Impaired social memories in 129P2 inbred mice are rescued by reduced Csk expression


  • L. Sinai,

    Corresponding author
    1. Mount Sinai Hospital Research Institute
    2. Molecular & Medical Genetics, Institute of Medical Science, University of Toronto, Toronto
      L. Sinai, Samuel Lunenfeld Research Institute, Mount Sinai Hospital, 600 University Avenue, Room 860, Toronto, Ontario, Canada M5G 1X5. E-mail:
    Search for more papers by this author
  • R. Mathew,

    1. Northern Ontario School of Medicine, Laurentian University, Sudbury, ON, Canada
    Search for more papers by this author
  • J. C. Roder

    1. Mount Sinai Hospital Research Institute
    2. Molecular & Medical Genetics, Institute of Medical Science, University of Toronto, Toronto
    Search for more papers by this author

L. Sinai, Samuel Lunenfeld Research Institute, Mount Sinai Hospital, 600 University Avenue, Room 860, Toronto, Ontario, Canada M5G 1X5. E-mail:


The C-terminal Src kinase (Csk) is an essential signaling factor guiding central nervous system (CNS) development. In the adult brain, Csk-mediated control of Src may also modulate glutamatergic synaptic transmission and N-methyl-d-aspartate receptor (NMDAR)-dependent synaptic plasticity. The regulation of N-methyl-d-aspartate (NMDA)-dependent plasticity by a myriad of kinase cascades has been investigated intensively during spatial and fear learning, while little is known about the regulatory kinases and role of NMDA-dependent plasticity during equally critical forms of social learning. We assessed social memory in Csk(+/+) and Csk(+/−) mice backcrossed onto 129P2, an inbred strain with wild-type impairments in social memory. Reduced Csk expression in Csk(+/−) mice was associated with increased NMDAR subunit 2B (NR2B) phosphorylation in the amygdala (AM) and olfactory bulb (OB), and with markedly improved social recognition memory and social transmission of food preference (STFP). In contrast, phosphorylation of NR2B was only slightly increased in the hippocampus of 129P2/Csk(+/−) mice, and the poor spatial object recognition memory of wild-type 129P2/Csk(+/+) mice was not rescued by reduced Csk expression. The Csk pathway appears to be a critical signaling cascade regulating social learning and memory, and presents a possible therapeutic target in diseases such as autism that are characterized by aberrant social behaviors.

The C-terminal Src kinase (Csk) acts to suppress Src family kinase activity and serves as a critical regulator of Src-mediated signaling in vivo. Genetic ablation of Csk leads to embryonic lethality (E9-10) and to severe defects in neuronal tube formation (Imamoto & Soriano 1993). The expression of Csk is down-regulated in the adult brain, with the exception of the olfactory bulb (OB)(Kuo et al. 1997). Despite this restricted adult expression, Xu et al. (2008) showed that Csk regulates excitatory synaptic transmission and plasticity in the hippocampus, and interacts with N-methyl-d-aspartate receptor (NMDAR) subunit 1 (NR1) in crude synaptosomal membrane fractions. Moreover, a recombinant Csk vector depresses NMDAR responses in acutely isolated CA1 pyramidal neurons, an effect that is blocked by the Src (40–58) peptide inhibitor (Imamoto & Soriano 1993).

Recombinant Src increases the open probability of NMDARs (Yu et al. 1997), and can increase NMDAR-mediated currents in acutely isolated CA1 neurons (Huang et al. 2001; Lu et al. 1998) by phosphorylating receptor subunits NR2A and NR2B (Salter & Kalia 2004). Furthermore, the intracellular application of both anti-Csk and Src activator peptide leads to enhanced NMDAR and α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor (AMPAR) currents and the enhancement of subsequent long term potentiation (LTP) (Xu et al. 2008). A link between Src kinase activity and several forms of learning and memory has been established (Kalia et al. 2004; Purcell et al. 2003). We surmised that reduced Csk expression in the adult brain could lead to enhanced NMDAR phosphorylation and enhanced NMDAR-dependent learning and memory. While most studies investigating the relation between kinase signaling, NMDAR function and behavior have focused on learned fear, object recognition or spatial memory, various forms of social behaviors, including social recognition and transfer of food preference, are also vital for survival. We characterized social memory in male 129P2 mice in several different amygdala- (AM) and olfactory-dependent social learning tasks (Richter et al. 2005).

