VU Research Portal

Behavioral Phenotyping of Complex Traits in Inbred and Mutant Mice Maroteaux, G.P.

2014

document version Publisher's PDF, also known as Version of record

Link to publication in VU Research Portal

citation for published version (APA) Maroteaux, G. P. (2014). Behavioral Phenotyping of Complex Traits in Inbred and Mutant Mice.

General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal ?

Take down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

E-mail address: [email protected]

Download date: 23. Sep. 2021 Behavioral phenotyping of complex traits in inbred and mutant mice

Gregoire Pierre Maroteaux Cover and chapters Artwork: Anton Rammelt Printed by www.offpages.nl ISBN 978-94-6182-464-6 Copyright © G. Maroteaux, 2014 All rights reserved. No part of this thesis may be reproduced, stored in a retrieval system, or transmitted in any form or by any means without prior permission of the author. The printing of this thesis was sponsored by: VRIJE UNIVERSITEIT

Behavioral phenotyping of complex traits in inbred and mutant mice

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad Doctor aan de Vrije Universiteit Amsterdam, op gezag van de rector magnificus prof.dr. F.A. van der Duyn Schouten, in het openbaar te verdedigen ten overstaan van de promotiecommissie van de Faculteit der Aard- en Levenswetenschappen op donderdag 25 september 2014 om 9.45 uur in de aula van de universiteit, De Boelelaan 1105

door

Gregoire Pierre Maroteaux

geboren te Versailles, Frankrijk promotor: prof.dr. M. Verhage copromotoren: dr. R.O. Stiedl dr. S. van der Sluis leescommissie: prof.dr. J.D. Armstrong prof.dr. B.M. Spruijt prof.dr. A.B. Smit dr. M. Kas dr. T. Pattij

TABLE OF CONTENTS

Chapter 1 General introduction 11

Behavioral phenotyping 13 Genetically engineered mice: 15 key to human genetic disease mechanisms Importance of the synapse in psychiatric disorders 18 Aim of this thesis 22

Chapter 2 High throughput phenotyping of avoidance 25 learning in mice discriminates different genotypes and identifies a novel gene

Introduction 28 Material & methods 29 Results 32 Discussion 40 Supplemental 43

Chapter 3 Heterozygous Munc18-1 mice exhibit an increased 57 anxiety-like phenotype but no cognitive impairment

Introduction 60 Material & methods 60 Results 62 Discussion 68 Conclusions 71 Supplemental 72

Chapter 4 Functional characterization of the PCLO S4814A 89 variant associated with major depressive disorder reveals cellular but not behavioral differences

Introduction 92 Results 93 Discussion 106 Material & methods 107 Supplemental 114 Chapter 5 Random mutagenesis by transposon-based gene 125 trap insertion in mice identifies a role of Ubn1 in avoidance learning and fear conditioning

Introduction 128 Material and methods 129 Results 133 Discussion 139 Supplemental 143

Chapter 6 General discussion 145

General aim 147 Limitations of standalone tests 148 Home cage relevance 149 The validation of new paradigms 152 Limitations of a new paradigm 154 Accuracy of animal models 155 Future directions 156 Final remarks 157

Chapter 7 Material & Methods 159

Laboratory animals 161 The test battery 161 Automated home cage observation and data analyses 169 Statistical analysis 170

References 173 Abbrevations 187 Samenvatting 189 Summary 193 Acknowledgment 195

Chapter General introduction 1

1 General introduction 13 In this chapter, the reasons for the development of a new approach of behavioral The last decades were marked by the expansion of new technologies in genetics, brain n need of a new behavior approach phenotyping are introduced, followed by the relevance of investigating behavior in mice andcomparing different genetic backgrounds or targeted mutations as well as inrandom mutation in different genes. Thereby the focus will be directedtowards the importancegenes of in neurologic presynaptic and psychiatric disorders. I For decades, scientists strove to develop a variety of appropriate behavioralphenotyping. tests for The rodent most commonly used behavioralmeasures in testsacute tests of are short basedduration, normally on 5 min locomotorto less examplesfrequentlylike activityanxiety assessment 60 in an open min.field (OF), anClassical elevated plus maze (EPM), or a dark- light box (DLB) (Crawley and Goodwin, 1980; timereaction5-choiceserial Hallthe and(BM) task and mazeBarnes theperformance inL,as such behavior, 1932; Pellow et al., 1985) or learned task (5-CSRTT) (Barnes, 1979; Robbins, 2002). To describe the effect of a genetic background, a mutation or a drug on behavior, a battery of tests of is behaviorrequired to suchtap into as different motor, aspects sensory, cognitive like and the circadianSHIRPA protocol functions.(Rogers et Such al., test 1997) batteries,or IMPReSS Brown protocol et (Brown al., and 2005),Moore, are 2012; designed to cover muscle, functions.cerebellar, Those sets of experimental sensory procedures contain anddifferent tests across severalneuropsychiatric screens of increasing complexity and specialization (Nolan standardized et protocols to improve al., comparability and 2000).reproducibility across studies. These However, test batteries the have succession of tests in those batteries requires repeated human-animal interaction. Carry- over effects can be caused by the order in whichtests are performedanimals. on Moreover,the sameenvironmental cohort effects,of the lack of consistency in the protocols between imaging and molecular biology, to increase the precision, the wealth and of the newlyreproducibility generated results. However, behavioral phenotyping – due notto its yet complexity experienced– has this revolution and throughput still (Crabbe et al., 1999). is hampered by poor reproducibility and low Behavioral phenotyping In 1839, Darwin described the behavioral phenotype of 12 species of finchesof Islands. the He demonstrated Galapagos the correlation between feeding behavior, morphology of the beak, and the environment of finches(Darwin, 1839) exemplifying the impact of genetic radiation on morphological differences underlying behavioral adaptation even before of the the genetic understanding basis (Mendel, 1865). Behavioral phenotyping is the description and analysis internally coordinated responses (actions of or inactions) of whole living organisms (individuals or groups) to internal and/or external stimuli the (Levitis with et interaction al., and physiology 2009). The expression, gene behavioral on response dependent of process an complex a is organism predictive be can responses observed normally from deviations behavioral Distinct environment. signs for neurological pathologies and psychiatric disorders in human. Thus, the understanding behavior of complexity The essential. is behaviors those in involved mechanisms biological the of diseases. psychiatric and neurological human mimic to models animal of use the requires laboratories, discrepancies in the data collection methods (visual scoring vs. computerized) and operational definition (e.g. definitions of grooming, rearing, lingering) are also sources of variation in the results that lead to different interpretations (Crabbe et al., 1999; Turner and Burne, 2013; Wahlsten et al., 2003; Würbel, 2002). Considering the huge number of variables that directly or indirectly affect behavioral testing, it is no surprise that behavioral testing is a bottleneck in screening for genes involved in neurological disorders and the development of novel drug treatments. Consequently, there is growing consensus that behavioral testing needs a boost toward automation to increase its effectiveness and standardization in the study of animal models of human diseases and therapeutic developments (Crabbe and Morris, 2004; Smith and Eppig, 2012; Tecott and Nestler, 2004).

New technology: home cage recording With the development of advanced video-tracking software, computer power and data storage capacity, new behavioral technologies have emerged, allowing automated observation of animals in their home cage over long periods of time (de Visser et al., 2006; Goulding et al., 2008; Jhuang et al., 2010). Automation of observation allows repetitive, objective, and consistent measurement over days or even weeks, rather than minutes or hours. This substantially increases the scientific efficiency with respect to the researcher’s investment of time and effort. Continuous tracking has increased the behavioral throughput and decreased confounding factors like handling, and the frequent adjustment to novel environments that is required in batteries of classical behavioral tests. Furthermore, continuous recording allows investigation of multi-dimensional aspects of behavior like habituation, baseline and challenged behavior (de Mooij-van Malsen et al., 2009; de Visser et al., 2006; Kas et al., 2008), offering the possibility to study longitudinally the progression of behavior, which is of utmost relevance to study models of neurological, psychiatric, and neurodegenerative disorders. These disorders show increasing prevalence in our society (Wittchen, 2012), altering everyday life. Their first symptoms are identifiable in humans through subtle or drastic changes in day to day behavior (e.g: food intake, sleep and activity disorder) (Dziegielewski, 2010). Long-term tracking improves the resolution of behavioral observations enabling to detect subtle changes in behavior over time which is not possible in standalone tests as they display reaction to a novel environment. In the last decade, new approaches to study mouse behavior were developed. Commercial software packages, like EthoVision, are able to track movement patterns (Spink et al., 2001). Learning software is in development to refine the quality and the resolution of the analysis of the observe behavior (Jhuang et al., 2010). Furthermore some studies reveal characteristic features like rest and awake bouts, grooming, rearing and jumping, of animal models for disease (Steele et al., 2007). In the IntelliCage, implanted radio frequency identification (RFID) chips are used together with RFID readers in strategic locations in the cage (doors, corners) to provide information on localization of a specific individual mouse in the cage. These techniques are used to assess learning ability of place discrimination as well as behavioral reward acquisition (Endo et al., 2011; Mechan et al., 2009). Finally, photo-beam, lick-o-meter and weight platform in a home cage are used to detect differences in eating,

14 1 General introduction 15

enetically engineered mice: ouse inbred strains key to human genetic disease mechanisms Scientists have always used a large questions diversity (http://www.nih.gov/science/models/). of Yeast, organisms worms, as and modelsflies became to modelsstudytothecell cycle, developmental excellentprocesses, answer andgenetics. investigateHowever,to biological thediversity and complexity ofpsychiatric and neurologic disorders, scientists need organisms closer to humans, sharing the basic brain structures and complex allowing behavioral the functions.investigation Startingof more in the late 70s, modelsmicehumangeneticistsgeneticfor diseases using spontaneous mutationsinbredmice(Schlager in generated a collection of andDickie, 1967).Quickly, themouse became thefavorite animal model inbiomedical research mainly driven by economic factors. Mice have a short theirgeneration small size facilitatestime their housing.and Additionally, theya have large geneticshort variations amonglifespan, inbredstrains providing anexcellent basis studyto complex genetic interactions. Furthermore, like humans, mice naturally develop diseases (cancer, hypertension, diabetes) affectingphysiology theirbehavior.sequencingandtheMoreover, humantheofthenandmousethe genome showed that 99% of the protein-coding genes are shared Genome betweenSequencing the Consortium two et speciesal., 2002). (MouseTherefore, molecular and genetic researchersdeveloped tools to manipulate the mouse genome in a temporal and spatial manner, allowing to study genes or groups of genes involved in human behavior and diseases (Smith and Eppig, 2012).Thousands ofinbred, recombinant, chromosome-substitution, transgenic, homozygous or heterozygous, conditional, and inducible mutant strains humanare now diseases available in to mouse investigate models (e.g. see Arozena MGI et al., 2008; Loos http://www.informatics.jax.org) et al., 2012; Roberts et al., 2007; Verhage et (Acevedo- al., 2000). M In 1902, William Castle recognized the value using mouse of inbred strains homozygous (Castle, 1919). This mice study is and considered to studiedrepresent the beginning inheritance of modern mouse genetics (Crawley, 2007). After 20 consecutive brother and sister mating, a strain is isogenic (genetically identical), as 98.6% of the loci, in each mouse, are homozygous(Strachan and Read, 1999). Inbred strains have proven their value in genetic and immunologyfor their isogenicity within strain and their heterogeneity1976). Historically, inbredbetween strains were inbredbred to selectstrains and increase(Mitchell, specific phenotypes. Those G drinking and activity states in different mutant strains(Goulding et al., 2008). However, these new automated methodologies, which may even include experimentalthorough validation.manipulations, Combining needautomated home cage recording data with an extensive test batteryallows comparison between the new method and the current gold standards based on classical tests. Eventually, these techniques will improve the monitoring ofhuman neurologicalmouse diseases,model followingby for thedevelopment symptomsof chronicof disorders like depression, epilepsy and pathological anxiety, and progressive diseases like Huntington,Parkinson and Alzheimer (Balci et al., 2013; Steele et al., 2007). characteristics are now used to study the genetics underlying those phenotypes. For instance, AKR/J, DBA2/J and C57BL/6J mice are more prone than other strains to diet-induced obesity (Alexander et al., 2006; Collins et al., 2004). C57BL/6J mice are also known to have an increased susceptibility for substance abuse (alcohol and morphine) and thus constitute an interesting model for the genetics of addiction (Belknap et al., 1989; Peirce et al., 1998). Different inbred strains are often compared to investigate genetic influence on specific behavior like avoidance learning (Buselmaier et al., 1981). However, phenotypic differences between inbred strains need to be taken in account while designing experiments, specific traits may influence the outcome of an experiment. For example, C3H mice are known to develop retinal degeneration, thus are inappropriate for any experiments requiring vision (Wong and Brown, 2006). Therefore, characterizing inbred strains in a new automated home cage and avoidance learning experiment is essential to estimate potential strain-related differences (e.g. locomotor activity) before assessing the phenotypic consequences of a mutation in a certain genetic background (Martin and Fisher, 1997). Additionally, the homogeneity of their genome makes inbred strains suitable to study specific genes using gene targeting.

Targeting specific genes involved in biological processes Genetic manipulation in mice has shown to be a crucial impetus for the study of specific genes involved in physiology and behavior. In 1981, Gordon and Ruddle successfully inserted DNA in a mice genome, creating the first transgenic mouse (Gordon and Ruddle, 1981). Transgenic methods were followed by targeted gene mutation, using homologous recombination. Targeted mutations (or gene knockout techniques) have revolutionized biomedical research. However, two different genetic processes which are generally overlooked may lead to misinterpretations of the results: epistasis and linkage interactions. (i) Epistasis occurs due to the interaction between the induced mutation and unlinked loci of the genetic background leading to differences in phenotypic consequences depending on the genetic background in which a mutation occurs. (ii) The flanking genes problem occurs when the observed effect of a closely linked gene is attributed to the induced mutation, causing false interpretation (Crusio, 2004). The flanking gene refers to the region, adjacent to the targeted gene, belonging to the genetic background of the embryonic stem cell (ES) (usually derived from 129S1 mice) inserted with the construct of interest. Even after backcrossing, flanking genes remain linked to the targeted locus, it is called genetic linkage, which can mislead the interpretation of the results (Bolivar et al., 2001; Crusio, 2004; Gerlai, 1996; Wolfer et al., 2002). For instance, a study on the isoform-specific p97FE65 (transcriptional activator in a nuclear signaling pathway) knockout mice revealed that the phenotype observed in hippocampus-dependent learning and memory was in fact due to the neighboring gene Gprc5b (retinoic acid-inducible orphan G protein– coupled receptor) mutation present in the ES cell used to create the targeted mutation, and not due to the deletion of p97FE65 itself (Cool et al., 2010). There are more examples showing that flanking genes influenced the behavioral phenotype rather than the target gene (Bolivar et al., 2001) and even if after backcrossing the strain repeatedly for 12 generations to obtain a congenic strain, the remaining genetic contribution of the donor strain will be low (~1%), but the probability that flanking gene will show an effect is still present (Crusio, 2004).

16 1 General introduction 17 Cre Cre B

Gene X

2 Gene X 2 2 Cre deletion Gene X 1 1 Cre/Loxp system. (A) Cre/Lox recombination is a genetic tool that enables site-specificrecombination. 1 1 Also, Also, genes essential for development or survival, or ubiquitous genes with different roles igure 1: It allowsItinserting, deleting, inverting translocatingand specificany portion (Sauer DNA Henderson,and of 1988), using the enzyme Cre recombinase, which is able to recognize, catalyze and recombineThis techniqueshort requiresLoxP a sequences.sequence of interest frame between two LoxP sites, and the Cre under a ubiquitous or cell-specific promoter (forrespectively general or tissue-specific recombination), or an injectablecarrying viralthe Cre particlerecombinase (Nagy, 2000; Orban et al., 1992). LoxP in the same direction, the Cre recombinase will excise the Floxed gene. (B) Crossing a mouse with a tissuefloxed will generesult in andoffspring a with mouse the floxed expressing gene deleted Cre in in the a specific specific tissue wheremodified Cre after http://cre.jax.org/introduction.html).is expressed.(scheme F in different tissues, have been hard to study using targeted gene mutations as theyleadto wouldgene mutations havetobeen hardgene targeted usingtissues, study in different targeted for identified limitations the overcome to Therefore, phenotypes. unspecific or lethal a spatial lethality) enabling embryonic technique, avoid recombination to Cre/lox the e.g. developed animal, biologists the molecular of mutations, life the of time any (at temporal and type-specific) (cell gene flanking the avoiding while 2007) Kieffer, and (Gavériaux-Ruff expression gene of gene flanking modification the of problem the resolves system Cre/lox the using gene a Deleting 1). (Fig. problem it Therefore, congenic. remains strain the same, the are parents both of background genetic the as flanking possible identify to Cre/lox, with mutation gene targeted classical compare to interesting is However, specifically. genes of hundreds study to enable mutations gene Target interference. gene genes, predominant of number asmall on focused community scientific the techniques, those with unstudied. genes of wealth a leaving phenotypes, clinical/disease by discovered were which A Random gene mutation using insertional mutagenesis: Sleeping Beauty transposon The Sleeping Beauty (SB) transposon is the most studied cut-and-paste transposon in vertebrates (Geurts et al., 2006; Ivics et al., 1997; Keng et al., 2005; Takeda et al., 2008). It is frequently used for insertional mutagenesis to generate new mouse germ lines (Miskey et al., 2005). Derived from the inactive Tc1/mariner transposable elements, it is used for general gene transfer and insertional mutagenesis. The transposition process requires two parts of the SB system: an insertional mutagen called the transposon (DNA vector) and the enzyme to mobilize the transposon: transposase. In order for the transposase to catalyze the transposition, the transposon has to be flanked by two specific sequences called inverted repeat/direct repeat (IR/DRs). These IR/DRs are binding sites for the transposase (Largaespada, 2009). The SB system requires the mating of one transgenic mouse line carrying the transposon, and a second one carrying a transposase (Dupuy et al., 2005). Multiple copies of transposons aggregate, creating a concatemer in the donor site. The generated offspring is a double transgenic male mouse carrying the transposon and the transposase. These mice are referred as seeder males as they are the source of sperm carrying the new SB transposon vector insertion. The next step toward a randomly mutated mouse is to mate the seeder male with a wild-type female, it will result in a generation 1 (G1) offspring carrying a new transposon insertion. The insertion of the SB transposon is essentially random. However, it has a tendency to insert itself in a sequence directly adjacent to TA dinucleotides (Ivics et al., 1997), and a higher frequency to insert close to the donor site, on the same chromosome. This phenomenon is called local hopping (Carlson et al., 2003). SB vectors include a 5’ gene trap sequence with a splice acceptor and a poladenylation, allowing the disruption of a gene even when inserted into an intron. A reporter gene (e.g. green fluorescent protein, GFP) will be expressed only if the transposon has landed in the same orientation as the transcription unit, enabling the identification of the G1 mouse holding a new transposon insertion in a gene with a simple visual inspection of the GFP expression just after birth (Horie et al., 2003). Finally, a linker mediated PCR (LM-PCR) is used to locate the landing site of the transposon (Fig. 2) (Largaespada, 2009). Combined with advanced automated phenotyping, this recently developed technique in mice is promising to identify genes involved in home cage behavior and avoidance learning. Since the sequencing of the genome, molecular tools became available to answer the question which gene, group of neurons, nucleus or brain structure is involved in specific behavioral functions. Genetic manipulation of animal models is a pillar of research on multifactorial psychiatric and neurological diseases by providing ample possibilities to induce genetic dysfunctions in an animal to model human disease phenotypes.

Importance of the synapse in psychiatric disorders The brain is the central control unit of the nervous system; behavior, cognition, and emotions all originate in the brain. Any dysfunction in this complex organ can result in neurological disorders including psychiatric disorders. The synaptic junction is the communication site between individual the neurons of the brain and has been the center of interest in neuroscience for decades. Yet, the molecular and cellular mechanisms involved in synaptic formation,

18 1 General introduction 19 E4 Breed to homozygocity E3 WT Splicing ranscription T ranslation T Breed out transposon & transposase Seeder mouse G1 E2 Insertionalmutagenesis. (A) Schematic representation of the SB gene trap transposon vector design by ransposase runcated gene rap vector T T T E1 rap vector T Endogenous gene igure2: A B F etal.,2008) (Takeda built Takeda withmain2parts. The trap firstcomposedis one, the splice 5’gene a of acceptor (SA), internal ribosome entry site (IRES), reporter transgene (lacZ) and a polyadenylation (pA) site. When landed in a gene, this part will be expressed insteadcalled theof poly-A thetrap unit,endogenous composed of gene.an internalThe promoter second(P), a reporterpart (green fluorescentof splice protein)thedonor and(SD) transposon a but no is polyadenylation site. The poly-A trap will be expressedlands inubiquitously the same if orientationthe astransposon the transcription unit. This allows for fast screening of thehaving G1 a mice,disrupted as genesthe expressones the poly-A trap (GFP). (B) Steps to generate a mouse line with a SB transposon insertion. Crossing a transgenic mouse carrying the transposon vector with another transgenic mouse carryinga transposase vector results in the seeder male generation, carrying bothwild-type withaoffspring. obtain ubiquitousbred malesG1 the femalescreen thefor are to mice are expression transposon G1 and transposase. Seeder of GFP if the transposon landed in a transcript unit (Largaespada, 2009). Then the mouse is bred to get rid of the donor site and the transposase. Finally, the mouse is bred to get obtain homozygous mice. stability, and neurotransmitter and neuromodulator release functions are still not completely understood. Since some proteins are essential to maintain synaptic transmission and plasticity, it is not surprising that mutations in coding-genes are linked to neurodevelopmental disorders and neurodegenerative diseases (Waites and Garner, 2011). The machinery of the presynapse involved in releasing synaptic vesicles containing neurotransmitters requires a fine coordination between the proteins present on the surface of the vesicle, in the cytosol of the neuron (e.g., MUNC18-1), and on the active zone of the synaptic bouton (e.g. PCLO). I showed a particular interest in the genes coding for MUNC18-1 and PCLO proteins in this thesis. MUNC18-1 has been intensively studied in-vitro during in the past 15 years for its crucial role in vesicle release. Munc18-1 was recently identified as a potential actor in epilepsy and mental retardation (Hamdan et al., 2011; Saitsu et al., 2010). However, no studies were made on the available heterozygous mouse model. Pclo was recently identified as potential candidate in major depressive disorder (MDD) in a genome-wide association study (GWAS) (Sullivan et al., 2009).

MUNC18-1 - a key protein in neurotransmission The release of neurotransmitters is the result of a highly specialized mechanism involving many proteins (Südhof, 2004). The complete deletion of certain pre-synaptic genes, like Munc18-1, results in lethal postnatal phenotype (Fernández-Chacón and Südhof, 1999; Verhage et al., 2000). Munc18-1 was identified as an essential gene involved in neurotransmitter release (Verhage et al., 2000), and is part of the neuron-specific SEC1-family of membrane trafficking proteins (Hata and Südhof, 1995). Munc18-1 allows the appropriate location of the syntaxin-1 protein at the presynaptic bouton (Arunachalam et al., 2008; Malintan et al., 2009; McEwen and Kaplan, 2008; Medine et al., 2007; Rowe et al., 1999) and controls SNARE-complex formation enabling the fusion of the vesicle to the presynaptic membrane (Rizo and Südhof, 2002; Südhof and Rothman, 2009; Toonen and Verhage, 2007)). Nineteen different de novo heterozygous mutations in humans have been linked to severe intellectual deficits without epilepsy (Hamdan et al., 2009) as well as early infantile epileptic encephalopathy (EIEE) also known as Ohtahara syndrome (Saitsu et al., 2008), suggesting Munc18-1 haploinsufficiency as underlying cause of those neurologic disorders (Fig. 3, generously shared by Dr. Ruud F. Toonen). The consequences of Munc18-1 mutations have been intensively studied in vitro demonstrating its essential contribution to vesicle release mechanisms (Meijer et al., 2012; Rizo and Südhof, 2002; Südhof, 2012; Toonen and Verhage, 2007; Verhage et al., 2000). Yet, no behavioral phenotyping of Munc18-1 heterozygous mice is available, although heterozygous Munc18-1 mice serve as potential model for the human disorders.

PCLO, a pillar of the active zone, implicated in major depressive disorder Identifying the genetic basis of complex traits is of utmost importance to characterize psychiatric and neurological disorders. The possibility to compare individuals with and without a certain disease on the basis of single nucleotide polymorphisms (SNPs) opened new opportunities to investigate genetic involvement of disorders. GWAS involve the identification of common genetic variants associated to specific traits (Manolio, 2010). A recent GWAS highlighted the complexity of psychiatric disorders by quantifying the contribution of common SNPs in 5 major psychiatric

20 1 General introduction 21 Pclo gene Exon 16 Exon 20 PCLO c. 1434G>A p.W478X 4 Exon 19 Exon Exon 14 Exon c. 1720A>C p.T574P c. 1162C>T p.R388X 5 20 3 mutations in the 592 (Hamdan et al., 2009), 3 de novo Exon 18 Exon Exon 14 Exon c. 1654T>C p.C552R 7 c. 1206delT p.402X 6 Exon 18 Exon Exon 11 Exon c. 1631G>A p.G544D 1 c. 961A>T p.K321X 2 BP1 (isoform a ((top, GenBank accession (Fejtova and Gundelfinger, 2006; X α (Saitsu et al., 2010), 2 R406H P480L G544D C552R T574P Exon 16 Exon (Otsuka et al., 2010). 7 c. 1439C>T p.P480L 5 Exon 9 Exon c. 747dupT p.Q250SfsX6 2 245 360 480 Exon 8-14 Exon c. Del exon8-14 5 gene, might influence the function of the protein. Exon 15 Exon in MDD. Taking the fact that 82% of the mouse the of 82% thatfact the Taking MDD. in Exon 9 Exon c. 1328T>G p.M443R 1 c. 703C>T p.R235X 2 (Saitsu et al., 2008), 1 was more frequentMDDpatientsinwasmore (Minelli etal.,2012) PCLO PCLO domain2 domain3a domain3b domain2 C180Y L183R Exon 14 Exon (Hamdan et al., 2011), mutations associated with EIEE/EOEE, West syndrome and Mental mutations and deletions in ST c. 1217G>A p.R406H 2 6 PCLO 134 Exon 6 Exon c. 388_389delCT p.L130DfsX11 2 de novo de novo Exon 7 Exon c. 548T>G p.L183R 5 BP1 X SD Intron3 (Milh et al., 2011), 5 c. 169+1G>A p.57X 3 V48D domain 1 Exon 7 Exon c. 539G>A p.C180Y 1 Exon 3 Exon c. 902+1G>A p.Q301fsX1 5 Exon 3 Exon 11 2 3 2 4 3 5 6 4 7 5 8 6 9 7 10 8 11 12 13 9 14 15 10 16 11 17 12 13 18 14 15 19 16 17 18 c. 157G>T p.E53X 2 Overview of ST 4 Exon 5 Exon c. 251T>A p.V84D 1 BP1 consisting of at most 20 exons (untranslated region and coding region are open and filledrectangles, BP1 protein show no clear spatial clustering.

X X

missense trunctations & deletions & trunctations igure 3: B A (Deprez et al., 2010), respectively). The locations of mutations are indicated by arrows. (B) The positions of retardation. (A) Summary of the number, NM003165)) and b ST (bottom, NM_001032221)) associated with EIEE. Schematic representation ST of (Sullivan et al., 2009) suggesting that the exonic variant, carrying an alanine instead of a serine inthe zinc-finger coding domain of the PICCOLO (PCLO) is localized at the synaptic bouton in the active zone where docking, and recyclingfusion take place. The active zone is located on the plasma membrane of a synapse and is composed of a lattice of proteins forming the cytomatrix of the active zone (CAZ), (Fenster etal., 2000). atThere are least 7 major components of the CAZ, MUNC13, RIM (Rab3-interacting molecule), BASSOON, PCLO/ACZONIN, ELKS and LIPRINS- 4 diseases (schizophrenia, bipolar major disorder, depressive autism disorder, spectrum disorder, and attention deficit and hyperactivityGenomics disorder) Consortium (Cross-Disorder et al., Group2013). In ofthe last years, the 7 GWAS Psychiatric et al.,for 2010; MDD Major Depressivewere Disorder publishedWorking Group (Lewisof the Psychiatric GWAS Consortium et al., 2013; Muglia et al., 2010; Shi et al., 2011; Shyn et al.,2012). However, none of 2011; the studies hasSullivan identified variantset that metgenome-wide al., significance. 2009; Wray et al., Yet, out of the top 200 SNPs, 11 were localized in a 167-kb region overlapping the F Schoch and Gundelfinger, 2006; Siddiqui and Craig, 2011). A replicate replicate case-control showed study A 2011). Craig, Siddiqui and 2006; Gundelfinger, Schochand thatthenucleotide substitution in possiblesupportinginvolvementof the gene is homologous to the human one, the creation of a knock-in mouse by replacing Serine in position 4742 in the C2A domain with an Alanine seemed the appropriate approach to study the behavioral effects of this non-synonymous SNP variation (Fenster and Garner, 2002) in a mouse model. Depression-like behavior in mice is commonly studied using assays like the forced swim test. Thus, the investigation of the PcloSA/SA mouse is necessary to provide evidence whether the Pclo amino acid substitution is mechanistically involved in MDD.

Aim of this thesis The goal of the 4 years of research in this PhD project was to contribute to the improvement of the characterization of mouse models for psychiatric and neurologic disorders. This project describes the effects of genetic background and specific or random gene mutations on complex behavioral traits in mice. For that purpose, I first focused on the description and analysis method of automated home cage tracking, enabling 7-day continuous behavior recording, thereby reducing human-animal interferences to a minimum. Thereafter, I characterized mouse models for neurologic and psychiatric disorder, and finally I studied the effects of random gene-trap mutation on spontaneous behavior and cognitive functions in 5 mutant mouse lines. The aim of Chapter 2 was to characterize a new approach of avoidance learning task taking place in an unsupervised automated home cage using the complex behavioral response of 8 of the most commonly used inbred strains. The home cage contained a shelter with 2 entrances. The mice developed a pattern of visits in the shelter and a preference for one of the two entrances. This preference was then challenged using a mild aversive stimulus. We observed striking strain differences in the complex response involving cognitive function and behavioral flexibility as well as unexpected behavioral responses, like hyperactivity in strains known to be hypoactive in classical test. Finally, as proof-of-principle for fast gene function screening, 43 different genetically modified strains were compared with respect to their avoidance learning. One gene, Specc1/cytospinB, turned out to be involved in the temporal response to the aversive stimulus. Behavioral phenotyping of mice lacking one copy of Munc18-1 is central to Chapter 3. We studied the possible involvement of Munc18-1 de novo heterozygous mutations in early infantile epileptic encephalopathy (EIEE) and intellectual deficits without epilepsy. To confirm the involvement of Munc18-1 in cognition and epilepsy, we performed an extensive battery of tests concerning home cage behavior and a battery of classical behavior test. These tests were run on classical recombinant heterozygous and Cre/loxP mutants. We observed a mild phenotype compared to human phenotype; the mutant mice showed no tremors, and cognitive functions were not affected. Instead, Munc18-1 heterozygous mice displayed increased anxiety levels and an active stress-coping style. To study potential mechanisms contributing to human complex psychiatric diseases, in Chapter 4 we focused on the large multi-domain protein PCLO which is involved in the synaptic active zone assembly. Several GWAS indicated a non-synonymous exonic variant of PCLO as an actor in MDD. After generating the knock-in mouse model PcloSA/SA, we characterized the behavioral phenotype of this homozygous mouse. No differences in spontaneous home cage behavior, anxiety, cognition and depression-like behaviors were observed between the knock-in mice and their littermates, suggesting that the PcloSA/SA variant may be involved but not sufficient to induce symptoms of MDD in C57BL/6J mice.

22 1 General introduction 23 illustrate the power of the combination of classical behavioral hapter 5 C summarizes the major results of 4 years of work on behavioralphenotypingon work ofyears 4 ofresults majorsummarizes the gene, coding for a nuclear protein and essential in the HIRA/UBN1/CABIN1/ASF1a hapter 6 hapter C Ubn1 The finalresults of Finally, tests and home cage recordings. Using the transposon technique, we generatedmutated 5 randomly mouse strains, including 5 differentgenes. mutations Each strain was screened in in the automated functional, home cage protocol for previously avoidance learning unstudied deficit. complex, was found to be associated with avoidance learning deficits in mice. UBN1 deficits in mice also reduced anxiety-like behavior and disrupted cued (auditory) fear conditioning. discussesand relevancetheimproving of behavioral functionaltestsfuturetheforof genomics, here specifically the functional characterization ofbehavioral the functions and dysfunctions. role of individual genes for specific

Chapter

High throughput phenotyping of avoidance learning in mice discriminates different genotypes and identifies a novel gene 2 Grégoire Maroteaux*1, Maarten Loos*2,4, Sophie van der Sluis1,3, Bastijn Koopmans4, Emmeke Aarts1, Koen van Gassen1, Aron Geurts5, The NeuroBSIK Mouse Phenomics Consortium6, David A. Largaespada5, Berry M. Spruijt7, Oliver Stiedl1,2, August B. Smit2* and Matthijs Verhage1,3*

1Department of Functional Genomics, 2Department of Molecular and Cellular Neurobiology and 3Department of Clinical Genetics; Center for Neurogenomics and Cognitive Research (CNCR), Neuroscience Campus Amsterdam; VU University Amsterdam and VU Medical Center, 1081 HV, Amsterdam, The Netherlands 4Sylics (Synaptologics BV), Burmanstraat 7, 1091 SG, Amsterdam, The Netherlands 5Department of Genetics, Cell Biology and Development, The Cancer Center, Minnesota, University of Minnesota, USA 6The NeuroBSIK Mouse Phenomics consortium collaborators are: Brussaard AB, Borst JGG, Elgersma Y, Galjart N, van der Horst GT, Levelt CN, Pennartz CMA, Smit AB, Spruijt BM, Verhage M, de Zeeuw CI. 7Department of Biology, Faculty of Beta Sciences, University Utrecht, Utrecht, The Netherlands

*these authors made equal contributions

Published in Genes Brain Behav. 2012 Oct;11(7):772-84

26 High throughput phenotyping of avoidance learning in mice discriminates different genotypes and identifies a novel gene 2 27 . This mutant specc1/cytospinB is involved in this aspect of cognition. specc1/cytospinB : FDR: False Discovery Rate; GEE: generalized estimating equations : avoidance learning, behavioral flexibility, home cage, high-throughput behavioral stract showed profound and specific delay in avoidance learning.Together, these data suggest that different genotypes express distinct learning and/or memory of associations between shelter entranceand aversive stimuli, and that Ab Recognizing and avoiding aversive situations are These central abilities aspects are of essential mammalian for cognition. health andgenetic basis. modeledWe these abilities survivaleightin common mouse inbred strains covering ~75% and are expected to have of the a species’ prominent natural variation and in gene-trap mice (>2000 mice), automated using an assay unsupervised, with an instrumented home cage two entrances. (PhenoTyper) Mice containingvisited this a shelter shelter 20-1200 withtimes/24 h and significant entrance. preference strong aversiveone often mild Subsequently,71%for and stimulusa of all mice developed a (shelter illumination) was automatically delivered when mice used Differenttheir genotypes preferred developedentrance. different coping strategies. Firstly, the numberthe ofpreferred entriesentrance decreased via in DBA/2J, C57BL/6J and 129S1/SvImJ, indicating that these genotypes associated one specific entrance with the aversive stimulus. Secondly, mice started sleeping outside (C57BL/6J, DBA/2J), indicating they associated the shelter in general with the aversive stimulus. Some mice showed no evidence for an and associationthe aversivebetween light, the but entrancedid show markedly shorter shelter residence illumination,times inindicating response they to did perceive illumination as aversive. Finally, using this screenedassay, we 43 different mutants, which yielded a novel gene, Abbreviations Keywords screening, common inbred, SPECC1 Introduction Recognizing and coping with aversive situations are essential abilities for health and survival in mammals. Successful avoidance strategies require distinct cognitive abilities: learning and remembering stimuli that predict aversive situations or harm, and also the flexibility to suppress or alter routine- or preferred behaviors. Aberrancies in those behaviors are core symptoms of many affective disorders (Dziegielewski, 2010) and are associated with many psychiatric disorders such as schizophrenia ((Szöke et al., 2005), for a meta-analysis: (Lesh et al., 2011)) and ADHD (for a meta-analysis: (Willcutt et al., 2005)), and with substance (ab)use (Kreek et al., 2005; Wills et al., 1994). Twin studies indicate that variation in harm avoidance and behavioral flexibility have a prominent genetic basis, with heritability estimates of 39-50%, (Finkel and McGue, 1997; Gagne and Saudino, 2010; Groot et al., 2004; Polderman et al., 2009; Stins et al., 2004; Yamagata et al., 2005). The long allele of the serotonin transporter gene has been implicated as a causal variant in harm avoidance, but evidence for non-involvement is currently prevailing (Munafò et al., 2009; Schinka, 2005; Wray et al., 2009). Genome-wide association studies have not yet produced convincing alternatives (Verweij et al., 2010). Hence, to date we know little about which genetic variation is involved in these avoidance traits. To unravel the mechanisms that drive avoidance behavior and to identify potential genetic factors that modulate these, mice are a suitable resource. Classical behavioral tests, such as passive avoidance (Baarendse et al., 2008; Crawley et al., 1997; McGaugh, 1966), have been used to describe avoidance responses to an electric foot shock. Different studies in avoidance responses suggest that natural genetic variation substantially influences this type of behavior (Crawley et al., 1997). In addition, mice can also establish multiple, distinct associations with an aversive stimulus (as for instance in a fear conditioning (Stiedl et al., 1999)). Using such paradigms, several features of avoidance learning and behavioral flexibility have been elucidated. However, classical stand-alone tests typically employ rather intense learning stimuli (i.e., electric shock) or require animals to be motivated by food deprivation. Unspecific stress levels are further enhanced by animal handling when placing mice in the test system, thereby potentially impacting on test results (Hurst and West, 2010). One prominent approach to avoid these limitations is to test animals in their home cage environment without human interference. Moreover, we expect that in a home cage environment even subtle stimuli are salient enough to drive avoidance learning. Innovations in automated detection of rodent behavior allow studying various aspects of spontaneous behavior (de Visser et al., 2006, 2005; Goulding et al., 2008; Jhuang et al., 2010; Voikar et al., 2010), and have paved the way for systematic (high-throughput) analysis of mouse behavior, also tapping in on the fast expansion of available genetically modified mice. Such automated, unbiased approaches to natural behaviors can also be applied to study cognitive traits, like learning and memory, but novel methods are required to automate analysis of these traits. Towards this goal, we characterized the complex behavioral response indicative of avoidance learning in mice using an unsupervised, automated high throughput system. We adopted an experimental procedure developed by De Heer et al (de Heer and Spruijt, 2008) in an instrumented home cage, the PhenoTyper, containing a shelter with two entrances. This assay exploits the natural tendency of mice to develop a preference for one of two shelter

28 High throughput phenotyping of avoidance learning in mice discriminates different genotypes and identifies a novel gene 2 29 l μ After After

. l used in a nested nested a in used l ad libitum μ M of each primer IR/DR(R) IR/DR(R) primer each of M μ 2 , involved in avoidance learning. , , and 1 unit of platinum polymerase (Invitrogen). The Taq 2 specc1/cytospinB cages for six and a half days. All experimental procedures were approved approved were procedures experimental All days. half a and six for cages ® M M dNTPs, 2 mM MgCl μ aterial & methods leeping beauty transposon mice by the national animal research committee and complied with the European Council Directive. European the with complied and committee research animal national the by S Transposon seeder mice were obtained by crossing transposon concatamerwith GT3A/tTA SB10 mice mice expressing the transposase mice (Geurts had et been back-crossed al., to 2006). C57BL/6J mice Both (Harlan) GT3A/tTA for >12 and Seeder generationsmice SB10 were matedbefore with C57BL/6J crossing.mice to generate mutant offspring. Integration of the transposon was detected by linker-mediated PCR. Two microgram of withNlaIII overnight 37°C,attail cleaned usingQiaquickup Qiagen’s DNA PCRpurification kit and elutedwas digested in 50 µl water. Linker-oligonucleotide mix was prepared by mixing 50 µl linker+ (25 µM), linker-50 (25µl µM), and 1 µl 5 M NaCl, heated in a boiling bath for 5 min and slowly cooled to room temperature. Sequence of the linker oligonucleotides was: linker+: 5’-TAATACGACTCACTATAGGGCTCCGCTTAAGGGACCATG-3’ linker-: 5’Phos-GTCCCTTAAGCGGAG-3’NH Digested Digested genomic DNA (400 ng) and 6 µl linker-oligonucleotide mix were used for overnight ligation with T4 ligation kit (MRC Holland, Amsterdam). The template was amplifiedin a 50 M 8 inbred strains 129S1/SvImJ, A/J, BALB/C, C3H/HeJ, C57BL/6J, DBA/2J, FVB/N and between NOD/LtJ. 7 All and male, 18 aged weeks old, were obtained in house, from housed Harlan under 12-hr (Horst, dark-light cycle The with access Netherlands) to water or and bred food entrances,andapplies automated detection shelterof entrances. Automation cannowbeused to sanction specifically visitsto the preferred entrance(illumination by of applyingthe shelter with bright a light). This mild learning paradigm, aversive which is not confounded stimulus by human interference and/or highly stressful stimuli, specifically addresses cognitive aspects of avoidance behavior and produces a wealth of Using information this on assay, other weaspects screenedof behavior. >2000 mice between and obtained the aversive evidence for stimulus specific and associations theassociations between preferred the stimulus shelterand the shelter entrance,in general. Finally, and as proof of more principlehigh for generalizedthroughput gene function analysis, we screened identified a novela gene, panel of 43 different mutants, and PCR machine was programmed for touchdown PCR at 94°C for 2’, 25 cycles of 94°C for 15”, 60°C for for 60°C 15”, for 94°C of cycles 25 2’, for 94°C at PCR touchdown for programmed was machine PCR 2 and 1:50 diluted was reaction primary The 1’30. for 72°C cycle), per (-0.5°C 30” 0.25 with supplemented except conditions, exact same the under PCR were: oligonucleotides the Sequencesof (0.25 µM). primer nested KJC1linker (0.25 and µM) primary PCR reaction supplemented with primers New long IR/DR(R) (100 mM), linker primer (100 mM), 200 acclimation to the facility for at least one week, each mouse was individually housed in one of the the of one in housed individually was week,mouse each one least at for facility the to acclimation PhenoTyper forty-eight New long IR/DR(R): 5’-GTTATGCTAGATGGCCAGATCTAGCTTGTGG AAGG-3’ linker primer: 5’-GTAATACGACTCACTATAGGGC-3’ IR/DR(R) KJC1: 5’-CCACTGGGAATGTGATGAAAGAAATAAAAGC-3’ linker nested primer: 5’-AGGGCTCCGCTTAAGG GAC-3’

Secondary PCR products were separated by electrophoresis and cleaned by Invisorb spin DNA extraction kit (Invitek) and eluted in 25 µl water. Three microliter was ligated in pGEM-T easy cloning kit (Promega), transformed into DH5α E-coli bacteria, plated on LB-AMP-X-gal plates and incubated overnight at 37°C. Colonies were inoculated and cells were grown overnight at 37°C. Plasmid were isolated and sequenced with BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems) and T7 and SP6 oligonucleotides. Sequence of the oligonucleotides was:

T7 primer: 5’-GTAATACGA CTCACTATAGGGC-3’ SP6 primer: 5’-ATTTAGGTGACACTATAG-3’

New mutants, where the gene trap had landed in an active gene, were bred to loose the transposase and the transposon donor site, which was confirmed by genomic PCR using oligonucleotides Specc1 F and Specc1 R or Specc1 R and IR/DR(R) KJC1, DMSO (10%), BSA (1x), dNTP’s (400 µM) and 1 unit Taq DNA polymerase (New England Biolabs), resulting in respectively a wild-type amplified DNA product of 522 bp, and a mutant product of 632 bp. PCR cycle: 95°C for 5’, 35 cycles of 95°C for 30”, 55°C for 30” 65°C for 1’ and 65°C for 5’. The sequences of the oligonucleotides were:

Specc1 F: 5’- CTGGGTAGCAGAATGTACGTCC-3’ Specc1 R: 5’- GCCTAGTAATGCCCTGATCTTC-3’ IR/DR(R) KJC1: 5’-CCACTGGGAATGTGATGAAAGAAATAAAAGC-3’

Specc1 male mutant mice, bearing a gene trap cassette in the third intron of the Specc1 gene deleting expression of two splice variants of the Specc1 gene, as outlined in Fig 7a.Mice were analyzed in the PhenoTyper at the age of 8 weeks and compared to their wild-type littermates.

Instrumented home cage Activity in the home cage was automatically recorded by video-tracking in specially designed cages (PhenoTyper model 3000, Noldus Information Technology, www.noldus.com/ phenotyper). Each cage contains a top unit with built-in hardware for video-tracking, i.e., an infrared-sensitive video camera. The latter provide constant and even illumination of the cage. An infrared filter placed in front of the camera prevents interference with room illumination. This method allows continuous behavioral recordings in both dark and light periods. EthoVision was used as video tracking and trial control software. For the purpose of the high-throughput screen, Noldus Information Technology developed a special version of the software (EthoVision HTP 2.1.2.0) based on EthoVision XT 4.1 (Noldus et al., 2001). Forty eight PhenoTyper cages were connected in a specially designed computer network. Four cages were connected to a PC running video tracking and trial control software. The 12 data collection PCs were connected to a central PC (running software for experiment design and progress monitoring) and a data storage server (for storage and analysis of track files). The cages (L=30 x W=30 x H=35 cm) were made of transparent Perspex walls with an opaque Perspex floor covered with bedding based

30 High throughput phenotyping of avoidance learning in mice discriminates different genotypes and identifies a novel gene 2 31 analysis software TM was used to create a database to © T 4.1 manual). During 4.1 daysfirst the mouse the On Tcould4 enter shelter. freelythe in X Comparisons between groups were made for multiple variables using a test of equal The preferenceThe indexdeterminedwas follows:as [(number entriesof through preferredthe ata-analysis tatistical analysis (SynaptologicsAmsterdam;www.synaptologics.com). shelterdetectionBV, visit the of bin Time was selected to be 5min unless stated otherwise. MatLab on cellulose. A feeding station and a water bottle were attached on to 2 adjacent walls outside of the cage. A shelter (height: 10 cm, diameter: 9 cm; non-transparent one material) of was fixedthe in corners. Two bright white LED, producingwalls, shining inwardsno through the Perspexheat, wall into the shelterwere from two anglesmounted to provide the on two cage aversive light stimulus. In EthoVision, the shelter was defined as a ‘hiddenzone’; the program distinguished the parameters ‘in shelter’ and ‘on shelter’.a rate Video-trackingof 15was samples/second.performed The at following parameters were entries in usedeach entrancefor of the analysis:shelter, the distance frequencymoved, and theof time spent in the shelter, in total for the 6.5 days (for more information on the algorithms used by the program see the Ethovision day 4, the preferred entrance was defined by the system as the most used entrance. On days 5 and 6, each time the mouse used its preferred entrance, bright light (500 lx) was automatically switched on in the shelter as long as the mouse through stayedeither inside.entrance Asthe lightsoon was as turnedit off.left Thethe lightshelter didentered notthrough turn onthe when thenon-preferred mouse entrance. The shelter mouse movedlight into a 3 cm area around theflashed preferred entrance of the onceshelter area on days 5 andwhenever the 6 to signal the aversive learning trial. D Raw track files from the PhenoTyper were processed using AHCODA proportion. For a given level of analysis, statistical analysis was based on estimated FDR entrance) – (number of entries through non-preferred entrance)] / (total number of entries). Theaversion index was defined as follows: [(time spent in the illuminated shelter after entering through the sanctioned entrance) – (time spent in the dark shelter after entering through the non-sanctioned entrance)] / (total time spent in shelter). group all the data needed for the analysis. S Before any analysis was performed, data were examineddeviation from the strainfor mean) (see Suppl.outliers Table S12 for details).(>3 A small number timesof mice were the standard define as outside sleeper because they were FVB;MicesleepingNOD).7with3no BALB;outsideC57; 3their5 less andinshelterthan50% (n=22; A/J; 4 outside more than 25% ofdifferenceanalysis.The the for exclude theirentrance)were each for 50%(exactly time4 day onpreference between strains in the total number of entrances across the first 4 days was determined using Generalizedestimating equations 1986).(GEE) (Zeger, The binomial test wasused determine to specifictestingwhethera 4,probability entrance takedaythesignificantlypreference was on to higherthan 50%The(FDR differences 0.014). < between the genotypes in the probability to use the preferred entrance on day 4 were tested using tests of equal proportions. (Verhoeven et al., 2005), alpha-levels were corrected by the minimum positive FDR with a threshold set at 5%. Kruskal-Wallis and Mann-Whitney U test were used for non-continuous data, like the onset and variability. GEE were used to model the frequency of the use of the entrance in the shelter, to model the change in preference resulting from the manipulation and the difference between strains in this change. Repeated measures ANOVA was used to discriminate the change of time spent in the shelter using one or the other entrance. All statistical analyses were performed in PASW Statistics© 17.0.2 and R© version 2.11.0.

Results To analyze avoidance learning, mice were tracked in instrumented home cages (PhenoTyper) that contained a shelter with two entrances. Mice were left in these cages for 7 days without human interference and normal 12-h dark-light cycle. During the first 4 days the animals developed a natural preference for one of the two shelter entrances. The PhenoTyper software automatically detected the preferred entrance on day 4 (Fig. 4a). During day 5 and 6, the use of this preferred entrance was sanctioned during each entry by a bright light that illuminated the shelter and remained on until the animal left the shelter again. The use of the other entrance was not sanctioned. On the last half day (dark phase of day 7), the aversive stimulus was no longer applied (probe trial, Fig. 4a). Our test set up is described in Fig. S1. We first studied basic sheltering behavior and development of a preferred entrance, and subsequently tested response strategies to an aversive stimulus in 8 common inbred strains (129S1/SvImJ, n= 53; A/J, n= 33; BALB/C, n= 33; C3H/HeJ, n= 22; C57BL/6J, n=87; DBA/2J, n= 36; FVB/N, n= 27; NOD/LtJ, n= 29). These strains together cover ~75% of the natural genetic variation in mice (Roberts et al., 2007). Across the entire cohort, the two entrances were equally often preferred (52% versus 48%; Fig 4b), ruling out a general entrance bias due to the layout of the arena. Subsequently, we screened collections of gene-trap mice using the Sleeping Beauty transposon/transposase system (Geurts et al., 2006), and null mutant strains, in total 42 strains, typically n=12 homozygous male mutants (in case of severe mutations: heterozygous) vs. 12 wild-type littermates (see Table S1 for genotypes and group sizes).

Genotypic differences in shelter visit pattern and preference During the first 4 days without interference, mice developed a characteristic pattern of shelter visits, with frequent entries, often with short durations, during the dark phase, and fewer entries during the light phase (Fig. 5a). Among the inbred lines, substantial differences were observed in the number of visits (GEE, χ2 = 4024, df = 8, p <0.001; Fig. 5a-b; see also Suppl. Tables S2-S3). The number of entries averaged over the total light and dark phases (12 h each, Fig. 5c) confirms the differences between the genotypes in visit frequencies (dark phase: GEE, Genotype x day: χ2 = 224, df = 21, p <0.001; light phase: GEE, Genotype x day: χ2 = 114, df = 21, p <0.001), and differences between light and dark phase for each genotype (Suppl. Table S4). Differences between genotypes were also similar in light and dark phases (Fig. 5c; similar strain ranking). Differences in shelter visit frequency among the genotypes correlated well with differences in general activity (Fig. 5d, Pearson correlation coefficient: 0.79, p<0.0001, n =320; see also Suppl. Table S5).

32 High throughput phenotyping of avoidance learning in mice discriminates different genotypes and identifies a novel gene 2 33 Day 7 Probe trial Maroteaux et al. 2012 Stably changed preferance 25% 59% 50% 48% 68% 58% 61% 45% 52% Right zzZZZ responses 0% 80% 100% 6 Resting outside Aversive response Aversive cognitive response Detecting cognitive/aversive 0% 4 Day 5 → 6 Left 75% 41% 50% 52% 32% 42% 39% 55% 48% Applying aversive stimulus Preferred entrance 0% 20% Non-preferred entrance J 6 A/J FVB/NJ DBA/2J all mice BALB/C NOD/LtJ C3H/HeJ C57BL/ 129S1/SvImJ Day 1 → 4 Protocol and data analysis. (a) The protocol is divided in 2 phases. First the habituation phase when the Free exploration Defining preference a igure 4: b mouse can freely explore the cage and the computer learningdetects phase whenthe the preferredmouse experiencesentrance. the aversiveSecond, stimulus the whenever avoidanceit uses its preferred potentiallyentrance. resultsThis in a modification of the behavior, typically, a decrease in the frequency of entering the shelter through the preferred entrance and in the time spent in the illuminated shelter. (b) a Percentagepreference of formice thehaving left or the right entrance. All strains showentrance exceptthatNODrespectivelyandC3H equalpreferencehaveleftrightthea(75%)theandfor (68%) entrance. proportion to prefer the right or the left F After 4 days, 129S1/SvImJ (129S1) mice developed the strongest preference (group average: 83 ± 2% of entries via their preferred entrance), followed by BALB/C (BALB) and DBA/2J (DBA) (76 ± 3 % and 73 ± 2%, Fig. 5e). Only 36% of the C3H/HeJ (C3H) mice developed a significant preference (see Suppl. Table S6), although the group average probability to enter the preferred entrance is rather high (69 ± 3%; Fig. 5e). This is explained by the low number of entries of C3H, and thus, low statistical power (see also Suppl. Fig. S2-S3). The total time spent outside decreased during the first three days in some strains, indicating a habitation effect. The time spent outside the shelter had stabilized by day 4 (Suppl. Fig. S4).

