KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society
INTERNATIONAL BEHAVIOURAL AND NEURAL GENETICS SOCIETY
KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior THIRD ANNUAL GENERAL MEETING June 22 - June 23, 2000 Brighton, UK
Executive Committee
President: Wim E. Crusio (Orléans, France!) President-elect: Douglas Wahlsten (Edmonton, Canada) Past-President: Hans-Peter Lipp (Zürich, Switzerland) Secretary: Enrico Alleva (Rome, Italy) Treasurer: Robert Gerlai (Indianapolis, IN,, USA) Members-at-large: Terry R. McGuire (Piscataway, N.J., USA) Osvaldo Giorgi (Cagliari, Italy) Catherine Belzung (Tours, France)
Local Host
Wim E. Crusio (Orléans, France)
Programme Committee
Philip Gorwood (Chair; Colombes, France) Joshua Dubnau (Cold Spring Harbor, NY, USA) Seth Grant (Edinburgh, UK) Richard Paylor (Houston, TX, USA) Marina Picciotto (New Haven, CT, USA)
This meeting was generously supported by: Current Trends Ltd., London, UK Elsevier Science, Amsterdam, The Netherlands Ely Lilly and Company, Indianapolis, IN, USA Parke-Davis Pharmaceutical Research, Division of Warner-Lambert Co., Ann Arbor, MI, USA
1 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society Program
Thursday, June 22, 2000
7.30 am-5.30 pm Registration
8.00-8.30 am Opening session
8.30-9.30 am Plenary Lecture. Jacqueline N. Crawley (Bethesda, MD, USA). What’s wrong with my mouse? Behavioral phenotyping strategies and applications.
9.30-10.30 am Symposium. Genetics of eating disorders. Chair: David A Collier (London, UK)
10.30-11.00 am David A. Collier, Andreas Karwautz, Janet Treasure (London, UK). Gene-environment studies of anorexia nervosa in discordant sister pairs.
11.00-11.30 am Iain C. Campbell, Nigel Brown, Ann Ward and Janet L. Treasure (London, UK). Energy sensing in subjects recovered from anorexia nervosa.
10.30-11.00 am Coffee Break
11.00-12.30 am Contributed papers. Genetic analysis of complex phenotypes. Chair: Douglas Wahlsten (Edmonton, Ont., Canada)
11.00-11.15 am Fred van Leuven (Leuven, Belgium). Transgenic mouse models for Alzheimer's disease.
11.15-11.30 am P. Gorwood, F. Limosin, P. Batel, C. Boni, M. Hamon, and J. Adès (Colombes, France). The DAT1 gene is involved in severe alcohol withdrawal in male alcoholics.
11.30-11.45 am R. Gerlai, P. Pisacane, and S. Erickson (Indianapolis, IN, USA). Gene targeting and compensation: Behavioral effects of null mutations in the heregulin - ErbB system.
11.45-12.00 am Richard Lathe (Edinburgh, UK). The enteroceptive hippocampus.
12.00-12.15 am Catharine H. Rankin, Jacqueline K. Rose, Kenneth Eng, and Karla Kaun (Vancouver, BC, Canada). Genetic dissection of habituation of the tap withdrawal response in C. elegans.
12.15-12.30 am Dennis A. Stephenson, Julie Gilchrist, Dionne Peterson, Sherry Tuner, Corike Nuibe, and George A. Carlson (Great Falls, MT, USA). ENU induced behavioral mouse mutants which may in- volve APP or PrP.
12.30-2.00 pm Lunch break
2.00-3.00 pm Poster Session I
1/ M. Adrover, C. Blanco, E. Rial Verde, V. Cheli, G. Sanchez, L. Alché, E. Kornisiuk, E. Martín, A. Epstein, and D. Jerusalinsky (Buenor Aires, Argentina). Expression of NMDA receptor NR1 sub- unit sequences carried by a HSV-1 viral vector in rat hippocampus interfered with habituation.
2/ S.P. Baron (Ann Arbor, MI, USA). A titrating, continuous visual signal-detection operant schedule for mice.
2 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society
3/ S. Viñas, S. Lewis, S. Barreau, C. Ducottet, A. Aubert, and C. Belzung (Tours, France). Strain differences in two animal models of depression : the forced swimming test and the chronic mild stress procedure.
4/ J. Adriaan Bouwknecht, Theo H. Hijzen, Jan van der Gugten, Rene Hen, and Berend Olivier (Utrecht, The Netherlands). Ethanol intake is not elevated in male 5-HT1B receptor knockout mice.
5/ Igor Branchi, Zoë Bichler, Marie-Claude Gonzalez, and Danièle Migliore-Samour (Orléans, France). YAC polytransgenic mouse, an animal model to assess possible alterations of choliner- gic basal forebrain in Down syndrome.
6/ B.J. Caldarone, C.H. Duman, S.L. King, and M.R. Picciotto (New Haven, CT, USA). Fear condi- tioning, latent inhibition, and habituation in mice lacking high affinity nicotinic receptors.
7/ S. Chiavegatto, V.L. Dawson, L.A. Mamounas, S.H. Bora, V.E. Koliatsos, T.M. Dawson, and R.J. Nelson (Baltimore, MD, USA). Altered gene expression of central serotonin 5-HT1A and 5-HT1B re- ceptors in male nNOS null mice.
8/ J.P. Hatcher, P.T. Davey, S. Bingham, P. Overend, A.A Parsons, and J.B. Davis (Harlow, UK). Vanilloid receptor-1 has a major role in thermal hyperalgesia.
9/ J.P. Hatcher, D.C. Rogers, C. Reavill, J.J. Hagan (Harlow, UK). Use of SHIRPA to investigate the behavioural phenotype of the Coloboma (cm/-) mouse.
10/ M. Hafezparast, S.J. Nicholson, A.S. Witherden, N. Bermingham, S. Ball, J. Peters, D.C. Rogers, J.E. Martin, E.M.C. Fisher (London, UK). Loa (legs at odd angles) a mouse model of motor neu- rone dysfunction: mapping and progress towards isolation of the causal gene.
11/ Josh Dubnau, Scott Gossweiler, Ulli Certa, Rod Scott, Clemens Broger, Martin Neeb, Jerry Yin, Jan Mous, and Tim Tully (Cold Spring Harbor, NY, USA). Functional genomics of long-term mem- ory.
12/ A. Holmes, J.G. Hohmann, E. Yared, R.A. Steiner, and J.N. Crawley (Bethesda, MD, USA). Vari- ability in an anxiety-related phenotype in galanin overexpressing transgenic mice.
13/ T. Hough, P.M. Nolan, J. Peters, E.M.C. Fisher, J. Martin, M. Browne, S. Rastan, L. Vizor, S.D.M. Brown, and A.J. Hunter (Harwell, UK). Clinical biochemistry screens can complement behavioural screens in mutagenised mice.
14/ Christopher Janus, David Westaway, Jacqueline Pearson, Peter St. George-Hyslop (Toronto, Ont. Canada). Impaired spatial leaning and memory in APP CRND8 transgenic mice.
3.00-4.00 pm Plenary lecture. Mario de Bono (Cambridge, England). The genetics of food-induced social foraging in C. elegans.
4.00-4.30 pm Coffee Break
4.30-6.00 pm Symposium. Behavioural-neurogenetic analysis of aggressive behaviour. Chair: Pierre L. Roubertoux (Orléans, France).
4.30-5.00 pm S.C. Maxson, A. Canastar, and C. Bishop (Storrs, CT, USA). Sex reversals (XX and XY females and XX and XY males), mating behaviors and aggressive behaviors in mice.
5.00-5.30 pm A.H. Veenema, G.A. Van Oortmerssen, A.J.H. De Ruiter, B. Bohus, J.M. Koolhaas, and F Sluyter (Nijmegen, The Netherlands). Neurobehavioral effects of the Y chromosome in wild house mice.
3 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society
5.30-6.00 pm Stéphane Mortaud, Laurent Nicolas, Isabelle Le Roy, and Pierre L. Roubertoux (Orléans, France). Attack behavior in mice: implication of the sts gene mapped on the pairing region of the X-Y chromosomes.
6.00-6.45 pm IBANGS Business Meeting (Members only)
7.30-10.30 pm Banquet (Brighton Metropole Hotel)
Friday, June 23, 2000
8.00 am-5.30 pm Registration
8.00-9.00 Plenary Lecture. R. Bourtchouladze, T. Abel, A. Morozov, P. Nguyen, I. Muzzio, and E.R. Kandel (New York, NY, USA). Genes important for long-term memory and synaptic plasticity.
9.00-10.00 am Contributed paper session. Analysis of single-gene polymorphisms. Chair: Jacqueline N. Crawley (Bethesda, MD, USA)
9.00-9.15 am F. Cirulli, S. Capogrossi Colognesi, M. Bianchi, A. Panerai, and E. Alleva (Rome, Italy). Reduced sensitivity to morphine place conditioning in IL-6 KO mice. Interleukin-6 (IL-6) is a cytokine in- volved both in inflammatory responses and in brain function.
9.15-9.30 am J.M. Delabar, M. Rachidi, C. Lopes, C. Chabert, P. Roubertoux, C. Vayssettes, A.L. Dele- zoide, and E.M. Rubin (Paris, France). Abnormal cerebellar folial pattern in Yac transgenic mice contain- ing a patterning gene from the Down syndrome chromosomal region-1.
9.30-9.45 am C. Dubertret, P. Gorwood, L. Gouya, J.C. Deybach, and J. Adès (Colombers, France). The haplotype relative risk method and the genes coding for dopamine receptors in schizophrenia.
9.45-10.00 am Alicja L. Markowska, Alena Savonenko, and Katrin Andreasson (Baltimore, MD, USA). Overexpression of cyclooxygenase-2 (cox-2) leads to cognitive impairment.
10.00-10.30 am Coffee Break
10.30-12.30 am Symposium. Human Neurodegenerative Diseases: Deciphering the Cognitive Impair- ment Dilemma Using Genetically Engineered Mice. Chair: Christopher Janus (Toronto, Ont., Canada)
10.30-11.00 am Fred van Leuven (Leuven, Belgium). Transgenic mice and Alzheimer pathology: no need for plaques or tangles?
11.00-11.30 am Paul Chapman (Cardiff, UK). Models of Alzheimer’s disease: More than just messing around with genes?
11.30-12.00 am D. Westaway (Toronto, Ont., Canada). Transgenic mouse models of Alzheimer’s Dis- ease: long-haul or home-stretch?
12.00-12.30 am Christopher Janus (Toronto, Ont., Canada). Behavioural mouse models of Alzheimer's Disease: A hit-or-miss approach?
12.30-2.00 pm Lunch break
4 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society
2.00-2.30 pm Poster Session II
15/ T. Lemberger, T. Mantamadiotis, O. Kretz, D. Gau, T. Steckler, and G. Schuetz (Heidelberg, Germany). Conditional mutagenesis of CREB in dopaminoceptive neurons.
16/ S. Lewis, M.A. Simoneau, S. Bailly, F. Guillou, B. Baron, and C. Belzung (Tours, France). Behav- iour of mice expressing human transferin in brain.
17/ A.L. Vyssotski, G. Dell’Omo, D.L. Vyssotski, D.P. Wolfer, L. Minichiello, R. Klein, I.I. Poletaeva, and H.-P.Lipp (Zurich, Switzerland). Mice lacking the neurotrophin receptor TrkB in the forebrain show intact spatial memory but impaired behavioral flexibility in a semi-naturalistic set-up.
18/ K.L. McIlwain and R.E. Paylor (Houston, TX, USA). Effects of testing experience in a mouse behavioral test battery.
19/ Claudia F. Plappert, Peter K.D. Pilz, and H.-U. Schnitzler (Tübingen, Germany). Sensitization of the acoustic startle response in mice differs between two inbred strains and is influenced by gly- cine.
20/ C. Reavill, P. Nelson, J. Latcham, and J.J. Hagan (Harlow, UK). Prepulse inhibition and startle responses in dishevelled (Dvl1-/-) mice.
21/ Michael Regulski, Grigori Enikolopov, and Tim Tully (Cold Spring Harbor, NY, USA). Genetics of NO signaling in the adult Drosophila melanogaster brain.
2.30-3.30 Plenary Lecture. Peter McGuffin (London, UK). Heredity, Hazards and the Origins of De- pressive Disorder.
3.30-4.00 pm Coffee Break
4.00-6.00 Symposium. The Cerebellum, Calcium and Behavior. Chair: M. Meyer (Martinsried, FRG)
4.30-5.00 pm H. Daniel (France). Calcium signalling in cerebellar Purkinje cells of mice lacking mGluR1 or InsP3R1 receptors: Role in long-term depression.
4.30-5.00 pm J. Barski and M. Meyer (Martinsried, FRG). Adaptation and learning of motor coordination in mice depends on calbindin-D28k in cerebellar Purkinje cells.
5.00-5.30 pm S.N. Schiffmann, G. Cheron, A. Lohof, P. d’Alcantara, M. Meyer, M. Parmentier, and S. Schurmans (Brussels, Belgium). Calretinin, cerebellar network activity and motor coordination.
5.30-6.00 pm B. Schwaller (Fribourg, Switzerland). The lack of parvalbumin affects the body and the mind.