Social communication in rodents is based on odors produced by exhaled gases, pheromones, feces and urine (Sanchez-Andrade & Kendrick 2009). The olfactory system is closely connected to both the hippocampus and AM, two neural structures central for social and emotional behaviors. The social recognition pathway starts with the detection of pheromones by the vomeronasal organ that projects to the accessory olfactory bulb (AOB) and then to the medial (MeA) and cortical (CoA) nuclei of the AM (Richter et al. 2005). Olfactory recognition memories require the activation of ionotropic glutamate receptors [NMDA (N-methyl-d-aspartate) and AMPA], followed by nitrous oxide release (Gao et al. 2009, Sanchez-Andrade et al. 2005). Blocking either NMDARs or AMPARs in the AM or OB will disrupt long-term memory formation (Brennan & Keverne 1997)

Several inbred strains, including C3H/HeJ, AKR/J, A/J and 129S1/SvImJ, do not show any preference for social novelty (social discrimination task) (Moy et al. 2007). While we have shown that 129P2 exhibited normal olfaction, they showed impaired social discrimination, social recognition, as well as poor social transmission of food preference (STFP). These aberrant olfactory social behaviors were rescued in 129P2/Csk(+/−) mice, possibly by Src-mediated enhancement of NMDAR activity and synaptic plasticity.

Materials and methods


The 129/Sv-Csktm1Sor/J mice were a kind gift from X. Yu (University of Toronto), and were originally purchased from the Jackson Laboratory (Bar Harbor, ME). These mice were crossed for 10 generations to 129P2 mice purchased from Harlan Laboratories (Mississauga, Canada).

Groups of 4–5 littermates were housed by sex in polycarbonate cages, and given ad libitum sterile food (Purina mouse chow) and water. The vivarium was maintained under controlled temperature (21°C ± 1°C) and humidity (50–60%), with a 12-h diurnal cycle (lights on: 0700–1900 h). Experiments were conducted in behavioral testing rooms during the standard light phase (between 0900 and 1400 h) using male mice only. The male juvenile mice used for long-term social recognition studies (Section Long-term social recognition) were less than 5 weeks of age. In most experiments, these juveniles were from the same litter and a particular juvenile was used once and then not used for any additional behavioral tests. Other male juvenile mice of the same background were obtained from our own breeding pairs and were used for multiple behavioral tests (Section Text procedures). Age- and size-matched 129P2 male mice were used for the sociability and social discrimination tasks.

The behavioral procedures were performed under dim light conditions (∼50 lux), except anxiety tasks that required brighter light (∼200 lux). All procedures were approved by the Animal Management Committee of Mount Sinai Hospital and Toronto Centre for Phenogenomics in compliance with the requirements of the Province of Ontario.

Western blot analysis

Polyclonal antibodies against pY1472 NR2B and native NR2B were purchased from Millipore (Billerica, MA, USA) and Santa Cruz (Santa Cruz, CA, USA) respectively. Preparation of synaptosomal fractions (P2) from AM, OB and the hippocampus, and immunoblotting were performed as described previously (Xu et al. 2008). Briefly, different brain areas were isolated and homogenized in buffer containing sucrose (320 mm), Tris-HCl (10 mm, pH 7.4), NaHCO3 (1 mm, pH 7.4), ethylenediaminetetraacetic acid (EDTA, 2 mm), sodium orthovanadate (1 mm) and 1% (v/v) protease inhibitor cocktail (Santa Cruz). Homogenate (H) was centrifuged at 1000 g for 15 min to remove nuclei and other large debris (P1). The supernatant (S1) was centrifuged at 10 000 g for 30 min to obtain a crude synaptosomal pellet fraction (P2). For densitometric quantification of Western blots, the immuno-reacted protein bands were analyzed by ImageJ software version 1.44d (Rasband, W.S., ImageJ, U. S. National Institutes of Health, Bethesda, MD, USA,, 1997–2005) and expressed relative to a standard sample.

Test procedures

Mice were used in multiple (six) behavioral tests with the exception of the juveniles used in long-term social recognition memory test (Section Long-term social recognition). The order of testing for behavioral experiments was the same for all mice: (1) open field locomotion and anxiety at 6–7 weeks, (2) elevated plus maze at age 7–8 weeks, (3) olfactory test with buried food at age 8–9 weeks, (4) social choice and social novelty tasks at age 9–11 weeks, (5) long-term social recognition memory at 11–12 weeks, (6) social transmission of food preference at 12–13 weeks and (7) displaced object recognition. We designed this order to test naÏve animals in anxiety-like behaviors. During tests 1–3, animals were also handled daily and habituated to the examiner.