Genotypic differences in avoidance behavior upon a mild aversive stimulus Avoidance learning was studied by automatically applying a mild aversive stimulus (shelter illumination with bright light) during days 5 and 6 each time mice entered the shelter using their preferred entrance, but not the other (see Fig. 4). Prior to the analysis of the effect of the aversive stimulus, we discarded two different categories of mice that were not suitable for the experiment. The first category consisted of all mice presenting no shelter preference (exactly 50:50 between the two entries, 3 mice in total of three different strains; C57BL6/J (C57), DBA and NOD/LtJ (NOD)). The second category comprised mice that did not visit the non-sanctioned entrance during the first 12h of day 5 (the beginning of the test phase; Suppl. Fig. S5 and Suppl. Table S7). These mice had not experienced that only the preferred entrance was sanctioned and therefore had no incentive to change their preference. Upon introduction of the aversive stimulus, the number of entries via the preferred entrance decreased for most genotypes (Suppl. Fig. S3). However, this response is a combination of specific aspects, recognizing and actively avoiding the sanctioned (wrong) entrance, and more general aspects, making fewer entries via either entrance (see Suppl. Fig. S3) and/or avoiding the shelter altogether (see below). Therefore, we used a more specific measure of the cognitive aspects: the fraction of entries via the preferred (and later sanctioned) entrance over the total number of entries, referred to as the ‘preference index’. This also allows comparisons between light and dark phases (when substantially different absolute numbers of entrances are made, Fig. 5a-c). A decrease of this index indicates that a mouse has established a specific association between its preferred entrance and the aversive stimulus and is a direct consequence of a decision process that occurred before an entry was made. Therefore, we consider changes in the preference index as a specific cognitive response. 129S1, DBA and C57 mice showed the strongest decrease of this index, i.e., the strongest cognitive response (Fig. 6a-c/i). The group average for DBA mice went from 0.51 (on day 4) to -0.4 (on day 6) over the two days that the preferred entry was

Figure 5: Spontaneous use of the shelter entries. (a) Entries through either the left (L) or right (R) entrance during the first 4 days for three mice of three different genotypes. (b) Total number of entries for every 15 min for 8 inbred strains. (c) Comparison of the total number of entries in dark and light phase for each strain. (d) Relation between the number of entries and the distance moved accumulated on the first 4 days. Cloud borders encompass the extreme individuals of each strain. Grey backgrounds in the different panels represent dark phases. Data shows mean ± S.E.M. (e) Average probability (± S.E.M) of each strain to enter the shelter through the preferred entrance on day 4. The white dashed line represents the 50% chance level to take one entrance over the other. A test of equal proportion was used to compare the mean number of entries between strains (FDR= 0.015). ◄

34 High throughput phenotyping of avoidance learning in mice discriminates different genotypes and identifies a novel gene 2

35

NOD/LtJ

FVB/NJ 4 DBA/2J

A/J 129S1/SvJ C3H/HeJ C57BL/6J

96

C3H/HeJ BALB/C

Maroteaux et al. 2012 Light phase Light

Day A/J

3 129S1/SvImJ 84 1 2 3 4

Day 4 0

0

Days

0.2 0.4 0.6 0.8 200 100 probability entrance Preferred e 2 72 Dark phase Dark Day 1 60 1 2 3 4 Day 3 0

500 400 300 200 100 # entries/12hrs # c 48 A/J (N=33) C3H/HeJ (N=22) (N=36) DBA/2J C57BL/6J (N=87) C57BL/6J 0.5 1 1.5 2 2.5 Hours 4 36 Day 2 3 24 Days General activity (distance, km) 2 12 Day 1 1 0.5 1 1.5 2 2.5 3 3.5 NOD/LtJ (N=29) FVB/NJ (N=27) 129S1/SvImJ (N=53) 129S1/SvImJ (N=33) BALB/C 0

0 5

0 3 2 1

10 15 20 L R L R L R

# entries x 1000/4days x entries # # entries/15min # Entrance d b a sanctioned (Fig. 6a/i). After these two days, 91% of DBA and 79% of C57 mice showed significant cognitive responses (False Discovery Rate (FDR) < 0.01; Fig. 6a,c/iii). 129S1 also showed a strong cognitive response, but the preference index did not decrease under 0 (Fig. 6b/i). The cognitive response was highly significant for these three genotypes on both day 5 and 6 (as compared to day 4, see Suppl. Table S8, GEE p<0.001). The cognitive response of BALB, NOD and A/J mice was slower and only reached significance by day 6 (Fig. 6d,h,f/i; Suppl. Tables S8-S9). The other two genotypes (FVB and C3H) showed no significant cognitive response (Fig. 6g,e/i), although several individuals of these strains did (Venn diagrams in Fig. 6g,e/iii). The cognitive response of all genotypes is highly comparable during light and dark phases (Fig. 6/i). Mice that show weak/no cognitive responses might do so because they do not experience shelter illumination as aversive. To address this possibility, we measured the time a mouse spent in the shelter after using the sanctioned/wrong entrance (i.e., the time spent in the now illuminated shelter). A change in the time spent in the now illuminated shelter is a measure for how aversive the light stimulus is. This might be different for individual mice and is driven by multiple individual factors such as anxiety, visual sensitivity and bright light aversion. To visualize this measure, we calculated the aversion index (see methods). A decrease in this parameter (‘aversion index’) indicates that the mouse experiences shelter illumination as aversive. Indeed, the aversion index was significantly decreased for all genotypes, except for C3H and FVB, two visually impaired strains (Wong and Brown, 2006) and for A/J (Fig. 6e,g,f/ii; see Suppl. Table S10). This indicates that most genotypes did experience illumination as aversive, but that for three strains, on average, shelter illumination is probably not sufficiently aversive to evoke avoidance responses, although several individuals of these strains showed an aversion to the light (Venn diagrams in Fig. 6e,g,f/iii). During the 2 days that the aversive stimulus was applied, mice might also have habituated to illumination, limiting changes in the aversion index and explaining why this index levels off between days 5-7 (Fig. 6/ii). To reveal the full potential of this avoidance test and visualize the complex response of each strain, we plotted these two indexes (12 h bins, dark and light phase) against each other, in one graphic (Fig. S6). This graphic shows the different strategies displayed by the different strain to cope with the aversive stimulus. In addition to the two principal responses (cognitive and aversive responses), we observed that some mice showed an alternative response, avoiding the shelter altogether and resting (sleeping) outside the shelter. These mice apparently established a general association between the aversive stimulus and the entire shelter, rather than a specific association between the stimulus and one specific entrance, as outlined above. This generalized association was most prominent for DBA mice. On day 6, 60% of these mice showed a significant increase in the time

Figure 6: Effect of the aversive stimulus on shelter entrance. (a-h/i) Preference index for every 12h, using a GEE model to define the effect of the manipulation. (a-h/ii) The aversion index. (a-h/iii) Venn diagrams show the percentage of mice significantly changing their behavioral response from days 4 and 6, using a test of equal proportion to define significance of increase or decrease (the yellow square represents the whole group; the dark green: the decrease in preference FDR= 0.011; the light green: the decrease of time in shelter using the sanctioned entrance FDR= 0.023; in red the increase of time resting outside FDR= 0.05). The overlap of the squares represents the population that shows 2 or 3 of the behaviors between day 4 and 6. The squares are proportional to the size of the group (for more details on the groups see Table S10). ◄

36 High throughput phenotyping of avoidance learning in mice discriminates different genotypes and identifies a novel gene 2 37 Days Days 43% Maroteaux et al. 2012 56% 52% 46% BALB/C (N=25) BALB/C NOD/LtJ (N=28) 1 2 3 4 5 6 7 1 2 3 4 5 6 7 54% 32% 25 28 0 0 0 0 0.6 0.4 0.2 0.6 0.4 0.2 0.6 0.4 0.2 0.6 0.4 0.2 -0.2 -0.4 -0.2 -0.4 -0.2 -0.4 -0.2 -0.4 h d i ii iii i ii iii % mice increase in with time spent outside resting a significant 37% 74% Days Days 33% 63% FVB/NJ (N=27) C57BL/6J (N=86) C57BL/6J 1 2 3 4 5 6 7 1 2 3 4 5 6 7 37% 79% 27 86 0 0 0 0 0.6 0.4 0.2 0.6 0.4 0.2 0.6 0.4 0.2 0.6 0.4 0.2 -0.2 -0.4 -0.2 -0.4 -0.2 -0.4 -0.2 -0.4 g c i ii iii i ii iii % mice decrease in preference with a significant Days Days 60% 66% A/J (N=29) 62% DBA/2J (N=35) DBA/2J 45% 1 2 3 4 5 6 7 1 2 3 4 5 6 7 45% 91% 29 35 0 0 0 0 0.6 0.4 0.2 0.6 0.4 0.2 0.6 0.4 0.2 0.6 0.4 0.2 -0.2 -0.4 -0.2 -0.4 -0.2 -0.4 -0.2 -0.4 f b iii i ii iii i ii % mice shelter in spent time in decrease with a using preferred entrance significant Days Days 36% 67% 32% 26% C3H/HeJ (N=19) C3H/HeJ 129S1/SvImJ (N=42) 129S1/SvImJ 1 2 3 4 5 6 7 1 2 3 4 5 6 7 19 37% 42 69% 0 0 0 0

0.6 0.4 0.2 0.6 0.4 0.2 0.6 0.4 0.2 0.6 0.4 0.2

-0.2 -0.4 -0.2 -0.4 -0.2 -0.4 -0.2 -0.4

Non-pref Non-pref Non-pref Non-pref Preferred Preferred Preferred Preferred

Total # of mice Total

Aversion Index Aversion Preference Index Preference Aversion Index Aversion Preference Index Preference e a iii ii i iii ii i spent resting outside (Fig. 6b/iii). A/J and NOD also showed prominent increases in outside resting (45% and 46%, respectively Fig. 3f,h/iii). For other genotypes, a smaller percentage of individuals started resting outside, but this response was observed for some individuals of all genotypes (Venn diagrams in Fig. 6/iii).

Specc1/cytospinB gene-trap mice show substantially delayed avoidance learning Finally, we searched for novel genes involved in avoidance learning and tested whether our newly developed automated assay could be used for this. For this, we applied the avoidance learning assay to a collection of gene-trap mice, which were generated using the Sleeping Beauty transposon (Geurts et al., 2006). This approach yields random inactivation of typically single genes, unique for every individual new integrant, without flanking gene complications known to confound behavioral assessment of knockout mice (Wolfer et al., 2002). Out of 43 screened single gene mutants, we identified one integrant that showed a specific defect in avoidance behavior. Integration of the transposon was localized (see Experimental Procedures) to the third intron of specc1/cytospinB a gene with no functional annotation yet (Fig. 7a). The location of the trap is predicted to prevent expression from 2 of the 3 transcription initiation sites, including the most widely expressed transcripts (canonical sequence, Fig. 7a). Loss of transcripts was confirmed by qPCR on whole brain mRNA according to the strategy outlined in Fig. 7a. In the brain, the gene was mainly expressed in the hippocampus (Allen brain atlas). Homozygous specc1/cytospinB mutant mice (n=11) were viable and showed no morphological abnormalities. The mutants had normal sensory-motor development and responded normally to visual cues. However, they exhibited a remarkable delayed response to the aversive stimulus and changed neither their preference (cognitive response) (see Fig. 7b/i Suppl. Tables S8-S9) nor the time spent in the shelter after entering via the preferred entrance (aversive response) during the 2 days that the aversive stimulus was applied (Fig. 7b/ii). During these two days, control littermates (n=11) showed cognitive- and aversive responses very similar to the responses previously observed for the founder strain (C57, Fig. 6c and 7b; GEE and RM ANOVA; see Suppl. Tables S8 and S9). Only during the last day, day 7, specc1/cytospinB mutant mice showed a cognitive and aversion response which normal mice show 2 days earlier (Fig. 7b). While avoidance learning was drastically delayed in the specc1/cytospinB mutant mice, other aspects of their behavior were normal. The overall activity pattern (Fig. 7c, d, f), their moving velocity (Fig. 7g), the pattern of entering the shelter, also relative to the day-night cycle (Fig. 7h), the total number of entries on days 1-4, i.e., until the introduction of the aversive stimulus (Fig. 7e), as well as the strength of their preference for one of the two entrances (Fig. 7f) were all normal.

Figure 7: Comparison of SPECC1 –deficient (KO) and wild-type (WT) controls. (a) SPECC1 gene-trap construction and different splice variants. (b/i) Preference index for every 12h. (b/ii) Aversion index. (c) Relation between the distance moved and the frequency of entrance use. (d) Total distance moved in the first 4 days (km). (e) Total number of entries in the first 4 days. (f) Preferred entrance probability on day 4. (g) Mean velocity in the dark and light phase (cm/sec). (h) Circadian rhythm of the total entries in 15 min bins. ◄

38 High throughput phenotyping of avoidance learning in mice discriminates different genotypes and identifies a novel gene 2 39 3’ 3’ C AAAA WT KO 4 0

0.8 0.6 0.4 0.2 D4 probability (%) probability D4 f Peptide Translation Transcription Maroteaux et al. 2012 normal mRNA normal protein WT KO Specc1/Cytospin-B Distance Distance (km) NH2 0 3 5’

800 600 400 200 12 13 14 15 16 17 1000 entries # WT iso92415 11 e 3’ C Days AAAA 0 0.5 1 1.5 2 2.5 WT KO 0

0.8 0.6 0.4 0.2 1.6 1.4 1.2 1.0 0 3’ x1000 entries #

1.0 1.4 0.6 0.2 2 Distance moved (km) moved Distance c d Peptide Translation no protein Transcription Transcription 5’ NH2 5’ 1 WT iso108709 Mutant iso108709 3’ C iso92415

Days

3’ 8 6 4 2 0 # entries # h Peptide Translation no protein Transcription Transcription 5’ C57BL/6J (86) (11) SPECC1-WT SPECC1-KO (11) 1 2 3 4 5 6 7 NH2 5’ iso49836 iso108709 1 2 3 Transposon 4 5 6 7 8 9 10 0 0

0.6 0.4 0.2 0.4 0.2 WT iso49836 Mutant iso49836

-0.2 -0.4 -0.2 -0.4 Dark Light

WT KO WT KO Preferred 5’ Non-pref

Non-pref Preferred

Preference Index Preference 6 4 2 0 Aversion Index Aversion Velocity (cm/sec) Velocity Mutant mRNA Mutant protein normal mRNA normal protein splice variant exon ii i b a g Discussion In this study, we investigated aversive learning in mice using a fully automated assay that relies on mouse-triggered sanctioning of shelter entrance with a mild aversive stimulus. Our data show that the complex adaptive behavioral response of mice can efficiently and successfully be detected, analyzed and visualized even in large cohorts of (mutant) mice. Different genotypes and single gene mutants exhibit marked and quantitative differences in distinct aspects of this behavioral response. The current data show that the natural preference of mice to reside in a dark/sheltered compartment can be exploited, as part of the natural behavioral repertoire (Kas et al., 2008), to generate high content behavioral information on cognitive traits, behavioral flexibility and anxiety. In addition, the current data also revealed several striking new features of sheltering behavior, such as marked, genotype-dependent differences in visit frequencies, the existence of sharp peaks in shelter visit activity and genotype-dependent timing of these peaks relative to the light/dark cycle. This circadian variation in sheltering behavior may contribute to inconsistencies in anxiety tests among strains when performed at different circadian time (Brunner et al., 2002; Jones and King, 2001). Our assay produced three principal features that could be firmly established for all genotypes except two visually impaired genotypes, FVB and C3H (Wong and Brown, 2006), and for transposon gene trap mice (in C57 background): (i) specific association between one shelter entrance and the aversive stimulus, that remained after conditioning was discontinued on day 7, was observed in most individuals of 8 inbred strains and gene trap mice (ii) generalized associations, observed in all genotypes defined by the apparent association between the shelter in general and the aversive stimulus, without discriminating between the two entrances, and (iii) aversion responses (staying shorter in the shelter when it is illuminated), observed in most mice of all genotypes. As 65% of visually normal mouse strains showed significantly reduced shelter residence time after using the sanctioned (now illuminated) preferred entrance, it can be concluded that the mild aversive stimulus can be used successfully for those strains, and genetic resources derived thereof. The avoidance response patterns observed in this study are a composite response consisting of aversion to the bright light inside the shelter, cognitive aspects (specific recognition of the sanctioned entrance and avoiding it) and behavioral flexibility aspects (suppression of the tendency to display learned behavior and switching to an alternative). Interestingly, some strains with high aversion response to light in the light-dark box test proved to be poor performers in our paradigm (i.e., A/J) and, conversely, low responders were good performers here (i.e., C57 (de Mooij-van Malsen et al., 2009)). This indicates that the cognitive response is not merely a read-out of differences in aversion to the light that could be the result of factors such as light-induced anxiety and visual capabilities. Similarly, higher anxiety in DBA than in C57 mice is not a predictor of good fear learning (Baarendse et al., 2008; Stiedl et al., 1999). Instead, it is conceivable that the three genotypes whose cognitive response was most pronounced (i.e., 129S1, DBA and C57) have superior learning performance and/or behavioral adjustability, which make them suitable for future drug screening programs identifying cognition modifying compounds. The similar performance of C57 and DBA mice is of particular interest as DBA mice are generally inferior to C57 in aversively motivated tasks, have been reported to be hippocampus impaired, while under appetitive conditions they perform similar to C57 mice in

40 High throughput phenotyping of avoidance learning in mice discriminates different genotypes and identifies a novel gene 2 41 specc1/ produced a delay in avoidance learning, whereas many general activity parameters, Knocking out the major splice variants of the functionally not annotated gene To To our knowledge, no systematic assessment of behavioral adjustability has been The change in preference induced by the aversive light stimulus was still observed at -actinin, many microtubule-associated proteins and several rho/ras GTPase-activating and λ cytospinB circadian rhythms and sensory-motor abilities were unaffected. This delay can be caused by a specific impairment in learning the association between the entranceor andcould thebe lightthe stimulus result of a difference in theeither perceived case, aversiveness this of the example light stimulus. illustrates In learning the sensitivity phenotypes of using this subtle paradigm aversive to The encoded learning pick protein up contains stimuli prominent avoidance in protein-protein ainteraction domains, home domaincalponin-homologyanda a domain.coiled-coil cage Itsexpression levelrelatively is lowadultinbrain but environment. is upregulated in several behavioral paradigms and disease models (European Institute,Bioinformatics 2012; GeneNetwork, 2012). Calponin-homology domains are also found in spectrin, performedthesameinpanelinbred of strain used forcomparison.here Generalized responses, i.e., avoiding the shelter altogether and sleeping outside, might be a mechanisticallyresponse distinctfrom the specific association. It can be considered as a moreaversive direct situationresponse andto might an be similar to generalized fear responses with reduced flexibility/discrimination.Interestingly,homecagerecording revealed difference activity in level some for straincompareclassicalto behavioral test. instanceFor FVB, knownhyperactivebe to novelain openfield comparedto other strains (Millstein and Holmes, 2007) showed lesser activity in the PhenoTyper. Locomotor activity is known to be influenced by environmentalstates. andThis emotional relation is not linear as shown factor onin activitydifferent depending outcomeson the ofenvironment corticotropin-releasing(hypoactivity during noveltyhyperactivity exposure, in familiar environmentbut (Britton and Indyk, 1990)). day 7, when the aversive stimulus was no longer applied. This in implicatesbehavior an in adaptiveresponse tochange the light stimulus, which contributes face validityaspect ofto thisthe task.cognitive Panels of inbred mouse strains have been characterized in different assays of learning and memory, such as contextual fear conditioning Bothetask et (V. al.,J. Bolivar2004) andet spatialal., navigation2001; in Morris water mazemaze (O’Leary(Lad etet al.,al., 2010)2010). and DifferentBarnes experimental conditions, anxietywater andin the stressorsMorris water such maze, as complicate the comparisons between tests. Not surprisingly, strain rankings vary among these tasks, consistent with the idea that no single task reveals the fullrichness oflearning and memory phenotypes inwidea range ofgenotypes (Wahlsten et al., 2005). Therefore, the present paradigm is an important new additionaversivemildstimulus,a uses without runssinceit humaninterference derives cognitive andthe to existing paradigms, response from the ratio of preferred entries excluding confounding effects of general activity. spatial learning (see (Youn et al., 2012)).al.,spatialetunderscoresThislearning importance (Youn (seethe avoiding ofunspecific stressors to assess cognitive function in mice for determining its heritable underpinnings. GDP exchange factors, which implies that Specc1/CytospinB might, like these other beproteins, involved in cytoskeleton remodeling. Interestingly, cytoskeleton remodeling small GTPases have a strong link to cognitive abilities in different mammals(vanGalen and Ramakers,including 2005)). Moreover,oftheSNPs13top reported humansbyVerweij etal.(2010) (reviewed in for harm avoidance in humans, rs971718 on chr 17 (p-value 3.8x10-5, effect size -.23) is located 6 Mb upstream from SPECC1 (Verweij et al., 2010). Together, the results described in this manuscript show that subtle learning stimuli in a home cage environment provide sufficient salience to drive avoidance learning. Due to the efficiency achieved with automated home cage screening, we anticipate that learning protocols in home cage settings will gain popularity in large scale mouse phenotyping efforts. Such efforts, especially when combined with drug screening or lesion studies, will further validate the avoidance learning paradigm as a novel cognition test in mice.

42 High throughput phenotyping of avoidance learning in mice discriminates different genotypes and identifies a novel gene 2 43 Storage Graphics Statistics Results and Y coordinates X FDR p<0.05 BALB/C DBA/2J A/J C57BL/6J NOD/LtJ 129S1/SvJ C3H/HeJ FVB/NJ MoPhDb . TM AHCODA Analysis -Processing Days Total # of entries Total EthoVision EthoVision Acquisition HTP 2.1.2.0 HTP 1 2 3 4 5 6 0

700 600 500 400 300 200 100 # entries # Total number of Total entries per day during the 6 experimental days. C3H, DBA, 129S1, FVB, NOD displayed a Overview of the screening system with 48 cages connected to 12 computers running EthoVision HTP 1: 2: S S x 6 = 48 PhenoTyper upplemental igure igure igure stable number of entries along the experiment. C57 mice showed an increase until day 4 and then a decrease in the the in decrease a then and 4 day until experiment. the showedalong C57 increase entries mice an of number stable number of mice entries. showed A/J a decrease until day 3 then an increase until day 5 and finallya small decrease details). statistical for S8 and S3 Table (see entries of number the in decrease constant a showed mice BALB 6. day on F 2.1.2.0 software, controlling hardware in the PhenoTyper and saving track files containing F S and zone visits of mice at 15 frames/s. Track files were error2.1.2.0 softwarecorrected offline, and and statistically smoothened analyzed and visualized using with AHCODA EthoVision HTP Number of entries using the preferred entrance 300 Dark phase

250

200

150

100 # entries 50

129S1/SvJ 0 1 2 3 4 5 6 7 A/J 100 BALB/C Light phase C3H/HeJ C57BL/6J 50 DBA/2J FVB/NJ NOD/LtJ 0 1 2 3 4 5 6 7 Days

Figure S3: Number of preferred entrances during the 6.5 days of experiment. The vertical dashed lines delineates the test period. Three strains (C57, 129S1, DBA) showed a decrease in the use of the preferred entrance at the beginning of the test period in the dark phase. BALB, A/J, NOD, FVB mice showed a decrease starting on day 5. C3H did not show any change in preferred entrance use. The differences were smaller in the light phase, during which mice did not seem to change their use of the preferred entrance. Only A/J mice showed an increase in the use of the preferred entrance during the test period in the light phase.

44 High throughput phenotyping of avoidance learning in mice discriminates different genotypes and identifies a novel gene 2 45 6 12 BALB/C DBA/2J 129S1/SvJ A/J BALB/C C3H/HeJ C57BL/6J DBA/2J FVB/NJ NOD/LtJ 5 9 A/J C57BL/6J NOD/LtJ 4 6 Days 129S1/SvJ C3H/HeJ FVB/NJ Time in hours Time 3 3 2 1 0 0

80 60 40 20

100 % mice entering the non-sanctionned entrance non-sanctionned the entering mice %

9 8 7 6 5 4

11 10 Time spent outside (h) outside spent Time Cumulative plot of the percentage of mice that entered the shelter through the non-sanctioned Time spent outside the shelter. A repeated measures ANOVA showed a general decrease in the time 5: 4: S S igure igure entrance during the dark phase of day 5. After 6 h of the dark phase of day 5, 100% of theFVBC57, non-sanction and NOD entrancevisited at least once. The last DBA mouse visiting this entrance showed this response atof 11 theh dark phase. Not all 129S1, A/J, BALB and C3H mice visited the micenon-sanctioned were excluded from the entrancefurther analysisafter because12 theyh. were notThose aware of the presence of a non-sanctioned entrance and therefore could not learn to change their preference. F spentoutside during firstthe3 days (between days (F(1,312)=20.7; and 2 1 p<0.001), and days (F(1,312)=6.56; and 3 2 p=0.01)). This overall effect is mainly dueto A/J and FVB that show(respectively a p=0.003 and significant p<0.001). decrease between day 1 and 2 F 0.2 C57BL/6J 0

-0.2

-0.4 0.2

0 DBA/2J -0.2

-0.4

-0.6 0

-0.2 129S1/SvJ ← Aversive response (Aversion Index)

-0.4

-0.6

-0.8 0

-0.2

-0.4 BALB/C -0.6 0 Aversive response (Aversion Index)→ C3H/HeJ -0.2

-0.4

-0.6 0

A/J -0.2 -0.4 0

-0.2 NOD/LtJ -0.4 0

FVB/NJ -0.2

-0.4 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 Cognitive response (preference index) →

Figure S6: Index Plot. We plotted in one graph the deliberate change in preference to enter through the non- sanctioned entrance (X-axis), and the change in time spent in the shelter after entering through the non-sanctioned entrance (Y-axis). To simplify the visualization, the two parameters plotted here are the inverse of the preferred index and the aversion index. These two parameters are plotted against each other before the introduction of the aversive stimulus (day 4), during the two days when the stimulus is operational (days 5-6), and in dark phase of day 7 when the stimulus was no longer applied (12 h bins, dark and light phase). Aversion (upward shift) is expected to develop instantaneously in response to the aversive stimulus, whereas the cognitive response (rightward shift) is expected to develop gradually, reflecting the avoidance learning process. C57 mice showed such a response: during the first 24 h (day 5) the response involved both aversion (upward) and cognitive (rightward) aspects, followed by a stabilization of the aversion behavior whereas development of the cognitive aspect of the response continued (only a further rightward shift) during the following 24 h (day 6). A similar pattern was observed for DBA mice, until the end of day 5, but the preference change stabilized on day 6 already. 129S1 mice showed most of the cognitive response preceding the aversive response. This can be explained by the fact that during the dark phase of day 5 the response is mainly driven by a decrease of the use of the preferred entrance and that during the light phase almost no changes occur. The main part of the aversive and cognitive responses takes place during day 6. BALB mice display an upward shift only during the dark phase of day 5 (aversive response mainly). The remaining genotypes (C3H, A/J, FVB and NOD) showed no significant aversion response (although individual mice did), nor a cognitive response.

46 High throughput phenotyping of avoidance learning in mice discriminates different genotypes and identifies a novel gene 2 47 Z

8 6 6 13 13 12 12 12 12 12 12 61 35 24 H

11 13 12 12 12 12 12 19 14 14 25 10 KO

7 2 8 9 6 4 21 12 T

7 6 6 11 13 13 12 12 12 12 12 12 12 12 21 verage 14 23 10 59 28 20 W A entries in 15 min train ng3 S mcry1mcry2 menin nf1 munc18 pkci r192q scap tsc tscb6 loxm18 null mcry1 mcry2 Snap25 specc1 tmed trim3 ube3 Ubn1 tarp8 syt1

Z

9 9 4 4 11 15 12 81 22 26 45 47 111 68 84 44 H aximum number M of entries in 15 min 5 5 9 11 11 13 17 12 12 12 12 16 23 23 25 10 10 36 28 24 24 20 KO T 7 6 6 11 13 12 12 12 12 12 19 18 14 14 25 25 10 20 40 W Maximum number of shelter entries and average number of shelter entries in 15 min for each strain. List of the number of mutagenized mice (WT: wild-type; KO: knock-out; HZ: heterozygous) screened 1: 2: S S able able train T in the PhenoTyper. S T ckii clasp2 C3H/HeJ C57BL/6J DBA/2J FVB/NJ NOD/LtJ alpha7 beta2 bko cb1 brevican clip115 clip170 DJ1 kazugaa1 kazu-t kazu-tlnic403 doc2A A/J BALB/C 129S1/SvImJ 218l 305d doc2ab fgf13 Fx kazubcatloxp Kcnd2 lach Table S3: Habituation pattern. This table compares the total number of entries on days 1, 2 and 3 to day 4. 129 and C3H mice showed a stable number of entries along the 4 days. C57 and DBA showed an increase between days 1 and day 4, whereas A/J, BALB, FVB and NOD mice displayed a decrease in the number of entries.

Mean 95% Wald Confidence (J) Difference Std. Interval for Difference Habituation days (D4-J) Error df Sig. Lower Upper day 1 to 4 129S1/SvImJ 1 -13.72 14.518 1 0.345 -42.17 14.74 stable 2 6.09 9.287 1 0.512 -12.11 24.30 3 -12.09 11.392 1 0.288 -34.42 10.23

A/J 1 -74.58 21.953 1 0.001 -117.60 -31.55 decrease 2 -20.09 21.614 1 0.353 -62.45 22.27 3 34.09 15.497 1 0.028 3.72 64.47

BALB/C 1 -208.36 39.886 1 0.000 -286.54 -130.19 decrease 2 -125.39 34.855 1 0.000 -193.71 -57.08 3 -77.88 20.455 1 0.000 -117.97 -37.79

C3H/HeJ 1 -1.82 3.283 1 0.580 -8.25 4.62 stable 2 2.82 2.594 1 0.277 -2.27 7.90 3 .55 2.369 1 0.818 -4.10 5.19

C57BL/6J 1 57.87 8.873 1 0.000 40.48 75.26 increase 2 53.30 7.448 1 0.000 38.70 67.90 3 28.23 6.446 1 0.000 15.60 40.86

DBA/2J 1 3.92 6.361 1 0.538 -8.55 16.38 stable 2 7.44 7.521 1 0.322 -7.30 22.19 3 9.78 4.455 1 0.028 1.05 18.51

FVB/NJ 1 -54.48 11.246 1 0.000 -76.52 -32.44 decrease 2 15.52 10.302 1 0.132 -4.67 35.71 3 10.89 6.323 1 0.085 -1.50 23.28

NOD/LtJ 1 -58.31 14.684 1 0.000 -87.09 -29.53 decrease 2 18.07 12.182 1 0.138 -5.81 41.95 3 24.24 9.777 1 0.013 5.08 43.40

48 High throughput phenotyping of avoidance learning in mice discriminates different genotypes and identifies a novel gene 2 49

pper 19.17 82.13 89.65 69.09 243.73 140.78 349.56 U 136.578 ifference N 35 53 33 33 27 22 28 86 onfidence D C nterval for ower 95% Wald 51.13 67.53 I 14.56 60.19 103.91 121.66 181.62 L 252.90 ig. S 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 .328 .073 .445 .028 .003 .000 .000 .000 ig. (2-tailed) S 1 1 1 1 1 1 1 1 df rror E 1.18 7.52 3.73 8.33 4.58 4.88 15.84 24.66 td. S orrelation C .138 .615 .701 .475 .382 .249 .706 -.219 ight) L ean earson 60.11 74.92 74.83 16.86 P 131.22 M 212.67 301.23 120.24 ifference ark- D D ( GEE model comparing the total number of entries in the dark and light phase for the first 4 days. All Correlationbetween distancethe movedDBA,numbertheand entries. ofFVB, mice NODC57, showed 4: 5: S S able able DBA/2J FVB/NJ NOD/LtJ C57BL/6J T T the strains showed a highly significant 0.0001)(p< difference between the entries in the shelter during the dark and the light phase. highlysignificant correlation (p<0.0001) between the distance moved and the number entriesof during the first 4 days; BALB mice displayed a weaker correlation but still significant (p<0.05), whereas and A/J C3H showed no significant correlation. BALB/C C3H/HeJ C3H/HeJ C57BL/6J DBA/2J FVB/NJ NOD/LtJ A/J 129S1/SvImJ BALB/C 129S1/SvImJ A/J Table S6: Percentage of mice showing a significant preferred entrance on day 4. On day 4, Over 80% of 129S1, A/J, BALB and DBA mice had a significant preference (i.e significantly higher than 50%) for one of the 2 entrances. Of C57, FVB, NOD , 65.7, 53.1 and 61.4% respectively had a preference on day 4. Only 36% of C3H mice had developed a significant preference on day 4.less than half the population (36%) that showed a significant preference on day 4.

% of mice having a significant (FDR =0.014) preferred entrance 129S1/SvImJ 86.8 A/J 75.8 BALB/C 93.9 C3H/HeJ 36.4 C57BL/6J 65.5 DBA/2J 80.6 FVB/NJ 48.1 NOD/LtJ 69.0

Table S7: Number of mice per strain that failed to use the non-sanctioned entrance 12 h after the beginning of the test phase on day 5.

mice not using the non-sanction entrance on day 5 dark phase 129S1/SvImJ 11 A/J 4 BALB/C 8 C3H/HeJ 3

50 High throughput phenotyping of avoidance learning in mice discriminates different genotypes and identifies a novel gene 2 51 pper 0.11 0.15 0.13 0.17 0.51 0.19 0.21 0.35 0.10 0.33 0.32 0.38 0.76 0.22 0.39 0.24 0.26 0.68 0.08 0.04 U ifference onfidence D C 0.11 0.17 0.18 0.35 0.10 0.23 ower 0.05 0.05 0.02 0.02 0.20 0.09 0.09 0.00 -0.01 -0.05 -0.07 -0.03 -0.03 -0.04 nterval for L 95% Wald I

ig. S 0.155 0.917 0.197 0.126 0.179 =0.018) 0.281 0.013 0.025 0.001 0.030 0.020 0.006 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 α ( 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 df rror E 0.172 0.162 0.108 0.109 0.016 0.037 0.014 0.037 0.041 0.023 0.025 0.052 0.039 0.024 0.024 0.028 0.043 0.050 0.020 0.048 td. S ean 0.17 0.13 0.12 0.12 0.16 0.14 0.21 0.18 0.43 0.25 0.43 0.36 0.05 0.03 0.28 0.07 0.30 0.08 0.04 0.00 M (J-d4) ifference D

ay 5 5 5 5 5 5 5 5 5 5 6 6 6 6 6 6 6 6 6 6 (J) D Results of the generalized estimated equations (GEE) on the proportion of entries through the preferred preferred the through entries of proportion the on (GEE) equations estimated generalized the of Results 8: 8: S able able SPECC1-KO SPECC1-WT NOD/LtJ FVB/NJ DBA/2J C57BL/6J C3H/HeJ entrance. Day 4 is used as the reference day. The simple contrast test showed significant (FDR= 0.018) differences differences 0.018) showed test The simple contrast significant(FDR= day. Day 4 is used as entrance. the reference through entries of proportion the in difference significant a and BALB/C, and DBA C57, 129, for 5 and 4 days between the preferred entrance between days 4 and DBA 6 C57, and for A/J, BABL/C. 129, The mean differenceis positive because we observe an increase in the probability between days 4 and 5 and day change. significant showa not did 6. SPECC1-WT SPECC1-KO whereas testing, mice daysof both during increase significant showed a T BALB/C A/J 129S1/SvImJ Table S9: GEE results for pair-wise comparison of the effect of the sanctioned entrance on the number of entries through the preferred entrance between strains.

Mean 95% Wald Confidence Difference Sig. Interval for Difference (I) Strain (J) Strain (I-J) Std. Erro df (α=0.032) Lower Upper 129S1/SvImJ a/j 0.05 0.04 1 0.214 -0.03 0.13 balb/c 0.01 0.044 1 0.81 -0.08 0.1 c3h/hej 0 0.046 1 0.947 -0.09 0.09 c57bl/6j 0.13 0.034 1 0 0.07 0.2 dba/2j 0.13 0.044 1 0.003 0.04 0.22 fvb/nj 0.05 0.045 1 0.295 -0.04 0.13 nod/ltj 0.02 0.037 1 0.673 -0.06 0.09

a/j balb/c -0.04 0.038 1 0.305 -0.11 0.04 c3h/hej -0.05 0.04 1 0.19 -0.13 0.03 c57bl/6j 0.08 0.025 1 0.001 0.03 0.13 dba/2j 0.08 0.037 1 0.033 0.01 0.15 fvb/nj 0 0.038 1 0.949 -0.08 0.07 nod/ltj -0.03 0.028 1 0.231 -0.09 0.02

balb/c c3h/hej -0.01 0.044 1 0.758 -0.1 0.07 c57bl/6j 0.12 0.032 1 0 0.06 0.18 dba/2j 0.12 0.042 1 0.005 0.04 0.2 fvb/nj 0.04 0.043 1 0.399 -0.05 0.12 nod/ltj 0 0.034 1 0.889 -0.06 0.07

c3h/hej c57bl/6j 0.14 0.034 1 0 0.07 0.2 dba/2j 0.13 0.044 1 0.003 0.05 0.22 fvb/nj 0.05 0.045 1 0.267 -0.04 0.14 nod/ltj 0.02 0.037 1 0.616 -0.05 0.09

c57bl/6j dba/2j 0 0.031 1 0.923 -0.06 0.06 fvb/nj -0.09 0.033 1 0.009 -0.15 -0.02 nod/ltj -0.12 0.02 1 0 -0.16 -0.08

dba/2j fvb/nj -0.08 0.043 1 0.054 -0.17 0 nod/ltj -0.11 0.034 1 0.001 -0.18 -0.05

fvb/nj nod/ltj -0.03 0.035 1 0.369 -0.1 0.04

specc1-wt specc1-ko -0.09 0.044 1 0.036 -0.18 -0.01

c57bl/6j specc1-ko -0.07 0.038 1 0.048 -0.15 0

c57bl/6j specc1-wt 0.02 0.026 1 0.522 -0.03 0.07

52 High throughput phenotyping of avoidance learning in mice discriminates different genotypes and identifies a novel gene 2 53 ig. S 0.129 =0.041) 0.015 0.102 0.051 0.031 0.012 0.377 0.710 0.018 0.354 0.297 0.057 0.032 0.895 0.001 0.001 0.594 0.205 0.036 0.430 0.364 0.034 0.007 0.002 0.006 0.006 0.004 0.004 0.000 0.000 0.000 α ( F 1.115 1.322 7.359 7.395 1.243 1.726 7.545 3.433 5.253 2.031 5.974 5.657 11.175 2.925 2.475 1.087 2.910 7.984 0.017 1.066 4.922 1.860 2.799 13.515 0.636 0.549 2.408 11.425 0.290 12.593 21.428 ean 1.821 0.121 5.170 quare 2.341 5.637 0.167 0.571 0.124 0.471 2.108 2.632 0.776 4.456 0.259 0.703 0.705 4.426 0.001 0.228 0.892 2.096 4.067 2.096 4.805 0.805 0.029 0.082 0.244 0.620 0.042 0.644 M S df 3.717 3.617 3.932 3.474 4.275 1, 283 4.222 2.607 4.224 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 7.000 7.000 7.000 34.142 29.100 4, 1174 4.9, 1380 4.9,

um S

III quares 1.821 0.121 5.170 2.341 0.571 2.108 2.632 1.042 2.774 2.992 2.328 6.242 4.924 2.296 0.962 5.464 8.828 4.456 0.707 4.426 0.001 S 2.096 4.805 0.029 0.082 0.866 0.620 0.042 14.129 23.383 22.524 ype of T

phase Dark vs. Light vs. Dark Dark vs. Light vs. Dark Dark vs. Light vs. Dark Dark vs. Light vs. Dark Dark vs. Light vs. Dark Dark vs. Light vs. Dark Dark vs. Light vs. Dark Dark vs. Light vs. Dark Dark vs. Light vs. Dark days overall overall overall overall overall overall overall overall overall overall overall overall overall overall overall overall overall overall D1 vs. D2 D1 vs. D2 D1 vs. D2 D2 vs. D3 D5 vs. D6 D5 vs. D6 D5 vs. D6 D3 vs. D4 D4 vs. D5 D4 vs. D5 D3 vs. D4 D4 vs. D5 D4 vs. D5 Resting time in the shelter after using the sanctioned entrance: Repeated measure ANOVA (using 10: S

ource

phase days * phase phase

phase * Strain_R days * phase Strain_R * phase * days days * Strain_R

days days days phase

phase days phase days S days * phase days * phase

days * phase

all strains all 129S1/SvImJ A/J BALB/C C3H/HeJ able

T Greenhouse – Geisser correction) revealed a significant main effect of days (F(4, 1174)=13.515, p<0.0001)significant and a interaction between the days 1.86, = p<0.004).and No 1174) theeffects strains (F(29.1, of the light and dark phase were observed for this aversion index. Importantly, 5 strains show a significant effect of days on the time spent in the illuminatedthe sanctioned shelter entrance) (after either using already between days 4 and 5 (129S1, p =0.032; p<.0001),C57, or between days 5 and 6 (129S1, DBA, C3H, with p-values of respectively: p<0.01, p<0.01, p=0.034). BALB,A/J, NOD, however, did not change their time spend in shelter after using the preferred entrance over the 2 days of the experiment. Table S10: Continued

Type III Sum Mean Sig.

Source days phase of Squares df Square F (α=0.041) days overall 14.599 4.323 3.377 10.810 0.000 D4 vs. D5 7.303 1.000 7.303 32.806 0.000 phase overall Dark vs. Light 0.379 1.000 0.379 7.178 0.009 C57BL/6J days * phase overall 1.927 4.762 0.405 2.187 0.058

days overall 9.587 3.466 2.766 3.769 0.009 D3 vs. D4 1.976 1.000 1.976 5.352 0.027 D5 vs. D6 3.384 1.000 3.384 10.425 0.003 phase overall Dark vs. Light 0.256 1.000 0.256 4.393 0.044 DBA/2J days * phase overall 2.226 4.383 0.508 1.734 0.139 D5 vs. D6 Dark vs. Light 3.658 1.000 3.658 4.971 0.032

days overall 3.407 4.162 0.819 2.532 0.042 phase overall Dark vs. Light 0.022 1.000 0.022 0.453 0.507 days * phase overall 2.095 3.911 0.536 3.046 0.021

FVB/NJ D1 vs. D2 Dark vs. Light 3.274 1.000 3.274 7.489 0.011 D2 vs. D3 Dark vs. Light 6.450 1.000 6.450 11.015 0.003 days overall 1.176 4.153 0.283 1.571 0.185 phase overall Dark vs. Light 0.001 1.000 0.001 0.010 0.922

NOD/LtJ days * phase overall 0.060 4.152 0.014 0.080 0.990 days overall 1.746 3.542 0.493 1.118 0.352 days * Strain overall 4.014 3.542 1.133 2.570 0.052 D4 vs. D5 1.235 1.000 1.235 6.925 0.016 phase overall Dark vs. Light 0.026 1.000 0.026 0.596 0.449 phase * Strain overall 0.003 1.000 0.003 0.026 0.873

SPECC1-WT vs. KO days * phase overall 1.708 4.257 0.401 1.892 0.115 days * phase * Strain overall 0.262 4.257 0.062 0.290 0.893 days overall 1.975 2.964 0.666 0.987 0.412 phase overall Dark vs. Light 0.019 1.000 0.019 0.352 0.566

SPECC1-KO days * phase overall 1.439 3.214 0.448 1.656 0.193

days overall 3.785 2.986 1.267 3.373 0.031 D4 vs. D5 1.956 1.000 1.956 27.691 0.000 phase overall Dark vs. Light 0.008 1.000 0.008 0.245 0.631

SPECC1-WT days * phase overall 0.531 3.294 0.161 0.567 0.656

54 High throughput phenotyping of avoidance learning in mice discriminates different genotypes and identifies a novel gene 2 55 C 1 5 7 5 2 4 13 19 +B+ A 1 1 5 2 2 2 0 0 otal C T 5 7 5 7 2 9 14 20 B+ C 1 3 7 + 9 4 25 10 20 A ombination of responses +B 3 8 9 11 12 21 1 23 56 0 0 0 0 0 0 0 C A entries umber of N ) C

8 (32%) 6 (32%) 15 (36%) 13 (45%) 13 (46%) 32 (37%) 10 (37%) 21 (60%) outside ( outside 1 1 5 2 0 0 0 0 time increase in resting resting in increase shelter 7 (37%) 13 (45%) 15 (54%) 17 (63%) 14 14 (56%)

23 (66%) 64 (74%) 29 (69%) decrease in time spent (B) 1 1 1 5 2 0 0 0 utside sleep ) O A 5 (26%) 9 (33%) 13 (52%) 12 (43%) 32 (91%) 18 (62%) 28 (67%) 68 (79%)

decrease in preference ( 1 1 0 0 0 0 0 0 istance moved D 19 35 27 25 29 28 42 86 (100%) roup size G Number of outliers per strain (>3x standard deviation of the strain mean) for the 4 main parameters. Numbermiceofthat changed significantly (A) their preference, (B) their time spent(C) and their time 12: 11: S S able able NOD/LtJ FVB/NJ DBA/2J C57BL/6J C3H/HeJ BALB/C A/J T resting outside between day 4 and 6. The right part of the table shows the number of mice per strain showing a combination of the 3 differentresponses. This table is complementaryto theVenn diagrams of fig 6/iii. T 129S1/SvImJ C3H/HeJ C57BL/6J BALB/C DBA/2J A/J FVB/NJ 129S1/SvImJ NOD/LtJ

Chapter

Heterozygous Munc18-1 mice exhibit an increased anxiety-like phenotype but no cognitive impairment 3 Grégoire Maroteaux1, Torben Hager1,4, Robert Hensbroek5, Maarten Loos4, Bastijn Koopmans4, Jiun Youn1, Sophie van der Sluis1,3, Oliver Stiedl1,2 and Matthijs Verhage1,3,5#

1Department of Functional Genomics, 2Department of Molecular and Cellular Neurobiology and 3Department of Clinical Genetics; Center for Neurogenomics and Cognitive Research (CNCR), Neuroscience Campus Amsterdam; VU University Amsterdam and VU Medical Center, 1081 HV, Amsterdam, The Netherlands 4Sylics (Synaptologics BV), Burmanstraat 7, 1091 SG, Amsterdam, the Netherlands 5Rudolf Magnus Institute for Neurosciences, University Medical Center Utrecht, Universiteitsweg 100, Utrecht 3584 CG, The Netherlands

58 Heterozygous Munc18-1 mice exhibit an increased anxiety-like phenotype but no cognitive impairment 3 59 Munc18-1 Munc18-1 heterozygous , haploinsufficiency Munc18-1 STXBP1 , MUNC18-1 haploinsufficiency in mice does neitherprovoke , epilepsy, genetic background, flanking genes abolishes synaptic transmission, resulting in postnatal Munc18-1 knockout mice. Human munc18-1 Munc18-1 Munc18-1 heterozygous mice as potential model for the human disease. Munc18-1 Haploinsufficiency, : protein controls syntaxin-1 localization and SNARE complex formation, making it an stract Ab Munc18-1 essential protein for the precise regulation of synaptic vesicle secretion in The mammalian brain. complete deletion of death in homozygous visually detectable epileptic seizures nor cognitive deficits. Keywords heterozygous mice showed no epileptic symptoms C57BL/6J background.in a 99.6% However, novelty-induced activity and anxiety-like behaviorincreased in both tested classical in knockout and a floxed familiarmice. Furthermore, environment were is responsible for early infantile epileptic encephalopathy (EIEE), Othahara syndrome and also severe intellectual deficits without epilepsy. Using anclassicalwebehavioral tests,extensiveof range wide behaviorala phenotypingcageandtest throughputhome battery, including high- investigated mice showed an increased anxiety phenotype coping response in without classical attentional tests or spatial as learning well impairment.haploinsufficiency as To inconclude, a mice increases modified anxiety fear and promotes a more activeClassical stress coping style. knockout flanking genescompared to floxed mice. However, enhance the anxious phenotype in novelty-response Introduction The regulation of synaptic vesicle secretion in mammalian brain synapses is precisely regulated and involves specific interactions between SNARE protein and Sec1/Munc18 protein. MUNC18-1 is a highly conserved protein, which was identified in mammals by (Hata et al., 1993). It is a key regulator of all fusion reactions (Südhof and Rothman, 2009), its expression is mainly localized in neurons and neuroendocrine cells (Garcia et al., 1994; Pevsner et al., 1994), where it plays the role of molecular chaperon for syntaxin-1 proteins allowing their appropriate localization and expression at the pre-synaptic button (Arunachalam et al., 2008; Malintan et al., 2009; McEwen and Kaplan, 2008; Medine et al., 2007; Rowe et al., 1999). MUNC18-1 controls SNARE-complex formation, and thus, the fusion of vesicles to the presynaptic membrane (Rizo and Südhof, 2002; Südhof and Rothman, 2009; Toonen and Verhage, 2007). The salient role of MUNC18-1 was evidenced by completely abolished synaptic transmission resulting in postnatal death of Munc18-1 knockout mice (Verhage et al., 2000). De novo heterozygous mutations in the human MUNC18-1 gene, STXBP1, cause early infantile epileptic encephalopathy (EIEE) or Othahara syndrome (Saitsu et al., 2008) which are characterized by tremors, hypotonia and hyperventilation, but also severe intellectual deficits without epilepsy (Hamdan et al., 2009). Those clinical studies suggest that MUNC18-1 haploinsufficiency is the underlying cause of these different symptoms. Until now, Munc18-1-deficient and heterozygous mice have been intensively studied in in vitro approaches to unravel its role in presynaptic neurotransmitter release. Heterozygous mice seem to be a potential model for the human haploinsufficiency phenotype. To test the effects of haploinsufficiency of Munc18-1 for behavioral functions, we performed a comprehensive behavioral screening of Munc18-1 HZ mice, using an extensive behavioral test battery covering a wide spectrum of behavioral paradigms including anxiety, emotional and spatial learning tests. Additionally, animals were studied in their home cage for 6.5 days without interruption or human interference to thoroughly investigate spontaneous behavior as well as avoidance learning.

Material & methods Subjects A Munc18-1 mice line was generate as described (Verhage et al., 2000). In total 5 batches of 23 to 26 animals per group with 12±1 per genotype (WT vs. HZ). Mice were aged between 8 to 12 weeks. To generate conditional Munc18-1 mice (Heeroma et al., 2004), floxed Munc18-1 mice were generated by insertion of LoxP sites flanking the second exon using homologous recombination. Mice were bred as Munc18-1 heterozygous by using standard mouse husbandry, and backcrossed for at least ten generations to a C57BL/6J background (C57BL/6JCrl, Charles River). All experiments were approved by the local animal research committee and complied with the European Council Directive (86/609/EEC).

Automated home cage observation and data analysis Mice were transferred to specially designed home cages (PhenoTyper model 3000, Noldus Information Technology, Wageningen, The Netherlands) in the second half of the subjective

60 Heterozygous Munc18-1 mice exhibit an increased anxiety-like phenotype but no cognitive impairment 3 61

TM ad male male mice Munc18-1 male mice were male mice (12 weeks Munc18-1 Munc18-1 -Y coordinates-Y of the center malemice were subjected to X Munc18-1 Munc18-1 T 4.1, Noldus Information Technology, Wageningen, The X . All tests were performed during the light phase in the following order, starting from Additional groups of mice were used to repeat or perform new tests. 24 est battery tatistical analysis Netherlands) as described in detail previously subjective (Maroteaux dark phase (19:00 h). etResulting track al.,files, containing 2012) starting at the first light phase (14:00 h – 17:00 h). The behavior of mice was video-tracked for three days (EthoVisiondays threevideo-tracked forwas mice ofbehavior The h). 17:00 – h (14:00 phaselight HTP 2.1.2.0, based on EthoVision of gravity (COG) at a resolution of 15 images per second, were processed using AHCODA analysis software (Synaptologics BV, Amsterdam,parameters. 115 activity The parameters were Netherlands)generated as described in to detail previously extract al., (Loos2014). Theet last and behavioralfirst 10 min of each dark and light phase were not included in summary staticsto ensure that a potential asynchrony of the data streams and light regime in the testing facility would not affect these statistics. T The test battery was performed on 26 male mice (13 weeksold. Mice WTwere acclimated theirto newand housing for one week 13prior testing.to They HZ)were aged between 8 and 10 kept under a 12-h light-dark cycle (light phase on at 7 a.m.) with access to libitum food and water the least stressful: Weighing, grip strength meter (GSM), elevated plus maze (EPM), open field (OF), dark-light box (DLB), and Barnes maze (BM). mice were subjected to a fear conditioning (FC) experiment. 24 floxed the 5-choice serial reaction time task (5-CSRTT) (Table 1). A full description of the providedmethods in Chapter Material is7: & Methods. S Before any analysis was performed, data were examined for outliers (>3 times the SD from the strain mean) and outlier data were removed. All statistical analyses were performed using IBM SPSS statistic 20 (IBM corporation, Armonk, NY, USA). Genotype differencesusing wereparametric compared tests (T-test, ANOVA, repeated-measures ANOVA) whenever normality and homoscedasticitycriteria weremet. Otherwise, nonparametric tests wereperformed (Kruskal- Wallis, Mann-Whitney U-test). Nonparametric data are presented as box plots (ends of the box denoting the 25 and 75% interquartile range and the whiskers providing the quartileupper and ±1.5 lower times the interquartile range, respectively, while the line in themedian). errorprobabilityAnacceptedwas statisticallybox levelasp<0.05 of significantdenotes throughout the the study of the classical behavioral tests. For all givenstatistical level ofanalysis analysis was of basedPhenoTyper data,on estimated FDR corrected by minimum(Verhoeven positive FDR with eta threshold set at 5%. al., 2005), alpha-levels were (12 WT, 12 HZ) went through the test battery. 12 HZ and 12 WT old) were subjected to a passive avoidance test. 11 HZ against 12 WT comparedinthe modified Barnes maze. 11 HZ and 12 WT Table 1: Behavior tests, sequence and number of mice used.

Experiment classical KO mice floxed mice WT HZ WT HZ Comments/duration/sequence Body weight 13* 13* 12* 12* age: 7-8 weeks Novelty induced hypophagia 13* 12(1)* 10(2)* 11(1)* 3 days habituation prior test, 10min max Grip strength 13* 13* 12* 12* 5 session front paws Elevated plus maze 13* 13* 10(2)* 12* 5 min Open field 12(1)* 12(1)* 12* 12* 10 min Dark-light box 12(1)* 12(1)* 12* 12* 10 min Rotarod 13* 13* 12* 12* 10 trials over 2 days Fear conditioning 11(2)* 12(1)* - - Training, context-, tone- memory Acoustic startle 13* 13* 12* 12* 260 trials pseudo randomized, acoustic startle & pre-pulse inhibition (65-115db), pulse 120db pre-pulse (65-115db) Passive avoidance 12 12 - - Training, retention & extinction tests Barnes maze 13* 13* 12* 12* 2 trials per day for 6 days+ reversal or 1 weeks delay Modified Barnes Maze 12 11 - - performed at reversed day-night cycle 5-Choice SRTT 12 12 - - food restricted to 90-95% of initial weight Forced swim test 13* 13* 12* 12* 10 min and 6 min sessions automated home cage 64 59 12* 12* 7 days protocol no human interference

*same mice as used in previous test; (x) number of outliers;

Results Munc18-1 heterozygous mice (HZ) are viable and show no obvious phenotypic or morphological abnormalities compared to their control littermates. Their breeding follows a Mendelian distribution (breeding WT female x HZ male gave 43% of heterozygous mice, n=491). In a routine screening for general health, 8-week-old HZ mice were significantly lighter (21±0.3 g) than their wild-type (WT) littermates (24±0.3 g) (T(24)=6.357, p<0.0001) (Fig. S7). Cellular MUNC18-1 protein levels were analyzed by Western blot analysis and revealed a gene dose effect, producing decreased MUNC18-1 level, relative to control littermates (100%) (Toonen et al., 2006). and confirms the haploinsufficiency in these mice. At postnatal day 56 (8 weeks) no gross abnormalities were observed in HZ mice that would confound behavioral phenotypes in Munc18-1 HZ mice. A grip strength meter test measuring fore limbs strength revealed no difference between genotypes (see Suppl. Fig. S8).

Munc18-1 haploinsufficiency influences activity patterns and anxiety traits in mice We used an automated home cage (PhenoTyper) paradigm to analyze spontaneous behavior in Munc18-1 HZ mice (as described in (Loos et al., 2014)) and studied 115 parameters divided in 6 main categories (see Table S13). Munc18-1 HZ mice were significantly different from WT mice in 49 out of 115 parameters, especially, habituation and activity parameters and the tendency to climb on

62 Heterozygous Munc18-1 mice exhibit an increased anxiety-like phenotype but no cognitive impairment 3 63 HZWT HZ HZ HZWT gene

2 1 0 duration (h) duration

HZWT Munc18-1 visit shelter Long Dark Light C Munc18-1 0

80 60 40 20

120 100 OnShelter zone visits visits zone OnShelter HZWT H HZWT

0.8 0.6 0.4 0.2 0.0 habituation ratio dark ratio habituation Activity duration - duration Activity HZWT B Dark Light HZ mice had a 2-fold lower time on the the on time lower 2-fold a had mice HZ

0 HZ WT

40 30 20 10 duration (min) (min) duration

HZ mice and WT were littermates detected OnShelter zone OnShelter Munc18-1 G 3 mice showed significant differences with their HZWT Munc18-1 = 59, ** p<0.01, ***p<0.001= 59, HZ Munc18-1

1.0 0.8 0.6 0.4 0.2 0.0 light Index light Activity dark Activity 2 F Days = 64 and N WT HZWT

heterozygote model without flankinggene variation. In these mutants, 3 2 dark - (h) 1 0 Time in feeding zone feeding in Time E 1 -1 -1 HZ mice show activity differences from WT mice during habituation and in the light-dark Munc18 HZ mice also showed a difference in their activity balance between the dark and light light and dark the between balance activity their in difference a showed also mice HZ HZWT strain, derived from an independent ES cell targeting experiment using pure C57BL/6J Munc18 0

0

30 20 10

80 60 40 20 duration (s) - dark dark - (s) duration

Munc18-1 (m) moved Distance These findings werereproduced in an independent approach by the analysis of the floxed Mean long arrest long Mean igure 8: A D Munc18-1 shelter compared to WT mice (p=0.0001 and p=0.0002, Fig. 8G and H). and 8G Fig. p=0.0002, and (p=0.0001 mice WT to compared shelter background (Heeroma et al., 2004). In these mutants, a single copy of the was inactivated using cross breeding with a Cre-deleter strain and generating bred a further to C57BL/6J, 32 out of 115 spontaneous behavior parameters were significantlyHZ differentfloxed. between The heterozygous WT floxed and control littermates, that are comparable with the classical HZ mice, in 15 parameters regarding phasecomparison based theonspontaneous activity.Distance movedbins thein 1-h inPhenoTyper acrossdays 3 afterplacement indicates an initial significantly increased activity in HZ versus WT mice (A). The habituationratio (between day 1 and 3) during the dark phase was significantly higher in WT than in HZ mice (B). The long sheltervisit duration wassignificantly increased in HZ than in WT mice(C). Long arrest duration was increased in average inHZcompared WTto during thedark phase (D).Thetime spent inthefeeding zonewas increased inHZ(E). The dark/light activity index was increased in HZ compared to WT (F). The time spent (G) and number of visits (H) on shelter were significantlyreduce in HZ comparedto WT. Statistics were done on log10 transformed data. Graphuse mean and SEM of raw data. N F in in habituation parameters. HZ mice showed increased activity during the first steeper a few and hours (habituation) 3 day inand 1 day between the activity their in difference (HZ: larger a cage, visits home shelter longer made also mice HZ B). 8A, Fig. (p<0.0001, duration activity their in decrease of home the exploration during arrests longer 8C), showed Fig. p=0.001, min; 80±2 WT: min, 96±4 Fig. (p=0.001, zone feeder the in longer stayed and 8D) Fig. p=0.001, s; 25.1±1 WT: s, 29.5±1 (HZ: cage 8E). observed was difference profound most The phase. dark the during active more relatively were mice shelter. of top on climb to tendency the regarding the shelter. Profound between differences the Profound shelter. completed, are habituation of aspects most after 3, day On 8F). Fig. (p=0.0006, phase the kinematics phenotype (distance and the velocity of the movements, long and short) and in the activity (aversion for the top of the shelter). However, floxed mice displayed a smoother dark and light difference as well as a less contrasted habituation. Taken together the spontaneous behavior data suggest that classical HZ mice might be more susceptible to novelty e.g introduced in the cage and favored activity during the dark phase (light is more anxiogenic) those characteristic are not found in the floxed HZ mice. However, both classical and floxed HZ mice avoided to climb on the shelter which is a more exposed zone than other cage areas. These three main differences suggest an enhanced anxiety-like phenotype in Munc18-1 HZ mice.

