Sex differences in the activity of signalling pathways and expression of G-protein-coupled receptor kinases in the neonatal ventral hippocampal lesion model of schizophrenia

Abstract

Animals with the neonatal ventral hippocampal lesion (NVHL) demonstrate altered responsiveness to stress and various drugs reminiscent of that in schizophrenia. Post-pubertal onset of abnormalities suggests the possibility of sex differences in NVHL effects that may model sex differences in schizophrenia. Here we demonstrate that novelty- and MK-801-induced hyperactivity is evident in both male and female NVHL rats, whereas only NVHL males were hyperactive in response to apomorphine. Next, we examined the sex- and NVHL-dependent differences in the activity of the ERK and Akt pathways. The basal activity of both pathways was higher in females than in males. NVHL reduces the level of phosphorylation of ERK1/2, Akt, and GSK-3 in both sexes, although males show more consistent downregulation. Females had higher levels of G-protein-coupled kinases [G-protein-coupled receptor kinase (GRK)] 3 and 5, whereas the concentrations of other GRKs and arrestins were the same. In the nucleus accumbens, the concentration of GRK5 in females was elevated by NVHL to the male level. The data demonstrate profound sex differences in the expression and activity of signalling molecules that may underlie differential susceptibility to schizophrenia.

Introduction

Schizophrenia is a severe mental disorder with poorly understood aetiology. One influential concept is the neurodevelopmental hypothesis of schizophrenia. It posits that schizophrenia pathology results from neurodevelopmental abnormalities caused by genetic and/or environmental factors. The neurodevelopmental deficits lead to modifications of critical signalling pathways and ultimately result in clinical manifestations of schizophrenia. The adult onset of behavioural symptoms induced by neonatal insults is reproduced in rats that undergo neonatal lesion of the ventral hippocampus [neonatal ventral hippocampal lesion (NVHL)]. In adulthood, NVHL leads to behaviours mimicking symptoms of schizophrenia, such as exaggerated behaviour responses to novelty, stress, and direct or indirect dopamine (DA) agonists, disrupted prepulse inhibition of the acoustic startle response, impaired social behaviour, and working-memory deficits (Lipska, 2004; Lipska et al. 1993; Tseng et al. 2009). Behavioural abnormalities in NVHL animals may be linked to the abnormal activity of the mesocortical and mesolimbic dopaminergic systems. Neurons in the prefrontal cortex (PFC) and nucleus accumbens (NAc) show exaggerated responses to the stimulation of the ventral tegmental area (Goto & O'Donnell, 2002, 2004; Tseng et al. 2006, 2007). These responses may be due to altered dopaminergic modulation of cortical pyramidal neurons leading to their enhanced excitability (Tseng et al. 2007) and/or to alterations in the GABAergic cortical transmission (François et al. 2009).

Although NVHL rats show exacerbated DA-related behaviours, biochemical data about dopaminergic mechanisms affected by NVHL are inconsistent. Rats with NVHL show altered basal and stimulated DA release in NAc (Corda et al. 2006; Lillrank et al. 1999a; Wan et al. 1996). Conflicting findings regarding alterations in DA receptor expression following NVHL have been reported (Black et al. 1998; El-Rawas et al. 2009; Flores et al. 1996a, b; Lillrank et al. 1999b; Lipska et al. 2003; Schroeder et al. 1999). DA receptors modulate, directly or indirectly, multiple signalling pathways. Dopaminergic drugs and DA depletion or elevation affect the activity of the ERK pathway (Ahmed et al. 2008; Beaulieu et al. 2006; Bezard et al. 2005; Bychkov et al. 2007a; Cai et al. 2000; Chen et al. 2004; Choe et al. 2002; Gerfen et al. 2002; Nagai et al. 2007; Pavon et al. 2006; Pozzi et al. 2003; Shi & McGinty, 2007; Valjent et al. 2000, 2004; Wang et al. 2005; Zhang et al. 2004). Previous studies have documented that the activity of the Akt pathway is under the influence of the dopaminergic system (Alimohamad et al. 2005; Beaulieu et al. 2004, 2006; Brami-Cherrier et al. 2002; Bychkov et al. 2007a; Chen et al. 2007; Gil et al. 2003; McGinty et al. 2008; Novikova et al. 2005; Roh et al. 2007; Shi & McGinty, 2007; Svenningsson et al. 2003; Wei et al. 2007). In addition to G-protein-dependent modulation of the signalling pathways, arrestin-mediated deactivation of the Akt pathway via DA D₂/D₃ receptors has also been reported (Beaulieu et al. 2007a, b). Therefore, alterations in the dopaminergic transmission caused by NVHL may be reflected in the abnormal activity of the cortical signalling pathways. Surprisingly, the effect of NVHL on the activity of these signalling pathways has never been examined.