Open field, elevated plus-maze and olfactory tests

Motor activity was measured in the open field for 30 min and quantified by total distance traveled and number of rearings. The open field test was performed in a transparent plexiglass open field (41 × 41 × 31 cm). Anxiety-like behaviors were measured by time spent in the center (31 × 31 cm) of the open field relative to the periphery, and by the time spent in the open arms of an elevated plus maze relative to the total time of 5 min (% of total time). The elevated plus maze consisted of two open arms (25 × 5 cm; 70 lx) and two closed arms (25 × 5 × 30 cm); 1.3 lx) extending from a central platform (5 × 5 cm) and elevated 50 cm from the ground. These procedures have previously been described in detail (Labrie et al. 2009).

Olfaction was assessed using a buried-food procedure following food deprivation as previously described (Clapcote et al. 2007). In the experiment, clean polycarbonate cages (30 × 17 × 12 cm) with fresh corncob bedding were used for each subject. One piece of Lab Diet rodent chow (PMI Nutrition International, Brentwood, MO) was buried in a random location beneath an evenly distributed layer of 3 cm of corncob bedding. The latency to find the buried food was recorded, with a maximum period of 10 min.

Social choice and social novelty tasks

We used the social approach task that was originally designed by Crawley (Nadler et al. 2004) to assess social choice and preference for social novelty. In this task, the tendency to spend time with another mouse vs. a novel inanimate object (an empty wire cage) and the preference to explore a novel mouse over a pre-exposed mouse are measured. The apparatus consisted of a clear Plexiglas box (53 cm long × 25.6 cm wide × 23 cm high) divided into three chambers; the outer chambers were 19.5 cm in length and the central chamber was 13 cm in length. The outer chambers were divided from the central chamber by clear plexiglass partitions (7.3 cm wide × 23 cm high) containing an opening (11 cm wide × 23 cm high). An empty wire cage (Galaxy Cup, Spectrum Diversified Designs, Inc., Streetsboro, OH) was used to hold a stranger mouse. The wire cage was 11 cm in height, with a bottom diameter of 10.5 cm and bars spaced 1 cm apart. A beaker full of water was placed on top of the cage to prevent it from moving. This cage was located in the center of one of the two outer chambers while the other outer chamber had an identical but empty wire cage. The test consisted of four experimental trials. First, the test mouse was placed in the middle chamber to explore the arena. After 5 min, the plexiglass partitions were removed to allow the mouse to freely enter the outer chambers. An observer measured the amount of time spent in each compartment and the number of entries into each compartment. Immediately after this habituation period, the mouse was placed into the center arena and the partition was reset. An unfamiliar mouse (mouse 1) was enclosed in the wire cage and placed in one of the outer chambers with the location (right or left outer chamber) alternated across subjects. The empty wire cage was placed in the middle of the opposite outer chamber. Following the placement of the wire cages, the partitions between the inner and outer chambers were again removed and the test mouse was allowed to explore the entire arena for 10 min. Time spent in each chamber, the number of entries and time sniffing each wire cage (a sniff was defined when the mouse snout was as close as <1 cm of the wire cage) were measured by an observer using Noldus software (Observer 5.0, Noldus Information Technology, Wageningen, The Netherlands).

At the end of this 10-min trial, each mouse was tested in a fourth session for 10 min to measure social preference for a novel mouse. A new unfamiliar mouse was placed in the wire cage that had been empty during the previous session. Following the placement of the wire cages containing the familiar and unfamiliar mouse, the partition between the inner and outer chambers were removed and the mouse was allowed to explore the entire arena. The test mouse could choose between the first familiar mouse (mouse 1) and the novel unfamiliar mouse (mouse 2). The time spent investigating the two mice, the time spent in each chamber and the number of chamber entries were measured.

Long-term social recognition

For long-term social recognition memory experiments, we used a procedure described previously (Kogan et al. 2000). Briefly, adult male mice were housed separately prior to an experimental session. After 15 min, a male juvenile mouse was added to the cage for an initial interaction trial of 2 min. Twenty hours later, either the same or a novel juvenile was placed back into the adult's cage for a 2-min test trial. Social investigation of the juvenile, consisting mainly of sniffing and licking of the anogenital region of the juvenile, was monitored continuously by an observer who scored the duration of investigative behavior with Noldus software (Observer 5.0, Noldus Information Technology). The social investigations included direct contact with the juvenile while inspecting any part of the body (such as grooming, licking, sniffing of the mouth, ears, tail, anogenital area), and close following (within 1 cm) of the juvenile.

