R

Tm

ASa

b

c

a

ARRA

KMLMAAACA

1

vedpsummrr

c9

0d

Behavioural Brain Research 220 (2011) 152–158

Contents lists available at ScienceDirect

Behavioural Brain Research

journa l homepage: www.e lsev ier .com/ locate /bbr

esearch report

he COGITAT holeboard system as a valuable tool to assess learning,emory and activity in mice

ntonia M. Posta,1, Thomas Wultscha,1, Sandy Poppa, Evelin Painsippb, Heike Wetzsteina,arah Kittel-Schneidera, Thomas A. Sontagc, Klaus-Peter Lescha, Andreas Reif a,∗

Molecular Psychiatry, Laboratory of Translational Neuroscience, Department of Psychiatry, Psychosomatics, and Psychotherapy, University of Würzburg, Würzburg, GermanyResearch Unit of Translational Neurogastroenterology, Institute of Experimental and Clinical Pharmacology, Medical University of Graz, Graz, AustriaDepartment of Experimental Psychology, University of Regensburg, Regensburg, Germany

r t i c l e i n f o

rticle history:eceived 5 July 2010eceived in revised form 25 January 2011ccepted 30 January 2011

eywords:ouse behavior

earningemory

ttentionlzheimer’s disease

a b s t r a c t

The comprehensive and stress-free assessment of various aspects of learning and memory is a prereq-uisite to evaluate mouse models for neuropsychiatric disorders such as Alzheimer’s disease or attentiondeficit/hyperactivity disorder (ADHD). COGITAT is an automated holeboard system allowing simultane-ous assessment of spatial working and reference-memory performance which we have adapted in thisstudy to enable its usage with mice. The holeboard apparatus consists of an open-field chamber with a25-hole floor insert, each hole being monitored by infrared light beams, located on three different levels,allowing the distinction between visits of holes, i.e. the animal reaches the bottom of the hole, or inspec-tions, which means only superficial exploration of the hole. Across trials, animals learn a pattern of fivebaited holes. Here, we show that C57BL/6 mice readily acquire this task within 5 days when submittedto six trials per day. A number of individual parameters – overall exploratory activity, number of visits

DHDognitive enhancernimal model

into or inspections of holes, number of baited, unbaited, or previously baited holes visited or inspected,reinspections of or revisits into any holes, number of pellets eaten, time to find pellets, and reference andworking memory errors—are obtained simultaneously and results are immediately available after theend of each experiment. The muscarinic antagonist scopolamine impaired task performance, while thecognitive enhancer metrifonate (an acetylcholinesterase inhibitor) reduced error rates. Overall, our dataindicate that this spatial learning task will be useful to characterize spatial memory in various genetic or

ode

pharmacological mouse m

. Introduction

Learning and memory processes are thought to underlie aariety of human psychiatric disorders, including neurodegen-rative disorders such as Alzheimer’s disease but also attentioneficit/hyperactvity disorder (ADHD). A number of behavioralaradigms have been developed over the past decades to assesspatial learning and memory performance in rodents. Commonlysed in this regard are mazes, including the Y-Maze, radial-arm

aze and water-maze [1–3]. These paradigms have providedany insights into the neurobiological mechanisms and the neu-

oanatomical substrates of learning and memory. However, theesults are often inconsistent from one experiment to another [4]

∗ Corresponding author at: Department of Psychiatry, Psychosomatics and Psy-hotherapy, Psychiatric Neurobiology, University of Würzburg, Füchsleinstrasse 15,7080 Würzburg, Germany. Tel.: +49 931 201 76402; fax: +49 931 201 77220.

E-mail address: reif a@klinik.uni-wuerzburg.de (A. Reif).1 These authors contributed equally to this work.

166-4328/$ – see front matter © 2011 Elsevier B.V. All rights reserved.oi:10.1016/j.bbr.2011.01.054

ls.© 2011 Elsevier B.V. All rights reserved.

and from one laboratory to another [5]. Furthermore, different test-ing procedures have had to be used to measure various cognitiveand motor capabilities, working and reference memory and seriallearning. These sequential studies interfere with each other and arebound to involve different circumstances and environmental con-ditions, which will affect the animals’ behavior in different waysand hamper a strictly comparable and complete view of the ani-mals’ memory function.