DA receptors belong to the G-protein-coupled receptor (GPCR) subfamily. The responsiveness of many GPCRs is controlled by a two-step desensitization mechanism involving phosphorylation of a ligand-activated receptor by a G-protein-coupled receptor kinase (GRK) and subsequent high-affinity binding of an uncoupling protein, arrestin (Gurevich & Gurevich, 2006b). Arrestin binding shields the cytoplasmic surface of the receptor precluding further G protein activation. DA D₁ and D₂ receptors have been shown to desensitize in a GRK- and arrestin-dependent manner (Kim et al. 2001; Macey et al. 2004, 2005), although the DA D₃ receptor is desensitization-deficient due to its resistance to GRK-mediated phosphorylation (Kim et al. 2001, 2005). The availability of GRKs and arrestins is an important regulator of the rate and extent of the receptor desensitization. Increased concentration of GRKs and arrestins promotes desensitization and reduces G-protein-mediated signalling, whereas reduced availability of GRKs/arrestins enhances receptor signalling in live animals (Bohn et al. 1999, 2003; Gainetdinov et al. 1999, 2003). Arrestins, in addition to mediating homologous desensitization of GPCRs, serve as scaffolds redirecting GPCR activity to multiple signalling pathways (Gurevich & Gurevich, 2006a). It is conceivable that NVHL alters the expression of GRK and/or arrestin isoforms, which would affect the signalling of DA receptors as well as that of other GPCRs. However, this possibility has never been investigated.

The majority of mechanistic studies of the NVHL model have been performed using male animals. However, sex differences both in schizophrenia (Angermeyer et al. 1990; Grossman et al. 2008; Hafner, 2003; Hafner et al. 1993; Larsen et al. 1996) and the activity of the brain signalling pathways (Kritze & Creutz, 2008; Zhou et al. 2009) have been documented. Therefore, it is conceivable that some of the NVHL effects are sex-specific. To further explore molecular signalling mechanisms of schizophrenia, we studied alteration in the ERK and Akt signalling following NVHL in male and female rats. We also examined the expression of GRKs and arrestins in both sexes. A comprehensive comparative assessment of signalling modifications induced by neonatal stress in males and females may improve our understanding of the molecular mechanism involved in schizophrenia.

Materials and methods

Neonatal lesion

Pregnant Sprague–Dawley rats were obtained at gestational day 14 (Charles River Laboratories, USA) and housed in the Vanderbilt University's animal facility under a 12-h light/dark cycle (lights on 06:00 hours) with food and water available ad libitum. All procedures followed NIH guidelines and were approved by the institutional IACUC. On postnatal day (PD) 7, male and female pups were anaesthetized by placing them on wet ice for 10–15 min. Pups were randomly assigned to receive NVHL or sham lesion. Ibotenic acid [3 μg in 0.3 μl PBS (pH 7.4)] or an equal volume of vehicle at rate 0.1 μl/min was infused bilaterally into the ventral hippocampus at coordinates A=−3.0 mm, L=3.5 mm, H=−5.5 mm from bregma. After surgery, the pups were warmed and returned to their dams. On PD 21, the animals were weaned, separated by sex and type of infusion, and housed three to a cage. The effect of the oestrous cycle on the NVHL-induced behaviour and activity signalling pathways was not specifically investigated in this study.

