CHAPTERS . DISCUSSION AND SUMMARY The Importance of the Frings Audiogenic Seizure-Susceptible Mouse Model of Epilepsy There are many genetic rodent models of epilepsy and a wide array of neurological phenotypes including seizures associated with each of them. The nonconvulsive seizure models include tottering (tg), lethargic (lh), stargazer (stg), and slow-wave epilepsy (swe) mice, and their phenotypes include ataxia and spike wave discharges. These phenotypes are due to mutations in different subunits of the voltage-gated calcium channel for the frrst three, and a member of the sodium/hydrogen exchanger family for the last one (Burgess et aI., 1997; Cox et al., 1997; Fletcher et aI., 1996; Letts et aI., 1998). As for convulsive seizure models, the weaver mouse shows a neurological phenotype of ataxia, tremor, and seizures in heterozygotes. This seizure phenotype is due to a mutation of the G protein-gated inwardly rectifying potassium channel GIRK.2 (Signorini et aI., 1997). Another convulsive model is the opisthotonus mouse which displays ataxia and convulsions in younger mice. This phenotype is due to mutations in the type 1 inositio11,4,5-triphosphate receptor (IP3R1) protein which is responsible for IP3-dependent release of calcium from intracellular stores (Matsumoto et aI., 1996). These proteins are all directly involved with ion conductance across neuronal membranes, and all share the predominant feature of ataxia along with seizure 71 phenotypes (reviewed by Jen and Ptacek, in press). It is interesting to fmd that massl bears no resemblance to any ion channel. This finding makes the Frings mouse even more unique among rodent models. First, it is a naturally occurring single gene model of generalized seizures with an identified and characterized genetic mutation that can now be manipulated to determine - exactly how MASS 1 functions in the normal brain and how loss of its function results in audiogenic seizures. Second, the function of MASS 1 is unlikely to be as an ion channel because of its lack of homology to any of the known ion channels, unless massl defmes a new class of ion channels. Seizure-susceptible trangenic and knockout mice have been produced that modulate targets other than ion channels. Examples include the calmodulin kinase II (X­subunit knockout mice which display limbic seizures (Butler et aI., 1995), neuropeptide Y knockout mice which display mild spontaneous seizures (Baraban et aI., 1997), glial glutamate transporter GLT-1 knockout mice which have lethal spontaneous seizures (Tanaka et al., 1997), and the serotonin 2C subunit (5HT2c) knockout mice which have an increased susceptibility to audiogenic seizures as well as rare spontaneous lethal seizures (Tecott et aI., 1995). All these models have complex neurological syndromes due to the manipulation or deletion of a critical protein during development and in adulthood. Therefore, the Frings mice are an important naturally occurring genetic model of a discreet seizure phenotype without other complicating neurological factors that will provide significant information on a novel mechanism of seizure-susceptibility as well as CNS excitability in general. 72 The fact that mass} is a novel gene that does not resemble an ion channel may lead to new pharmacological possibilities for the treatment of epilepsy. The majority of current antiepileptic drugs target ion channels and possess negative side effects that make them less than ideal therapeutic agents. In addition, at least 25 % of patients remain refractory to these drugs (Graves and Leppik, 1993). The most common targets include voltage-gated Na2+ channels which are blocked by most antiepileptic drugs including phenytoin, carbamazepine, valproate, and lamotrigine; the T-type calcium channel which is blocked by ethosuximide, and GAB A-mediated chloride channel currents which are enhanced by benzodiazepines and barbiturates (reviewed by White, 1997). These channels are essential for normal eNS activity, and, therefore, modulation of channel function produces myriad effects that are both beneficial for the treatment of epilepsy and detrimental to normal neuronal function. Mouse models like the Frings audiogenic seizure-susceptible mouse that are genetically linked to novel proteins that can be pharmacologically targeted may yield effective treatments without adverse effects. Dissertation Summary Results from this dissertation establish the Frings mouse as a Inonogenic lllodel of generalized seizures evoked by sensory stimulation. The breeding data for the intercross and backcross mice showed the appropriate ratios of unaffected and affected progeny expected for a single gene disorder. This result Ineans that the mutant mass} gene is sufficient to produce audiogenic seizures in Frings mice. However, the protein function could be a link in a signaling pathway, and loss of this link disrupts downstream effectors. In this case massi would have an indirect effect on neuronal hyperexcitability. 73 The genetic mapping began with 257 (Frings X C57BL/6J) X Frings backcross mice which enabled us to map 'the gene to the middle of chromosome 13. However, it soon became clear that [me mapping would be required to limit the aInount of DNA that would have to be screened to identify massi. Approximately 1200 (Frings X C57BL/6J)F2 intercross mouse genotypes were added to the original sample set which significantly decreased the interval of massi. The interval was then slnall enough to complete a dense physical map spanning the region. The physical map allowed us to identify new SSLP markers which further narrowed the critical region. These new SSLPs will facilitate other mapping projects by providing researchers with additional markers in this region of chromosome 13. Finally, the [me mapping and physical mapping combined limited the massi interval to two overlapping cosmid clones. Identification of candidate gene cDNAs began by using the two cosmid clones to screen mouse brain cDNA