Holeboard learning paradigms for rats have been described pre-viously [6–8]. Previous studies indicated that mice are able to learnto retrieve one food pellet located in a holeboard task setup [9,10].However, the simplicity of the task used precludes direct assess-ment of reference- and working-memory performances. Kuc et al.adapted a protocol similar to the one published by Van der Stayet al. [8,11]. However, their setup is limited in picturing the animalı̌s

behavior in detail, as it is unable to distinguish between superficialhead dips and deep inspectations of a visited hole. In an attempt toovercome these problems, we have modified an elaborated hole-board system for rats, named COGITAT system [6]. Our goal was tomodify and validate this system to be suitable for mice, and allow-

A.M. Post et al. / Behavioural Brain R

Start

Holeboard

ulb

llb

pellet

Fig. 1. The COGITAT holeboard contains a total of 25 holes, each consisting of acylindrical tube that is closed off at its lower end by a feeding plate.The system usesao(p

iaftiBwetom

2

2

GbTteiofaIrtatr(be

TL

n infrared sytem to record activity: Differential dips of the head at the upper levelf the tubes (ulb), which are equivalent to inspections, exploring deep into the holellb), but without eating the pellet, which are scored as visits, and collection of theellets (pellet).

ng a detailed and automatic simultaneous recording of cognitivend motor behavior. In order to validate this holeboard, we optedor the use of an anticholinergic compound, scopolamine, knowno cause cognitive impairment [12,13]. The acetylcholinesterasenhibitor metrifonate was used as a cognitive enhancer [14,15].y coupling the COGITAT system to videotracking software, weere able to measure learning and activity parameters in a single

xperiment. This substantially saves time, simplifies experimen-al protocols and enables to control memory parameters for levelsf activity allowing the precise determination of cognitive impair-ent.

. Methods

.1. The COGITAT hole board

During a test period, mice explored a modified COGITAT hole board (CognitronmbH, Göttingen, Germany; Fig. 1; size 660 mm × 670 mm, inner surface) borderedy a clear Plexiglas surround (height, 270 mm) giving access to distal spatial cues.he board contains an array of 5 × 5 holes (diameter, 35 mm; distance apart cen-er to center, 127 mm), each consisting of a cylindrical tube closed off at its lowernd by an adjustable feeding plate (50 mm below the upper surface) with a cavitynto which a food pellet (0.045 g; Bio-Serv, Frenchtown, NJ) fits exactly. The colorf the feeding plate and of the food pellet is a perfect match. The ground below theeeding plate and the cylindrical tubes was covered with vanilla odor to prevent thenimals from working out the distribution of the pellets by using olfactory stimuli.n a typical experimental run, five of the cylinders are baited with one pellet each. Aun (trial) is automatically ended after a predefined time period of 240 sec. Duringhis time-limited phase, the animals are provided with the opportunity of finding

nd eating the food pellets. At the same time, however, they also have the oppor-unity of recognizing the spatial pattern in which the pellets are presented and ofemembering this pattern. The system uses an infrared system to record activityFig. 1): differential dips of the head at the upper level of the tubes (upper lighteam [ulb] 10 mm beneath the upper surface), which are equivalent to inspections;xploration deep into the hole (lower light beam [llb], 20 mm above the level of

able 1ist of variables.

Variable Definition/explanation

ActivityTotal distance Total distance traveled in the course oOverall inspections The sum of inspected and visited holeUpper light beam disruptions The total number of disruptions of upLower light beam disruptions The total number of disruptions of low

AcquisitionTime to find pellets The time it took to find the pellets wiPellets eaten The number of pellets eaten

ErrorsReference memory errors, total The percentage of the sum of inspecti

inspected and visitedWorking memory errors, total The percentage of the sum of inspecti

inspected and visited

esearch 220 (2011) 152–158 153

the pellet), which are scored as visits; and collections of the pellets by eating them,which is detected by an infrared light barrier at the level of the pellet. There are mani-fold, partially inter-dependent parameters which can be recorded automatically andsimultaneously. In order to provide an overview, we selected seven variables fromthe automatic output that represent these manifold parameters as comprehensiveas possible, and then thematically assigned them to one of three behavioral domainsfor better readability: Activity, Acquisition and Errors. The domain “Activity” com-prises the overall inspections of holes in one trial as well as the number of upperand lower light beam disruptions as measures of general activity. The time to findall five pellets and the number of eaten pellets per trial are combined to the domain“Acquisition”. Finally, reference and working memory errors make up the domain“Errors”. Additional surveillance with a video camera (VideoMot2, TSE Systems, BadHomburg, Germany) offers the possibility of recording not only the correct path andspeed of the animals, but also their motor performance and general level of activity,during one trial and over the course of the whole experiment. This variable was alsoassigned to the Activity domain. All selected variables and the respective behavioraldomains can be found in Table 1.