Social transmission of food preference

Social transmission of food preference (STFP) was conducted as described previously (Kogan et al. 2000). Briefly, all mice were habituated for 2 days to eat only powdered rodent chow (Teklad rodent diet, Harlan Laboratories, Mississauga, Canada) from plastic jars placed in the opposite corners of the home cage. A demonstrator mouse (wild type, WT) was randomly chosen from each home cage of 4–5 mice and food deprived in a separate cage for 22 h but with free access to water. The next day, each demonstrator mouse received powdered rodent chow mixed with either 2% cocoa or 1% cinnamon for 1 h, or until at least 0.2 g of powdered food was consumed; one half of these demonstrators received cocoa- and the other half received cinnamon-flavored food. The demonstrator mice were then returned to their home cages to interact with the observer mice. After 30 min of interaction, the demonstrator mice were again separated from the observer groups. The observer mice were then food deprived overnight with free access to water and tested 24 h later for food preference. Each observer mouse was single caged and habituated to the novel environment. After 15 min the observer mice were given a free choice between powdered rodent chow food flavored with 2% cocoa or 1% cinnamon. The positions of the food jars with the cued flavors were balanced between cages to control for possible place preference. After 1 h, the food jars were removed and the weight of food eaten from each jar was determined. The weight of cued-flavored food eaten over the total weight consumed (both flavors) was used as an indicator of social transfer of food preference.

Spatial object recognition

The spatial object recognition task was performed in the same arena as the open field. We used four similar (upside down) funnels as objects to prevent mice from climbing. Mouse behavior was recorded and analyzed using the Observer 5.0 Noldus software (Noldus Information Technology). The testing procedure was adapted from previously described protocols (Roullet et al. 1996). On test day, each mouse was placed individually in the center of the empty open field for a 5-min habituation session. The mouse was then placed in a holding cage for 2 min. The four objects (funnels) were placed in specific positions, two near each corner of the open field and the other two near the center. The mouse was placed in the center of the open field and allowed to explore the objects for a 15-min session (training phase). Object exploration was measured by recording the time spent exploring the funnels across sessions. A mouse was considered to be exploring an object if its snout was in contact with the object. At the end of the training phase, the mouse was again placed in the holding cage for 3 min, and the position of two objects was changed from the center to periphery to assess the response to spatial displacement. The mouse was then returned to the open field, and the time spent exploring the displaced and non-displaced objects was recorded for 5 min. Reaction to a spatial change (preference index) of the objects was determined by the time spent exploring the two moved (displaced) objects over the total time spent exploring all objects (displaced and non-displaced).

Statistical analysis

All the treatment group means, in both behavior tasks and western blot results, were compared by one-way analysis of variance (anova). When the anova detected significant treatment effects, pair-wise differences between means for a given variable were evaluated using Tukey's post hoc multiple comparison test with significance set at P < 0.05. All statistics were calculated using MINITAB for Windows 13.32 (Minitab Inc., State College, PA). All values reported in the text and figures are expressed as mean ± standard error of mean (SEM).


Western blots

The expression of Csk was measured in the OB, AM and hippocampus in both WT (+/+) and heterozygous (+/−) Csk adult mice. Expression of Csk was detected in all three tissues but was highest in the OB (Fig. 1a) as shown previously (Kuo et al. 1997). The Csk kinase can reduce NMDA receptor currents through down-regulation of Src-mediated receptor phosphorylation (Imamoto & Soriano 1993), suggesting that reduced Csk could increase phosphorylation of NMDAR subunits. Indeed, densitometric quantification of the Western blots probed using phospho-anti-pY1472 (Fig. 1b) showed that reduced Csk was associated with enhanced basal phosphorylation of NR2B at Y1472 compared with WT mice in both the OB (WT intensity: 43.45 ± 5.670 N = 6; Csk(+/−): 71.85 ± 9.181, N = 6; F1,10 = 6.93, P < 0.05) (Fig. 1c) and AM (WT intensity: 39.41 ± 7.250, N = 4; 129P2/Csk(+/−): 74.22 ± 11.37 N = 4; F1,6 = 6.67, P < 0.05) (Fig. 1d) but only by 4% in the hippocampus (WT intensity: 93.15 ± 0.5674, N = 4; 129P2/Csk(+/−): 97.38 ± 0.9875, N = 4; F1,6=13.77, P < 0.05) (Fig. 1e).

Figure 1.