Behavioural testing

On PD 60, the spontaneous and apomorphine-stimulated locomotor activity was assessed in a circular (30 cm diameter) apparatus. The detector registered the number of 15° turns in any direction, and the data were converted into centimetres travelled. Spontaneous locomotion was recorded for 30 min, then rats were removed from apparatus, treated with D₁/D₂ dopaminergic agonist apomorphine (0.75 mg/kg s.c.) and locomotor activity was recorded for an additional 60 min. Seventy-two hours later, the rats were treated with non-competitive NMDA antagonist MK-801 (0.2 mg/kg i.p.) and tested for locomotor activity for 120 min.

Tissue preparation and histology

All rats were sacrificed by decapitation under pentobarbital anaesthesia 72 h after the last behavioural testing [half-life of MK-801 in the brain is 2 h (Vezzani et al. 1989)]. The heads were placed in a dish with liquid nitrogen to rapidly chill the brains, then the brains were quickly removed and frozen on dry ice. Samples of the PFC (plates 8–10 according to Paxinos & Watson, 1998); included all of the prelimbic and infralimbic cortices of these plates, NAc (included all of the shell and core of plates 13–15), rostral caudate putamen (CPu; all of the CPu of plates 13–15) were prepared from frozen 100-μm-thick sections as previously described (Ahmed et al. 2007; Bychkov et al. 2007a). This process allows for good anatomical precision while keeping the tissue frozen at all times. Tissue was immediately dissolved in cold lysis solution (Ambion, USA) containing the protein denaturing agent guanidine thiocyanate that inhibits enzymatic activity. Samples were stored at −80 °C until needed. We found that phosphorylated proteins are stable in lysis solution at −80 °C for several months. This sample preparation procedure offers good preservation of the protein phosphorylation state, sufficient to detect changes in the in-vivo phosphorylation levels induced by drugs or loss of DA as described previously (Bychkov et al. 2007a). The middle portion of the brain containing the hippocampus was fixed overnight in 4% paraformaldehyde, cryoprotected in 30% sucrose, and frozen on dry ice. The 35-μm-thick sections were cut on a cryostat and stained with Cresyl Violet to verify the size and location of the lesion.