5 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society
M. Adrover1, C. Blanco2, E. Rial Verde1, V. S.P. Baron1. A titrating, continuous visual Cheli1, G. Sanchez2, L. Alché3, E. Kor- signal-detection operant schedule for nisiuk1, E. Martín2, A. Epstein4, and D. Je- mice. rusalinsky1. Expression of NMDA receptor Interest in genetic influences on behav- NR1 subunit sequences carried by a HSV-1 ior has increased and off particular interest viral vector in rat hippocampus interfered have been the effects of gene manipulation on with habituation. cognitive behaviors such as learning. How- The NMDA receptor antagonist AP5 ever, disruptions of learning could be inter- (icv) selectively impaired place learning with- preted as disruptions of behaviors, such as out affecting visual discrimination learning attention, underlying and necessary for the (Morris et al., Nature 319:774, 1986). Infusion expression of learning. The goal of the current of NMDA receptor antagonists into the dorsal experiments was to develop a method of hippocampus after training caused retrograde measuring attention in mice. Experiments amnesia of an inhibitory avoidance task (Jeru- were performed in operant chambers contain- salinsky et al., Behav. Neural Biol. 58:76, ing three horizontally-placed nose-poke holes 1992). To further clarify the participation of the on one wall and an aperture on the opposite NMDA receptor in learning/memory process- wall to which a dipper of 0.01 ml milk could be ing, viral vectors derived from Herpes Simplex presented. To obtain milk reinforcement mice Virus type-1 were constructed carrying either were required to poke into a hole within 2 sec the sense NR1(+) or antisense NR1(-) se- of the offset of a light flash emitted from that quences of the NR1 subunit gene of NMDA-R hole. The initial signal duration of 100 ms was and the GFP gene. The expression was cor- decreased by 10 ms upon a correct response roborated indirectly by GFP expression, and and increased by 10 ms upon an incorrect directly by binding experiments with a radio- response or omission. Signals were randomly labelled NMDA antagonist and by Western presented from the three nose-poke holes. A blots. The vectors were then infused into dor- session consisted of 135 trials. Stable behav- sal hippocampus of adult male Wistar rats. ior was determined during trials 20 –135. After three days, they were put in an open C57BL/6J mice (n=6) maintained an average field and various behavioural parameters were stimulus duration of 20 ms (± 2) whereas observed to study habituation. The animals 129S3/SvJ mice (n=4) maintained an average infused either with the vehicle, the GFP vector stimulus duration of 68 ms (± 18.6) (t-test; or the NR1(+) vector behaved as naive con- p<0.001 between strains). When the inter-trial trols in the training session. All of them interval was increased from 12 (± 3) to 20 sec showed a significant decrease in both cross- (± 3) the mean signal duration maintained by ings -from one quadrant to another- and rear- the C57BL/6J mice increased to 44.5 ms (± 9) ings; but the NR1(+) injected group showed (t-test; p<0.001), whereas the signal duration an even higher statistically significant de- maintained by the129S3/SvJ mice was not crease in both crossings -from one quadrant altered. The results of these preliminary stud- to another- and rearings than control animals. ies are consistent with those reported with a The rats infused with the NR1(-) vectors per- five-choice serial reaction time task in mice formed as control animals in the training ses- (Humby et al., EJN 11, 1999) and indicate that sion, except for a diminished number of the current task can be interpreted as a meas- grooming; in the test session they showed no ure of attention and should be useful in the decrease in crossings but in rearings. Hence, study of the effects of genetic manipulation on this might be interpreted as either difficulties in cognition. completely recording the trace or as retro- 1Parke-Davis Pharmaceutical Res. Div., War- grade amnesia. These results allow to suggest ner-Lambert Co., Ann Arbor, MI 48105, USA. that the NR1 subunit of the NMDA receptor in the hippocampus is directly involved in the J. Barski1 and M. Meyer1. Adaptation and mechanisms leading to habituation in such a learning of motor coordination in mice de- way that even a slight change in the availabil- pends on calbindin-D28k in cerebellar Pur- ity of this subunit interfered with some of the kinje cells. behavioural parameters recorded. Previously we have demonstrated a 1Inst. Cell Biol. & Neurosci. School Med., 2 3 specific type of ataxia in mice carrying a tar- Dept. Anatomy, School Veter., and Dept. geted nullmutation of the calbindin D28k gene Virology, School Sci., Univ. Buenos Aires, Ar- 4 which encodes a protein thought to act as a gentina. Centre de Genetique Moleculaire et cytosolic calcium buffer in various populations Cellulaire, Univ. Claude Bernard, Lyon. of neurons (Airaksinen et al., 1997, PNAS 94, France. 1488-1493). This behavioral phenotype is rec- ognized exclusively in environmental condi-
6 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society tions which are novel to the mice and require systems of memory utilize different strategies adaption of movement. It is, in contrast, im- and they recruit different neural structures possible to distinguish genotypes by inspec- raises questions: Do they share the same or tion when animals are moving in their stan- different cellular and molecular mechanisms? dard cage environment. This ataxia can be Can molecular genetics, with its ability to dis- described quantitatively by counting slips the cern homology relationship, discern common- mice make when walking along an elevated alities or differences in these two very different runway which is 100 cm long, 2 cm wide and memory systems? One evidence to shared which carries obstacles 0.5 cm in height every mechanisms has come from studies of stages 10 cm. Using this assay, ataxia is clearly de- in memory storage. Both explicit and implicit tectable already in young adult (6 week old) memory display at least two temporally distinct nullmutants and also, to a lesser extent, in memory processes: short-term and long-term heterozygous mice. The phenotype is not re- memory. Moreover, in each case, long-term flected in any gross developmental parama- memory requires new protein synthesis. ters such as weight or size and appears to be Studies in Aplysia and Drosophila pro- largely independent of structural changes in vided the initial evidence that cAMP signaling the CNS. Reasoning that the cerebellum is pathways play an important role in implicit likely to be involved and that cerebellar Pur- forms of learning in invertebrates. Here we kinje cells are the only cerebellar neurons ex- have taken a genetic approach to explicit pressing calbindin and furthermore, belong to memory storage. We have generated trans- the few neuronal populations containing very genic with reduced PKA activity in hippocam- high levels of this protein we focussed further pal neurons [R(AB)-mice], and mice express- analysis on the cerebellum. In cerebellar ing a dominant negative small GTPase Rap1, slices the nullmutation is accompanied by al- , in the forebrain under the control of the tered fast components of synaptically evoked tetO/tTA system. Rap1 mice show a reduction calcium transients but is without effect on the of coupling of the cAMP cascade to the MAP elicited electrical response. To prove the criti- kinase signaling. We explored both transgenic cal involvement of Purkinje cells we have em- mice in biochemical, physiological, and behav- ployed a conditional targeting strategy. To this ioral studies. end mice with a floxed calbindin allele and a We found that PKA and Rap1 play a transgenic line expressing Cre recombinase critical role in the hippocampus in initiating the under regulatory control of the L7 gene were molecular events leading to the consolidation generated. Preliminary analysis of crosses of short-term changes in neuronal activity into homozygous for the floxed allele and het- long-term memory. Taken together with stud- erozygous for the transgene reveal highly ies demonstrating the crucial role of CREB- specific and efficient recombination. Condi- induced transcription in implicit and explicit tional mutants still display the ataxic pheno- memory storage (Aplysia, Drosophila, and type described above. Furthermore, they mice), our results indicate that quite different seem to profit less than wildtype animals from memory processes use a restricted number of previous experience in the runway when re- mechanisms for converting short- to long-term tested a week later. Thus, Purkinje cells of the memory. cerebellar cortex and their fast calcium - 1Columbia University and 2NY State Psychiat- handling capabilities are critical for motor ad- ric Institute, 3HHMI, New York, NY 10032, aptation and possibly also learning, processes USA. 4University of Pennsylvania, Philadel- that presumably are essential for survival in phia, PA, USA. 5University of Alberta School natural habitats. of Medicine, Edmonton, T6G 2H7 Canada. 1Max-Planck-Institute of Neurobiology, D- 82152 Martinsried, FRG. J. Adriaan Bouwknecht1, Theo H. Hijzen1, Jan van der Gugten1, Rene Hen2, and Ber- R. Bourtchouladze1,2, T. Abel4, A. Moro- end Olivier1,3,4. Ethanol intake is not ele- zov1,3, P. Nguyen5, I. Muzzio1, and E.R. vated in male 5-HT1B receptor knockout Kandel1,2. Genes important for long-term mice. memory and synaptic plasticity. Rationale: Recently, the finding that 5- Memory is not a unitary process, but HT1B receptor knockout (5-HT1B KO) mice consists of at least two systems — explicit have increased ethanol intake could not be memory (hippocampus-dependent), the mem- replicated. We assessed ethanol consumption ory of conscious recollection of facts and in male wildtype (WT) and 5-HT1B KO mice events, and implicit memory (hippocampus- derived from the original population (Crabbe et independent), the memory of perceptual and al. 1996). Objectives: To investigate whether motor strategies. The fact that these two major elevated ethanol consumption is present using
7 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society the original paradigm in our colony with ge- velopmental and adult phase of neurobiologi- netic makeup as similar as possible to the cal and behavioral deficits in YAC mice has original study. Methods: Mice had continuous been carried out. Abnormalities of basal fore- access to two pipettes, one filled with water brain cholinergic neurons have been investi- and one with increasing concentrations of gated using specific neuronal markers, and ethanol (0, 3, 6, 10 or 20% v/v). Fluid intake behavioral impairments have been monitored was determined daily. Results: Ethanol intake in a passive avoidance and a two-object rec- (g/kg body weight) did not differ between ognition task. These analyses, performed on genotypes. However, body weights (20-25%) YAC mice bearing different fragments of the and water intake (50%) were consistently ele- DCR, provided information to identify which vated in 5-HT1B KO mice. Conclusions: Phe- part, or even gene, of such region is responsi- notypic effects on ethanol intake could not be ble for the different features of cognitive im- replicated. These data confirm other studies, pairment displayed by DS subjects. suggesting ethanol intake in 5-HT1B KO mice 1Génétique Neurogénétique Comportement, is not increased. Hence, the initial finding of CNRS UPR 9074, Orléans, France elevated ethanol intake in 5-HT1B KO mice may have been due to phenotypic differences B.J. Caldarone1, C.H. Duman1, S.L. King1, in fluid intake. and M.R. Picciotto1. Fear conditioning, la- 1Dept. of Psychopharmacology, Faculty of tent inhibition, and habituation in mice Pharmacy, Utrecht University, Sorbonnelaan lacking high affinity nicotinic receptors2. 16, 3584 CA Utrecht, The Netherlands. Although nicotine has been reported to 2Center for Neurobiology and Behavior, Co- improve performance in several tests of cogni- lumbia University, New York, New York tion, the specific nicotinic receptor subtypes 10032, USA. 3PsychoGenics Inc., 4 Skyline that mediate these effects are largely un- Drive, Hawthorne, NY 10532, USA. 4Yale Uni- known. As a first step in understanding which versity School of Medicine, Dept. of Psychia- receptor subtypes regulate the cognitive ef- try, 34 Park Street, New Haven, CT 06508, fects of nicotine, baseline performance of USA. knockout mice lacking the beta-2 subunit of the nicotinic receptor and wildtype controls Igor Branchi1, Zoë Bichler1, Marie-Claude was evaluated in fear conditioning, latent inhi- Gonzalez1, and Danièle Migliore-Samour1. bition, and habituation tasks. In the fear condi- YAC polytransgenic mouse, an animal tioning task, mice were administered 3 pair- model to assess possible alterations of ings of a tone with foot shock and tested for cholinergic basal forebrain in Down syn- freezing to the context and tone 24 hr later. drome. Young (2-4 months) knockout and wildtype Down syndrome (DS) the most frequent ge- mice did not differ in fear conditioning, al- netic cause of mental retardation (1:700 live though aged (9-20 months) knockout males births), is due to an extra copy of chromosome exhibited less freezing to the context and tone 21. Structural and functional abnormalities in compared to aged wildtype males. No differ- the central nervous system of DS individuals ences in fear conditioning were observed be- have been widely reported. The first aim of DS tween aged knockout and wildtype females. research in recent years has been to identify Latent inhibition of fear to a pre-exposed tone the genetic bases of the different phenotypic was also assessed. Both knockout and wild- aspects. For this purpose, YAC transgenic type mice displayed similar levels of latent mouse models of DS have been developed by inhibition, although overall levels of freezing inserting in the murine genome a Yeast Artifi- were lower in knockout mice. Locomotor ha- cial Chromosome (YAC) bearing a fragment of bituation to a novel environment did not differ the human Down syndrome Chromosomal between wildtype and knockout mice in either Region (DCR). This region was reported to young or aged animals. These results support play an important role in the development of the previous study showing learning deficits in typical DS features, including mental retarda- aged beta-2 knockout mice (Zoli et al., 1999, tion. In YAC mice, trisomy is due to two copies EMBO J., 18, 1235) and suggest that other of mouse chromosome 16 and one copy of cognitive tasks may not be influenced by beta- homologous fragment of human chromosome 2 receptors. 21 included in the YAC. Several works re- 1Department of Psychiatry , Yale University ported a neurodegeneration of cholinergic School of Medicine, New Haven, CT, USA. neurons in DS brain, and, more recently, a 2This work was supported by NIH grants treatment based on acetylcholinesterase in- DA10455, DA00436, DA11733, and DA84733. hibitors has been shown to improve cognitive abilities in DS patients. An assessment at de-
8 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society
