Impaired recognition memory in male mice with a supernumerary X chromosome

Physiology & Behavior 96 (2009) 23–29

Contents lists available at ScienceDirect

Physiology & Behavior j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / p h b

Impaired recognition memory in male mice with a supernumerary X chromosome Lars Lewejohann a,⁎, Oliver S. Damm b, C. Marc Luetjens b, Tuula Hämäläinen c, Manuela Simoni b, Eberhard Nieschlag b, Jörg Gromoll b, Joachim Wistuba b a b c

Department of Behavioural Biology, University of Münster, Badestrasse 13, D-48149 Münster, Germany Centre of Reproductive Medicine and Andrology, Domagkstrasse 11, 48149 Münster, Germany Institute of Biomedicine, Department of Physiology, University of Turku, Kiinamyllynkatu 10, 20520 Turku, Finland

a r t i c l e

i n f o

Article history: Received 3 April 2008 Received in revised form 7 July 2008 Accepted 6 August 2008 Keywords: XXY X chromosome Klinefelter syndrome Mice Novel object task Recognition memory

a b s t r a c t Several aberrant chromosomal constellations are known in men. Of these the karyotype XXY (Klinefelter syndrome, KS) is the most common chromosomal disorder with a prevalence of about one in 800 live-born boys. KS is associated with hypogonadism and is suspected to cause variable physical, physiological and cognitive abnormalities. As a supernumerary X chromosome is also associated with infertility, sound animal models for KS are difficult to obtain. In this study, male mice with two X chromosomes (XXY⁎) were derived from fathers carrying a structurally rearranged Y chromosome (Y⁎) that resulted in physical attachment of a part of the Y chromosome to one X. These animals display certain physiological features that resemble closely those of human KS and can also be utilized to study X chromosomal imbalance and cognition. Therefore 15 XXY⁎ males and 15 XY⁎ controls were subjected to a battery of behavioral tests, including a general health check, analysis of spontaneous exploration and locomotor activity, measures for anxiety-related behavior and the “novel object task” to test memory performance. Physiologically, XY⁎ males did not differ from C57Bl/6 wild type mice carrying a normal Y chromosome, which provided a valid control group. All mice appeared healthy. XXY⁎ mice did not differ from their wild type littermates with respect to locomotion, exploration and anxiety-related behavior. XXY⁎ male mice, however, exhibited no significant recognition memory performance in contrast with wild type XY⁎ males that readily fulfilled a given task. These findings support the hypothesis that the presence of a supernumerary X in male mice influences cognitive abilities. We suggest that the altered endocrine state and/or changes in the dosage of X-linked genes in the XXY⁎ mouse model affect brain function, in particular those regions responsible for cognition and learning behavior. © 2008 Elsevier Inc. All rights reserved.

1. Introduction Phenotypically male individuals bearing a supernumerary X chromosome develop from karyotypically aberrant gametes. These gametes derive from germ cells that suffered errors during the meiotic disjunction of the heterosomes. The most common (1:800 newborns) aberrant human male phenotype is the Klinefelter syndrome (47, XXY). In men, a supernumerary X chromosome resulting from nondisjunction events leads to dramatic phenotypical changes compared with normal XY males. This sex-chromosomal aberration especially affects endocrinological regulation and reproductive function resulting in infertility, often accompanied by cognitive and behavioral defects. This indicates how decisive the chromosomal balance between sex chromosomes is for the phenotype [1]. In Klinefelter patients, disturbances in social cognitive processing caused by possible changes in the neural network have been assumed to

⁎ Corresponding author. Tel.: +49 251 83 21014. E-mail address: [email protected] (L. Lewejohann). URL: http://www.ethologie.de (L. Lewejohann). 0031-9384/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.physbeh.2008.08.007

be functional consequences related to the X chromosomal abnormality [2,3]. Murine models that display conditions found in Klinefelter patients caused by a karyotype containing a supernumerary X chromosome have been previously reported [4–8]. The generation of XXY mice utilizes males with a spontaneously mutated Y chromosome, designated Y⁎, which acquired a new centromere distally and lost the normal centromere. Using the published breeding scheme, we experienced that we did not succeed in generating 41, XXY males (with a completely separated Y). We obtained a few females of the karyotype XY⁎X and a very few of those gave birth to male pups of the KT 41, XYY⁎X (4 animals/ 5 years). These XYY⁎X males were described as breeders of 41, XXY males. However in our colony those XYY⁎X males either died before puberty or turned out to be infertile owing to testicular mixed atrophy. We assume that in our colony, in contrast to the published model, the sexchromosomal aberrations in the female XY⁎X and in XYY⁎X males influenced meiotic progress more strongly, not resulting in sufficient numbers of fertile animals to achieve XXY males. As an alternative model, we obtained mice with a supernumerary X chromosome (karyotype 41, XXY⁎) by breeding XY⁎ males to XX females resulting in male offspring with a supernumerary X chromosome closely attached to