Munc18-1 haploinsufficiency increases anxiety-like behavior in mice As the integrated phenotyping in the home cage suggested anxiety-related phenotypes, we subsequently tested Munc18-1 HZ mice in classical anxiety tests in the elevated plus maze, an open field and a dark-light box. The haploinsuffiency of Munc18-1 did not affect general activity on the elevated plus maze (HZ: 28.6±0.82 m; WT: 27±1.12 m) or open field (HZ: 44.7±3 m; WT: 38±2.2 m). However, in each of these three tests, HZ mice showed a delayed latency to enter anxiety-associated zones, the open arms in the elevated plus maze, the center zone of the open field arena and the bright compartment in the dark-light box. Moreover, HZ mice spent less time in, made fewer visits and moved less than their control littermates in these anxiety-associated zones (Table. 2). Especially in the dark-light box, Munc18-1 HZ mice showed increased anxiety-like behavior compared with their WT controls. In addition to these classical tests, we also tested how soon after being introduced into a novel home cage Munc18-1 HZ and WT mice started to consume a cracker (novelty-induced hypophagia). Interestingly, in this case of positive reinforcement, HZ and WT showed the same delay. Similar results were obtained with Munc18-1floxed HZ mice on elevated plus maze: a longer latency to visit the open arms, fewer entrances and less time spent on the open arms. However, Munc18-1 floxed HZ mice did not show an anxiety phenotype in dark-light box and open field test (see Table S13). Taken together, these results reveal an increased anxiety phenotype in Munc18-1 HZ mice, which is stronger in classical Munc18-1 HZ than in floxed HZ mice and no anxiety phenotype when positive reinforcement was used.

Munc18-1 haploinsufficiency affects fear response in mice Subsequently we analyzed fear learning and subjected the mice to a contextual and auditory fear conditioning task (FC) which involves the amygdala and the hippocampus. Mice were trained to associate a specific environment (conditioning context) and a tone with a foot shook. The locomotor activity and the rate of exploration were measured partly during training and when mice were re-exposed 24 h later to the chamber, where they previously received the shock, then 2 h later to a different chamber first without and then with tone exposure. The two genotypes showed no significant differences in baseline activity (p=0.98) and exploration rate when placed in the fear conditioning box for the first time. No hearing or nociception impairment was detected in HZ mice during the training phase. The reactions to the 30-s tone or the 2-s foot shock did not differ between the two genotypes. However, during the context-dependent memory test, 24 h later, HZ mice showed a significantly higher exploratory activity (p=0.006) than WT mice.

64 Heterozygous Munc18-1 mice exhibit an increased anxiety-like phenotype but no cognitive impairment 3

65

*significance under 0.05 under *significance

0.867 22 0.2 65.7 360.1 12 70.6 376.4 12 (s) Latency hypophagia Novelty-induced

0.015* 22 -3 11 52.9 11 7 19.5 13 (s) BC enter to Latency

0.001* 22 4 1.34 6.5 11 1.42 14.4 13 (m) moved Distance

0.000* 22 4.5 2.4 17.5 11 3.3 36.5 13 (n) entries BC

0.000* 22 5.3 18.7 84.3 11 25.9 257.9 13 (s) comp. bright in Time box Dark-light

0.018* 22 2.6 0.4 1.6 12 0.2 2.8 12 (%) distance Center

0.273 22 1.1 0.22 0.8 12 0.13 1.08 12 (m) moved distance Center

0.09 22 -2 3.03 44.7 12 2.2 38 12 (m) moved distance Total

0.233 22 1.2 1.3 6.3 12 1 8.3 12 (n) center Visits

Open field Open Latency to enter center (s) center enter to Latency 0.025* 22 -2 41.8 257.4 12 28.9 135.3 12

Open arm visits (n) visits arm Open 0.016* 24 2.6 1.3 5.2 13 1.7 10.8 13

Latency to open arms (s) arms open to Latency 0.000* 23 -4 38.6 217.3 13 5.5 52 13

Duration on open arms (s) arms open on Duration 0.008* 24 2.9 6.5 21.4 13 10.8 57.7 13

Elevated plus maze plus Elevated Distance moved (m) moved Distance 0.22 24 -1 0.82 28.6 13 1.12 27 13

est T easure M p-value df t M . E . S ean M N M . E . S ean M N

T W Z H -test T

Anxiety measures in mice were increased by by increased were mice in measures Anxiety 2: able T heterozygous mutation. heterozygous Munc18-1 There was no difference in the novel environment 2 h after the context-dependent memory test. However, HZ mice were again significantly more active when the tone was presented (p=0.006). The tone-induced activity effect was less strong than that of the context. No significant differences were found in exploration during the tone-dependent memory phase (Fig. 9A-B). Furthermore, we tested associative emotional learning using a passive avoidance paradigm. Mice were placed in a box with two compartments separated by a door. During training, mice were placed into the bright comportment. After 60 s, the door between the two compartments opened to allow a mouse to enter the dark compartment (DC). The door was closed as soon as a mouse entered the dark compartment. Subsequently, mice experienced a 2-s foot shock. In the first retention test 1 (R1), 24 h later, none of the WT animals entered the DC while HZ mice had a median transfer latency of ~400 s. Again 24 h later all mice were manually placed (forced exposure) in the DC without delivering a shock for a total of 8 min (minus the time spent in the dark compartment during R1) so that the total time in the DC was identical in all mice. Again 24 h later, mice were placed in the light compartment once daily for 9 days and the latency to enter the DC was measured. Generalized estimating equation (GEE) model analysis revealed a strong difference between HZ and WT in the latency to enter the DC and the time spent in the DC. HZ mice were faster to enter (p<0.001) and stayed longer (p<0.0001) in the DC compared to the WT in each retention test (Fig. 9C-D). However, both groups had the same number of visits in the DC. These data show that HZ mice experienced the foot shock as aversive but displayed faster fear extinction than WT mice. Together with the fear conditioning data, this suggests that HZ mice cope differently with an aversive situation. This may be due to a learning impairment or more active (reactive) stress-coping style.

Munc18-1 haploinsufficiency does not alter spatial learning and attention in mice Syndromes described in humans with mutations in the Munc18-1 gene typically involve mental retardation (Deprez et al., 2010; Hamdan et al., 2011; Saitsu et al., 2010). Therefore, we tested cognitive abilities of WT and HZ mice by measuring spatial learning and memory, impulsivity and attention. To evaluate spatial learning, we tested heterozygotes carrying a classical null allele as well as heterozygous floxed mice in the Barnes maze, a large round maze with 24 holes on the rim with one target hole connected to an escape box. WT and HZ showed a similar learning curve in locating the target hole by reducing the latency to enter the escape box through the habituation trials (H1) to the end of training phase (T10). In the first cohort of mice, we tested long term memory. One week after the last training (T10), mice were placed back on the Barnes maze. Both genotypes remembered the location of the escape box, they moved and took a similar amount of time to find the escape box (Fig. 10A, B). In the second cohort, reversal learning and behavioral flexibility was tested in Munc18-1 floxed mice. After nine trials the escape box was re-located to the hole diametrically opposed on the Barnes maze; both groups took more time to find the escape box on the first trial but then learned the new location faster and without genotype difference (see. Fig. S9). Furthermore, we tested Munc18-1 heterozygous mice in the modified Barnes maze due to its increased complexity (Youn et al., 2012). The modified Barnes maze assesses spatial memory with much lower possibility for the mice to develop a serial strategy. The performance in the modified Barnes maze was the same for the two genotypes. A clear learning effect was observed,

66 Heterozygous Munc18-1 mice exhibit an increased anxiety-like phenotype but no cognitive impairment 3

67

+tone New context New Munc18-1 Munc18-1

Retention

New context New Context 24h F.E

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 Training Context

0 24h 0

80 60 40 20 Exploration (%) Exploration

600 500 400 300 200 100

munc18-1 HZ Time in dark compartment (s) compartment dark in Time B

D

+tone New context New Retention

munc18-1 WT

New context New Context

24h

Post-shock Shock - US - Shock Training

F.E Tone - CS - Tone T R1 R2 R3 R4 R5 R6 R7 R8 R9 R10

Munc18-1 Munc18-1 HZ mice show lower fear to conditioned stimuli than WT controls. Activity (A) and exploration Context

0 0

50 10

Activity (cm/s) Activity

To investigate To deficit in attention and impulsivity, we trained a separate cohort of 600 500 400 300 200 100 Transfer latency (s) latency Transfer A igure igure 9: (B) (B) during fear conditioning from training to retention tests indicate no time and (C) latencies context. novel in the Transfer activity exposure tone during and test retention contextual the during differencesduring training but (R1-10) tests retention to (T) training from experiments avoidance passive during (D) compartment dark the in spent indicated significant differences between genotypes with faster extinction in HZ exposure mice. to F.E the dark refers (shock) compartment to 24 h the after forced R1, so that all mice spent a **p<0.01. total 12; = NWT NHZ *p<0.05; (C). plots box =11, SEMfor ± showmeans R2. except Data time before compartment of 8 min in the dark HZ and WT littermates in the 5-choice serial reaction did time not taskdiffer in(5-CSRTT). the numberHZ ofand sessions WTto acquiremice the task (p=0.87), whichlearning iscurves in observed line in with the Barnes the maze experiments. Attention performance in terms of responsevariability,accuracy, number omission of anderrorsdidnotdiffer between genotypes differences impulsiveNo Supplementalresponding S14).in (Fig.10F-Hand motivation and Table to execute the task were detected.impulsivity between Hence, WT and no HZ mice differences were detected inin the attention5-CSRTT. Taken together, performance Barnes and with a decrease in the latency and in the distance to reach the escape box. No difference in the reduction of the latency was observed between the two performedgenotypes. 24 h A after 1-min the probelast trial.trial Both was genotypes stayed in the target amountquadrant forof a timesimilar (Fig. 10E). Taken together these data showed allelethat the in loss of mice one does not impair cognitivememory (Fig. 10A, B) functions and behavioral flexibility (Fig. required 10C, D, E). for spatial learning, long-term F C A C 300 200 WT WT HZ HZ 250 160 200 120 150 80 Latency (s) Latency (s) 100

50 40

0 0 H1 H2 T1 T2 T3 T4 T5 T6 T7 T8 T9T10 L1 L2 H1 H2 T1 T2 T3 T4 T5 T6 T7 T8 T9 T10T11T12

B D E 12 20 60 10 16 40 8 12 20 6 % of time in the target quadrant 8 0 4 WT HZ Distance moved (m) 2 Distance moved (m) 4

0 0 H1 H2 T1 T2 T3 T4 T5 T6 T7 T8 T9T10 L1 L2 H1 H2 T1 T2 T3 T4 T5 T6 T7 T8 T9 T10T11T12

F G H

WT WT WT

HZ HZ HZ

0 20 40 60 80 100 0 2 4 6 8 0 0.2 0.4 0.6 0.8 1 Response accuracy (%) Premature responses (%) Correct response latency (s)

Figure 10: No deficit in Barnes maze and 5-choice serial reaction time test performance of Munc18-1 HZ versus WT mice. Latency (A) and Distance (B) to find the escape box of the Barnes maze with 24 peripheral holes, and latency (C) and distance (D) to find the escape box of the modified Barnes maze with 44 holes in 3 circles shown across trials in HZ and WT mice. Time spent in the target quadrant during the probe trial of the modified Barnes maze (E). Percentage of correct responses (F), percentage of premature (impulsive) responses (H) and latency for correct responses (H) in the 5-choice serial reaction time task. There was no significant difference between genotypes in any measure depicted here. H1-2: habituation trials; T1-12: training trials with escape hole, L1-2: Long-term memory tests.

maze and 5-CSRTT data demonstrate that Munc18-1 heterozygous mice are not impaired in spatial learning, memory, impulsivity or attention.

Discussion In this study we are the first to behaviorally characterize a mouse model of haploinsufficiency of STXBP 1 (MUNC18-1) causing Othahara syndrome, seizures and mental retardation in humans (Saitsu et al., 2010). We investigated the behavioral phenotype of Munc18-1 heterozygous male mice using an extensive behavioral test battery, including long-term home cage activity recording. Despite the high anxiety level observed in the anxiety tests, HZ mice exhibited an increased activity in fear conditioning as well as an increased distance moved and shorter transfer latencies in passive avoidance indicating lower fear responses and faster extinction. However, no deficits in learning and memory in the Barnes maze and the 5-CSRTT were observed.

68 Heterozygous Munc18-1 mice exhibit an increased anxiety-like phenotype but no cognitive impairment 3 69 HZ Z mice Z H Munc18-1 HZ mice show unc18-1 unc18-1 M HZ mice displayed Munc18-1 Munc18-1 HZ mice show both, increased state and trait anxiety. trait and state increased both, show mice HZ mice are still carrying in their genomeof 129Sv 0.6% haploinsufficiency effects on mouse behavior which Munc18-1 Munc18-1 Munc18-1 floxed mice did not. In the home cage behavior, classical and floxed Munc18-1 HZ mice showed 15 similar significant differences in comparison to their respective Mice are nocturnal animal and require a certain period of habituation to a novel environment, environment, a to novel of habituation period a certain require and animal nocturnal Mice are unc18-1 haploinsufficiency increased anxiety unc18-1 haploinsufficiency promotes an active coping style in response to stress enetic background and flanking genes increase novelty response of classical classical of response novelty increase genes flanking and background enetic led to increased anxiety level on the elevated plus maze inmice. classicalThe classicaland HZ floxed showed increase anxiety in the open field andwell, inwhereas the dark-light box as M In our study we investigated Munc18-1 (the light and dark between contrast a stronger exhibited HZ mice classical that fact the therefore to reaction a light (excessive phase novel is pattern and a more habituation anxiogenic) stronger environment) compared to their controls, suggests that this classical HZ mice are even than anxious the floxedHZ more mice. Thosetraits wereanxiety not in detected tests asclassical they days several over in of the PhenoTyper HZ persisted behavior the anxiety-like too short. Since are that conclude we 2012a) al., et (Fonio WT littermates such as a clear avoidance for the top of the shelter (more exposed and elevated zone) and changes in kinematics with smaller distance and lower velocity in the short and long movement events (more hesitant movement). These two main differences, commonto bothclassical and floxed HZ, takentogether with the high anxiety levels of the elevatedcan be plus interpreted maze as a higher anxiety behavior in both novel and familiar environment. M Most commonly in fear conditioning and passive avoidance, increased transfer latencyfreezing are the main criteriaand to demonstrate delayedfear memory in mice (Misane et al., 2013). However, a fear response can be expressed in rodents in active 2 different suppression ways. of i) By any freezing, movement an to avoidcoping style. attention ii) during By trying threat, to also escape calledfrom the possible, reactive threat proactive orcoping styleattacking (Koolhaas theet al., threat 2007). Inif our escapestudy, is not an increased activity and exploration rate in fear conditioning experiments, when exposed to the context or the tone, 24 h after training. Additionally, after forced exposure, HZ faster displayedfear extinction in the passive avoidance task. Together these data could be interpreted as memory deficits. However, several cognitivetests involving memory, attention, impulsivity (Barnes maze, 5-CSRTT) did not confirm this.We conclude that a more proactive way of coping in responsepanicky-like to behavior high according emotionalto the load two-tier that model resulted(Koolhaas deficit hippocampus-dependentin in et contextualal.,more fearresponses. Alternatively,2007) we cannot exclude rather than a thattherecallfearof impairedmaylead to fear expression thenegativemiceHZduetoin effect of elevated anxiety (“stress”) on cognition (Diamond et al., 2007). G An intrinsicAnconfounding effectclassical of flanking-gene knockoutthe is problem (Büeleral., et 1992; Crusio, 2004). Classical HZ background which could explain the high anxiety in DLB and OF as well as the strong contrast between dark and light activity and stronger habituation pattern compared to their controls which is not displayed by floxed HZ. However, in the home cage 15 out 115 parameters showed the same differences between HZ and WT in classical and floxed strains and 75 more showed the same direction of variation between WT and HZ (e.g.: activity duration: classical WT≥HZ; floxed WT≥HZ). Flanking genes can lead to dramatic effect on behavior, if the embryonic stem cell carries a drastic mutation (Kumar et al., 2004). Genetic background is known to influence behavior, for example fmr1KO mice in mixed background of C57BL/6J-129 showed impaired spatial learning abilities in Morris water maze (MWM) (D’Hooge et al., 1997). However, fmr1KO mice in a C57BL/6J-FVB/NJ background show no deficit in spatial learning in the MWM or the BM (Yan et al., 2004). This variation in phenotype for a same mutation in different genetic backgrounds may be due to compensatory (epistatic) mechanisms present in certain mouse strains. In our case, the presence of 0.6% of 129Sv background in classical Munc18-1 HZ mice could be responsible for the enhanced novelty-induced trait anxiety displayed in OF, DLB and PhenoTyper, however Munc18-1 haploinsufficiency is still responsible for a higher anxiety and more active coping style in fear conditioning and passive avoidance.

Epilepsy in Munc18-1 HZ is not apparent but cannot be excluded De novo heterozygous mutation of STXBP1 in humans is responsible for the Ohtahara and West syndrome with epileptic seizures (Deprez et al., 2010; Lacaze et al., 2013; Saitsu et al., 2010). Compared to other mammals, the mouse is better known genetically and easier to manipulate which makes it and excellent animal model for genetic studies of epilepsy. Genetic deletion of both alleles of Munc18-1 in cultured neurons leads to a complete loss of synaptic transmission and results in postnatal death of homozygous Munc18-1 pups (Verhage et al., 2000). A partial loss of Munc18-1 in autaptic cultures did not affect synaptic physiology under basal conditions. However, repeated stimulations increases synaptic depression due to a smaller ready releasable pool (Toonen et al., 2006). Heterozygous Munc18-1 deletion also impairs neuromuscular synaptic function (Toonen et al., 2006), although both classical and floxed Munc18-1 HZ mice, a potential model for epilepsy, in our study did not exhibit any seizure-like phenotype. Several reasons may explain this absence of epileptic seizure. First, epileptic seizures are strain- dependent with DBA/2J mice being more sensitive to induced audiogenic seizure and sudden death (Faingold and Randall, 2013). Second, seizures often lead to visible behavioral alterations such as unusual body postures and lack of motor control in combination with immobility (Gram, 1990) while mild seizures cannot be detected on the behavioral level and require EEG measurements (Platt and Riedel, 2011). However, there is a high comorbidity between epilepsy and anxiety (Kanner, 2011; Vazquez and Devinsky, 2003) and since no EEG measurements were performed in our study, we cannot exclude that Munc18-1 haploinsufficiency does not provoke epilepsy in our mouse models as in certain human cases (Saitsu et al., 2010) or that mild seizure went undetected during the behavioral tests (Platt and Riedel, 2011).

No cognitive abnormality in Munc18-1 haploinsufficient mice Many patients carrying a mutation in STXBP1 present non-epileptic movement disorders and mental retardation (Hamdan et al., 2011; Mignot et al., 2011; Milh et al., 2011). Yet our models did not show cognitive impairment in Barnes maze and 5-CSRTT experiments. However, mice

70 Heterozygous Munc18-1 mice exhibit an increased anxiety-like phenotype but no cognitive impairment 3 71 HZ mice are more in an organism. Munc18-1 protein appears to be sufficient Munc18-1 3 display increased anxiety behavior in α 1, MUNC18-1 α that lead to severe mental retardation and sometimes leads to an increaseananxietyleads tobehavior withoutcognitive Munc18-1 Munc18-1 mutations affect only one copy of the gene which makes them all , the reduced amount of on a wide range of mouse behaviors. In humans, clinical studies show de novo Munc18-1 Munc18-1 onclusions to maintain most of the cognitive function in the mice. However, epileptic seizure (Deprez et al., 2010; Hamdan al.,et 2008). al., These 2011, 2009; Milh et al., 2011; autosomal Saitsudominant. In et our mouse model, despite deletion of one copy of the gene, healthyi.e. one copy of anxious and show a more proactive coping strategy when facing a more severe stressor. As for severalAlzheimer models (Radde etal., 2008), ourmodel does notmimic thehuman condition but is still of great value to understand the importance of 19 different point mutations in C In summary, the present study provided a detailed analysis of singlethe allele effectof of the deletion of a are commonly used as cognitive show deficitimpaired spatial models, learning in forMWM and example BM as Neurogranin/RC3well et al.,as 2001).anxiogenic knockoutHeterozygous phenotypemice for (MiyakawaNa,K-ATPase elevated zero maze and impaired spatial learning and memory in MWM (Moseley et al., 2007). CamKII knock-out mice show a learning deficit ininactivation singleMWMcopyofa of (Giese et al.,1998) In our case the impairment. Three reasons prevents us to conclude on this part, first there a publication bias, to our knowledge no inconclusive studies of mice models though to impairmentbe models of are cognitive published in those background carries some terms. compensatory mechanisms Second,and third, the human phenotype wecan be cannot due to a higher rule complexity of cognitive functions. out that the genetic Supplemental

0.07

0.06

0.05

0.04

(N/g) 0.03 WT 0.02 HZ

0.01

Fore limbs strength / body weight 0 1 2 3 4 Measurement

Figirue S7: Munc18-1 heterozygous (HZ) mice has a significantly lower body weight than their wild-type (WT) controls. Mice were weighed before the start of the test battery. HZ mice were approximately 3 g lighter than their WT littermates.

The age range was 8±10 weeks (mean ± SEM) and did not differ between genotypes. NWT=12; NHZ=12; ***p<0.001.

25 ***

20

15

Weight (g) Weight 10

Figirue S8: Munc18-1 HZ and WT mice had no significant differences in grip strength. The force that was applied 5 to disconnect the fore limbs of mice from the grid they were holding on to, was tested 4 times consecutively (at 10 s interval), HZ mice showed no difference compared to WT mice (interaction measurement x genotype: F(3,72)=2.211; p=0.094). N =12; N =12. 0 WT HZ WT HZ

72 Heterozygous Munc18-1 mice exhibit an increased anxiety-like phenotype but no cognitive impairment 3

73

HZ WT habituation trials; T1-9: training trials; R1-7: reversal trials. N trials. reversal R1-7: trials; training T1-9: trials; habituation =12. N =12;

genotype on the learning curves (trials: F(3.2,67)=3.7; p=0.014, trials x genotypes: F(3.2,67)=2.7; p=0.47) and thus HZ appeared to relearn the new escape location faster. H1-2: H1-2: faster. location escape new the relearn to appeared HZ thus and p=0.47) F(3.2,67)=2.7; genotypes: x trials p=0.014, F(3.2,67)=3.7; (trials: curves learning the on genotype

p=0.043, trials x genotype: F(3,59)=1.13; p=0.342). The fact that HZ showed more variation in the distance traveled to find the escape hole, results in an interaction effect of the the of effect interaction an in results hole, escape the find to traveled distance the in variation more showed HZ that fact The p=0.342). F(3,59)=1.13; genotype: x trials p=0.043,

p=0.285). HZ and WT learned the new location of the escape hole at similar speed and decreased their latency to find the escape hole at the same rate (trials: F(2.8,58.73)=2.954; F(2.8,58.73)=2.954; (trials: rate same the at hole escape the find to latency their decreased and speed similar at hole escape the of location new the learned WT and HZ p=0.285).

(trials: F(1,22)=11.304; p=0.003, trials x genotype: F(1,22)=1.54; p=0.228) as well as in the distance to find the escape box (trials: F(1,22)=5.123; p=0.034, trials x genotype: F(1,22)=1.2; F(1,22)=1.2; genotype: x trials p=0.034, F(1,22)=5.123; (trials: box escape the find to distance the in as well as p=0.228) F(1,22)=1.54; genotype: x trials p=0.003, F(1,22)=11.304; (trials:

trial (T9) the escape hole location was changed to the opposite side of the table. A similar increase was observed in both groups between T9 and R1 in the latency latency the in R1 and T9 between groups both in observed was increase similar A table. the of side opposite the to changed was location hole escape the (T9) trial 9 the After

th

with declining latency to enter the escape box (left panel; F(4.5,94.35)=16.812 ; p<0.0001) and distance traveled to find the escape box (right panel; F(5.05,106) = 9.61; p<0.0001). p<0.0001). 9.61; = F(5.05,106) panel; (right box escape the find to traveled distance and p<0.0001) ; F(4.5,94.35)=16.812 panel; (left box escape the enter to latency declining with

floxed HZ and WT mice did not differ in spatial learning on the Barnes maze. During the learning phase both genotypes showed similar learning curves curves learning similar showed genotypes both phase learning the During maze. Barnes the on learning spatial in differ not did mice WT and HZ floxed Munc18-1 9: S igirue F

Trials Trials

1H 1T 3T 5T 7T 9R 2R 4R 6R7 R6 R5 R4 R3 R2 R1 T9 T8 T7 T6 T5 T4 T3 T2 T1 H2 H1 1H 1T 3T 5T 7T 9R 2R 4R 6R7 R6 R5 R4 R3 R2 R1 T9 T8 T7 T6 T5 T4 T3 T2 T1 H2 H1

0 0.0

0.2

50

0.4 Distance (m)

Latency (s)

100

0.6

150

0.8

WT

200

HZ 1.0

250 1.2

Learning Habituation Reversal Learning Habituation Reversal 74

Table S13: Munc18-1 floxed mice showed increased anxiety only in elevated plus maze. Although no difference was detected in locomotor activity, HZ mice performed poorly in the EPM compared to WT mice. HZ mice showed an increased latency to enter , spent less time and less visits of open arms. In the open field, dark-light box and novelty-induced hypophagia HZ floxed mice did not differ from WT mice.

Wild-type mice Heterozygous mice T-test Test Measure (unit) N Mean S.E.M N Mean S.E.M T DF P value Elevated plus maze Distance moved (m) 10 16.8 0.96 12 16.1 1.54 -0.4 20 0.73 Duration open arms (s) 10 48.57 6.65 12 16.87 4.67 -4 20 0.001* Latency in open arms (s) 10 16.02 3.98 12 102.6 24.06 3.24 20 0.004* Visits of open arms (n) 10 11.5 1.44 12 3.83 0.81 -4.8 20 0.000*

Open field Latency center (s) 12 78.77 9.66 12 102.3 18.89 1.11 22 0.28 Visits in the center (n) 12 17 2.23 12 14.58 1.38 -0.9 22 0.368 Total distance moved (m) 12 61.2 2.56 12 60.37 2.55 -0.2 22 0.823 Distance in center (m) 12 2.11 0.3 12 1.99 0.2 -0.3 22 0.735 Distance in center (%) 12 3.42 0.4 12 3.26 0.24 -0.3 22 0.737

Dark-light box Duration in bright comp. (s) 12 133.9 11.8 12 122.9 12.74 -0.6 22 0.534 Bright comp. visits (n) 12 37.08 4.13 12 34.17 3.17 -0.6 22 0.581 Distance moved (m) 12 10.9 0.87 12 9.92 0.96 -0.8 22 0.458 Latency to enter bright comp. (s) 12 6.84 1.73 12 6.7 1.45 -0.1 22 0.951

Novelty-induced hypophagia Latency (s) 10 148.5 14.73 11 116.5 32.56 -0.9 19 0.397 Heterozygous Munc18-1 mice exhibit an increased anxiety-like phenotype but no cognitive impairment 3 75 )

0.12 0.33 0.63 0.63 0.82 0.68 -value P ANOVA (

n=12 Z mice heterozygous (HZ) mice. 91±1.4 1.6±0.1 5.5±1.2 52.9±0.9 H 0.48±0.05 0.82±0.04 munc18-1

mice n=12 T 54±2.5 4.6±1.3 2.6±0.6 91.6±1.8 W 0.79±0.05 0.42±-0.04 5-CSRTT task performance in wild-type (WT) and 14: S Correct response latency (s) latency response Correct Reward latency (s) Response variability (s) Impulsive responses (n) Omission errors (%) Response accuracy (%) easure (unit) able M T Motivation Attention and inhibitory control 76

Table S15: Spontaneous home cage behavior of munc18-1 wild-type and heterozygous mice. 115 parameters derived from X-Y coordinates recorded for 3 consecutive days without human interference. Alpha level was correct for FDR to p=0.032

Wild-type mice Heterozygous mice Statistics

Mean SEM N Mean SEM N DF T-test P-value Activity duration dark 6682.27 260.45 64 6112.78 293.01 59 118.03 1.45 0.149 Mean activity duration dark 24.76 0.54 63 23.06 0.47 57 117.93 2.3 0.0234 Activity number dark 286.73 11.17 64 281.41 12.86 59 104.74 0.59 0.5568 Feeding zone duration dark 8272.85 381.26 64 10622.25 521.01 59 111.29 -3.53 0.0006 OnShelter zone duration dark 1980.05 148.63 64 1098.71 152.77 59 118.59 3.15 0.002 OnShelter zone number dark 93.07 7.22 64 55.31 10.85 59 117.82 4.2 0.0001 Spout zone duration dark 1579.79 84.53 64 2342.05 342.08 59 105.55 -3.07 0.0027 Activity duration light 1359.18 107.09 64 805.43 102.23 59 96.81 3.59 0.0005 Mean activity duration light 22.6 1.98 62 15.61 0.88 54 96.1 4.8 <0.0001 Activity number light 65.62 4.2 64 50.23 4.76 59 101.5 2.69 0.0083

Spontaneous behavior (activity) Feeding zone duration light 2352.73 143.82 64 2592.55 279.17 59 87.22 -0.76 0.4471 OnShelter zone duration light 245.03 31.25 64 129.57 36.18 59 103.96 4.54 <0.0001 OnShelter zone number light 14.35 1.67 64 6.58 1.17 59 117.6 5.17 <0.0001 Spout zone duration light 321.99 27.99 64 852.37 672.92 59 90.28 3.2 0.0019 Heterozygous Munc18-1 mice exhibit an increased anxiety-like phenotype but no cognitive impairment 3

77

0.0002 -3.81 115.57 59 0.01 0.9 64 0.01 0.83 index dark-light duration zone Spout

0.4959 -0.68 97.3 59 0.02 0.91 62 0.01 0.89 index dark-light duration zone OnShelter

0.0987 1.67 83.1 55 0 0.5 63 0 0.51 index dark-light distance movement short Mean

Spontaneous behavior (dark-light ratio) (dark-light behavior Spontaneous

0.666 -0.43 99.93 54 0 0.51 62 0 0.51 index dark-light distance movement long Mean

0.8813 -0.15 104.38 56 0.02 0.55 62 0.01 0.54 index dark-light number visit shelter Long

0.111 1.61 112.44 57 0.01 0.33 62 0.01 0.35 index dark-light duration visit shelter Long

0.4238 0.81 41.93 21 0.02 0.48 43 0.01 0.49 index dark-light duration visit shelter short Mean

0.0882 -1.72 108.77 59 0.01 0.81 64 0.01 0.78 index dark-light duration zone Feeding

Long arrest number arrest Long 0.0152 -2.47 113.07 57 0.01 0.85 63 0.01 0.81 index dark-light

Mean long arrest duration arrest long Mean 0.1946 1.31 99.81 54 0.01 0.43 62 0.01 0.45 index dark-light

Long arrest duration arrest Long 0.0723 -1.81 114.2 58 0.02 0.81 64 0.01 0.78 index darklight

Mean short arrest duration arrest short Mean 0.0387 2.09 99.75 56 0 0.49 63 0 0.5 index dark-light

Activity number Activity 0.0064 -2.78 116.35 57 0.01 0.85 63 0.01 0.81 index dark-light

Mean activity duration activity Mean <0.0001 -4.24 105.01 54 0.01 0.62 62 0.01 0.54 index dark-light

Activity duration Activity 0.0003 -3.72 117.58 57 0.01 0.89 63 0.01 0.83 index dark-light Table S15: Continued 78

Wild-type mice Heterozygous mice Statistics

Mean SEM N Mean SEM N DF T-test P-value Activity duration habituation ratio dark 1.79 0.03 64 1.56 0.03 59 121 5.04 <0.0001 Mean activity duration habituation ratio dark 1.85 0.02 63 1.71 0.02 57 116.78 4.31 <0.0001 Activity number habituation ratio dark 1.93 0.03 64 1.8 0.03 59 119.21 2.97 0.0036 Mean short arrest duration habituation ratio dark 2.09 0.01 63 2.18 0.02 57 103.08 -4.52 <0.0001 Long arrest duration habituation ratio dark 2.13 0.04 64 2.21 0.05 59 108.02 -1.07 0.2868 Mean long arrest duration habituation ratio dark 2.12 0.03 63 2.31 0.05 57 101.11 -3.54 0.0006 Long arrest number habituation ratio dark 2 0.04 64 1.92 0.04 59 118.93 1.47 0.1434 Feeding zone duration habituation ratio dark 2.1 0.04 64 2.26 0.06 59 98.85 -2.16 0.0335 Mean short shelter visit duration habituation ratio dark 2.04 0.03 63 2.13 0.05 57 103.54 -1.51 0.1346 Long shelter visit duration habituation ratio dark 2.04 0.04 61 2.31 0.08 55 98.3 -3.09 0.0026 Mean long movement distance habituation ratio dark 2.01 0.01 63 2 0.01 57 100.19 0.51 0.6113 Mean short movement distance habituation ratio dark 1.98 0.01 63 1.95 0.01 57 101.8 3.41 0.0009 OnShelter zone duration habituation ratio dark 2.08 0.08 63 1.63 0.08 59 117.96 4.7 <0.0001 Spout zone duration habituation ratio dark 1.95 0.04 64 2.59 0.7 59 69.4 0.03 0.9735 Activity duration habituation ratio light 1.86 0.09 64 1.83 0.08 56 116.1 0.24 0.8143 Mean activity duration habituation ratio light 2.08 0.11 62 1.84 0.05 54 108.23 2.6 0.0107 Activity number habituation ratio light 1.8 0.04 64 1.97 0.07 57 101.01 -1.78 0.0783 Mean short arrest duration habituation ratio light 1.98 0.01 63 2.06 0.02 56 94.13 -3.32 0.0013 Spontaneous behavior (habituation) Long arrest duration habituation ratio light 2.28 0.29 64 2.61 0.17 58 115.95 -2.51 0.0136 Mean long arrest duration habituation ratio light 2.27 0.07 62 2.39 0.08 51 102.38 -1.21 0.2291 Long arrest number habituation ratio light 1.83 0.06 64 2.19 0.15 53 84.2 -2.44 0.0168 Feeding zone duration habituation ratio light 2 0.07 64 2.83 0.2 58 85.75 -4.44 <0.0001 Mean short shelter visit duration habituation ratio light 1.93 0.05 36 2.08 0.05 15 38.71 -2.33 0.0252 Long shelter visit duration habituation ratio light 2.01 0.02 63 1.98 0.02 58 118.18 0.79 0.4321 Mean long movement distance habituation ratio light 1.99 0.01 62 1.98 0.02 54 97.47 0.38 0.7069 Mean short movement distance habituation ratio light 2 0.01 63 1.99 0.01 55 88.59 0.55 0.583 OnShelter zone duration habituation ratio light 1.87 0.09 55 1.63 0.08 36 68.39 2.27 0.0266 Spout zone duration habituation ratio light 2.02 0.09 59 1.87 0.14 41 74.06 1.6 0.1144 Heterozygous Munc18-1 mice exhibit an increased anxiety-like phenotype but no cognitive impairment 3

79

0.1868 1.33 116.83 59 0.04 2.31 64 0.05 2.4 threshold movement Long

0.001 3.4 99.86 59 38.57 317.67 64 36.05 502.2 light number movement Short

0.0188 2.38 115.02 55 0.02 1.98 63 0.02 2.06 light distance movement short Mean

0.0005 3.62 102.47 59 44.03 319.12 64 50.01 551.97 light distance movement Short

0.0005 3.59 99.48 59 25.9 194.65 64 24.17 305.26 light number movement Long

0.7158 0.37 112.13 54 0.29 13.03 62 0.27 13.18 light distance movement long Mean

0.0018 3.22 98.17 59 313.93 2261.56 64 322.91 3710.35 light distance movement Long

0.0117 2.57 97.14 59 9.63 79.1 64 5.89 94.83 light number arrest Long

0.0092 -2.65 111.22 54 3.65 47.71 62 5.15 39.06 light duration arrest long Mean

0.9924 -0.01 118.48 59 976.48 4110.73 64 879.26 4098.14 light duration arrest Long

Spontaneous behavior (kinematics) behavior Spontaneous

<0.0001 3.54 99.74 59 55.28 446.87 64 55.47 738.17 light number arrest Short

0.0373 -2.11 108.35 56 0.03 2.21 63 0.03 2.12 light duration arrest short Mean

0.0015 3.26 98.3 59 56 512.32 64 54.54 797.16 light duration arrest Short

0.3356 0.97 104.04 59 113.45 2092.25 64 92.75 2301.41 dark number movement Short

0.0005 3.56 110.03 57 0.02 1.98 63 0.02 2.08 dark distance movement short Mean

Short movement distance movement Short 0.1789 1.35 106.1 59 157.55 2101.5 64 135.31 2541.6 dark

Long movement number movement Long 0.3674 0.91 104.34 59 83.66 1469.42 64 67.39 1584.49 dark

Mean long movement distance movement long Mean 0.761 0.3 116.92 57 0.26 13.35 63 0.25 13.46 dark

Long movement distance movement Long 0.3858 0.87 103.2 59 1315.19 18455.6 64 941.26 19697.13 dark

Long arrest number arrest Long 0.6629 0.44 104.36 59 17.86 415.34 64 15.67 413.68 dark

Mean long arrest duration arrest long Mean 0.0039 -2.95 117.14 57 1.25 32.37 63 1.19 27.75 dark

Long arrest duration arrest Long 0.8306 -0.21 66.04 59 761.57 13350.86 64 703.4 11254.96 dark

Short arrest number arrest Short 0.3308 0.98 103.65 59 172.2 3230.56 64 134.97 3555.48 dark

Mean short arrest duration arrest short Mean 0.235 -1.19 111.19 57 0.03 2.17 63 0.02 2.13 dark

Short arrest duration arrest Short dark 3934.75 147.95 3647.01 64 157.19 119.76 59 0.1851 1.33

Long arrest threshold arrest Long 0.15 6.62 64 0.16 6.76 118.29 59 0.5417 -0.61

Long movement max. velocity max. movement Long 19.44 0.23 64 0.26 17.83 59 <0.0001 4.74 113.6

Long movement fraction of total movement total of fraction movement Long 0.01 0.45 64 0.01 0.46 120.08 59 0.2538 -1.15 80

Table S15: Continued

Wild-type mice Heterozygous mice Statistics

Mean SEM N Mean SEM N DF T-test P-value Short shelter visit duration dark 647 42.83 64 552.21 36.87 59 119.62 1.68 0.0961 Mean short shelter visit duration dark 9.76 0.48 63 9.28 0.51 57 113.6 0.87 0.3866 Short shelter visit number dark 79.68 5.13 64 81.32 7.21 59 106.25 0.72 0.4756 Long shelter visit duration dark 20132.72 770.39 64 19123.18 883.51 59 117.3 0.86 0.3909 Long shelter visit number dark 7.93 0.38 64 6.74 0.35 59 117.19 1.9 0.0594 Mean long shelter visit duration 4968.57 149.11 63 6168.19 429.4 58 97.12 -3.09 0.0026 Short shelter visit threshold 4.66 0.09 64 4.41 0.1 59 118.29 1.99 0.049 Long shelter visit fraction of total visits 1.1 0.01 64 1.08 0.01 59 120.55 2.56 0.0118 Long shelter visit threshold 9.96 0.08 64 9.93 0.1 59 115.55 0.22 0.8249 Short shelter visit duration light 171.69 14.67 64 95.14 13.96 59 103.5 3.88 0.0002 Mean short shelter visit duration light 9.73 0.54 43 9.69 0.65 21 46.13 -0.17 0.8633 Spontaneous behavior (sheltering) Short shelter visit number light 22.31 2.16 64 11.67 1.77 59 116.19 4.34 <0.0001 Long shelter visit duration light 36214.98 885.65 64 37117.32 989.27 59 118.28 -0.68 0.4981 Long shelter visit number light 6.59 0.26 64 5.57 0.25 59 112.85 2.27 0.0254 Heterozygous Munc18-1 mice exhibit an increased anxiety-like phenotype but no cognitive impairment 3

81

Spout zone change in response to light to response in change zone Spout 0.1562 1.43 84.82 36 0.01 -0.03 51 0.02 0

Spout zone change in anticipation light anticipation in change zone Spout 0.4877 -0.7 69 52 0.02 0.05 62 0.01 0.04

Spout zone change in response to dark to response in change zone Spout 0.835 -0.21 89.07 58 0.01 -0.01 63 0.01 -0.01

Spout zone change in anticipation dark anticipation in change zone Spout 0.949 0.06 100.12 51 0.01 0 61 0.01 0.01

Spontaneous behavior (pattern) behavior Spontaneous OnShelter zone change in response to light to response in change zone OnShelter 0.6164 -0.5 70.68 36 0.01 -0.03 51 0.01 -0.04

OnShelter zone change in anticipation light anticipation in change zone OnShelter 0.0013 3.31 110.96 52 0.01 0.04 62 0.01 0.08

OnShelter zone change in response to dark to response in change zone OnShelter 0.4108 0.83 116.11 58 0.01 -0.01 63 0.01 0

OnShelter zone change in anticipation dark anticipation in change zone OnShelter 0.3897 0.86 108.72 51 0.01 0 61 0.01 0.01

Feeding zone change in response to light to response in change zone Feeding 0.1831 -1.34 72.65 36 0.05 0.06 51 0.04 -0.03

Feeding zone change in anticipation light anticipation in change zone Feeding 0.9177 0.1 102.25 52 0.04 -0.2 62 0.03 -0.2

Feeding zone change in response to dark to response in change zone Feeding 0.1279 -1.53 118.89 58 0.03 0.04 63 0.03 -0.01

Feeding zone change in anticipation dark anticipation in change zone Feeding 0.617 -0.5 102.77 51 0.04 0.04 61 0.03 0.01

Activity change in response to to light to to response in change Activity 0.7936 0.26 120.31 59 0.01 -0.07 64 0.01 -0.07

Activity change in response to to dark to to response in change Activity 0.3092 -1.02 112.7 59 0.01 0.2 64 0.01 0.18

Activity change in anticipation of light of anticipation in change Activity 0.6809 0.41 119.89 59 0.01 0.03 64 0.01 0.03

Activity change in anticipation of dark of anticipation in change Activity 0.9123 0.11 110.3 59 0.01 0.02 64 0.01 0.02 82

Table S16: Spontaneous behavior of floxed munc18-1 wild-type (WT) and heterozygous (HZ) mice. 115 parameters derived from X-Y coordinates recorded for 3 consecutive days without human interference. Alpha level was corrected by FDR to p=0.036.

WT mice HZ mice Statistics Munc18-1 floxed Mean SEM N Mean SEM N DF T-test P-value Activity duration dark 8189.84 467.76 11 8131.61 431.61 12 20.66 0.09 0.928 Mean activity duration dark 23.26 0.81 11 22.6 0.85 12 21 0.57 0.5741 Activity number dark 346.49 14.88 11 353.71 13.29 12 20.36 -0.37 0.7159 Feeding zone duration dark 9015.71 705.94 11 10658.15 581.77 12 18.3 -1.81 0.086 OnShelter zone duration dark 2417.38 184.78 11 592.21 156.87 12 13.13 5.71 0.0001 OnShelter zone number dark 145.39 14.68 11 26.99 6.47 12 15.38 7.23 >0.0001 Spout zone duration dark 1806.37 96.62 11 2120.32 155.75 12 19.75 -1.82 0.0837 Activity duration light 1064.49 149.36 11 546.78 217.43 12 14.28 1.85 0.0847 Mean activity duration light 19.33 1.08 11 14.22 2.36 12 13.77 1.89 0.0805 Activity number light 55.18 6.78 11 38.97 9.72 12 16.49 1.39 0.1843

Spontaneous behavior (activity) Feeding zone duration light 1898.13 202.07 11 1886.63 221.46 12 20.96 0.04 0.9698 OnShelter zone duration light 192.29 35.47 11 12.52 14.13 12 12.21 3.62 0.0034 OnShelter zone number light 12.76 3.11 11 1.34 0.81 12 19.07 4.91 0.0001 Spout zone duration light 250.12 50.8 11 96.94 49.01 12 15.29 2.11 0.0514 Heterozygous Munc18-1 mice exhibit an increased anxiety-like phenotype but no cognitive impairment 3

83

0.1876 -1.37 19.25 12 0.03 0.91 11 0.02 0.86 index dark-light duration zone Spout

0.7271 0.36 13.32 12 0.04 0.91 11 0.01 0.92 index dark-light duration zone OnShelter

0.0728 1.89 20.23 12 0 0.5 11 0 0.51 index dark-light distance movement short Mean

Spontaneous behavior (Dark-Light ratio) ratio) (Dark-Light behavior Spontaneous

0.7454 -0.33 19.72 12 0.01 0.51 11 0.01 0.51 index dark-light distance movement long Mean

0.336 0.99 20.36 12 0.03 0.44 11 0.03 0.49 index dark-light number visit shelter Long

0.2083 1.3 20.92 12 0.01 0.27 11 0.01 0.3 index dark-light duration visit shelter Long

0.0191 2.88 8.58 5 0.01 0.46 10 0.01 0.51 index dark-light duration visit shelter short Mean

0.4086 -0.84 19.73 12 0.02 0.85 11 0.02 0.82 index dark-light duration zone Feeding

Long arrest number number arrest Long 0.2137 -1.29 16.14 12 0.03 0.89 11 0.01 0.86 index dark-light

Mean long arrest duration duration arrest long Mean 0.0898 1.82 14.3 11 0.04 0.38 11 0.02 0.47 index dark-light

Long arrest duration duration arrest Long 0.4801 -0.72 20.86 12 0.02 0.86 11 0.02 0.84 index dark-light

Mean short arrest duration duration arrest short Mean 0.0518 2.12 14.54 12 0.01 0.48 11 0 0.5 index dark-light

Activity number number Activity 0.3702 -0.92 17.12 12 0.02 0.88 11 0.01 0.86 index dark-light

Mean activity duration duration activity Mean 0.1267 -1.62 14.48 12 0.04 0.61 11 0.02 0.55 index dark-light

Activity duration duration Activity 0.5814 -0.57 13.42 12 0.04 0.9 11 0.01 0.88 index dark-light Table S16: Continued. 84

WT mice HZ mice Statistics Munc18-1 floxed Mean SEM N Mean SEM N DF T-test P-value Activity duration habituation ratio dark 0.92 0.06 11 0.8 0.07 12 20.89 1.42 0.1702 Mean activity duration habituation ratio dark 0.88 0.04 11 0.78 0.06 12 17.61 1.33 0.2005 Activity number habituation ratio dark 1.05 0.04 11 1.02 0.03 12 20.36 0.47 0.6399 Mean short arrest duration habituation ratio dark 1.01 0.02 11 1.11 0.03 12 16.99 -2.91 0.0097 Long arrest duration habituation ratio dark 1.06 0.05 11 1.14 0.06 12 20.97 -1.02 0.32 Mean long arrest duration habituation ratio dark 1.06 0.05 11 1.12 0.05 12 20.99 -0.79 0.4364 Long arrest number habituation ratio dark 1 0.04 11 1.03 0.05 12 18.74 -0.37 0.7182 Feeding zone duration habituation ratio dark 1 0.05 11 1.06 0.05 12 20.14 -0.77 0.4495 Mean short shelter visit duration habituation ratio dark 1.01 0.04 11 1.11 0.07 12 19.22 -1.36 0.1899 Long shelter visit duration habituation ratio dark 0.93 0.05 11 1.01 0.06 12 20.78 -1.06 0.2996 Mean long movement distance habituation ratio dark 1.01 0.02 11 1 0.03 12 19.82 0.3 0.7654 Mean short movement distance habituation ratio dark 1 0.01 11 0.96 0.01 12 20.72 2.02 0.0566 OnShelter zone duration habituation ratio dark 1.19 0.1 11 0.81 0.19 12 15.25 1.73 0.1037 Spout zone duration habituation ratio dark 0.9 0.05 11 0.9 0.05 12 21 -0.02 0.9869 Activity duration habituation ratio light 0.76 0.07 11 1.64 0.38 12 11.69 -2.88 0.0142 Mean activity duration habituation ratio light 0.86 0.07 11 1.26 0.2 12 13.54 -2.14 0.0512 Activity number habituation ratio light 0.88 0.07 11 1.27 0.15 12 15.86 -2.5 0.0236 Mean short arrest duration habituation ratio light 0.98 0.02 11 1.15 0.06 12 15.53 -2.82 0.0125 Spontaneous behavior (habituation) Long arrest duration habituation ratio light 1 0.1 11 1.42 0.25 12 14.83 -1.71 0.1086 Mean long arrest duration habituation ratio light 1.04 0.1 11 0.99 0.17 12 13.44 0.28 0.7813 Long arrest number habituation ratio light 0.99 0.1 11 1.33 0.22 9 14.21 -1.55 0.142 Feeding zone duration habituation ratio light 0.93 0.09 11 1.33 0.28 10 12.98 -1.52 0.1525 Mean short shelter visit duration habituation ratio light 0.87 0.07 10 1.12 0.01 12 9.35 -3.46 0.0068 Long shelter visit duration habituation ratio light 1.01 0.01 11 0.97 0.01 3 20.68 2.57 0.0179 Mean long movement distance habituation ratio light 1.03 0.03 11 0.98 0.03 12 18.65 1.32 0.2023 Mean short movement distance habituation ratio light 1 0.02 11 1.01 0.02 12 20.43 -0.3 0.7678 OnShelter zone duration habituation ratio light 1.04 0.12 11 0.79 0.33 12 2.51 0.74 0.5239 Spout zone duration habituation ratio light 1.01 0.22 11 1.05 0.78 3 4.84 -0.06 0.9521 Heterozygous Munc18-1 mice exhibit an increased anxiety-like phenotype but no cognitive impairment 3

85

0.0009 4.01 16.95 12 0.05 1.25 11 0.11 1.7 threshold movement Long

0.0095 2.96 15.43 12 53.21 178.08 11 53.47 417.36 light number movement Short

0.0009 4.19 13.77 12 0.02 0.94 11 0.06 1.2 light distance movement short Mean

0.0023 3.55 17.91 12 50.28 168 11 85.19 496.87 light distance movement Short

0.0211 2.56 15.8 12 33.32 108.79 11 31.94 232.32 light number movement Long

0.0791 1.85 19.73 12 0.43 12.46 11 0.58 13.75 light distance movement long Mean

0.0094 2.92 17.33 12 393.83 1349.49 11 503.13 3191.49 light distance movement Long

0.0377 2.3 13.45 12 12.81 40.3 11 7.59 78.68 light number arrest Long

0.016 -2.8 12.09 12 11.62 51.87 11 1.96 28.48 light duration arrest long Mean

0.9331 -0.09 19.12 12 310 2359.63 11 212.34 2327.66 light duration arrest Long

Spontaneous behavior (kinematics) (kinematics) behavior Spontaneous

0.0105 2.91 15.56 12 77.26 257.84 11 78.3 598.3 light number arrest Short

0.0003 -4.55 16.21 12 0.06 1.3 11 0.02 1.03 light duration arrest short Mean

0.0356 2.3 15.89 12 90.92 332.67 11 77.23 616.7 light duration arrest Short

0.0034 3.31 20.51 12 114.9 2115.4 11 166.66 2749.42 dark number movement Short

0.0001 5.32 15.07 12 0.02 0.94 11 0.06 1.24 dark distance movement short Mean

Short movement distance distance movement Short 0.0001 4.84 19.07 12 138.48 1983.54 11 317.05 3407.18 dark

Long movement number number movement Long 0.9215 0.1 20.99 12 126.82 1872.09 11 124.24 1889.22 dark

Mean long movement distance distance movement long Mean 0.18 1.39 20.99 12 0.58 13.1 11 0.61 14.25 dark

Long movement distance distance movement Long 0.3959 0.87 20.53 12 2121.93 24511.25 11 1889.68 26899.87 dark

Long arrest number number arrest Long 0.3316 1.01 13.54 12 11.07 459.62 11 28.16 488.68 dark

Mean long arrest duration duration arrest long Mean 0.0061 -3.07 19.62 12 1.28 30.52 11 1.32 24.9 dark

Long arrest duration duration arrest Long 0.0394 -2.22 18.03 12 518.11 14022.04 11 658.92 12162.84 dark

Short arrest number number arrest Short 0.0715 1.9 20.97 12 222.25 3677.02 11 236.14 4276.54 dark

Mean short arrest duration duration arrest short Mean 0.0008 -4.09 16.27 12 0.04 1.19 11 0.02 1.03 dark

Short arrest duration duration arrest Short 0.8388 0.21 19.54 12 166.62 4404.26 11 209.59 4459.45 dark

Long arrest threshold threshold arrest Long 0.13 5.41 11 6.08 0.2 12 0.0106 -2.82 19.58

Long movement max. velocity velocity max. movement Long 0.47 19.77 11 17.74 0.51 20.65 12 0.0076 2.96

Long movement fraction of total movement movement total of fraction movement Long 0.01 0.43 11 0.01 0.49 12 0.003 -3.48 16.6 86

Table S16: Continued.