Representation of Csk expression and NR2B phosphorylation in different brain regions. Csk protein expression levels in the olfactory bulb, amygdala and hippocampus in WT and heterozygote Csk mice (a). NR2B phosphorylation at residue Y1472 is increased in the amygdala and the olfactory bulb in 129P2/Csk(+/−) mice (b). NR2B phosphorylation in the olfactory bulb (c) NR2B phosphorylation in the amygdala (d) NR2B phosphorylation in the hippocampus (e). NR2B phosphorylation levels in the olfactory bulb, amygdala and hippocampus in wild-type and heterozygote Csk mice. *P < 0.05 and **P < 0.01.

Activity in an open field

Exploration in an open field can assess both motor capacity and affective state. We measured total distance traveled and number of rearings as indices of motor ability. The 129P2/Csk(+/−) did not differ from WT littermates in total distance traveled (Fig. 2a), or in the number of rearings (Fig. 2b).

Figure 2.

Activity measurements are similar between 129P2/Csk(+/+) and129P2/Csk(+/−) mice. Each data point represents activity for 5 min. (a) Total distance traveled and (b) number of rearing.

Elevated plus maze and open field

The time spent in the center of the open field indicated no difference in overall activity (WT: 0.7481% ± 0.3658, N = 18; 129P2/Csk(+/−): 1.560% ± 0.6719, N = 14; F1.30 = 1.26, P > 0.05) (Fig. 3a). However, the 129P2/Csk(+/−) mice spent significantly more time exploring the open arms of the elevated plus maze than WT mice (WT = 6.547 ± 2.688%, N = 12; Csk(+/−) = 24.04 ± 6.884%, N = 11; F1,21 = 5.96, P < 0.05), suggesting a decrease in one anxiety-like phenotype in Csk(+/−) mice (Fig. 3b).

Figure 3.

Anxiety-like behavior is decreased in the elevated plus maze but not in the open filed. Percentage time spent in the center of the open field is not significantly different between groups (a). However, 129P2/Csk(+/−) mice spent more time in the open arm on the elevated plus maze when compared with 129P2/Csk(+/+) *P < 0.05 (b). Each time point represents activity during 5 min.

Social approach tests

Social choice and social novelty tests

The social choice test assesses the general tendency of a mouse to approach a conspecific under nonthreatening conditions relative to the time spent exploring an inanimate object. In the three-compartment chamber, we measured the total time spent sniffing a wire cage containing a stranger mouse in one compartment and the time spent sniffing an identical but empty wire cage in the opposite compartment (Moy et al. 2008). We observed a significantly higher preference for social approach (mouse) over exploration of the empty wire cage (object) in both WTs (stranger mouse: 157.6 ± 26.85 seconds; object: 84.78 ± 15.89 seconds; N = 14, P < 0.05, F1,26 = 5.45) and Csk(+/−) mice (stranger mouse: 169.1 ± 31.09 seconds; object: 57.47 ± 13.86 seconds; N = 12, P < 0.005, F1,22 = 10.75) (Fig. 4a).

Figure 4.

Short-term social recognition was rescued in 129P2/Csk(+/−) mice Time spent sniffing each cage during the test for (a) sociability and (b) preference for social novelty in 129P2/Csk(+/+) and129P2/Csk(+/−) mice. Both groups showed a preference for a social subject over a nonsocial empty cage (a). Only 129P2/Csk(+/−) showed a preference for social novelty (b). *P < 0.05 and **P < 0.01.

In the test for social novelty preference, only 129P2/ Csk(+/−) mice showed a significant preference for the novel stranger (mouse 2) over the pre-exposed familiar mouse 1 as evidence by the increase in time spent sniffing the wire cage containing the novel mouse. In fact, 129P2 WT mice did not appear to distinguish between the first and second mouse as indicated by interaction time (familiar mouse 1: 73.00 ± 9.388 seconds; novel mouse 2: 80.08 ± 11.72 seconds; N = 22, P > 0.05, F1,24 = 1.71), suggesting an impairment in short-term social recognition memory in the 129/P2 inbred strain. This deficit was rescued in Csk(+/−) mice (familiar mouse 1: 65.82 ± 17.98 seconds; novel mouse two: 138.2 ± 20.69 seconds; N = 11, P < 0.05, F1,20 = 6.97) (Fig. 4b).

As previously reported (Moy et al. 2004), the number of entries into each outer side chamber (containing a mouse or empty cage) was the least sensitive index of social approach in the choice tests (Table 1), and is usually used as a control measurement for exploration of the new environment. The number of entries into each chamber was not different between genotypes across the three different sessions. The entry measure provided evidence that all of the subjects tested in this study explored the social test box. Therefore, the lack of preference for social novelty in 129P2/Csk(+/+) could not be attributed to hypoactivity (Table 1).