24

L. Lewejohann et al. / Physiology & Behavior 96 (2009) 23–29

the Y⁎ chromosome designated as 41, XXY⁎. While the XXY model has been examined as a model for Klinefelter disease [8], XXY⁎ mice have not been analyzed in detail so far. Although mice with the karyotype XXY⁎ have two X chromosomes and a part of the mutated Y⁎ chromosome, these mice are phenotypically normal males. This indicates that at least the primary genomic signal, the gene sry (sexdetermining region on the Y chromosome) had been activated, resulting in male sexual determination. It is not clear whether the phenotype is close to the Klinefelter model and/or whether these mice resemble features of XX males [9], a numeric sex-chromosomal aberration for which a reliable mouse model has not yet been described. Human XX males are different from Klinefelter patients, more often exhibiting gynecomastia and smaller body size on average, resembling that of women more closely. To date, all XX males analyzed have been infertile; most of them hypogonadal but epigenetic differences have also been confirmed [10]. So far, no studies have been performed addressing cognition in male patients with the karyotype 46, XXsry. For this study we obtained male mice with the aberrant karyotype 41, XXY⁎ (identical with those male mice previously described as 40, XXY⁎ [11] see below) from our colony as well as age-matched, apparently normal fertile XY⁎ littermates to conduct behavioral phenotyping. Here we wanted to analyze whether male mice with a supernumerary X chromosome and a part of the Y chromosome show differences in a nonconditional test of recognition memory when compared with their male littermates with a normal number of chromosomes. This will help to clarify whether XXY⁎ mice can be used as an easy to obtain model to analyze further the linkage between sex-chromosomal imbalance and cognition processes. In order to assure that preconditions for recognition memory were identical in both groups, we conducted a general health check, and also tested for exploration and anxiety. In males with Klinefelter syndrome as well as in XX males, infertility due to germ cell loss, impaired androgenization, hypogonadism, elevated serum LH levels, reduced testicular weight and Leydig cell hyperplasia have been described [1,10]. To compare physiological qualities of the mice with features reported for human Klinefelter patients or XX males, we recorded body weight and weight of the organs of the reproductive tract, as well as of the brains and pituitaries. Furthermore, we measured serum levels of testosterone and gonadotropins (FSH, LH) in order to compare the endocrine changes related to the different karyotypes. If the XXY⁎ mice resemble the Klinefelter and/or XX male phenotype, an easy to obtain animal model to study genetic mechanisms underlying these diseases as well as to survey possible treatments might be at hand. 2. Materials and methods 2.1. Animals Fifteen male mice exhibiting the karyotype 41, XXY⁎ and fifteen male littermates as controls (karyotype 40, XY⁎) bred of the strain B6Ei.Lt-Y⁎ were obtained from our departmental colony founded by mice imported from Charles River Laboratories (Stock JAX 002021, Sulzfeld, Germany). Wild type male mice of the strain C57Bl/6 for comparative analyses were obtained from Harlan-Winkelmann (Borchen, Germany). 41, XXY⁎ animals are identical to those male mice previously described as 40 XXY⁎[11]. Whereas the linked chromosomes XY⁎ have been interpreted as one chromosome, our Fluorescence In Situ Hybridization (FISH) results revealed a strong Y signal indicative of the presence of a substantial part of the Y chromosome having been translocated to a close association with one of the X chromosomes in these animals (see Fig. 1). We therefore suggest designating the karyotype of these males more precisely as 41, XXY⁎. As controls we used XY⁎ littermates that were proven to be a physiologically valid control group by comparison of morphometric data and endocrine state with that of 15 age-matched C57Bl/6 XY mice (see Table 1). Mice were housed in standard (37 (l)× 21 (w) × 15 (h) cm) cages, with up to five animals of mixed karyotypes per cage. Ear punctures allowed