WT mice HZ mice Statistics Munc18-1 floxed Mean SEM N Mean SEM N DF T-test P-value Short shelter visit duration dark 871.85 125.67 11 1000.64 131.9 12 21 -0.71 0.4874 Mean short shelter visit duration dark 7.31 0.58 11 7.62 0.99 12 18.2 -0.29 0.7768 Short shelter visit number dark 110.99 9.33 11 121.77 22.8 12 15.56 -0.49 0.6325 Long shelter visit duration dark 16158.49 891.16 11 14249.64 724.8 12 19.75 1.66 0.1124 Long shelter visit number dark 5.41 0.55 11 4.21 0.48 12 20.98 1.73 0.0988 Mean long shelter visit duration 5164.85 244.83 11 6068.01 409.2 12 19.44 -2.01 0.0581 Short shelter visit threshold 4.53 0.13 11 4.43 0.17 12 19.82 0.47 0.644 Long shelter visit fraction of total visits 0.07 0 11 0.05 0.01 12 20.52 2.64 0.0156 Long shelter visit threshold 10.33 0.14 11 10.39 0.14 12 20.99 -0.29 0.7717 Short shelter visit duration light 121.35 27.04 11 48.91 38.36 12 13.64 1.48 0.1605 Mean short shelter visit duration light 7.02 0.57 10 9.64 1.11 12 7.96 -2.33 0.0484 Spontaneous behavior (sheltering) Short shelter visit number light 17.83 3.56 11 7.48 3.71 5 15.69 1.98 0.065 Long shelter visit duration light 38274.44 446.35 11 37983.87 947.84 12 15.58 0.28 0.7852 Long shelter visit number light 5.75 0.55 11 5.25 0.39 12 19.21 0.78 0.4423 Heterozygous Munc18-1 mice exhibit an increased anxiety-like phenotype but no cognitive impairment 3

87

Spout zone change in response to light light to response in change zone Spout 0.8078 0.25 5.98 5 0.04 -0.01 6 0.02 0.01

Spout zone change in anticipation light light anticipation in change zone Spout 0.0001 -4.65 20.72 12 0.01 0.08 11 0.01 0.01

Spout zone change in response to dark dark to response in change zone Spout 0.0909 -1.78 19.07 12 0.01 0.02 11 0.01 0

Spout zone change in anticipation dark dark anticipation in change zone Spout 0.1224 -1.61 19.88 12 0.03 0.05 11 0.02 -0.02

Spontaneous behavior (pattern) (pattern) behavior Spontaneous OnShelter zone change in response to light light to response in change zone OnShelter 0.5532 -0.62 8.21 5 0.04 -0.04 6 0.03 -0.07

OnShelter zone change in anticipation light light anticipation in change zone OnShelter 0.0027 3.44 19.18 12 0.01 0.02 11 0.02 0.09

OnShelter zone change in response to dark dark to response in change zone OnShelter 0.4468 -0.78 17.97 12 0.01 0 11 0.01 -0.01

OnShelter zone change in anticipation dark dark anticipation in change zone OnShelter 0.9116 -0.11 17.72 12 0.01 0 11 0.02 0

Feeding zone change in response to light light to response in change zone Feeding 0.111 1.78 8.34 5 0.12 -0.2 6 0.1 0.07

Feeding zone change in anticipation light light anticipation in change zone Feeding 0.2352 1.23 19.46 12 0.06 -0.19 11 0.04 -0.1

Feeding zone change in response to dark dark to response in change zone Feeding 0.2149 -1.28 20.04 12 0.04 0.01 11 0.05 -0.07

Feeding zone change in anticipation dark dark anticipation in change zone Feeding 0.4487 0.78 16.22 12 0.06 0.02 11 0.11 0.12

Activity change in response to to light light to to response in change Activity 0.4968 0.69 20 12 0.02 -0.09 11 0.01 -0.08

Activity change in response to to dark dark to to response in change Activity 0.1993 -1.33 19.99 12 0.02 0.3 11 0.02 0.26

Activity change in anticipation of light light of anticipation in change Activity 0.4057 -0.86 14.17 12 0.03 0.13 11 0.01 0.11

Activity change in anticipation of dark dark of anticipation in change Activity 0.3706 -0.92 20.68 12 0.01 0.01 11 0.01 -0.01

Chapter

Functional characterization of the PCLO S4814A variant associated with major depressive disorder reveals cellular but 4 not behavioral differences Asiya Giniatullina1*, Gregoire Maroteaux1*, Lieke Geerts1*, Bastijn Koopmans2, Maarten Loos2, Remco Klaassen3, Ning Chen3, Roel C. van der Schors3, Pim van Nierop3, Wilko Altrock4, Ka Wan Li3, Brenda Penninx5, Eco de Geus6, Dorret Boomsma6, Patrick F. Sullivan7, L. Niels Cornelisse9, Ruud F. Toonen1, Sophie van der Sluis9, Oliver Stiedl1,8, Danielle Posthuma1,9, August B. Smit3, Alexander J. Groffen9 and Matthijs Verhage1,9

1Departments of 1Functional Genomics, 3Molecular and Cellular Neurobiology, Center for Neurogenomics and Cognitive Research, Neuroscience Campus Amsterdam, VU University Amsterdam, the Netherlands 2Sylics (Synaptologics BV), Amsterdam, the Netherlands 4Department of Neurochemistry and Molecular Biology, Leibniz Institute for Neurobiology, Magdeburg, Germany 5Department of Psychiatry, and EMGO Institute for Health and Care Research and Neuroscience Campus Amsterdam, VU University Medical Center/GGZ in Geest, Amsterdam, The Netherlands, Department of Psychiatry, University Medical Center Groningen, Groningen, The Netherlands, Department of Psychiatry, Leiden University Medical Center, Leiden, The Netherlands 6Department of Biological Psychology, VU University Amsterdam, Amsterdam, the Netherlands 7Department of Genetics, University of North Carolina, Chapel Hill, NC, USA 8Department of Molecular and Cellular Neurobiology, Center for Neurogenomics and Cognitive Research, Neuroscience Campus Amsterdam, VU University Amsterdam, the Netherlands 9Department of Clinical Genetics, Section Medical Genomics, VU University Medical Center, Amsterdam, the Netherlands

*equally contributing authors Characterization of Pclo SA/SA mouse model 4 91 Pclo S4814A variation PCLO variation (S4814A) has a small contribution to variation in amouse knock-in model expressing the PCLO mice. We conclude that the

PCLO SA/SA Pclo in MajorDepressivein geneticmany Disorder Likevariations(MDD). PCLO variant.In the highly homogeneous background of inbred mice, two functional effects stract SA/SA of the SA-variation were observed at the cellular level: an increased synaptic Piccolo level and a 30% increased excitatory synaptic transmission in cultured neurons. Other aspects of Piccolo function were unaltered: calcium-dependent phospholipid binding, synapse formation in vitro and synaptic accumulation of synaptic vesicles. Moreover, anxiety, cognition and depressive- like behavior were normal in the overall heritability and the association does thatnot the alwaysmolecular replicate.effects ofIt suchhas variations been penetrate to proposed a variabledue extentto inphenotypic theand populationgenotypic heterogeneity at the population level. More mayrobust be exposedeffects by studying such variations in isolation, in a more homogeneous context. We testedthis proposal by modeling the Pclo in complex trait association studies, the Ab Genome-wideassociation studies have suggested arole for anonsynonymous exonic variation presynapticthein gene produces cellular phenotypes, which inphenotypes. modelexplainingproposea We (subtle)how cellular phenotypespenetrate notdo mice, however, do not thebehavioralto level but,genetic dueto translate and phenotypic heterogeneity andnon-linearity, still into behavioral produce association signals in human population studies. Keywords: Major Depressive Disorder, Piccolo, C2-domain, synaptic plasticity, Introduction Piccolo (also known as Aczonin) is the largest protein (~550 kDa) of the cytomatrix at the active zone (CAZ) of brain synapses (Fenster et al., 2000). The CAZ is an electron-dense region in electron micrographs and integrates many proteins that regulate neurotransmitter secretion in presynaptic terminals. Piccolo and its interaction partners are positioned to support synaptic vesicle supply to the site of release and hereby support synaptic reliability during intense activity (Limbach et al., 2011). Piccolo protein consists of two Zinc-fingers, one PDZ, three coiled-coil (CC) and two C2 domains (C2A and C2B). The C2A domain of Piccolo has been characterized in its calcium and phospholipid binding properties (Garcia et al., 2004). Partial deletion of the PCLO gene in mice produced an effect on body size but not on synaptic function (Mukherjee et al., 2010), but acute Piccolo knock-down in cultured neurons resulted in enhanced synaptic vesicle release, which was explained by Piccolo actions on the presynaptic cytoskeleton and modulation of Synapsin dynamics (Leal-Ortiz et al., 2008) More recently, double knock-down of Piccolo and the related CAZ protein Bassoon showed that the two proteins together are necessary for the maintenance of synapse integrity, as their loss induced aberrant degradation of multiple presynaptic proteins (Waites et al., 2013). Piccolo is believed to regulate neurotransmitter release by facilitating activity-dependent F-actin assembly and the dynamic recruitment of key signaling molecules into presynaptic boutons (Waites and Garner, 2011) as well as scaffolding of actin regulatory molecules, including Abp1, GIT1, and Profilin (Fenster et al., 2003; Kim et al., 2003; Wang et al., 2009). A role for PCLO in Major Depressive Disorder (MDD) was proposed in the first large scale genome- wide association study for MDD (Sullivan et al., 2009). MDD, or unipolar disorder, is characterized by low or depressed mood, loss of interest or pleasure in almost all activities (anhedonia) and low energy/fatigability. The heritability of MDD is calculated to be around 31-42% (Sullivan et al., 2000) . The first large scale genome-wide association study did not find an association that met genome-wide significance, but 11 of the top 200 findings localized to a single 167 kb segment within the PCLO gene. The third most significant single nucleotide polymorphism (SNP, rs2522833) coded for a non-synonymous amino acid change (ala-4814-ser) in the C2A domain of Piccolo. Matching patients and controls more closely made this variation reach statistical significance. The association with MDD was independently reproduced by others(Aragam et al., 2011; Hek et al., 2010)A recent replicate case–control study (Minelli et al., 2012) investigated depression-related personality traits in healthy subjects as a function of the PCLO rs2522833 genotype. This study showed that rs2522833 homozygotes were more frequent among MDD patients than in controls (p<0.01). Personality traits (self-report) test revealed that the C allele carriers were associated with increased vulnerability to depression, including higher harm avoidance and lower novelty seeking. In particular, C allele carriers were more fearful and fatigable, less impulsive/more deliberate and less extravagant/more frugal. These findings provided further support for the involvement of the PCLO C2A domain in Major Depression pathogenesis, although the association was not always replicated in subsequent GWAS studies (Kohli et al., 2011; Lewis et al., 2010; Muglia et al., 2010; Rietschel et al., 2010; Shi et al., 2011; Shyn et al., 2011; Wray et al., 2012)and suggest that it may act by influencing personality traits that increase the risk of developing MDD. Many genome-wide linkage studies have been performed to understand the genetic basis of MDD, but none of them found genes that reach genome-wide significance. This has

92 Characterization of Pclo SA/SA mouse model 4 93 19%; total +/+ Pclo 63.5%, -and phosphatidylserine- 2+ SA/+ -dependent lipid binding 2+ Pclo a C -dependentphospholipid binding was 17.4%, 17.4%, 2+ mice) and compared them to wild-type SA/SA SA/SA Pclo Pclo domain does not alter A 2 C ). We performed extensive analysis of molecular, electrophysiological and +/+ Pclo mice are viable and fertile

variation in the mouse SA A / / S SA -dependent binding to negatively charged phospholipids is a common property of C2 The aim of our study was to test this proposal and investigateproposalandthisfunctional test differences to between wasstudy our of aim The 2+ esults he clo binding pocket and on the opposite surface of the C2A domain (Garcia et al., 2004).that The calcium-dependentfact lipid binding is normal in the mutant protein alsosequence variationconfirms does not thatcompromise overall this protein folding. P littermates ( domains in many presynaptic proteins, and of was Piccolo previously described (Garcia for et the al., C2Aassociation ofthe two variants ofthe domain isolated 2004). C2A domain ofPiccolo. Both C2A domain variants We started by were testing expressed as calcium-dependentGST-fusion constructs phospholipidand binding to phospholipidliposomesphosphatidylserine was tested. Similarcontaining Ca 20% anionic R T the two sequence variants in the Piccolo C2A domaininbred inmice. To the investigate highly the standardizedrole of contextthe depression-associatedof SNP rs2522833,knock-in mouse we modelcreated replacing S4742 in the C2A domain of Piccolo,mice, whichwith is an conservedalanine in (which we refer to as been explained by the fact that depression is a highly multi-genic disease, and susceptibility is conferred by a combination of several (many) genetic variants and external factors. A complex interplay probably exists between genetic variants and environmental factors, factorin may interact whichwith each othereach factor to a variable extent. Protective factors probably also contribute, alleviating the effects of each disease-causing factorto a variable extent. In such a complex scenario, it is difficult to assess the impact of individual factors in thehumanpopulation, contextinwhich eachof individual the contains differenta palette of contributing factors. However, assessment may be more successful in standardized whichmodels have a homogeneous suchgenetic background andas a highly standardizedinbred environment. mice, observedfor the two protein variants (Fig. 11). This finding is consistent with the location of the S4742A variation, which is separated by more than 30 Å from the Ca Ca To To generate a mouse model for the human rs2522833 variant, the sequencesmouse wereand aligned human(Fig. 12A).protein The human amino acid substitution S4841A corresponds to S4742A in mouse Piccolo. Flanking amino acid sequences are fully conserved between human and mouse (Fig. 12A). The corresponding codon in exon homologous19 of gene mousetargeting (Fig.PCLO 12B) was to producemutated a byknock-in mouse line, i.e., fully preservingthe natural expression pattern of the mutation Piccoloof alanine protein, to serine. with By the crossing heterozygous only themice, differenceexpectedoffspring Mendelian being wasdistribution produced the (Fig with 12C, behavioral parameters in these mice and tested calcium-dependent phospholipid binding bythe two variants of C2A domain in a cell-free system. 1,5

Piccolo +/+ 1,2 Piccolo SA/SA U A

, 0,9 g n i d n i b 0,6 d i p i L

0,3

0 0 0.01 0.1 1 10 free calcium , mM

Figure 11: Calcium titration for lipid binding by the C2A domain variants of Piccolo. Data are average of three measurements ± SEM, corrected for protein concentration.

126 mice; χ2(2)=0.547, p=0.76), indicating that homozygous PcloSA/SA were viable. Homozygous mice were fertile and no gross abnormalities were observed at any stage of development.

Proteomic analysis reveals differences in retrieval of Piccolo variants from synaptosomal lysates We took advantage of the Pclo knock-in mice to study changes in the Piccolo protein levels, interaction partners and retrieval efficiency upon S/A mutation. Using immunoprecipitation, we investigated whether the PCLO sequence variation affected its retrieval and its interactions in synaptosomal preparations from adult mouse hippocampal and cortical tissue. First, we compared the amount of Piccolo precipitated from these synaptosomal preparations. The amount retrieved from PcloSA/SA mice was significantly higher than from thePclo +/+ synaptosomes (Table 3). Since the total levels of Piccolo were comparable in the brains of Pclo+/+ and PcloSA/SA mice (Fig 12D), these data suggest that the Piccolo protein is more enriched in the synaptosomal preparation of PcloSA/SA brain than of Pclo+/+ brain. We confirmed this finding in 3 independent immunoprecipitations from Pclo+/+ and PcloSA/SA synaptosomes: on average the yield of synaptosomal preparations from Pclo+/+ brain was 89 +/- 2% of PcloSA/SA synaptosomes (Fig 12E). Several known and potential new interactors of Piccolo were detected in proteomic analysis (Table 3). The previously reported Piccolo interacting partners, Bassoon, CAST, Munc13, Rim (Wang et al., 2009) or Siah1 (Waites et al., 2013) were not detected in our analyses. The suggested difference in association between different Piccolo variants with the Serine- threonine kinase receptor-associated protein (STRAP, Table 3) was subjected to subsequent confirmation by mixing recombinant STRAP with brain lysates of PcloSA/SA and Pclo+/+ brains and immunoprecipitation of either STRAP or Piccolo. These experiments failed to confirm the (differential) association between the two proteins.

94 Characterization of Pclo SA/SA mouse model 4 95

< PCLO < Heavy chain control beads control antibody. 4819 4819 4747 4747 α

M13F WT anti-PCLO synaptosomes, from 3 NruI

SA/SA control beads control AvrII

Pclo

AvrII EcoRV SpeI EcoRI anti-PCLO S4742A H L D N T S H L D N T A H L S H L D N T A H L D N T EcoRI

ct cac ctc gat aac act Tct cac ctc Gct cac ctc gat aac act Tct cat ctg gac aac act Gct cat ctg gac aac act S4742A 30

60 45

150

SNP rs2522833 AscI Spel IP: Mice: brain was 89±2% of EcoRI 19 +/+ D mice produced resulted in Mendelian distribution of Pclo SA/+ Pclo 18 /SaIIAvrII EcoRV no Ab no Ab no Ab +/+ exon 19. Theneomycin cassette exon19. was flanked Frtby sites for Flp-mediated PCLO 17 Spel SA/+ V L I D L S S T I D L S S V L aca att gat tta tct agc gta ttg T I D L S S V L tct agc aca gta ttg att gat tta S S T V L I D L tct agc aca gta ttg att gat tta S S T V L I D L tct agc aca gta ttg att gat tta 4734 4734 4806 4806 mouse brains. Beads without addition of antibody were used as negative control. (E) Spel +/+ is expressed in synapses: Piccolo was immunoprecipitated from synaptosomes isolated AvrII EcoRV SA/SA expected observed EcoRI EcoRV NotI Pclo S4742A targeting S4742A strategy forknock-inmouseline generation. (A)Sequence alignment showsthat 16 and Pclo SA/SA M13R Pclo SA/SA Exon engineered allele wt allele Pclo 0 90 80 70 60 50 40 30 20 10 100 wt allele human wt allele human minor allele human wt allele mouse engineered allele mouse igure12: excision. Regions used for homologous recombination in the gene replacement vector are shown by thick grayand dark gray lines. (C) Crossing heterozygote theresidue S4814inhumans isconserved inmice andcorresponds strategy(B)S4742.Targeting to introduce to a singlenucleotide substitution into independent immunoprecipitations. Piccolo was immunoprecipitated and detected using gp44 F from offspring. (D) Average yield of synaptosomal preparations from adult E C B A To confirm potential new interaction partners of Piccolo, we performed an independent immunoprecipitation from Pclo+/+ mouse cortex synaptosomes, using two Piccolo antibodies: the same antibody used in for comparison between Pclo+/+ and PcloSA/SA (rb44α, raised in rabbit, above) and an antibody raised against the same Piccolo epitope in guinea pig (gp44α). Proteins present in amounts significantly higher than in control condition for both precipitations were: Rufy3, Rho guanine nucleotide exchange factor 2 and IQ motif and SEC7 domain-containing protein 1 (Table 3, right). Rufy3 was also found in high levels in the precipitation in Table 3, in both Pclo+/+ and PcloSA/SA condition, suggesting Rufy3 as a novel Piccolo interaction partner.

PcloSA/SA neurons have normal neuronal morphology Piccolo is considered to be one of the first components to arrive in newly generated synapses (Zhai et al., 2001) and is proposed to be involved in the structural organization of synapses. Therefore we investigated if the sequence variation affects synaptogenesis in cultured neurons. Hippocampal neurons from Pclo+/+ and PcloSA/SA littermates were cultured for 14 days on glia feeder layers and subsequently fixed, stained for presynaptic marker Vamp2 and dendritic marker MAP2 and analyzed using an automated analysis algorithm (SynD, Schmitz et al., 2011). In a separate experiment, Pclo+/+ and PcloSA/SA neurons were fixed at DIV 15 and stained for Piccolo, presynaptic marker Synapsin, and a dendritic marker, MAP2 (Fig. 13A), and analyzed in the same manner. At both time points, the number of synapses, the synapse area and dendrite length were similar between Pclo+/+ and PcloSA/SA neurons (Fig. 13B-D). Scholl analysis of neuronal complexity in culture (Fig 13E) also indicated that Pclo+/+ and PcloSA/SA neurons developed in a very similar manner in culture.

Synaptic Piccolo/VAMP2 ratio is 30 % higher in PcloSA/SA synapses Protein quantification (Table 3) suggested that synaptic Piccolo levels were higher inPclo SA/SA brains. Therefore we quantified staining intensities of synaptic Piccolo protein relative to the synaptic vesicle marker VAMP2 (Fig. 14A). Since Piccolo is known to modulate Synapsin dynamics (Leal- Ortiz et al., 2008), we also quantified synaptic Synapsin levels (Fig 14B). PcloSA/SA neurons showed a trend towards lower VAMP2 levels (Fig 14C) and higher Piccolo levels (Fig. 14D), but population averages of both these trends were not significant. However, when compared at the single synapse level, i.e., by dividing Piccolo intensity over VAMP2 intensity for each individual synapse, the Piccolo/VAMP2 ratio was significantly higher in PcloSA/SA neurons (29% higher, t-test p=0.005, Kruskal-Wallis p=0.01, Fig 14E). Correlation of Piccolo and VAMP2 intensity per synapse show an overall overlap between Pclo+/+ and PcloSA/SA, and a subgroup of PcloSA/SA synapses with very high Piccolo and low VAMP2 levels (Fig. S10). A trend towards higher Synapsin levels in PcloSA/SA neurons (p=0.098), was observed for Synapsin intensity measured over all synapses.

PcloSA/SA neurons show increased excitatory synaptic transmission Knock-down of Piccolo promotes activity dependent synaptic vesicle release in hippocampal neurons (Waites et al., 2011). Therefore, we tested if the increase in Piccolo/VAMP2 ratio in PcloSA/SA neurons affected synaptic transmission by electrophysiological analysis in autaptic hippocampal neurons. The excitatory postsynaptic current (EPSC) amplitude in naïve PcloSA/SA neurons was 30% higher compared to Pclo+/+ neurons (Fig. 15A-B, left, p=0.047), while

96 Characterization of Pclo SA/SA mouse model 4 97

+/+ Pclo (µm) (µm) (top), and +/+ Pclo DIV15 DIV14 Distance from soma from Distance Distance from soma Distance 0 50 100 150 200 0 50 100 150 200

8 7 6 5 4 3 2 1 0 8 7 6 5 4 3 2 1 0

number of intersections of number number of intersections of number neurons (Fig. 15E). This was E SA/SA expressing neurons (Fig. 15J-M). SA/SA Pclo DIV15 DIV15 DIV15 Pclo +/+ SA/SA neurons (Fig. S11). Vesicle replenishment +/+ SA/SA +/+ SA/SA

2 1 0 1 0 SA/SA 0

50 1,5 0,5

1,2 0,8 0,6 0,4 0,2 dendrite length dendrite

(mm) synapse area synapse 200 150 100 (µm2) synapse number synapse Pclo n=16. (A) Representative images of DIV15 littermates for a number of behavioral parameters. They parameters. behavioral of number a for littermates SA/SA +/+ DIV14 DIV14 DIV14 Pclo +/+ SA/SA slow: p=0.018) (Fig. 15F-H). A similar, but not significant +/+ SA/SA +/+ SA/SA Pclo m. No differences were detected in either DIV14 or DIV15 neurons in (B) τ μ 2 1 0 0

1 0

50

1,5 0,5

1,2 0,8 0,6 0,4 0,2

synapse area synapse (µm2) 200 150 100 length dendrite n=18, (mm) synapse number synapse B C D +/+ n=21. DIV15 hippocampal neurons were stained for synaptic marker Synapsin SA/SA Pclo Pclo mice to their their to mice Synapsin MAP2 SA/SA fast: p<0.0001; τ n=16, +/+ Pclo neurons. In line with these findings, the vesicularrelease probability (inferred Pclo SA/SA PCLO sequence variation did not produce any effect of neuronal synaptogenesis or dendrite length mice have normal body weight and general health expressing neurons. In summary, these data indicate enhanced release probability in neurons causes increased neurotransmitter release and faster synaptic depression. (bottom) neurons, scale bar 20 SA / Pclo SA SA/SA SA/SA SA/SA clo igure 13: showed no differencesin general health, body weight,muscle strength, or motor coordination (see Supplementary information). Their behavioral phenotype was firststudied in an automated This eliminates a postsynaptic contribution to the altered evoked current Pcloamplitude in naïve from the ratio of EPSC and RRP charge), was elevated in confirmed by faster rundown of EPSC size during high frequencybi-exponential stimulation as( fitted with a following high frequency stimulation was unaffected (Fig. 15I). Moreover, spontaneous vesicle releasefrequency andkinetics werealsounaffected in effect, was observed for paired pulseratio in We compared compared We Pclo and dendritic marker MAP2, Pclo P synapse number per cell (DIV14 p=0.766, DIV15 p=0.944), (C) synapse(C)cellper(DIV14p=0.554,areaDIV15 p=0.407), synapse DIV15number p=0.944), cellper(DIV14p=0.766, (D) dendrite length (DIV14 DIV15 p=0.567, orp=0.495), (E) Scholl analysis. All data shown as mean values±SEM. the release kinetics were unaffected (Fig. 15A and Fig. S11A). The size of thepoolreadily (Fig. 15E-F), releasable estimated by back-extrapolation from the cumulative current released during 2.5a second 40Hztrain stimulation (Schneggenburger etal., 1999), wassimilar between F and branching. DIV14 hippocampal neurons were stained for synaptic markerand VAMP2,Piccolo, dendritic marker MAP2, and A Table 3: Immunoprecipitation results from P77 mouse hippocampus and cortex (crude synaptosomal fraction). polyclonal Piccolo antibody (gp44α) and rabbit polyclonal Piccolo antibody (rb44α). Average spectral counts Left: Rabbit polyclonal antibody against Piccolo, rb44α (Cases-Langhoff et al., 1996). Shown mean spectral from 4 precipitations for control condition (no antibody) and 2 immunoprecipitations with each antibody are count from 4 precipitations, of proteins present in significantly higher levels in Pclo+/+, PcloSA/SA, or both, relative shown. All proteins with spectral counts of 1 and higher were included if their levels were significantly higher in to the control condition (no antibody), proteins with average spectral count in both conditions of 1 or less not at least one of the antibody precipitations compared to the control condition. P values were calculated using included. Beta-binomial p-value shown for comparison of Pclo+/+ and PcloSA/SA conditions. Right: independent the beta-binomial model for spectral count analysis. precipitations from Pclo+/+ cortical tissue crude synaptosomal fractions were performed with guinea pig

+/+ SA/SA +/+ Protein Pclo vs Pclo , rb44 antibody Pclo only accession Control p value Control Protein name number (no antibody) Pclo+/+ PcloSA/SA +/+ - SA/SA (no antibody) gp44 antibody rb44 antibody p value control-gp44 p value control-rb44 Piccolo Q9QYX7 0 9.25 15.75 0.0176 8 22 0,00001 0,0001 8 Rufy3 Q9D394 0 2.5 2 0.6165 0 2 12,5 0,0042 0,0002 Serine-threonine kinase receptor-associated Q9Z1Z2 0 2.5 0.25 0.0127 n/a n/a n/a n/a n/a protein (STRAP) Rho guanine nucleotide exchange factor 2 Q60875 n/a n/a n/a n/a 0,25 5,5 4,5 0,0047 0,0091 Dynamin-1-like protein E9PUD2 n/a n/a n/a n/a 0 13 0 0,0015 1 Creatine kinase B-type Q04447 n/a n/a n/a n/a 0 3 0,5 0,0024 0,1656 Mapk1 P63085 0 2.5 1.25 0.1030 n/a n/a n/a n/a n/a 26S proteasome non-ATP-ase regulatory subunit 7 P26516 0 2 0.75 0.1245 n/a n/a n/a n/a n/a α-1,3/1,6-mannosyltransferase ALG2 Q9DBE8 0 1.5 1 0.1995 n/a n/a n/a n/a n/a Guanine nucleotide-binding protein-like 1 P36916 0 1.75 2.5 0.4032 n/a n/a n/a n/a n/a Dual specificity protein phosphatase 3 Q9D7X3 0 1 1.5 0.5256 n/a n/a n/a n/a n/a Proteasome activator complex subunit 4 Q5SSW2 0 1.75 1.25 0.5627 n/a n/a n/a n/a n/a Low MW phosphotyrosine protein phosphatase Q9D358 0 1.5 2 0.5923 n/a n/a n/a n/a n/a Ser/Thr protein phosphatase PP1β P62141 0 4.5 5.5 0.6011 n/a n/a n/a n/a n/a Ser/Thr protein phosphatase PP1γ P63087 0 2.75 4 0.3693 n/a n/a n/a n/a n/a Ser/Thr protein phosphatase PP1α P62137 0 2 2.5 0.7961 n/a n/a n/a n/a n/a Ras-related C3 botulinum toxin substrate 1 P63001 0 1.75 2 0.9999 n/a n/a n/a n/a n/a Elavl4 Q61701 0 1.75 1.75 0.9999 n/a n/a n/a n/a n/a Leucine zipper protein 1 Q8R4U7 n/a n/a n/a n/a 0 4,5 0 0,0328 1 IQ motif and SEC7 domain-containing protein 1 E9PUA3 n/a n/a n/a n/a 0 1,5 1,5 0,0378 0,04 AP-2 complex subunit beta Q9DBG3-2 n/a n/a n/a n/a 3,25 5 8,5 0,458 0,0333 Glyceraldehyde-3-phosphate dehydrogenase P16858 n/a n/a n/a n/a 2,75 2,5 9 0,5003 0,0138 F-actin-capping protein subunit alpha-2 P47754E n/a n/a n/a n/a 2,25 2 7 0,8513 0,0478 Centrosomal protein of 170 kDa H7BX26 n/a n/a n/a n/a 0 0 10 1 0,0063 Rho guanine nucleotide exchange factor 17 O54991 n/a n/a n/a n/a 0 0 4,5 1 0,0339 Alanine aminotransferase 2 Q8BGT5 n/a n/a n/a n/a 0 0 2,5 1 0,0042 Contactin-associated protein 1 Q80U35 n/a n/a n/a n/a 0 0 2 1 0,0389 cGMP-dependent protein kinase 1 P0C605 n/a n/a n/a n/a 0 0 1,5 1 0,04 WD repeat-containing protein 47 Q8CGF6 n/a n/a n/a n/a 0 0 1,5 1 0,04

98 Characterization of Pclo SA/SA mouse model 4 99 1 1 8 n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 0,04 0,04 0,04 0,1656 0,0138 0,0333 0,0339 0,0091 0,0389 0,0478 0,0063 0,0042 0,0002 p value control-rb44 ). Average spectral counts α 1 1 1 1 1 1 n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 0,458 0,8513 0,0015 0,0378 0,0328 0,5003 0,0001 0,0024 0,0047 0,0042 p value control-gp44 only +/+ Pclo 7 2 9 0 0 10 1,5 1,5 1,5 2,5 8,5 4,5 4,5 0,5 n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 12,5 0,00001 rb44 antibody ) and rabbit polyclonal Piccolo antibody (rb44 α 3 5 2 2 0 0 0 0 0 0 13 22 1,5 5,5 2,5 4,5 n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a gp44 antibody

8 0 0 0 0 0 0 0 0 0 0 0 n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 2,75 3,25 2,25 0,25 ontrol C (no antibody) polyclonal Piccolo antibody (gp44 from 4 precipitations for control condition (no antibody) and 2 immunoprecipitations withshown. Alleach proteins withantibody spectral countsare of 1 and higher were included if their levels were significantly higher in at least one of the antibody precipitations compared to the control condition. the beta-binomialP values model werefor spectralcalculated count analysis.using

SA / SA n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 0.1245 0.1995 0.6011 0.6165 0.0127 0.7961 0.0176 0.5923 0.5627 0.1030 0.3693 0.5256 0.4032 0.9999 0.9999 p value +/+ - SA / , orboth,, relative SA 1 2 2 2 4 1.5 5.5 2.5 2.5 n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 1.75 1.25 1.25 0.75 0.25 15.75 clo SA/SA P , rb44 antibody Pclo

, +/+ SA/SA 1 +/+ 2 2 1.5 1.5 2.5 2.5 2.5 4.5 n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 1.75 1.75 1.75 1.75 2.75 9.25 clo P Pclo conditions. Right: independent

Pclo vs

SA/SA

+/+ Pclo Pclo 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a ontrol C and +/+ (no antibody) Pclo 7 3 (Cases-Langhoff et al.,1996). Shown mean spectral X X 26 α X rotein P62141 P62137 P26516 P36916 P16858 Q61701 P63001 P63087 P63085 O54991 Q9Z1Z2 P P0C605 E9PUA3 H7B Q60875 E9PUD2 Q9D358 Q80U35 Q9D7 number Q5SSW2 Q8BGT5 Q9D394 Q04447 P47754E Q8R4U7 Q8CGF6 Q9DBE8 Q9QY accession Q9DBG3-2 β γ α cortical tissue crude synaptosomal fractions were performed with guinea pig +/+ Pclo Immunoprecipitation results from P77 mouse hippocampus and cortex (crude synaptosomal fraction). -1,3/1,6-mannosyltransferase ALG2 rotein name able 3: precipitations from T Left: Rabbit polyclonal antibody against Piccolo, rb44 P cGMP-dependent protein kinase 1 WD repeat-containing protein 47 Dynamin-1-like protein Creatine kinase B-type Mapk1 7 subunit regulatory non-ATP-ase proteasome 26S α Guanine nucleotide-binding protein-like 1 Dual specificity protein phosphatase 3 Proteasome activator complex subunit 4 Low MW phosphotyrosine protein phosphatase Ser/Thr protein phosphatase PP1 Ser/Thr protein phosphatase PP1 Ser/Thr protein phosphatase PP1 Ras-related C3 botulinum toxin substrate 1 Elavl4 Leucine zipper protein 1 IQ motif and SEC7 domain-containing protein 1 AP-2 complex subunit beta Glyceraldehyde-3-phosphate dehydrogenase F-actin-capping protein subunit alpha-2 Centrosomal protein of 170 kDa Rho guanine nucleotide exchange factor 17 Alanine aminotransferase 2 Contactin-associated protein 1 Piccolo Rufy3 Serine-threonine kinase receptor-associated protein (STRAP) Rho guanine nucleotide exchange factor 2 countfromprecipitations,4 ofproteins present insignificantly higher levels in to the control condition (no antibody), proteins with average spectral count in both conditions of 1 or less not included. Beta-binomial p-value shown for comparison of SA/SA A Piccolo+/+ Piccolo C D 2000 A 2000

1500 1500

Piccolo 1000 1000

500 500 Vamp2 intensity (AU) (AU) intensity Vamp2 Piccolo intensity (AU) intensity Piccolo Vamp-2 0 0 +/+ SA/SA +/+ SA/SA

B E F 1,5 2500 ** MAP2 2000 1 1500

1000

Merge 0,5

B 500 Piccolo/Vamp2 ratio Piccolo/Vamp2 0 (AU) intensity Synapsin 0 +/+ SA/SA +/+ SA/SA Synapsin Merge

Figure 14 Synaptic PCLO expression relative to VAMP2 is changed in PcloSA/SA neurons. DIV14 hippocampal neurons were stained for synaptic marker VAMP2, dendritic marker MAP2, and Piccolo, Pclo+/+ n=16, PcloSA/SA n=21. DIV15 hippocampal neurons were stained for synaptic marker Synapsin and dendritic marker MAP2, Pclo+/+ n=18, PcloSA/SA n=16. (A) Representative images of DIV14 Pclo+/+ (left), and PcloSA/SA (right) dendrites, scale bar 10μm. (B) Representative images of Synapsin signal in DIV15 Pclo+/+ (left), and PcloSA/SA (right) neurons, scale bar 10μm. No differences were detected in (C) Piccolo signal intensity (p=0.294), (D) VAMP2 intensity (p=0.337), or (F) Synapsin intensity (p=0.098). Significant difference was found in (E) Piccolo/VAMP2 levels ratio (t test p=0.005, Kruskal- Wallis p=0.01). All data shown as mean values±SEM.

home cage screen using the PhenoTyper (Maroteaux et al., 2012), in which their spontaneous activity and responses to several behavioral challenges were recorded for 7 days without human intervention (Maroteaux et al., 2012). Subsequently, PcloSA/SA mice and their Pclo+/+ littermates were introduced to a battery of classical behavioral tests with an increasing emotional load.

PcloSA/SA mice have normal spontaneous behavior in the home cage Major depression disorder affects everyday life and can disturb general activity, sleeping and eating habits. To test whether PcloSA/SA mice showed such differences in a familiar environment, we used an unsupervised automated home cage, the PhenoTyper. The spontaneous activity of both groups of mice was assessed during the first 3 days. PcloSA/SA and Pclo+/+ mice showed no significant differences in their circadian rhythm, with more activity in the dark phase than in the light phase (Fig. 16A) and two activity peaks, one at the beginning and one at the end of the dark phase, which are characteristic of mice with C57Bl/6J background (Maroteaux et al., 2012). The automated behavioral phenotyping in the homecage produces a large quantity of parameters, which are summarized in 115 most informative parameters. No differences were observed between PcloSA/SA and Pclo+/+ mice except two (p-value of 0.04), which is within the expected false positive discovery range (see table S18 for details). The six parameters related to anxiety traits were all normal, as shown in Fig. 16B-G. Taken together, PcloSA/SA mice showed normal spontaneous behavior in a familiar environment.

100 Characterization of Pclo SA/SA mouse model 4 101

+/+ n=46. Pclo Piccolo +/+ Piccolo SA/SA Piccolo SA/SA +/+ SA/SA +/+ SA/SA Pclo 0.2 Hz Time (s) Time 0

5 4 3 2 1 0 3.0 2.5 2.0 1.5 1.0 0.5 3.5

RRP (nC) RRP half width (ms) width half

D Spontaneous release Spontaneous n=44, M +/+ 0 10 20 30 40 50 40 Hz 0 Pclo

80 60 40 20

120 100 Normalized amplitude (%) amplitude Normalized I Piccolo +/+ Piccolo SA/SA Piccolo +/+ SA/SA hippocampal autaptic neurons show a n=46. (F) Normalized release during a

40 Hz Time (s) Time * SA/SA 5 0 SA/SA

***

40 35 30 25 20 15 10 +/+ SA/SA (pA) amplitude +/+ SA/SA

Pclo Pclo Spontaneous release Spontaneous L 0.0 0.5 1.0 1.5 2.0 2.5 n=21. All data are shown as means ± SEM. 0 8 6 4 2

and 10 0.6 0.4 0.2 0.0

Cumulative charge (nC) charge Cumulative 0.09 0.06 0.03 0.00

n=44,

SA/SA +/+

slow (s) slow Decay fast (s) fast Decay τ C τ +/+ slow component (p=0.018). τ Pclo Pclo Pclo H G cells. (A) Averaged first evoked current traces. (B) First evoked n=47. (C) n=47. Readily releasable pool (RRP) size is estimated by back- n=20, +/+ SA/SA SA/SA +/+ * SA/SA n=46. (E) Vesicular release probability, Pvr, was calculated by dividing Pclo 5 0 +/+ SA/SA

Pclo 20 15 10

Pclo SA/SA Piccolo +/+ Piccolo SA/SA Piccolo (Hz) frequency

40 Hz

Time (s) Time Spontaneous release Spontaneous 5 0

25 20 15 10 K EPSC amplitude (nA) amplitude EPSC 0.0 0.25 0.5 n=46, 0

B 80 60 40 20 +/+

neurons. cells is increased with 30% (p=0.047), without a significant effect on kinetics of 120 100 Normalized amplitude (%) amplitude Normalized n=44, Pclo F SA/SA SA/SA Pclo +/+ Pclo Pclo Pclo 25 ms 5 nA * fast component (p<0.0001) and (H) a n=42. (I) Representative spontaneous release traces. (J) Spontaneous release frequency and (K/L) τ +/+ SA/SA variation. Electrophysiological characterization of SA/SA mice show normal avoidance learning in the home cage

SA / Pclo Pclo

SA 200 ms

0.08 0.06 0.04 0.02 0.00 0.16 0.14 0.12 0.10 Pvr clo igure 15: 100 pA PCLO +/+ PCLO SA/SA PCLO E PCLO +/+ PCLO SA/SA PCLO A kinetics are normal in P n=40, F mild increase in synaptic transmission in 40 Hz train stimulation shows altered release kinetics, fit withwhich (G) cana be best described by a double exponential release (data not shown). (I) Also, recovery from this train stimulation, evaluated by subsequent 0.2 Hz stimulation, is unaffected. On day 5 and 6 of the (Maroteaux PhenoTyper et protocol,al., 2012). mice The underwentPhenoTyper contains an a avoidanceshelter a preferencewith for learning onetwo of theentries. two test entrances Mice(without biasdevelop for left or right, see Maroteaux et al.,2012). The preferred entrance is automatically detected on day 4. Subsequently, on day 5 and 6, each entry through the preferred entrance, but not the bright other illuminationentrance, isof sanctionedthe shelter. by A preference index (Maroteaux behavioralet response al.,to this challenge,2012) which describesinvolves cognitive the aspects (recognizing that one extrapolation from the cumulative charge released during a 2.5 second 40 Hz train stimulation. Averaged traces are shown. The dotted lines represent linear fits on the last 40 stimuli.by the (D) RRP size is not significantly affected EPSC charge by the RRP charge estimated in (D). current amplitude of J entry is sanctioned while the other is not) and aspects of behavioral flexibility (adjusting to the challenge and shifting preference). Mice of both groups developed a preferred entrance up to day 4 to a similar extent and actively shifted their preference toward the other entrance of the shelter also to a similar extent (Fig. 16H). Taken together, these results indicate no abnormality in avoidance learning in a familiar environment for PcloSA/SA mice.

A SA/SA 6000 +/+

4000 Distance [cm]

2000

0 1 2 3 Days 3 x10 E C D 2.0 1 B 1200 25 20 1.5 0.8 800 15 0.6 1.0 10 0.4 light [s] Light [s] 400 Light [s]

0.5 Dark / light 5 0.2 feeding duration Activity Activity duration Mean activity Meanactivity bout

0 0 0.0 Activity durationindex 0 +/+ SA/SA +/+ SA/SA +/+ SA/SA +/+ SA/SA

Shelter task F 10 150 H G 0.4 8 100 6 0.2 Dark Light 4 50 0.0 2 On shelter visits On shelter visits −0.2

0 0 Preference index +/+ +/+ SA/SA +/+ SA/SA −0.4 SA/SA

1 2 3 4 5 6 7 Days

Figure 16: Home cage behavior. (A) Circadian activity. (B) Activity duration light is the total time a mouse was active during the light phase. (C) Mean activity bout-light, is the average of time a segment of activity lasts in the light phase. (D) Feeding duration-light, is the time spent in the feeder zone of the cage during the light phase. (E) Activity duration index is the ratio of activity between the dark and the light phase. (F) On shelter visits-light, is the number of visits on top of the shelter in the light phase. (G) On shelter visits-dark, is the number visits on top of the shelter in the dark phase. PcloSA/SA mice showed no difference in spontaneous behavior compared to Pclo+/+ (H) Preference index represents the change of preference cause by the shelter task on day 5 and 6. The horizontal dash line represents the no preference line. Grey background bars represent the dark phase. All bar graph are means and SEM.

102 Characterization of Pclo SA/SA mouse model 4 103

+/+ SA/SA -val Pclo 0.13 0.12 0.41 0.93 0.77 0.25 0.74 0.87 0.85 0.78 0.56 0.34 0.24 0.44 0.84 P t-test Pclo and SA/SA 1.35 8.18 1.24 1.86 8.92 1.00 17.51 sem 19.95 10.38 30.33 67.86 72.89 23.08 126.80 285.89 Pclo SA / SA 5.53 7.80 34.15 11.80 73.33 25.93 121.12 133.17 28.82 143.81 mean 277.96 833.90 144.80 1964.29 4656.66 5.85 2.77 1.80 1.80 0.95 9.86 19.13 sem 12.42 35.67 23.67 41.79 54.97 69.28 90.25 261.24 +/+ 5.67 31.33 12.47 131.13 10.25 49.73 49.76 mutation does not affect anxiety-like behavior in mice. 30.46 151.96 mean 141.82 143.68 949.81

1833.13 260.46 4426.94 SA/SA mice show no difference in anxiety behavior. Pclo SA/SA Pclo arameters Latency [s] latency open arms [s] # Visits open arms Total Total distance [cm] Latency [s] # Visits Time in lit side [s] Distance [cm] P Latency center [s] Distance in open arms [cm] arms open in Distance # Visits center Total Total distance [cm] Time in center [s] Distance in center [cm] Time in open arms [s] reacted in a similar way to the different anxietytests. mice in four anxiety-related tests (Table 4). The elevated plus maze, open +/+ +/+ Pclo Pclo and

SA/SA mice show normal cognitive behavior mice have normal anxiety-related behavior Anxiety-related test performances.

and SA SA / / Pclo SA SA SA/SA ests able 4: clo clo Novel home cage- induced hypophagia Elevated plus maze Dark-light box T Open field T Both P field and dark-light box are all based on conflicting situations with aThe negative animal has to choose betweenreinforcement. a safe dark or confined place where it can be in contact with the walls (thigmotaxis) and the tendency to explore an open and brightly lit environment and be exposed. No significant differences were observed in the total distance moved, the latency and the time spent in the exposed environment (i.e. the open arms, center of the field andcompartment, lit respectively). The fourth test, the novelty induced hypophagia uses the conflict appetitivebetweenhighlypalatable(pieceanenvironmentnovelof andfood). reward a P Cognitive impairment is an important component in the diagnostic criteria of MDD (Austin et al., 2001; Harvey, 2007). Therefore we performed several cognition tests in MDD and anxiety disorders have a high comorbidity(Hirschfeld, 2001). Pclo Therefore, we tested mice. The Novel object recognition test is designed to evaluate novelty recognition between two objects in three phases. In the first phase (sample phase) both groups of mice showed no preferencesfor either two objects, in adjacent corners, which have similar shapes but different mice showed no significant differences in the latency to eat the piece of food, in the new home cage. Therefore, we conclude that colors. In the second phase (the place phase) object 1 is moved to the corner opposite to object 2. Both groups now spent more time visiting the displaced object (positive visit and duration index) implying they remembered object 2 and its location. In the last phase, object 1 is exchanged with a new object with a different shape. PcloSA/SA and Pclo+/+ mice both spent more time visiting the new object 1 (Fig. 17A-B). Hence, the PcloSA/SA mutation did not affect memory or novelty seeking in the novel object recognition test. We tested spatial learning and memory in the Barnes maze, in which the mouse has to remember the location of the escape hole among 23 other holes, on a highly illuminated circular platform surrounded by distinctive visual cues (Barnes, 1979). No differences between PcloSA/SA and Pclo+/+ mice were observed in the latency to find and enter the escape box (from day 1 to day 5, PcloSA/SA: 161±20s to 61±14s and Pclo+/+: 163±21s to 54±11s) or in the distance travelled to reach it (Fig. 17C-D). On day 6, we performed a probe trial, removing the escape box, and measured the probability of visiting holes in the target octant (octant where the escape box was previously located). Both groups showed a similar probability, higher than chance (0.125) to visit the target octant (PcloSA/SA: 0.19±0.02 and Pclo+/+: 0.22±0.02) (Fig. 17E). Hence, PcloSA/SA mice showed no deficit in spatial learning and memory in the Barnes maze. In the T-maze a spontaneous alternation task was performed. During exploration, normal mice alternate between the two arms, which requires the recollection which arm was visited last (Deacon and Rawlins, 2006). The two groups of mice showed a similar percentage of alternation, higher than 80% (Fig. 17F). Taken together, these data demonstrate that PcloSA/SA mice show normal spatial recognition, learning and memory.

PcloSA/SA mice have normal sensorimotor gating, fear memory and depression-like behavior We also tested sensorimotor gating, a brain mechanism which filters inputs to the brain, known to be affected in several human brain disorders. The Pre-pulse inhibition (PPI) test measures sensorimotor gating in mice. We measured the acoustic startle response (ASR) in a range of 65 to 120db. Both PcloSA/SA and Pclo+/+ showed an increasing startle response with the increasing sound intensity (Fig. 17G). PPI with a delay of 30 ms (Fig. 17H) and 100 ms (Fig. 17I) were not altered in PcloSA/SA compared to the Pclo+/+. Hence, PcloSA/SA mice have no significant deficits in sensorimotor gating. A fear conditioning test was used to identify deficits in acquisition, expression and extinction of fear-related memories. During the fear conditioning training both groups received a condition stimulus in a given context (fear conditioning box) and subsequently experienced a foot shock. All mice perceived the tone showing increased scanning behavior) and responded to the shock by acutely increasing their activity (from 9 to 34 cm/s for PcloSA/SA and from 9.2 to 38.5cm/s for Pclo+/+). When presented again to the context, 24h later, PcloSA/SA mice as well as Pclo+/+ reduced their activity (to 5.2cm/s for PcloSA/SA and 5.5 cm/s for Pclo+/+) and increased the number of freezing events, showing an association between the foot shock and the context box. When placed in the new context, activity as well as freezing events were normal, and the conditioning tone did not affect their activity or freezing behavior (Fig. 17J-K). Hence, PcloSA/SA mice show normal association of fear with the context. In the most widely used depression assay, the forced swim test, both groups were immobile for more than 60% of the time. Furthermore, both PcloSA/SA and Pclo+/+ took respectively 82 ± 8s

104 Characterization of Pclo SA/SA mouse model 4 105 SA/SA SA/SA SA/SA +/+ +/+ +/+ 0 0

80 60 40 20 80 60 40 20

0.4 0.3 0.2 0.1 0.0

100 (s) strategy Change 100 % Alternation Probe trial (probability) trial Probe [db] intensity M 65 67 71 79 Av E F 0

80 60 40 20 SA/SA %PPI (100ms) %PPI +/+ I +/+ SA/SA

0 80 60 (%) 40 Floating 20 L Days Days

represents the group mean shown with SEM. + tone + ■

intensity [db] intensity new context new

65 67 71 79 Av

1 2 3 4 5 context New 1 2 3 4 5 Context 24h 0

0 0

60 40 20 80

100 (30ms) %PPI 800 600 400 200 200

Context Escape latency (s) latency Escape (cm) box to Distance

8 6 4 2 0 H events Freezing D C K

+/+ SA/SA

+/+ SA/SA tone + 1 2 context New Object

mice vs control mice. 1. Learning and memory. Novel object recognition. (A) Visit New context New Context represent individual data points and +/+ 24h ○ SA/SA 1 2 Place intensity [db] intensity

Pclo

Post-shock Shock -Us Shock SA/SA

1 2 Tone-CS Sample 65 75 85 95 105 115

Memory in Context 0 0

0 0 5

0.8 0.4 0.8 0.4 40 10 35 cm/s

-0.4 -0.8 -0.4 -0.8

600 400 200 Duration index Duration Visit index Visit ASR [IU] ASR igure 17: A 2. Startle response and fear memory G B 1. Learning and memory 1. Learning and J index. (B) Duration index. Barnes maze. (C) Escape latency. (D) distance travelled to the escape box. (E) Probe trial. (F) T-maze alternation task. 2. Startle response and fear memory. (G) Acoustic startle (I)response. 100ms delayPercentage betweenof the PPI pre-pulsewith and (H)the pulse.30ms (J) andFear conditioning with training Numberand retention of test. (K)freezing event (2s threshold) (L) before changingForced the swimmingswim strategy. test. (M) Percentage of time floating. (N) Latency F and 85s ± 6s before changing their swimming strategy from active to more passive (Fig. 17O-P). Therefore, PcloSA/SA variation does not affect sensorimotor gating or fear response and memory and depression-related behavior.

Discussion In this study, we described and characterized a knock-in mouse model of major depressive disorder that carries the PCLO allelic variation rs2522833, which has shown suggestive association with MDD. This characterization revealed several abnormalities in molecular/cellular but not behavioral parameters tested. We found a slight but significant increase in Piccolo protein levels in synaptosomal lysates and in the synaptic expression of the Piccolo protein relative to a synaptic marker VAMP2. Secondly, excitatory synaptic transmission in cultured neurons from PcloSA/SA mice was increased, without an accompanying increase in readily releasable pool size or release probability. These differences were all similar in percentage change (30%), as might be expected from a subtle change in protein structure. In contrast to these cellular changes, the analysis of PcloSA/SA mouse behavior showed consistently no significant differences in general health, spontaneous or challenged home cage behavior, coordination, anxiety, learning, memory, fear response or despair. In cultured hippocampal neurons of PcloSA/SA mice, we observed a significant increase in the ratio of Piccolo to VAMP2 levels at individual synapses, which was driven by a sub-population of synapses in the PcloSA/SA group. VAMP2 is a SV membrane protein and the increased ratio therefore suggests that in PcloSA/SA synapses, SV clustering is slightly impaired. On a population average, Piccolo and VAMP2 expression levels did not change (Fig 14C,D). This correlation at the single synapse level and lack thereof at the population level is consistent with recent observations that synaptic protein levels are regulated in concert depending on local activity and local proteasomal break-down of synaptic proteins including Piccolo (Lazarevic et al., 2011). The synaptic levels of Synapsin tended to follow changes in Piccolo, as expected from their functional relation previously described in hippocampal synapses (Leal-Ortiz et al., 2008). PcloSA/SA neurons exhibited 30% higher synaptic vesicle release upon stimulation with a single action potential. The enhanced evoked release could be attributed to an increase in synaptic release probability with a concomitant effect on synaptic depression during high frequency stimulation. A similar trend towards stronger paired pulse depression was observed, but this effect was not significant. Because this parameter was tested at the end of the experimental protocol, the cells were not naïve anymore. The current study had sufficient power to detect moderate to large effects (Cohen’s d of .59 for electrophysiology, .99 for imaging, and 1.04 for behavioral experiments, details in the Materials and Methods), but was unable to detect smaller effects, which may still exist. In fact, because the effect size for the association between rs2522833 and MDD observed in the GWAS (Sullivan et al., 2009) was very small, also the phenotypic effect of this genetic variation is likely to be small. Nevertheless, at least two significant cellular differences were detected in the highly homogeneous context of inbred mice. In the respect, this study is one of the very first in successfully identifying functional consequences of single nucleotide coding variation associated with complex traits.

106 Characterization of Pclo SA/SA mouse model 4 107 gene, Pclo in MDD, if present, is (Sullivan et al., 2009), this PCLO PCLO knock-in mice was performedmicewasknock-in using Pclo TGTGCTAGATAAATCAATCAATACCTGG-3’. CATCTGGACAACACTCCTCGGTGGTATC-3’ AGC GCT mice might not show a strong behavioral phenotype without additional genetic or SA/SA Pclo Upon completion of the ES clone expansion, Southern analysisconfirmation with 5’ probe, 3’ The 5’ and 3’ arm fragments were cloned in the LoxNwCD vector sequentially, and the The S4742A point mutation, located in exon 19 of 3’ homology arm, was introduced by site- BAC clone RP23-451C1 was used as template for amplification of 5’ homology arm and the Althoughthe initial GWAS reported suggestive evidence for aterial & methods clo knock-in mouse generation environmental factors. The molecular changes we observed might slightly increase the risk for MDD under certain circumstances or in combination with other molecular changes. probe, probe, and neo probe, as well point as mutation sequencing S4742A were performed. Blastocyst insertion and orientation were confirmed byrestriction digestionfinal and vector end-sequencing. was obtained by The standard molecular cloning. Aside from the homology arms,final the vector also contained loxP sequences flanking the Neo expression cassetteselection (for positive of the ES cells), and resistant a ES clones DTA derived expressionfrom random insertion cassette events). The (forfinal vector was negative confirmed by both selection restriction of endonuclease G418- digestion and by end sequencing linearizinganalysis. the final vector priorNotI to electroporationwas (fig. 6). used for directed mutagenesis. Primers used for introducing the S4742A mutation (mutated sequence in bold): 5’-CCAGGTATTGATTGATTTATCTAGCACA forward: M P For the creation of a mouse line carrying S4742A knock-in within exon19 of mouse 3’ homology arm. The following primers were used (lowerendonuclease sites):case letters represent restriction 5’ arm forward: 5’-tcttgtgcggccgcATCACACTTGTATCTCACAACAGCCTTCTC-3’ 5’ arm reverse 5’-acaacctaggTTGTGAATCCAAGATGCTTAATACATTTGTGG-3’ 3’ arm forward: 5’-caaaggcgcgccAGCACAATTCTCAGTATCATGGCTG-3’ 3’ arm reverse: 5’-caacatcgcgATGTTATTCATGCAAACAGAAGTGTGC-3’ we contractedwe Caliper LifeSciences. generationThe of homologous recombination in C57BL/6 mouse embryonic stem cells and subsequent injection of selected ES cells into wild-type blastocysts. 5’-GATACCACCGAGGAGTGTTGTCCAGATG reverse: unlikely to be very large. It is also widely accepted that MDD is a multifactorial disease, brought on by a combination of many genetic and environmental factors. This might be another reason why wasnotsupported subsequentresultsbyfrom GWAS studies (KohliLewisetal., et2011; al., 2010; from or2012), al., et Wray 2011; al.,et Shyn 2011; al.,et Shi 2010; Rietschelal., et2010; al.,Muglia et the recently published mega-analysis(Ripke et al., 2013) including 9240 controls. MDD Based on these casesstudies and our currentand results, the role of9519 injections were performed with two clones that were confirmed to be correctly targeted and had a single neo insertion. Male chimeras were bred with C57BL/6 females to generate heterozygotes. The heterozygote PcloSA/+ mice were bred with Cre delete to excise the Neo sequence. This allele was named SA (for S4742A). Subsequently PcloSA/+ mice were maintained in the C57BL/6 strain. Homozygous males and females were used for experiments. All animal procedures were performed according to the Dutch law and ethical guidelines of the VU University Amsterdam and complied with the European Council Directive (86/609/EEC).

Cloning We amplified the Pclo C2A domain from human cDNA with forward primer ggggatccGTC TCTCATCCAATTACAGGA and reverse primer CTCAAAGAACAGACTGAAAGCGTAgcggccgctaa. The fragment was cloned into pGEX4T3 vector via BamHI and NotI sites. The amplified fragment had the S4841A SNP (which is not surprising in view of the minor allele frequency of rs2522833), so we needed to mutate it to obtain the major allele (‘+/+’). The following primers were used for this purpose: Forward 5’- GTATTGATTGATTTATCTAGCACATCTCACCTCGATAACACTCCAAG-3’ and reverse 5’- CTTGGAGTGTTATCGAGGTGAGATGTGCTAGATAAATCAATCAATAC -3’ (mutated residues in bold).