Quantitative Western blotting

Samples for Western blotting were prepared as described previously (Ahmed et al. 2007; Bychkov et al. 2007a). Electrophoresis and transfer onto Immobilon-P (Millipore, USA) membrane were performed essentially as previously described (Bezard et al. 2005) using Criterion cassettes (Bio-Rad, USA) with 26 wells per gel. Equal amounts of protein from each animal were loaded on the gel (0.625–5 μg). Arrestins were detected with arrestin2-specific (Mundell et al. 1999) (1:9000) or arrestin3-specific (Orsini & Benovic, 1998) (1:900) affinity-purified rabbit polyclonal antibodies. We used rabbit polyclonal antibodies (Santa Cruz Biotechnology, USA) to quantify GRK2 (1:500), GRK3 (1:300), GRK5 (1:500), and GRK6 (1:300). To measure phosphorylated proteins, the following antibodies (all from Cell Signaling Technologies, USA) and dilutions were used: mouse anti-phospho-p44/42 MAPK (Thr²⁰²/Tyr²⁰⁴) (1:2000), rabbit anti-phospho-Akt (Thr³⁰⁸) (1:1000); rabbit anti-phospho-Akt (Ser⁴⁷³) (1:1000), rabbit anti-phospho-glycogen synthase kinase (GSK)-α/β (Ser^21/9) (1:1000), rabbit anti-phospho-mTOR (Ser²⁴⁴⁸) (1:1000), rabbit anti-phospho-PDK1 (Ser²⁴¹) (1:1000). To detect total proteins, anti-p44/42 MAPK, anti-Akt, anti-GSK-3α, anti-PDK1, anti-mTOR, and anti-Rictor rabbit antibodies were used (Cell Signaling Technologies), all in 1:1000 dilution. For arrestins, dilutions of standards containing a 1:1 mix of Escherichia coli-expressed purified bovine arrestin2 and arrestin3 (Gurevich et al. 2002) in sample buffer were loaded onto each gel along with samples. For quantification of GRKs, we used bovine GRK2 and GRK3 (Kim et al. 1993), human GRK5 (Kunapuli et al. 1994), and GRK6 (Loudon & Benovic, 1994), purified as described in the given references. We used purified total and phospho-ERK2 (pERK) (expressed in E. coli, purified, and measured as described by Waas et al. 2003) as standards to determine absolute ERK concentrations. To identify bands corresponding to phosphorylated MEK1/2, Akt at Thr³⁰⁸ and Ser⁴⁷³, phosphoinositide-dependent kinase 1 (PDK1) at Ser²⁴¹, and mTOR at Ser²⁴⁴⁸, we used lysates prepared from HEK293 cells stimulated with FGF (10 ng/ml) for 5 min. To identify bands on the blot corresponding to GSK-3α and GSK-3β, we used lysates prepared from HEK293 cells stimulated with insulin (60 nM) for 5 min. Appropriate dilutions of proteins were loaded onto each gel alongside the samples to generate calibration curves. Arrestin2, 3; GRK2, 3, 5, 6, total and pERK are expressed in ng/mg protein. Other proteins are expressed in arbitrary units.

For Western blots the grey values of the bands were measured on X-ray film using Versadoc 4 (Bio-Rad). To account for the background, the local background subtraction algorithm (the background determined in the area immediately surrounding the band of interest) provided by Versadoc software was used. The optical densities of the bands corresponding to arrestins, GRKs, pERK and total ERK were converted into nanograms of the respective protein per milligram of protein using the calibration curves constructed with protein standards of known concentrations. For other proteins, for which no purified standard proteins were available, lysates of HEK293 cells were stimulated with either FGF (10 ng/ml, 5 min; for MEK1/2, and Akt) or insulin (50 mg/ml, 30 min; for GSK-3 and mTOR). The standards aided in the band identification, ensured linearity, and reduced blot-to-blot variability. Calibration curves were fitted to linear equations using Prism 4.0 (GraphPad Software, USA). The values for arrestins, GRKs, pERK and total ERK were expressed in ng/mg of total protein; for all other proteins in arbitrary units/mg of total protein. The arbitrary unit for MEK1/2 and Akt was defined as the amount of respective protein (phosphorylated or total) in 1 μg and for GSK-3 and mTOR in 3 μg of the total lysate protein of stimulated HEK293 cells.

Statistics

StatView software (SAS Institute, USA) was used for statistical analysis. Locomotor activity and biochemistry data were analysed by two-way ANOVA with lesion and sex as main factors. Where significant effects of factors or lesion × sex interaction were detected, the data were post-hoc analysed for male and female separately by two-tailed Student's t test. Differences of p<0.05 were considered significant. Each significant finding was verified in two independent experiments based on specific hypothesis (to account for multiple comparisons).

Results

The lesion and behaviour

Injection of ibotenic acid in ventral hippocampus of neonatal rats resulted significant damage of the ventral hippocampus with boundaries AP 4.3–5.8 and DV 4.2–9.0 (Fig. 1a). The typical extent of the NVHL in this work corresponds to maximal or near maximal extents of the lesion seen in other studies (Tseng et al. 2007). The dorsal hippocampus, the amygdalo-hippocampal transitional area, and the entorhinal cortex regions were spared in all cases.

Fig. 1.