the discrimination of individual mice but behavioral experiments were carried out with the experimenter being unaware of the karyotypes of the subjects. All animals were kept in a light–dark cycle of 12:12 h. The cages contained a thin layer of wood shavings and paper towels as nesting material. Food (Altromin 1324, Altromin, Lage, Germany) and tap-water were available ad libitum. Room temperature was maintained at 24 °C. Mice were looked after daily, but handled only once a week while transferring them to clean cages. All procedures and protocols met the guidelines for animal care and animal experiments in accordance with national and European (86/609/EEC) legislation (animal licences No. A54/02 and A79/06 RP Muenster). 2.2. Karyotyping The presence of sex chromosomes was analyzed by interphase Fluorescence In Situ Hybridization (FISH) of blood samples. Blood samples of 200–300 µl were drawn via retrobulbar bleeding. 750 µl Biocoll (Biochrom AG, Berlin, Germany; [12]) was used as separating solution. The cells were cracked with KCl and methanol/glacial acetic acid and stored at −20 °C until further processing. Of every sample, 0.5 µl was dropped onto a slide and dried on a heating plate. The slides were denatured in 70% formamide 2× SSC at 68 °C and dehydrated through an ethanol series. Probes used were 1200-XMCY3-02 (X chromosome) and 1189-YMF-02 (Y chromosome; Cambio, Cambridge, UK). Probes were incubated overnight while hybridizing. Afterwards the slides were placed in 50% formamide 2× SSC at 43 °C for 15 min, before being transferred into heated 2× SSC for 10 min and then for another 10 min at room temperature. The slides were counterstained with 30 µl Hoechst 33258 and mounted in Vectashield (Vector Laboratories, Burlingame, USA). Staining was evaluated with a fluorescence microscope (Axiovert 200, Zeiss, Oberkochem, Germany). 2.3. Hormone analysis Serum testosterone was measured by RIA [13]. Each sample was processed in duplicate after extraction with diethyl ether. Intra- and inter-assay coefficients of variation were 4.3% and 5.6%, respectively. The detection limit of the assay was 0.69 nmol/l. Gonadotropin levels were measured in serum samples and in homogenates of pituitaries frozen in liquid nitrogen and subsequently stored at −80 °C until homogenization. Whole pituitaries were transferred in 2 ml Eppendorf tubes filled with 1 ml ice-cold phosphate buffered saline (PBS) and homogenized by applying ultrasound for 30 s. Afterwards 4 ml PBS and 10 µl of protease inhibitor (P8340, Sigma-Aldrich, Munich, Germany) were added and this solution was stored frozen until measurement. Follicle-stimulating hormone (FSH) and Luteinizing hormone (LH) concentrations were measured with an AutoDelfia-ImmunoAssay System (1235 AutoDelfia; LKB Wallac, Turku, Finland, [14]). The intra- and inter-assay coefficients of variation were for FSH 4.3% and 10.4% at 4.8 µg/l respectively and for LH 19% at 0.04 µg/l and b5% above 1 µg/l and 12.5% at 0.24 µg/l and 7.8% at 0.78 µg/l, respectively [15,16]. The FSH assay range was 0.04–25 ng/ml and the LH assay range was 0.02–12.5 ng/ml. 2.4. Organ and tissue collection Animals were anesthetized and killed by decapitation after experiments were completed. Trunk blood was collected and serum was stored at −20 °C for later evaluation. Testes, epididymides, brains and pituitaries were dissected out and weighed. Organs were either fixed in Bouin's fixative or snap-frozen in liquid nitrogen and subsequently stored at −80 °C. 2.5. Histology Organs were fixed for several hours and afterwards routinely embedded in paraffin and cut in 5 µm serial sections. Periodic acid-

L. Lewejohann et al. / Physiology & Behavior 96 (2009) 23–29

25

Fig. 1. A, B: Karyotyping cell nuclei of peripheral blood cells by FISH: left column, green probe marks the Y chromosome (arrows), middle column, the red probe marks the X chromosome (arrowheads), right column overlay; the nuclei are counterstained by DAPI. A: Animals exhibiting the karyotype XXY⁎ were found to present the Y-chromosomal signal in close attachment to one of the X chromosomes. B: XY⁎ littermates exhibited two separate signals. C: Testicular histology: while the testes of the adult XY⁎ controls exhibit full spermatogenesis (right micrograp), the seminiferous epithelium in the tubules of XXY⁎ males lacks all germ cells; all seminiferous tubules reveal Sertoli cell only (SCO; left micrograph). Scale bar 100 µm.

Schiff/haematoxylin staining was used for histological analysis. Histological slides were analyzed with an Axiovert 200 microscope (Zeiss, Oberkochem, Germany) attached to a CCD-camera (Axicom, Zeiss) controlled by specific camera software (Axiovision 3.1, Zeiss). 2.6. General health check Health and neurological status were assessed using a 10-min protocol including tests as described in standard check lists such as SHIRPA [17] and the Fox battery [18]. Animals were weighed, inspected for physical appearance, and underwent neurological testing including acoustic startle, visual placing, grip strength and

reflex functions to ensure that behavioral findings were not the result of deteriorating physical conditions of the animals. 2.7. Barrier test Spontaneous exploratory behavior was measured using the barrier test [19]. A standard sized cage (37 × 21 × 15 cm) was divided by a 3 cm high Plexiglas barrier into two equal compartments. Mice were placed in one of the compartments according to a pseudo-random schedule. Latency was measured either as the time the mouse took to climb over the barrier into the other compartment or as a maximum time of 300 s in case the mouse did not climb over the barrier. The apparatus was thoroughly cleaned after each trial by wiping the surfaces with a

Table 1 Dataset of endpoints analyzed Serum Hormones Bodyweight Head–tail Skull width Bi-testis Bi-epididymis Accessory sex Brain weight Pituitary FSH [g] length [cm] [mm] weight [mg] weight [mg] glands [mg] [mg] weight [mg] [ngl/ml] XXY⁎ XY⁎ controls C57Bl6 XY