2.2. Animals

For behavioral experiments, 30 male C57Bl/6j mice aged between 8 and 10weeks (Charles River, Sulzfeld, Germany) were used. Animals were grouphousedupon arrival under identical conditions and allowed to acclimate to our facility fortwo weeks. The animals were kept on a 12:12-h light:dark cycle (lights on at 06:00 h),and humidity and room temperature were 55% and 21 ◦C, respectively. They hadfree access to regular mouse chow (ssniff, Soest, Germany) and water during theacclimation period. Seven days prior to bahavioral testing, mice were weighed andand individually housed with free access to water only. Mice were administered areduced food ration, which consisted of a daily food ration calculated individuallyfor each animal (food was given every day after trials between 3 and 4 p.m.) in sucha way that a weight reduction of no more than ∼10–15% of the initial weight wasachieved. In contrast to rats, when feeding was offered before test procedure [6],mice did not search intensively enough during the trials (data not shown). Thus,mice and rats require different feeding strategies to motivate them for the task. Thefood restriction procedure was maintained throughout behavioral testing. In addi-tion, mice were given a daily ration of 2–3 sugar pellets prior to behavioral testingto become familiar with the food reward. All animal protocols have been reviewedand approved by the review board of the Government of Lower Franconia and theUniversity of Würzburg and were conducted according to the Directive of the Euro-pean Communities Council of 24 November 1986 (86/609/EEC). The experimentswere designed in such a way that the number of animals used and their sufferingwas minimized.

2.3. Drug treatment

The effects of the non-selective muscarinic antagonist scopolamine and theacetylcholinesterase inhibitor metrifonate on learning and memory were tested inthe COGITAT holeboard task. Scopolamine hydrochloride was dissolved in saline andadministered at 0.05 ml/10 g of bodyweight in a concentration of 0.1 mg/kg subcu-taneously. Metrifonate was dissolved in saline and administered at 0.1 ml/10 g ina concentration of 50 mg/kg i.p. Each drug was administered 30 min prior to thedaily session. Control animals reveived a saline injection at a volume of 0.05 ml/10 gintraperitoneally. Mice were trained for 5 days and went through 6 trials per day.

2.4. Test procedure

Experiments were carried out between 8 a.m. and 4 p.m. Five holes of the COG-ITAT system were serially baited in a L-shaped pattern with food pellets not visibleto the animals while moving. Each daily session consisted of six 240 s trials withan intertrial interval of 30 min. Animals were tested in a random order. A trial was

f one trial (cm)s, with and without food collectionper light beams in one trialer light beams in one trial

thin a time limit

ons and visits to unbaited holes in relation to the total number of holes

ons and visits to previously baited holes in relation to the total number of holes

1 rain R

ctfieoaPtctEst

2

weec

3

TtmatiFbatdtalwdcmhd

3

Asf3stf3

rF3gam

lftn

54 A.M. Post et al. / Behavioural B

ompleted as soon as the animal had collected and eaten all of the pellets withinhe allotted time span of 240 s or when the time span had elapsed, whichever camerst. Spatial cues available for the animals inside the holeboard enclosure were thentrance with the starting box, the four corners of the enclosure, the upper edgesf the 25 holes, and the side walls for orientation with the vibrissae (thigmotaxis)s tactile stimuli. Outside the enclosure, distal cues visible through the transparentlexiglas walls were, besides the experimenter (who maintained the same positionhroughout the experiments), a wall, a window, a black curtain and a rack with theages of the remaining experimental animals. In the center above the hole boardhe video camera (VideoMot2, TSE Systems, Bad Homburg, Germany) was attached.ach animal was subjected to six trials per day. During the first experimental ses-ion, a trial duration of 240 s provided the animals with the opportunity to explorehe hole board and to find the five hidden pellets.