Table 1.  Number of entries into each side chamber during the test for sociability and preference for social novelty recognition in 129P2/Csk(+/+) and129P2/Csk(+/−) mice
Sessions Number of entries
  1. There are no significant differences in number of entries among groups.

HabituationChamber 16.577.5
 Chamber 26.216
 Center chamber12.7813.16
SociabilitySocial chamber23.8523.08
 Nonsocial chamber16.2818.16
 Center chamber16.3520.25
Short-term social novelty recognitionSocial chamber 118.5716.75
 Social chamber 220.0720.58
 Center chamber18.2118.75

Social recognition test

The social recognition test relies on the known reduction in interaction time between two mice over repeated presentations and increased interactions between unfamiliar conspecifics. In this version of the social recognition test, an adult mouse was first habituated to a new home cage for 15 min, followed by a 2-min presentation of a novel juvenile mouse (presentation 1, P1). The adult mouse was then re-exposed to the same juvenile 24 h later (presentation 2, P2). There was a significant reduction in the duration of investigation time between P1 and P2 in the 129P2/Csk(+/−) mice (P1: 94.93 ± 5.018 seconds; P2: 67.96 ± 5.262 seconds; N = 12, P < 0.005, F1,22 = 13.76) but not in the 129P2/Csk(+/+) WT mice (P1: 89.12 ± 5.519 seconds, P2: 87.41 ± 5.881 seconds; N = 16, P > 0.05, F1,30 = 0.04), indicating that 129P2/Csk(+/−) mice were better able to recognize the pre-exposed juvenile (Fig. 5). As a control, we examined the response to a novel (non-pre-exposed) juvenile presented after P2. In this test, there was no significant reduction in the investigation duration 129P2/Csk(+/+) WT (93.29 ± 4.931 N = 8, P > 0.5), and 129P2/Csk(+/−) (100.6 ± 5.483 N = 8, P > 0.3, anova) (Fig. 5), indicating that the reduction at P2 was specific to familiar juveniles. Thus, social recognition memory was improved in 129P2/Csk(+/−) mice.

Figure 5.

Long-term social memory is impaired in 129P2(Csk+/+) mice and restored in 129P2(Csk+/−) littermates. (a) Familiar juveniles were exposed to both groups at test trials. There was significant reduction in investigation duration, indicating long-term social memory in the 129P2/(Csk+/−) mice, **P < 0.01 but not in the 129P2/(Csk+/+) mice.

Social transmission of food preference

Social transmission of food preference (STFP) is an innate form of olfactory memory that allows mice to gage food safety by the health of conspecific demonstrators (Drew et al. 2007). The observer mouse interacts with a healthy demonstrator mouse that has recently eaten a novel food. During this phase, there were no differences in the interaction time between observer and demonstrator mice of different genotypes (the amount of time spent in mouth-to-mouth contact): 129P2/Csk(+/+) (19.33 ± 0.8819 N = 9) and 129P2/Csk(+/−) WT (19.78 ± 1.024 N = 9). When observer mice are presented with a choice between the food eaten by the demonstrator and another novel food, observer mice prefer the food eaten by a healthy demonstrator. Mice were presented with two unfamiliar flavors of powdered food (rodent chow flavored with 2% cocoa or 1% cinnamon). The 129P2/Csk(+/−) mice strongly preferred the flavor previously consumed by the demonstrator mouse (the cued food) as measured by weight consumed in a free choice test (80.45 ± 5.468% of food intake, N = 12, P < 0.05, F1,21 = 7.31), while WT mice showed no preference for the cued food (47.03 ± 11.47%, N = 11, P > 0.05) (Fig. 6a).

Figure 6.

Social olfactory memory enhancement in 129P2(Csk+/−) mice measured by the social transmission of food preference test. 129P2(Csk+/−) mice showed and enhanced preference for familiar food, but 129P2(Csk+/+) did not show any preference. Data are presented as of percent preference intake of familiar food. *P < 0.05.

Buried-food test

As the social memory and STFP tests rely on the ability of mice to discriminate odors, it was important to determine if all mice had similar olfactory function. For this purpose, we measured the latency to find a buried-food pellet.

Both genotypes were able to find the buried food in equal time (WT: 2.255 ± 0.3755 seconds, N = 12; 129P2/Csk(+/−) = 1.878 seconds ± 0.1606 N = 8; P > 0.05, F1,18 = 0.61). Thus, differences in olfactory-dependent social memory tests were not due to differences in olfactory sensitivity (Fig. 6b).