libraries. The library screenings did not produce a single positive cDNA clone from the massi region. Therefore, large scale sequence walking was employed as an alternate strategy for identifying a candidate cDNA from the massi region. Exon screening programs identified putative exons within the genomic sequence which were confrrmed by RT -PCR analysis in mouse brain RNA. Various PCR approaches were utilized to clone and sequence the entire cDNA. COlnparing sequences of PCR products from Frings and C57BL/6J spanning the nonrecolTlbinant interval identified 17 single nucleotide polymorphisms between the two strains. One of these polymorphisms was a single base pair deletion found at position 7009 which changed the valine to a stop codon ending the open reading frame of the gene. The 7009AG deletion was only found in Frings mouse DNA and not in 15 other seizure­resistant and seizure-susceptible control mice tested. 74 At this time we concluded that this gene was the massl gene that was responsible for audio genic seizure-susceptibility in Frings mice. The frrst piece of evidence supporting this conclusion was that this single gene spanned the entire 36 Kb nonrecombinant interval that had to contain the maSSllTIutation. Second, we did not fmd any other putative genes within the region; however, it is possible that a second, or even third, gene that has not yet been identified is located within the intronic spaces. Third, the 7009AG mutation was identified within this gene and was found exclusively in Frings mice. Finally, the 7009AG is a significant mutation that is expected to produce a truncated protein in Frings mice that would be missing the fmal 709 amino acids of the protein. This expectation was confirmed to occur in vitro. Characterizing the mass 1 gene determined that there are three putative transcript lengths. The longest transcript was called massl.l and it is 9.4 Kb including the 5' and 3' untranslated regions. Two alternate putative 5' untranslated regions leading into the open reading frame of the cDNA were also identified. The second, massl.2, was 7.1 Kb and the third, massl.3, was 3.7 Kb. These transcripts have not all been confrrmed to exist, and there may be a tissue specific pattern of expression for the alternate transcripts. Northern blots would have provided this information, but multiple experiments did not show convincing bands even after long exposure times. The most logical explanation for these results is the low abundance of massl InRNA and inefficient transfer of the longer transcripts. However, all possible hybridization conditions were not exhaustively tested and one may prove to be successful in future experiments. 75 Northern blots did consistently show indistinct bands that probably correspond to the 9.4 and 7.1 Kb transcript lengths. However, the 3.7 Kb was not identified at any time. Furthermore, the massl.l putative human homolo g has been cloned (A. Clark, unpublished results) which further supports that this is a real transcript. However, massl.2 and massl.3 have not yet been cloned from human RNA so those transcripts have not been confinned. Finally, massl.2 results from splicing to an alternate exon 7, labeled exon 7a in Figure 4.3, which leads into 5' untranslated sequence. This suggests that massl.2 is not an artifact produced by incomplete splice of the mRNA. In contrast, the putative 5' untranslated region of.massl.3 is continuous with its first exon, exon 21; therefore this transcript could be the result of PCR amplification of unedited mRNA. Another variation identified in the massl transcripts is the exon 27 long (27L) and short (27S) splice variants. There is a putative AGGT consensus splice site at position 7117, and 27S is the result of a splice occuring at this site. The 27S transcript has this splice site connected to the 5' splice site of exon 28 removing 83 base pairs from the transcript. This changes the open reading frame and results in a truncated protein not significantly longer than the Frings truncated protein. Interestingly, both 27L and 27S are expressed in human brain RNA, and these 83 base pairs are 100% 76 - . identical between the mouse and human sequences whereas the rest of the sequence is on average 80-85% identical (A. Clark, unpublished results). Because this splice . variant changes the reading frame and would produce a truncated protein, it would seem unlikely that this variant is translated in vivo. The expression pattern of the mass} gene is notable and requires further discussion. Searches for genes with expression primarily in the brain, kidney, and lung identified functionally divergent genes ranging from Kir7.1 inwardly rectifying potassium channel (Doring et al., 1998), to the HNF-3/Forkhead Homologue-4 transcription factor (Tichelaar et aI., 1999) to a novel serine protease (Yamamura et aI., 1997). Therefore, it would seem that the mass} expression pattern reveals little about its possible function. Although the Frings mouse expresses mutant mass} in brain, lung, and kidney, no defects have been observed in the function of either the kidney or lung. However, detailed studies on these tissues have not been conducted and may yield interesting results in these mice. The expression of mass} mRNA in adult tissues is low and has been a problem throughout the project. When the mass} gene was fmally identified by large scale sequencing, the original brain libraries that were screened to identify candidate cDNAs were tested for mass} and it was determined that one contained no detectable copies of the mass} gene and the other library had too few copies to guarantee at least one positive clone under normal screening conditions. This was the fITst indication that the abundance of the mass} gene was low in the mouse brain. Further study of the gene in brain, kidney, and lung supported this observation because RT-PCR detected low 77 mass} mRNA expression in mouse tissues, Northern blots did not show conclusive mass} RNA bands after long