LH [ng/ml]

Pituitary hormones Testosterone FSH LH [nmol/l] [ng/pituitary] [ng/pituitary]

29.81 ± 0.5 10.57 ± 0.1 27.85 ± 0.55 10.25 ± 0.1

17.8 ± 0.2 18 ± 0.3

23.5 ± 0.9 162.5 ± 5

88.9 ± 4 90.9 ± 3.4

168.6 ± 14 163.6 ± 5.9

437.8 ± 4.1 440.7 ± 4.6

1.5 ± 0.2 1.3 ± 0.1

85.1 ± 9.0 3.2 ± 0.7 5.6 ± 0.9 46.4 ± 4.8 0.5 ± 0.1 12.0 ± 3

28.43±0.66 10.27 ± 0.1

18.1 ± 0.4

200.2 ± 8.2

102.6 ± 3.8

168.0 ± 7.5

451.1 ± 4.6

1.3 ± 0.1

54.8 ± 5.5 0.5 ± 0.1

9.5 ± 2.5

170.5 ± 26.7 191.5 ± 22.2

671.7 ± 93 409.4 ± 47.6

528 ± 127.8 394.4 ± 48.4

Values are mean ± SEM. Measurements resulting in significant differences between XXY⁎ and XY⁎ and between XY⁎ and C57/Bl6 XY are given in bold (upper and lower row).

26

L. Lewejohann et al. / Physiology & Behavior 96 (2009) 23–29

Fig. 2. Body weight, head–tail length and bi-testicular weight. Data are given as box plots. Each box represents the 25th–75th percentile, and the horizontal line across the box is the median (50th percentile). Whisker lines extending below and above represent the extremes lying within 1.5 times the interquartile range (box height). While body weight and head– tail length were significantly higher, bi-testicular weight was dramatically reduced in XXY⁎ males. Statistics: XXY⁎ n = 15, XY⁎ n = 15; ANOVA ⁎p b 0.05; ⁎⁎p b 0.01, ⁎⁎⁎p b 0.001.

disinfecting agent. The same cleaning procedure applied for the tests described below. 2.8. Elevated plus-maze Anxiety-related behavior was measured by means of the elevated plus-maze [20,21] in which mice had the choice of moving into opposing arms, which were either shielded or open. The maze was elevated 50 cm above the floor and had arms 30 cm long and 5 cm wide. At the beginning of a trial, mice were put into the center of the maze randomly facing one of the arms. Each entry into an open or shielded arm was counted and the time animals spent in either type of arm was measured for 5 min using a Palm handheld computer and proprietary software for behavioral recording. 2.9. Open field In the open-field test [22] mice had the opportunity to explore a square arena (30 by 30 cm, walls 40 cm high) for 3 min. Locomotor activity and the ratio between exploration and fear of open space, as measured by the time spent near the walls or in the center of the arena, were assessed automatically using a tracking system [23].

2.10. Novel object task Memory performance was analyzed with the “novel object task” (NOT) [24]. The test is based upon the tendency of mice to investigate a novel object rather than a familiar one. In a first trial the mice were allowed to explore two similar objects presented in an open-field arena (see above) for 3 min. The five objects to be discriminated in the object recognition task consisted of a biologically neutral material such as plastic or metal, and animals could not move them around in the arena. Objects are not known to have any ethological characteristic for the mice. After 1 h the mouse was again placed into the open field now containing one object similar to both objects presented in the first trial and one novel object. The number of visits and the time spent exploring the objects were recorded using the “NOT” [25] software and a handheld computer. To avoid object or place preferences, place and novelty-status were changed for each object at regular intervals. Fecal boli were removed, and the walls and the floor of the open-field arena were cleaned with a detergent after each tested animal. In order to compare groups, a recognition index was calculated for each individual by dividing the amount of time spent exploring the novel object by the total time of object exploration.

Fig. 3. Endocrine status. Data are given as box plots (see Fig. 1). Testosterone appears reduced in, XXY⁎ males (t = trend) and the LH levels and FSH levels in serum were significantly higher in these animals compared with their littermate controls. Statistics: LH, FSH: XXY⁎ n = 12, XY⁎ n = 11; testosterone: XXY⁎ n = 15, XY⁎ n = 15; ANOVA t = p b 0.1; ⁎⁎p b 0.01.