Antibodies Piccolo rabbit or guinea pig polyclonal antibody, raised against the same antigen (rb44α or gp44α), was used for immunocytochemistry (1:1000), Western blotting (1:6000), or immuno- precipitations (2µg antibody in each 10mg protein Cases-Langhoff et al., 1996); MAP2 chicken polyclonal 1:20000 (Abcam); Synapsin1-2 rabbit polyclonal 1:1000 (P610; gift from T.C. Sudhof, Stanford University School of Medicine, Stanford, CA); VAMP2 mouse monoclonal 1:1000 (SySy);

Western Blotting To resolve Piccolo on gel, we made use of precast gradient (3-8%) NuPAGE Tris-Acetate Gels in Novex Tris-Acetate SDS Running Buffer (Invitrogen). The gels were run at 150V and transferred to PVDF membranes at 200 mA overnight or at 350 mA for 3 hours at 4°C (transfer buffer with 0.1% SDS). Membranes were blocked in 5% milk (Merck) and 0.1% albumin (Invitrogen) in TBST (20 mM Tris, 137 mM NaCl and 0.1% Tween 20) and incubated with primary antibody mix for 2h at 4°C in TBST. After 3 washes in TBST, membranes were incubated with alkaline phosphatase conjugated secondary antibody mix for 1 hour at 4°C (Dako, 1:10000). Finally, the membranes were washed in TBST and incubated for 5 min with enhanced chemifluorescence substrate, followed by imaging on a FLA5000 image reader (Fujifilm).

Phospholipid binding assays For liposome aggregation experiments, we prepared small unilamellar liposomes composed of 80% phosphatidylcholine (18:1 [Δ9-Cis] 1,2-dioleoyl-sn-glycero-3-phosphocholine) and 20% phosphatidylserine (18:1 1,2-dioleoyl-sn-glycero-3-phospho-L-serine sodium salt). Lipids were acquired as solutions in chloroform from Avanti Polar Lipids. Liposomes were prepared

by mixing chlorophorm stock solutions of lipids, drying the lipid mix under N2 stream, and

108 Characterization of Pclo SA/SA mouse model 4 109 , 20mg total α _affi and gp_44_ α -100). After solubilization at 4°C for 30 for 4°C solubilizationat After -100). X L protein, and mixed with an equal volume ofequalvolume an withmixed and protein, L μ g/ μ , calibrated using Fura-2 fluorescence spectroscopy)together with 20 µl of 2+ Several controls were included to confirm that the increase in absorbance was specific: the The DLS assays were performed on SmartSpec3000 spectrophotometer (Bio-Rad), rotein interaction partners A350 increase was not observed after addition of GST, or in absence of phosphatidylserine, and incubating thewith GST-proteins calcium in the absence of liposomes did not cause an increase in the absorbance at 350 nm (excluding the possibility that the absorbance increase was caused by protein aggregation). Single C2 domain was not capable of aggregating necessarywhichcapableprotein,is GST-fusion have dimerizationliposomes:useof viacysteines theGST. in it was P Tissue preparation assay).HippocampiBradford (determinedby protein total of mg experiment,each5 used For we and cortices were removed from 11-week old mice, frozen on dry ice and obtainstored at sufficient-80°C. amountTo ofprotein, for eachprecipitation reaction hippocampusfrom andthree cortex mice were combined. On the day of the experiment,a tissue glass was homogenized Potter-Elvehjem in homogenizer containing ice-cold sucrose, 5 homogenizationmM Hepes at pH 900 7.4) rpm with buffer 12 up and down (320strokes of the piston. mM The lysate was centrifuged at 1000 × g, 4°C, for 10 min. The supernatant was further centrifuged at 16,000 ×gfor 30 min obtain to a pellet enriched in synaptosomes. This pellet was resuspended in 10mM 5concentration ofprotein a HEPESsolution to liposomes. This was used for background reading for the first 160 s.GST-C2A Thenprotein wethat addedprovoked 10liposome µg aggregationof and a corresponding increase in A350. The increase in absorbance (average values of baseline absorbance subtracted values infrom the last 200average s of measurement) was used as read-out of protein-lipid binding. extraction buffer (45 mM HEPES, 300 mM NaCl, 2% Triton 2% NaCl, mM 300 HEPES, mM (45 buffer extraction immediatelysupernatantswereThe min. 20 for 4°C ×g, 16000centrifuged at wassample the min, used for immuno-precipitation or pull-down protein from crude synaptosomal fractions of adult mice experiments. (P77) using a Piccolo-specific antibody We precipitated native Piccolowith an epitope outside the C2A domain, and determined protein levels by mass spectrometry. Immunoprecipitation The supernatant was mixed with antibody (Piccolo rb2_44_ protein measured with Bradford assay for each immunoprecipitation reaction) and incubated at 4 °C overnight, followed by binding to Agarose Protein A beads (Santa Cruz) for 1 hour. measuringabsorbancequartzinterval.cm 350nm scuvette1 at (A350)mixed1awe calcium atIn solution in buffer A (with calcium chelated with EGTAto arange of concentrations between 0 and 2 mM free Ca reconstitutingthe liposomes inbuffer 100 mM NaCl(25mMA 7.4, DTT)and mM 3 Hepes 1 to pH The mg/ml. solution was mixed and vortexed, sonicated on ice (6 times 6 seconds, with at least 1 minute interval between each cycle). The sonicated liposomes we spun down at 13000xg for 30 mins, and supernatant was used for experiments. Proteomics After washing the beads 4 times in 10mM HEPES solution, proteins were eluted from the beads by boiling samples for 5 mins with 20 µl SDS loading buffer. The samples were subjected to separation by SDS-PAGE, and stained with colloidal coomassie. Each sample was split in 5 gel slices according molecular mass (see Fig. S13). To remove the coomassie staining, gel slices were washed twice with 50 mM ammonium bicarbonate/50% acetonitrile, once with 100% acetonitrile, and dried in a speedvac. Proteins were subjected to in-gel digestion by trypsin (Promega), during overnight incubation at 37°C. The resulting peptide solution was dried in a speedvac (and stored at -20C), dissolved in 20 µl 0.1% TFA, and analyzed on an LTQ-Orbitrap Discovery instrument (Thermo Fisher Scientific). Protein identification and relative quantification were performed with MaxQuant software (version 1.3.0.2, Cox and Mann, 2008). Cortex and hippocampus samples were analyzed simultaneously. MS/MS spectra were annotated against the Uniprot Mouse reference proteome database (version 01/2013). Database searches were performed with trypsin/P specificity allowing up to two missed cleavages. N-acetylation of proteins and oxidized methionine were searched as variable modifications. No fixed modifications were used. Mass tolerance was 6ppm for precursor ions and 0.5 Da for fragment ions. The false discovery rates (FDR) for peptide and protein identification were restricted at 5% false discoveries. Spectral counts, normalized to the total number of spectral counts measured in each sample, were used for relative protein quantification. Statistical evaluation of differences in protein abundance was performed with the unpaired beta-binomial test for spectral counts (Pham and Jimenez, 2012). By performing ten parallel precipitations (four each from Pclo+/+ and PcloSA/SA brain tissue, and two precipitations from Pclo+/+ brain tissue without addition of the antibody), we ensured that all conditions and quantities were equal. For subsequent analysis, we included all proteins that had spectral counts of 1 and higher, and were detected in significantly higher levels in the antibody precipitation compared to control (no antibody), Table 3. To analyze the statistical significance of these results, we made use of the beta-binomial model for spectral count analysis (Pham et al., 2010), a statistical tool specifically developed for mass spectrometry data analysis, where results (spectral counts) are discrete values that do not follow normal distribution. For the confirmation of binding between Piccolo and STRAP1, synaptosomal fractions from Pclo+/+ and PcloSA/SA brain tissue were prepared as described above. Piccolo was immunoprecipiated using gp44α antibody, and beads were incubated with lysate of HEK293 cells transfected via calcium phosphate method with STRAP1-flag construct (source). Beads were washed, built in gel loading buffer and resulting protein mix was subjected to gel electrophoresis, Western blotting. The proteins were detected using anti-flag antibody.

Cell culture Dissociated hippocampal cultures were obtained from newborn mice as described previously (de Wit et al., 2009). Hippocampi and cortices were dissected in HBBS (Invitrogen) supplemented with 7 mM Hepes and digested with 0.25% trypsin (Invitrogen) at 37°C for 20 min. After trituration with a fire-polished glass pipette, cells were plated at a density of 25,000 cells/well on top of a pre-grown rat glia monolayer on 25-mm glass coverslips. Cultures were grown in Neurobasal

110 Characterization of Pclo SA/SA mouse model 4 111 -gluconic acid, 10 + , , and penicillin/streptomycin X T 4.1, Noldus Information Technology, Wageningen, Technology, InformationNoldus 4.1, T X -100 and 4% fetalcalf serum4% and(FCS), -100 neuronsthewere X analysis software (Synaptologics BV, Amsterdam, The TM littermates were used. They were all males aged between +/+ Pclo and 15

SA/SA Pclo -Y -Y coordinates of the center of gravity (COG) at a resolution of 15 coordinates X lectrophysiology -100 in PBS. After blocking in 0.1% TritonPBS. in After blocking-100 0.1% in tainings Netherlands) to generate behavioral parameters. 115 activity described parametersin detail previously (Loos wereet al., 2014). The last generated and first 10 min as of each dark and light (all (all from Invitrogen). For autaptic (island) cultures, 25-mm glass coverslips collagen/poly-d-lysine were substrate coated using with a stamp cells/well. 2,000 at top on plated with were neurons which after days, evenly 4–7 for islands these on cultured spaced squares. Glial cells were S Neuronal cultures were fixed with4% formaldehyde in PBS and permeabilized with0.5% Triton X medium medium supplemented with 18 B27, mM Hepes, 0.5 mM GlutaMA mM NaCl, 4.6 mM MgCl2, 4 mM 15K2-ATP, mM Creatine phosphate, 10U/ml phosphocreatine kinase and 1 mM EGTA. pH of both solutions was setand tobuffers 7.3 were ensuredto have an osmolaritymOsm.300Cellsofstimulated were signalsandtemperaturerecordedroomwere at withMulticlampa 700Bamplifier and Digidata1440A (Axon Instruments). Electrical stimulation was done by depolarization from -70 to 30 mV for 0.5 ms. belowOnly 10 cells MΩ, witha leak a currentseries below resistance500 pA and EPSC amplitudes of at leastforanalysis. 500Analysis waspAdone with Clampfit were 10 (Axon Instruments),used Mini Analysis (synaptosoft) and Matlab (Mathworks). Behavioral testing 16 Homozygote 8-10 8-10 weeks at the beginning of the experiment. Animal were single housed cycle in with 12h accessdark- light to food and water ad libitum. Micehome were cages transferred to (PhenoTyper specially designed model 3000, Noldus Netherlands)inthesecond half ofthesubjective h). The behavior light17:00 of– h phase (14:00 Information Technology, Wageningen, The mice was video-tracked for three days as described in detail previously ((Maroteaux et al., 2012) EthoVision on based 2.1.2.0, HTP EthoVision The Netherlands), starting at the first subjective darkcontaining phase (19:00 h). Resulting track files, per s, were processed using AHCODA stainedwithprimary antibodies (atconcentrationsRTforat 1.5h indicated below), washed with withAlexaFluor-conjugatedRT at PBS,incubated andh 1 secondaryfor antibodies (Invitrogen). Secondary antibodies we always usedDabco–Mowiol (Invitrogen) at and examined 1:1,000 on a dilution.confocal microscope (Zeiss Coverslips LSM510). were acquiredImages with were a 40× oil objective, mountedNA 1.3, and 0.7× zoom. Basic neuronal with parameters were analyzed in Matlab using SynD (Schmitz et al., 2011). E Whole-cell voltage clamp electrophysiology was performed at room temperature on autaptic hippocampal neurons that had been in culture for 14 to 18 days with borosilicate glass pipettes (2-4MΩ). Extracellular solution contained 10 mM HEPES, 10 mM glucose, mM140 NaCl, 2.5 mM KCl, 4 mM MgCl2 and 4 mM CaCl2. Intracellular solution contained 125 mM K phase were not included in summary statistics, to ensure that a potential asynchrony of the data streams and light regime in the behavior facility would not affect these results. After the end of the PhenoTyper protocol, mice were housed in the animal facility in individual transparent Perspex cages with 12 h dark-light cycle and food and water ad-libitum. Following this week, after analysis of the PhenoTyper data, Ubn1 mice (controls and mutant) were introduced in the test battery going form the least (grip strength meter) to the most stressful test (fear conditioning). The sequence is described in Table 5. A full description of the Methods can be found in chapter 7: Material & Methods

Statistical analysis for behavioral testing Before any analysis was performed, data were examined for outliers (>3 times the SD from the strain mean). All statistical analyses were performed using IBM SPSS statistic 20 (IBM, Armonk, NY, USA). Genotype differences were compared using parametric tests (T-test, ANOVA, repeated-measures ANOVA) whenever normality and homoscedasticity criteria were met. Otherwise, nonparametric tests were performed (Kruskal-Wallis, Mann-Whitney U-test). Nonparametric data are presented as box plots (ends of the box denoting the 25 and 75% interquartile range and the whiskers providing the upper and lower quartile ±1.5 times the interquartile range, respectively, while the line in the box denotes the median). An error probability level of p<0.05 was accepted as statistically significant throughout the study of the classical behavioral tests. For all given comparisons of PhenoTyper results, statistical analysis was based on estimated false discovery rate (FDR) (Verhoeven et al., 2005) with alpha-levels correct by minimum positive FDR with a threshold set at 5%.

Table 5: Behavior tests, sequence and number of mice used

Pclo Experiment +/+ SA/SA Comments/duration/sequence automated home cage 16* 15* 7 days protocol no human interference Body weight 16* 15* age: 7-8 weeks Novelty induced hypophagia 16* 15* 3 days habituation prior test, 10min max Grip strength 16* 15* 5 session front paws Elevated plus maze 16* 15* 5 min Open field 16* 15* 10 min Novel object and place object task 16* 15* 3 phases: sample (S), place (P) recognition and object (O) recognition Dark-light box 16* 15* 10 min Accelerating Rotarod 16* 15* 10 trials over 2 days T-maze 16* 15* successful alternation: enter the previously non-visited goal arm Barnes maze 16* 15* 2 trials per day for 6 days+ reversal or 1 weeks delay Fear conditioning 16* 15* Training, context-, tone- memory Acoustic startle 16* 15* 260 trials pseudo randomized, acoustic startle (65-115db), & pre-pulse inhibition pulse 120db pre-pulse (65-115db) Forced swim test 16* 15* 10 min and 6 min sessions

112 Characterization of Pclo SA/SA mouse model 4 113 =.05, =.05, power=.80 (t-test for two independent samples). α ower analysis P To determine the minimum effect sizes that we could measure, we performed power analysis of our experimental groups, given We calculated effect sizes interms ofdifference units of between standard the deviation two(Cohen’s d, means whichelectrophysiological divided is the by experiments the (n=46, n=47) standard Cohen’s deviation, d behavioralforandconsidered .2 isofexperiments d Cohen’s A (n=15, wasn=16)(n=16,n=18) 1.04. .99, Cohen, .59, for 1988). imaging For experiments a small effect, of .5 is a moderate effect and .8 is a large effect (Cohen, 1988). Supplemental

4500

4000

3500

3000

2500

2000

Piccolo intensity, AU 1500

1000

500

0 0 500 1000 1500 2000 2500 3000 3500 4000 4500 VAMP2 intensity, AU

Figure S10: Protein level intensity per synapse. Black – Pclo+/+, grey – PcloSA/SA.

A B 1.2 o i t )

a 1 s 14 r

m ( e

12 0.8 s l h t 10 u d

p 0.6 i

d w

8 f e

l 0.4 r 6 i Piccolo +/+ a

P 0.2 4 Piccolo SA/SA 2 0 EPSC ha 0 0 0.5 1 +/+ SA/SA Interstimulus interval (s)

Figure S11:. Electrophysiological analysis of Pclo+/+ and PcloSA/SA hippocampal autaptic neurons. (A) First evoked release kinetics is normal in PcloSA/SA neurons, as assessed by EPSC half width. (B) Paired pulse ratios at various interstimulus intervals, is not significantly affected in PcloSA/SA neurons. Pclo+/+ n=27, PcloSA/SA n=35.

PcloSA/SA general health and motor coordination are normal A general health check of PcloSA/SA mice and their Pclo+/+ littermates was performed at age 6 to 7 weeks. No abnormalities were observed. Pclo+/+ and PcloSA/SA mice had similar body weight, 25.1 ± 0.33g and 24.9 ± 0.46g, respectively (Fig. S12A). Motor coordination and muscle tonus are prerequisites to measure anxiety, cognition or despair in standalone behavioral tests and therefore, we tested motor coordination by training mice to walk on a Rotarod, a rotating cylinder of which the rotating speed is increased during trials. Both groups had a similar latency

114 Characterization of Pclo SA/SA mouse model 4 115 ○ SA/SA +/+

0.0 2.4 2.2 2.0 1.8 1.6

front and hind paws hind and front Strength (N) Strength mice showed no D SA/SA Pclo SA/SA +/+

1.5 1.3 1.1 0.0

front paws front Strength (N) Strength mice. (A) Body weight. (B) Rotarod distance. +/+ C Pclo mice (1.2 ± 0.02). Next, mice were allowed to 10 +/+ 8 SA/SA +/+ : 0.6±0.1m to on 0.9±0.1m) the first and the second Pclo mice vs Day 2 +/+ 6 SA/SA ) (Fig. S12C and S12D). Hence, +/+ Pclo Pclo Trials 4 Pclo Day 1 2

0.4 0.0 1.2 0.8 Distance [m] Distance B and 1.9 N± 0.04 in SA/SA Pclo+/+ MW empty beads PcloSA/SA MW MW empty PcloSA/SA beads Pclo+/+ Strength and coordination in Gel scans of A. Immunoprecipitation from wild-type or knock-in adult mouse brain synaptosomes SA/SA : 0.6±0.1m to 1.2±0.2m and +/+ (1.2 ± 0.03N) was similar to that of 13: 12: S S

0 SA/SA Pclo

26 24 22 28 SA/SA Weight (gr) Weight igure igure Pclo grab the grid with four paws. Again, no difference was observed in thein instrength (2.0 ± 0.05N F Grip strength for (C) front paws and (D) tall four paws together. representBox extremeplots individuals. show Line graphs showmedian, means and SEM.the 4 quartiles and to fall off the cylinder throughout the trials and thus their( distance travelled on the cylinder F day of the test (Fig. S12B). To evaluate the to musclemetal gridstonus linked into a theforce meter limbs,(grip strengthmice meter). wereFirst, presentedmice were allowedthe to grid grab with the front paws only and were pulled Pclo off by the tail. The strengthrecorded for impairment in motor coordination and muscle function. (Results in 3).Table A PCLO IP Rb44a beads P2 TX-P2

wt mut wt mut wt mut wt mut wt mut wt mut transfected HEK

IP PCLO IB flag

flag-Rasal

flag-Strap

Figure S14: No interaction was detected between Piccolo immunoprecipitated from synaptosomal fractions and STRAP1 overexpressed in HEK293 cells.

116 Characterization of Pclo SA/SA mouse model 4

117

0.59 0.54 28.56 48.25 199.02 63.53 237.15 light duration zone Spout

0.45 0.77 26.55 0.94 6.76 1.45 7.96 light number zone OnShelter

0.52 0.65 28.78 15.79 115.06 15.47 128.58 light duration zone OnShelter

S

0.55 0.61 26.92 149.96 1639.82 108.57 1752.85 light Feeding zone duration zone Feeding (activity) behavior pontaneous

0.92 0.1 28.84 4.02 48.23 4.24 48.8 light number Activity

0.33 1 15.69 0.99 16.55 5.54 21.11 light duration activity Mean

0.26 1.14 24.03 98.69 796.18 204.52 1019.99 light duration Activity

0.64 -0.47 23.77 200.97 1565.16 105.34 1465.05 dark duration zone Spout

0.94 0.07 28.41 16.81 128.83 14.04 130.29 dark number zone OnShelter

OnShelter zone duration zone OnShelter dark 0.44 0.77 28.9 168.99 2176.96 165.9 2354.34

Feeding zone duration zone Feeding dark 0.32 1.02 25.48 485.2 7201.41 337.57 7797.98

Activity number Activity dark 0.97 0.04 28.79 14.44 318.82 15.23 319.63

Mean activity duration activity Mean dark 0.46 -0.75 28.91 0.8 23.75 0.7 22.97

Activity duration Activity dark 0.62 -0.49 27.9 267.58 7643.46 315.64 7438.82

arameter P SEM ean M SEM ean M hase P -value: P -test: T f D

T W I . K

Parameters. Description of the spontaneous behavior of of behavior spontaneous the of Description Parameters. AHCODA 17: S able T mice Pclo

TM SA/SA 118

Table S17: Continued

WT K.I Parameter Phase Mean SEM Mean SEM Df T-test: P-value: Activity duration darklight index 0.86 0.04 0.9 0.01 16.66 -1.08 0.29 Mean activity duration darklight index 0.55 0.04 0.59 0.01 16.8 -1.09 0.29 Activity number darklight index 0.86 0.01 0.86 0.01 28.99 0.02 0.98 Mean short arrest duration darklight index 0.48 0 0.49 0 27.53 -0.6 0.55 Long arrest duration darklight index 0.82 0.01 0.79 0.03 18.75 0.83 0.42 ight ratio) L Mean long arrest duration darklight index 0.46 0.02 0.41 0.03 24.91 1.76 0.09 ark-

D Long arrest number darklight index 0.84 0.01 0.86 0.01 28.87 -1.16 0.26 Feeding zone duration darklight index 0.82 0.01 0.81 0.02 26.42 0.12 0.91 Mean short shelter visit duration darklight index 0.52 0.02 0.49 0.02 20.97 0.91 0.37 Long shelter visit duration darklight index 0.34 0.01 0.35 0.01 25.85 -0.71 0.49 Long shelter visit number darklight index 0.58 0.02 0.56 0.03 26 0.83 0.41 Mean long movement distance darklight index 0.52 0.01 0.52 0.01 26.68 -0.21 0.84 pontaneous behavior ( S Mean short movement distance darklight index 0.51 0 0.51 0 23.54 -0.63 0.53 OnShelter zone duration darklight index 0.94 0.01 0.94 0.01 28.86 -0.04 0.97 Spout zone duration darklight index 0.84 0.02 0.87 0.02 27.66 -1.08 0.29 Characterization of Pclo SA/SA mouse model 4

119

0.98 -0.02 22.47 0.34 1.36 0.25 1.35 light ratio habituation duration zone Spout

0.94 -0.07 26.22 0.12 0.93 0.1 0.92 light ratio habituation duration zone OnShelter

0.53 -0.63 26.22 0.02 1.02 0.02 1.01 light ratio habituation distance movement short Mean

0.78 0.28 28.15 0.02 1.03 0.03 1.04 light ratio habituation distance movement long Mean

0.29 1.09 25.87 0.03 0.96 0.02 1 light ratio habituation duration visit shelter Long

0.58 0.56 18.9 0.1 0.92 0.09 0.99 light ratio habituation duration visit shelter short Mean

0.18 -1.4 17.78 0.43 1.51 0.1 1 light ratio habituation duration zone Feeding

0.69 0.4 27.66 0.12 1.03 0.14 1.1 light ratio habituation number arrest Long

0.06 -2.02 18.96 0.25 1.46 0.08 0.98 light ratio habituation duration arrest long Mean

0.07 -1.89 17.88 0.39 1.66 0.09 1.03 light ratio habituation duration arrest Long

S

pontaneous behavior (habituation) behavior pontaneous

0.65 0.46 28.38 0.02 0.98 0.02 0.99 light ratio habituation duration arrest short Mean

0.46 -0.75 21.59 0.12 1.02 0.06 0.92 light ratio habituation number Activity

0.4 0.87 14.82 0.06 0.91 0.43 1.23 light ratio habituation duration activity Mean

0.76 0.31 18.45 0.15 1.01 0.41 1.13 light ratio habituation duration Activity

0.51 0.67 28.98 0.07 0.96 0.07 1.03 dark ratio habituation duration zone Spout

OnShelter zone duration zone OnShelter habituation ratio dark ratio habituation 0.11 1.64 27.95 0.11 1.27 0.1 1.51

Mean short movement distance movement short Mean habituation ratio dark ratio habituation 0.71 0.37 26.02 0.01 1 0.01 1.01

Mean long movement distance movement long Mean habituation ratio dark ratio habituation 0.47 -0.74 25.2 0.01 1.06 0.02 1.04

Long shelter visit duration visit shelter Long habituation ratio dark ratio habituation 0.46 -0.75 27.94 0.06 1.11 0.05 1.05

Mean short shelter visit duration visit shelter short Mean habituation ratio dark ratio habituation 0.46 0.75 25.77 0.04 0.95 0.06 1.01

Feeding zone duration zone Feeding habituation ratio dark ratio habituation 0.68 0.42 25.87 0.06 0.97 0.04 1

Long arrest number arrest Long habituation ratio dark ratio habituation 0.53 0.63 26.16 0.04 0.92 0.03 0.96

Mean long arrest duration arrest long Mean habituation ratio dark ratio habituation 0.7 0.39 26.2 0.05 1.06 0.03 1.08

Long arrest duration arrest Long habituation ratio dark ratio habituation 0.33 0.98 24.89 0.05 0.97 0.03 1.03

Mean short arrest duration arrest short Mean habituation ratio dark ratio habituation 1 0.01 1.03 0.02 25.68 -1.39 0.18

Activity number Activity habituation ratio dark ratio habituation 0.98 0.04 0.97 0.04 29 0.13 0.9

Mean activity duration activity Mean habituation ratio dark ratio habituation 0.91 0.03 0.88 0.03 28.71 0.83 0.41

Activity duration Activity habituation ratio dark ratio habituation 0.89 0.04 0.85 0.04 28.98 0.66 0.52 120 Table S17: Continued

WT K.I Parameter Phase Mean SEM Mean SEM Df T-test: P-value: Long movement fraction of total movement 0.42 0.01 0.42 0.01 26.41 -0.07 0.95 Long movement max. velocity 19.93 0.48 19.63 0.34 25.8 0.5 0.62 Long arrest threshold 5.1 0.18 5.08 0.2 28.71 0.11 0.92 Short arrest duration dark 3906.38 141.25 3882.16 104.59 26.21 0.14 0.89 Mean short arrest duration dark 0.98 0.02 0.99 0.03 27.05 -0.4 0.69 Short arrest number dark 3959.63 175.99 3895.98 161.14 28.68 0.27 0.79 Long arrest duration dark 10586.57 332.24 10807.83 633.42 23 -0.32 0.75 Mean long arrest duration dark 24.61 1.13 25.49 1.12 28.85 -0.57 0.58 Long arrest number dark 430.5 16.73 424.26 18.04 28.92 0.26 0.8 Long movement distance dark 25623.94 1637.8 27133.1 1827.92 28.99 -0.64 0.53 Mean long movement distance dark 14.63 0.32 15.6 0.46 27.28 -1.76 0.09 Long movement number dark 1751.81 100.29 1739.9 89.84 28.48 0.09 0.93 Short movement distance dark 3209.56 219.98 3283.19 229.7 28.99 -0.24 0.81 Mean short movement distance dark 1.28 0.04 1.34 0.05 28.66 -0.93 0.36 Short movement number dark 2516.27 116.5 2468.57 105.32 28.63 0.31 0.76 Short arrest duration light 542.06 48.08 482.58 48.97 28.75 0.9 0.37 Mean short arrest duration light 1.04 0.03 1.04 0.03 28.7 -0.05 0.96 Short arrest number light 521.37 45.51 463.13 47.94 28.54 0.92 0.37 pontaneous behavior (kinematics) S Long arrest duration light 2298.56 167.7 3424.8 995.32 15.85 -1.12 0.28 Mean long arrest duration light 28.45 1.82 38.29 5.97 20.19 -1.87 0.08 Long arrest number light 78.37 6.14 63.71 7.89 25.41 1.49 0.15 Long movement distance light 2432.47 317.5 2397.18 333.4 28.98 0.08 0.94 Mean long movement distance light 13.55 0.5 14.34 0.51 28.86 -1.12 0.27 Long movement number light 179.68 19.7 167.49 23.39 27.99 0.42 0.68 Short movement distance light 484.1 39.56 424.83 46.89 27.36 1 0.33 Mean short movement distance light 1.22 0.04 1.26 0.04 28.77 -0.66 0.52 Short movement number light 396.86 30.96 337.98 33.12 28.09 1.34 0.19 Long movement threshold 1.74 0.06 1.83 0.08 28.43 -0.9 0.37 Characterization of Pclo SA/SA mouse model 4

121

0.3 -1.05 28.95 0.32 5.14 0.28 4.71 light number visit shelter Long

0.36 0.93 22.25 1147.58 37047.25 587.05 38251.93 light duration visit shelter Long

0.57 0.57 27.4 2.56 15.48 3.66 17.79 light number visit shelter Short

S

pontaneous behavior (sheltering) behavior pontaneous

0.2 -1.32 20.43 0.9 7.85 0.55 6.55 light duration visit shelter short Mean

0.75 0.32 28.79 21.44 119.9 20.17 128.6 light duration visit shelter Short

0.26 1.16 25.59 0.16 9.91 0.1 10.12 threshold visit shelter Long

0.68 -0.42 26.67 0 0.06 0.01 0.06 visits total of fraction visit shelter Long

0.46 -0.76 27.71 0.14 4.55 0.11 4.42 threshold visit shelter Short

Mean long shelter visit duration visit shelter long Mean 0.49 0.7 28 212.78 5231.28 174.95 5420.1

Long shelter visit number visit shelter Long 0.78 0.28 28.65 0.51 6.5 0.45 6.68 dark

Long shelter visit duration visit shelter Long 0.82 -0.23 27.23 819.54 19808.93 608.23 19573.29 dark

Short shelter visit number visit shelter Short 0.56 0.59 28.99 9.73 103.98 10.26 111.96 dark

Mean short shelter visit duration visit shelter short Mean 0.68 -0.42 28.86 0.64 7.28 0.55 6.94 dark

Short shelter visit duration visit shelter Short 0.72 0.36 25.12 60.46 786.45 87.96 825.14 dark 122

Table S17: Continued

WT K.I Parameter Phase Mean SEM Mean SEM Df T-test: P-value: Activity change in anticipation of dark 0.01 0.01 0 0.01 19.3 0.22 0.83 Activity change in anticipation of light 0.11 0.02 0.06 0.02 25.09 1.87 0.07 Activity change in response to to dark 0.27 0.02 0.28 0.02 27.9 -0.31 0.76 Activity change in response to to light -0.09 0.01 -0.11 0.02 24.6 0.95 0.35 Feeding zone change in anticipation dark 0.08 0.09 0.14 0.07 23.8 -0.49 0.63 Feeding zone change in response to dark -0.16 0.04 -0.05 0.04 28.86 -2.06 0.05 Feeding zone change in anticipation light -0.2 0.06 -0.06 0.07 26.77 -1.54 0.13 Feeding zone change in response to light -0.02 0.09 -0.17 0.11 16 1.07 0.3 OnShelter zone change in anticipation dark 0.01 0.01 0.01 0.01 26.65 -0.64 0.52 OnShelter zone change in response to dark 0.01 0.01 0.01 0.01 29 -0.22 0.83 OnShelter zone change in anticipation light 0.11 0.01 0.12 0.02 25.68 -0.22 0.83

pontaneous behavior (pattern) OnShelter zone change in response to light -0.08 0.02 -0.02 0.02 15.48 -1.95 0.07 S Spout zone change in anticipation dark -0.01 0.02 -0.03 0.03 26.11 0.52 0.61 Spout zone change in response to dark 0 0.01 -0.03 0.01 21.07 2.15 0.04 Spout zone change in anticipation light 0.03 0.02 0 0.02 24.38 1.01 0.32 Spout zone change in response to light -0.02 0.02 0 0.05 12.02 -0.39 0.7

Chapter

Random mutagenesis by transposon-based gene trap insertion in mice identifies a role of Ubn1 in avoidance learning 5 and fear conditioning Gregoire Maroteaux1, Bastijn Koopmans4, Joost Hoetjes1, Sophie van der Sluis2, Maarten Loos4, Joke Wortel2, Junji Takeda5, Oliver Stiedl1,3, Guus Smit3, Matthijs Verhage1,2

1Department of Functional Genomics, 2Department of Clinical Genetics, 3Department of Molecular and Cellular Neuroscience; Center for Neurogenomics and Cognitive Research (CNCR), Neuroscience Campus Amsterdam; VU University Amsterdam and VU Medical Center, 1081 HV, Amsterdam, The Netherlands; 4Sylics (Synaptologics BV), P.O. Box 71033, 1008 BA, Amsterdam, The Netherlands; 5Department of Social and Environmental Medicine, Graduate School of Medicine, Osaka University, Suita, Osaka, Japan

126 Screening for genes involved in avoidance learning 5 127 showed a showed -/- ), fibroblast growth factor

Dpp10 mice showed a normal preference -/- ) littermates. Furthermore, we demonstrated demonstrated we Furthermore, littermates. ) +/+ . ) ) of each behavior spontaneous strain altered displayed -/- ). Those 5 strains were behaviorally phenotyped using an in associative memory. Ubn1 Ubn1 a voltage-gated potassium (Kv) channels, Tetratricopeptide repeat domain Human and mice share ~99% of protein-coding genes which make mice ideal This study shows on the one hand the robustnessotherthehandthecircadianone On the behavior. studyshowsThis on ) and Ubinuclein 1 ( Using mouse germlines carrying the Sleeping beauty transposase together with a Kcnd2 We We showed that the mutants ( ), Ttc39c stract Fgf13 ethods: onclusions: esults: esults: less anxious behavior in the elevated plus maze and a weaker fear response in the fear conditioning. fear the in response fear weaker a and maze plus elevated the in behavior anxious less C hand regardless of the gene in which the transposon landed, we could in observeactivity modificationbouts in the different strain of mice. Finally, out ofobserved 5 the involvementrandomly of mutated genes, we that that Ubn1 was involved in avoidance learning as Ubn1 development, yet a deficit in the response to the aversive stimulus. Additionally, Ubn1 Additionally, stimulus. toyet a thein aversive response thedeficit development, M 13 ( transposon, we generated randomly mutated mouse strains, carrying a transposable element in 5 functionally unstudied genes: Dipeptidyl-peptidase 10 ( modelsinvestigateto genefunctions. Moreover,thediversity mutationof techniques, allowing spatial and temporal modification of gene expression in mice, has provided priceless answers on basic biological processes involved However, inmost of genethe studied function,genes were drugdiscovered after and clinicalphenotype alterations, diseasesobservations leaving a large number of ingenes unstudied. mechanisms.dramatic Ab Background: especially activity bouts compared to their control ( control their to compared bouts activity especially 39c ( automated high-throughput home cage-based system to screen spontaneous and avoidance learning behavior of mice in seven covering a large panel days of characteristic behaviors followed by an extensive behavioral test batteryR Introduction With the complete sequencing of the human genome in 2006 (Gregory et al., 2006; Lander et al., 2001) now understanding the function of the proteins encoded by these genes becomes imperative. One approach to do so is to study a model in which a particular gene is missing. The fast expansion of molecular genetics in mice, and the fact that ~99% of the human protein- coding genes have their homologues in the mouse genome, makes it an ideal model to investigate genes of interest and study their function at different levels in this organism (Mouse Genome Sequencing Consortium et al., 2002). The diversity of mutation techniques allows precise modifications of any gene in both spatial and temporal manners. These techniques proved to be powerful tools to provide answers to basic biological questions regarding disease mechanisms, drugs and gene function (Ledford, 2012). Until now, however, most of the studied genes were discovered after the observation of a clinical/disease phenotype or some other strong biological implication, which leaves a large number of genes unstudied. However, Izsvak and his team engineered the Sleeping Beauty (SB) transposase, member Tc1/mariner transposable elements found in fish, to have an activity in cultured mammalian cells (Ivics et al., 1997). In 2001, mouse germlines were generated carrying the SB transposase and a transposon (Dupuy et al., 2001; Fischer et al., 2001; Horie et al., 2001) which enabled to generate relatively large numbers of random transgenic mice. Using this technique, we generated many transgenic mouse strains, including five that carry a mutation in genes that have to date only been characterized to a very limited extent. Here we describe those five mutations in such genes. The first one encodes Dipeptidyl-peptidase 10 (Dpp10), which is a member of the S9B family but without detectable protease activity. However Dpp10 is involved in synaptic transmission by forming a complex with voltage- gated Kv4.2 channels and modulates their proprieties (Cotella et al., 2012) and thus is part of synaptogenesis. Mutations in this gene are associated with asthma and obesity (Melén et al., 2010). Dpp10 also emerged as a positional and functional candidate in autism (Marshall et al., 2008). The second gene encodes the fibroblast growth factor 13 (Fgf13) and belongs to the Fgf family, which comprises 22 members in humans and mice. Fgfs are extracellular polypeptides involved in multiple developmental and metabolic processes (e.g. embryonic development, cell growth, morphogenesis tissue repair and tumor growth). However, Fgf13 belongs to the subfamily of iFgf (Fgf11-14), which are intracellular proteins that act in a tyrosine kinase FGF receptor-independent manner (Goldfarb, 2005). Fgf13 is located in region associated with the x-linked Borjeson-Forssman-Lehmann syndrome (BFLS) characterized by mental retardation, obesity and seizure (Itoh and Ornitz, 2008). The third mutation hit Kcnd2, a voltage-gated potassium (Kv) channels. Voltage-gated ion channels are known to regulate neurotransmitter release, heart rate and other important physiological processes. Kcnd2 was shown to be a major contributor of the cardiac repolarization phase and Brugada syndrome (Niwa and Nerbonne, 2010; Postma et al., 2000). Truncation of Kv4.2 channels encoded by Kcnd2 produce aberrant neuronal excitability characteristic of temporal lobe epilepsy (Singh et al., 2006). The fourth gene is Tetratricopeptide repeat domain 39c (Ttc39c), with currently unknown function. However, some studies suggest that it plays a role in the anaphase (Blatch and Lässle, 1999).

128 Screening for genes involved in avoidance learning 5 129 -SB X association, chromatin organization Ubn1 cassette to profile the expression pattern of the -SB had been back-crossed to C57BL/6J mice (Harlan) LacZ X ), a subunitaHIRA/UBN1/CABIN1/ASF1athe ),of (HUCA) histone is essential for HIRA- Ubn1 Ubn1 Behavioral phenotyping is limited by classical standalone tests which are time-consuming aterial and methods aboratory animals leeping Beauty transposon mice and show limited reproducibility. Therefore, we developed a protocol in an automatedthroughput homehigh- cage-based system to screen spontaneous and avoidance learning behavior of mice in seven days (Loos et al., 2014; insertionalMaroteaux et random al., mutagenesis2012) In and this highstudy, we content combined in phenotyping avoidance to learning identify before genesproceeding involvedwith a investigationspectrum identified aroleof of Ubn1 in avoidanceclassical learning and fear behavioralconditioning in mice. tests. This and thus proliferation arrest in primary human cells et(Tang al., 2012). M L Mice were all males between 7-9 weeks old in a C57BL/6J background.and bredThey in werethe VU generateduniversity animal facility. A week prior to the experiment they werehoused single in classical Plexiglas cage in our light-darkfacilities, cycle. All with experiments food were approvedand by waterthe local ad animal researchlibitum compliedcommittee with inthe Europeanand Councila Directive (86/609/EEC). 12 h S Transposon seeder mice wereconcatamer and mice SY1-4 SY1-2 (Synaptologics Amsterdam, BV, The Netherlands) obtained with C by crossing gene-trap containing transposon The fifth gene is fifthTheUbinuclein gene is ( 1 chaperonecomplex. Ubn1 is a nuclear protein that interacts with cellular and viral transcription factor (Aho et al., 2000). mice expressing the transposase (Horie et al., 2001) that acts to mobilize the transposonfacilitate and its random reintegration into the genome of the seeder mouse offspring.and The SY1-2 SY1-4 concatamer mice were generated by injecting plasmida gene-trap into containing the SY1 vector pronuclei of B6C3F2 lines fertilized were oocytes generated, which of were C57/BL6J mice. evaluated Lines Five for SY1-2 and SY1-4 plasmid transgenic showed the integrationhighest signal by and were Southern selected for blotting. furtherSY1 vector breeding.is a modificationThe of the IF3 plasmid (Keng et al., 2005). In brief, The vector ofconsists a gene-trap cassette generated by a slice acceptor fused an to stopinternal codons ribosome in entrythree frames,site and a endogenous gene; the gene-trap cassette is followed expressionby cassettea with greenits own fluorescentpromoter (chicken proteinbeta-actin), (GFP) flanked by a splice acceptor, allowingpoly(A)-trapfor selection transpositionof eventsexpressionGFP by (Kengal.,et2005) and generating GFP-expression throughout the body if the gene activetrap (polyA-containing)cassette gene.lands The C into an for >12 generations before crossing to the SY1-2 or SY1-4 mice. Seeder mice C57BL/6Jwere micemated withto generate mutant offspring. Integration of the transposonlinker-mediated was PCR withdetected 2 differenttechniques by depending of the construct: Linker-mediated PCR for GT3A construct: Applied to detect transposon insertion in Fgf13, Kcnd2 and Ubn1. For genotyping: 2 μg of tail DNA was digested with NlaIII overnight at 37°C, cleaned up using Qiagen’s Qiaquick PCR purification kit and eluted in 50 µl water. Linker-oligonucleotide mix was prepared by mixing 50 µl linker+ (25 µM), 50 µl linker- (25 µM), and 1 µl 5 M NaCl, heated in a boiling bath for 5 min and slowly cooled to room temperature. Sequence of the linker oligonucleotides was:

linker+ 5’-TAATACGACTCACTATAGGGCTCCGCTTAAGGGACCATG-’3

linker- 5’Phos-GTCCCTTAAGCGGAG-3’NH2 Digested genomic DNA (400 ng) and 6 µl linker-oligonucleotide mix were used for overnight ligation with T4 ligation kit (MRC Holland, Amsterdam, Netherlands). The template was amplified in 50 μl primary PCR reaction supplemented with primers New long IR/DR(R) (100

mM), linker primer (100 mM), 200 μM dNTPs, 2 mM MgCl2, and 1 unit of platinum Taq polymerase (Invitrogen). The PCR machine was programmed for touchdown PCR at 94°C for 2 min, 25 cycles of 94°C for 1 s, 60°C for 30 s (-0.5°C per cycle), 72°C for 90 s. The primary reaction was diluted 1:50 and 2 μl used in a nested PCR under the same exact conditions, except supplemented with 0.25 μM of each primer IR/DR(R) KJC1 (0.25 µM) and linker nested primer (0.25 µM). Sequences of the oligonucleotides were:

New long IR/DR(R) 5’-GTTATGCTAGATGGCCAGATCTAGCTTGTGG AAGG-3’ linker primer 5’-GTAATACGACTCACTATAGGGC-3’ IR/DR(R) KJC1 5’-CCACTGGGAATGTGATGAAAGAAATAAAAGC-3’ linker nested primer 5’-AGGGCTCCGCTTAAGG GAC-3’

Secondary PCR products were separated by electrophoresis and cleaned by Invisorb spin DNA extraction kit (Invitek) and eluted in 25 µl water. 3 µl was ligated in pGEM-T easy cloning kit (Promega), transformed into DH5α E-coli bacteria, plated on LB-AMP-X-gal plates and incubated overnight at 37°C. Colonies were inoculated and cells were grown overnight at 37°C. Plasmid were isolated and sequenced with BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems) and T7 and SP6 oligonucleotides. Sequence of the oligonucleotides was:

T7 primer 5’-GTAATACGA CTCACTATAGGGC-3’ SP6 primer 5’-ATTTAGGTGACACTATAG-3’ SY1-2, SY1-4 LM PCR

Linker-mediated PCR for TakZ construct: Applied to detect transposon insertion in Dpp10 and Ttc39c. Applied to detect transposon insertion in 50 ng of tail DNA was digested with AluI, HpaII, RsaI, or TaqI (or BfaI, HaeIII, MboI if not successful) for 3 h. After heat inactivation 5 ng were used for ligation with linker for 2 h at 16°C. Linker-oligonucleotid mix was prepared by mixing, 2,5 µl of Spl-Top (1 mM), 2.5 µl Spl-Sau (1 mM) or 2.5 µl Spl-Blt (1 mM) or 2.5 µl Spl-Taq1/HpaI (1 mM) and 100 µl 10 mM Tris-HCl pH 7.5 5 mM MgCl solution, and soaked in boiling water bath following by gradual cooling to room temperature.

130 Screening for genes involved in avoidance learning 5 131 5’-CGAATCGTAACCGTTCGTACGAGAATTCGTACGAGAATCG CTGT CCTCTC CAACGAGCCAAGG-3’ 5’-GATC CCTTGGCTCGTTTTTTTTTGCAAAAA 5’-CG CCTTGGCTCGTTTTTTTTTGCAAAAA 5’-CCTTGGCTCGTTTTTTTTTGCAAAAA 5’-CGAATCGTAACCGTTCGTACGAGAA-3’ 5’-TCGTACGAGAATCGCTGTCCTCTCC-3’ 5’-AGTGTATGTAAACTTCTGACCCACTGG-3’ 5’-CTTGTGTCATGCACAAAGTAGATGTCC-3’ 5’- CTGGGTAGCAGAATGTACGTCC-3’ 5’- GCCTAGTAATGCCCTGATCTTC-3 5’-CCACTGGGAATGTGATGAAAGAAATAAAAGC-3’ 5’- GTAAGCCAACTTCAACAATCTGAGG -3’ 5’- -3’ CATATCAATAATGCTTCCTGACAGC 5’CTGGAATTTTCCAAGCTGTTTAAAGGCA CAGTCAA – 3’

New mutants, where the gene trap had landed in an active gene were bred to lose the

Spl-top Sau Taq1/HpaII Blnt Theligated genomic DNAwaspurified, Qiagenby PCR purification kit, eluted waterµl 40 in and overnight digested with KpnI (TakR) or BamHI (TakZ) at 37°C. After purification of the digested DNA in 50 µl water, 1 µl was used to amplified thetemplate reaction insupplement a with 12.5 primers µl T/DR primary(10 µM),reaction Spl-P1 PCR (10 µM), 10 mM dNTPs,Hotstar and Taq 0.3 (Qiagen). Units The PCR machine was programmed for touchdown min, 30PCR cycles of at 94°C for 1 min,95°C 55°C for for1 min, 72°C15 for 1 min. 1 µl of the primary reaction were used in a secondary PCR under the same conditions, except supplemented with T/bal (10 µM) and Spl-P2 (10 µM) primers. Spl-P1 Spl-P2 T/DR T/bal transposase and the transposon donor site,oligonucleotides specific interrupted the for locus (see below) which IR/DR(R)and KJC1, DMSO (10%), was confirmed by genomicBSA (1x), PCRdNTP’s (400 µM) usingand 1 unit Taq DNA polymerase (New England Biolabs) resultingrespectively in a wild-type amplified DNA product of 522 bp and a mutant product of 632 bp. PCR cycle: 95°C for 5 min, 35 cycles of 95°C for 3 s, 55°C for 30 min, 65°C for 1 min and 65°C for 5 min. The sequences of the oligonucleotides specific for the interrupted loci were: Specc1 F Secondary PCR products were separated by electrophoresis and cleaned by Invisorb spin DNA extraction kit (Invitek) and eluted in 25 µl water. One microliter was used for sequencing with BigDye Terminator Cycle v3.1 Sequencing Kit (Applied Biosystems) and oligonucleotide.T/bal Specc1 R IR/DR(R) KJC1 FGF (GT3A vector) was confirmed by genomic PCR using oligonucleotides FGF13-F and FGF13-R or FGF13-F and Long IR/DR(L2), DMSO (10%), BSA (1x), dNTP’s polymerase (400 (New EnglandµM) Biolabs) resultingand in respectively1 a wild-typeunit amplified Taq DNA productDNA of 369 bp and a mutant product of 298 bp. PCR cycle: (WT) 95°C for 5’ min, 35 cycles of 95°C for 3 s, 52°C for 30 s, 65°C for 1 min and 65°C for 5 min or (KO) 95°C for 5’ min, 35 cycles of 95°C for3 s, 55°C for 30 s, 65°C for 1 min and 65°C for 5s. The sequences of the oligonucleotides were: FGF13-F FGF13-R Long IR/DR(L2) Using for Dpp10 (TakZ); Dpp10-F, Dpp10-R, T/bal, for Kcnd2 (GT3A); Kcnd2-F, Kcnd2-R, T/bal, for TTc39 (TakZ); Ttc39c-F, Ttc39c-R, T/bal and for UNB1 (GT3A); Ubn1-F, Ubn1-R, T/bal oligonucleotides with dNTP’s (400 µl) and 0.1 Unit Phire Hotstart DNA polymerase (Thermo Scientific) resulting in respectively a wild-type amplified DNA product of 650bp, 680bp, 610bp or 585bp and a mutant product of respectively 500bp, 425 bp 420bp or 406bp. PCR cycle: 98°C for 1 min, 33 cycles of 98°C for 5 s, 60°C for 1 s, 72°C for 15 s and 72°C for 1 min. The sequences of the oligonucleotides were: Dpp10-F 5’-CCATCCATCCACTTCTTAACACATCC-3’ Dpp10-R 5’-GGATGCAATACAGATGAAACCAGC-3’ T/bal 5’- CTTGTGTCATGCACAAAGTAGATGTCC-3’ Kcdn2 F 5’-GTTCCTCTTCTCCATTTATACAGTCCAGG-3’ Kcdn2 R 5’- CAATGTAGAGTTCTTTGTCAATTCTGTCCC-3’ T/bal 5’-CTTGTGTCATGCACAAAGTAGATGTCC-3’ Ttc39c F 5’-GTTCCCATTTGAAGCTGTGTGAGC-3’ Ttc39c R 5’-AATACCTGAGAGGAACAGATTAAAGG-3’ T/ba 5’-CTTGTGTCATGCACAAAGTAGATGTCC-3’ Ubn1-F 5’-AAGGTAGGCATTTTCTAAATAACAATAGC-3’ Ubn1-R 5’-GGTACTGGGAAAACTCAAGTTATAGCC-3’ T/bal 5’-CTTGTGTCATGCACAAAGTAGATGTCC-3’

Behavior test battery Mice were transferred to specially designed home cages (PhenoTyper model 3000, Noldus Information Technology, Wageningen, The Netherlands) in the second half of the subjective light phase (14:00 h – 17:00 h). The behavior of mice was video-tracked for seven days as described in detail previously ((Maroteaux et al., 2012) EthoVision HTP 2.1.2.0, based on EthoVision XT 4.1, Noldus Information Technology, Wageningen, The Netherlands), starting at the first subjective dark phase (19:00 h). Resulting track files, containing X-Y coordinates of the center of gravity (COG) at a resolution of 15 coordinates per s, were processed using AHCODATM analysis software (Synaptologics BV, Amsterdam, The Netherlands) to generate behavioral parameters. 115 activity parameters were generated form the first three days as described in detail previously (Loos et al., 2014). The last and first 10 min of each dark and light phase were not included in summary statistics, to ensure that a potential asynchrony of the data streams and light regime in the behavior facility would not affect these results. After the end of the PhenoTyper protocol, mice were housed for a week in the animal facility in individual transparent Perspex cages with 12 h dark-light cycle and food and water ad-libitum. Following this week, after analysis of the PhenoTyper data, Ubn1 mice (controls and mutant) were introduced in the test battery with administration order going from the least (grip strength meter) to the most stressful test (fear conditioning). The sequence is described in Table 6. A full description of the Methods can be found in chapter 7: Material & Methods

Statistical analysis Before any analysis was performed, data were examined for outliers (>3 times the SD from the strain mean). All statistical analyses were performed using IBM SPSS statistic 20 (IBM,

132 Screening for genes involved in avoidance learning 5 133 age: weeks 7-8 mice (n=14). 10 trials over 2 days 5 session front paws –/– 10 min, starting in a corner omments/duration/sequence 2 180 s trials per day for 6 days C 6 sessions, sample and test on 2days 7 days protocol no human interference 3 days habituation prior test, 10 min max 10 min, starting from the dark comportement 5 min, starting in the center facing a closed arm Training, context- and tone-dependent memory (n=12) and ubn1 +/+ ays 13 13 15 17 21 16 14 0*-7 D 22-27 20-21 28-29 Behavior tests, temporal sequence of ubn1 esults est able 6: ariability of spontaneous behavior in gene-trap mutant mice Dark-light box Accelerating rotarod T Elevated plus maze Grip strength Novelty-induced hypophagia Open field T-maze Automated home cage test Body weight Barnes maze Fear conditioning * Day 0 corresponds to a mean mouse age of 49 - 63 days. T R Using a Sleeping Beauty gene trap we generated a large collection of novel mutanttested mice theseand in an automated high throughput behavioral spontaneousscreen behaviorfor phenotypescognitive, (Table 7; Fig. anxiety18). Here, and we describe 5 of these mutants,showedenvironment.spontaneousthemcagealteredhometheirof Four behavior (habituation in and baseline) and one a cognitive deficit. V We compared the automated an homozygousused we mutants, gene-trap phenotypesin behavioralidentify To (+/+).littermates transposon gene-trap mutants (–/–) home cage to system to perform their tests on a wild-type long time scale (days) withoutDuring human7 daysinterference. the behavior of the animal is tracked in habituation (day 1 and3) and2), challenged baseline situations (day (day 4 to 7). The spontaneous behavior including habituation and baseline activity, was segmented in general behavior (raw data) and 6 distinct activity classes Armonk, NY, USA). Genotype differencesANOVA, were repeated-measures compared using ANOVA) parametric whenever were tests met. normality (T-test, Otherwise, and nonparametric tests homoscedasticity were U-test). performedNonparametric criteria data are (Kruskal-Wallis,presented as box plots Mann-Whitney (ends of the box denoting the 25 75%and interquartile range and the whiskers providing the the upperinterquartile andrange, respectively,lower while quartilethe line ±1.5 in the times box denotes the probabilitymedian). level of An p<0.05 was error accepted as statistically significant throughout the study of the classical behavioral tests. For all given comparisons of PhenoTyper results, wasstatistical based analysison estimated false discovery rate (FDR) (Verhoeven et al., correct by 2005)minimum positivewith FDR with aalpha-levels threshold set at 5%. Table 7: Genes and transposons insertion

Gene name Dpp10 Fgf13 Kcnd2 Ttc39c Ubn1 location 123,332,142 59,062,145 21,215,503 12,599,926 5,050,068 size (bp) 713,417 505,926 514,302 137,124 36,217 # exons 27 10 6 15 19 # isoforms 2 6 1 2 1

Chromosome 1 X 6 18 16 Orientation reverse reverse forward forward forward

Transposon type TakZ largaespada largaespada TakZ TakZ size (bp) 10,067 3,220 3,220 10,067 10,067 landing site 123,830,742 59,276,994 21,684,413 12,687,132 5,054,718 exon preceding 4 3 1 5 2

Disrupted isoforms 2 4 1 2 1

holding together 22 key parameters (Loos et al., 2014) describing a mouse-centered analysis (for review (Benjamini et al., 2010)). After habituation, Dpp10-/- mice showed an increased baseline activity, and repressed activity when exposed to a novel environment suggested by the smaller distance moved in the first dark phase and a larger distance moved in the third dark phase compared to their controls. Fgf13-/- showed a lower dark/light index, suggesting a more active behavior during the light phase with in general longer short visits to the shelter. Kcnd2-/- show no significant difference with their control littermates. Overall, Ttc39c-/- were more active during the light phase with higher maximum velocity in long movements and shorter long visits in the shelter. However WT Ttc39c displayed unusual performance when compared to a C57BL/6J reference group (data not shown), which make their results hard to interpret. Ubn1-/- mice showed a clear hyperactivity behavior during the dark phase in baseline condition (Table 8). Taken together these results indicate that spontaneous behaviors, especially activity, are highly sensitive to gene disruption.