Iain C. Campbell1, Nigel Brown1, Ann P. Chapman1. Models of Alzheimer’s dis- Ward1, and Janet L. Treasure1. Energy ease: More than just messing around with Sensing in subjects recovered from Ano- genes?2 rexia Nervosa. The identification and manipulation of Leptin is a lipostatic hormone also in- Alzheimer’s disease (AD) related genes has volved in the control of metabolism, and its provided a major impetus for the creation of deficiency leads to massive obesity. We hy- animal models. Not surprisingly, mouse mod- pothesised that overactivity in this hormonal els have led the way. Mice overexpressing system contributes to the development of AN. several different mutant forms of amyloid pre- In rodents, glucose metabolism is an energy cursor protein (APP), presenilin 1 (PS1) and sensing system and the activity of this path- risk factors for AD (such as ApOE) have been way is linked to leptin production. Further- created, crossed and tested for histopathology more, in normal weight and obese subjects, ranging from amyloid plaque deposition to cell leptin secretion increases in the late evening and synapse loss. Because AD is character- in an manner related to daily food intake indi- ised in humans by loss of cognitive function, cating it is related to energy balance. Accord- the generation of tests for the behavioural and ingly, we tested the general hypothesis that physiological consequences of AD-related abnormalities in the glucose-insulin-leptin axis gene mutation is both critically important and are predisposing factors for AN. 18 females challenging. We have analysed both mice and recovered from an episode of AN (to avoid rats overexpressing mutated (670-671NL) starvation related changes) were given a human APP, examining learning, memory and standard meal, and glucose, insulin, b- attention while also measuring synaptic physi- hydroxybutyrate and leptin responses, com- ology. Aged transgenic mice and rats both pared with age and BMI matched controls demonstrate behavioural impairments, though (see Ward et al, 1998). Bloods were collected the severity is greater in the mice, either as a at 15 minute intervals from 11:00 and then at result if increased gene dosage, greater sus- 12:45, 13:45 and 14:45. The meal was pre- ceptibility, or both. On the other hand, the sented at 11:15 and had to be eaten by 12:15. range of behavioural tests that can be applied Subjects were allowed diet soft drink or black to rats suggest that future attempts to charac- coffee with the meal and free access to min- terise their deficits may produce a greater un- eral water. Recovered AN subjects showed a derstanding of the nature of behavioural proc- normal BMI-leptin correlation and, as for con- esses affected by AD-related gene mutations trols and obese subjects, showed no immedi- than would be possible by studying mice ate leptin response. We observed both in- alone. Aged transgenic mice also demonstrate creases and decreases in post meal leptin altered synaptic physiology, being more sensi- responses, which may be related to the en- tive to trauma, more excitable, and less likely ergy balance of the individual. There is an in- to demonstrate long-term synaptic enhance- verse correlation between changes in plasma ment. The opportunity to conduct long-term leptin and glucose in the post meal period in recordings from chronically implanted rats will the controls: we propose it is a manifestation be of great value in correlating the physiologi- of an energy sensing system in humans. cal and behavioural consequences of AD. Thus, as glucose shows small rebound in- 1Cardiff School of Biosciences, Cardiff Univer- creases following the post prandial hypogly- sity, Cardiff, UK. 2This work was supported by cemia, leptin decreases, signalling that energy grants from the Wellcome Trust and the U.S. requirements are being fulfilled. Perhaps more National Institute of Aging (NIA). importantly, this relationship between leptin and glucose is absent in the recovered AN S. Chiavegatto1, V.L. Dawson1, L.A. Ma- group. From this, an hypothesis on the aetiol- mounas1, S.H. Bora1, V.E. Koliatsos1, T.M. ogy of AN can be proposed: 1) AN subjects Dawson1, and R.J. Nelson1. Altered gene are unable to accurately monitor the availabil- expression of central serotonin 5-HT1A and 2 ity of body energy; 2) because they do not 5-HT1B receptors in male nNOS null mice. accurately sense levels of body energy, sub- Male mice with targeted disruption of jects prone to AN are able to escape from the the neuronal isoform of nitric oxide synthase normal homeostatic process which increases (nNOS) display a marked increase in aggres- food intake when the demand for energy is sive behavior with no difference in blood tes- increased: as a consequence they are able to tosterone concentrations. Several lines of re- deplete their energy stores and lose weight. search have established a role for brain sero- 1Institute of Psychiatry, King's College, Lon- tonin (5-HT) in aggression. Studies using se- don. SE5 8AF. UK. lective 5-HT receptor agonists and genetically engineered mice strongly suggest that 5-HT1A
9 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society and 5-HT1B receptors play important roles in consisted in one white and one black com- aggression. We investigated the neurochemi- partments communicating through a neutral cal profile of these mice. Serotonin turnover (grey) zone. Morphine injection was always (5HIAA/5-HT) was reduced in the cortex associated to the white compartment. Condi- (18.1%), hypothalamus (18%) and midbrain tioning trials lasted 30 min each and the ex- (16.4%) of nNOS-/- mice (p<0.05). We thus periment ended with a 15 min preference test examined the 5-HT terminals by immunocyto- performed in a drug-free state. Significant dif- chemistry. There was no significant alteration ferences in time spent in the different com- in the density and pattern of 5-HT terminals in partments of the apparatus between KO and the nNOS-/- mice suggesting that the selective wild-type littermates emerged during the ha- disturbance in the serotonergic system in the bituation session. Specifically, IL-6 KO mice brain of nNOS-/- mice is not due to a loss of 5- spent more time in the white compartment, HT axons or structural alterations. We used a while controls spent most of the habituation semi-quantitative RT-PCR methodology and session in the black area of the apparatus. On focused on the mRNA levels of four 5-HT re- the test session, significant differences be- ceptors 5-HT1A, 5-HT1B, 5-HT2A and 5-HT2C tween IL-6 null mice and their controls were implicated in aggression/impulsivity. The no longer evident since this last group spent mRNA levels of the 5-HT receptors were nor- significantly more time in the white compart- malized according to the mRNA levels of ment, compared to pre-conditioning levels. No GAPDH in each brain area. The expression of conditioned place preference was shown by postsynaptic 5-HT1A receptors was dramati- IL-6 KO mice, in line with previous reports cally decreased in hypothalamus (49 ± 4%) showing a reduced sensitivity to the analgesic and increased in both the hippocampus (41 ± effects of morphine in these null mice and re- 6%) and amygdala (94 ± 34%), whereas the duced density of mu receptors. Overall these mRNA level of 5-HT1B receptor was decreased data indicate that IL-6 appears involved in the in frontal cortex (35 ± 8%) and increased in expression of emotional behaviours and in the hippocampus (39 ± 5%) in nNOS-/- mice. responding to exogenous opiates. 1 The expression of the postsynaptic 5-HT2A Section of Behavioural Pathophysiology, Lab. and 5-HT2C receptors was similar between WT FOS, Istituto Superiore di Sanità, Rome, Italy and nNOS-/- mice in all brain areas studied. In and 2Dept. of Pharmacology, University of Mi- the midbrain, the mRNA of 5-HT receptors lan, Italy. and the mRNA of tryptophan hydroxylase (5- HT synthetic enzyme) and 5-HT transporter David A. Collier1, Andreas Karwautz1, Janet were not different between genotypes. These Treasure1. Gene-environment studies of data suggest that selective disturbance of the anorexia nervosa in discordant sister pairs. 5-HT1A and 5-HT1B receptors expression might Previously, a number of investigators be responsible for the aggressive phenotype have implicated the 5-HT2A gene as an aetio- of nNOS-/- mice. logical factor in anorexia nervosa. In order to 1The Johns Hopkins University, Baltimore, clarify the role of this gene, we examined both MD, USA. 2Supported by FAPESP (Brazil), genetic and environmental risk factors for AN NIH and National Alliance for Schizophrenia using a within family, case-control design of and Depression. 45 discordant sister pairs. The anorexia ner- vosa phenotype was associated with the per- F. Cirulli1, S. Capogrossi Colognesi1, M. sonality traits of harm-avoidance, persistence, Bianchi2, A. Panerai2, and E. Alleva1. Re- perfectionism, ineffectiveness as well as sev- duced sensitivity to morphine place condi- eral axis I & axis II disorders. Novelty seeking, tioning in IL-6 KO mice. Interleukin-6 (IL-6) interoceptive awareness and self-directedness is a cytokine involved both in inflammatory were all reduced. The sisters with AN differed responses and in brain function. from their healthy sisters in terms of personal In this experiment the positively- vulnerability traits and exposure to high paren- reinforcing properties of IP morphine were tal expectations and sexual abuse. Factors assessed using a place conditioning paradigm within the dieting risk domain were not signifi- in transgenic male mice not expressing IL-6 cant, but instead there was evidence of poor (IL-6 KO) and in control (IL-6 +/+) subjects. feeding in childhood. We found no difference Following an habituation session, IL-6 KO in the distribution of genotypes or alleles of the mice and their wild-type littermates were ex- DRD4, COMT, the 5HT2A, and 5HT2C recep- posed to a conditioning procedure, each tor genes. However, these results have to be mouse receiving four pairings of morphine (10 seen as preliminary because our power calcu- mg/kg) with specific environmental cues (floor lations indicate that there is insufficient power texture, wall colour). The testing apparatus
10 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society to detect the expected effect on risk with the Institute of Mental Health, Bethesda, MD sample size employed. 20892, USA. 1Eating Disorders Unit, The Institute of Psy- chiatry, London SE5 8AF, UK. H. Daniel1. Calcium signalling in cerebellar Purkinje cells of mice lacking mGluR1 or 1 Jacqueline N. Crawley . What’s wrong with InsP3R1 receptors: Role in long-term de- my mouse? Behavioral phenotyping pression. strategies and applications. Activation of subtype 1 metabotropic Targeted gene mutation presents a glutamate receptors (mGluR1) and subtype 1 powerful new tool for understanding the role of inositol 1,4,5-trisphosphate receptors genes in behavior. Rigorous experimental de- (InsP3R1) play a key role in the induction of sign is required for the behavioral phenotyping long-term depression (LTD) of synaptic trans- of transgenic and knockout mice. Use of well mission at parallel fiber (PF)-Purkinje cell syn- established, quantitative behavioral tasks, ap- apses in the cerebellar cortex, as confirmed propriate Ns, correct statistical methods, con- by impairment of this form of plasticity in sideration of background genes contributed by mGluR1 deficient transgenic mice (Conquet et the breeder parents, and consideration of litter al., 1994) and in InsP3R1 deficient transgenic and gender issues, will yield meaningful com- mice (Inoue et al., 1998). Activation of parisons of -/-, +/-, and +/+ genotypes. mGluR1 stimulates intracellular cascades, Our laboratory attempts to design meth- involving protein kinase C (PKC) activation ods to optimize the behavioral characterization and calcium release from InsP3-sensitive in- of mutant mice. Initial observations evaluate ternal stores through the production of InsP3. general health, neurological reflexes, sensory Indeed, both of these cascades, at least in abilities, and motor functions. Specific tests certain experimental conditions, are likely to include measures of home cage behaviors, be involved in LTD induction. body weight, body temperature, appearance The aim of the present study was to of the fur and whiskers, righting reflex, acous- investigate in in vitro thin cerebellar slice tic startle, eye blink, pupil constriction, vibris- preparations of these mutants and wild-type sae reflex, pinna reflex, Digiscan open field mice, possible interactions between various locomotion, rotarod motor coordination, hang- sources of calcium in dendrites of Purkinje ing wire, footprint pathway, hot plate tactile cells, that can be involved in the induction of response, visual cliff, and acoustic startle. LTD. In addition to the patch-clamp recordings Hypothesis testing then focuses on at of PF-mediated EPSPs in current clamp least three well-validated tasks within each mode, cytosolic free calcium concentration 2+ relevant behavioral domain. Specific tests will [Ca ]i in proximal of wild-type and mutant be described and illustrated for the domains of Purkinje cells was measured with fluores- learning and memory, feeding, nociception, cence images recorded with a-CCD camera, and behaviors relevant to discrete symptoms or with fluorometric method with a photomulti- of human anxiety, depression, schizophrenia, plier detector system. This was achieved by and drug addiction. Adaptation of standard rat dialysing the calcium sensitive dyes fluo-3 tests for use in mice, and development of new (100 µM) or bis fura-2 (100 µM) into Purkinje tasks for mice, are necessary in many behav- cells through the patch-clamp electrode. ioral domains. In mGluR1 deficient Purkinje cells, the An example of our approach will be de- calcium signal from mGluR-InsP3 pathway scribed for the behavioral phenotyping of was absent, whereas the calcium responses galanin overexpressing transgenic mice, gen- due to direct activation of voltage-dependent erated in the laboratory of our collaborator calcium channels and to release from ryano- Robert Steiner at the University of Washington dine-sensitive stores were at least qualitatively in Seattle, as a model for the striking galanin preserved. In addition, the lack of functional overexpression in Alzheimer’s disease. Gala- mGluR1 did not alter, at least qualitatively, the nin transgenic mice were normal on measures InsP3-dependent control of calcium release of general health, reflexes, sensory and motor from internal stores, since it was possible to 2+ abilities. As compared to wildtype littermates, detect [Ca ]i dendritic changes induced by galanin transgenics displayed impairments on photolytic release of 25 µM InsP3. In these cognitive tasks, supporting the growing litera- mutants, the combination of calcium influx ture that galanin is an inhibitory modulator in through voltage-dependent calcium channels cholinergic pathways relevant to learning and and calcium release from InsP3-sensitive in- memory. ternal stores following photolytic release of 1 Section on Behavioral Neuropharmacology, InsP3 (but not from ryanodine-sensitive inter- Experimental Therapeutics Branch, National nal stores) rescued a PKC-dependent LTD.
11 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society
This demonstrates that impairment of LTD in iors. To identify components of these path- mGluR1-deficient mice is not due to major ways we have conducted screens for muta- abnormalities in the signal transduction path- tions which abolish social behavior. Most of ways involved in LTD induction dowstream the mutants identified thus far appear behav- mGluR1, but is solely due to the lack of func- iorally normal, but forage in isolation. tional mGluR1. 1MRC Laboratory of Molecular Biology, Hills In Purkinje cells from mice with a dis- Road, Cambridge, CB2 2QH, England. rupted InsP3R1 gene, calcium responses due to direct activation of voltage-dependent cal- J.M. Delabar1, M. Rachidi1, C. Lopes1, C. cium channels were preserved. Application of Chabert2, P. Roubertoux2, C. Vayssettes1, a selective agonist of mGluR (1S,3R-ACPD) A.L. Delezoide3, and E.M. Rubin4. Abnormal 2+ induced no variation of [Ca ]i in dendrites and cerebellar folial pattern in Yac transgenic soma, indicating that in these cells, calcium mice containing a patterning gene from the release from internal stores does not occur Down syndrome chromosomal region-1. after InsP3R1 activation. Nevertheless, in Down syndrome is the major cause of some mutant cells, clear increases in dendritic mental retardation in humans. A considerable 2+ and somatic [Ca ]i were detectable during portion of the Down syndrome pathology has 1S,3R-ACPD application, but these rises ap- been mapped to the Down syndrome chromo- peared to reflect calcium influx via voltage- somal region-1 (DCR-1) in 21q22.2. A library gated calcium channels, because they were of Tg mice has been recently constructed with temporally associated with spike firing. Finally Yac clones from the DCR-1 (Smith et al 1997). in these cells, the release of calcium from ry- In this study the neuropathological analysis of anodine-sensitive stores was preserved since two lines carrying one copy of the Yac 230-E8 caffeine, a ryanodine receptor agonist, first (650kb) reveals a major modification of the caused reliable and reversible elevation of cerebellar folial pattern: lobules III to VI are 2+ [Ca ]i in dendrites and soma, and second in- clearly larger than normal; this hypertrophy creased calcium signals evoked by cell depo- increases the size of these lobules by 1.3- larization. 1.7x. Granule cell density and layering appear 1 Neurobiologie et Neuropharmacologie du to be normal in the cerebellum. Among the 5 Developpement, IDN-CNRS, France. known genes that contains the Yac 230-E8, one, C21orf5, has two orthologs involved in M. de Bono1. Genetics of natural variation patterning (in C.elegans and in drosophila). in C. elegans social behaviour. This gene is strongly expressed in frontal cor- Successful foraging is a pre-requisate tex, hippocampus and cerebellum of human ( for the evolutionary success of a species. Dif- normal and DS) and mouse embryos. A simi- ferent natural isolates of C. elegans exhibit lar expression pattern is observed in normal distinct foraging patterns in response to food. and Tg mice with human and mouse probes Animals from social strains aggregate and respectively. These observations suggest that feed together on a lawn of bacteria. Animals C21orf5 could play an important role in the from solitary strains show no aggregation and patterning of the cerebellum and in the patho- feed in isolation. About one third of wild iso- genesis of Down syndrome. lates are solitary and two thirds are social. 1CNRS UMR 8602 , UFR Necker Enfants- Variation at a single major genetic locus, npr- Malades, 156 rue de Vaugirard, 75730 Paris, 1, is responsible for this natural difference in France. 2CNRS UPR 9074, Genetique Neuro- foraging. npr-1 encodes a potential seven genetique Comportement, Orleans, France. transmembrane domain neuropeptide receptor 3Service de foeto-histopathologie Hop. R. De- with two natural variants that correlate with bre, Paris, France. 4Genome Sciences De- social and solitary behavior. The two variants partment, Lawrence Berkeley National Labo- differ at a single amino acid position that may ratory, Berkeley, CA 94720, USA. alter receptor activity. Social foraging occurs in response to C. Dubertret1, P. Gorwood1, L. Gouya1, J.C. signals derived from food. To understand how Deybach1, and J. Adès1. The haplotype rela- this behavior is generated we are defining the tive risk method and the genes coding for food signals that elicit social behavior, and the dopamine receptors in schizophrenia. molecular and cellular signaling pathways that The involvment of dopamine in the aeti- positively and negatively regulate social forag- ology of schizophrenia is suggested by num- ing. Studies of previously identified C. elegans ber of neurobiological and pharmacological mutants suggest that social foraging is regu- data. The role of the different dopamine recep- lated by pathways that are genetically distinct tors in schizophrenia is remaining unknown, from those controlling other C. elegans behav- as association studies showed conflicting re-