L. Lewejohann et al. / Physiology & Behavior 96 (2009) 23–29

27

Fig. 4. Novel object task. A) The median time spent exploring known (gray bars) and novel (white bars) objects is given for XXY⁎ and XY⁎ males. Individual performance is illustrated by the connected dots. Statistics: XXY⁎ n = 15, XY⁎ n = 15; paired Wilcoxon test (one-tailed); ⁎⁎p b 0.01. B) The recognition index (time spent with novel object divided by total time) is given as percentage for XXY⁎ males (white box plot) and for XY⁎ controls (gray box plot). The dotted line at 50% indicates the value at which both objects would be explored for a similar amount of time. Statistics: XXY⁎ n = 15, XY⁎ n = 15; unpaired Wilcoxon test (two-tailed); ⁎p b 0.05. C) Correlation with fitted line between recognition indices and testosterone concentrations. Statistics: XXY⁎ n = 15, XY⁎ n = 15; Pearson's product-moment correlation: XXY⁎: p N 0.93, XY⁎: p b 0.01; correlation coefficients are given in the figure. While testosterone levels and recognition indices are correlated in XY⁎, no correlation was found in XXY⁎ animals.

2.11. Statistics The statistical software “R” Version 2.6.2 was used for graphical presentation and statistical analysis [26]. Behavioral data were analyzed by use of non-parametric statistics [27] since several data sets showed a non-Gaussian distribution. For the same reason, graphs are presented as box plots representing the 25th–75th percentiles, and location measures are given as medians (50th percentile). A significance-level (α) of 0.05 was selected. Comparison of two samples was performed using the unpaired two-sample Wilcoxon test. Paired data from the novel object test were analyzed using the paired Wilcoxon test. One-tailed tests were used on the assumption that, if at all, the novel object would be preferred over the known object and only a preference for the novel object would be considered meaningful with regard to object recognition. Physiological measurements that did not violate assumptions of parametric statistics were analyzed using one-way ANOVA (~ Students t test). Pearson's productmoment correlations were calculated for correlation analysis of memory performance with physiological measures.

from further behavioral analysis. However, some mice were lacking whiskers (4 XXY⁎, 5 XY⁎). 3.2. Endocrine state XXY⁎ mice exhibited significantly elevated serum levels of LH in the pituitary (F(1,21) = 6.6245, p b 0.02) as well as in serum (F(1,21) = 13.42, p b 0.01) compared with XY⁎ littermate controls. Serum FSH (F(1,21) = 14.19, p b 0.002) but not pituitary FSH levels were higher in XXY⁎ compared with XY⁎ males (Fig. 3, Table 1). Serum testosterone was lower in males with a supernumerary X chromosome but statistically only a trend (F(1,28); p N 0.052) was found (Fig. 3, Table 1). Additionally, XY⁎ males were compared with C57Bl/6 males in order to validate their suitability as a control group. C57Bl/6 males did not differ from XY⁎ males with regard to those hormones that differed significantly between XY⁎ and XXY⁎ males. Pituitary FSH, however, was significantly elevated in C57Bl/6 males (F(1,26) = 5.84, p b 0.03) while XY⁎ and XXY⁎ did not differ in their pituitary FSH concentrations (Table 1). 3.3. Exploratory and anxiety-related behavior

3. Results 3.1. Phenotypes and general health check Cell nuclei obtained from peripheral blood were analyzed by FISH for the number of sex chromosomes (Fig. 1A, B). Animals exhibiting the karyotype XXY⁎ were found to present three signals of which the Y-chromosomal signal was always observed in close attachment to one of the two X chromosomes, while XY⁎ littermates exhibited two separate signals. The males bearing a supernumerary X chromosome had significantly higher body weight (F(1,28) = 6.88, p b 0.02), a greater head–tail length (F(1,28) = 10.4, p b 0.004) and decreased bi-testicular weight (F(1,28) = 761.4, p b 0.001) (Fig. 2). All other organ weights and body characteristics were not different from those of littermate controls (Table 1). Testicular histology revealed complete absence of germ cells (Sertoli cell only, SCO) from testes of males with karyotype XXY⁎, while all controls had complete spermatogenesis (Fig. 1C) as did all 15 males of the strain C57Bl/6. In addition, C57Bl/6 male mice exhibited higher bi-testicular (F(1,28) = 15.6, p b 0.001) and bi-epididymal (F(1,28) = 5.18, p b 0.03) weights compared with XY⁎ males (Table 1). All animals assigned to behavioral tests passed the health check, showing no severe defects that would have led to exclusion

In the barrier test no statistical difference was detectable between XXY⁎ and XY⁎ mice, indicating no difference in spontaneous exploratory behavior mediated by karyotype. Anxiety-related behavior as measured by the proportion of open arm entries in the elevated plus-maze test did not reveal statistical differences between either group. XXY⁎ mice did not differ from XY⁎ controls regarding locomotor behavior as measured by path length, stops, and velocity in the open-field test; neither did the proportion of time spent in the center of the open field, as another measure of anxiety-related behavior, reveal any significant difference between karyotypes. 3.4. Novel object task In the novel object task XY⁎ male mice significantly preferred the novel object over the known object, indicating recognition memory processes (Fig. 4A). In contrast, the XXY⁎ males neither preferred the known nor the novel object, indicating poor recognition memory (Fig. 4A). This difference between karyotypes was confirmed by calculation of the recognition index (RI) for each individual by dividing the time spent exploring a novel object by the total time spent exploring objects. XY⁎ male mice had a median RI of 0.7 that was well