.5. Statistical procedure

For statistical evaluation a repeated measures analysis of variance (ANOVA),ith Greenhouse-Geisser adjustments (groups and trials as factors), was used for

ach variable. Scheffé tests served as post-hoc-analyses of between group differ-nces. The results are displayed as means ± SEMs of the individual trials of theorresponding experimental periods.

. Results

During the animals’ acquisition of this learning task, the COGI-AT system automatically provides 23 different parameters abouthe animals’ overall exploratory behavior and their learning and

emory capabilities in a comparison between vehicle, scopolaminend metrifonate pretreated animals. Additionally, we measuredhe total distance moved by the animals by means of video track-ng software (VideoMot2, TSE Systems, Bad Homburg, Germany).rom all available parameters, we chose 8 variables to distinguishetween three different behavioral domains: Activity, Acquisitionnd Errors. The Activity domain comprises the variables total dis-ance moved, overall inspections as well as the frequency of theisruptions of the upper and lower lightbeams (ulb, llb, respec-ively). For the Acquisition domain, the variables time to find pelletsnd pellets eaten were taken. Finally, for the Error domain weooked at the reference memory error (rme) as well as at the

orking memory error (wme). The reference memory error isefined as the number of visits of unbaited holes in relation to theumulative number of upper light beam disruptions. The workingemory error is defined as the number of visits of previously baited

oles in relation to the cumulative number of upper light beamisruptions.

.1. Activity

Fig. 2 shows the general activity of the mice during the task.ll three groups readily searched for the hidden pellets. Fig. 2ahows the total distance moved by the animals while searchingor the food pellets. The data reveal significant effects for trial (F(2,6) = 1.794, p < 0.001) and group (F(2, 36) = 8.902, p = 0.009) withcopolamine treated animals moving longer distances as comparedo metrifonate treated animals (p = 0.012). No between group dif-erences can be seen in the latency to enter the holeboard (F(2,6) = 0.086, p = 0.919, data not shown).

Repeated measures ANOVA of overall inspections (Fig. 2b)evealed significant effects for trial (within-subject effect:(2, 36) = 26.562, p < 0.001) and trial × group interaction (F(2,6) = 3.520, p < 0.001) as well as significant differences betweenroups (Fig. 1a, F(2, 36) = 25.275, p < 0.001): scopolamine treatednimals showed increased levels of activity when compared toetrifonate (p < 0.001) and vehicle treated (p < 0.001) animals.

When looking at inspections of the holes at the upper

ightbeam stage (Fig. 2c), we also found significant effectsor trial (within-subject effect) (F(2, 36) = 23.060, p < 0.001) andrial × group interaction (F(2, 36) = 6.156, p < 0.001) as well as sig-ificant between-group differences (F(2, 36) = 7.571, p = 0.002).

esearch 220 (2011) 152–158

Metrifonate treated animals exhibit fewer hole visits at the ulbstage as compared to vehicle controls (p = 0.014) and scopolaminetreated animals (p = 0.004). There were no differences betweenscopolamine and vehicle treated animals.

In contrast, at the lower lightbeam stage (Fig. 2d) the within-subject effect for trial (F(2, 36) = 15.304, p < 0.001) and thetrial × group interaction (F(2, 36) = 5.294, p < 0.001) are signifi-cant, but no between-group differences could be observed (F(2,34) = 2.844, n.s.). Consequently, animals from the three pharma-cological groups did not differ for this variable.

3.2. Acquisition

Fig. 3 shows the general ability of mice to learn the task(Fig. 3a displays the time needed to complete the task, whileFig. 3b shows the number of pellets found by the animals). Atraining or learning effect is present in all three groups over thetime course for both acquisition related variables [main effect fortrial (within-subject effect), time to find pellets (F(2,36) = 39.732),p < 0.001) and pellets eaten (F(2, 36) = 44.615), p < 0.001]. Therewere also significant trial x group interactions for both variables[trial × group, time to find pellets (F(2, 36) = 7.037), p < 0.001] andpellets eaten (F(2, 36) = 4.526), p < 0.001]. In addition, we foundsignificant between-group differences for both variables [maineffect for group (between subjects effect), time to find pellets(F(2, 36) = 16.576), p < 0.001] and pellets eaten (F(2, 36) = 16.798),p < 0.001].