Neonatal ventral hippocampal lesion (NVHL) was produced by the infusion of ibotenic acid into the ventral hippocampus of neonatal (postnatal day 7; PD 7) rat pups of both sexes. (a) A Nissl-stained coronal section of the adult rat brain demonstrating a typical extent of the lesion. (b) The graph shows elevated activity in the neonatally lesioned rats in response to novel environment (novelty), apomorphine or MK-801 administration. The rats were lesioned at PD 7 and tested at PD 56. The data were analysed by two-way ANOVA with group (control vs. lesion) and sex as main factors. Single-line brackets indicate the difference between the sexes; double-line brackets indicate the differences between the control and lesioned groups; groups across sexes. * p<0.05, ** p<0.01 for respective control group by Student's t test.

NVHL significantly increased spontaneous (novelty-induced) locomotor activity in both male and female postpubertal rats [F(1, 44)=19.6, p<0.0001]. The level of spontaneous locomotion in female rats of both control and NVHL groups was significantly higher than in males [F(1, 44)=11.13, p<0.05] (Fig. 1b). However, there was no significant interaction between the factors. Apomorphine injection caused significant increase in locomotor activity. The two-way ANOVA analysis yielded significant effect of lesion on the apomorphine-induced locomotor activity [F(1, 44)=4.26, p=0.045]. There was no significant effect of sex (p=0.052) on apomorphine-stimulated locomotion and no lesion × sex interaction. A more detailed analysis demonstrated that only males with NVHL had significantly higher apomorphine-stimulated locomotor activity than control males, whereas females with NVHL were not significantly different from control females (Fig. 1b). Furthermore, whereas both control and NVHL male rats demonstrated significant increase in the locomotor activity upon apomorphine administration (p=0.023 and 0.0015 for control and NVHL, respectively, according to Wilcoxon's signed ranks test), only control females significantly increased their locomotor activity in response to apomorphine (p=0.041 and 0.48, respectively). Thus, novelty-induced locomotor activity was higher in female rats, and both sexes respond to NVHL by increasing spontaneous locomotion. In contrast, the apomorphine-stimulated locomotion was quantitatively similar in normal males and females, but only males respond to NVHL by increased apomorphine-induced locomotion. Conversely, MK-801 increased locomotor activity in control or NVHL male and female animals to a similar degree (Fig. 1b).

Sex difference and the effect of NVHL on the activity of the MEK-ERK pathway

Next, we examined the activity of the MEK-ERK pathway in males and females with NVHL. The levels of phospho-MEK1/2 were higher in females than in males in all brain regions (Fig. 2a, b). We detected no effect of lesion in any region (in NAc, the lesion effect approached significance with p=0.052). However, the level of phospho-MEK1/2 was reduced by the lesion in NAc in males and CPu in females (Fig. 2b). Similarly to MEK1/2, the phosphorylation of ERK1/2 was higher in males than in females in all brain regions, effect of sex [PFC: F(1, 24)=11.61, p=0.002; NAc: 13.45, p=0.0012; CPu: 18.4, p=0.0003] (Fig. 2c, d). The phosphorylation of ERK1/2 was significantly reduced by NVHL in PFC [F(1, 24)=5.27, p=0.031], and NAc (F=10.92, p=0.003), but not in CPu (Fig. 2d). The activation of ERK1/2 was altered by NVHL in both males and females, since there was no significant lesion × sex interaction. Separate analysis in males and females demonstrated that NVHL significantly reduced the ERK1/2 phosphorylation in males only (in PFC and NAc). There were no differences in the levels of total MEK1/2 or ERK1/2 between males and females or in response to NVHL (Fig. 2a,c).

Fig. 2.