Displaced object recognition

In the displaced object recognition task, many strains of mice, including C57Bl/6, will spend significantly more time investigating familiar objects that have been displaced from an earlier encounter, indicating a memory for the spatial organization of objects. Wild-type 129P2 mice exhibited impaired detection of a spatial change in previously explored objects, and this deficiency was not rescued in 129P2/Csk(+/+) mice. An initial 15-min free exploratory period showed no innate preference for any individual objects in either WT or Csk(+/−) 129P2 mice (Fig. 7a) or in the positive control strain C57Bl6/6 (data not shown). In contrast to C57Bl/6 mice, neither 129P2 genotype showed any exploratory preference for the displaced over the non-displaced objects, indicating that this particular deficit was not rescued by reduced Csk expression (Fig. 7b).

Figure 7.

Reduced Csk expression can not restore spatial object recognition deficit in 129P2(Csk+/+) mice. (a) Mean duration (seconds) of object exploration in the training phase (10 min) of object recognition. (b) Mean time (seconds) spent exploring the displaced and non-displaced objects in the spatial change session. ***P < 0.001.


Fear conditioning and spatial navigation are by far the most frequently studied rodent behaviors (Crawley 2007), and much is known about the neural substrates, signaling cascades, and the synaptoplastic mechanisms that underlie these learned behaviors. Associating places or cues with food or danger has an obvious survival benefit. In addition, mammals learn necessary survival skills regarding food or danger from other members of the species through social learning, but the neuronal pathways and modulatory signals have not been studied extensively like other forms of learning and memory. Our findings suggest that decreased Csk expression leads to enhanced NR2B phosphorylation in brain regions important for NMDA-dependent social memories. These social memory paradigms may help define behavioral endophenotypes relevant to the study of aberrant social behavior in humans.

Enhanced NR2B phosphorylation could decrease the threshold for learning. Indeed, the Src family kinases are a well-studied example of a neuronal signal that enhances NMDA activity, NMDA-dependent synaptic plasticity and memory under various learning situations. The C-terminal Src kinase (Csk) acts to inhibit Src, suggesting that Csk inhibition or genetic ablation could improve memory through Src disinhibition. To test whether Src kinase disinhibition could improve behavioral phenotypes in the 129P2 background strain, 129P2/Csk(+/+) and 129P2/Csk(+/−) littermates were first compared on tests of motor activity, anxiety, and olfactory sensitivity, as these traits or abilities are either necessary for social behavior (olfaction) or greatly influence performance (motor activity, anxiety). While 129P2/Csk(+/+) and 129P2/Csk(+/−) mice exhibited similar mobility and anxiety in the open field and similar olfactory thresholds, NMDA-dependent social choice, novelty preference/discrimination, social recognition and transfer of food preference were all rescued by reduced Csk activity. Concomitant with improved social memory, reduced Csk activity in Csk(+/−) mice was associated with elevated NR2B phosphorylation at Y1472, possibly leading to increase synaptic plasticity in relevant neuronal circuits within the OB-AM pathway. In contrast, the hippocampus-dependent displaced object recognition task was not rescued, nor was hippocampal NR2B tyrosine phosphorylation greatly increased, suggesting that NMDA phosphorylation could enhance certain forms of learning and underscoring the possible role of amygdalar NR2B phosphorylation in social memory.

The 129P2/Csk(+/−) mice showed a decreased anxiety-like phenotype in the elevated plus maze but not in the open field test. Previous studies suggested that an imbalance between excitatory and inhibitory neurotransmitter systems leading to glutamatergic hypoactivity can explain fearfulness in olfactory bulbectomised OB rats (Kelly et al. 1997). Enhanced NR2B phosphorylation did reduce innate anxiety in Csk heterozygotic mice in the elevated plus maze. However, no effect was observed in the open field. Nonetheless, general anxiety did not appear to contribute to the reduced social behaviors in WTs as they showed the same exploratory tendencies in tests of social behavior and object recognition.

Olfaction is thought to be one of the most important senses for rodents, governing social recognition and discrimination, foraging, and protection against toxic ingestion. Wild-type 129P2 mice were impaired in several dimensions of sociability, neither preferring social novelty nor showing long-term social recognition memory. Thus, while the impairment in novelty detection in 129P2 mice could be explained by a lack of interest for novelty in general, it is possible that impaired social discrimination could contribute to this impairment. In addition, it is important to consider that the buried-food test is a gross measurement for main olfaction which may not be able to detect subtle but important accessory olfactory deficits important for social recognition. An olfactoryhabituation/dishabituation task could determine whether the animal can detect and differentiate different odors, including both social and nonsocial odors.