Post Hoc Scheffé tests revealed significant differences betweenthe scopolamine treated and the vehicle groups for both variableswith scopolamine treated mice taking more time to find pellets(p = 0.003) and eating fewer pellets (p = 0.001) compared to vehi-cle treated mice. There was also a significant difference betweenscopolamine and metrifonate treated mice for both variables, withmetrifonate treated mice taking less time to find pellets (p < 0.001)and eating more pellets (p < 0.001) than scopolamine treated mice.No difference between metrifonate and vehicle treated mice wasobserved.

3.3. Errors

Repeated measures ANOVA for reference memory error (rme)revealed a significant main effect for trial (within-subject effect rme(F(2, 36) = 24.182), p < 0.001) and a significant trial x group interac-tion (trial × group rme (F(2, 36) = 6.908), p < 0.001), shown in Fig. 4.There also is a significant main effect for group (between-subjectseffect rme (F(2, 36) = 4.728), p = 0.021) with vehicle treated animalsmaking more reference memory errors than metrifonate treatedanimals (p = 0.025).

The repeated measures ANOVA for working memory error(Fig. 4, wme) also reveals a significant main effect for trial(inner subject effect wme (F(2, 36) = 10.039), p < 0.001), a signifi-cant trial × group interaction (trial × group wme (F(2, 36) = 7.977),p < 0.001) and a significant main effect for group (between-subjectseffect wme (F(2, 36) = 46.488), p < 0.001). All of the 3 wme Post HocScheffé tests reveal significant effects: metrifonate treated animalscommit fewer working memory errors than vehicle treated animals(p < 0.001), and vehicle treated animals commit fewer work-ing memory errors than scopolamine treated animals (p < 0.001).Finally, scopolamine treated animals commit more working mem-ory errors than metrifonate treated animals (p < 0.001).

4. Discussion

Using the novel modified COGITAT holeboard/videotrackingsystem, we were able to detect changes in behavioral measuresrelated to exploration and memory. While some of the obtained

A.M. Post et al. / Behavioural Brain Research 220 (2011) 152–158 155

F ng the( e upph /kg s.c

pwtamattOaIhowfavwdlrradtrmctw

ig. 2. Data on the Activity behavioral domain. The overall exploratory activity duriPanel b) is shown. The inspections (Panel c) indicate short dipping of the head at tholes (Panel d). Full circle, metrifonate 50 mg/kg i.p.; triangles, scopolamine. 0.1 mg

arameters can also be obtained by other methods, such as theater maze or the radial arm maze, it is essential to stress that

here is a fundamental difference between the water and radialrm mazes on one and the hole board on the other side. In the wateraze, stress forces the animals to use the best strategy enabling the

nimal to escape from the water as quickly as possible [2,16]. In con-rast, both the radial arm maze and the hole board offer the animalshe opportunity of exploring a new environment and finding food.n the holeboard in particular, mice display exploratory [11,17]s well as food-searching behavior [11] without being forced to.mportantly, the system described here differs from conventionaloleboards in that infrared beams are positioned on three levelsf the holes. This provides the possibiliy to determine precicelyhether mice explore parts of the board only superficially, per-

orm deep inspections or eat the hidden pellet. However, therere big differences between rats [6] and mice in terms of moti-ation to enter the holeboard and actively search for the pellets,hich was a crucial point when establishing this task. A restrictedaily feeding procedure during the experiments had to be estab-

ished to motivate the animals to perform the task. In contrast toats, mice still engaged in the search for pellets if they expected toeceive their entire daily food ration after the trials. In contrast, ratsct the opposite way, as they do not search for pellets when theiraily ration of food is given at the end of the trials [6]. Thereforehe COGITAT procedure, which has been previously only used in

ats [6], had to be modified slightly to meet the specific require-ents of mice. These modifications included the above mentioned

hanges in the feeding procedure of the animals. However, whenhe animals were motivated to search for the hidden food pellets, itas not only possible to measure differences in working memory

course of the experiment in terms of locomotion (Panel a) and overall inspectionser surface of the holes, in contrast to intensive searching for pellets deep inside the.; open circle, vehicle.