Avoidance learning is absent in Ubn1–/– To identify cognitive deficits, we used the avoidance learning paradigm of the automated home cage involving the shelter with 2 entrances (described in (Maroteaux et al., 2012)). Until day 4, the mouse develops a preference for one of the entrance which will be automatically be sanctioned by a bright light each time it is used on day 5 and 6. The preference index reflects the cognitive aspect (discriminating the sanctioned entrance from the other one) and behavioral flexibility (actively changing the preference) involved in response to the challenge. All mutants developed a preference for one of the entrances during the first 4 days without bias for left or right, similar to their control littermates (Fig.19). Ttc39c–/– mice started with a lower preference index than Ttc39c+/+ (p=0.018; see Fig. 19D and Table S19). However, they showed a similar reaction to the aversive stimulus. During the shelter task on day 5 and 6, all the KO

134 Screening for genes involved in avoidance learning 5 135 123 123 iso: 1 33473 Normal iso: 2 WT ISO 6: ISO WT 456 19 13 14 15 iso: 1/2/3/4 Mutant Mutant No protein No protein 6 Normal 119306 WT ISO 5: ISO WT 789 45 101112 WT ISO 2: 112603 ISO WT WT ISO 2: 169401 WT Mutant Normal iso: 5 Mutant 16 17 18 Normal No protein Mutant No protein 7 8 9 10 11 12 , reversed strand; the transposon landed after 131415 6 X 6 iso: 6 WT ISO 1: WT WT ISO81542 WT 13 1415 131319 WT ISO 4: ISO WT Normal 161718 Mutant Mutant 5 No protein Normal No protein 789 119833 192021 Normal WT ISO 3: ISO WT 2 3 4 5 WT ISO 1: 25294 WT WT ISO 1: 112606 ISO WT mRNA Protein Normal Normal 7 8 9 1011 12 222324 mRNA Protein Translation Translation Transcription Transcription 145767 WT ISO 2: ISO WT 2526 4 5 6 124402 1 3 WT ISO 1: ISO WT 10 mRNA mRNA Protein Protein Translation Translation iso 1 iso 2 Transcription Transcription 27 mRNA Protein 1 2 Gene-trap constructions and their different splice variances. (A) Dipeptidyl-peptidase 10 (Dpp10) is Translation 1 2 3 4 Transcription A. Dpp10, Chromosome 1 A. Dpp10, Chromosome E. UBN1 Chromosome 16 C. Kcnd2, Chromosome 6 B. Fgf13, Chromosome X D. Ttc39: Chromosome 18 D. igure 18: exonaffecting3, out splice4 of 6 variants.(C) Potassium voltage-gated channel (Kcnd2) located on chromosome 6. The transposon landed after exon 1, affecting39B the(Ttc39) onlylocated spliceon chromosomevariant. 18. (D)The Tetratricopeptide transposon landedrepeat after Ubinuclein domain exon 1 5 (Ubn1) affectinglocated bothon chromosome splice 16. variants.The transposon (E) landed before sequenceexon the ATG), expression3 could(containing be disrupted. theThe blue rectanglestarting represents the transposon F located on chromosome 1, reversed strand; the transposon landed after exon (B)4, Fibroblastaffecting growth factor 13 (Fgf13)both is located splice on chromosome variants. Table 8: Spontaneous behavior measured in the automated home cage.

Classes Parameters Dpp10 Fgf13 Kcnd2 Ttc39c Ubn1 General behavior activity during the first dark phase (12h) - activity during the third dark phase (12h) + +

1: kinematics: Long movement max velocity + description of movement Long arrest threshold + characteristics Mean long arrest duration - light - Long movement threshold -

2: sheltering: Long shelter visit duration - dark - Use of the shelter Short shelter visit threshold + Long shelter visit fraction of total visits Long shelter visit threshold

3: habituation: Activity duration - habituation ratio dark + ratio of day 3 over day 1 Activity duration - habituation ratio light

4: DarkLight index:dark value/ Activity duration - darklight index - - (dark value + light value

5: activity pattern: Activity change in anticipation of dark complex circadian pattern Activity change in anticipation of light Activity change in response to to dark Activity change in response to to light

6: activity: Activity duration - dark + + characteristics and quantity Mean activity duration - dark of individual activity bouts OnShelter zone number - dark + Activity duration - light + + Mean activity duration - light +

(+) implies a significant (<0.05) positive difference for -/- compared to +/+ (-) implies a significant (<0.05) negative difference for -/- compared to +/+ ( ) implies no significant difference between --/ and +/+

mice of each strain displayed a decrease in the preference index and a change of preference (preference index below 0). The KO mice for Dpp10, Fgf13, Kcnd2 and Ttc39 also showed a decrease in their preference followed by a change in their preference similar to their respective WT mice (Table S20). However, Ubn1–/– mice did not react to the stimulus and kept the same preferred entrance throughout 48 h of training (significant interaction between genotypes and days: p<0.0001; see Table S20), indicating a deficit in avoidance learning. After 48 h the challenge stopped and the former preferred entrance was no longer sanctioned. A majority of WT mice (>60%) kept the newly acquired preference index, which shows that the change is stable. Dpp10–/–, Fgf13–/– and Kcnd2–/– mice also exhibited a stable change in the preference. In contrast, Ttc39–/– did not have a stable change in the preference index. After 12 h without

136 Screening for genes involved in avoidance learning 5 137

–/– Kcnd2 appears -/- +/+ -/- +/+ Ubn1 mice after the -/- +/+ 15 27 14 14 12 failed to avoid the mice and(C) Ttc39 shelter task shelter task –/– 9 Ubn1 21 12 20 10 +/+ n= KO mice have normalhavemiceavoidance KO mice,(B)Fgf13

–/– Ttc39 1 2 3 4 5 6 7 Ubn1 1 2 3 4 5 6 7 Dpp10 Fgf13 Kcnd2 Ttc39c Ubn1 Strain Kcnd2

Dpp10 0 0 0.6 0.4 0.2 0.6 0.4 0.2 -0.2 -0.4 -0.6 -0.2 -0.4 -0.6 E F D Fgf13and mice learned the task but as soon as the aversivethe as soon as butlearnedtask micethe , –/– +/+ -/- -/- +/+ -/- +/+ Dpp10 Ttc39 mice showed no difference in the development or the change of –/– mice did not significantly during thenot preference did their micechange shelter task shelter task shelter task –/– Ttc39 mice showed a different development of the preference and a –/– Ubn1 mice showed a clear change of preference on day 6 which was stable on day 7. were using both entrances with a similar frequency (Preference index +/+ –/– Ttc39c Ubn1 mice showed a reduced avoidance of the sanctioned entrance and did not change their ) controls. (D) Ttc39 –/– Dpp10 1 2 3 4 5 6 7 1 2 3 4 5 6 7 Kcnd2 1 2 3 4 5 6 7 Fgf13 +/+ Ubn1 0 0 0 0.6 0.4 0.2 0.6 0.4 0.2 0.6 0.4 0.2 -0.2 -0.4 -0.6 -0.2 -0.4 -0.6 -0.2 -0.4 -0.6 Preferenceindex fortransposon5 mouse mutants. (A) mice developed a clear preference as habit-like response that was transiently changed A B C –/– igure19: sanctioning period, termination of sanctioning did not affect their behavior either (see Table S21). Table (see either behavior their affect not didsanctioning of terminationperiod,sanctioning theseresultsshowthat together, Taken preference, but restored their preference for the punished entrance compared to from -0.25 on day 6 to -0.003 on day 7). day on -0.003 to 6 day-0.25 on from shelter task. (E) transient change in preference depending only on the negative reinforcement. sanctionedentrance(Fig.subsequent 19A-F).In experiments, batteryweuseda classical of tests to to be involved in avoidance learning as mice carrying the transposon in miceshowed nodifference in the development, the change and the stability of the changed preference compare to their wild-type ( F aversive stimulus, upon negative reinforcement and immediately returned to the original pattern as soon as the stopped.reinforcementAs stimulus was stopped they went back to their first unconditioned preference, suggesting that Ttc39 preference (>0) whereas learning. In contrast, to identify whether Ubn1–/– showed a learning deficit or were not sufficiently aroused by the aversive stimulus to learn to avoid the sanctioned entrance.

Ubn1–/– deletion has anxiolytic properties on the elevated plus maze As home cage monitoring suggested hyperactivity of Ubn1–/– mice (Fig. 20A, B) without affecting basic shelter behavior (Fig. 20C-E), we investigated locomotor activity in classical assays such as EPM. Ubn1–/– mice covered larger distances on the open arms (p=0.023) but not the closed arms (p=0.222) and in total spent significantly more time on the open arms (p=0.020). However, Ubn1–/– showed no significant difference, compare to Ubn1+/+, in other anxiety test (e.g.: open field, the dark-light box and novelty-induced hypophagia, ). As a result, Unb1–/– mice were less affected by the anxiogenic zone of the elevated plus maze (open arms), although they responded similarly to their control littermates in all other anxiety-related tests (Table 9). Ubn1-/- mice were less anxious than their WT controls.

A

60 -/- +/+ 50

40

30

20 Distance moved (m) 10

0 0 12 24 36 48 60 72 hours

B C D E 3 80 0.8 800

60 0.6 600 2 40 0.4 400 1 20 0.2 200 Time in shelter (h) total entries 4 days Preferenceratioday n4 Activity duration − dark (h) 0 0 0 0 -/-+/+ -/-+/+ -/-+/+ -/-+/+

Figure 20: Ubn1–/– increased the activity during the dark phase. (A) Distance moved during 3 days. (B) Duration of the activity segments during the dark phase is significantly longer in –/– than in +/+ mice. (C) Time spent in the shelter, (D) strength of the preference on day 4, and (E) total number of entries in the shelter for the first 4 days did not differ. The use of the shelter is not affected by the increased activity of Ubn1-/- mice.

138 Screening for genes involved in avoidance learning 5 139

+/+ –/– Ubn1 Ubn1 = 0.043) = α .825 0.741 0.221 0.516 0.186 0.02* 0.536 0.210 0.456 compared compared 0.045 0.648 0.504 0.023* 0.038* –/– -test ( -test T Ubn1 (n=14)

–/– 10.1±2 175±13 12.4±1.1 36.3±3.7 112.8±11.9 56.3±10.4 12.93±1.53 11.82±3.95 44.8±2.76 93.4±15.75 18.33±4.03 259.8±54.4 1.69±20.46 48.98±10.14 bn1 U mice increased their time freezing, +/+ (n=12) +/+ 5±1.1 Ubn1 167±21 11.2±1.4 6.8±2.4 33.1±2.9 79.9±9.9 145±36.8 40.2±2.3 23.4±7.71 279±66.8 1.48±25.8 8.46±1.96 11.75±2.03 bn1 82.35±21.85 U reduced their activity and increased their freezing +/+ Ubn1 easures # Visits and M Latency (s) Latency (s) –/– # Visits center Latency center (s) Duration lit box(s) Distance moved (m) Distance moved (m) Distance moved (m) # Visits in Open Arms Distance in center (m) Duration in open arms (s) arms open in Duration Ubn1 Latency in open Arms (s) Distance in open arms (m) arms open in Distance mice have a deficit to associate the tone but not the context to shock. to context the not but tone the associate to deficit a have mice mice showed reduced freezing to the tone. Based on the activity in the different different the in Based activity the on tone. the to freezing showed mice reduced –/– –/– mice (Fig. 21A-C). Spatial learning and memory were not affected in in affected not were memory and learning Spatial 21A-C). (Fig. mice mice both responded to the tone by reducing their activity (p=0.005) whereas the mice in the Barnes with maze similar activity in both genotypes (Fig. 21E, F). In the fear Ubn1 +/+ –/– Ubn1 Performance of Ubn1 mice in four anxiety tests. +/+ results in deficient cued fear memory, but unaffected spatial memory –/– Ubn1 Ubn1 Ubn1 iscussion bn1 ests able 9: T Elevated plus maze T Novelty induced hypophagia induced Novelty Dark light box Open field behavior, behavior, indicating an association between the shock and the context. Two hours later were theyplaced in a new context for 180 s with increased activity the in conditioning both tone genotypes. was presented Thereafter, for 180 s and training phases training of fear conditioning this deficit cannot betoattributed shock impaired and tone performance, cognitive normal their despite that, show results these together, Taken perception. spatial learning and memory, and contextual fear conditioning function, that all involve hippocampal conditioning conditioning protocol, during the training phase for fear conditioning, both genotypes exposed a a as to to tone their foot explicit conditioned exposure stimulus before shock. were whereas whereas In this study, we applied a novelscreeningidentifyappliedapproachweatogenesinvolved newavoidance study,in this In learning.Using theinsertional mutation technique, wegenerated new5 mouse strains carrying D U There was no differences in body weight, motor coordination and limb strength between strength limb and coordination in body weight,motor differences no was There and and and to to shock caused a transient massive activity increase. When placed in the conditioning context 24 h after training for 180 s, A B C D 1.2 1.1 2.2 28 1.0 1.0 2.0 26 0.8 0.9 1.8 0.6 0.8 1.6 24 +/+ front paws Strength (N) 0.4 0.7 Strength (N) 1.4 Distance (m) -/-

body weight (g) 1.2 22 0.2 0.6 front and hind paws Day 1 Day 2 0 0.0 0.0 0.0 +/+ -/- 2 4 7 9 +/+ -/- +/+ -/- Trials E F 200 10 +/+ -/- 8 150

6 100 4 50

2 Escape latency (s)

Distance to target hole (m) 0 0 1 2 3 4 5 1 2 3 4 5 Days Days G H Training Retention Training Retention 50 0.4 +/+ 24h -/- 0.3 24h 40 0.2

10

Activity (cm/s) Activity 0.1 % freezing duratoin

0 0.0

Context Context Context Context Tone - CS Shock - USPost-shock New contextNew context+tone New contextNew context+tone

Figure 21: Ubn1 deficiency impairs simple associative memory. (A) Body weight, (B) Rotarod performance, grip strength of (C) front paws and (D) all four paws did not differ between genotypes. (E) The distance and (F) the latency to find the escape hole of the Barnes maze did not differ between Ubn1+/+ and Ubn1–/– mice. (G) There was no difference in the activity values during fear conditioning training and retention tests except in tone-dependent (cued) fear. (H) Percentage of freezing duration indicated no difference during training, in the conditioning context and in the novel context, but Ubn1–/– mice froze less than Ubn1+/+ mice in response to the tone.

transposons in genes that were not previously subjected to functional studies; Dpp10, Fgf13, Kcnd2, Ttc39c and Ubn1. By investigating the spontaneous behavior in a setup without human interference, we observed a general increase in activity in 4 of the 5 strains compared to their respective controls. Whereas Kcnd2–/– mice showed no alteration, Fgf13–/– and Ttc39c–/– mice were more active during light phase, and Dpp10–/– had reduced activity when first introduced in the home cage (day 1) followed by hyperactivity once habituated (day 3), and Ubn1–/– increased their activity in the dark phase as well once habituated (day 3). However, the difference in activity did not interfere with the basic parameters involved in the avoidance task, such as the number of entries, the time spent in the shelter, or the probability to use the preferred

140 Screening for genes involved in avoidance learning 5 141 mice –/– , thereby, deleting This suggests that . mice developed their Specc1 –/– mice were unresponsive to –/– Ttc39c mice showed low anxiety could mice in the Barnes maze and in –/– Ubn1 –/– genotypes showed a decreased activity Ubn1 mice did. gene, resulted in a substantially delayed –/– by the insertion of a transposon could alter Ubn1 mice displayed reduced anxiety-like behavior mice did not show any deficit comparedto their –/– Ubn1 Ubn1 –/– Specc1 Ubn1 and mice were more active in the home cage, no circadian –/– –/– mice. A differential performance in various anxietytests –/– and Kcnd2 Ubn1 Ttc39c –/– during the tone-dependent memory test Ubn1 +/+ , Fgf13 Ubn1 mice displayed a normal preference development, yet they showed a did not decrease their activity or increased the number of freezing –/– –/– –/– Ubn1 Ubn1 mice appears to be weaker when the stressor is of low intensity. . Whereas Dpp10 –/– Interestingly,previousain study the insertion thetransposonof in Ubn1 is a subunit of the HIRA/UBN1/CABIN1/ASF1a (HUCA) histone chaperone complex. avoidance response (Maroteaux et al., 2012). In contrast, expression of two splice variants of the compared to their control littermates on the elevated plus maze but not in other anxiety tests (open field, dark-light box, and novelty-induced hypophagia). Reduced anxiety-like behavior was reported in several studies mutants of and itwas correlated hyperactivityto onthe Alzheimer’s plus maze (Keers et al., 2012; Ognibene et disease models or al., circadian2005). Despite the fact rhythm that clock rhythm abnormalities were observed and hyperactivity was not monitored in any of the stand- alone tests. Here, the increased exploration of open anxiolytic-likearms is phenotype a in robust and situation-specific events compared to respective control littermates, It deposits histone H3.3 into chromatin chromatin assembly (Dulac, 2010; Fischer et al., and 2007; Ray-Gallet et al., 2011). is Post-translational linked to modification gene histoneof protein acetylation,by methylation phosphorylation, activation, DNA Methylation repression and and RNAi are several types of epigenetic modifications(Gräff et al., associated 2011). Therefore, with disruption cognitive of functions explain that mild aversive stimulus (bright light) in the home cage was provokeaversivea notassociation withstrongthepreferred entrance andfearinconditioningenough thecontext to was clearly associated to the foot shock but not to the tone. the HUCA complex and thus the histone localization and their post-translational modification, henceforth affect cognitive function like associative learning. entrance.Screeninganimal behavioravoidance in learning discriminate allowed togene, oneus Ubn1 the aversive stimulus. These two examples show the complexity of this task of andresponses the that diversitycan occur in this test. contextual fear conditioning. In addition, both is not new and the elevated plus maze appears to be the most sensitive assay (see Henderson et al., 2004; Wu et al., 2010) and suggests that these tests investigate different anxiety aspects. Moreover learning and memory was unaffected in deficit inresponding to the aversive stimulus of the avoidance task. Additionally, Ubn1 preference in a significantly different way than their controls but responded to the avoidance task in a similar way. However, their change of preference was only as transient as that of their control mice. displayed reduced anxiety-like behavior in the elevated plus maze and a weaker association to the tone and to the foot shock in the fear condition. Ubn1 Altogether, the associative memory of hippocampal and amygdala functions are unaffected because they are essential for contextual fear conditioning (Maren, 2011). Hence, the fact that Ubn1 and increased scanning behavior when training. the Yet, tone was presented during fear conditioning In this study we developed a novel approach to characterized functionally unknown genes, using both molecular genetics and advanced behavioral phenotyping. By generating 5 new transgenic mouse strains we first noticed that spontaneous behavior is very sensitive to gene modification as 4 out of 5 strains show different activity behavior compare to their respective controls. Second, we revealed the involvement of Ubn1, a subunit of the HUCA histone chaperone complex, in the formation of simple associative memory. The disruption of Ubn1 by the insertion of the transposon is likely to be indirectly, via epigenetic alteration, responsible for a deficit in avoidance learning, an anxiolytic effect in the elevated plus maze and impaired cued (auditory) fear conditioning while more complex hippocampus-dependent forms of learning such as contextual fear conditioning and spatial learning were unaffected. Studying 5 random functionally unknown genes and finding out that one of them is involved in cognitive processes, suggests that a large proportion of genes in the genome are directly or indirectly related to cognition which support the high complexity of brain function.

142 Screening for genes involved in avoidance learning 5 143 ig. ig. ig. .161 .141 .971 .914 S S S .021 .019 .756 .778 .984 .848 0.355 0.018 0.055 0.487 0.284 1 1 1 1 1 3 3 3 3 3 2 2 2 2 2 df df df quare quare quare S S S enotype x days enotype x days enotype x days G G G hi- hi- hi- .012 .559 .037 .503 3.915 .060 .000 7.593 1.962 7.954 2.435 5.345 3.803 3.250 C C C 10.114 Wald Wald Wald Wald ig. ig. ig. S S S .735 .923 .018 .760 .944 .090 .000 .000 .000 0.429 0.070 0.004 0.000 0.000 0.000 1 1 1 1 1 3 3 3 3 3 2 2 2 2 2 df df df days days days quare quare quare S S S hi- hi- hi- .115 .093 .005 .009 21.121 7.049 13.132 2.764 2.869 C C C 8.048 23.071 22.253 45.033 38.830 70.886 Wald Wald Wald Wald ig. ig. ig. .131 .110 S S S .574 .108 .238 .270 .466 .009 .800 0.156 0.231 0.281 0.837 0.879 0.003 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 df df df enotype enotype enotype G G G quare quare quare S S S hi- hi- hi- .315 .532 1.161 9.113 .064 1.219 1.437 1.395 2.577 2.559 2.276 2.010 C C C 0.023 6.864 0.042 Generalized estimation equation for the preference index during the probe phase (days 6-7) Generalized estimation equation for the preference index during the development phase (days1-4) Generalized estimation equation for the preference index during the test phase (days 4-6). Wald Wald Wald Wald 20: 19: 18: S S S upplemental able able able T T T UBN1 TTC39 Kcnd2 Kcnd2 TTC39 UBN1 fgf13 fgf13 Dpp10 Dpp10 Dpp10 fgf13 Kcnd2 TTC39 UBN1 S

Chapter General discussion 6

General discussion 6 147 De novo was insufficient -deficient mice, PCLO Ubn1 in a mouse model. in mice was responsible for an S4814A substitution produces subtle subtle produces substitution S4814A Munc18-1 PCLO Munc18-1 that daily activity bouts are easily affected by random easilyaffected by activityaredailybouts that hapter 5 hapter C a new paradigm to investigate complex cognitive abilities involved in , we studied haploinsufficiency of was inspired by previously published results from a genome-wide association study study association a genome-wide from results published by previously was inspired hapter 2, hapter 3 C C gene in Major Depression Disorder (MDD). By engineering the mutation in mice, we were able able were we mice, in mutation the engineering By (MDD). Disorder Depression Major in gene hapter 4 hapter In In C Finally, by screening five gene-trap mutant lines generated with the sleeping beauty The studies reported in these four chapters demonstrate the usefulness of high-content eneral aim avoidancelearning, takinghomeplacethecageenvironment, in detailedwasinbredstrains 8 in and one gene-trap transposon mouse strain. This study showed that a mild stimulus in a home cage is sufficient to trigger avoidance behavior with striking genotype-dependent behavioral differences. The strengths of this assay were shown in the subsequentcharacterization chapters of mouse models throughof human psychiatric the diseases. G The study of neurological and psychiatric genes involved diseases in the requires behavioral symptoms thorough as well investigationas their interaction of withto the thefacilitate environment the development of efficient therapeutic strategies. High-throughputtechnologies screening have been developed However, in behavioral phenotyping genetics, is still and a slow, in labor-intensive molecularlab reproducibilityprocess (Crabbe with et al., and poor1999). Genetic betweenand cellularneuroscientific research would therefore research. benefit from comprehensive and systematic behavioral screeningto enhancecollection and reproducibility speed of behavioral of data. Furthermore, data limited knowledge exists about genetic factors contributing to differences in learning, memory andto flexibility,affective andand psychiatric their disorders link (Dziegielewski, 2010; Kreek et al., 2005;Lesh et al., 2011; Szöke et al., 2005; Willcutt et al., 2005; Wills et al., 1994). The complexitycompels of themental disordersdevelopment of long-term, unbiased behavior assays environment for thetaking mice. The aim of thisplace thesis was to describein an efficient waya of performing familiar automatedinvestigatebehavioral(PhenoTyper),tocage complexhomephenotypingand the in behavioral traits of inbred and mutant mice using advanced data analysis. mutations in this gene are known to humans. cause We epileptic concluded seizures that and/or haploinsufficiency mental retardation of in increase in anxiety promoting an active stress coping mentalstyle. deficits nor seizuresHowever, as reported in humans.mice showed neither transposon, we demonstratedtransposon,wein changes in synaptic expression levels and synaptic transmission not sufficient to affect mice behavior. mice However, in mice. affect alteration to any behavioral to provoke sufficient not transmission synaptic and levels expression synaptic in changes in in humans, the suggesting of involvement a point non-synonymous mutation (rs2522833) in the PCLO that the of to exchange one demonstrate amino acid in the C2a domain of mutations. However, higher cognitive functions appeared more robust. and high-throughput behavioral phenotyping to pathologies investigate and mouse genes models involved of in neurologic cognitivediscuss the added value of automated functions. phenotyping and its limitations: In I will thoroughly discuss the following paragraphs, the applications, the Iimplications, and the future will perspective of this technique. however, did show impairment in avoidance learning and fear conditioning. Limitations of standalone tests Mood disorders like anxiety, depression and schizophrenia are complex diseases with a “spectrum” nature (many symptoms and comorbidity) making them extremely challenging to study. Scientists, therefore, use animals to model and understand these neuropsychiatric disorders, with the ultimate goal of finding therapeutic targets. However, current animal models display substantial limitations. Indeed, many symptoms established in psychiatric disorders, like delusion, guilt and hallucinations, are specific to human nature and cannot be reproduced with certainty in animals. However, behavioral scientists designed a wealth of tests meant to measure in isolation single behaviors or symptoms that are part of the spectrum of different neuropsychiatric disorders. For example, conflict situations between fear induced by novelty and the instinct to explore, are used in elevated plus maze, dark light box and open field. These 3 tests have one or several safe zones (respectively closed arms, dark compartment and walls) and one or several exposed zones (respectively open arms, lit compartment, and center). These tests have been extensively used to study effects of multiple procedures (mutations, drugs, lesions) on anxiety-related behavior (Bolivar et al., 2000; Luhmann et al., 2005; Simon et al., 1994; Treit and Fundytus, 1988; Turri et al., 2001). The apparent simplicity of analysis, the low cost, and the simplicity of the protocol have made them the gold standard of rodent anxiety measure.

The habituation period bias Classical behavioral tests were designed to catch a specific behavior that provides a reasonable interpretation in the least amount of time. Those tests were thoroughly developed to last between commonly 5 min to maximally 60 min allowing testing a reasonable amount of mice per day (commonly 8 to 30 mice). However, this short test duration with coercion comes with a major drawback: mice are introduced to a completely new environment and are supposed to explore and respond to a specific task in that time given. Thus, classical behavioral tests have no baseline condition and the reactions are always in response to challenging conditions. Consequently, all data of classical tests are acquired in a state of arousal (“stress”) of the animal. The limitation of short test duration was shown in a study comparing anxiety between BALB/c mice and a first generation of captive wild mice. The animals had free access from their home cage to a modified circular open field (2.5 m diameter) for 45 h. BALB/c mice showed a transient period of high anxiety for 30 min (low percent of time in the center) followed by a long lasting calm behavior, whereas the wild mice were exhibiting the opposite pattern, low anxiety behavior during the first 30 min preceding a consistent anxious behavior compared to BALB/c during the rest of the test. The first 30 min of this experiment reflect a transient period of habituation followed by the inversion of behavioral performance in each strain (Fonio et al., 2012b) ), demonstrating a strain- specific response to a novel environment which may change dramatically over time.

Motivation bias Novelty exposure also induces a motivational bias. For example, a mouse introduced on an elevated plus mazes (EPM), will search for a possible escape, visiting all the arms (open or closed). It is generally accepted that a mouse visiting the open arms less than its control is more anxious. However, a recent study demonstrates that a mouse, with free access from its home

148 General discussion 6 149 ); the daily use of the shelter, development of a of development shelter, the of use daily the ); hapter 2 hapter C Automated Automated home cage phenotyping allows the observation of a freely moving animal, the Another interesting study comparing learning abilities of DBA/2J and C57 mice on a Recent studies on deliberate choice (Fonio et al., 2012a, 2009) demonstrate that the ome cage relevance extraction of everyday life parameters concerning e.g. eating, sleeping, drinking and exploration, exploration, and drinking sleeping, eating, e.g. concerning parameters life of everyday extraction and their progression and patterns in time (de Visser et al., 2005; Endo et al., 2011; Goulding et al., 2008; Jhuang et 2008). Under al., such 2010; the Paylor, conditions, of motivation the animal is the animal is intrinsic; not to forced react to and a its is novel reaction environment not biased by any handling. In automated home cage systems like the behavior is PhenoTyper, tracked for multiple consecutive days allowing to observe daily parameters together with certain ( behavior specific shelter as such characteristics H Psychiatric disorders are properly characterized complex by a few multi-factorial-dependentrelatively basic behavioral measures. In disordersthe order underlyingto mechanismsunderstand of and these disorders it is cannotessential to not onlymouse models butalsogenerateto characterize be them under thesuitable conditions. proper Psychiatric diseases are chronic illnesses and their diagnosis, treatment and recovery are longdepend on thelasting environmental factorsprocesses including pharmacotherapies.and This is why designing new ethologically relevant behavioral protocols with more translational power in which the animal can be monitored without human interference (or any major unwanted stressor)phases (habituation,in baseline,different and challenged situations) is a necessity. modified Barnes maze (spatial learning task) using positive (foodreward) and negative (bright light and wind) reinforcement, showed that DBA/2J mice, generally known to perform poorly in emotional and spatial learning task compared to C57 (Podhorna and Brown, 2002), actually perform equally well when positive reinforcement is used (Youn et exampleal.,showing thatlearning abilities2012). highlyaredependent motivationaltheon Thisstatethatand is another different motivations competing with each other will determine Together, those the studies performanceshow the need outcome.to improve behavior standards in a more dynamic and animal centered way. free access to a novel environment shows deterministic rathersuggesting higher cognitive than functions for stochasticexploration of a novel arena. This progression,approach indicates that the way the data are analyzed requires adjustment so that researchers comparethedynamicscan ofthebehavior describe(animal-centered approach)and rather than comparing and interpreting means of arbitrarily selected time bins (experimenter-center approach). cage to an EPM, will show multiple exploratory behaviors closed (e.g. arm returns stretch rather than attend spending time postures) on the and open arms (Roy et al., 2009). This mousesituation; thethe showsby biased severely is EPM the describeusuallyanxietyin of state the that forced on to the EPM tries to escape and therefore explores givenchoicethe(withaccess), free mousewillavoid aopenthe arms, whichall does meannot that the possibilities. However, if the animal is more anxious only that the motivation is different: escape under forced condition versus reduced exploration of a novel environment with threatening features, i.e. areasanxiogenic (open arms). Thus the question is whether a mouse is anxious or just a reasonable? preference for one of the 2 entrances of the shelter, and clear discrimination of genetic background pattern in this etiologically relevant spontaneous behavior, were not described before.

From general to specific behavioral investigations Behavioral differences between inbred strains are often used in classical tests to unravel genetic involvement in specific behaviors. In the home cage environment, C57 showed the most characteristic pattern of entering the shelter with 2 peaks of entries, one at the beginning and one at the end of the dark phase. This behavior was observed in all mouse lines (WT and mutants) backcrossed to C57 studied in this thesis (Chapter 2: Specc1; Chapter 3: Munc18-1; Chapter 4: Pclo and Chapter 5: Dpp10, Kcnd2, Fgf13, Ttc39c, Ubn1). Interestingly, Tecott and his team report a similar circadian pattern in the active and inactive state of OB and Htr2c WT mice on the C57 background (Goulding et al., 2008). This shows that the organization of the activity pattern in freely moving mice in a familiar environment is robust and reproducible within and between laboratories, even if the setup is different. Moreover, other inbred strains showed different pattern of activity and shelter use, suggesting that this behavior is genetically driven (Chapter 2). However, the distinct characteristics of activity measures tracked during longitudinal monitoring appeared to be more sensitive to mutations (Loos et al., 2012). Many of the mutant mice studied showed differences in habituation (Ubn1, Dpp10), disrupted dark-light activity contrasts (Fgf13, Ttc39c) (Chapter 5) or length of activity bouts (Munc18-1) (Chapter 3) compared to their respective littermate controls. Taken together these data suggest that the circadian behavioral pattern is robust and depends on the overall genetic background, whereas the fine-tuning of the activity bouts are affected by mutations even in genes not known to be directly involved in activity behavior. The ability to distinguish global (circadian) activity and bouts of activity at the same time in home cage behavior allows the study of progressive effects of diseases and treatments on metabolism (feeding) and physiology (locomotion). Several automated home cage setups have proven efficient in discriminating behavioral differences between several inbred or mutant strains (de Visser et al., 2006; Goulding et al., 2008; Jhuang et al., 2010) and genetic models of human diseases, such as Rett syndrome (Moretti et al., 2005), Huntington’s disease (Morton et al., 2005), Down syndrome (Faizi et al., 2011) and Alzheimer’s disease (Jyoti et al., 2010).

Automated home cage - a flexible environment High-throughput phenotyping systems are generally well suited for studying spontaneous behavior including sleeping, eating, drinking and movement analysis. In contrast, Crabbe and Morris expressed that those systems are not suited for protocols requiring training periods (e.g., spatial learning) (Crabbe and Morris, 2004). However, in this thesis was described and analyzed an avoidance learning test integrated in a high-throughput system that requires cognitive abilities such as learning and memory and behavioral flexibility (Chapter 2). We observed that mice generally learned to switch preference in less than 12 h, whereas classical tests to study discrimination learning require either a long period of training (e.g. 5-CSRTT) or a strong aversive stimulus as in fear learning (passive avoidance and fear conditioning). Spatial learning tests might not be implemented yet, but it is only a matter of time until there is an adequate way to do so. In our laboratory, we are already working on a modular home cage

150 General discussion 6 151 ),and HZ mice had hapter2 ), i.e., showing C Munc18-1 hapter 3 C long-term monitoring experiments. heterozygous mice ( in vivo Munc18-1 Acutestress responses canbeexpress different3in ways: fight, flightor freeze. in However, Uninterrupted Uninterrupted recording of behavior for several consecutive days comes with other he situation influences the motivation linked to different test environments, and the possibility of this systemto pointsallows trackcollectionthe bodydataonposturesof stretch-attending such severalas postures during body risk assessment (Hager et al., 2014a). T Highly anxious inbred strain like BALB/c and A/J are known novelfor their environments low exploration (Crawley rate et in al., 1997; Lad reproduced et results al.,from tests 2010; in Moy a familiar et environment al., (Tang 2007).A/J miceet beingal., However, the 2002) most we withactive BALB/c strain and during all PhenoTyper experiments. inbredThis strain suggestsare differentthat in theirresponse to stress; the poor performance of certain instrainslocomotor activity-based tests seems thus mostly duestress to coping style ( high anxiety in EPM and in themice express PhenoTyperlow fear response withoutin classical obviousfear conditioning cognitiveand passive avoidance deficits. tasks.concluded We Yet, that these mice these are coping with stress in a more active way; a conclusionsupported that was by autonomic measures (heart rate) in fear similarbaseline experiments.heart theirasrate controls during theexploration phase, butdisplayed highera heart rate response during emotional challenges (fear 2014b). Thisconditioning, again shows the usefulnessnovelty) of (Hager et al., most of the studies on fear responses, increase in freezing and reduction of locomotor activity are the primary parameters used to measure stress in mice. During this thesis, theI encounteredunusual phenotype expressed by since the interpretation of hypoactivity under novel situations is used as index of trait anxiety, this may be confused with state anxiety. advantages. advantages. The reduced interaction between human and animal allows to observe animals all at the same time of the day, eliminating the bias cycle. that circadian canthe of subphases occur different in due day the to across the testing sequential or disturbance rhythm of circadian the A large number of the experiments are conducted during the daytime, because it is work hours (note: some facilities have reversed day night cycles). A major issue is that mice are nocturnal animals meaning that each time they are picked up for an altered. experiment Moreover, most their of resting the period experiments are is done mouse by mouse implying that at tested one, last the that the state circadian different a in first is day) the of beginning the at (tested one the end of the day (just before its natural active phase). Such differencesin timing will certainly sequentially. tested are mice group-housed if concern a more even is This behavior. the influence Such aspects of testing will contribute to variation in the performance, if not an intrinsic part of the behavior in the firstplace. In an automated home cage,the mouse is no performlonger a task at a forced point in time decided to by the experimenter: it has the choice deliberate to do (animal-centered). it do to when and so The validation of new paradigms The validity of a behavioral test is assessed by the combination of criteria: (i) face validity – does the observed behavior in the test corresponds to what is observed in human in a similar situation?, (ii) construct validity – does the test measures what it is supposed to measure?, (iii) predictive validity – can the triggered behavior predict behavior or treatment outcomes (e.g., drugs) seen in other tests (Belzung and Lemoine, 2011; Goswami et al., 2013).

Face validity In Chapter 2, I described and analyzed a new paradigm, using a mild stimulus (bright light) to provoke avoidance of the preferred entrance of the shelter. The test was designed to measure the avoidance response of mice and their ability to switch preference. In regard to the results on 8 inbred strains, we conclude that six strains show a response to the stimulus and two (FVB and C3H) showed no response as they are known to be visually impaired (Wong and Brown, 2006). In the other six, three were good performers (C57, 129S1, DBA) and three poor performers (A/J, NOD, BALB/C). The three well-performing strains showed avoidance for the sanctioned entrance and a stable change of preference for the non-sanctioned entrance, indicating avoidance learning and behavioral flexibility. Based on these results, we can conclude that our test measures the intended behavior (i.e.: avoidance learning using ethologically valid stimulus).

Construct validity To determine construct validity, automated home cage recording faces one challenge. The only data that it can be compared to are derived from classical behavioral tests which are generated under different experimental conditions. Therefore, direct comparison between results is inherently flawed. At the same time, while comparing 5 commonly used renowned measures of anxiety from the elevated plus maze, dark-light box and open field from 3 different laboratories (Lad et al., 2010; Podhorna and Brown, 2002; Wahlsten et al., 2006), and analyzing relations between behaviors on these tests using the Mouse phenome database (http://phenome.jax. org/), we found no significant correlations between the different measures in these established assays (see Table 10). This lack of overlap between established anxiety tests contributes to the notion that multiple behavioral dimensions underlie tests of anxiety (Henderson et al., 2004; Turri et al., 2001) (Chapter 2). Construct validity is hard to establish in experimentally new tests. However, this issue merits further studies comparing different home cage and paradigms.

Predictive validity Classical tests have been validated with drugs such as anxiolytics which increase the time spent in the open arms of the EPM, in the center of the open field, and in the lit compartment of the DLB. In an automated home cage environment, the administration of drugs is more complicated as one needs an accurate, non-invasive and stress-free way to deliver the drugs. Drug administration via the drinking water is inaccurate due to variable consumption across the diurnal cycle and injections are very stressful and require the interference that one aims to avoid. Osmotic mini- pumps could be a solution. However, their volume is very limited, allowing only hydrosoluble drugs

152 General discussion 6

153

Wahlsten Brown Brown Brown

Closed arm duration arm Closed light in Duration center the in Duration index Avoidance

Elevate plus maze plus Elevate box Dark-light field Open maze plus Elevated

5 5 5 5 N Schalkwyk Team

0.937 0.421 0.22 0.071 (2-tailed) Sig. center Duration Parameter

-0.05 0.473 0.665 -0.846 orrelation C earson P field pen O est T

6 6 6 N Wahlsten Team

0.364 0.736 0.855 (2-tailed) Sig. duration arm Closed Parameter

-0.456 0.178 0.097 orrelation C earson P maze plus levate E est T

Brown Team N 7 7

Duration in light in Duration Parameter Sig. (2-tailed) Sig. 0.506 0.724

0.305 orrelation C earson P box ark-light D est T -0.165

Brown Team N 7

Duration in the center the in Duration Parameter Sig. (2-tailed) Sig. 0.057

-0.74 orrelation C earson P field pen O est T

Correlation between anxiety measures of the most used anxiety tests and different laboratories. different and tests anxiety used most the of measures anxiety between Correlation 10: able T to be delivered at appropriate concentrations. Liposoluble drugs like Diazepam require a large volume of DMSO which degrades the mini-pumps membrane (data not shown). Furthermore, as mice have a very high metabolic turnover, the slow delivery of the osmotic pump would require a very high concentration of the drug as the drug is metabolized extremely fast. Another solution is to orally administer the drugs, by preparing pellets of food (e.g. butter) containing an exact amount of lipophilic drugs. Preliminary tests with diazepam dissolved in butter showed sedative effects surprisingly even in a very low amount of diazepam (0.5 mg/kg) (data not shown) and not an increase in activity as is observe on all classical anxiety tests (Crawley, 1985). This shows again the influence of the emotional state on the animal’s behavior and the differential effect of drugs on activity as a consequence of arousal induced by the human handling and administration of the drug. It was as previously observed in rats with the neuropeptide corticotropin-releasing factor (CRF) that increases activity in the home cage while reducing activity in a novel situation due to different basal emotional states (Britton and Indyk, 1990).

Limitations of a new paradigm Automated home cage phenotyping offers a spectrum of improvements of behavioral testing. However, this technology comes with its own limitations.

Individual housing Mice are social animals with a structured hierarchy based on physical dominance. However, tracking multiple animals in an automated home cage is still problematic. Until now, no setup is capable of tracking efficiently more than one animal in a cage (de Visser et al., 2006; Goulding et al., 2008; Jhuang et al., 2010). Some cages use RFID (radio-frequency identification) chips implanted subcutaneously in the animals allowing group housing with individual recognition (Endo et al., 2011; Ogi et al., 2013). However, this requires a setup with different compartments and RFID readers on each door giving the information that whether or not the animal is in this particular compartment and does not allow a fine tracking of the animal.

Improvement of the resolution The tracking system used in the PhenoTyper only locates the center of gravity of the animal, which only allows tracking coarse movement. Refining the track by adding head and tail points will allow to measure posture and scanning behavior by measuring the length and the angle between the 3 points. Our laboratory is already working on a 3-point tracking system and on the description of anxiety-related body posture parameters like stretch-attend posture and head scanning behavior (Hager et al., 2014a). Behaviors like rearing, grooming, foraging, and housekeeping are still hard to track, even with 3 points: such subtly different behaviors need more than 3 points resolution to be reliably detected. A theoretical solution could be multiple camera recording, using two synchronized cameras to record at different point of views (top and side) or 3 dimensions (slightly different angles). These systems will require more computer power, programming, and analytical skills and software (e.g. machine learning approaches) to retrieve, extract and process all available information in a reliable most informative manner for improved data mining.

154 General discussion 6 155 ” (Crabbe” andMorris, ), we studied the effect of gene that is suspected to be to suspected is that gene Pclo Munc18-1 . Our results demonstrated that the , two potential models for cognitive disorders hapter 3 4 C and the average neuroscientific laboratory actually has in mice in (human homologue of hapter 3 in miceinwas responsible higherfora anxiety moreandactivea C STXBP-1 Munc18-1 Munc18-1 , we reproduced in mice a point mutation in the the in mutation a point mice in , we reproduced . This suggests that other factors are involved in the development of such disorders. disorders. such of development the in involved are factors other that suggests This . mutation in that spontaneous activity was easily influenced by mutations and that cognitive PCLO hapter 4 hapter C de-novo hapter 5 In In However, by randomly mutating five functionally previously unstudied genes, we showed C ccuracy of animal models A Micehave provenbe excellentto models forphysiological andcellular studies. During the80’s, researchers generated mouse strains, developing tumors Myc by in overexpressingthe lymphoid the cell oncogene lineage. These models al., were 1985). Micereferred models to were alsoas developed“oncomice” to mimic (Adamsintracerebral hemorrhageet (Wang et 2008),al., type 1 diabetes (Homo-Delarche and Drexhage, 2004), malaria (Hisaedaand Huntingtonet al.,disease 2004)(Lin et al., 2001, 1999). However, the spectrum disordersnature of psychiatric makes them challengingdelusion, hallucination, guilt, or to lack of empathy. Thus most model,of the existing animal models are specificallycharacterized by only the few for transposable symptoms: anxiety, depression particular (implied by lack of symptomsmotivation), impaired learning, and memory. like Studies on high cognitive functions like learning, memory, anxiety, fear, and despair already show that animal model are essential to understand the mechanisms underlying complex behavior. Deficitsshown to increase in depression related behavior hippocampal in mouse models (Santarelli neurogenesis et al., 2003; Snyder were et al., 2011; Surget et al., 2011). In were characterized. To model cognitive retardation with or without epilepticby a seizure caused the haploinsufficiency of Budget limitation Finally, one of the main limitations of automated home cage Indeed behavioral as testingstated is by the Crabbe costs. and Morris: “ greateraccessenthusiastic to students largeaequipmentthan to budget. 2004),which is very limiting for the development of this kind of new technology. However, the variation induced by deploying different experimenters has been studied by a few laboratories (Crabbeet al., 1999), as it contributes toidiosyncratic results obtained in different laboratories. Automation of behavioral phenotyping reduces the variation and will introducedbecome essential to facilitateby objective long-termexperimenters, quantification. absenceoneofcopy of stress-copingstyle. However we did not observe seizures or cognitive impairment as expected based on the human phenotype. involved in MDD in human. Using an extensive test battery tapping a wide range of behaviors, we a of tapping wide range behaviors, battery test in MDD an involved in Using extensive human. C2A the in alanine an of instead a serine carrying mice in behavior the of alteration any see not did of domain in functions are dependent on genes that are not necessarily directly involved in brain functions. Future directions Continuous recording of mice in their home cage opens new angles on how to analyze and describe behavior and to develop new paradigms to challenge mice cognitive abilities (Chapter 2). However, we are only in the early stages of new, less anthropomorphically biased and more animal-centered approaches on behavior (Fonio et al., 2009). The success of automated recording depends on the collaboration of behavioral scientist, software programmers and statisticians. The simplicity of classical behavioral tests that was once their strength has now shown its limits. Continuous recording allows scientists to finally start tackling the high complexity of behavior and see it from a more dynamic angle rather than from the classical static segregated point of view. However, current home cage recording are already asking for upgrades. As discussed previously, recordings could use more resolution (e.g., multi-point tracking), and modular extension (e.g., connection to a separate test compartment) in which different cognitive tests (e.g., spatial memory, working memory, positive and negative reinforcement learning) could be developed. Another interesting extension would be to couple automated home cage recordings with in vivo recording techniques (ECG, EEG, optogenetics and electrophysiology). Automated home cage recordings have the potential to bring a whole new dimension to the study of animal behavior and the screening of gene candidates for human disorders. Our study on Munc18-1 haploinsufficiency, in Chapter 3, requires deeper investigation. The absence of epileptic seizure in our model is still to be confirmed, as seizures were observed in heterozygous Munc18-1 mice with a mixed background (C57-129S1) (personal communication from R. Hensbroek). Backcrossing our model into a mixed background might bring back the seizures in our model. However, to prove the presence of those seizures, the most efficient way is to perform EEG recordings. From our data in Chapter 4, the point mutation in Pclo is apparently not sufficient to induce depressive-like behavior in mice. Through our battery of tests, involving both automated home cage testing as well as a series of classical tests, we did not observe any symptoms related to MDD like despair (forced swim test), low working memory (Barnes maze) and dysregulation in the circadian pattern (home cage recording). However, we did not test anhedonia (e.g. sucrose preference) or reduced social interaction. Furthermore, MDD is also exhibited in humans in more complex symptoms which are as yet hard to measure in a mouse, like feelings of worthlessness, guilt, and suicidal ideation. The Pclo point mutation might still be responsible for the dysregulation of higher cognitive function which would require symptom-specific genetic analyses rather than symdrom-based analyses. Behavioral characterization of randomly mutated mice using the Sleeping Beauty technique, in Chapter 5, shows the importance of studying all the genes in the genome, rather than focusing research on only a few popular genes. With five random mutations, we managed to observe that the circadian activity patterns were robust, whereas alternations in activity bouts were more sensitive to genetic mutation. This will need further investigation to understand what the roles of these and other genes are in spontaneous behavior. We also observed that a ubiquitous gene like Ubn-1 plays an indirect role in learning and memory. However, to confirm the phenotype, subsequent experiments need to be done like 5-choice serial reaction time task or a Barnes maze with a reversal learning to study behavioral flexibility.

156 General discussion 6 157 inal remarks F Overall, automated home cage phenotyping seems to considerably complement the study of ethologically valid behavior in mice, and is an important extension of classical behaviorAutomated hometests. cage tests facilitate the study of spontaneous daily behavior which until now received less attention and allow determining many new measures of increased discriminatory power for thorough assessment of interactions. It geneswill be useful and (nature) reveal new insights and for the environmenthuge number not of models (nurture) characterizedthat are yet and in their a satisfactory from manner physiology to to drawpathology. However, conclusionsthe vast about amount of the data changeoverstillgenerated by requiressuch system a thorough expertise to optimal spectrum extract of measures. the Automated biologicallyphenotyping will relevantcontinue to parametersdevelop increasing technical alongandtechnological intocomputation(e.g., with power) possibilities. While anstill under development,studiesautomatedshowthethesisthatthisreporteddoinphenotyping on starts to become an essential tool in the characterization translationalof valuegenetically of disease/disorder models. modified mice for higher

Chapter Material & Methods 7

Material & Methods 7 161 section of section Materials and Methods Materials at tery b oratory animals b he test a ovelty-induced hypophagia rip strength levated plus maze each chapter includes a list of the test used as well as the variation in the tests specific to the chapter. the to specific tests the in variation the as well as used test the of list a includes chapter each Specific for each chapter T G Neuromuscular function was assessed by sensing the peak amount of force (N) in mice graspingapplied a pull bar connected to a force meter (1027DSMinstruments, Grip Strength Columbus,Meter, Columbus OH, USA). Mice were front paws only, followed allowed by grasping 4 to times with graspfront and hind paws. the The median pullof each 5 repetitions bar was taken as grip strength. 4 times with Here Here all tests used The arein described. chapters the different L E We performed the elevated plus maze introducedonto the center of an Elevated plus maze (EPM) facing test a closed arm (arms 30 x 6 cm, as describe in (Loos illuminatedwashigh,elevatedabovesingleEPMcmground). Thewallscmwhitethe witha35 50 et al., 2009). Mice were fluorescent light bulb from above(open arms70 lx, closed arm 30 lx) and exploratory behavior Forthree days prior to the testing mice day, were familiarized to a highly palatable snack (a few crumbsofcream cracker) placed into familiara metal food cup inthe home. Onthe testing day noveltyinduced hypophagia was assessed by transferring mice to a novel clean cage with fresh bedding containing the metal cup with the familiar snack. The latency to start eating the snack was recorded manually. If a subject did not eat within 600 s, the maximum time was assigned. N was video tracked for 5 min (Viewer 2, Biobserve GmbH, St. Augustin, Germany). The border between center and arm entries was defined at 3 cm into each arm. Zone visits were analyzed using the elevated plus maze plugin of the tracking software, which was set to count zone visits if both the nose and body reference point had crossed the zone a zone border. The dependent measures were the number of open arm visits, time spent on the open arms and the first latency to explore an open arm. To counteract the detection of distance moved due to jitter of body reference point produced by grainy video signal, track correction option was set to 1.

Open field Mice were introduced into a corner of a white square open field (50 x 50 cm, walls 35 cm high) illuminated with a single white fluorescent light bulb from above (200 lx) and exploration was tracked for 10 min (Viewer 2, Biobserve GmbH, St. Augustin, Germany). The surface area was divided into nine equally sized squares, and the center square was used as center area. Zone visits to the center area were counted using the body reference point. To counteract the detection of distance moved due to jitter of body reference point produced by grainy video signal, track correction option was set to 1.

Novel object and place object task The test was conducted in open field boxes, the day after the open field test. The testing day was divided into three phases: sample (S), place (P) recognition and object (O) recognition. During the initial familiarization stage, two identical objects were placed in two adjacent corner of the arena equidistant (10 cm) from both walls. Each mouse was introduced in the opposite left corner of the arena for 10 min to explore the objects and returned to the home cage afterwards. For P recognition testing 2 hours later, mice were introduced for 5 min in the same arena with identical

162 Material & Methods 7 163 object place sample ark-light box D Mice were introduced into the dark compartment (<10 lx, length x width x height: 25 x 25cm) ofx a30 dark light box. 60 s later the motorized door opened providing access to an identical sized compartment which was brightly lit (~625 lx) and left open for 10 min. Visits to, and time spent in the light compartment were counted protruded whenat least 2 cm into thethe light compartment bodyaway from the door. reference point of a mouse objects, of which one (object 1) was placed in the opposite corner from object 2 instead of the adjacent For corner. O recognition testing 2 hours after the P testing, mice were introduced for 5 min in the same arena with a familiar object in the same corner and a novel object in the same corner than object 1 in the place recognition phase. Objectobject,detectedcontainingthe ascm) exploration7.5 x was (7.5 zone square a scored detectedin wasanimal the of whenhead the by video tracking software (Viewer 2, Biobserve GmbH, St. Augustin, Germany). The objects for a mouse to discriminate consisted of plastic Lego stacks of identical color but different shape (cubepyramid).ormousemousethe color to From the ofobjects (red,yellow blue)or wellas as the role of the pyramid shape (familiar or novel) were randomized and counterbalanced across genotypes.Theand arena objects werecleaned withethanol waterand 70% aftereach trial. The basic measure was the time (s) taken by the mice to explore the objects in the P and O trial. The performance was evaluated by calculating a discrimination index (O1-O2/O1+O2), where O1 = time spent exploring the object in the new location (P) or the novel object O2 (O), = time spent exploring the object in familiar location (P) or familiar object (O). Accelerating rotarod Motor function and motor learning was evaluated using an accelerating rotarod (Roto-rod series 8, IITC Life Science, Woodland Hills, CA, USA). On day one, mice received two habituation trials of 120 s (acceleration of from 0 to 20 rpm in 120 s) followed by 3 training trials (acceleration of from 0 to 40 rpm in 180 s). On day 2, mice received 5 additional training trials. Previous observations indicated that never reach the maximum programmed rpm during habituation and training sessions. The maximum rpm reached in each trial was the dependent measure.

T-maze Short-term spatial memory was assessed in a T-maze (white PVC, arms 30 x10 cm) according to a protocol described before (Deacon and Rawlins, 2006). A sample trials was started by placing a mouse into the start arm (base of the T) to explore the maze. During a sample trial, a removable central partition (17 cm long) protruded from the center of the back wall into the start arm, forcing mice to choose left or right arm while positioned in the start arm. After an entry into a goal arm, a guillotine door at the entrance of the goal arm was lowered, and the mouse was contained in the arm for 30 s. To prepare for the test trial, the central partition was removed and a guillotine door at the end of the start arm was lowered. The test trial was initiated by placing the mouse in the start arm and removing both the guillotine door from the goal arm and the guillotine door at the end of the start arm. A successful alternation was scored if mice choose to enter the previously non-visited goal arm. A total of 6 sample and test trials were performed, distributed across two days with at least 1 h in between.

Barnes maze Apparatus and room: The Barnes maze consisted of a circular grey platform (diameter 120 cm) elevated 100 cm above the floor with 24 holes (4.5 cm diameter) spaced at equal distance 5 cm away from the edge of the platform. One hole was designated as escape hole, and equipped with a cylindrical entrance (4.5 cm diameter x 5 cm depth) mounted underneath the maze providing access to an escape box (15.3 x 6.4 x 6.1 cm) containing a metal stairway for easy access that was not visible unless mice approached the hole closely. Other holes were equipped identical cylindrical

164 Material & Methods 7 165 Data analysis: The path travelled by a mouse was video tracked by an overhead camera Protocol: Mice received training sessions twice a day, typically in the morning and afternoon.morningand the typically in day, receivedatrainingsessionsProtocol:Mice twice and analyzed using Viewer 2 software with Barnes maze plugin (Viewer 2, Biobserve GmbH, St. Augustin, Germany). The distance and latency to reach the target location were recorded, wellas as hole visits defined by crossing of the headreference point into a holezone drawn 1 cm around each hole. Multiple consecutive hole visits were counted as single number hole of visit,single holeand visitsthe to holes other that the target hole were maycounted developas a errors.serial Micesearch strategy to locate a target hole. A hole visits was marked visit,as ifserial the previous visit has occurred 1 or 2 holes away. As index of a serial search strategy the percentage of serial hole visits was calculated as follows: (total hole visits) / (serial hole visits). To detect a spatial search strategy, the Barnes maze was divided into octants, and all 24 holes were assigned to one of 5 zone categories based on the (i.e. distance target, away 1st,from the 2nd, target 3rd hole and 4th zone). The proportioncalculated as follows:of (total numberhole of visitsvisits to a hole to in a givena zone) / specific[(total number of zone hole was visits)*(number of zones in the category)]. This proportion was calculated for duringall training zones,as well as probeboth trials, and analyzed to detect a spatial search strategy. Mice were introduced in an opaque cylinder (10 cm diameter, 25 cm high) placed in the center of the maze, after which the experimenter left the room and closed the door. The cylinder was pulled upwards 30 s later, and mice could explore the maze to latencylocate to enterthe the escapeescape hole hole.exceeded If 300 s,the mice were gently guided toward the escape hole. During the first 2 habituation sessions, the escape box contained cagemouse’s own enrichmenthome cage, ofand a once in the cylinder, mice were left in there for 60 s. After each mouse, the platform and escape box were thoroughly cleaned with 70 % ethanol. The platform was rotated 90° after each trial to avoid the use of any odors cue. During a 300 s probe trial the escape hole was identical to all 23 other holes. entrances, entrances, but without escape box. Visual extra-maze cues (50 x 50 cm) composed of black and white patterns were mounted on the walls ~70 cm away from the maze. Three fans surrounding the maze (60 cm away from the maze spaced apart) ~120° produced a variable airflow acrossthe entire maze by a slow 90° horizontal movement, proving both an aversive environment as well as dispersion of any odor cues. Several fluorescenttube lights mounted at the ceiling provided sound. background provided ceiling the to mounted speaker A lx). (1000 illumination bright Modified Barnes maze The modified Barnes maze was performed as describe in (Youn et al., 2012), using a large round platform of 122 cm of diameter with 44 holes arranged in such a way that no serial exploration is obvious. Mice were trained to target the target hole using visual extra maze cues placed on the wall (as for the classical Barnes maze). All holes contained white double-floored cup underneath (5 cm Ø). The platform can be divided in 4 quadrants, each containing 2, 3 and 6 holes in the inner, middle and outer ring respectively. The target hole was always placed in the middle ring. To prevent odor cues, the target location was varied between each animals and the platform was clean with 70% ethanol solution and rotated between trials. All the cups were removed and washed under running water once a day. A dark cylinder (6.8 x 12 cm Ø x length) served as transport container from the home cage to the Platform, to minimize the handling stress. A video camera placed above the center of the platform, monitored the performance of mice during trials. Images were recorded and analyzed by a computer located in an adjacent room by using Viewer software (Viewer 2, Biobserve GmbH, St. Augustin, Germany). The experimenter was not present in the experimental room during trials but observed the experiments on the computer screen. The distance and latency to reach the target location were recorded and tracked by the software

Fear conditioning Fear conditioning was performed as described before (Misane et al., 2005), employing a computerized fear conditioning system (TSE, Bad Homburg, Germany). Mice were trained in a Plexiglas cage (36 x 21 x 20 cm) placed within a constantly illuminated (120–500 lx) dark gray fear conditioning box (context 1). A high-frequency loudspeaker (Conrad, KT-25-DT, Hirschau, Germany) provided constant background noise in the conditioning box (white noise, 68 dB). Context and tone conditioning consisted of exposure to the conditioning context for 180 s followed by 30 s tone exposure used as CS (10 kHz, 75 dB, pulsed 5 Hz). Termination of tone was followed after 15 s by the onset of the US (foot shock: 0.7 mA, 2 s, constant current) delivered through a stainless steel floor grid (4 mm diameter, distance 9 mm). Mice were removed from the fear conditioning box 30 s after shock termination. The fear conditioning box was thoroughly cleaned with 70% ethanol before the placement of each mouse. The light in the experimental room was turned off during training. Context-dependent memory was assessed 23 h after conditioning in the fear conditioning box (context 1) for 180 s without tone or US presentation under otherwise identical conditions as during training. Tone-dependent memory was tested in a novel context (context 2) 3 h after context-dependent memory tests. Context 2 was an identically sized Plexiglas cage with a plain floor (no shock grid) in a white surrounding (380–480 lx) outside the fear conditioning box. No background noise was provided and lights in the experimental room were on. In the tone-dependent memory test, a 180-s exposure to a novel context without stimulation (pre-CS phase) was followed by a 180-s period of tone presentation (CS phase). The context 2 box was cleaned with 1% acetic acid in the same way as the fear conditioning box. During training and testing, activity (cm/s) was measured by a photo beam detection system (10 Hz detection rate, resolution of 1.3 x 2.5 cm).