Down-regulation of Csk rescued the aberrant STFP and both impaired social discrimination and long-term olfactory social memory in the 129P2 inbred strain, possibly by increasing the NR2B phosphorylation in the AM and OB. Indeed, both sensory memory tasks are known to be NMDA-dependent, indicating that Csk is an important regulator of NMDA-dependent olfactory memory formation. Similarly, forebrain-specific over-expression of the NR2B subunit in the AM, the CA1 region of the hippocampus, the striatum and the cortex of transgenic animals resulted in superior memory on a variety of tasks, including STFP (White & Youngentob 2004). Because NR2B phosphorylation was only slightly increased in the hippocampus and could not rescue the aberrant displaced object recognition memory in mice, we suggest that reduced Csk protein levels specifically upregulated NMDAR function in the AM and OB, resulting in enhanced olfactory social recognition in 129P2/Csk(+/−) mice.

Our study shows the importance of choosing the optimal strain to study the contribution of individual genes to a specific behavior or group of behaviors. We chose a strain that is a poor performer in several tasks of social memory so that a possible rescue could be observed. Similarly, strains that perform well in many behavioral tasks are ideal for detecting phenotypic impairments, while average performing strains are beneficial if the behavioral outcome of a genetic manipulation is uncertain. We performed a battery of different cognitive tasks dependent on the hippocampus, AM and OB using the deficient 129P2 strain. Reducing Csk expression could also enhance other behavioral tasks, such as fear conditioning and object recognition. To explore this possibility, the genetic background strain could be changed to one with behavioral deficits in these tasks or the training time could be reduced so that only the superior learner mice could learn the task.

We do not yet know the allelic combinations that confer these abnormal behavioral phenotypes in 129P2 mice. Here we argue that increased NR2B activity rescued learning impairments in this strain, suggesting a role for WT deficits in NMDA-dependent plasticity. In the future, it is essential to perform quantitative trait locus (QTL) and microarray analysis on the 129P2/Csk(+/+) and 129P2/Csk(+/−) mice to identify targets involved in olfactory social memory (Schimanski & Nguyen 2004). Also, it would be of great interest to investigate whether this mouse strain shows impaired synaptic plasticity in the AM and OB and if this impairment is rescued by CSK reduction. These mice could also be used in QTL analysis to determine whether QTLs for AM or OB LTP and QTLs for AM and OB-dependent learning and memory overlap. If this were the case, a causal relationship between LTP and memory might be strengthened.

Differences in neurocellular organization may also underlie these functional deficits in social behaviors and social learning in 129P2 mice. It has been shown that Csk phosphorylates Src family kinases (SFKs), including Src and Fyn, and that this suppresses the ability of Fyn and Src to phosphorylate downstream targets. Expression of Csk is normally highest in the embryonic brain, while Fyn activity is low (Inomata et al. 1994). During myelination, Csk levels decrease substantially compared with those in the embryonic brain. The brain of fyn−/− mice had fewer oligodendrocytes and reduced numbers of myelin sheaths (Sperber et al. 2001). Given the integrated activities of SFKs and their upstream regulators in development, it is possible that Csk plays some indirect role in myelination, a subject of a future study.

In addition to the importance of Csk in synaptic plasticity, learning, and memory, Csk might be an interesting target molecule to regulate social tendencies and behaviors. Similar to our 129P2 mice, 129S1/SvImJ (Moy et al. 2007) mice exhibited a preference for sniffing a social subject vs. a nonsocial object in the three chamber social task, indicating normal social approach. However, these mice failed to show a preference for social novelty in the three chamber social approach task. Failure to display preference for social novelty by this inbred strain could be a relevant endophenotype for the symptoms of autism. Aberrant social behaviors in people with autism may emerge as indiscriminate approaches to strangers in addition to an overall lack of social approach (Loveland et al. 2001). A lack of preference for social novelty in mice might, therefore, model autism-spectrum behaviors where these individuals prefer to remain with familiar people.

Our results provide evidence of enhanced NR2B phosphorylation in the AM and OB, and improved olfactory social memory in 129P2 mice when Csk protein levels were reduced during development and adulthood. Further investigation of this pathway could facilitate the development of new treatments for diseases characterized by aberrant social and social recognition behaviors, such as autism.


This study was supported by grants-in-aid to J.C.R. from the Canadian Institutes of Health Research (CIHR; Grant No.MOP-13239) and the OSOTF studentship award from Samuel Lunenfeld Research Institute.