(which can also be done in the water maze), but simultaneouslyalso in reference memory. It has to be mentioned that food restric-tion can be seen as a stressful factor that might influence behavioraltesting, especially when it is as complex as a learning task. How-ever, there are studies reporting better performance in comparabletasks (radial maze) in mice that were calorically restricted in amanner similar to our food restriction paradigm (however thiswas even done over a longer period of time as compared to ourstudy) as opposed to mice that were allowed to feed ad libitum[28]. This data suggests that the more ecologically valid way ofoffering food in a restricted manner improves cognitive abilities inmice.

As c57BL/6 mice are generally viewed as “good performers” inspatial learning tasks [18], and as they are a common backgroundfor genetic modification and thus are often used in behavioralparadigms, we decided to use this mouse strain for our valida-tion experiments. To disrupt learning, we used the anticholinergiccompound scopolamine, known to cause cognitive impairment[12,13], and the acetylcholinesterase inhibitor metrifonate as a cog-nitive enhancer [14,15]. The effective periods for both compoundsencompassed at least the approximately 180 min of testing perday [19,20]. Scopolamine, a non-selective muscarinic antagonist,has been shown to induce performance impairments in a varietyof learning and memory tasks in rodents [11,21,22]. In this study,daily pretreatment with doses of 0.1, 0.3 and 1.0 mg/kg of scopo-

lamine resulted in significant performance deficits in c57Bl/6 mice.The highest dose, however, also induced a significant decreasein activity during training (distance travelled, not shown) whichis consistent with previously reported results [11,23]. Hence, weopted to use the lowest dose of 0.1 mg/kg, as this dose resulted

156 A.M. Post et al. / Behavioural Brain Research 220 (2011) 152–158

F bilityi en pet

ie

esnf(tamnastimai

ig. 3. Data on the Acquisition behavioral domain. These panels show the general an the given time of 240 s per trial, whereas (B) illustrates how many of the 5 hiddriangles, scopolamine. 0.1 mg/kg s.c.; open circle, vehicle.

n spatial learning impairments without inducing any motor sideffects.

In contrast to scopolamine, daily pretreatment with metrifonatenhanced the holeboard performance of c57Bl/6 mice. As can beeen in Fig. 2a, metrifonate and vehicle treated animals had a sig-ificantly decreased total distance moved on the holeboard, as they

ound more hidden pellets (Fig. 3b) in a shorter period of timeFig. 3a), indicating more effective search strategies. This assump-ion is backed up by the overall inspections of a hole variable,s the scopolamine treated animals inspected holes significantlyore often than the two other groups (Fig. 2b–d). Increasing the

umber of inspections increased their chances of finding pellets,lthough this required increased effort and lacked a sophisticatedearch strategy. This increase in speed diminished their attention

o the task and to the pattern and resulted in a higher rate ofnspections of unbaited holes. As a consequence, obvious reference

emory deficits are present (see Fig. 4) in scopolamine treatednimals, for which they are, however, able to compensate by anncreased locomotor and exploratory activity, as can be seen in

of mice to learn the task. (A) Shows the time taken to find all hidden food pelletsllets were found within the 240 s time span. Full circle, metrifonate 50 mg/kg i.p.;

Fig. 2a. However, this compensation does not correlate to learningsuccess—measured as the number of pellets consumed—as com-pared to vehicle or metrifonate treated animals (Fig. 3a). Thesefindings correspond well to a similar setup used by Kuc and asso-ciates [11]. With respect to working memory error, scopolaminetreated animals performed poorer compared to vehicle treated ani-mals. In contrast, metrifonate treated animals performed superiorto both groups on days 4 and 5 (Fig. 4a) arguing for the importanceto measure reference memory performance. It could be argued thatthe weaker performance of scopolamine treated animals mightbe a motivational problem resulting from an attenuated feelingof hunger. However, scopolamine treated mice not only searchedlonger (which could be an argument for being less hungry) but alsomade more mistakes and were more active which clearly suggests

that they were trying to find the pellets. In the end, also scopo-lamine treated mice were able to complete the task almost as wellas the two other groups, which again suggests that the problemdoes not result from decreased motivation. Taken together, thisset of experiments demonstrate that our system is able to disen-

A.M. Post et al. / Behavioural Brain Research 220 (2011) 152–158 157

Fig. 4. Data on the Errors behavioral domain. Two forms of errors are depicted. (A) Shows the percentage of the sum of inspections and visits to unbaited holes in relation tot B) shor . Fullv

tmoc

lipttedtstc

he total number of holes inspected and visited (reference memory errors), whereas (elation to the total number of holes inspected and visited (working memory errors)ehicle.

agle various aspects of learning and memory not accessible byore simplistic paradigms, thereby allowing better interpretation

f the animals’ behavior especially important in mouse models ofognitive disorders.