166 Material & Methods 7 167 P/SO; RF Concepts,RF Dundonald,P/SO; combinationUK),in X Thedigital moviewas usedoffline for automatic analysis, using custom-developed software coustic startle and prepulse inhibition orced swim test Acoustic Acoustic startle and prepulse inhibition (PPI) were measured during one 45 min session in four Plexiglas cylinders in ventilated sound-attenuating chambers (Med Associates, St. Albans, VT), During CA).Irvine, (NewportCorporation, tables heavyvibration-free passive separate on placed startle the of intensity The dB. 75 of background noise white provided speaker a separate testing, chamber, closed a in cylinders Plexiglas the inside placed microphone a with calibrated was stimuli intensities pulse and prepulse reported the Hence, off. switched was noise background white while are lower than the actual cumulative sound pressure level during testing (e.g., a 75 dB pulse in addition to a 75 dB background noise add up to 78 dB). The session started with a period habituation of 5 min, followed by a total of 260 trials with pseudo randomized interval periods (5–15 s) consisting of acoustic startle trials with white noise bursts at various intensities (65, 75, 80, 70, and 115 110, dB; 105, 10 100, trials 95, per inhibition and 85, 90, trials intensity) prepulse with white 71 intensity; and 79 dB; 30 per trials prepulse 67, 65, (0, intensities bursts prepulse at various noise startle produced boxes 4 all that such order pseudo-randomized in dB) 120 always intensity startle stimuli prepulse noise white of onset trials, prepulse the of half In time. same the exactly at stimuli time) rise/fall programmed no ms; (40 stimuli startle and time) rise/fall programmed ms 1 ms; (20 was separated by a 30 ms, in the other half by 100 ms interval. In each startle the after interval trial, ms 100 the the during collected was units) highest machine relative (in startle peak intensity stimulus, from which the individual mean highest startle intensity peak during the 100 ms null- period prior to prepulse stimuli was subtracted. The startle sensitivity was determined for each mouse by determining the ST at which a statistically significantstartle response was measured (one-sample t-test, with average null-period as reference value). The equipment was calibrated to allow for a wide range of startle intensities, however, the force generated by some mice the at highest pulse intensities could exceed the dynamic the range when of Therefore, analyses. the subsequent in PPI equipment of percentage (maximum the reducing of artificially units) 2047 PPI no 15) of out 5 than more (e.g., 33% than more was trials pulse dB 120 censored such of number intensity startle [(mean * 100 = PPI follows: as calculated was PPI of percentage The calculated. was pulse)]. startle (mean intensity / prepulse) with pulse startle(mean intensity - pulse) F Mice were placed in a rectangular transparent Perspex tank (22 x 14 x 35.5 cm, length x width x height), filled upto a height of 30 cm with water (25 °C).Traditionally, around manytank FST setups.is However, used round tanks in distort the image of the mouse in the tank particularly onthesides thereby confounding theprecision ofquantitative motion-detection analysis. Two swim sessions of 10 and 6 min sepearted by 24h, were conducted. After the session, mice were placed on a clean dry tissue under a warm light bulb (max. 5 min) in their home cage.resolution digitalAcamera (Sunkwang high- B140 A for motion detection (R.F. Jansen and O. Stiedl; see http://www.falw.vu/~ngc/FST.html). The program for automatic analysis, using a custom-developed motion detection algorithm, is with Virtualdub software (v1.9; www.virtualdub.org) was used to record the swim session (AVI files) at arate of 25 frames per sec. made available upon contact (O. Stiedl, [email protected]). The software calculates activity based on frame-to frame pixel changes.

Passive avoidance The experiments were performed with a computer-controlled passive avoidance (PA) system (Model 256000, TSE Systems, Bad Homburg, Germany). The PA box is divided by a black wall, which contains an automated sliding door (9 x 11.5 cm, width x height) with two equal compartments (30 x 25 x 30 cm), a bright compartment with a light intensity of 1000 lx and a dark compartment with a light intensity of 10 lx. The floor of the PA box is a stainless steel grid through which a single foot shock was be applied during training. Before each experiment the box was cleaned with 70% ethanol solution. The approach mirrored previous experiments (Baarendse et al., 2008) with the following procedures: Training: During training, a mouse was placed in the bright compartment were it had a 60 s time interval to explore (initial exploration phase). After 60 s the door opened to provide access to the dark compartment. One second after entering the dark compartment with all four paws (monitored by photo beams; detection rate 100 HZ) the door was closed. Three seconds later, a 2 s shock of 0.7 mA (constant current) was delivered through the grid floor. After shock exposure, the mouse remained in the dark compartment for 60 s before it was returned to its home cage, to prevent an association between the handling and the shock. The mice were guided gently into a transfer container (8 x 10 x 8 cm, width x height x depth) filled with nesting material from their home cage. Retention tests: 24 h after training, mice were placed in the bright compartment to perform retention test 1 (R1). After 15 s the door opened between the two compartments, allowing each mouse to move freely between the bright and the dark compartment. The door closed after 600 s had elapsed, after which each mouse was returned to the home cage via the transfer container. PA retention tests were repeated daily for 8 days to determine the extinction of the avoidance response across retention tests.

5-choice serial reaction time task At 8 to 9 weeks of age, mice were food-restricted to gradually decrease their body weight to 90–95% of their initial body weight before daily training in operant cages commenced (5 days each week). Water was available ad libitum throughout the experiment. Mice were trained to perform the 5 choice serial reaction time task (5-CSRTT) on an individually paced schedule, as described previously (Loos et al., 2011, 2009). During the first week, mice underwent 1

168 Material & Methods 7 169 -Y -Y coordinates of X

servation b T Noldus4.1, Information Technology, Wageningen, The Netherlands), starting at X utomated home cage o and data analyses Mice were transferred to specially designed home cages Information (PhenoTyperTechnology, Wageningen,model The 3000, Netherlands) Noldus in the second half of the light subjective phase (14:00 h – 17:00 h). The describedbehavior of in mice detail was video-tracked previously for ((Maroteauxthree EthoVision days et as al., 2012) EthoVision HTP 2.1.2.0, based on A habituationperformmagazineweek,trainingnext4sessions.trained to theandan were miceIn 5-CSRTT to commenced only and reward, a earn to holesstimulus the into instrumentalresponse training when they earned at least 50 rewards within one session. trial Duringstarted 5-CSRTTwith a training response a of the subject into the illuminated magazine, which magazine lightswitched and off after an ITI of 5 s a stimulus in 1 of the 5 stimulus holes was presented for limited a duration (stimulus duration). A response in the correct stimulus hole within the limited holdtimeafterterminations of4 ofthestimulus switched onthemagazine light anddelivered a food pellet. Both an incorrect response into a non-illuminated stimulus hole and an omission of a correct response resulted in a time-out period, during which all stimulus lights and the house light were turned off. When the time-out period ended, both the house light and the magazine light were switched on, and the subject could start the next trial. An impulsive response into a non-illuminated stimulus hole during the delay period also resulted in a time-out period, but a subsequent response into the illuminated magazine restarted the same trial. The of percentage omission errors was defined as [100 x(omissions) incorrectresponses)]. Response responses)(number correctaccuracydefined was x [100 of as / /(omissions + number of correct and (number of correct and incorrect responses)]. Impulsivity in terms of the percentage impulsive responses was defined as [100 x (number of impulsive s, 5-CSRTT first incorrect16 responses)].the at responses)+correct set In session, stimulus was duration the / (number of omissions + which was decreased in subsequent sessions to 8, 4, 2, 1.5 and 1 s as soon as the subject reached criterion performance (omissions < 30%, response accuracy > 60%, started trials > 50) or after 10 sessions. Intra-individual variability in correct response latencies (response variability in short) was defined by the standard deviation of the correctresponse latencies. Thetotal numbersessions of required to reach the stimulus duration of 1 s was used as measure of required training sessions. Dependent measures were calculated from the 6th until the 10th session at stimulus duration of 1 s, and the average of these sessions was used as standard 5-CSRTT performance. In theweek following the 10th andsession, 12.5 the ITIs), was with programmed each vary to 7.5 (5, interval occurring an equal number of times within session. Strains that completed fewer than 50 trials on average in combination with long magazine latencies (greater than 4 s), indicativetogether ofreduced motivation, wereexcluded. Individual mice wereexcluded analysesfrom if they initiated fewer than 30 trials on average, had long magazine latencies, or made no correct or incorrect responses during two or more standard sessions. the first subjective dark phase (19:00 h). Resulting track files, containing the center of gravity (COG) at a resolution of 15 coordinates per second, were processed using AHCODATM analysis software (Synaptologics BV, Amsterdam, The Netherlands) to generate behavioral parameters. 115 activity parameters were generated as described in detail previously (Loos et al., 2014) The last and first 10 min of each dark and light phase were not included in summary statics, to ensure that a potential asynchrony of the data streams and light regime in the testing facility would not affect these statistics.

PhenoTyper

PhenoTyper

r PhenoType

per PhenoTy

Statistical analysis Before any analysis was performed, data were examined for outliers (>3 times the SD from the strain mean). All statistical analyses were performed using IBM SPSS statistic 20 (IBM corporation, Armonk, NY, USA). Genotype differences were compared using parametric tests (T-test, ANOVA, repeated-measures ANOVA) whenever normality and homoscedasticity criteria were met. Otherwise, nonparametric tests were performed (Kruskal-Wallis, Mann- Whitney U-test). Nonparametric data are presented as box plots (ends of the box denoting the 25 and 75% interquartile range and the whiskers providing the upper and lower quartile ±1.5 times the interquartile range, respectively, while the line in the box denotes the median). An error probability level of p<0.05 was accepted as statistically significant throughout the study of the classical behavioral tests. For all given level of analysis in the PhenoTyper data, statistical analysis was based on estimated FDR (Verhoeven et al., 2005), P-values were correct by minimum positive FDR with a threshold set at 5%.

170

References & 173 ie, L., Fujita, Y., Kwan, E., Davletov, B., Monnier, P.P., X Wells, S., Potter, P., Kelly, M., Cox, R.D., Brown, S.D.M., 2008. ENU mutagenesis, a way ., A van Grootheest, G., Jansen, R.F., Pieneman, A.W., Ögren, S.O., Verhage, M., Stiedl, O., 2008. Han, L., Tassew, N.G., He, Y., Wang, L., Bolivar, V.J., Vedder, M.J., Geistfeld, J.G., 2004. Genetic and behavioral differences among five ., Moore, M.W., 2012. Towards an encyclopaedia of mammalian gene function: the International Chambon, P., Angelis, M.H. de, 2005. EMPReSS: standardized phenotype screens for functional Mitchell, P., Goodwin, G.M., 2001. Cognitive deficits in depression Possible implications for Indyk, E., 1990. Central effects of corticotropin releasing factor (CRF): evidence for similar L M.N.,L.,Flaherty,Mappingquantitative2001. knockout/congenicof withstrains.loci Genometrait ., Chang,Leibowitz,Dourmashkin, G.Q., 2006. Distinct S.F., J.T., phenotypes obesity-proneof AKR/J, Noordewier, B., Lamé, M., 1989. Genetic dissociationNoordewier,Lamé,B.,M., 1989. multiple of morphine effects among C57BL/6J, ., Harris, A.W., Pinkert, C.A., Corcoran, L.M., Alexander, W.S., Cory, S., Palmiter, R.D., Brinster, R.L., ., 1979. Memory deficits1979. associated with senescence: neurophysiological a and behavioral study the in Lipkind, D., Horev, G., Fonio, E., Kafkafi, N., Golani, I., 2010. Ten ways to improve the quality of descriptions descriptions of quality the improve to ways Ten 2010. I., Golani, N., Kafkafi, E., Fonio, G., Horev, D., Lipkind, ., .J.J., Lemoine, M., 2011. Criteria of validity for animal models of psychiatric disorders: focus on anxiety Pooler, O., Flaherty, L., 2001. Inbred strain variation in contextual and cued fear conditioningbehavior.fear cued contextualand variationin strainInbred L.,2001. Flaherty, O., Pooler, ., Wang, K.-S., Pan, Y., 2011. Genome-wide association analysis of gender differences in major Genome-widedepressive major 2011.differencesassociationgender in analysis of Y., K.-S., Pan, Wang, ., M Lässle, M., 1999. The tetratricopeptide repeat: a structural motif mediating protein-protein interactions.protein-proteinmediating motif structural a tetratricopeptiderepeat: The 1999. M., Lässle, P ., P ., M M ., rozena, . R . ., Cook, M., Flaherty, L., 2000. List of transgenic and knockout mice: behavioral knockoutmice:profiles. GenomeMamm. transgenicand of ListL., 2000.Flaherty, Cook,M., K Y ., ., . .- A M A D D C . ook, L ., .J., Oakeshott, S., Shamy, J.L., El-Khodor, B.F., Filippov, I., Mushlin, R., Port, R., Connor, D., Paintdakhi, A., . N D . .W. . M Buisson, M., Pajunen, T., Ryoo, Y.W., Giot, J.F., Gruffat,Giot, Sergeant,J.F., H., Buisson,Ubinuclein, 2000. J., Ryoo,M., Y.W., novel Pajunen,Uitto, A., a T., C C V V S S G G ., ., F S forwardunderstandto genefunction. GenomicsAnnu.49–69. Rev. Hum. Genet. 9, 1985. The c-mycThe1985.oncogene drivenimmunoglobulinby enhancers induces lymphoid malignancy transgenicin mice.Nature318,533–538. nuclearproteininteracting withcellular viralandtranscription factors.Cell1165–1176. Biol. J. 148, DBA2J and C57BL/6J mice compared to control strains. Int. J. Obes. 2005 30, 50–59. doi:10.1038/sj.ijo.0803110DBA2Jmicecompared C57BL/6Jcontroland strains.to 50–59. Obes. Int. 200530, J. disorder in the Netherlands NESDA and doi:10.1016/j.jad.2011.04.054 NTR population-based samples. J. Affect. Disord. 133, 516–521. Gaisano, Sugita, H.Y., S., 2008. Munc18-1 is critical for plasma membrane localization of syntaxin1 but not of SNAP-25722–734.PC12incells.doi:10.1091/mbc.E07-07-0662 Mol. Biol. Cell 19, functionalPsychiatry J. neuropathology. doi:10.1192/bjp.178.3.200200–206. 178,Br. Differential involvement of the dorsal hippocampusdoi:10.1002/hipo.20356 Hippocampus11–19. 18, in passive avoidance in C57BL/6J and DBA/2J mice. High-Throughput2013.Menalled,AutomatedPhenotyping D., L., Brunner,Ramboz,Kwak, S.,Howland, S., D., of Two Genetic Mouse Models of Huntington’s Disease. PLoS Curr. 5. doi:10.1371/currents.hd.124aa0d16753f8 8215776fba102ceb29 rat. J. Comp.Physiol.rat. J. Psychol. 74–104. 93, DBA/2JC3H/HeJand inbred mousestrains. Physiol.69–74. 46, Behav. disordersdepression.and doi:10.1186/2045-5380-1-9Biol. Mood Anxiety 9. Disord.1, of whole-animal movement. Neurosci. Biobehav. Rev. 34, 1351–1365. doi:10.1016/j.neubiorev.2010.04.004 1351–1365. 34, Rev. Biobehav. Neurosci. movement. whole-animal of BioEssays News Rev. Mol. BIES5>3.0.CO;2-N Cell. Dev. Biol. 21, 932–939. doi:10.1002/(SICI)1521-1878(199911)21:11<932::AID- Res.1549–1552.11,doi:10.1101/gr.194001 Off. J. Int.J. Off. Mamm. doi:10.1007/s003350010051 Genome Soc. 260–274. 11, Mamm.Genome 12,651–656.doi:10.1007/s003350020039 inbredmousestrainscommonly productionthe usedin transgenic of knockout and mice. Genes BrainBehav. doi:10.1111/j.1601-183x.2004.00064.x 149–157. 3, interactionswithenvironmental noveltywith andcaffeine. Psychopharmacology (Berl.) 366–370. 101, annotationmousethegenome.of1155–1155. doi:10.1038/ng1105-1155 Nat. Genet. 37, MousePhenotyping Consortium. Dis.Model. Mech.289–292.doi:10.1242/dmm.009878 5, eferences cevedo- dams, J. ho, lexander,J., ragam, runachalam, ustin, R A A A A A A A Baarendse, Balci, Barnes, Belknap,J. Belzung, Benjamini, Benjamini, Blatch, Bolivar, Bolivar, Bolivar, Bothe, Britton, Brown, Brown, Brunner, D., Nestler, E., Leahy, E., 2002. In need of high-throughput behavioral systems. Drug Discov. Today 7, S107–112. Büeler, H., Fischer, M., Lang, Y., Bluethmann, H., Lipp, H.P., DeArmond, S.J., Prusiner, S.B., Aguet, M., Weissmann, C., 1992. Normal development and behaviour of mice lacking the neuronal cell-surface PrP protein. Nature 356, 577–582. doi:10.1038/356577a0 Buselmaier, W., Vierling, T., Balzereit, W., Schwegler, H., 1981. Genetic analysis of avoidance learning by means of different psychological testing systems with inbred mice as model organisms. Psychol. Res. 43, 317–333. Carlson, C.M., Dupuy, A.J., Fritz, S., Roberg-Perez, K.J., Fletcher, C.F., Largaespada, D.A., 2003. Transposon mutagenesis of the mouse germline. Genetics 165, 243–256. Cases-Langhoff, C., Voss, B., Garner, A.M., Appeltauer, U., Takei, K., Kindler, S., Veh, R.W., De Camilli, P., Gundelfinger, E.D., Garner, C.C., 1996. Piccolo, a novel 420 kDa protein associated with the presynaptic cytomatrix. Eur. J. Cell Biol. 69, 214–223. Castle, W.E., 1919. Studies in heredity in rabbits, rats, and mice, by W. E. Castle. Carnegie Institution or Washington,, Washington,. Cohen, J., 1988. Statistical Power Analysis for the Behavioral Sciences. L. Erlbaum Associates. Collins, S., Martin, T.L., Surwit, R.S., Robidoux, J., 2004. Genetic vulnerability to diet-induced obesity in the C57BL/6J mouse: physiological and molecular characteristics. Physiol. Behav. 81, 243–248. doi:10.1016/j. physbeh.2004.02.006 Cool, B.H., Chan, G.C.-K., Lee, L., Oshima, J., Martin, G.M., Hu, Q., 2010. A flanking gene problem leads to the discovery of a Gprc5b splice variant predominantly expressed in C57Bl/6J mouse brain and in maturing neurons. PloS One 5, e10351. doi:10.1371/journal.pone.0010351 Cotella, D., Radicke, S., Cipriani, V., Cavaletto, M., Merlin, S., Follenzi, A., Ravens, U., Wettwer, E., Santoro, C., Sblattero, D., 2012. N-glycosylation of the mammalian dipeptidyl aminopeptidase-like protein 10 (DPP10) regulates trafficking and interaction with Kv4 channels. Int. J. Biochem. Cell Biol. 44, 876–885. doi:10.1016/j.biocel.2012.02.011 Cox, J., Mann, M., 2008. MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nat. Biotechnol. 26, 1367–1372. doi:10.1038/nbt.1511 Crabbe, J.C., Morris, R.G.M., 2004. Festina lente: late-night thoughts on high-throughput screening of mouse behavior. Nat. Neurosci. 7, 1175–1179. doi:10.1038/nn1343 Crabbe, J.C., Wahlsten, D., Dudek, B.C., 1999. Genetics of mouse behavior: interactions with laboratory environment. Science 284, 1670–1672. Crawley, 2007. What’s wrong with my mouse: behavioral phenotyping of transgenic and knockout mice. John Wiley & Sons. Crawley, J., Goodwin, F.K., 1980. Preliminary report of a simple animal behavior model for the anxiolytic effects of benzodiazepines. Pharmacol. Biochem. Behav. 13, 167–170. Crawley, J.N., 1985. Exploratory behavior models of anxiety in mice. Neurosci. Biobehav. Rev. 9, 37–44. doi:10.1016/0149-7634(85)90030-2 Crawley, J.N., Belknap, J.K., Collins, A., Crabbe, J.C., Frankel, W., Henderson, N., Hitzemann, R.J., Maxson, S.C., Miner, L.L., Silva, A.J., Wehner, J.M., Wynshaw-Boris, A., Paylor, R., 1997. Behavioral phenotypes of inbred mouse strains: implications and recommendations for molecular studies. Psychopharmacology (Berl.) 132, 107–124. Cross-Disorder Group of the Psychiatric Genomics Consortium, Lee, S.H., et al., 2013. Genetic relationship between five psychiatric disorders estimated from genome-wide SNPs. Nat. Genet. doi:10.1038/ng.2711 Crusio, W.E., 2004. Flanking gene and genetic background problems in genetically manipulated mice. Biol. Psychiatry 56, 381–385. doi:10.1016/j.biopsych.2003.12.026 D’Hooge, R., Nagels, G., Franck, F., Bakker, C.E., Reyniers, E., Storm, K., Kooy, R.F., Oostra, B.A., Willems, P.J., De Deyn, P.P., 1997. Mildly impaired water maze performance in male Fmr1 knockout mice. Neuroscience 76, 367–376. Darwin, C., 1839. The zoology of the voyage of H.M.S. Beagle, under the command of captain Fitzroy, R.N. during the years 1832 to 1836. de Heer, Spruijt, 2008. Learning (in) the PhenoTyper: an integrative approach to conducting cognitive behavioural challenges in a home cage environment. Presented at the Measuring behavior 2008. De Mooij-van Malsen, J.G., Yu, K.L., Veldman, H., Oppelaar, H., van den Berg, L.H., Olivier, B., Kas, M.J.H., 2009. Variations in ventral root axon morphology and locomotor behavior components across different inbred strains of mice. Neuroscience 164, 1477–1483. doi:10.1016/j.neuroscience.2009.09.008

174 References & 175 BP1 mutations. X ., Dobbin, M., Tsai, L.-H., 2007. Recovery of learning and memory is associatedmemoryislearning Recoveryand of L.-H., 2007. Tsai, Dobbin,M.,., X 1-adrenergic receptor by xamoterol as a potential cognitive enhancer. Neurobiol. β Institute,2012. European Bioinformatics Institute |Homepage |EBI [WWW Document]. .2005.00181.x X Südhof, Genetics 1999. T.C., of synaptic vesicle function: toward the complete functional ., R 2010. DSM-IV-TR in Action.inDSM-IV-TR 2010.JohnWiley Sons.and ., Rawlins, J.N.P., 2006. T-maze alternation in the rodent. Nat. Protoc. 1, 7–12. doi:10.1038/ Campbell, A.M., Park, C.R., Halonen, J., Zoladz, P.R., 2007. The temporal dynamics model of F . S Randall,M., 2013. Effects of age, sex, and sertraline administration on seizure-inducedrespiratory ., Garner, C.C., 2002. Gene structure and genetic localization of the PCLO gene encoding the .J., ., Kessels, M.M., Qualmann, B., Chung, W.J., Nash, J., Gundelfinger, E.D., Garner, C.C., 2003. Interactions 2003. C.C., Garner, E.D., Gundelfinger, J., Nash, W.J., Chung, B.,Qualmann, M.M.,Kessels, Fritz, S., Largaespada, D.A., 2001. Transposition and gene disruption in the male germline of the Wienholds, E., Plasterk, R.H., 2001. Regulated transposition of a fish transposon in the mouse germ Akagi, K., Largaespada, D.A., Copeland, N.G., Jenkins, N.A., 2005. Mammalian mutagenesis using hacón, M Chung, W.J., Zhai, R., Cases-Langhoff, C., Voss, B., Garner, A.M., Kaempf, U., Kindler, S., Gundelfinger, E.D., E.D., Gundelfinger, S., Kindler, U., Kaempf, A.M., Garner, B., Voss, C., Cases-Langhoff, R., Zhai, W.J., Chung, Gundelfinger, E.D., 2006. Molecular organization and assembly of the presynaptic activezone of ., . L ., C M ., . Sananbenesi, F., Wang,Sananbenesi, F., Weckhuysen, S., Holmgren, P., Suls, A., Van Dyck, T., Goossens, D., Del-Favero, J., Jansen,A.,Del-Favero,Goossens,Verhaert,J., D., Dyck, T., Suls,A.,Weckhuysen, Van Holmgren, S., P., ., ., D . McGue, M., Sex 1997. differences and nonadditivity in heritability of the Multidimensional Personality D ., L.,vandenBos, R., Kuurman,Kas, M.J.H., W.W., Spruijt, B.M., 2006.Novel approach thebehaviouralto . D E 2010. Brainfunction2010. chromatinand plasticity.Nature728–735.465, doi:10.1038/nature09231 L., van den Bos, R., Spruijt, B.M., 2005. Automated home cage observations as a tool to measure the effects the measure to tool a observationsas cage home Automated 2005.B.M., Spruijt,R., Bos, den vanL., C D ., Benjamini, Y., Golani, I., 2009. Freedom of movement and the stability of its unfolding in free explorationunfoldingmovementstabilityFreedomitsfreetheofandofin Golani, 2009. I.,Benjamini, Y., ., . R .J., .J., . Bader, P.L.,C.,Bader, Tun, Encarnacion, A., Kleschevnikov, A., N.,Belichenko,Priestley,Saw, M.,R.W.,P., Tsien, . S Maekawa, F., Võikar, V., Haijima, A., Uemura, Y., Zhang, Y., Miyazaki, W., Suyama, S., Shimazaki,Miyazaki,Suyama,S., K., W., Wolfer, Zhang, Y., Haijima, A.,Uemura,Y., V., Võikar, Maekawa, F., A ., L S ., S A S ., A A ., ., D C E T M isser, isser, Dis. 43, 397–413. doi:10.1016/j.nbd.2011.04.011397–413.Dis. 43, characterization of inbred mice: automated doi:10.1111/j.1601-183 home cage observations. Genes Brain Behav. 5, 458–466. neurotransmitterrelease.Results 49–68. Probl.43, CellDiffer. of wheelofrunning cage flooron locomotion. doi:10.1016/j.bbr.2004.12.004 382–388. Brain Res. Behav. 160, Garner, C.C., 2000. Piccolo, a presynaptic zinc finger protein structurally related to bassoon. Neuron 25, 203–214. 25, Neuron bassoon. to related structurally protein finger zinc presynaptic a Piccolo, 2000. C.C., Garner, nprot.2006.2 presynapticactiveprotein zone Neurosci.Piccolo.Int.J. Dev. Int. Off. Soc.Neurosci.161–171. J. Dev. 20, K., Lagae, L., Jordanova, A., Van Coster, R., Yendle, S., Berkovic, S.F., Scheffer, I., Ceulemans, B., De Jonghe, P., 2010. Clinical spectrum of early-onset epileptic encephalopathies associated with ST Neurologydoi:10.1212/WNL.0b013e3181f4d7bf1159–1165. 75, between Piccolo and the actin/dynamin-binding protein Abp1 link vesicle endocytosis to presynaptic active Biol.zones.Chem.doi:10.1074/jbc.M21079220020268–20277. J. 278, emotionalmemory processing: synthesisa neurobiologicalthe on basisstress-inducedof amnesia, flashbulb traumaticand memories,60803. doi:10.1155/2007/60803 NeuralYerkes-Dodson theandPlast.law. 2007, anatomyorganelle.anof Physiol.doi:10.1146/annurev.physiol.61.1.753 Annu. 753–776.Rev. 61, a highlyamobile somatic Sleeping Beauty transposon system.221–226.Naturedoi:10.1038/nature03691 436, QuestionnaireScales.Pers. Soc. J. Psychol. 72,929–938. withchromatin178–182. remodelling.doi:10.1038/nature05772 Nature447, mouse. Genes. N. Y. N 2000 30, 82–88.2000 30, N mouse.Genes.Y. N. D.P., D.P., Yada, T., Tohyama, C., Lipp, H.-P., Kakeyama, M., 2011. Automated test of behavioral flexibilitybehavioral usingina sequencing mice IntelliCage.taskin BrainRes.Behav.221, doi:10.1016/j.bbr.2011.02.037 172–181. EMBLEBI.http://www.ebi.ac.uk/URL- (accessed 3.3.11). line.Proc.Natl. doi:10.1073/pnas.121569298Acad.A. 6759–6764.S.Sci. 98,U. arrest in the DBA/1 mouse model of sudden unexpected death in epilepsy 78–82.doi:10.1016/j.yebeh.2013.04.003(SUDEP). Epilepsy Behav. EB 28, Mobley, W.C., Shamloo, M., 2011. Comprehensive behavioral phenotyping of Ts65Dn mouse model of Down syndrome: activation of of mice.ofProc. Natl. Acad.A. 21335–21340.S.Sci. 106,U. doi:10.1073/pnas.0812513106 V V e e eacon, eprez, iamond, ulac, upuy, upuy, upuy, upuy, ziegielewski, uropeanBioinformatics ndo, ejtova, enster, enster, ernández- inkel, ischer, ischer, onio, aingold, aizi, enster, enster, D F D F D F D F D F D F D D F F D E E F F F Fonio, E., Benjamini, Y., Golani, I., 2012a. Short and long term measures of anxiety exhibit opposite results. PloS One 7, e48414. doi:10.1371/journal.pone.0048414 Fonio, E., Golani, I., Benjamini, Y., 2012b. Measuring behavior of animal models: faults and remedies. Nat. Methods 9, 1167–1170. doi:10.1038/nmeth.2252 Gagne, J.R., Saudino, K.J., 2010. Wait for it! A twin study of inhibitory control in early childhood. Behav. Genet. 40, 327–337. doi:10.1007/s10519-009-9316-6 Garcia, E.P., Gatti, E., Butler, M., Burton, J., De Camilli, P., 1994. A rat brain Sec1 homologue related to Rop and UNC18 interacts with syntaxin. Proc. Natl. Acad. Sci. U. S. A. 91, 2003–2007. Garcia, J., Gerber, S.H., Sugita, S., Südhof, T.C., Rizo, J., 2004. A conformational switch in the Piccolo C2A domain regulated by alternative splicing. Nat. Struct. Mol. Biol. 11, 45–53. doi:10.1038/nsmb707 Gavériaux-Ruff,C ., Kieffer, B.L., 2007. Conditional gene targeting in the mouse nervous system: Insights into brain function and diseases. Pharmacol. Ther. 113, 619–634. doi:10.1016/j.pharmthera.2006.12.003 GeneNetwork, 2012. GeneNetwork Search [WWW Document]. URL http://www.genenetwork.org/ (accessed 3.3.11). Gerlai, R., 1996. Gene-targeting studies of mammalian behavior: is it the mutation or the background genotype? Trends Neurosci. 19, 177–181. Geurts, A.M., Wilber, A., Carlson, C.M., Lobitz, P.D., Clark, K.J., Hackett, P.B., McIvor, R.S., Largaespada, D.A., 2006. Conditional gene expression in the mouse using a Sleeping Beauty gene-trap transposon. BMC Biotechnol. 6, 30. doi:10.1186/1472-6750-6-30 Giese, K.P., Fedorov, N.B., Filipkowski, R.K., Silva, A.J., 1998. Autophosphorylation at Thr286 of the alpha calcium- calmodulin kinase II in LTP and learning. Science 279, 870–873. Goldfarb, M., 2005. Fibroblast growth factor homologous factors: evolution, structure and function. Cytokine Growth Factor Rev. 16, 215–220. doi:10.1016/j.cytogfr.2005.02.002 Gordon, J.W., Ruddle, F.H., 1981. Integration and stable germ line transmission of genes injected into mouse pronuclei. Science 214, 1244–1246. Goswami, S., Rodríguez-Sierra, O., Cascardi, M., Paré, D., 2013. Animal models of post-traumatic stress disorder: face validity. Front. Neurosci. 7, 89. doi:10.3389/fnins.2013.00089 Goulding, E.H., Schenk, A.K., Juneja, P., MacKay, A.W., Wade, J.M., Tecott, L.H., 2008. A robust automated system elucidates mouse home cage behavioral structure. Proc. Natl. Acad. Sci. 105, 20575–20582. doi:10.1073/ pnas.0809053106 Gräff, J., Kim, D., Dobbin, M.M., Tsai, L.-H., 2011. Epigenetic regulation of gene expression in physiological and pathological brain processes. Physiol. Rev. 91, 603–649. doi:10.1152/physrev.00012.2010 Gram, L., 1990. Epileptic seizures and syndromes. Lancet 336, 161–163. Gregory, S.G., et al., 2006. The DNA sequence and biological annotation of human chromosome 1. Nature 441, 315–321. doi:10.1038/nature04727 Groot, A.S., de Sonneville, L.M.J., Stins, J.F., Boomsma, D.I., 2004. Familial influences on sustained attention and inhibition in preschoolers. J. Child Psychol. Psychiatry 45, 306–314. Hager, T., Jansen, R.F., Pieneman, A.W., Manivannan, S.N., Golani, I., van der Sluis, S., Smit, A.B., Verhage, M., Stiedl, O., 2014a. Display of individuality in avoidance behavior and risk assessment of inbred mice. (submitted to Frontiers in Behav. Neurosci.). Hager, T., Maroteaux, G., Pont, P. du, Julsing, J., van Vliet, R., Stiedl, O., 2014b. Munc18-1 haploinsufficiency results in enhanced anxiety-like behavior as determined by heart rate responses in mice. Behav. Brain Res. 260, 44– 52. doi:10.1016/j.bbr.2013.11.033 Hall, C., L, E., 1932. A study of the rat’s behavior in a field. A contribution to method in comparative psychology. Univ. Calif. Publ. Psychol. 6, 1–12. Hamdan, F.F., Gauthier, J., Dobrzeniecka, S., Lortie, A., Mottron, L., Vanasse, M., D’Anjou, G., Lacaille, J.C., Rouleau, G.A., Michaud, J.L., 2011. Intellectual disability without epilepsy associated with STXBP1 disruption. Eur. J. Hum. Genet. EJHG 19, 607–609. doi:10.1038/ejhg.2010.183 Hamdan, F.F., Piton, A., Gauthier, J., Lortie, A., Dubeau, F., Dobrzeniecka, S., Spiegelman, D., Noreau, A., Pellerin, S., Côté, M., Henrion, E., Fombonne, E., Mottron, L., Marineau, C., Drapeau, P., Lafrenière, R.G., Lacaille, J.C., Rouleau, G.A., Michaud, J.L., 2009. De novo STXBP1 mutations in mental retardation and nonsyndromic epilepsy. Ann. Neurol. 65, 748–753. doi:10.1002/ana.21625

176 References & 177 PPswe/ β ., Khilnani, V., Poggio, T., Steele, A.D., Serre, T., 2010. Automated home-cage X 2001. The comorbidity of major depression and anxiety disorders: recognition and Drexhage, H.A., 2004. Immune cells, pancreas development, regeneration and type 1 ., ., F Turri, M.G.,Turri,DeFries, Flint,J.C.,analysis 2004. QTL J., multiple of behavioral measures anxietyof A . ., Roelandse,M., Wierda, K., van Aerde, K.I., Toonen, R.F.G., Hensbroek, R.A., Brussaard, A., Matus, D M 2011. Anxiety2011.DisordersEpilepsy: in Forgotten PsychiatricThe 90–91.Comorbidity.11, Epilepsy Curr. . . ., de Mooij-van Malsen, A.J.G., Olivier, B., Spruijt, B.M., van Ree, J.M., 2008. Differential genetic Maekawa, Y., Iwakawa, D., Okada, H., Himeno, K., Kishihara, K., Tsukumo, S., Yasutomo, K., 2004. 2007. Cognitive Impairments in Major Depression and Bipolar Disorders. Psychiatry Edgmont 4, 12–14. 4, Edgmont Psychiatry Disorders. Bipolar and Depression Major in Impairments Cognitive 2007. Garrote, E., Yu, ., R N H West,R.S.,anxietylaboratory inTaming 2010. 825–826. mice.Nat. doi:10.1038/nmeth.1500 Methods 7, ., ., ., ., King, S.M., 2001. Influence of circadian phase andtest illumination on pre-clinical models of anxiety. ., M Kuroiwa, A., Ikawa, M., Okabe, M., Kondoh, G., Matsuda, Y., Takeda, J., 2001. Efficient chromosomal Yusa, K., Yae, K., Odajima, J., Fischer, S.E.J., Keng, V.W., Hayakawa, T., Mizuno, S., Kondoh, G., Ijiri, T., Plano, A., Riedel, G., Platt, B., 2010. EEG, activity, and sleep architecture in a transgenic A Pedroso, I., Breen, G., Aitchison, K.J., Nolan, P.M., Cichon, S., Nöthen, M.M., Rietschel, M., Schalkwyk, D . ., H H Hackett, P.B., Plasterk, R.H., Izsvák, Z., 1997. Molecular reconstruction of Sleeping Beauty, a Tc1-like . Mulder, C.L., Luijendijk, H.J., van Duijn, C.M., Hofman, A., Uitterlinden, A.G., Tiemeier, H., 2010. The H elarche, Slaughter, C.A.,Slaughter, Synaptic1993.vesicle Südhof, fusionT.C., complex contains homologueunc-18 bound to ., Ornitz, D.M., 2008. Functional evolutionary history of the mouse Fgf gene family. Dev. Dyn. Off. Publ. Publ. Am. Off. Dyn.Ornitz, FunctionalDev. D.M.,2008.family.evolutionary gene Fgf mouse thehistory of Südhof, T.C., 1995. A novel ubiquitous form of Munc-18 interacts with multiple syntaxins. Use of the yeast two- yeast the of Use syntaxins. multiple with interacts Munc-18 of form ubiquitous novel A 1995. T.C., Südhof, ., L ., ., A P ., Ko, J., Shin, H., Lee, J.-R., Lim, C., Han, J.-H., Altrock, W.D., Garner, C.C.,J.-R.,Garner,Lee,Gundelfinger, Altrock,C.,Shin, H.,Lim,J.-H.,Premont, Han, E.D., J.,W.D., R.T., Ko, Yae, K., Yae, .W., Hayakawa, Mizuno, T., S., Yusa, K., Uno, Y., Kokubu, C., Kinoshita, Akagi, T., K., Jenkins, N.A., D .J. ., N K ., K R ., ., ., A ., Y Y K N M S PSEN1A246EAlzheimer’s diseaseAlzheimers mouse. J. doi:10.3233/JAD-2010-100879Dis.22,JAD 873–887. doi:10.5698/1535-7511-11.3.90 syntaxin.doi:10.1038/366347a0 347–351. Nature366, regulation of motor activity and anxiety-related behaviors in mice using an automated home cage task. Behav.task. cageautomatedhome anxiety-relatedan activityusing and motormicebehaviors regulation ofin doi:10.1037/0735-7044.122.4.769 Neurosci. 122, 769–776. hybrid system to study interactions between proteins involved in membrane traffic. J. Biol. Chem. 270, 13022–13028. 270, Chem. Biol. J. traffic. membrane in involved proteins between interactions study to system hybrid A., Verhage, M., 2004. Trophic support delays but does not prevent cell-intrinsic degeneration Neurosci.J. 623–634.deficient 20, of munc18-1.doi:10.1111/j.1460-9568.2004.03503.xfor Eur. neurons L.C., Fernandes, C., 2012. Reduced anxiety and depression-like e38263. mousedoi:10.1371/journal.pone.0038263afterhours. behaviours7, ONEPLoS in the circadian Period mutant Copeland, N.G., Horie, K., Takeda, J., 2005. Region-specific2005. the using saturationJ.,mice germline mutagenesis in Takeda, K.,Horie,Copeland, N.G., SleepingBeauty transposondoi:10.1038/nmeth795 system.Nat. Methods 763–769. 2, PCLO gene and depressive disorders: replication in a population-based study. Hum. Mol. Genet. 731–734. 19, doi:10.1093/hmg/ddp529 Kaang, B.-K., Kim, E., 2003. The GIT family of proteins forms multimers and associates withcytomatrix proteinBiol.thePiccolo.Chem.6291–6300. J. 278,doi:10.1074/jbc.M212287200 presynaptic in mice. Behav. Genet. 34, 267–293. doi:10.1023/B:BEGE.0000017872.25069.44mice.Genet.inBehav. 267–293. 34, managementprimaryin care.Prim.CareCompanion Clin.Psychiatry J. 244–254.3, Escapeofmalaria parasites fromhost immunity requires CD4+ CD25+ regulatory cells.T Nat. Med. 29–30. 10, doi:10.1038/nm975 diabetes.Immunol.Trends doi:10.1016/j.it.2004.02.01225,222–229. transposition of a Tc1/mariner- like transposon Sleeping Beauty in mice. Proc. Natl. Acad. Sci. U. S. A. 98, 9191– transposonA.98,like SleepingS. Proc. mice. Natl.Acad. U.BeautySci.transposition Tc1/mariner- in a of doi:10.1073/pnas.1610717989196. transposonfish, fromtranspositionits and human in cells. 501–510. Cell 91, Physiol.72,Behav.99–106. Matsuda, Y., Plasterk, R.H.A., Takeda, J., 2003. applicationCharacterizationgeneticto screening mice. inMol.Cell. Biol. 23,9189–9207. of Sleeping Beauty transposition and its behaviouralphenotyping mice.ofNat. Commun. doi:10.1038/ncomms1064. doi:10.1038/ncomms10641, Assoc. Anat. 237, 18–27. doi:10.1002/dvdy.21388 Assoc.18–27. Anat. 237, ata, eeroma,J. ek, enderson, arvey, arvey, irschfeld, isaeda, omo- orie, orie, urst,J. ata, ata, anner, as, im, eers, eng, V toh, vics, Z., H K H K H K H H K K H H H H H H Jhuang, Jyoti, H Jones, I I Kohli, M.A., Lucae, S., Saemann, P.G., Schmidt, M.V., Demirkan, A., Hek, K., Czamara, D., Alexander, M., Salyakina, D., Ripke, S., Hoehn, D., Specht, M., Menke, A., Hennings, J., Heck, A., Wolf, C., Ising, M., Schreiber, S., Czisch, M., Müller, M.B., Uhr, M., Bettecken, T., Becker, A., Schramm, J., Rietschel, M., Maier, W., Bradley, B., Ressler, K.J., Nöthen, M.M., Cichon, S., Craig, I.W., Breen, G., Lewis, C.M., Hofman, A., Tiemeier, H., van Duijn, C.M., Holsboer, F., Müller-Myhsok, B., Binder, E.B., 2011. The neuronal transporter gene SLC6A15 confers risk to major depression. Neuron 70, 252–265. doi:10.1016/j.neuron.2011.04.005 Koolhaas, J.M., de Boer, S.F., Buwalda, B., van Reenen, K., 2007. Individual variation in coping with stress: a multidimensional approach of ultimate and proximate mechanisms. Brain. Behav. Evol. 70, 218–226. doi:10.1159/000105485 Kreek, M.J., Nielsen, D.A., Butelman, E.R., LaForge, K.S., 2005. Genetic influences on impulsivity, risk taking, stress responsivity and vulnerability to drug abuse and addiction. Nat. Neurosci. 8, 1450–1457. doi:10.1038/nn1583 Kumar, R.A., Chan, K.L., Wong, A.H.W., Little, K.Q., Rajcan-Separovic, E., Abrahams, B.S., Simpson, E.M., 2004. Unexpected embryonic stem (ES) cell mutations represent a concern in gene targeting: lessons from “fierce” mice. Genes. N. Y. N 2000 38, 51–57. doi:10.1002/gene.20001 Lacaze, E., Gruchy, N., Penniello-Valette, M.-J., Plessis, G., Richard, N., Decamp, M., Mittre, H., Leporrier, N., Andrieux, J., Kottler, M.-L., Gerard, M., 2013. De novo 15q13.3 microdeletion with cryptogenic West syndrome. Am. J. Med. Genet. A. doi:10.1002/ajmg.a.36085 Lad, H.V., Liu, L., Paya-Cano, J.L., Parsons, M.J., Kember, R., Fernandes, C., Schalkwyk, L.C., 2010. Behavioural battery testing: evaluation and behavioural outcomes in 8 inbred mouse strains. Physiol. Behav. 99, 301–316. doi:10.1016/j.physbeh.2009.11.007 Lander, E.S., et al., International Human Genome Sequencing Consortium, 2001. Initial sequencing and analysis of the human genome. Nature 409, 860–921. doi:10.1038/35057062 Largaespada, D.A., 2009. Transposon mutagenesis in mice. Methods Mol. Biol. Clifton NJ 530, 379–390. doi:10.1007/978-1-59745-471-1_20 Lazarevic, V., Schöne, C., Heine, M., Gundelfinger, E.D., Fejtova, A., 2011. Extensive remodeling of the presynaptic cytomatrix upon homeostatic adaptation to network activity silencing. J. Neurosci. 31, 10189–10200. doi:10.1523/JNEUROSCI.2088-11.2011 Leal-Ortiz, S., Waites, C.L., Terry-Lorenzo, R., Zamorano, P., Gundelfinger, E.D., Garner, C.C., 2008. Piccolo modulation of Synapsin1a dynamics regulates synaptic vesicle exocytosis. J. Cell Biol. 181, 831–846. doi:10.1083/jcb.200711167 Ledford, H., 2012. Mice guide human drug trial. Nature 483, 389. doi:10.1038/483389a Lesh, T.A., Niendam, T.A., Minzenberg, M.J., Carter, C.S., 2011. Cognitive control deficits in schizophrenia: mechanisms and meaning. Neuropsychopharmacol. Off. Publ. Am. Coll. Neuropsychopharmacol. 36, 316– 338. doi:10.1038/npp.2010.156 Levitis, D.A., Lidicker, W.Z., Freund, G., 2009. Behavioural biologists don’t agree on what constitutes behaviour. Anim. Behav. 78, 103–110. doi:10.1016/j.anbehav.2009.03.018 Lewis, C.M., Ng, M.Y., Butler, A.W., Cohen-Woods, S., Uher, R., Pirlo, K., Weale, M.E., Schosser, A., Paredes, U.M., Rivera, M., Craddock, N., Owen, M.J., Jones, L., Jones, I., Korszun, A., Aitchison, K.J., Shi, J., Quinn, J.P., Mackenzie, A., Vollenweider, P., Waeber, G., Heath, S., Lathrop, M., Muglia, P., Barnes, M.R., Whittaker, J.C., Tozzi, F., Holsboer, F., Preisig, M., Farmer, A.E., Breen, G., Craig, I.W., McGuffin, P., 2010. Genome-wide association study of major recurrent depression in the U.K. population. Am. J. Psychiatry 167, 949–957. doi:10.1176/appi.ajp.2010.09091380 Limbach, C., Laue, M.M., Wang, X., Hu, B., Thiede, N., Hultqvist, G., Kilimann, M.W., 2011. Molecular in situ topology of Aczonin/Piccolo and associated proteins at the mammalian neurotransmitter release site. Proc. Natl. Acad. Sci. U. S. A. 108, E392–401. doi:10.1073/pnas.1101707108 Lin, C.H., Tallaksen-Greene, S., Chien, W.M., Cearley, J.A., Jackson, W.S., Crouse, A.B., Ren, S., Li, X.J., Albin, R.L., Detloff, P.J., 2001. Neurological abnormalities in a knock-in mouse model of Huntington’s disease. Hum. Mol. Genet. 10, 137–144. Lin, X., Cummings, C.J., Zoghbi, H.Y., 1999. Expanding our understanding of polyglutamine diseases through mouse models. Neuron 24, 499–502. Loos, M., Koopmans, B., Aarts, E., Maroteaux, G., van der Sluis, S., Neuro-BSIK Mouse Phenomics consortium, Verhage, M., Smit, A.B., 2014. The genetic influence on spontaneous home-cage behavior measured in 11 common inbred mouse strains. Submitt. Genes Brain Behav.