28

L. Lewejohann et al. / Physiology & Behavior 96 (2009) 23–29

above chance level (0.5) and significantly different from the median RI of XXY⁎ males that was 0.57 (Fig. 4B). Testosterone concentrations correlated significantly with object recognition memory in XY⁎ (R = 0.644, p b 0.01) mice but not in XXY⁎ (R = −0.02, p N 0.9) mice (Fig. 4C). None of the other hormonal and physiological measures showed a comparable correlation with object recognition memory. 4. Discussion We examined XXY⁎ male mice and compared these animals with their XY⁎ littermate controls. Interestingly, XXY⁎ mice were bigger and weighed significantly more than controls. Although body metrics of mice can, if at all, only cautiously be translated to humans, it is noteworthy that a comparable pattern of increased height and weight can be found in Klinefelter patients [1]. In contrast to Klinefelter patients, XX males are significantly smaller than XY and XXY males [9]. Of the organs weighed, we found only testicular weight to be dramatically reduced in XXY⁎ mice. Furthermore those small and firm testes were lacking all germ cells. In addition, the endocrine situation in XXY⁎ mice was hypergonadotropic for LH and serum FSH and the animals appeared hypogonadal, since at least a trend for lowered serum testosterone was found in the XXY⁎ males. These conditions resemble closely the well-documented hypergonadotropic hypogonadism associated with the Klinefelter syndrome in humans [1]. Apart from these phenotypic features, analysis of general and sensory health status revealed no severe health deficits in the experimental animals. This is also in agreement with current knowledge on the health status of Klinefelter patients [28]. XXY⁎ mice did not differ from their fertile XY⁎ littermates with respect to locomotion, exploration and anxiety-related behavior. Different levels of anxiety and/or exploratory behavior may influence performance in the novel object task. Therefore these results indicate no confounding effects that might have disturbed the measurement of object recognition memory. In the novel object task XXY⁎ male mice exhibited no significant object recognition memory and consequently performed significantly worse than XY⁎ males. The test does not use explicit reinforcement, and lengthy training is avoided in order to test inherent rather than acquired cognition capacities. Although we advise caution in translating findings from cognitive abilities of mice to men, it is obviously worthwhile to investigate the underlying principles that may contribute to cognitive deficits in both, XXY⁎ mice and Klinefelter patients. The X chromosome has accumulated a disproportionate number of genes linked to mental functions and is thought to play a crucial general role in intelligence. Xlinked genes are supposedly involved in social-cognition and emotional regulation (for review see [29]). Most X-linked genes are inactivated on one X chromosome in the presence of an additional X chromosome. However, in humans about 15% of X-linked genes escape from silencing [30]. The Klinefelter phenotype is hypothesized to be due to X chromosomal genes that escape this inactivation and thus are expressed in excess [31]. Cognitive deficits, although usually not severe, are well documented for Klinefelter patients [2,3,32]. In adult males tested in a battery of cognitive tests, significant deficits on various cognitive performance scores were found [31]. In accordance with this, in Klinefelter boys, X-linked effects in areas of cognitive strength and talent are suggested to result in impairment in executive skills of social processing [2,3]. Furthermore, it has been hypothesized that in KS the chromosomal imbalance caused by the presence of the supernumerary X chromosome and the hypogonadism might induce alterations in the subcortical pathways that are linked to cognitive deficits [32]. Knowledge about learning and memory capacities of XX males is scarce, owing to the considerably low numbers of patients investigated so far. Notably, the prevalence of Klinefelter patients among intellectual disabled patients was published to be 4 times higher than that in newborn babies while no such pattern was found for XX males suggesting that the affection of learning and memory capacities in XX males might be different [9]. In humans a link between testosterone and spatial cognition has been demonstrated. Testosterone substitution in hypogonadal