When discussing the results of behavioral tests, housing andighting conditions during testing should always be considered asnfluencing variables, especially as we conducted the experimentsresented here during the light (and therefore inactive) phase ofhe day. We have however shown in a previous study [24] thathe time of day of the behavioral testing does not have a largeffect on results. Furthermore, many studies (see for example [25])

emonstrated that learning and memory can reliably be tested inhe light phase, and thus we are confident that our results are con-istent. As for the housing conditions, mice were singly housed forhe experiments, and thus it could be argued that this isolationauses cognitive deficits. However, one study comparing housing

ws the percentage of the sum of inspections and visits to previously baited holes incircle, metrifonate 50 mg/kg i.p.; triangles, scopolamine. 0.1 mg/kg s.c.; open circle,

effects on the forced swim test and the tail suspension test foundhigher occurrence of depression-like behavior in mice that werehoused in groups and as a consequence were subjected to socialstressors (as opposed to single housed mice)[26]. Moreover, grouphousing increases phenotypic variance [27] which can be a problemin behavioral testing.

5. Conclusion

In summary, the COGITAT holeboard appears to be well suitedfor testing mice and allows a direct assessment of reference-

and working-memory performance simultaneously. Here, we usedestablished pharmacological interventions to depict the abilityof c57Bl/6 mice to complete a learning task. The automatic andcomputerized simultaneous analysis of a variety of behavioralparameters, together with their statistical and graphical presenta-

1 rain R

tlis

A

m1Fd

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

58 A.M. Post et al. / Behavioural B

ion, offers an objective characterization of cognitive, memory, andearning capabilities. It also enables to measure learning and activ-ty parameters in a single experiment, substantially saving time andimplifying experimental protocols.

cknowledgements

This study was supported by the Deutsche Forschungsge-einschaft (KFO 125, SFB 581, SFB TRR 58 A1, A5 and Z2, GRK

156, RE1632/5-1) and the Bundesministerium für Bildung undorschung (BMBF 01GV0605). There is no conflict of interest toisclose.

eferences

[1] Crusio WE, Schwegler H, Lipp HP. Radial-maze performance and structural vari-ation of the hippocampus in mice: a correlation with mossy fibre distribution.Brain Res 1987;425:182–5.

[2] Morris R. Developments of a water-maze procedure for studying spatial learn-ing in the rat. J Neurosci Methods 1984;11:47–60.

[3] Shillito EE. The effect of chlorpromazine and thioridazine on the exploration ofa Y-maze by rats. Br J Pharmacol Chemother 1967;30:258–64.

[4] Eijkenboom M, Van Der Staay FJ. Spatial learning deficits in rats after injectionof vincristine into the dorsal hippocampus. Neuroscience 1999;91:1299–313.

[5] Bear MF. Homosynaptic long-term depression: a mechanism for memory? ProcNatl Acad Sci USA 1999;96:9457–8.

[6] Heim C, Pardowitz I, Sieklucka M, Kolasiewicz W, Sontag T, Sontag KH. Theanalysis system COGITAT for the study of cognitive deficiencies in rodents.Behav Res Methods Instrum Comput 2000;32:140–56.

[7] Oades RD, Isaacson RL. The development of food search behavior by rats: theeffects of hippocampal damage and haloperidol. Behav Biol 1978;24:327–37.

[8] van der Staay FJ, van Nies J, Raaijmakers W. The effects of aging in rats on work-ing and reference memory performance in a spatial holeboard discriminationtask. Behav Neural Biol 1990;53:356–70.

[9] Laarakker MC, Raai JR, van Lith HA, Ohl F. The role of the alpha 2A-adrenoceptor in mouse stress-coping behaviour. Psychoneuroendocrinology2009;35:490–502.

10] Ohl F, Holsboer F, Landgraf R. The modified hole board as a differential screen

for behavior in rodents. Behav Res Methods Instrum Comput 2001;33:392–7.