12 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society sults. We thus reexamined the dopamine hy- In Drosophila, spaced training results in pothesis controling for one of the major bias of several short-term forms of memory, as well association studies (i.e. statification bias) and as in long-term memory, which is CREB- and one of the major limitation of parametric link- protein synthesis-dependent. In contrast, age studies (i.e. estimating the unknown char- memory after massed training is less stable, acteristics of the type of inheritability), using CREB independent, and insensitive to protein the haplotype relative risk method. We synthesis inhibitors. We have used these be- searched for non-randomly transmitted alleles haviorally specific training protocols and muta- at the DdeI DRD1, Taq1D DRD2, Bal1 DRD3, tions that disrupt memory, in combination with FspI DRD4, DRD5 microsatellite and (TCAT)n Affymetrix gene chip technology, to character- repeat TH loci. These polymorphisms were ize a genomic response to memory formation. tested in 37 trios, containing the schizophrenic We have identified several memory candidate proband and both parents. Our results do not genes. support a major role of Taq1D DRD2 (c²=1.09, 1Cold Spring Harbor Laboratories, 1 Bungtown df=1, p=0.30), Bal1 DRD3 (c²=0.03, df=1, Road, Cold Spring Harbor, NY 11724, USA p=0.86), DRD4 (c²=0.24, p=0.62), DRD5 mi- crosatellite (c²=, df=11, p=0.89), and (TCAT)n R. Gerlai1,2, P. Pisacane1, and S. Erickson1. repeat TH (c²=0.47, df=4, p=0.98) polymor- Gene targeting and compensation: Behav- phisms. We found a significant excess of ioral effects of null mutations in the transmission of the allele 1 (c²=3.69, df=1, heregulin - ErbB system. p=0.05), with a closed-to-significant excess of Genetic redundancy is a problem in genotype 1*1 at the DdeI DRD1locus gene targeting studies because functionally (c²=5.29, df=1, p=0.07). The present analysis relevant sister proteins can compensate for suggest a small but significant effect of DRD1 the lack of protein product of a targeted gene. gene in the susceptibility to schizophrenia, but We chose a molecular system in which we the sample size needs to be increased before hope to demonstrate both the lack and pres- concluding for evaluation its real impact. ence of compensation after dysruption of par- 1Hopital Louis Mourier, Université de Paris VII, ticular single genes. Mammals may not be rue des Renouillers 178, 92701 Colombes, able to compensate for the lack of heregulin, a France. single ligand for multiple ErbB receptors, how- ever, compensation is expected when a single Josh Dubnau1, Scott Gossweiler1, Ulli ErbB receptor is knocked out. To investigate Certa1, Rod Scott1, Clemens Broger1, Mar- this, we disrupted the heregulin-1 , ErbB2, or tin Neeb1, Jerry Yin1, Jan Mous1, and Tim ErbB3 locus in a targeted manner and ana- Tully1. Functional genomics of long-term lyzed mice heterozygous for the mutation. memory. Heregulin and its receptors were shown to be Three features of long-term memory are involved in embryonic brain development and, conserved across animal phyla. First, memo- more recently, in plastic changes associated ries are initially stored in a short-term labile with adult brain function in rodents. Although form but can progress to a longer lasting, sta- they have never been shown to play roles in ble form. Second, long-term but not short-term mammalian behavior, we decided to charac- memory requires a new program of gene ex- terize the mutant mice behaviorally using a pression. And third, many tasks require re- battery of simple tests. Despite the absence of peated training sessions interspersed with rest gross morphological defects, heregulin mutant intervals (spaced training), rather than re- mice exhibited elevated activity levels in the peated training without rest intervals (massed open field, showed improved rotorod perform- training), to produce long-term memory. Nu- ance, and finished T-maze spontaneous alter- merous attempts at identification of genes and nation task faster compared to control wild pathways that are induced during memory type littermates, findings that suggest a con- consolidation have so far focused either on in- sistent hyperactivity across tests. ErbB2 and vitro models of neuronal plasticity or in-vivo ErbB3 mutant mice, whose strain origin was pharmacological manipulations of neuronal identical to that of heregulin mutants, showed activity. These studies have led to a laundry no sign of the behavioral alterations. We sug- list of genes that likely play important roles in gest that the abnormalities seen in heregulin neuronal plasticity in a broad sense. In con- mutant mice were due to mutation at that lo- trast, only a handful of transcripts and proteins cus and the lack of alterations seen in ErbB2 have been discovered that are induced or re- and ErbB3 mutant mice is the result of com- pressed during memory consolidation per se. pensation by unaltered sister receptors. Hence identification of genes involved specifi- 1Genentech Inc, 1 DNA Way, South San Fran- cally with long-term memory has been difficult. cisco. CA 94080, USA. 2Present address: Eli
13 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society
Lilly, Neuroscience Department, Drop Code whether the A9 allele is a risk factor or the 0510, Lilly Corporate Center, Indianapolis, IN A10 allele is a protective factor against com- 46286, USA. plicated withdrawal. Conclusions. The role of the A9 allele of the DAT gene in withdrawal P. Gorwood1, F. Limosin1, P. Batel1, C. complications is replicated in this independent Boni1, M. Hamon1, and J. Adès1. The DAT1 sample, probably involving severity of symp- gene is involved in severe alcohol with- toms rather than a specific type of symptom drawal in male alcoholics. (seizures and delirium are generally consid- Introduction. Two German case-control ered as the most severe symptoms of with- studies (Sander et al., 1997 ; Schmidt et al., drawal). 1998) showed that the A9 allele may increase 1Hopital Louis Mourier, Université de Paris VII, the risk for severe withdrawal symptoms for rue des Renouillers 178, 92701 Colombes, alcohol-dependent patients who stop drinking France. (specially withdrawal seizures and delirium). These studies were consistent as one was M. Hafezparast1, S.J. Nicholson1, A.S. retrospective and qualitative, and the other Witherden1, N. Bermingham1, S. Ball2, J. prospective and quantitative. We analyzed Peters2, D.C. Rogers3, J.E. Martin4, E.M.C. lifetime withdrawal symptoms in 120 French Fisher1. Loa (legs at odd angles) a mouse alcohol-dependent patients, in order to specify model of motor neurone dysfunction: map- which symptoms and/or patients may be more ping and progress towards isolation of the specifically involved in complicated alcohol- causal gene. withdrawal symptoms. Method. Patients were Motor neurone diseases (MNDs) are assessed with the DIGS (Diagnostic Interview neurodegenerative disorders that kill 1 in 500 for Genetic Studies) for lifetime psychiatric adults in UK. Some of these diseases have and addictive disorders. Expected allele fre- genetic aetiology. For example, approximately quencies were based on 65 control subjects 20% of cases of amyotrophic lateral sclerosis without psychiatric or addictive morbidity, and (ALS) are familial and so far only one gene, matched for male gender and French origins. superoxide dismutase 1 (SOD1), has been The A9 allele of the DAT1 was revealed with identified in 10-15% of familial ALS. Thus the the previously described PCR. Results. In our majority of the genes involved in the pathology sample, the A9 allele was more frequent in of MND remain to be identified. patients who had at least once withdrawal sei- We have a mutant mouse, called legs at zure or delirium (p=3D0.028), showing the odd angles (Loa), that exhibits an autosomal same trend as did the first study. There was a dominant motor function loss in the hind limbs. significant linear trend for increased number of Homozygotes for the Loa die within 24 hours withdrawal symptoms in patients with the A9 of birth. However, the gross anatomy of the allele (p=3D0.03). Furthermore, patients who progeny is apparently normal, indicating nor- took alcohol at least once in order to reduce mal development. Heterozygous Loa/+ mice withdrawal symptoms were more frequent in on the other hand are viable, have a normal the group of patients who have the A9 allele life span, and can be identified by a character- (p=3D0.027). Although the distribution of istic clasping of the hind limbs when sus- symptoms is not superimposable for the two pended by the tail. Compared with their wild groups (with versus without the A9 allele), no type littermates, these mice perform signifi- specific symptom (or clinical group of symp- cantly more poorly in the rotarod test of bal- toms) was found increased in one group. We ance and coordination, but show increased also performed a factorial analysis in order to spontaneous locomotor activity. These fea- detect a role of the DAT gene on latent tures of Loa mice are more profound in older classes. The factorial analysis of the 120 pa- animals, indicating that the Loa mutation re- tients showed the existence of 3 factors, ex- sults in a progressive motor function deficit. plaining 55% of the total variance. It is note- Histopathological analysis has revealed a sig- worthy that the DAT1 gene was associated nificant decrease in the number of anterior only in one specific dimension, untitled "se- horn cells in Loa/+ mutants compared with vere neuropsychiatric withdrawal symptoms". wild type littermates, but there is no evidence Interestingly, the frequency of the A9 allele of muscle denervation in Loa/+ mice, indicat- (54.0%) in the group of 65 healthy controls ing that the pathology of the Loa mutation is a (without alcohol abuse or dependence) neuropathy. matched for sex and origins was intermediate A large intraspecific backcross between between alcohol-dependent with withdrawal Loa/+ and the C57BL/6 inbred mouse strain complications (71.4%) and those without such was set up. Only affected N1 animals were symptoms (42.3%), addressing the question backcrossed to C57BL/6 to generate more
14 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society than 1,000 N2 affected mice. The N2 progeny latencies to withdrawal to a thermal stimulus were used to map the Loa mutation to an in- post-carrageenan (F = 5.54; df = 2,52; terval of approximately 1.6 cM in the distal P<0.01). Follow up analysis revealed that region of Mmu12. This interval is flanked by whilst both +/- and +/+ mice showed D12Mit17 and D12Mit181. We have con- significantly decreased latencies post- structed YAC/PAC/BAC contigs of the regions carrageenan compared to baseline (P<0.05), flanking the critical region and are generating -/- mice showed no such difference indicating more STS markers for chromosome walking no thermal hyperalgesia in these mice. These and bridging the gap that exists within our results confirm and extend the findings of contig of the Loa region. The STS markers are Davis et al. (FENS Meeting; Brighton 2000) also analysed for simple sequence repeat and and show that VR1 receptors have a major single nucleotide polymorphisms to narrow the role in the mediation of thermal hyperalgesia critical region. The mouse and human com- and that their role in thermal nociception is parative map, EST and other bioinformatics dependent on temperature. resources are routinely surveyed for identifica- 1Neurosciences Research, SmithKline tion of candidate genes near or within the criti- Beecham Pharmaceuticals, Third Avenue, cal region. Any candidate genes that map to Harlow, Essex CM19 5AW, UK. the Loa critical region will be further analysed and if not excluded, they will be sequenced in J.P. Hatcher1, D.C. Rogers1, C. Reavill1, J.J. affected and wild type mice for the identifica- Hagan1. Use of SHIRPA to investigate the tion of the Loa gene. behavioural phenotype of the Coloboma 1Department of Neurogenetics, Imperial Col- (cm/-) mouse. lege School of Medicine, Norfolk Place, Lon- Coloboma mice have a 1.1 – 2.2 cM don W2 1PG, UK. 2MRC Mammalian Genetics deletion on chromosome 2 which encom- Unit, Didcot, Oxon, OX11 ORD, UK. passes the Snap gene encoding the pre- 3Neuroscience Research, SmithKline synaptic nerve terminal protein, SNAP-25. The Beecham Pharmaceuticals, New Frontiers mutation is semi-dominant with the heterozy- Science Park, Harlow, UK. 4Department of gotes exhibiting a behavioural phenotype of Histopathology, The Royal London Hospital, spontaneous hyperactivity, constant head Whitechapel, London E1 1BB, UK. bobbing and a prominent eye dysmorphology (Hess and Wilson, 1992, J. Neurosci. 12: J.P. Hatcher1, P.T. Davey1, S. Bingham1, P. 2865-2874; Heyser et al., 1995, Dev. Brain Overend1, A.A Parsons1, and J.B. Davis1. Res. 89: 264-269). It has been suggested that Vanilloid receptor-1 has a major role in such a mutation may be a model of Attention thermal hyperalgesia. Deficit Hyperactivity Disorder (Hess et al., The vanilloid receptor-1 (VR1) is a non- 1996, J. Neurosci. 16: 3104-3111). We have selective ligand-gated cation channel used the SHIRPA protocol (Rogers et al., expressed in sensory neurones, which is 1997, Mamm. Genome 8: 711-713) to investi- suggested to have a major role in thermal gate the behavioural phenotype of these mice. nociception. Here we report studies in which Coloboma mice (cm/-) were found to have a we have investigated the responses of number of differences in the primary observa- homozygous (-/-, n=10) and heterozygous (+/- tion screen when compared to wildtype con- , n=10) VR1 knockout mice , and their wildtype trols including reductions in body weight and littermate controls (+/+, n=10) in the hotplate size, smaller eye openings and increased test and in the carrageenan model of thermal spontaneous activity. In a rotarod task, cm/- hyperalgesia. In the hotplate test there was a mice showed significantly shorter latencies significant effect of genotype on withdrawal than their wildtype littermates indicating that latency (F=7.91; df = 2,54; P<0.001). Follow they may have motor co-ordination deficits. In up analysis showed no difference between -/- a test of locomotor activity, cm/- mice showed mice and +/+ controls at 50oC (P=0.24). At a a significant increase in activity over the 60 hotplate temperature of 52.5oC there was a minute test period. This increase in activity suggestion of an effect although this failed to was further enhanced upon a second expo- reach significance (P=0.09). However at 55oC sure to the same apparatus. Our results indi- a significant difference was found (P<0.01). cate that cm/- mice show phenotypic differ- There was no significant difference between ences from wildtype controls and may serve +/- mice and +/+ mice at any hotplate as a useful model for ADHD. temperature. The same mice were then tested 1Neurosciences Research, SmithKline in the carrageenan model of thermal Beecham Pharmaceuticals, Third Avenue, hyperalgesia, where analysis of variance Harlow, Essex CM19 5AW, UK. showed a sigificant effect of genotype on