men activates a cortical network, resulting in enhanced cerebral glucose metabolism and reflecting improved visuospatial capability [33,34]. In rodent models, testosterone has been shown experimentally to play a role in learning disruption and the androgen receptor seems to play a role in mediating these effects on cognition and physical response in the brain such as dendritic spine formation of the prefrontal cortex and the hippocampus [8,35,36]. In castrated male rats testosterone is involved in nonspatial learning and testosterone administration potentiates disruptions induced by scopolamine [35]. In our mouse model, although not statistically significant, testosterone levels were reduced by about 50% in XXY⁎ animals while serum LH and FSH levels were significantly elevated, both indicating a hypergonadotropic hypogonadal state of XXY⁎ males which appears to resemble closely the endocrine state of some Klinefelter patients [1]. Our results indicate a strong correlation between testosterone and performance in the novel object task in controls but not in XXY⁎ mice. This indicates that higher levels of testosterone in some of the XXY⁎ mice fail to bring about an increase in memory performance. Interestingly, this is in line with observations from human Klinefelter patients who perform worse than controls regardless of testosterone substitution [31]. However whether testosterone levels directly influence non-conditional learning behavior, or whether a disturbance of the androgen receptor is responsible, has to be clarified in further experiments addressing the rescue of this phenotype by testosterone administration. Additionally, we do not exclude a genetic basis of the observed differences and suggest analysis of changes in the expression of X-linked genes associated with recognition memory and learning processes to be considered in future studies. We thus suggest that the basis for cognitive deficits is to be found in structural differentiation of the brain that occurs earlier in life, rather than in the hormonal milieu of the adults. The same mechanisms leading to sexual differentiation in individuals with balanced sex-chromosomal content probably lead to maladjusted brain structures owing to X-linked gene dosage effects or even X-linked imprinting, reviewed in [37]. Lue et al. [8] examined behavioral deficits in 41, XXY mice (possessing a full Y chromosome) by testing conditional learning of a Pavlovian association and found the rate of learning to be significantly slower in XXY mice compared with their XY littermates. They concluded that there is an impaired medial temporal lobe function in XXY mice. Several causes might induce this failure: low plasma testosterone levels, defect of the androgen receptor or overdosed X-linked gene expression in specific brain regions responsible for learning [8]. Taken together, our results are highly consistent with results from the XXY mouse – thought to be a good model for the Klinefelter syndrome – indicating that the part of Y⁎ chromosome present, although closely attached to a X chromosome, is functionally comparable to a fully separated Y chromosome. We believe that the far more easy to generate XXY⁎ mice resemble a suitable model for Klinefelter syndrome. Taking into account that the genetic mechanisms that lead to Klinefelter syndrome are thought to be in the composition of those genes that escape silencing, it is very promising that so far only very few genes have been found to escape silencing in mice [38]. Acknowledgements We wish to thank Martin Heuermann and Günter Stelke for animal caretaking. The authors are indebted to Reinhild SandhoweKlaverkamp, Petra Köckemann and Jutta Salzig for excellent technical assistance. Furthermore we thank Susan Nieschlag M.A. and Dr. Trevor G. Cooper for language editing. The study was supported by the Deutsche Forschungsgemeinschaft, DFG grant No. WI 27-23/1-1 and by a grant of the Medical Faculty Münster (IMF LU 1 2 03 05).

L. Lewejohann et al. / Physiology & Behavior 96 (2009) 23–29

References [1] Lanfranco F, Kamischke A, Zitzmann M, Nieschlag E. Klinefelter's syndrome. Lancet 2004;364:273–83. [2] van Rijn S, Aleman A, Swaab H, Kahn R. Klinefelter's syndrome (karyotype 47,XXY) and schizophrenia-spectrum pathology. Br J Psychiatry 2006;189:459–60. [3] Temple CM, Sanfilippo PM. Executive skills in Klinefelter's syndrome. Neuropsychologia 2003;41:1547–59. [4] Hunt P, Eicher E. Fertile male mice with three sex chromosomes: evidence that infertility in XYY mice is an effect of two Y chromosomes. Chromosoma 1991;100:293. [5] Mroz K, Carrel L, Hunt PA. Germ cell development in the XXY mouse: evidence that a compromised testicular environment increases the incidence of meiotic errors. Hum Reprod 1998;14:1151–6. [6] Mroz K, Hassold TJ, Hunt PA. Meiotic aneuploidy in the XXY mouse: evidence that X chromosome reactivation is independent of sexual differentiation. Dev Biol 1999;207:229–38. [7] Lue Y, Rao PN, Sinha Hikim AP, Im M, Salameh WA, Yen PH, et al. XXY male mice: an experimental model for Klinefelter syndrome. Endocrinology 2001;142:1461–70. [8] Lue Y, Jentsch JD, Wang C, Rao PN, Hikim AP, Salameh W, et al. XXY mice exhibit gonadal and behavioral phenotypes similar to Klinefelter syndrome. Endocrinology 2005;146:4148–54. [9] de la Chapelle A. Analytic review: nature and origin of males with XX sex chromosomes. Am J Hum Genet 1972;24:71–105. [10] Vorona E, Zitzmann M, Gromoll J, Schüring AN, Nieschlag E. Clinical, endocrinological, and epigenetic features of the 46,XX male syndrome, compared with 47, XXY Klinefelter patients. J Clin Endocrinol Metab 2007;92:3458–65. [11] Eicher EM, Hale DW, Hunt PA, Lee BK, Tucker PK, King TR, et al. The mouse Y⁎ chromosome involves a complex rearrangement, including interstitial positioning of the pseudoautosomal region. Cytogenet Cell Genet 1991;57:221–30. [12] Huang GC, Zhao M, Jones P, Persaud S, Ramracheya R, Löbner K, et al. The development of new density gradient media for purifying human islets and isletquality assessments. Transplantation 2004;77:143–5. [13] Chandolia RK, Weinbauer GF, Simoni M, Behre HM, Nieschlag E. Comparative effects of chronic administration of the non-steroidal antiandrogens flutamide and Casodex on the reproductive system of the adult male rat. Acta Endocrinol (Copenh) 1991;125:547–55. [14] Rochira V, Balestrieri A, Faustini Fustini M, Borgato S, Beck Peccoz P, Carani C. Pituitary function in a man with congenital aromatase deficiency: effect of different doses of transdermal E2 on basal and stimulated pituitary hormones. J Clin Endocrinol Metab 1999;87:2857–62. [15] Haavisto AM, Pettersson K, Bergendahl M, Perheentupa A, Roser JF, Huhtaniemi I. A supersensitive immunofluorometric assay for rat luteinizing hormone. Endocrinology 1993;132:1687–91. [16] van Casteren JI, Schoonen WG, Kloosterboer HJ. Development of time-resolved immunofluorometric assays for rat follicle-stimulating hormone and luteinizing hormone and application on sera of cycling rats. Biol Reprod 2000;62:886–94.