11] Kuc KA, Gregersen BM, Gannon KS, Dodart JC. Holeboard discrimination learn-ing in mice. Genes Brain Behav 2006;5:355–63.

12] Ingram DK, Spangler EL, Iijima S, Ikari H, Kuo H, Greig NH, London ED. Rodentmodels of memory dysfunction in Alzheimer’s disease and normal aging: mov-ing beyond the cholinergic hypothesis. Life Sci 1994;55:2037–49.

[

[

esearch 220 (2011) 152–158

13] Levin ED, McGurk SR, Rose JE, Butcher LL. Cholinergic-dopaminergic interac-tions in cognitive performance. Behav Neural Biol 1990;54:271–99.

14] Ringman JM, Cummings JL. Metrifonate (Trichlorfon): a review of thepharmacology, pharmacokinetics and clinical experience with a new acetyl-cholinesterase inhibitor for Alzheimer’s disease. Expert Opin Investig Drugs1999;8:463–71.

15] van der Staay FJ, Hinz VC, Schmidt BH. Effects of metrifonate, its transformationproduct dichlorvos, and other organophosphorus and reference cholinesteraseinhibitors on Morris water escape behavior in young-adult rats. J PharmacolExp Ther 1996;278:697–708.

16] Brucato FH, Levin ED, Mott DD, Lewis DV, Wilson WA, Swartzwelder HS. Hip-pocampal long-term potentiation and spatial learning in the rat: effects ofGABAB receptor blockade. Neuroscience 1996;74:331–9.

17] Fiore M, Petruzzi S, Dell’Omo G, Alleva E. Prenatal sulfur dioxide exposureinduces changes in the behavior of adult male mice during agonistic encoun-ters. Neurotoxicol Teratol 1998;20:543–8.

18] Orsini C, Buchini F, Conversi D, Cabib S. Selective improvement of strain-dependent performances of cognitive tasks by food restriction. Neurobiol LearnMem 2004;81:96–9.

19] Cummings JL, Cyrus PA, Bieber F, Mas J, Orazem J, Gulanski B. Metrifonate treat-ment of the cognitive deficits of Alzheimer’s disease. Metrifonate Study Group.Neurology 1998;50:1214–21.

20] Legault G, Smith CT, Beninger RJ. Scopolamine during the paradoxical sleepwindow impairs radial arm maze learning in rats. Pharmacol Biochem Behav2004;79:715–21.

21] Chen J, Long Y, Han M, Wang T, Chen Q, Wang R. Water-soluble derivative ofpropolis mitigates scopolamine-induced learning and memory impairment inmice. Pharmacol Biochem Behav 2008;90:441–6.

22] Eleore L, Lopez-Ramos JC, Yi PJ, Delgado-Garcia JM. The cognitive enhancer T-588 partially compensates the motor associative learning impairments inducedby scopolamine injection in mice. Behav Neurosci 2007;121:1203–14.

23] Anisman H, Cygan D. Central effects of scopolamine and (+)-amphetamine onlocomotor activity: interaction with strain and stress variables. Neuropharma-cology 1975;14:835–40.

24] Post AM, Weyers P, Holzer P, Painsipp E, Pauli P, Wultsch T, Reif A, Lesch KP.Gene–environment interaction influences anxiety-like behavior in ethologi-cally based mouse models. Behav Brain Res 2011;218:99–105.

25] Holmes A, Wrenn CC, Harris AP, Thayer KE, Crawley JN. Behavioral profilesof inbred strains on novel olfactory, spatial and emotional tests for referencememory in mice. Genes Brain Behav 2002;1:55–69.

26] Karolewicz B, Paul IA. Group housing of mice increases immobility andantidepressant sensitivity in the forced swim and tail suspension tests. Eur

J Pharmacol 2001;415:197–201.

27] Nagy TR, Krzywanski D, Li J, Meleth S, Desmond R. Effect of group vs. singlehousing on phenotypic variance in C57BL/6J mice. Obes Res 2002;10:412–5.

28] Idrobo F, Nandy K, Mostofsky DI, Blatt L, Nandy L. Dietary restriction: effectson radial maze learning and lipofuscin pigment deposition in the hippocampusand frontal cortex. Arch Gerontol Geriatr 1987;6:355–62.