15 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society
A. Holmes1, J.G. Hohmann2, E. Yared1, R.A. 3Supported by NSF grant IBN-9720143 and Steiner2, and J.N. Crawley1. Variability in an the NIHM IRP. anxiety-related phenotype in galanin over- expressing transgenic mice3. T. Hough1, P.M. Nolan2, J. Peters2, E.M.C. Galanin (GAL) is a 29 amino acid neu- Fisher3, J. Martin4, M. Browne5, S. Rastan5, ropeptide localized in brain regions relevant to L. Vizor2, S.D.M. Brown2, and A.J. Hunter1. the processing of emotional information, in- Clinical biochemistry screens can com- cluding hippocampus, amygdala, bed nucleus plement behavioural screens in mutagen- of the stria terminalis, hypothalamus, raphé ised mice. nuclei, and locus coeruleus. Consistent with In the past phenotypic analysis of this expression pattern, intracranial admini- mutagenised, transgenic and knockout mice stration of exogenous galanin produces in- often focused solely on behavioural tests. In creases and decreases in anxiety-related be- contrast the F1 offspring of mutagenised male haviors in rats, depending upon the site of mice from the SB/Harwell ENU programme administration and the behavioral test em- are subjected to a routine blood biochemistry ployed (Bing et al., 1993; Moller et al., 1999). screen as part of a systematic phenotype- To further explore the role of GAL in the me- driven search for novel mouse mutations. The diation of anxiety-related behaviors, we gen- resulting data are used not only to identify erated transgenic mice overexpressing the possible biochemical mutants, but also to pro- GAL gene linked to a dopamine beta- vide a complement to the behavioural hydroxylase promoter. GAL-tg mice are viable, screens. Consequently abnormal biochemi- and show normal neurological reflexes, motor, cal/physiological parameters can be ruled out and sensory abilities. GAL-tg mice exhibit in- as a precursor to behavioural anomalies. Fol- creased GAL mRNA in the locus coeruleus, lowing completion of all behavioural tests, at 8 elevated GAL peptide in the forebrain, and to 12 weeks, around 300 ml of blood is col- increased GAL fiber density in the hippocam- lected from each F1 mouse. Samples are col- pus. Homozygous and heterozygous GAL-tg lected in Li-Hep capillary tubes from the tail mice, and wild type littermate controls, were vein and centrifuged at 3000 rpm for 10 min- obtained for behavioral testing from the Uni- utes at 4°C. Approximately 125 ml of plasma versity of Washington in two separate is obtained. Kidney, liver, bone and lipid pro- batches. Heterozygotes and wild type litter- files, as well as glucose and bicarbonate tests mates were bred at the Jackson laboratory are performed on an Olympus AU 400 ana- (Bar Harbor, Maine). In experiment #1, female lyser. Male and female data are analysed University of Washington GAL-tg mice dis- separately. Two criteria are used to identify played a heightened anxiety-like behavior in potential outliers: mice with values > 3SD's the light/dark transition test, but not in the ele- from the running mean for any one parameter vated plus-maze test. In experiment #2, a sec- or >2SD's for groups of related parameters. ond batch of female University of Washington Such potential outliers are re-tested after one GAL-tg mice displayed heightened anxiety-like month. In addition, offspring of mice identified behavior in both the light/dark transition test from other components of the phenotype and the elevated plus-maze test. In experi- screen are also subjected to biochemical ment #3, female heterozygous GAL-tg mice analysis. To date 1600 F1's have been bred at the Jackson laboratory showed re- screened and around 500 mice comprising 25 duced anxiety-like behavior in the elevated mutant lines (10 mutants, 10 controls per line) plus-maze, heightened anxiety-like behavior in tested. Twelve F1 animals showing consistent the home base emergence test, and no sig- abnormalities in plasma biochemistry were nificant phenotype in the light/dark transition tested for inheritance of the mutation. Of these test nor in novel object exploration. The vari- 12, 3 have currently been confirmed as inher- ability in anxiety-related phenotype across ited. Our results have shown that incorpora- separate batches of GAL-tg mice is discussed tion of such a biochemical screen is useful in in relation to the effects of gene dosage, and characterising abnormal pheno- also in terms of experimental variables, choice types/mutations. For example certain classes of test, and other factors common to studies of behavioural mutants identified in this screen with mutant mice. exhibited associated changes in biochemical 1Section on Behavioral Neuropharmacology, parameters. ETB, National Institute of Mental Health, Be- 1Dept of Neuroscience, SmithKline Beecham thesda, MD, USA. 2Dept. Obstetrics & Gyne- Pharmaceuticals, New Frontiers Science Park, cology and Physiology & Biophysics, Univer- Harlow, UK. 2MRC Mammalian Genetics Unit sity of Washington, Seattle, WA, USA. and Mouse Genome Centre, Harwell, OX11 ORD, UK. 3Department of Neurogenetics, Im-
16 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society perial College, London W2 1PG, UK. digms. Experienced Tg mice performed in a 4Department of Morbid Anatomy, Queen Mary similar manner to controls, but when forced to and Westfield College, London E1 1BB, UK. acquire spatial information in a series of train- 5Dept of Biotechnology and Genetics, Smith- ing sessions, the mutated Tg mice were con- Kline Beecham Pharmaceuticals, New Fron- sistently worse than Tg wild-type mice. In the tiers Science Park, Harlow, UK. case of Tg(APP)CRND8 mice, spatial learning and memory was significantly impaired with Christopher Janus1. Behavioural mouse respect to littermate controls, coinciding with models of Alzheimer's Disease: A hit-or- the early onset of AD-related pathology in miss approach? these mice. The Tg(APP)CRND8 mice revealed The development of an animal model of yet a greater divergence from controls in the memory impairment in Alzheimer's Disease subsequent learning reversal tests, suggesting (AD) is pivotal for understanding mechanisms compromised behavioural flexibility. of the disease and for exploration of new I shall argue that more detailed analy- treatments. One of the main features of the ses of mice behaviour in the water maze may disease is progressive cognitive and memory be useful in characterisation of their cognitive decline which coincides with amyloid plaques, impairments, and may convincingly demon- neurofibrillary tangles, and neuronal loss. The strate sensorimotor disturbances. Analysis of heterogeneity of AD dementia presents a for- behavioural search strategies (thigmotaxis, midable challenge for the development of rep- search patterns) together with analysis in a resentative animal paradigms related to the cued (visible platform) version of the test may cognitive impairment observed in AD. How- help us better understand group and individual ever, the potential advantages are enormous variance. Exhaustive analysis of water maze since the model will allow testing of biochemi- mouse behaviour may first, lead to improved cal phenotypes, dissection of essential patho- appreciation of compromised cognitive func- logical defects, and dissection of modifier tions caused by AD transgenes, and second, genes and environmental factors. Difficulties may provide solid behavioural data for further in the development of the model are mainly experimentation which should ultimately yield imposed by species constraints of learning, a better understanding of AD-related impair- hence careful choice of species and measured ment in humans. learning behaviour is crucial. Further, learning 1Centre for Research in Neurodegenerative in animals most often is inferred from the Diseases, University of Toronto, Canada. analysis of their behavioural motor acts so impairments in the peripheral/sensorimotor Christopher Janus1, David Westaway1, systems may bias the analysis of cognitive Jacqueline Pearson1, Peter St. George- systems. Hyslop1. Impaired spatial leaning and With regard to commonly used labora- memory in APP CRND8 transgenic mice. tory animals, the development of transgenic Dominant mutations in the ß-amyloid (Tg) mice harbouring human genes implicated precursor protein (APP) are associated with in AD established the feasibility of reproducing familial Alzheimer's Disease (FAD). To ad- at least facets of the AD phenotype in a trac- dress the relationship between FAD mutations table ssystem. Also, the plethora of behav- and cognitive deficits, we created a novel line ioural learning tests is most widely repre- of Tg mice (TgCRND8) in a C3H/C57 genetic sented by a water maze test developed mainly background, expressing a compound mutant for studying learning and memory in the rat, form of the human but is used as a routine basis for the analysis APP695KM670/671NL+V717F. These mice of mutant mice. exhibit very early-onset deposition of Aß- In my talk I focus on behavioural studies containing amyloid plaques, from 3 months using Tg mouse models expressing familial onwards, with congophilic, neuritic plaques AD mutations (PS1and APP genes) in the wa- present from 5 months of age. The cognitive ter maze test. Tg mice expressing mutant hu- characteristics of the transgenic mice were man PS1 (alleles M146L and L286V) did not compared to their non- Tg littermates. Since it have any detectable neuropathologic changes is accepted that the hippocampal region is or show any noticeable sensorimotor or cogni- affected in the early stages of AD, we tested tive impairment in a conventional (place dis- TgCRND8 mice in the hidden platform (place crimination) version of the water maze test. discrimination task) version of the Morris water However, naïve mutant PS1 mice showed ini- maze (WM) at age coincident with the onset of tial impairment in the rate of acquisition of AD-related pathology. We report that spatial information, both in simple cue learning TgCRND8 mice showed a significant impair- and conventional place discrimination para- ment in the acquisition of the spatial informa-
17 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society tion and, unlike non- transgenic littermates, ciceptin/pain) is a relic of the early evolution of did not develop a bias for the spatial position the hippocampus. If the hippocampus com- of the submerged platform. In the subsequent putes novelty by comparing sensory and learning reversal test, the Tg mice were also memory inputs, as proposed by Vinogradova significantly impaired in re-learning of the new in 1975, then enteroceptive modulation pro- spatial position and showed no spatial bias for vides that the hippocampus computes sali- the new position in the probe trial. During both ence. learning tests Tg mice showed increased What then is the output of the hippo- thigmotaxic swim behaviour, but in neither of campus that reflects salience? Ablation and the above tests did they differ significantly stimulation studies have demonstrated that from non-Tg mice in their swim speed. We the hippocampus acts as an endocrine trans- conclude that expression of mutated human ducer, and governs adaptive changes in body APP in the TgCRND8 mice confers AD-related physiology and hypothalamic/pituitary/adrenal pathology at earlier times than reported for (HPA) axis activity. For instance, glucocorti- previous animal models of AD, and that pro- coid levels rise in response to a novel stimu- found impairments in spatial leaning and lus, or to a previously learned aversive taste, memory coincide with the onset of neuropa- the hormonal rise is abolished by hippocam- thology. Our data suggest that TgCRND8 pectomy. Because adrenal hormones (gluco- mice will be of great utility in assessing amy- corticoids and norepinephrine) enhance con- loid-directed therapies and the pathways link- solidation of memory traces, endocrine activity ing Aß synthesis and cognitive impairment. directed by the hippocampus may explain its 1Centre for Research in Neurodegenerative role in memory. In support, the requirement for Diseases, University of Toronto, Tanz Neuro- the hippocampus in memory acquisition can science Building, 6 Queen’s Park Cr. W. To- be circumvented, at least in part, by co- ronto, Ontario, M5S 3H2 Canada. administration of adrenal hormones. However, the hippocampus developed early in verte- Richard Lathe1. The enteroceptive hippo- brate evolution; more recent refinements may campus. replace HPA axis modulation by neuronal re- Hippocampal lesions produce deficits in lays that locally release norepinephrine in ac- specific types of learning, but its role in mem- tive regions of cortex. ory is unknown and the emphasis on memory 1Centre for Genome Research and Centre for processes may be misleading. To cast light on Neuroscience, University of Edinburgh, UK. hippocampal function we examined the spec- trum of genes expressed in the formation. T. Lemberger1, T. Mantamadiotis1, O. Gene-trapping suggested that perhaps one Kretz1, D. Gau1, T. Steckler2, and G. Schu- third of the mouse genome is expressed in the etz1. Conditional mutagenesis of CREB in hippocampus; the cases examined in detail all dopaminoceptive neurons. encoding membrane-associated signalling Dopaminergic neurotransmission in molecules (Steel et al., 1998, Hippocampus 8: basal ganglia and related pathways plays a 444-457). On further analysis the hippocam- major role in the regulation of movement, re- pus was found to contain one of the highest ward oriented behaviours, mnemonic func- densities of receptors for soluble ligands in the tions like habit formation and in drug addiction. brain: these are accessible, functional, and The CREB trancription factor is thought to be mediate physiological and cognitive changes one important nuclear targets of the dopamine in vivo. The hippocampus is thus a primary D1 receptor/cAMP/PKA signal transduction target for ligands that reflect body physiology cascade. To analyze the contribution of CREB including ion balance and blood pressure, im- in this signaling pathway in vivo, we have munity, pain, reproductive status, satiety and conditionally disrupted the CREB gene in neu- stress. It is thus hypothesized that the early rons expressing dopamine D1 receptors using hippocampus diverged from the olfactory sys- the Cre/loxP system. For this purpose, we tem (mediating exteroception), to sense solu- have generated transgenic lines using a 140 ble molecules in blood and cerebro-spinal fluid kb yeast artificial chromosome containing the (enteroception). dopamine D1 receptor gene to drive the ex- It is suggested that the hippocampus pression of the Cre recombinase. The D1-Cre may temporarily store this information via syn- mice have been crossed with CREBlox mice aptic long-term potentiation (LTP): it is tempt- in which exon 10 of the CREB gene is flanked ing to speculate that enhancement of LTP by by loxP sites. In the resulting ”D1-CREB” (D1- 'good' ligands (e.g.s estrogen/fertility; al- Cre / CREBlox/lox) mutant animals the pattern phaFGF/satiety) and impairment by 'bad' of recombination, detected as a disappear- ligands (e.g., interleukins/infection; no- ance of the CREB protein, localizes to the re-