29

[17] Rogers DC, Peters J, Martin JE, Ball S, Nicholson SJ, Witherden AS, et al. SHIRPA, a protocol for behavioral assessment: validation for longitudinal study of neurological dysfunction in mice. Neurosci Lett 2001;306:89–92. [18] Fox WM. Reflex-ontogeny and behavioural development of the mouse. Anim Behav 1965;13:234–41. [19] Prior H, Sachser N. Effects of enriched housing environment on the behaviour of young male and female mice in four exploratory tasks. J Exp Anim Science 1995;37:57–68. [20] Pellow S, Chopin P, File SE, Briley M. Validation of open: closed arm entries in an elevated plus-maze as a measure of anxiety in the rat. J Neurosci Methods 1985;14:149–67. [21] Lister RG. The use of a plus-maze to measure anxiety in the mouse. Psychopharmacology (Berl) 1987;92:180–5. [22] Walsh RN, Cummins RA. The open-field test: a critical review. Psychol Bull 1976;83:482–504. [23] Lewejohann L. Digital image processing in behavioral sciences; 2007. http://www. phenotyping.com/digital.html. Date of Access: 11/26/. [24] Ennaceur A, Delacour J. A new one-trial test for neurobiological studies of memory in rats. 1: behavioral data. Behav Brain Res 1988;31:47–59. [25] Lewejohann L. Novel object task; 2007. http://www.phenotyping.com/not.html. [26] R Development Core Team. R: a language and environment for statistical computing; 2007. http://www.r-project.org/. [27] Siegel S. Nonparametric statistics for the behavioral sciences. New York: McGrawHill; 1956. [28] Handelsman DJ, Liu PY. Klinefelter's syndrome—a microcosm of male reproductive health. J Clin Endocrinol Metab 2006;91:1220–2. [29] Skuse DH. X-linked genes and mental functioning. Hum Mol Genet 2005;14:27–32. [30] Brown CJ, Greally JM. A stain upon the silence: genes escaping X inactivation. Trends in Genetics 2003;19:432–8. [31] DeLisi LE, Maurizio AM, Svetina C, Ardekani B, Szulc K, Nierenberg J, et al. Klinefelter's syndrome (XXY) as a genetic model for psychotic disorders. Am J Med Genet B Neuropsychiatr Genet 2005;135B:15–23. [32] Itti E, Gaw Gonzalo IT, Pawlikowska-Haddal A, Boone KB, Mlikotic A, Itti L, et al. The structural brain correlates of cognitive deficits in adults with Klinefelter's syndrome. J Clin Endocrinol Metab 2006;91:1423–7. [33] Zitzmann M, Weckesser M, Schober O, Nieschlag E. Changes in cerebral glucose metabolism and visuospatial capability in hypogonadal males under testosterone substitution therapy. Exp Clin Endocrinol Diabetes 2001;109:302–4. [34] Zitzmann M. Testosterone and the brain. Aging Male 2006:195–9. [35] Leonard ST, Moerschbacher JM, Winsauer PT. Testosterone potentiates scopolamine induced disruptions of nonspatial learning in gonadectomized male rats. Exp Clin Psychopharmacol 2007;15:48–57. [36] Leranth C, Szigeti-Buck K, Maclusky NJ, Hajszan T. Bisphenol A prevents the synaptogenic response to testosterone in the brain of adult male rats. Endocrinology 2007:988–94. [37] Davies W, Wilkinson LS. It is not all hormones: alternative explanations for sexual differentiation of the brain. Brain Res 2006;1126:36–45. [38] Tsuchiya KD, Willard HF. Chromosomal domains and escape from X inactivation: comparative X inactivation analysis in mouse and human. Mamm Genome 2000;11:849–54.