178 References & 179 .2012.00820.x X BP1-relatedencephalopathy X Drecombinant inbred strains. Genes X .2011.00754.x X .2009.00528.x X D recombinant inbred strains. Genes Brain Behav. doi:10.1111/j.1601- X Huston, J.P., Hasenöhrl, R.U., 2005. Contralateral increase in thigmotactic scanning following Noor, Noor, A., Vincent, J.B., Lionel, A.C., Feuk, L., Skaug, J., Shago, M., Moessner, R., Pinto, D., Ren, Nguyen, T.H., Han, L., Latham, C.F., Osborne, S.L., Wen, P.J., Lim, S.J.T., Sugita, S., Collins, B.M., Loos, Loos, M., van der Sluis, S., B., Koopmans, Aarts, E., van Gassen, K., Geurts, A., D.A., Largaespada, Kaplan, J.M., 2008. UNC-18 promotes both the anterograde trafficking and synaptic function of 1966. Time-dependent1966. processes memoryin storage.Science 1351–1358.153, Wyss, A., Rieger, H., Mohajeri, M.H., A 2009. comparison of learning and memory characteristics 2010. Genomewide association studies and assessment of the risk of disease. N. Engl. J. Med. 363, Rickman, C., Chamberlain, L.H., Duncan, R.R., Munc18-12007. prevents the formation of ectopic ., ., ., ., ., ., ., ., ., Fisher, E.M., 1997. PhenotypicE.M., Fisher, analysis--making 1997. Genet.yourmostthemouse.ofTrends 254–256. 13, .J., T L R 1865.Versucheuber Pflanzen-Hybriden (Experiments Plant in Hybridization) 13:221–268. . G Burkhardt, P., de Wit, H., Toonen, R.F., Fasshauer, D., Verhage, M., 2012. Munc18-1 mutations that Moutard, M.-L., Trouillard, O., Gourfinkel-An, I., Jacquette, A., Arveiler, B., Morice-Picard, F., Lacombe, F., Morice-Picard, B., Arveiler, Jacquette, Gourfinkel-An,A., I., M.-L.,Moutard, O., Trouillard, . M O A N Himes, B.E., Brehm, J.M., Boutaoui, N., Klanderman, B.J., Sylvia, J.S., Lasky-Su, J., 2010. Analyses of 2011. Seeking a spotless mind: extinction, deconsolidation, and erasure of fear memory. Neuron 70, . H ., Staal, J., Pattij, T., Smit, A.B., Spijker, S., 2011. Independent genetic loci for sensorimotor gating . . N ., van der Sluis, S., Bochdanovits, Z., van Zutphen, I.J., Pattij, T., Stiedl, O., Smit, A.B., Spijker, S., 2009. et al., 2013. A mega-analysis of genome-wide association studies for major depressive disorder. Mol. C ., Staal, J., Pattij, Neuro-BSIK T., Mouse Phenomics Consortium, Smit, A.B., Spijker, S., 2012. Independent ., E T Villeneuve,N.,Chouchane, M., Kaminska, A., Laroche, C., Barthez, M.A., Gitiaux, C., Bartoli, C., Borges- A ., ., C .2011.00754.x G ., C ., ., ., S M E ., X S M M M M augh,J. geneticloci for sensorimotor gating and attentional performance in B Brain Behav. 11, 147–156. doi:10.1111/j.1601-183 147–156. 11,BrainBehav. and attentional performance in B 183 Activity and impulsive action are controlled by doi:10.1111/j.1601-183 different817–828. 8, Behav. genetic and environmental factors. Genes Brain unilateralbarrel-cortex doi:10.1016/j.bbr.2004.06.00639–43. lesionmice.BrainRes. inBehav.157, Psychiatrydoi:10.1038/mp.2012.21 497–511. 18, Meunier, F.A., 2009. Abrogating Munc18-1-SNARE complex interaction has limited impact on exocytosisPC12Biol.cells.Chem.doi:10.1074/jbc.M109.013508 J. 21637–21646.284, in 166–176. doi:10.1056/NEJMra0905980 166–176. 830–845.doi:10.1016/j.neuron.2011.04.023 Spruijt, B.M., Stiedl, O., Smit, A.B., Verhage, M., 2012. High throughput phenotyping of avoidance learning in mice mice in learning avoidance of phenotyping doi:10.1111/j.1601-183 throughput High Behav. 2012. M., Brain Verhage, A.B., Genes Smit, O., gene. Stiedl, B.M., Spruijt, novel a identifies and genotypes different discriminates Y., Thiruvahindrapduram,Y., B., Fiebig, A., Schreiber, S., Friedman, J., Ketelaars, C.E.J., Vos, Y.J., Ficicioglu, Kirkpatrick,C., S., Nicolson, R., Sloman, L., Summers, A., Gibbons, C.A.,Thompson, A.,Teebi, A.,Vardy, C., Chitayat,Crosbie, D., V., Weksberg,Luscombe, R.,S., Baatjes, R., Zwaigenbaum,B., Szatmari,2008. S.W., StructuralScherer, P., L., variation ofchromosomes Roberts,inautism spectrum Am. disorder. J. W., Fernandez, Hum.Genet. 82,477–488.doi:10.1016/j.ajhg.2007.12.009 syntaxin.3836–3846.doi:10.1091/mbc.E08-02-0160 Mol.Biol. Cell 19, of young and middle-aged wild-type mice in the IntelliCage. J. Neurosci. Methods 180, 43–51. doi:10.1016/j. jneumeth.2009.02.018 SNAREcomplexes livingindoi:10.1242/jcs.020230Cell4407–4415.cells. Sci. 120, J. strongly impair SNARE-complex binding support doi:10.1038/emboj.2012.72 normal synaptic transmission. EMBO J. 31, 2156–2168. shared genetic factors between asthma and obesity in children. J. Allergy Clin. Immunol. 126, 631–637.e1–8. doi:10.1016/j.jaci.2010.06.030 D., Chiron, C., Ville, D., Charles, P., LeGuern,ST2011.E.,Charles,Depienne,Chiron, C.,C., Ville,D., P., Héron,D., D., presenting as infantile spasms and generalized tremor in three patients. Epilepsia 52, 1820–1827. doi:10.1111/ j.1528-1167.2011.03163.x Correia, A., Cacciagli, P., Mignon-Ravix, C., Cuberos, H., Chabrol, B., Villard, L., 2011. Epileptic and nonepilepticEpilepticandL.,Villard,2011. Mignon-Ravix, B.,Chabrol, H.,Cuberos,Cacciagli, C.,A.,Correia, P., wen, J. E G alintan, anolio, aren, arshall, artin,J. c c echan, edine, eijer, elén, ignot, endel, ilh, aroteaux, aroteaux, ipke, oos, oos, oos, uhmann, L L L L R M M M M M M M M M M M M M M M features in patients with early onset epileptic encephalopathy and STXBP1 mutations. Epilepsia 52, 1828–1834. doi:10.1111/j.1528-1167.2011.03181.x Millstein, R.A., Holmes, A., 2007. Effects of repeated maternal separation on anxiety- and depression-related phenotypes in different mouse strains. Neurosci. Biobehav. Rev. 31, 3–17. doi:10.1016/j.neubiorev.2006.05.003 Minelli, A., Scassellati, C., Cloninger, C.R., Tessari, E., Bortolomasi, M., Bonvicini, C., Giacopuzzi, M., Frisoni, G.B., Gennarelli, M., 2012. PCLO gene: its role in vulnerability to major depressive disorder. J. Affect. Disord. 139, 250–255. doi:10.1016/j.jad.2012.01.028 Misane, I., Kruis, A., Pieneman, A.W., Ögren, S.O., Stiedl, O., 2013. GABA(A) receptor activation in the CA1 area of the dorsal hippocampus impairs consolidation of conditioned contextual fear in C57BL/6J mice. Behav. Brain Res. 238, 160–169. doi:10.1016/j.bbr.2012.10.027 Misane, I., Tovote, P., Meyer, M., Spiess, J., Ogren, S.O., Stiedl, O., 2005. Time-dependent involvement of the dorsal hippocampus in trace fear conditioning in mice. Hippocampus 15, 418–426. doi:10.1002/hipo.20067 Miskey, C., Izsvák, Z., Kawakami, K., Ivics, Z., 2005. DNA transposons in vertebrate functional genomics. Cell. Mol. Life Sci. CMLS 62, 629–641. doi:10.1007/s00018-004-4232-7 Mitchell, M.S., 1976. Biology of the Mouse Histocompatibility-2 Complex. Principles of Immunogenetics Applied to a Single System. Yale J. Biol. Med. 49, 303–304. Miyakawa, T., Yared, E., Pak, J.H., Huang, F.L., Huang, K.-P., Crawley, J.N., 2001. Neurogranin null mutant mice display performance deficits on spatial learning tasks with anxiety related components. Hippocampus 11, 763– 775. doi:10.1002/hipo.1092 Moretti, P., Bouwknecht, J.A., Teague, R., Paylor, R., Zoghbi, H.Y., 2005. Abnormalities of social interactions and home- cage behavior in a mouse model of Rett syndrome. Hum. Mol. Genet. 14, 205–220. doi:10.1093/hmg/ddi016 Morton, A.J., Wood, N.I., Hastings, M.H., Hurelbrink, C., Barker, R.A., Maywood, E.S., 2005. Disintegration of the sleep-wake cycle and circadian timing in Huntington’s disease. J. Neurosci. Off. J. Soc. Neurosci. 25, 157–163. doi:10.1523/JNEUROSCI.3842-04.2005 Moseley, A.E., Williams, M.T., Schaefer, T.L., Bohanan, C.S., Neumann, J.C., Behbehani, M.M., Vorhees, C.V., Lingrel, J.B., 2007. Deficiency in Na,K-ATPase α isoform genes alters spatial learning, motor activity, and anxiety in mice. J. Neurosci. 27, 616–626. doi:10.1523/JNEUROSCI.4464-06.2007 Mouse Genome Sequencing Consortium, Waterston, R.H., et al., 2002. Initial sequencing and comparative analysis of the mouse genome. Nature 420, 520–562. doi:10.1038/nature01262 Moy, S.S., Nadler, J.J., Young, N.B., Perez, A., Holloway, L.P., Barbaro, R.P., Barbaro, J.R., Wilson, L.M., Threadgill, D.W., Lauder, J.M., Magnuson, T.R., Crawley, J.N., 2007. Mouse behavioral tasks relevant to autism: phenotypes of 10 inbred strains. Behav. Brain Res. 176, 4–20. doi:10.1016/j.bbr.2006.07.030 Muglia, P., Tozzi, F., Galwey, N.W., Francks, C., Upmanyu, R., Kong, X.Q., Antoniades, A., Domenici, E., Perry, J., Rothen, S., Vandeleur, C.L., Mooser, V., Waeber, G., Vollenweider, P., Preisig, M., Lucae, S., Müller-Myhsok, B., Holsboer, F., Middleton, L.T., Roses, A.D., 2010. Genome-wide association study of recurrent major depressive disorder in two European case-control cohorts. Mol. Psychiatry 15, 589–601. doi:10.1038/mp.2008.131 Mukherjee, K., Yang, X., Gerber, S.H., Kwon, H.-B., Ho, A., Castillo, P.E., Liu, X., Südhof, T.C., 2010. Piccolo and bassoon maintain synaptic vesicle clustering without directly participating in vesicle exocytosis. Proc. Natl. Acad. Sci. U. S. A. 107, 6504–6509. doi:10.1073/pnas.1002307107 Munafò, M.R., Freimer, N.B., Ng, W., Ophoff, R., Veijola, J., Miettunen, J., Järvelin, M.-R., Taanila, A., Flint, J., 2009. 5-HTTLPR genotype and anxiety-related personality traits: a meta-analysis and new data. Am. J. Med. Genet. Part B Neuropsychiatr. Genet. Off. Publ. Int. Soc. Psychiatr. Genet. 150B, 271–281. doi:10.1002/ajmg.b.30808 Nagy, A., 2000. Cre recombinase: the universal reagent for genome tailoring. Genes. N. Y. N 2000 26, 99–109. Niwa, N., Nerbonne, J.M., 2010. Molecular determinants of cardiac transient outward potassium current (I(to)) expression and regulation. J. Mol. Cell. Cardiol. 48, 12–25. doi:10.1016/j.yjmcc.2009.07.013 Nolan, P.M., Peters, J., Strivens, M., Rogers, D., Hagan, J., Spurr, N., Gray, I.C., Vizor, L., Brooker, D., Whitehill, E., Washbourne, R., Hough, T., Greenaway, S., Hewitt, M., Liu, X., McCormack, S., Pickford, K., Selley, R., Wells, C., Tymowska-Lalanne, Z., Roby, P., Glenister, P., Thornton, C., Thaung, C., Stevenson, J.A., Arkell, R., Mburu, P., Hardisty, R., Kiernan, A., Erven, A., Steel, K.P., Voegeling, S., Guenet, J.L., Nickols, C., Sadri, R., Nasse, M., Isaacs, A., Davies, K., Browne, M., Fisher, E.M., Martin, J., Rastan, S., Brown, S.D., Hunter, J., 2000. A systematic, genome-wide, phenotype- driven mutagenesis programme for gene function studies in the mouse. Nat. Genet. 25, 440–443. doi:10.1038/78140

180 References & 181 BP1 mutationsBP1 causeonlynotOhtahara syndromealsobutsyndrome-resultWest X de Geus, E.J.C., Hoekstra, R.A., Bartels, M., van Leeuwen, M., Verhulst, F.C., Posthuma, D., ., 2002. The 5-choice serial reaction time task: behavioural pharmacology and functional C Woolfe, A., Vassias, I., Pellentz, C., Lacoste, N., Puri, A., Schultz, D.C., Pchelintsev,N.A., Schultz,A.,Pellentz,Puri,D.C.,Adams,P.D., Lacoste,I.,C.,N., Vassias, A., Woolfe, Middei, S., Daniele, S., Adriani, W., Ghirardi, O., Caprioli, A., Laviola, G., 2005. Aspects of spatial Mattheisen, M., Frank, J., Treutlein, J., Degenhardt, F., Breuer, R., Steffens, M., Mier, D., Esslinger, Esslinger, C., D., R.,Steffens, Mier, M., Breuer, Degenhardt, J., F., Mattheisen,Treutlein, Frank,J., M., Brown, R.E., 2002. Strain differences in activity and emotionality do not account for differences in Savoie, Brown,V., R.E., 2010. Learning, memory and search strategies of inbred mouse strains with .J. Bezzina, C.R., de Vries, J.F., Wilde, A.A., Moorman, A.F., Mannens, M.M., 2000. Genomic organisationGenomicM.M.,Mannens, 2000.A.A., Wilde,A.F.,Moorman, J.F., Vries, C.R.,Bezzina,de ., Spink,A.J.,Tegelenbosch, R.A., EthoVision:versatile2001.automationtrackingvideoafor ofsystem ., T ., Oguni,H., Liang, J.-S., Ikeda, H., Imai, K., Hirasawa, K., Imai, K., Tachikawa, E., Shimojima, K., Osawa, Chui, D., Marth, J.D., 1992. Tissue- and site-specificTissue- recombination Natl. and1992. Proc. transgenic mice. DNA Marth,in J.D., D., Chui, Derr, Derr, R., Shendure, J., Kolata, T., Silver, L.M., 1998. A major influence of sex-specific loci on alcohol ., ., Jimenez, C.R., 2012. An accurate paired sample test for count data. Bioinforma. Oxf. Engl. 28, i596– Piersma, S.R., Warmoes, M., Jimenez, C.R., 2010. On the beta-binomial model for analysis of spectral D Hsu,S.C.,R.H.,Scheller,neural-specificn-Sec1: a 1994. syntaxin-binding protein. Proc. Natl. Acad. Sci. E .W., .W., ., 2008. Simultaneous behavioral characterizations: Embracing complexity. Proc. Natl. Acad. Sci. 105, P ., V ., Chopin, P., File, S.E., Briley, M., 1985. Validation of open:closedValidationelevatedM.,S.E.,ofFile,1985.a entriesan plus-mazeBriley, armas Chopin,in P., Duma, C., Goedert, M., Jucker, M., 2008. The value of incomplete mouse models of Alzheimer’sdisease.incompleteofmodels mouseof value The 2008. M.,Goedert, C.,Duma, Jucker, M., ., M T P . . ., ., . Riedel, G., 2011. The cholinergic system, EEG and sleep. Behav. Brain Res. 221, 499–504. doi:10.1016/j. C L ., T Itoh, K., Fushiki, S., 2013. Social behavior is perturbed in mice after exposure to bisphenol A: a novel ., ., . V M V A L . . S R P R Südhof, T.C., 2002. Snares and Munc18 in synaptic vesicle fusion. Nat. Rev. Neurosci. 3, 641–653. doi:10.1038/nrn898 641–653. 3, Neurosci. Rev. Nat. fusion. vesicle synaptic in Munc18 and Snares 2002. T.C., Südhof, T ., T allet, H G behavioralexperiments. Res.Behav.Methods Instrum. Comput. Psychon. J. Soc. 398–414.33,Inc different visual abilities the Barnesin Brain Res.maze. Behav. doi:10.1016/j.bbr.2010.08.030 assessmentemploying IntelliCage.an 223–228.3,Brain Behav. doi:10.1002/brb3.130 memory and behavioral disinhibition in Tg2576 transgenic mice as a model of Alzheimer’s disease. Behav. BrainAlzheimer’s Behav.disease.of model a as transgenicmicebehavioral disinhibitionmemoryTg2576 and in 225–232.Res.156, doi:10.1016/j.bbr.2004.05.028 Acad. Sci. U. S. A. 89, 6861–6865. Acad.A.S.Sci.89, U. M., Yamamoto, T., 2010. ST 2010. T., M.,Yamamoto, Japaneseof cohortEpilepsia study. doi:10.1111/j.1528-1167.2010.02767.x 51, 2449–2452. 20563doi:10.1073/pnas.0811546106–20564. preference in C57Bl/6 and DBA/2andpreferenceC57Bl/6 inbredinmice. Mamm. GenomeInt.942–948.J. Off. 9, Mamm. Genome Soc. measureanxietyNeurosci.ofrat.the in J. Methods 149–167. 14, U. S. A.1445–1449. S.91,U. i602.doi:10.1093/bioinformatics/bts394 count data in label-free tandem mass spectrometry-based proteomics. Bioinforma. Oxf. Engl. doi:10.1093/bioinformatics/btp67726, 363–369. bbr.2011.01.017 learningmemoryand performance betweenDBA/2and C57BL/6 mice.Genes96–110. 1,Brain Behav. Boomsma, D.I., 2009. Attention problems, inhibitory control, and18-year-oldandfactors: 12-,study 9-, ina twins.Neuropsychology intelligence doi:10.1037/a0014915 23,381–391. index overlapping genetic and chromosomal localisation of two members of the KCND ion channel family, Genet.KCND2 614–619. 106, and KCND3. Hum. Eur. J. Nucl. Med. Mol. Imaging 35 Suppl 1, S70–74. doi:10.1007/s00259-007-0704-y Nucl.Med. J. Mol. Imaging S70–74. SupplEur. 351, Jansen, L.E.T., Almouzni, G., 2011. Dynamics of histone H3 deposition in vivo reveal a nucleosome gap-filling mechanismmaintainH3.3forto chromatin integrity.Mol.Cell928–941. doi:10.1016/j.molcel.2011.12.00644, Walter, H., Kirsch, P., Erk, S., Schnell, K., Herms, S., Wichmann, H.-E., Schreiber, S., Jöckel, K.-H., Strohmaier, J., J., K.-H.,Jöckel,Strohmaier, S., Wichmann, H.-E.,Schreiber,S., Herms,Schnell,K., S., Erk, P., Kirsch,H., Walter, Roeske,Haenisch, D., B.,Gross, M., Hoefels, Schmäl,Schulze, S.,Lucae, C., T.G., E.B.,S., Binder, T.F., Wienker, Meyer-Lindenberg,Bettecken, B.,Brors,T., Nöthen,A.,A.,Müller-Myhsok,Juraeva, D., Zimmer,W., Maier, B., M.M.,Genome-wideCichon,2010. S.,association-, replication-,neuroimaging andimplicatesstudy HOMER1 etiologythein majorofdepression. Biol.Psychiatry 578–585.68,doi:10.1016/j.biopsych.2010.05.038 neurochemistry.Psychopharmacology (Berl.)doi:10.1007/s00213-002-1154-7 163,362–380. eary, L tsuka, gi, gnibene, rban, ’ oldus, adde, ay- ietschel, obbins, izo, J., J., izo, aylor, aylor, eirce, J. ellow, ham, evsner, J., evsner, ham, latt, B., odhorna, J., olderman, ostma, N O O O O O P P P P P P P P P P R R R R R Roberts, A., Pardo-Manuel de Villena, F., Wang, W., McMillan, L., Threadgill, D.W., 2007. The polymorphism architecture of mouse genetic resources elucidated using genome-wide resequencing data: implications for QTL discovery and systems genetics. Mamm. Genome Off. J. Int. Mamm. Genome Soc. 18, 473–481. doi:10.1007/s00335-007-9045-1 Rogers, D.C., Fisher, E.M., Brown, S.D., Peters, J., Hunter, A.J., Martin, J.E., 1997. Behavioral and functional analysis of mouse phenotype: SHIRPA, a proposed protocol for comprehensive phenotype assessment. Mamm. Genome Off. J. Int. Mamm. Genome Soc. 8, 711–713. Rowe, J., Corradi, N., Malosio, M.L., Taverna, E., Halban, P., Meldolesi, J., Rosa, P., 1999. Blockade of membrane transport and disassembly of the Golgi complex by expression of syntaxin 1A in neurosecretion-incompetent cells: prevention by rbSEC1. J. Cell Sci. 112, 1865–1877. Roy, V., Chapillon, P., Jeljeli, M., Caston, J., Belzung, C., 2009. Free versus forced exposure to an elevated plus- maze: evidence for new behavioral interpretations during test and retest. Psychopharmacology (Berl.) 203, 131–141. doi:10.1007/s00213-008-1378-2 Saitsu, H., Kato, M., Mizuguchi, T., Hamada, K., Osaka, H., Tohyama, J., Uruno, K., Kumada, S., Nishiyama, K., Nishimura, A., Okada, I., Yoshimura, Y., Hirai, S., Kumada, T., Hayasaka, K., Fukuda, A., Ogata, K., Matsumoto, N., 2008. De novo mutations in the gene encoding STXBP1 (MUNC18-1) cause early infantile epileptic encephalopathy. Nat. Genet. 40, 782–788. doi:10.1038/ng.150 Saitsu, H., Kato, M., Okada, I., Orii, K.E., Higuchi, T., Hoshino, H., Kubota, M., Arai, H., Tagawa, T., Kimura, S., Sudo, A., Miyama, S., Takami, Y., Watanabe, T., Nishimura, A., Nishiyama, K., Miyake, N., Wada, T., Osaka, H., Kondo, N., Hayasaka, K., Matsumoto, N., 2010. STXBP1 mutations in early infantile epileptic encephalopathy with suppression-burst pattern. Epilepsia 51, 2397–2405. doi:10.1111/j.1528-1167.2010.02728.x Santarelli, L., Saxe, M., Gross, C., Surget, A., Battaglia, F., Dulawa, S., Weisstaub, N., Lee, J., Duman, R., Arancio, O., Belzung, C., Hen, R., 2003. Requirement of hippocampal neurogenesis for the behavioral effects of antidepressants. Science 301, 805–809. doi:10.1126/science.1083328 Sauer, B., Henderson, N., 1988. Site-specific DNA recombination in mammalian cells by the Cre recombinase of bacteriophage P1. Proc. Natl. Acad. Sci. U. S. A. 85, 5166–5170. Schinka, J.A., 2005. Measurement scale does moderate the association between the serotonin transporter gene and trait anxiety: comments on Munafo et al. Mol. Psychiatry 10, 892–893; author reply 895–897. doi:10.1038/ sj.mp.4001708 Schlager, G., Dickie, M.M., 1967. Spontaneous Mutations and Mutation Rates in the House Mouse. Genetics 57, 319–330. Schmitz, S.K., Hjorth, J.J.J., Joemai, R.M.S., Wijntjes, R., Eijgenraam, S., de Bruijn, P., Georgiou, C., de Jong, A.P.H., van Ooyen, A., Verhage, M., Cornelisse, L.N., Toonen, R.F., Veldkamp, W.J.H., Veldkamp, W., 2011. Automated analysis of neuronal morphology, synapse number and synaptic recruitment. J. Neurosci. Methods 195, 185– 193. doi:10.1016/j.jneumeth.2010.12.011 Schneggenburger, R., Meyer, A.C., Neher, E., 1999. Released fraction and total size of a pool of immediately available transmitter quanta at a calyx synapse. Neuron 23, 399–409. Schoch, S., Gundelfinger, E.D., 2006. Molecular organization of the presynaptic active zone. Cell Tissue Res. 326, 379–391. doi:10.1007/s00441-006-0244-y Shi, J., Potash, J.B., Knowles, J.A., Weissman, M.M., Coryell, W., Scheftner, W.A., Lawson, W.B., DePaulo, J.R., Jr, Gejman, P.V., Sanders, A.R., Johnson, J.K., Adams, P., Chaudhury, S., Jancic, D., Evgrafov, O., Zvinyatskovskiy, A., Ertman, N., Gladis, M., Neimanas, K., Goodell, M., Hale, N., Ney, N., Verma, R., Mirel, D., Holmans, P., Levinson, D.F., 2011. Genome-wide association study of recurrent early-onset major depressive disorder. Mol. Psychiatry 16, 193–201. doi:10.1038/mp.2009.124 Shyn, S.I., Shi, J., Kraft, J.B., Potash, J.B., Knowles, J.A., Weissman, M.M., Garriock, H.A., Yokoyama, J.S., McGrath, P.J., Peters, E.J., Scheftner, W.A., Coryell, W., Lawson, W.B., Jancic, D., Gejman, P.V., Sanders, A.R., Holmans, P., Slager, S.L., Levinson, D.F., Hamilton, S.P., 2011. Novel loci for major depression identified by genome- wide association study of Sequenced Treatment Alternatives to Relieve Depression and meta-analysis of three studies. Mol. Psychiatry 16, 202–215. doi:10.1038/mp.2009.125 Siddiqui, T.J., Craig, A.M., 2011. Synaptic organizing complexes. Curr. Opin. Neurobiol. 21, 132–143. doi:10.1016/j. conb.2010.08.016 Simon, P., Dupuis, R., Costentin, J., 1994. Thigmotaxis as an index of anxiety in mice. Influence of dopaminergic transmissions. Behav. Brain Res. 61, 59–64.

182 References & 183 Wierda, K., Sons, M.S., de Wit, H., Cornelisse, L.N., Brussaard, A., Plomp, J.J., Verhage, M., 2006. Verhage, M., 2007. Munc18-1 in secretion: lonely Munc joins SNARE team and takes control. Trends control.Trends takesand team SNAREsecretion:joins Munclonelyin Munc18-1 2007. M.,Verhage, 2004. The synaptic vesicle cycle. Annu. Rev. Neurosci. 27, 509–547. doi:10.1146/annurev. Neale, M.C., Kendler, K.S., 2000. Genetic epidemiology of major depression: review and meta- ., ., Jackson, W.S., King, O.D., Lindquist, S., 2007. The power of automated high-resolution behavior de de Geus, G., E.J.C., James, Willemsen, M.R., Blackwood, Smit, Arolt, Zandbelt, J.H., T., Baune, V., B.T., Nestler, E.J., 2004. Neurobehavioral assessment in the information age. Nat. 462–466.Neurosci. 7, 2012.presynapticThe doi:10.1016/j.neuron.2012.06.012active11–25. zone.Neuron75, Soumier, A., Brewer, M., Pickel, J., Cameron, H.A., 2011. Adult hippocampal neurogenesis buffers Rothman, J.E., 2009. Membrane fusion: grappling with SNARE and SM proteins. Science 323, 474–477. Science474–477. proteins.323,SM andMembraneSNAREfusion:grappling with 2009. Rothman,J.E., Read, A.P., 1999. HumanMolecular Read,1999. A.P.,Genetics, ed.2ndWiley-Liss, NewYork. G Eppig, 2012.J.T., The Mammalian Phenotype Ontology as a unifying standard for experimental and ., G ., Izsvák, Z., Ivics, Z., 2008. Insertional mutagenesis of the mouse germline with Sleeping Beauty . Tegelenbosch,R.A., Buma,Noldus, M.O., L.P., 2001.The EthoVision video tracking system--a toolfor . ., ., ., ., ., ., ., F ., Tanti, A., Leonardo, E.D., Laugeray, A., Rainer, Q., Touma, C., Palme, R., Griebel, G., Ibarguen-Vargas, Y., R.,C.,Palme,Griebel,Ibarguen-Vargas,Y., G.,Touma, Q., A.,Rainer, Laugeray,E.D.,A., Leonardo, Tanti, van Baal, G.C.M., Polderman, T.J.C., Verhulst, F.C., Boomsma, D.I., 2004. Heritability of Stroop and C F F ., Radulovic, J., Lohmann, R., Birkenfeld, K., Palve, M., Kammermeier, J., Sananbenesi, Spiess, F., J., 1999. F . . Schürhoff, F., Schürhoff, F., F., Meary, Mathieu, Testsexecutive Ionescu,of A., Leboyer, 2005. functionsM., S., first- in . . T C D S C . Ogiwara, I., Kaneda, M., Tokonami, N., Mazaki, E., Baba, K., Matsuda, K., Inoue, Y., Yamakawa, K., 2006. A K.,2006.Yamakawa,Mazaki, N., Baba,E.,K.,Kaneda, Ogiwara,Matsuda, Tokonami,M.,I., K., Y., Inoue, H . ., P . L . T Orchard,S.M., Sanford,L.D., 2002. activitycageHomebehavioral and performance hybridinbredandin Puri,A., Ricketts, M.D., Rai, Hoffmann, T.S., J., Hoi, E., Adams,P.D., Schultz, D.C., Marmorstein, R., 2012. . ., R P R ., . ., .J., T T F L A A ., ., C A O A Y X Kv4.2truncation mutation patientain withtemporal lobe epilepsy.Neurobiol. 245–253.doi:10.1016/j.Dis. 24, nbd.2006.07.001 high-throughputphenotyping data.Mamm. Genome 23,653–668. doi:10.1007/s00335-012-9421-3 stressresponses depressiveand 458–461. behaviour. Naturedoi:10.1038/nature10287476, behavioralphenotyping transgenicof mice.Physiol. 73,731–744. Behav. analysisrevealedapplicationits by mouse prionmodelsandHuntington’sdiseases. toof Proc. Natl.Acad. Sci. A.doi:10.1073/pnas.0610779104S.1983–1988.104,U. Strain and substrain differences in context- andtone-dependent fear conditioningBrainRes.1–12. 104, of inbred mice. Behav. flanker performance doi:10.1186/1471-2202-5-49 12-year in children.old 49. Neurosci.BMC 5, neuro.26.041002.131412 doi:10.1126/science.1161748 analysis.Psychiatry1552–1562.Am. J. 157, Hen, R., Belzung, C., 2011. Antidepressants recruit new neurons to improve stress response regulation. Mol. Psychiatry1177–1188.doi:10.1038/mp.2011.48 16, mice.Brain555–569. Res.Behav.136, Holsboer, F., Hoogendijk, W.J., Hottenga, J.J., Hu, Y., Kohli, M., Lin, D., Lucae, S., Macintyre, D.J., Maier, W., McGhee, McGhee, W., Maier, D.J., Macintyre, P., S., Heutink, A.C., Heath, Q., Lucae, He, S.D., M., D., Gordon, A., Fava, Lin, K., Farmer, M., W.L., Domschke, Coventry, S., Kohli, Cichon, D., Y., Hu, J.J., Hottenga, W.J., Hoogendijk, F., Holsboer, D., Posthuma, K., Pirlo, R.H., Perlis, M.M., Nöthen, W.A., Nolen, W.J., Muir, G.W., Montgomery, P., F.G., K.A.,McGuffin, Zitman, M., R., Verhage, Dyck, van J.-Y., Tzeng, J.W., Smoller, A.B., A., Smit, Schosser, P., Rizzu, M., Rietschel, depressive major for association Genome-wide 2009. B.W.J.H., doi:10.1038/mp.2008.125 Penninx, D.I., Boomsma, N.R., 359–375. Wray, 14, N.G., Martin, Psychiatry Mol. piccolo. protein presynaptic the for role possible a disorder: degreerelatives schizophrenicof patients: meta-analysis.a Psychol. Med.771–782.35, transposition.Methods Mol.Biol. Cliftondoi:10.1007/978-1-59745-232-8_8 109–125.435, NJ Identification of an ubinuclein region1 required for stability and function of the human HIRA/UBN1/CABIN1/ ASF1ahistoneH3.3 chaperone complex. Biochemistry (Mosc.) 51,2366–2377.doi:10.1021/bi300050b doi:10.1038/nn1225 Neurosci.564–572. 30, doi:10.1016/j.tins.2007.08.008 Munc18-1 expression levels control synapse recovery by regulating readily releasable Acad.doi:10.1073/pnas.0608507103A.S.Sci.18332–18337.103, U.pool size. Proc. Natl. ingh, B.,ingh, mith, teele, ecott, oonen, nyder, nyder, J. tiedl, oonen, pink, tins, J. trachan, ang, akeda, J., ang, zöke, üdhof, üdhof, üdhof, urget, ullivan, ullivan, ullivan, S S S S S S S S S S S S S S T S T T T T T Treit, D., Fundytus, M., 1988. Thigmotaxis as a test for anxiolytic activity in rats. Pharmacol. Biochem. Behav. 31, 959–962. Turner, K.M., Burne, T.H.J., 2013. Interaction of genotype and environment: effect of strain and housing conditions on cognitive behavior in rodent models of schizophrenia. Front. Behav. Neurosci. 7, 97. doi:10.3389/ fnbeh.2013.00097 Turri, M.G., Datta, S.R., DeFries, J., Henderson, N.D., Flint, J., 2001. QTL analysis identifies multiple behavioral dimensions in ethological tests of anxiety in laboratory mice. Curr. Biol. CB 11, 725–734. Van Galen, E.J.M., Ramakers, G.J.A., 2005. Rho proteins, mental retardation and the neurobiological basis of intelligence. Prog. Brain Res. 147, 295–317. doi:10.1016/S0079-6123(04)47022-8 Vazquez, B., Devinsky, O., 2003. Epilepsy and anxiety. Epilepsy Behav. EB 4 Suppl 4, S20–25. Verhage, M., Maia, A.S., Plomp, J.J., Brussaard, A.B., Heeroma, J.H., Vermeer, H., Toonen, R.F., Hammer, R.E., van den Berg, T.K., Missler, M., Geuze, H.J., Südhof, T.C., 2000. Synaptic assembly of the brain in the absence of neurotransmitter secretion. Science 287, 864–869. Verhoeven, K.J.F., Simonsen, K.L., McIntyre, L.M., 2005. Implementing false discovery rate control: increasing your power. Oikos 108, 643–647. doi:10.1111/j.0030-1299.2005.13727.x Verweij, K.J.H., Zietsch, B.P., Medland, S.E., Gordon, S.D., Benyamin, B., Nyholt, D.R., McEvoy, B.P., Sullivan, P.F., Heath, A.C., Madden, P.A.F., Henders, A.K., Montgomery, G.W., Martin, N.G., Wray, N.R., 2010. A genome- wide association study of Cloninger’s temperament scales: implications for the evolutionary genetics of personality. Biol. Psychol. 85, 306–317. doi:10.1016/j.biopsycho.2010.07.018 Voikar, V., Colacicco, G., Gruber, O., Vannoni, E., Lipp, H.-P., Wolfer, D.P., 2010. Conditioned response suppression in the IntelliCage: assessment of mouse strain differences and effects of hippocampal and striatal lesions on acquisition and retention of memory. Behav. Brain Res. 213, 304–312. doi:10.1016/j.bbr.2010.05.019 Wahlsten, D., Bachmanov, A., Finn, D.A., Crabbe, J.C., 2006. Stability of inbred mouse strain differences in behavior and brain size between laboratories and across decades. Proc. Natl. Acad. Sci. U. S. A. 103, 16364– 16369. doi:10.1073/pnas.0605342103 Wahlsten, D., Cooper, S.F., Crabbe, J.C., 2005. Different rankings of inbred mouse strains on the Morris maze and a refined 4-arm water escape task. Behav. Brain Res. 165, 36–51. doi:10.1016/j.bbr.2005.06.047 Wahlsten, D., Metten, P., Phillips, T.J., Boehm, S.L., Burkhart-Kasch, S., Dorow, J., Doerksen, S., Downing, C., Fogarty, J., Rodd-Henricks, K., Hen, R., McKinnon, C.S., Merrill, C.M., Nolte, C., Schalomon, M., Schlumbohm, J.P., Sibert, J.R., Wenger, C.D., Dudek, B.C., Crabbe, J.C., 2003. Different data from different labs: lessons from studies of gene-environment interaction. J. Neurobiol. 54, 283–311. doi:10.1002/neu.10173 Waites, C.L., Garner, C.C., 2011. Presynaptic function in health and disease. Trends Neurosci. 34, 326–337. doi:10.1016/j.tins.2011.03.004 Waites, C.L., Leal-Ortiz, S.A., Andlauer, T.F.M., Sigrist, S.J., Garner, C.C., 2011. Piccolo regulates the dynamic assembly of presynaptic F-actin. J. Neurosci. Off. J. Soc. Neurosci. 31, 14250–14263. doi:10.1523/JNEUROSCI.1835-11.2011 Waites, C.L., Leal-Ortiz, S.A., Okerlund, N., Dalke, H., Fejtova, A., Altrock, W.D., Gundelfinger, E.D., Garner, C.C., 2013. Bassoon and Piccolo maintain synapse integrity by regulating protein ubiquitination and degradation. EMBO J. 32, 954–969. doi:10.1038/emboj.2013.27 Wang, J., Fields, J., Doré, S., 2008. The development of an improved preclinical mouse model of intracerebral hemorrhage using double infusion of autologous whole blood. Brain Res. 1222, 214–221. doi:10.1016/j. brainres.2008.05.058 Wang, X., Hu, B., Zieba, A., Neumann, N.G., Kasper-Sonnenberg, M., Honsbein, A., Hultqvist, G., Conze, T., Witt, W., Limbach, C., Geitmann, M., Danielson, H., Kolarow, R., Niemann, G., Lessmann, V., Kilimann, M.W., 2009. A protein interaction node at the neurotransmitter release site: domains of Aczonin/Piccolo, Bassoon, CAST, and rim converge on the N-terminal domain of Munc13-1. J. Neurosci. Off. J. Soc. Neurosci. 29, 12584–12596. doi:10.1523/JNEUROSCI.1255-09.2009 Willcutt, E.G., Doyle, A.E., Nigg, J.T., Faraone, S.V., Pennington, B.F., 2005. Validity of the executive function theory of attention-deficit/hyperactivity disorder: a meta-analytic review. Biol. Psychiatry 57, 1336–1346. doi:10.1016/j.biopsych.2005.02.006 Wills, T.A., Vaccaro, D., McNamara, G., 1994. Novelty seeking, risk taking, and related constructs as predictors of adolescent substance use: an application of Cloninger’s theory. J. Subst. Abuse 6, 1–20. Wittchen, H.-U., 2012. The burden of mood disorders. Science 338, 15. doi:10.1126/science.1230817

184 References & 185 .2010.00591.x X .2005.00173.x X mouse. Genes Brain Behav. 3, 337–359. doi:10.1111/j.1601- X Takahashi, Y., Kijima, N., Maekawa, H., Ono, Y., Ando, J., 2005. Genetic and environmentaletiologyGeneticofand 2005. J., Ando, Y., Ono, Maekawa,H.,N., Kijima, Y., Takahashi, Brown, R.E., 2006. Visual detection, pattern discrimination and visual acuity in 14 strains of mice. 2002. Behavioral phenotyping enhanced--beyond (environmental) standardization. Genes Brain Crusio, W.E., Lipp, H.P., 2002. Knockout mice: simple solutions to the problems of genetic backgroundgenetic of problemsthe tosolutions simplemice:Knockout 2002. H.P., Lipp, W.E.,Crusio, ., 1986. Longitudinal data analysis using generalized linear models. Biometrika 73, 13 –22. doi:10.1093/ Pergadia,M.L., Blackwood, D.H.R., Penninx, B.W.J.H.,Gordon,Nyholt, S.D., D.R.,Ripke,MacIntyre, S., James, M.R., Gordon, S.D., Dumenil, T., Ryan, L., Coventry, W.L., Statham, D.J., Pergadia, M.L.,Pergadia,L.,W.L.,Coventry,Statham,D.J.,Ryan,Madden, Dumenil,T., M.R.,S.D.,James, Gordon, ., S Lin, Y.-W., Min, M.-Y., Chen, C.-C.,Min, M.-Y., MiceLin,2010.Y.-W., lacking Asic3 showreduced anxiety-like behavior theon ., Asafo-Adjei, P.K., Arnold, H.M., Brown, R.E., Bauchwitz, R.P., 2004. A phenotypic and molecular Vardinon-Friedman, H., Vardinon-Friedman, Cases-Langhoff, C.,B., E.D., N.E., Becker, Ziv, Gundelfinger, C.C.,Garner, 2001. ., ., ., P ., A Ellenbroek, B.A., van Eck, I., Roubos, S., Verhage, M., Stiedl, O., 2012. Finding the right motivation: ., . H . ., ., .2004.00087.x L R R . L . . D X G A S . N N R and flankingand Trends genes.Neurosci. 336–340. 25, Genes Brain Behav. 5, 389–403.Genesdoi:10.1111/j.1601-183 5, BrainBehav. P.A.F., Heath, A.C., Montgomery, G.W., Martin, N.G., 2009. Accurate,Large-Scale5HTTLPRGenotyping of2009. andMartin, Heath,A.C.,N.G.,Montgomery, G.W.,P.A.F., Flanking Single Nucleotide Polymorphisms in an Association Study of Depression, Measures.Anxiety, Biol.Psychiatry anddoi:10.1016/j.biopsych.2009.04.030468–476. 66, Personality D.J., McGhee, K.A., Maclean, A.W., Smit, J.H., Hottenga, J.J., Willemsen, G., Middeldorp,E.J.C.,McGhee,Geus,Willemsen,C.M.,K.A.,G.,Hottenga,J.J.,D.J., de J.H., Smit, Maclean, A.W., Lewis, C.M., McGuffin, P., Hickie, I.B., van den Oord, E.J.C.G., Liu, J.Z.,E.M., Medland, Macgregor, S.E.,S., Statham, McEvoy, B.P.,D.J., Henders,Byrne, A.K., Heath, A.C., Montgomery, G.W., Martin, N.G., Boomsma, D.I., Madden, P.A.F., Sullivan, P.F., 2012. Genome-wide association study of major depressive results,disorder:meta-analysis, lessonsandnew learned. Mol.36–48.Psychiatrydoi:10.1038/mp.2010.109 17, elevated plus maze and reduced aggression. Genes Brain Behav. 9, 603–614. doi:10.1111/j.1601-183 603–614. 9, Behav. Brain Genesaggression. reduced and maze pluselevated Behav. 1, 3–8. 1, Behav. effortful Twin control.Res. Hum. Genet. Int. J. doi:10.1375/1832427054936790 300–306. Twin Off. Stud. Soc. 8, characterization of the fmr1-tm1Cgr 183 fragile genotype-dependentdifferences in effective reinforcements for spatial learning. Brain Behav. Res. 226, 397– doi:10.1016/j.bbr.2011.09.034 403. biomet/73.1.13 Assembling the presynaptic active zone: a characterization of an active one precursor vesicle. Neuron 29, 131–143. 29, Neuron vesicle. precursor one active an of characterization a zone: active presynaptic the Assembling amagata, an, Q.J., oun, J., Wolfer, Wolfer, Wong, Wray, Wray, Wray, Wray, Wu, W.- Wu, Würbel, Y Y Y Zeger, Zhai,

Abbrevations # 129S1: 129S1/SvImJ 5-CSRTT: 5-choice serial reaction time task A ADHD: Attention deficit hyperactivity disorder ASR: Acoustic startle response B BALB: BALB/C BDNF: Brain-derived neurotrophic factor BM: Barnes maze C C3H: C3H/HeJ C57: C57BL/6J CAZ: Cytomatrix of the active zone COG: center of gravity CRF: corticotropin-releasing factor CS: conditioned stimulus D DBA: DBA2/J DC: Dark compartment DLB: Dark-light box DMSO: Dimethyl sulfoxide DPP10: Dipeptidyl-peptidase 10 &

E ECG: Electro-cardiogram A EEG: Electro-encephalogram bbrevations EIEE: Early infantile epileptic encephalopathy EPM: Elevated plus maze EPSC: Excitatory postsynaptic current ES: Embryonic stem F FC: Fear conditioning FDR: False Discovery Rate FGF-13: Fibroblast growth factor 13 FST: Forced swim test FVB: FVB/N G GEE: generalized estimating equations GFP: green fluorescent protein Gprc5b: Retinoic acid-inducible orphan G protein–coupled receptor GSM: Grip strength meter GWAS: Genome-wide association study H HUCA: HIRA/UBN1/CABIN1/ASF1a complex HZ: Heterozygous I IR/DRs: Inverted repeat/direct repeat IRES: Internal ribosome entry site

187 K Kcnd2: Voltage-gated potassium (Kv) channels KI: Knock-in KO: Knock-out L LM-PCR: Linker mediated polymerase chain reaction M MDD: Major depressive disorder MWM: Morris water maze N NOD: NOD/LtJ O OF: Open field P pA: Polyadenylation PDK1: Phosphoinositide-dependent protein kinase-1 PPI: Pre-pulse inhibition R RFID: Radio frequency identification RIM: Rab3-interacting molecule RM ANOVA: Repeated measure ANOVA RRP: Readly releasable pool S SA: Splice acceptor SB: Sleeping beauty SD: Splice donor SD: Standard deviation SEM: Standard error of the mean SNP: Single nucleotide polymorphism STRAP: Serine-threonine kinase receptor-associated protein STXBP-1 : Syntaxin-binding protein 1 T TTC39: Tetratricopeptide repeat domain 39c U UBN-1: Ubinuclein 1 US: Unconditionned stimulus W WT: Wildtype

188 Samenvatting Gedragsfenotypering van complexe eigenschappen in inteeltlijnen en genetisch gemodificeerde muizen Neurologische en psychiatrische aandoeningen worden bij mensen over het algemeen herkend door verstoringen in het gedrag waar te nemen. Medisch gezien is het essentieel de onderliggende biologische mechanismen van deze gedragsstoornissen in kaart te brengen. De complexiteit van deze stoornissen en de onderliggende biologische mechanismen vereisen het gebruik van diermodellen en experimentele opstellingen die geschikt zijn om gedrag nauwkeurig te fenotyperen. De gebruikelijke gedragstesten zijn over het algemeen gebasseerd op motorische activiteit en zijn ontworpen om gedrag te meten gedurende een relatief korte periode (5 tot 60 minuten). Om het effect van een bepaalde genetische achtergrond, een specifieke genetische mutatie, of van chemische stoffen (medicatie, drugs) op gedrag te meten, is een gevarieerd scala aan gedragstesten nodig dat verschillende aspecten van gedrag representeert, zoals motorische activiteit, waarneming en zintuiglijke functies, cognitie, en circadiane functies. Het afnemen van zo’n testbatterij is tijdrovend en bovendien leiden herhaalde mens-dier interacties, omgevings factoren, en technische verschillen tussen laboratoria tot variatie in onderzoeksresultaten, met non-replicatie en verschillen in interpretatie van resultaten als gevolg. Er is dan ook sprake van groeiende consensus dat gedragsfenotypering geautomatiseerd zou moeten worden; dit zou de effectiviteit en de standaardisatie in het onderzoek met diermodellen naar menselijke ziektes en behandeling ten goede komen. & De afgelopen tien jaar zijn nieuwe technologieën ontwikkeld, die de mogelijkheid bieden S

om dieren geautomatiseerd en langdurig (meerdere dagen) te observeren in hun thuiskooi. amenvatting Deze manier van observeren maak het mogelijk om herhaaldelijk, objectief, en consistent resultaten te verkrijgen terwijl mens-dier interacties geminimaliseerd worden. Bovendien biedt langdurige observatie de mogelijkheid om multidimensionale aspecten van gedrag te bestuderen, zoals habituatie, basaal gedrag, en gedrag wanneer het dier op de proef gesteld wordt, wat de studie van verandering in, of progressie van, gedrag over tijd mogelijk maakt. Al deze gedragsaspecten kunnen van groot belang zijn voor de studie van neurologische, psychiatrische, en neuro-degeneratieve aandoeningen. In Hoofdstuk 2 beschrijven wij een studie waarin gebruik gemaakt wordt van een zogenaamd geautomatiseerd “high throughput” systeem; een systeem dat het gedrag van de muis nauwkeurig en frequent (15 keer per seconde) registreert. In deze studie richtten we ons op het karakteriseren van complexe gedragspatronen die in muizen indicatief zijn voor vermijdingsgedrag. We hebben een test ontwikkeld die gebaseerd is op de natuurlijke neiging van muizen om een voorkeur te ontwikkelen voor één van de twee ingangen van hun schuilplaats. Na de voorkeur vastgesteld te hebben, werd het gebruik van de voorkeursingang “bestraft” door de schuilplaats intern te verlichten als de muis deze had betreden via de voorkeursingang. Door middel van deze cognitie leertaak werd een grote hoeveelheid en verscheidenheid aan gedragsinformatie verkregen. Acht inteeltlijnen en 43 verschillende mutanten zijn aan deze leertaak onderworpen, en één nieuw kandidaat gen werd geïdentificeerd - specc1/cytospinB - die een rol lijkt te spelen in leren via vermijding (of wel: avoidance learning). Onze data laat zien dat complex gedrag van muizen, zelfs in grote cohorten van muizen en mutanten, efficiënt

189 en succesvol gedetecteerd, geanalyseerd, en gevisualiseerd kan worden. Diverse inteeltlijnen en genetische mutanten vertoonden duidelijke kwantitatieve afwijkingen in verschillende aspecten van deze gedragstaak. In Hoofdstuk 3 is onderzoek gedaan naar de mogelijke rol van Munc18-1 haplo-insufficiëntie bij muizen, als een model voor “early infantile epileptic encephalopathy (EIEE)”, ook wel het Othahara syndroom genoemd. Homozygote deletie van Munc18-1 resulteert in postnatale sterfte. Het MUNC18-1 eiwit speelt een belangrijke rol bij de secretie van synaptische blaasjes van zoogdieren. De novo hetrozygote (HZ) mutaties in het humane MUNC18-1 gen, STXBP1, zijn vermoedelijk betrokken bij ernstige cognitieve stoornissen waarbij ook epileptische aanvallen kunnen optreden. Door middel van een uitgebreide testbatterij en observatie in een geautomatiseerde thuiskooi, hebben we het gedrag van Munc18-1 haploinsufficiënte muizen onderzocht. Er werden geen duidelijke epileptische aanvallen of cognitieve beperkingen gevonden bij Munc18-1 HZ muizen. Echter, HZ muizen lieten een verminderde angstreactie (fear) en snellere uitdoving zien, ondanks hun eerder geobserveerde hogere angsttoestand (anxiety). Ook lieten MUNC18-1 HZ muizen onder stress een meer proactieve verwerkingsstrategie zien. De verlaagde hoeveelheid van het MUNC18-1 eiwit in muizen bleek echter nog voldoende te zijn om de meeste cognitieve functies te behouden. Genoom-wijde associatie studies (GWAS) suggereren dat een niet-synonieme variant in het exon van het pre-synaptische gen PCLO een rol speelt in depressie. In Hoofdstuk 4 ging de focus uit naar de effecten van een genetische variant van PCLO (S4814A) in de homogene C57BL/6J muizen op gedrags- en op moleculair en cellulair niveau. Het knock-in muismodel met een expressie van de PcloSA/SA variant liet een verhoogde synaptische Piccolo expressie zien, en had 30% hogere excitatoire synaptische transmissie in gekweekte neuronen. Echter, de angsttoestand (anxiety), cognitie, en depressie-gerelateerd gedrag van deze PcloSA/SA muizen bleek niet afwijkend. Het feit dat depressie een complexe, multifactoriële ziekte is, voortgebracht door een combinatie van vele genetische varianten en omgevingsfactoren, is mogelijk een reden waarom PcloSA/SA muizen geen sterk gedragsfenotype vertoonden. Toch sluiten we niet uit dat de moleculaire veranderingen die we geobserveerd hebben het risico op depressie onder bepaalde omstandigheden kan verhogen. Historisch gezien zijn de genen die tot nu toe het meest onderzocht zijn genen die in klinische observaties een dramatisch effect op het fenotype bleken te hebben. Omdat een groot aantal genen om deze reden nooit onderzocht werd, hebben we in Hoofdstuk 5 vijf random mutanten bestudeerd (Dpp10, Fgf13, Kcnd2, Ttc39c en Ubn1) die gecreëerd waren middels de Sleeping Beauty techniek. Deze 5 stammen werden gedragsmatig gefenotypeerd met behulp van nieuwe geautomatiseerde methodes waarin muizen in de thuiskooi leren middels vermijding (“avoidance learning”; zie Hoofdstuk 2). Deze methode, waarin gebruik wordt gemaakt van spontaan gedrag in de thuiskooi en de natuurlijke neiging van muizen om licht te vermijden, onthulde een hoge gevoeligheid van spontaan gedrag voor random genetische modificaties. Bovendien bleek Ubn1, een subunit van het HUCA histon chaperonne complex, geassocieerd te zijn aan de formatie van associatief geheugen. Deze studie van 5 willekeurige functionele genen, waarvan één betrokken bleek bij cognitieve processen, suggereert dat een aanzienlijke proportie van genen in het genoom direct of indirect betrokken is bij cognitie, en bevestigt de grote complexiteit van deze functie in het brein.

190 Tenslotte vat Hoofdstuk 6 de resultaten van 4 jaar werk over gedragsfenotypering samen en wordt de toegevoegde waarde besproken van geautomatiseerde, langdurige, en continue (“high-content” en “high-throughput”) gedragsfenotypering bij het bestuderen van muismodellen voor de rol van genen in humane neurologische aandoeningen en cognitieve functies. De verwachting is dat dergelijke geautomatiseerde gedragsfenotypering in thuiskooien de komende jaren een vlucht zal nemen en het gedragsonderzoek bij muizen ethologisch gezien zal verbeteren en aanvullen. Hoewel verdere validatie nodig is, laten de resultaten in dit proefschrift zien dat in het onderzoek naar humane ziektes en aandoeningen, automatische gedragsfenotypering een essentieel onderdeel zou moeten worden van de gedragskarakterisering van genetisch gemodificeerde muizen.

& S amenvatting

191

Summary Behavioral phenotyping of complex traits in inbred and mutant mice Neurological pathologies and psychiatric disorders are commonly detected through distinct human behavioral abnormalities. From a medical perspective, the understanding of the underlying biological mechanisms involved in those behaviors is essential. The complexity of both these abnormal behaviors and the biological mechanisms requires the use of animal models and appropriate phenotyping systems. Common behavioral tests in mice are based on locomotor activity measures and are designed to segregate behaviors over a short period of time (5 to 60 min). Thus, to describe the effect of a particular genetic background, a genetic mutation, or a drug on behavior, a battery of tests is required to tap into different aspects of behavior such as motor, sensory, cognitive and circadian functions. Those sets of experimental procedures are time consuming. Moreover, repeated human-animal interactions, environmental effects and other technical differences between laboratories are sources of variation in the results that lead to non-replication and difficulty in the interpretation of results. Consequently, there is growing consensus that behavioral testing needs a boost toward automation to increase its effectiveness and standardization in the study of animal models of human diseases and therapeutic developments. In the last decade, new behavioral technologies have emerged, allowing automated observation of animals in their home cage over long periods of time (e.g., several consecutive days). Automation of observation allows repetitive, objective, and consistent measurement with minimal human- & animal interactions. Furthermore, continuous recording allows investigation of multi-dimensional S

aspects of behavior like habituation, and baseline and challenged behavior, offering the possibility ummary to study longitudinally the progression or change of behavior. This is of utmost relevance to the study of models of neurological, psychiatric, and neurodegenerative disorders. In Chapter 2, we characterized complex behavioral responses indicative of avoidance learning in mice, using an unsupervised automated high throughput system. We adopted an assay exploiting the natural tendency of mice to develop a preference for one of two shelter entrances, by automatically detecting shelter entrance preference. Next, when the preferred entrance was used to enter the shelter, the animal was sanctioned by a mild aversive stimulus (illumination of the shelter with bright light). This learning paradigm specifically addressed cognitive aspects of avoidance behavior and produced a wealth of information on other aspects of behavior. Using this assay, we screened 8 inbred strains and a panel of 43 different mutants, and identified a new candidate gene, specc1/cytospinB, involved in avoidance learning. Our data show that the complex adaptive behavioral response of mice can efficiently and successfully be detected, analyzed and visualized even in large cohorts of (mutant) mice. Various inbred strains and single gene mutants exhibited marked quantitative differences in distinct aspects of this behavioral task. Chapter 3 investigated the potential of Munc18-1 haploinsufficiency in mice as a model for early infantile epileptic encephalopathy (EIEE), also known as Othahara syndrome. De novo heterozygous mutations in the human MUNC18-1 gene, STXBP1 are suspected to cause severe intellectual deficits with or without epileptic seizures.MUNC18-1 protein is essential for regulatory functions of synaptic vesicle secretion in mammalian synapses, a homozygous deletion of

193 Munc18-1 results in a postnatal death. Using an extensive test battery, including high-throughput home cage behavioral phenotyping and a wide range of classical behavioral tests, we investigated the behavioral effects of Munc18-1 haploinsufficiency in mice. Heterozygous (HZ) Munc18-1 mice showed no obvious epileptic seizures or cognitive impairment. However, despite the high anxiety level observed, HZ mice exhibited a lower fear responses and faster extinction. Munc18-1 HZ mice were, however, more anxious and showed a more proactive coping strategy compared to their WT litter mates when facing a more severe stressor. The reduced amount of MUNC18-1 protein appears to be sufficient to maintain most of the cognitive function in the mice. In Chapter 4, we focused on the characterization of a potential mouse model for Major Depressive Disorder (MDD). Genome-wide association studies (GWAS) have suggested a role for a non-synonymous exonic variation in the presynaptic gene PCLO in MDD. We investigated the PCLO variation (S4814A) and its effects on a molecular, cellular and behavioral level in the highly homogeneous background of C57BL/6J inbred mice. Knock-in mouse model expressing the PcloSA/SA variant showed an increased synaptic Piccolo level and a 30% increased excitatory synaptic transmission in cultured neurons. However, anxiety, cognition and depressive-like behavior were normal in PcloSA/SA mice. The fact that MDD is a multifactorial disease, brought on by a combination of many genetic and environmental factors might be a reason why PcloSA/SA mice did not show a strong behavioral phenotype. However, the molecular changes we observed might slightly increase the risk for MDD under certain circumstances. Historically, most of the studied genes were discovered after clinical observations in dramatic phenotype alterations, leaving the majority of genes in the genome unstudied. Thus in Chapter 5, we generated 5 randomly mutated mouse strains (Dpp10, Fgf13, Kcnd2, Ttc39c and Ubn1), using germlines carrying the Sleeping Beauty transposase together with a transposon. Those 5 strains were behaviorally phenotyped using a novel screening approach involving avoidance learning. The home cage-based system used to screen spontaneous and avoidance learning behavior described in e.g. Chapter 2, revealed the high sensitivity of spontaneous behavior to gene modifications. Moreover, we showed the involvement of Ubn1, a subunit of the HUCA histone chaperone complex, in the formation of simple associative memory. Studying 5 random functionally unknown genes and finding out that one of them is involved in cognitive processes, might suggest that a large proportion of genes in the genome are directly or indirectly related to cognition, which support the high genetic complexity of brain function. Finally, Chapter 6 summarizes the major results of 4 years of work on behavioral phenotyping and discusses the added value of high-content and high-throughput behavioral phenotyping to investigate mouse models of neurologic pathologies and genes involved in cognitive functions. Overall, the recently developed automated home cage phenotyping systems are expected to substantially improve and complement the study of ethologically valid behavior in mice. While additional validation is still required, the studies reported in this thesis do show that automated phenotyping is turning into an essential tool in the characterization of genetically modified mice for higher translational value of disease/disorder models.

194 Acknowledgment I want to conclude by expressing my gratitude to the people who played a part, directly or indirectly, in the achievement of this manuscript. First of all, I want to thank Prof. Matthijs Verhage for giving me the responsibility of such a promising project. I managed to bring this project where it is now, thanks to the guidance of my two co-supervisors Dr. Sophie van der Sluis and Dr. Oliver Stiedl. Oliver was always available to help me interpret my data and share his behavioral expertise and Sophie showed a great deal of patience and taught me how to handle statistics in the different problematic I encountered. My work really benefited of the different points of view of these three persons. I also want to thank Bastijn Koopmans as we made a great team in the mouse house, spending countless hours observing squeaky, dummy and all their friends running around and for his programming talents that made me the bug hunter I am now. Being able to discuss my thoughts, hypothesis and problem with the PT meeting team was always constructive and inspiring. I am grateful for the work of the technicians, especially Joost Hoetjes for genotyping my animals and for improving my Dutch with slapouwehoeren as well as Joke Wortel, Christian van der Meer and the Mousehouse team for taking care of the animals. Thank you, Els Borghols for your efficiency in all the administrative matters. I would like to thanks the committee for reading my thesis and accepting to challenge me during the defense. Thanks to Desiree Schut and Bastijn Koopmans for being my paranimf, I know you’ll be of great & support during my defense. A

Thanks to the people in room A-445, and the more and the less fruitful discussion I had with cknowledgment Tony, Jurjen, Rocio, Julia and Jens. We had a good balance between scientific, technical and the 4 o’clock BS discussion during those years. I really enjoyed my coffee and turkse pizza days with the kippenhok,M argherita, Marieke, Jula, Tatiana, Sabine, Asiya and Rhea. And of course it would not have been the same without the core of the lab, the technician’s room. I really enjoyed stopping by every day, especially Fridays, to have good moment with Desiree, Robbie, Joost, Bastijn, Joke, Frank, Ingrid. Thanks of the rest of the team on the 4th floor and also people from the 3rd floor. I also want to thanks people from outside the VU, with whom I had great times in Amsterdam. First my two awesome flatmates, Nutabi and her sweet and joyful craziness and Will for his Californian mellow and his recipe of rice and beans, it was great to share those years with you. My climbing buddies, Blaz, Rita, Marieke, Caspar, Will and Margherita always keep la patate! Anton, thanks for everything, Amsterdam wouldn’t have been the same without you. You introduced me to Amsterdam in a way that I could only love it. And again thanks for the illustration of this book. Thank you also to Michaëla, Ferdinand, Crelis and Aukje, my second family. Desiree and Michel, you made me love the Dutch way of life and took me as one of your own even thought I wasn’t speaking the language. I really want to thank you, and Marghe, Marieke, Robbie, Joost, Jana, Andreas, Tony, Tatiana, Danai, Bart, Emmeke, Loek and the fishermen’s friends for all the gezellig game nights, parties or bbq that we had.

195 Many thanks to my Mom and Dad who I guess are partly responsible for the fact that I could and wanted to do a PhD, Mat for being an awesome brother. Finally, thanks to Juliette, for believing in me, in us even with 500km of distance. If you read this and that your name is not in it, I’m sorry but be sure that you have all my gratitude.

196 Thank you!!!