18 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society gions expressing Cre: striatum, nucleus ac- ronal degeneration. This notion is consistent cumbens, cortex layer VI, hippocampus CA2. with findings from the recent studies suggest- Disruption of the CREB gene in the brain ing that the use of non-steroidal anti- leads systematically to the upregulation of the inflammatory agents, which inhibit the cyclo- CREM gene, a member of the ATF / CREB oxygenase activity (cox-2), is protective familly of transcription factors. To exclude that against the development of AD (McGeer et al., CREM can compensate for the loss of CREB, 1996). In the present study we assessed the we have also generated double mutants D1- behavioral phenotype of an initial line of thy- Cre / CREBlox/lox / CREMnull/null that should 1/hcox-2 transgenic mice, that overexpresses be completely devoid of any CREB-like activity human cox-2. Our results indicate that the in the D1 receptor expressing neurons. These young thy-1/hcox-2 mice appear not to be dif- mice are now being tested for striatum- ferent from age-matched non-transgenic lit- dependent memory, response to psychostimu- termates in spatial memory and in general lants and immediate-early gene induction. phenotype such as activity, emotional- 1German Cancer Research Center, Heidel- ity/anxiety, body balance and coordination, berg, Germany. 2Max Planck Institute for Psy- swimming ability, visual acuity, agility and sen- chiatry, Munich, Germany. sorimotor reflexes. Once aged, these thy- 1/hcox-2 mice showed accelerated impairment S. Lewis1, M.A. Simoneau1, S. Bailly1, F. in spatial memory tasks dependent upon hip- Guillou1, B. Baron1, and C. Belzung1. Be- pocampal function. This cognitive deficit coin- haviour of mice expressing human trans- cided with the overexpression of the Cox-2 ferin in brain. transgenic protein in the CA1 and CA3 regions Transgenic mice expressing the human of the hippocampus. These findings indicate transferin gene in oligodendrocytes exhibit an that the overexpression of cox-2, which plays important synthesise to myelin specific mark- a role in the inflammatory processes may ers in the central nervous system in a B6D2 cause the deficit in cognitive function by itself genetic background. We have compared be- or in interaction with aging. havioural characteristics of 3-month-old het- 1Dept. of Psychology, 2Neurology, and erozygous female transgenic mice with their 3Neuroscience, Johns Hopkins University, age-matched wildtype in five tests: Hugues School of Medicine, Baltimore, MD, USA. box test for the measure of trait anxiety, 4Supported by Monsanto/Searle Co., St. light/dark box and elevated plus-maze for the Louis, MO, USA. measure of state anxiety, Porsolt test for de- pression symptoms and spatial open-field test S.C. Maxson1, A. Canastar1, and C. for memory. No apparent motor deficit could Bishop2. Sex reversals (XX and XY females be observed in all these tests. No difference and XX and XY males), mating behaviors was found between transgenics and wildtypes and aggressive behaviors in mice. in anxiety tests. In Porsolt test, transgenic There are effects of one or more Y mice spent significantly less time escaping. In chromosomal genes on copulatory and attack the spatial open-field, wildtype mice react to a behaviors of mice. These have been demon- change in the spatial configuration of objects strated for several pairs of Y chromosomal by in increase of exploration, an effect that is variants. It has been recently proposed that not seen in transgenic mice, suggesting a sex reversed mice might also be used to fur- deficit in spatial memory in these mice. These ther investigate the effects of the Y chromo- results show that mice expressing human some on these behaviors. We tested XX and transferin display specific changes of their be- XY females on the C57BL6 and XX and XY havioural pattern, which could be linked to a males on the FVB background for copulatory specific pattern of expression of myelin in the and attack behaviors. Attack behavior was brain. tested in a neutral cage with an opponent of 1JE Psychobiologie des Emotions, UFR Sci- the same genotype. Copulatory behavior was ences, Parc Grandmont, F-37200 Tours, tested in the home cage with a female in hor- France. monally induced estrus. The XX and XY fe- males on the C57BL6 background did not dif- Alicja L. Markowska1, Alena Savonenko1, fer in any of these behaviors, and it appears and Katrin Andreasson2,3. Overexpression that the XX and XY males on the FVB back- of cyclooxygenase-2 (cox-2) leads to cog- ground also do not differ on any of these be- nitive impairment.4 haviors. Several explanations will be dis- Cognitive impairment in Alzheimer's cussed to possible account for the failure to disease (AD) is associated with inflammatory find any differences. processes, which play a critical role in neu-
19 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society
1Biobehavioral Sciences Graduate Degree Meanwhile the molecular underpinnings Program, The Department of Psychology, The of the genetic liability to depression remain University of Connecticut, Storrs, CT 06269- obscure. Some promising leads have 4154 USA. 2Departments of OB/GYN and emerged from studies of candidate genes, Human and Molecular Genetics, Baylor Col- particularly those involved in serotinergic lege of Medicine, 6550 Fannin street, Houston transmission. While systematic whole genome TX 77030 USA. 3This work was supported by scans have been undertaken in bipolar manic grants from The University of Connecticut Re- depression with controversial and conflicting search Foundation and by The University of results, linkage studies in unipolar depression Connecticut Inbred Mouse Fund (1) and by taking a systematic approach have only just NIH(2). begun. Some of the approaches being em- barked upon using linkage and linkage dis- Peter McGuffin1. Heredity, hazards and the equilibrium mapping will be described. origins of depressive disorder. 1Social, Genetic and Developmental Psychia- Depression is a common disorder try, Research Centre, Institute of Psychiatry, which, in addition to imposing a public health De Crespigny Park, Denmark Hill, London, burden, imposes an economic burden in west- SE5 8AF United Kingdom. ern industrialised countries that is rivalled only by cardiovascular disease. There is consistent K.L. McIlwain1 and R.E. Paylor1. Effects of evidence that depression is familial with the testing experience in a mouse behavioral risk in the first degree relatives of sufferers test battery.2 ranging from three to nine times that in the In recent years there has been an general population. Recent population based increased interest in the behavioral and hospital register based twin studies show phenotyping of genetically modified and in that familial aggregation is largely explained inbred strains of mice. Our laboratory uses a by genetic factors. That is, although the envi- specific battery of tests for the initial ronment accounts for between 30% and 60% assessment of phenotypic behavioral of the variation in liability, environmental fac- differences of transgenic and knockout mice, tors appeared to be entirely of the non-shared as well as inbred strains of mice. Our standard type. battery includes: open field activity, light-dark Older terminology suggested that de- exploration, rotorod, prepulse inhibition (PPI), pressive disorder could be sub-classified into acoustic startle habituation, conditioned fear, “reactive” forms largely resulting from envi- Morris water maze, and hot plate. Mice are ronmental insults and “endogenous forms” tested in the order listed, and this order for the that were “constitutional”, but there is little ge- test battery was devised from least invasive to netic evidence to support this and a recent most invasive, to decrease the chance that twin-family analysis suggests that it is unlikely behavioral responses are altered based on that there are two broad forms of depression, prior test history. The studies presented here one mainly genetic and the other non-genetic. were designed around two basic questions. It is therefore probably that most cases of de- The first study addresses whether or not there pression result from a combination of genetic are differences between mice that have liability and environmental adversity. However, undergone other previous testing and mice the interplay between genes and environment that are naïve to the test experience. The appears to be complicated. Thus, there is evi- second study asks what is the effect of the dence dating back to the mid-1980s that the testing order with respect to how an animal relatives of depressed subjects not only show performs on subsequent tests? In the first increased rates of depression, but also in- experiment, one set of C57BL/6J male mice creased rates of experiencing (or reporting) were evaluated on all of the tests described threatening life events. The familial clustering above. The behavior of these 'test battery' of life events has been supported by twin stud- mice was compared to aged matched naïve ies which also have produced the surprising mice that were only tested on one test from finding that life events are in some cases in- the battery. We found that on some tests the fluenced by genes. This has led to the hy- behavior of 'test battery' mice was significantly potheses that familial factors may influence different from the behavior of naïve mice, the liability to depression indirectly by predis- while on other tests there were no differences posing individuals to select a more “aversive” between test battery and naïve mice. For environment or, alternatively, there are some example, test battery mice responded inherited cognitive schema that predisposes differently in the open-field, rotarod, and both to depressive symptomatology and re- hotplate test, but behaved similar on the porting unpleasant happenings. prepulse inhibition and fear conditioning test.
20 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society
Experiments in the second study were carried 1UPR CNRS 9074, Génétique, Neurogéné- out on male 129/SvEvTac and C57BL/6J male tique, Comportement, Institut de Transgénose, mice. In the second study we used an 3 B rue de la Ferollerie; 45071, Orléans abbreviated battery of tasks, which included - Cedex 02, France. 2Supported by CNRS (UPR open field activity, light-dark exploration, PPI, 9074), Ministry for Research and Technology, and fear conditioning. Given that it is not Région Centre, Préfecture de la Région Cen- feasible to study all possible combinations, we tre, and Fondation pour la Recherche Médi- chose 4 representative orders. Preliminary cale (to I.L.R.) UPR 9074 is affiliated with results from this second study suggest that INSERM and University of Orléans. certain test variables are sensitive to test order (such as latency to enter in the light-dark Claudia F. Plappert1, Peter K.D. Pilz1, and paradigm) whereas others (i.e. PPI) are H.-U. Schnitzler1. Sensitization of the resistant. These two studies demonstrate that acoustic startle response in mice differs some behavioral tests appear to be sensitive between two inbred strains and is influ- to previous testing experience, while other enced by glycine.2 tests may be immune. Sensitization is evoked by strong, aver- 1Department of Molecular and Human sive stimuli and causes a general increase in Genetics, Baylor College of Medicine, the strength of behavioural responses. The Houston, TX, USA. 2This work was supported acoustic startle response (ASR) is a general by AG18232-01. coordinated muscle contraction elicited by loud acoustic stimuli which can be used to Stéphane Mortaud1, Laurent Nicolas1, Isa- study two different sensitization paradigms. belle Le Roy1, and Pierre L. Roubertoux1. First, acoustic sensitization evoked by the Attack behavior in mice: implication of the startle stimuli themselves slows down or even sts gene mapped on the pairing region of prevents the normally observed habituation the X-Y chromosomes.2 process. Second, electric sensitization evoked The sexual dimorphism of aggression by footshocks increases the ASR compared to has led to a search for its Y- chromosomal the response amplitude before footshocks. correlates. We have previously confirmed that Acoustic sensitization and footshock sensitiza- initiation of attack behavior against a con- tion interact if acoustic stimuli have a high specific male is Y- dependent in two strains of SPL. High level acoustic stimuli produce laboratory mice (NZB and C57BL/6J). We strong sensitization preventing footshocks have provided evidence that the pairing region from eliciting additional sensitization. of the Y co-segregates with attack behavior, in We were interested in two questions: 1.) these strains. In addition, the genetic corre- Is there a genetic influence on sensitization? If lates of attack behavior are not expressed so, sensitization should differ between geneti- when borne on the homologous pairing region cally different mouse strains. We examined on the X chromosome but only when carried two different strains, DBA/CN and BALB/CAN on the Y chromosome. Only one functional in both sensitization paradigms. 2.) Does gly- gene (coding for steroid sulfatase or STS) is cine, a major inhibitory transmitter in spinal mapped on this region as of yet, suggesting cord and brainstem, play a role in sensitiza- that it could be a candidate for attack behav- tion? If so, in spasmodic (spd) mutant mice ior. We estimated the genetic correlation be- with a defective glycine receptor alpha-subunit tween the concentration of STS protein in the sensitization should be changed compared to liver and initiation of attack behavior. We have the wildtype (WT). employed also mice in which gene invalidation For acoustic sensitization the mice were induced attack behavior. Pharmacological given 200 acoustic stimuli with either high SPL modulations of STS or of its metabolites modi- (25 dB above startle threshold) or low SPL fies the frequencies of attack in these male (individually determined so that a weak ASR mice, confirming the implication of STS in ag- amplitude of about 100 mV was elicited in gression. Recent investigations have demon- each animal). ASR amplitude was measured strated the involvement of STS in neurosteroid in a standard movement sensitive device. For biochemical pathways, and several lines of electric sensitization, footshocks (0.5 mA, 500 evidence indicate that neurosteroids interact ms duration, 1 shock/s) were presented be- with neurotransmitters. These conclusions and tween two series each consisting of 40 high our present results support the hypothesis that SPL startle stimuli. sulfatation of steroids may be the prime mover 1.) High SPL stimuli produced higher of a complex network, including genes shown ASR amplitudes in BALB than in DBA. The to be implicated in aggression by mutagene- course of ASR during repetitve stimulation sis. differed between the two strains. In BALB,
21 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society
ASR initially increased and then only de- identified. The current research is designed to creased slowly, while in DBA, ASR decreased investigate the role of a number of specific about exponentially over the startle stimuli. genes expressed in identified neurons in the Only with low SPL startle stimuli both strains response to tap and in the plasticity expressed showed about the same ASR amplitude and in the response. A number of strains of mutant the same exponential amplitude decrease. We worms were studied. The effect of the assume that in BALB sensitization by high mutation on spontaneous behavior, on SPL acoustic stimuli was very strong leading response to tap and on habituation to tap were to high ASR amplitudes, and conteracting ha- tested. Several mutations affected bituation, indicated by the initial ASR increase spontaneous behavior without affecting the and the slow decrease afterwards. In DBA evoked behavior, while others altered sensitization was only weak and habituation habituation to tap without affecting the initial was able to succeed leading to a fast ASR response to tap. Most of the strains tested decrease. With low SPL startle stimuli, that were deficient in some aspect of glutamate elicited the same weak ASR in both strains, transmission, with altered levels of presynaptic the amount of habituation was the same indi- glutamate or altered post-synaptic receptors cating same weak sensitization in both strains (AMPA-type, NMDA-type, GluCl-type or in this case. Footshocks presented after high metabotropic type). SPL startle stimuli elicited a much smaller 1Department of Psychology, University of ASR increase in BALB than in DBA. This con- British Columbia, Vancouver, BC, Canada. firms our hypothesis that in BALB acoustic 2This research was supported by an NSERC sensitization is strong preventing additional operating grant from the Canadian footshock sensitization. Government. 2.) In spd, high SPL startle stimuli elic- ited about 3-fold stronger ASR amplitudes as C. Reavill1, P. Nelson2, J. Latcham3, and in the WT, and the ASR showed no time de- J.J. Hagan1. Prepulse inhibition and startle pendent changes. With low SPL startle stimuli, responses in dishevelled (Dvl1-/-) mice. ASR in spd declined about exponentially over Behavioural analyses of mice with se- time, comparable to the WT in the high and lective chromosomal mutations are currently the low SPL condition. Footshocks elicited no being conducted in order to link genes with ASR increase in spd, but a strong increase in disease states. Dishevelled (Dvl1-/-) mice the WT. We assume that spd are at a high have been reported to show reduced prepulse sensitization level producing high ASR ampli- inhibition (PPI) (Lijam et al, 1997, Cell, 90: tudes, counteracting habituation and prevent- 895-905). Therefore we have studied PPI and ing footshock sensitization. startle responses in Dvl1-/- mice. Male and Our results indicate that sensitization is female mice were tested at 8 weeks of age in genetically influenced. To locate the genes custom built startle chambers. For PPI trials, that are responsable for the stronger sensiti- mice received either acoustic stimuli of white zation in BALB compared to DBA further ge- noise pulses (110dB/10 msec) or prepulses netic analysis (e.g. QTL analysis) of these (4, 12 or 20 kHz/80 or 90 dB/10 msec) each strains will be necessary. Furthermore glycine followed 100 msec later by a white noise plays a role in sensitization. We assume that a stimulus (110dB/10 msec). Dvl1-/- mice were glycinergic tonic inhibition exits on the path- further tested for PPI using a pulse stimulus of ways mediating sensitization that is not de- 100dB. In separate startle experiments mice scribed until now. received acoustic stimuli (4, 12, 20 kHz or 1University of Tübingen, Dept. of Animal white noise, 80, 90 or 110dB/10 msec). Re- Physiology, Morgenstelle 28, D-72076 Tübin- peated measures analysis of variance was gen, Germany. 2Supported by DFG. used to analyse untransformed PPI responses and log transformed startle responses. There Catharine H. Rankin1, Jacqueline K. Rose1, were no significant effects of genotype on PPI Kenneth Eng1, and Karla Kaun1. Genetic at a pulse of 110 dB (F[1,34] = 0.22; P = 0.64), dissection of habituation of the tap 100 dB (F[1,34] = 0.002; P = 0.97) or on star- withdrawal response in C. elegans.2 tle response (F[1,34] = 0.43; P = 0.52) com- Previous research has shown that the pared to 129/SvEv littermates. Also, there was tap withdrawal response in the nematode C. no significant effect of sex on PPI at a pulse of elegans is composed of two competing 110 dB (F[1,34] = 0.21; P = 0.65) or at 100 dB reflexes that are integrated to produce the (F[1,34] = 3.30; P = 0.08). Contrary to previ- response. The tap withdrawal response shows ously published data (Lijam et al, 1997) Dvl1-/- short- and long-term habituation. The neural mice did not show PPI deficits in this study. circuit underlying the response has been
22 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society
1Neurosciences Research, 2Statistical Sci- mice was supported by the high Ca2+- ences, and 3Laboratory Animal Sciences, SB saturation of calbindin-D28K in these cells as Pharmaceuticals, Third Avenue, Harlow, Es- demonstrated by the immunocytochemical sex, CM19 5AW, UK. and biochemical characterization of a para- doxical calretinin-like immunoreactivity. Michael Regulski1, Grigori Enikolopov2, The firing behavior of Purkinje cells and Tim Tully1. Genetics of NO signaling in analyzed in alert mice is severely affected in the adult Drosophila melanogaster brain. both 2-4 months- and 2-2.5 years-old Cr-/- Nitric oxide (NO) is a transcellular mes- mice with a 200% increase in the spontaneous senger involved with a wide variety of physio- simple spike firing rate, a 30-50% reduction in logical processes in vertebrates including complex spike duration and a 70-85% reduc- vasodilation, immune response and synaptic tion in the simple spike pause as compared plasticity. It has also been found to participate with their wild-type littermates. In contrast, in in developmental and behavioral plasticity in cerebellar slices, excitatory synaptic transmis- invertebrate nervous system including long- sion and short-term synaptic plasticity at paral- term memory formation in honeybees. We are lel fiber- or climbing fiber-Purkinje cell syn- interested in characterizing NO signaling in apses are unaltered, indicating that marked the fly adult brain. We have cloned and char- modifications of the firing behavior in vivo can acterized dNOS gene, which codes for a Dro- be undetectable in slice. sophila homolog of the nitric oxide synthase, These results show that calretinin plays the main source of NO. Immunocytochemical a major role at the network level in cerebellar staining indicates that DNOS protein is ex- physiology without modification of properties pressed throughout the adult fly brain. We of single cells examined so far and also show found that its expression levels rise for the first that the knock-out technique can unexpectedly two days after eclosion and then decline. We « knock-in » new form of molecules as a neu- think that these changes are linked to specific ron’s reponse to these network changes. events in the final developmental stages of the 1Lab. Neurophysiology, 5IRIBHN, Université fly brain. We carried out a screen for muta- Libre de Bruxelles, Brussels, Belgium. 2Lab. tions in dNOS. Analysis of the phenotypes Neurophysiology, Université Mons, Belgium. associated with these mutations will help us to 3Lab. Neurobiology, CNRS-URA 1857, Paris, understand NO functions in the fly adult brain. France. 4Dept. Neurochemistry, Max Planck 1Cold Spring Harbor Laboratories, 1 Bungtown Inst. Psychiatry, Martinsried, Germany. Road, Cold Spring Harbor, NY 11724, USA 6Supported by PAI, FMRE and FRSM (Bel- gium) and Boerhinger-Ingelheim. S.N. Schiffmann1, G. Cheron2, A. Lohof3, P. d’Alcantara1,5, M. Meyer4, M. Parmentier5, B. Schwaller1. The lack of parvalbumin af- and S. Schurmans5. Calretinin, cerebellar fects the body and the mind. network activity and motor coordination.6 Parvalbumin (PV) is a cytosolic low- The involvement of the cerebellum in molecular weight, high-affinity Ca2+/Mg2+- motor control has long been recognized. binding protein which is predominantly ex- However, the identification of this brain area pressed in specific subpopulations of as a primary site of motor learning is a largely GABAergic interneurons in various brain re- supported but still controversial hypothesis. In gions. Additionally, PV is expressed in fast- the cerebellum, the parallel fiber-Purkinje cell twitch muscle fibres of rodents where the con- synapse can undergo long-term synaptic plas- centration is estimated to be in the millimolar ticity suggested to underlie motor learning and range. To deduce possible functions of PV in resulting from variations in intracellular cal- these two tissues, PV-deficient mice (PV-/-) cium concentration ([Ca2+]i). Ca2+ binding pro- were generated by homologous recombina- teins such as calretinin, calbindin and parval- tion. In fast-twitch muscles of PV-/- mice the 2+ bumin are enriched in the cerebellum but their decay of [Ca ]i after 20 ms stimulation is role in information processing is poorly under- slower compared to wild-type (WT) mice and stood. leads to a prolongation of the time required to Mice deficient in calretinin (Cr-/-) have attain peak twitch tension and to an extension been generated by gene targeting. They are of the half-relaxation time. This is in good impaired in tests of motor coordination such agreement with Ca2+-measurements in PV- as the runway, the horizontal stationary rod containing hippocampal neurones or PV- and the wheel running test with a severe injected chromaffin cells. In both cases, PV as worsening with aging, suggesting functional a slow-onset buffer -the rate of Ca2+-binding deficits in cerebellar pathways. An impairment being determined by the off-rate of Mg2+- does in Ca2+ homeostasis in Purkinje cells of Cr-/- not affect the amplitude of Ca2+-transients, but
23 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society significantly increases the initial decay of opmental processes in Drosophila melano- 2+ [Ca ]i followed by a later exponential decay, gaster. Given this success with Drosophila, we thus even prolonging the transient. The have undertaken a mutagenesis programme knowledge about the metal-binding character- to identify mutations at loci that might exas- istics (affinities, kinetics) of PV is an absolute perate the mild phenotypic defects associated requirement for the understanding of the with 'knock-out' mutants at either the amyloid physiological role played in either muscle fi- beta (A4) precursor protein or prion protein bres or neurones. locus. Offspring from ENU (ethyl nitrosourea) The shape of cytosolic Ca2+-transients treated male mice were subjected to a serious in neurones are determined by the kinetics of of simple behavioral and observational Ca2+-entry systems (channels in the plasma screens. To date, we have screened in excess membrane or intracellular organelles), cytoso- of 1500 G1 progeny and observed sixteen lic Ca2+-buffers such as PV or calbindin-D28k heritable mutations. Three involved mutations (CB) and finally by extrusion systems charac- at the tyrosinase (albino) locus providing an terized by their Ca2+ extrusion rates. Simula- estimate of mutation frequency consistent with tion studies based on the experimental data published data. Several exhibited gait abnor- from hippocampal neurones or PV-injected malities while other displayed reduced body chromaffin cells indicate that slow buffers (PV) weight. Two abnormal gait mutants (lines 67 combined with high extrusion rates can mimic and 81) exhibit phenotypic variations that ap- a very rapid Ca2+ extrusion mechanism. pear to depend upon the presence or absence One fundamental aspect of synaptic of either App or Prnp gene. Line 67 exhibit transmission, short-term plasticity, has long episodes of slow deliberate movement which been known to be related to Ca2+ homeosta- last for a few days around weaning with a mild 2+ sis, in particular to levels of [Ca ]i remaining residual high stepping gait which persists into after neuronal electrical activity (“residual Ca2+ adult life. Other characteristics include plastic hypothesis”). Interneurones from different tail, 'praying behavior' and poor co-ordination brain regions (e.g. neocortex, cerebellum, hip- that are also episodic. Presumptive mutant pocampus) contain high levels of PV, both in homozygotes die before weaning and exhibit a the soma and processes. PV was hypothe- more severe phenotype with earlier onset. The sized to modify the time course of residual most prominent characteristic associated with Ca2+ removal in terminals following an action these presumptive homozygotes is a difficulty potential, and hence modulate short-term to right themselves when they roll on their plasticity. In the presentation, results from hip- back. Line 81 has a more complex phenotype pocampal and cerebellar electrophysiological however, the most persistent characteristic is recordings in PV-/- mice will be discussed. an intention tremor. PrP-null homozygotes are Furthermore, the lack of PV affects the spon- significantly more reactive to changes in envi- taneous locomotor activity in these mice and ronmental conditions (e.g., hyperactive, ag- is likely linked to the absence of PV in Purkinje gressive) than null heterozygotes. Detailed cells. They contain the highest known concen- descriptions of both mutants will be presented trations of PV and CB compared to all other along with video recordings in an attempt to neurones. Mice deficient for both, PV and CB get a better understanding of their clinical sig- displayed a drastic decrease in locomotion, nificance. characterized by a reduced speed and a low 1McLaughlin Research Institute, 1520 23rd percentage of fast movements. A recent find- Street South, Great Falls MT 59405, USA. ing on altered dendritic spine morphology of 2Stowers Institute for Medical Research, 1000 Purkinje cells in PV-CB double KO mice corre- East 50th Street, Kansas City, MO 64110 lates well with the data on the impairment of USA. 3This work was supported by the Ameri- spontaneous locomotor activity. can Health Assistance Foundation. 1Institute of Histology and General Embryol- ogy, University of Fribourg, Switzerland. Fred van Leuven1. Transgenic mice and Alzheimer pathology: no need for plaques Dennis A. Stephenson1,2, Julie Gilchrist1, or tangles? Dionne Peterson1, Sherry Tuner1, Corike We have generated different strains of Nuibe1, and George A. Carlson1. ENU in- transgenic mice that overexpress either wild- duced behavioral mouse mutants which type or mutant Amyloid Precursor Protein may involve APP or PrP.3 (APP), human wild-type or mutant Presenilins Induction of mutations that contribute to (PS1, PS2), human ApoE4 or human protein a phenotype (by enhancement or alleviating a tau. All constructs were based on the mouse characteristic) has proven to be highly effec- thy1 gene promoter to express the transgenes tive in dissecting pathways that affect devel- specifically in neurons and all mice were gen-
24 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society erated and maintained in the FVB back- cognitive and behavioral deficits observed in ground. all APP transgenic strains that we have gen- Remarkably, all APP transgenic mouse erated and characterized, from the late and strains displayed in essence a similar behav- selective development of amyloid plaques only ioral phenotype as the original APP/RK mice in old APP/london mice that produce high lev- (Moechars et al., 1996, EMBO J. 15:1265- els of Aß42. Whereas the occurrence of 1274; 1999, J. Biol. Chem. 274:6483-6492). plaque and vascular amyloid is explained by The major symptoms include: disturbed be- these higher levels of Aß42, the cognitive and havior of reduced exploration, neophobia, in- behavioral phenotypic traits must be linked to creased aggression, excitotoxicity with prema- other metabolites of APP, i.e. Aß40, ß-C-stubs ture death and hypersensitivity to kainic acid, and secreted APP, most likely in combination. but hypo-sensitivity to NMDA with reduced To define their respective contributions, other cognition and defective LTP. The differences and more complex transgenic mouse strains between the APP transgenic strains were are being generated. quantitative in nature, i.e. differences in inten- Double transgenic mice, i.e. sity and in severity, or differing in the age of APP/london x PS1[A246E], develop amyloid onset of the symptoms. These quantitative plaques when only 6-9 months old, concomi- differences correlated with the level of expres- tant with increased Aß42 levels. The other sion of the transgene and with its nature, i.e. APP-metabolites are relatively unchanged, wild-type APP less severe than mutant APP. which is concordant with the observation that All the symptoms mentioned are "early de- the early behavioral traits in the APP/Lo x PS1 fects", i.e. obvious at ages ranging from 3 and double tg mice are not essentially different 9 months. Measured in the Morris watermaze, from the single APP/london tg mice. Single the cognitive deficit was most marked in PS1 tg mice that overexpress either the wild- APP/london transgenic mice even when only 3 type human PS1 or the EOFAD mutant to 6 months, but was also observed in APP/wt PS1[A246E], have essentially no pathology or transgenic mice (Moechars et al., 1999, J. phenotypic abnormalities. Mice deficient in Biol. Chem. 274:6483-6492). PS1 are not viable, but primary cultures of The common early symptoms are in embryonal neurons grow and differentiate obvious contrast with the absence of amyloid normally, and were used to demonstrate that plaques in the APP/RK and APP/wt mice and production of the amyloid peptides is reduced with the late appearance of amyloid plaques dramatically in the absence of PS-1 (De together with vascular amyloid in the brain of Strooper et al., 1998, Nature 391:387-390). the APP/london transgenic mice. These le- To overcome the lethality of the PS1 sions developed only in the highest express- deficiency, we have now generated mice that ing APP/london transgenic strain, and only are neuronally deficient in PS1: mice with a when these mice were 12 months or older. "floxed" PS1 gene were crossed with trans- The biochemical correlate of the phenotype genic mice that overexpress cre-recombinase and of the occurrence of plaque and vascular in neurons by way of the thy1 gene promoter. amyloid became evident from analysis of the The viable offspring accumulate ß-C-stubs in brain levels of membrane-bound APP, of α- their brain from endogenous mouse APP, and ß- secreted APP, of the C-terminal "stubs" demonstrating in vivo the metabolic impact of (C99) and of Aß40 and Aß42. No single inter- PS1 deficiency on γ-secretase cleavage (Ilse mediate correlated completely with the com- Dewachter et al., unpublished results). These plex of phenotypic dysfunctions. On the other mice are being characterized and further hand, the development of amyloid plaques in crossed with APP/london transgenic mice to the APP/london mice was direct correlated to yield "triple" transgenic mice, i.e. homozygous high Aß42 levels. The histochemical and im- for floxed PS1, and heterozygous for thy1- munochemical characteristics of the amyloid Cre-recombinase and for thy1-APP/London. In plaques and of the vascular amyloid recapitu- essence these mice overexpress human late closely the amyloid pathology of AD pa- APP/london in the same neurons that are de- tients (Van Dorpe et al., 2000, Am. J. Pathol. ficient in PS1 and will be instrumental to an- in press). This includes immunoreactivity for swer the question if and how accumulated ß- hyper-phosphorylated tau in swollen neurites C-stubs are pathological or beneficial. This around the amyloid plaques, but without the constitutes the in vivo paradigm of the thera- neurofibrillary inclusions of tau-pathology seen peutic intervention in AD patients aimed at in AD patients. inhibition of γ-secretase to reduce production The combined observations in the APP of the amyloid peptides which would entail transgenic mouse strains demonstrate above accumulation of their obligate immediate pre- all the marked dissociation in time of the early cursors, the ß-C-stubs of APP.
25 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society
Although the APP/london mice faithfully tant or even deciding for the pathogenesis of recapitulate the plaque and vascular amyloid AD in carriers of ApoE4 alleles. We have gen- pathology of AD patients and also demon- erated transgenic mice that over-express hu- strate interesting cognitive and behavioral man ApoE4 in either neurons (thy1 gene pro- problems, the pathological aspect of neurofi- moter) or astrocytes (GFAP gene promoter). brillary tangle formation is lacking. This is a Transgenic mice with neuronal expres- paradox and could also explain the lack of an sion of human ApoE4 progressively exhibited early cholinergic deficit in the APP/london motoric problems that correlated with neuronal mice, whereby the cholinergic defect is re- hyper-phosphorylation of protein tau. Neurons stricted to disturbed cholinergic tracts associ- in brain and spinal cord reacted positively with ated with amyloid plaques in old APP/London monoclonal antibodies AT8, AT180 and PHF1, mice. which specify AD related epitopes. Increased To implement and understand the prob- protein Tau phosphorylation was dependent lem of tau-pathology in AD, we have gener- on the level of neuronal expression of human ated transgenic mice that overexpress human ApoE4 and on the age of the mice. In addition, protein tau and two suspected tau-kinases, ApoE4 transgenic mice developed axonopa- GSK-3ß and cdk5 with its activator p35. Over- thy, severe motor impairment and neurogenic expression of protein tau4R in neurons results muscle atrophy (Tesseur et al., 2000, Am J in transgenic mice that were psychomotori- Pathol, 156, 3, 951-964, Am. J. Pathol. in cally impaired and developed prominent press). Numerous inclusions stained positive axonopathy in brain and spinal cord. Axonal for ubiquitin, neurofilaments and synapto- dilations with accumulation of neurofilaments, physin in the white matter tracts of the CNS, mitochondria and vesicles are prominent sug- indicating impairment of axonal transport. This gesting that defective axonal transport causes was confirmed at the ultrastructural level and axonal degeneration. This effect was gene- was similar but not identical to defects in our dosage related, it proved that merely increas- transgenic mice that overexpress human pro- ing the concentration of the four-repeat tau tein tau. In sharp contrast, none of these protein isoform is sufficient to cause neuronal symptoms were detected in transgenic mouse injury without additional requirement of intra- lines that over-express human ApoE4 in as- neuronal neurofibrillary tangles (Spittaels et trocytes at similar levels. Our data link, for the al., 1999, Am. J. Pathol. 155: 2153-2165). first time, the genetic risk-factor ApoE4 to a Since patients with AD or other tauopa- pathological defect in AD, i.e. hyper- thy develop tau inclusions that consist of hy- phosphorylation of protein tau. The polymor- per- phosphorylated protein tau, tau kinases phisms in the ApoE gene promoter that consti- are thought to be actively involved. Overex- tute increased risk for developing AD could pression of GSK-3β in neurons of single and then also cause or define expression of ApoE double transgenic mice, provided evidence in neurons, and, combined with our results, that GSK-3β is indeed an effective protein tau- offer a mechanism for the pathogenic role of kinase in vivo. Hyperphosphorylation of mur- the ApoE4 allele in AD. ine and human protein tau was exemplified by Experiments in "multiple" transgenic the appearance of isoforms with slower elec- mice are ongoing and other relevant genes trophoretic mobility, immunoreactive with are being implemented to determine which of monoclonal antibodies AT-8 and AT-180, the APP metabolites is causing the early signs among others that are certified to recognize of the "amyloid"-related phenotype, how "tau- typical phosphorylated tau epitopes in AD pathology" is related and to be implemented brain. Further analysis at the ultrastructural and how the neuronal "ApoE4-tau" connection level will reveal whether and how tau phos- is operating, to eventually determine the im- phorylation is involved in tangle formation or in portance, if any, of the intraneuronal tangles, the other neuro-degenerative processes. the other lesion essential for diagnostics of ApoE4 is an important genetic risk- AD. 1 factor for AD, but besides the epidemiological Experimental Genetics Group, Center for evidence, the molecular contribution of ApoE4 Human Genetics, Flanders Institute for Bio- to the neurodegenerative pathogenesis is not technology, K.U. Leuven-Campus Gasthuis- known. Rodent neurons do not express any berg O&N 06, B-3000 Leuven, Belgium ApoE, as opposed to astrocytes, while some 1 evidence is available that human brain regions Fred van Leuven . Transgenic mouse mod- in which neurons express ApoE might be most els for Alzheimer's disease. vulnerable for developing neurofibrillary pa- ApoE4 is an important genetic risk- thology. We tested the hypothesis that the factor for AD, but besides the epidemiological expression pattern of human ApoE is impor- evidence, the molecular contribution of ApoE4
26 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society to the neurodegenerative pathogenesis is not Netherlands, selective breeding in wild house known. Rodent neurons do not express ApoE, mice has resulted in two lines: [1] a fast at- as opposed to astrocytes, while some evi- tacking, aggressive line characterized by short dence is available that human brain regions in attack latencies (SAL) and [2] a slow attack- which neurons express ApoE might be most ing, non-aggressive line with long attack la- vulnerable for developing neurofibrillary pa- tencies (LAL). However, aggression is not the thology. We tested the hypothesis that the only behavior in which these lines vary and it expression pattern of human ApoE is impor- is generally believed that they differ more fun- tant or even deciding for the pathogenesis of damentally with each line showing its particu- AD in carriers of ApoE4 alleles. We have gen- lar neurobehavioral characteristics. An impor- erated transgenic mice that over-express hu- tant genetic candidate that might explain this man ApoE4 in either neurons (thy1 gene pro- neurobehavioral variation is the Y chromo- moter) or astrocytes (GFAP gene promoter). some. To investigate Y chromosomal effects, Transgenic mice with neuronal expression of congenic lines have been developed (named human ApoE4 progressively exhibited motoric SAL.LY, LAL.SY), which only differ from their problems that correlated with neuronal hyper- parental lines (SAL, LAL) with regard to the Y phosphorylation of protein tau. Neurons in chromosome (i.e. the nonpairing part). Com- brain and spinal cord reacted positively with parisons between SAL.LY and SAL on the one monoclonal antibodies AT8, AT180 and PHF1, side and LAL.SY and LAL on the other side which specify AD related epitopes. Increased may then reveal possible neurobehavioral ef- protein Tau phosphorylation was dependent fects of the Y chromosome. This presentation on the level of neuronal expression of human reviews the results of previous studies of Y ApoE4 and on the age of the mice. In addition, chromosomal effects on behavior (aggression, ApoE4 transgenic mice developed axonopa- defensive burying), neurochemistry (apomor- thy, severe motor impairment and neurogenic phine-induced stereotypy) and neuroanatomy muscle atrophy (Tesseur et al, Am J Pathol, (the sizes of the hippocampal inter- and in- 2000, 156: 951-964). Numerous inclusions trapyramidal mossy fibers terminal fields) in stained positive for ubiquitin, neurofilaments these selection lines. Moreover, it investigates and synaptophysin in the white matter tracts of whether Y chromosomal variation affects the the CNS, indicating impairment of axonal behavior of these mice in four paradigms sug- transport. This was confirmed at the ultrastruc- gested to study to some extent anxiety and/or tural level and was similar but not identical to depression levels: the elevated plus maze, defects in our transgenic mice that overex- open field, sudden silence test and Porsolt's press human protein tau. In sharp contrast, swim test. none of these symptoms were detected in 1Department of Animal Physiology, University transgenic mouse lines that over-express hu- of Groningen, Haren, The Netherlands and man ApoE4 in astrocytes at similar levels. Our 2Department of Psychoneuropharmacology, data link, for the first time, the genetic risk- University of Nijmegen, Nijmegen, The Neth- factor ApoE4 to a pathological defect in AD, erlands. i.e. hyper-phosphorylation of protein tau. The polymorphisms in the ApoE gene promoter S. Viñas1, S. Lewis1, S. Barreau1, C. Du- that constitute increased risk for developing cottet1, A. Aubert1, and C. Belzung1. Strain AD could then also cause or define expression differences in two animal models of de- of ApoE in neurons, and, combined with our pression : the forced swimming test and results, offer a mechanism for the pathogenic the chronic mild stress procedure. role of the ApoE4 allele in AD. The present study was aimed at inves- 1Experimental Genetics Group, Center for tigating strains differences in animal models of Human Genetics, Flanders Institute for Bio- depression. We focused on two models: the technology, K.U. Leuven-Campus Gasthuis- forced swimming test and a chronic mild berg O&N 06, B-3000 Leuven, Belgium stress procedure. The effects of chronic mild stress were observed on sexual behavior, su- A.H. Veenema1, G.A. Van Oortmerssen1, crose consumption and fur state of the mice A.J.H. De Ruiter1, B. Bohus1, J.M. Kool- (cleaned or not). Mice from 9 different inbred haas1, and F Sluyter2. Neurobehavioral ef- strains were used: BALB/cByJ, C57BL/6J, fects of the Y chromosome in wild house 129Sv/ter, AKR/OlaHsd, CBA/J, mice. C3H/HeNHsd, AJOlaHsd, SJL/JHanHsd and In mice artificial selection is one of the DBA/2OlaHsd. most powerful tools to demonstrate the contri- Results show: 1) in the forced swim- bution of genetic variation to individual differ- ming test, immobility duration was the highest ences in aggressive behavior. In Haren, The in C57, AKR and 129 mice while swim dura-
27 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society tion was higher in BALB and DBA mice; 2) in released into an oudoor pen of 10x10 m. Four the chronic mild stress procedure, the main feeder boxes were placed in distant corners, effect was seen on fur state. Indeed, this state four boxes in proximity to two central shelters remained optimal during the 4 weeks of stress offering protection against wind and rain. in C57 mice while it worsened in BALB and Predators were barred from the pen by nets DBA mice after 1 week and in AJ and SJL af- and electrical fences. A daily feeding cycle ter 3 weeks of stress. started at 8 p.m. and lasted overnight till 8 We conclude that the choice of a given a.m. During this time, a mouse would receive strain is an important factor, that should be a food reward of about 0.5 g at every box but considered when studying the effects of a only during the first visit. As it was not certain given pharmacological treatment or a given whether poorly learning mice would receive mutation. Moreover, present results show that enough food, additional free food was placed the two animal models of depression may not inside the shelters every third day. measure the same psychological state, be- It was found that all 3 groups (WW, WM cause they are not sensitive to the same fac- and MM) gradually learned to visit all eight tors of variation. feeders during a night, without significant dif- 1JE Psychobiologie des Emotions, UFR Sci- ferences between the groups. This indicates ences, Parc Grandmont, F-37200 Tours, that the TrkB mutants did not suffer from spa- France. tial memory deficits nor from basic learning inabilities. However, differences between mu- A.L. Vyssotski1,2, G. Dell’Omo1,3, D.L. Vys- tants and wildtypes emerged gradually at sotski2, D.P. Wolfer1, L. Minichiello4, R. those days with free food inside the shelters. Klein4, I.I. Poletaeva5, and H.-P.Lipp1. Mice While the wildtypes soon abandoned to visit lacking the neurotrophin receptor TrkB in the outside feeders during such nights, the the forebrain show intact spatial memory TrkB mutants continued to patrol the boxes in but impaired behavioral flexibility in a their habitual way. For example, on day 21, semi-naturalistic set-up.6 the average number of boxes visited by the Mice lacking the neurotrophin receptor mutants was 7.0, while the average visits to TrkB in the forebrain have been found to be those boxes dropped to 0.86 in the wildtypes unable to learn the Morris water maze task, to (Mann-Whitney, p<0.007). The heterozygous be impaired in radial maze learning, while they animals performed in-between (average visits appeared less or not affected in more simple 4.7, different from wildtypes, Mann-Whitney, learning tasks such contextual fear condition- p< 0.002). ing (Neuron 24, 401-414, 1999). In order to In conclusion, these data show rather clarify how such an impairment would affect convincingly that TrkB deficient mice were the learning ability in a natural environment, able to learn but, once having learned a task, we have developed a system consisting of were almost unable to switch quickly to an- eight computer-controlled feeder/trap units. other behavioral strategy. This implies that the This set-up permits to deliver (or withhold) TrkB neurotrophin receptor must play a fun- food reward to individual transponder tagged damental role in behavioral flexibility. In addi- mice visiting the traps for obtaining their daily tion, natural learning set-ups monitoring a col- food. The system operates inside a mouse lective of mice permit to recognize such slowly colony kept permanently outdoors (or in- emerging behavioral changes (or lack thereof) doors), and records continuously visits of much better than single test episodes of indi- every mouse. After an individual mouse has vidual mice. 1 entered a feeder box, other mice are barred Institute of Anatomy, University of Zürich, 190 from entering until the occupant has left. Thus, Winterturerstrasse, CH-8057 Zürich, Switzer- 2 the following variables are recorded for each land. P.K. Anokhin Institute of Normal Physi- individual: i) time of visit; ii) place of visit; iii) ology, Russian Academy of Sciences, Mos- 3 food reward or not (correct choice and later re- cow, Russia. Laboratory of Veterinary Medi- entries). The task resembles a radial maze cine, Istituto Superiore di Sanità, Rome, Italy. 4 test but is more complicated because of pro- European Molecular Biology Laboratory, Hei- 5 tracted food delivery schedules and the ne- delberg, FRG. Laboratory of Genetics and cessity of adjusting behavior when faced with Physiology of Behavior, Chair of Higher Nerv- a box occupied by another mouse feeding in- ous System Activity, Moscow State University, 6 side. On the other hand, it emulates the eve- Russia. Supported by Swiss National Science ryday learning requirements of mice quite Foundation (31-46691.96, 31-57139.99) and nicely. EC BIO4-98-0297/BBW98.0125. Forty mice ( 11 wildtypes, WW; 21 het- erozygous, WM; and 8 mutants, MM) were
28 KNOCKOUTS & MUTANTS III: Genetically Dissecting Brain and Behavior Third Annual General Meeting of the International Behavioural and Neural Genetics Society
David Westaway1. Transgenic mouse mod- to immunization with Αβ42, although there are els of Alzheimer’s Disease: long-haul or quantitative and qualitative differences from home-stretch? studies reported by Schenk and co-workers in Although there is intense interest in PDAPP mice. The advantages and shortfalls recreating an accurate murine model of Alz- of TgCRND8 mice as (i) a credible AD model heimer Disease (AD), there are considerable as per the criteria described above and (ii) a obstacles associated with this goal. It has vehicle to decipher the impact of amyloid been suggested that accurate models should deposition and amyloid-directed therapeutics fulfill some simple expectations including: 1) on neuropathology and cognitive function will Progressive neuropathology culminating in be discussed. several accepted pathologic hallmarks of AD 1Centre for Research in Neurodegenerative (plaques, tangles, loss of cholinergic neurons). Diseases and Department of Laboratory Medi- 2) Presence of cognitive deficits as well as cine and Pathobiology, University of Toronto. neuropathologic changes noted in 1. Such cognitive changes should be robust, and evi- dent in different behavioural paradigms target- ing the same memory system. Ideally, behav- ioural paradigms employed to test mice should address neuroanatomic structures that are affected in AD (e.g., the hippocampus). 3) In the case of experiments employing familial Alzheimer Disease (FAD) mutations, pheno- typic changes documented in 2 and 3 should be correlated with the presence of the FAD mutations, and should be absent or less overt in mice expressing wt gene alleles expressed at equal (or greater) steady-state levels. 4) Recapitulation of phenotypic changes as per items 1-3 in independent Tg lines harbouring the same construct, to exclude the contribu- tion of insertional mutations or linked loci. 5) Confirmation of key phenotypic traits in inde- pendent laboratories. As there are few AD models address- ing these criteria, this laboratory has also en- tered the fray to create transgenic (Tg) mice with AD-related pathology. Using a mutant form of APP695 inserted into the cos.Tet prion promoter expression vector, microinjections were carried out into a mouse genetic back- ground partially protective against the effects of APP over-expression. This strategy allowed the establishment of a Tg line expressing high levels of mutant APP and designated TgCRND8. Aβ is readily detectable in the brains of TgCRND8 mice by western blotting, from 120 days onwards. Fields of amyloid plaques can be detected in the cortex and hip- pocampus by immunohistochemistry from 90 days of age. Congophilic and birefringent Aβ deposits with similarities to senile plaques of AD are seen from an early age, but although dystrophic neurons are present, neurofibrillary tangles have yet to be detected. Co- expression of either mutant PS1 or PS2 greatly accelerates amyloid deposition in TgCRND8 mice, to the extent that dense fields of plaques are evident from between 43 to 90 days of age, depending upon the particular presenilin mutation. TgCRND8 mice respond
29