Modulation of activity of the MEK-ERK pathway by sex and neonatal ventral hippocampal lesion (NVHL). (a) Representative Western blots for phosphorylated (upper panel) and total (lower panel) MEK1/2 in the PFC. Note a much high levels of phospho-MEK1/2 in females compared to males (no differences in total MEK1/2 level). Left five lanes (standards): lysates of HEK293 stimulated with FGF (10 ng/ml for 5 min (amount of total protein in μg loaded per lane is indicated for each blot). (b) Quantification of Western blot data for phospho-MEK1/2. Brackets indicate the difference between the sexes across control and lesion groups; * p<0.05 for respective control group by Student's t test. (c) Representative Western blots for phosphorylated (upper panel) and total (lower panel) ERK1/2. Note a much high levels of pERK1/2 in females compared to males reduced concentration of pERK in neonatally lesioned animals of both sexes (no differences in the total ERK1/2 level). Standards (left five lanes) are the same as for MEK1/2 (amount of total protein in μg loaded per lane is indicated for each blot). (d) Quantification of Western blot data for pERK1/2. Single-line brackets indicate the difference between the sexes across control and lesion groups; double-line brackets indicate the differences between the control and lesioned groups; groups across sexes.* p<0.05, *** p<0.001 for respective control group by Student's t test.

Sex difference and the effect of NVHL on the Akt pathway

Akt activity

Figure 3a represents the proteins and their phosphorylation sites in the Akt pathways investigated in the study. The Akt pathway mediates the effects of DA via DA D₂/D₃ receptors (Beaulieu et al. 2007a, b) and is affected by DA depletion and dopaminergic drugs (Bychkov et al. 2007a). Akt is regulated by phosphorylation at multiple sites (Fig. 3a). The phosphorylation at Thr³⁰⁸ by PDK1 is obligatory for activation (Chan et al. 1999). The role of phosphorylation at Ser³⁴⁷³ by mTOR complex 2 (mTOR in complex with Rictor; Fig. 3a) remains poorly understood, but may precede phosphorylation at Thr³⁰⁸ and promote recognition and activation of Akt by PDK1 (Sarbassov et al. 2005; Yang et al. 2002). Furthermore, Ser³⁴⁷³ phosphorylation may regulate Akt substrate specificity (Guertin et al. 2006; Jacinto et al. 2006; Manning & Cantley, 2007). Female rats displayed higher level of Akt phosphorylation at Thr³⁰⁸ than males in all brain regions examined [PFC: F(1, 24)=12.08, p=0.002; NAc: 17.52, p=0.0003; CPu: 22.35, p<0.0001] (Fig. 3b, c). The Akt phosphorylation at Ser⁴⁷³ was also significantly higher in females than in males [PFC: F=18.1, NAc: 24.96, CPu: 30.2; p<0.001] (Fig. 3b, d).

Fig. 3.

Modulation of activity of the Akt pathway by sex and neonatal ventral hippocampal lesion (NVHL). (a) Simplified diagram of the Akt pathway indicating the proteins (in rectangles) and phosphorylation sites (in ellipses) examined in the study. Arrows indicate phosphorylation by upstream kinases at specific sites. (b) Representative Western blots for Akt phosphorylated at Ser⁴⁷³ (upper panel), Akt phosphorylated at Thr⁴⁰⁸ and total Akt (lower panel) in the NAc. Left five lanes (standards): lysate of HEK293 stimulated with FGF (10 ng/ml) for 5 min (amount of total protein in μg loaded per lane is indicated for each blot). (c) Quantification of Western blot data for phospho-Akt (Thr³⁰⁸) (means±S.E.M.). The data were analysed by two-way ANOVA with group (control vs. lesion) and sex as main factors. Single-line brackets indicate the difference between the sexes; double-line brackets indicate the differences between the control and lesioned groups across sexes. * p<0.05, ** p<0.01 for respective control group by Student's t test. (d) Graph showing quantification of the Western blot for data phospho-Akt (Ser⁴⁷³) (means±S.E.M.). Single-line brackets indicate the difference between the sexes; double-line brackets indicate the differences between the control and lesioned groups across sexes.

NVHL leads to significant reduction in the Akt phosphorylation at Thr³⁰⁸ in NAc [F(1, 24)=6, p=0.02] and at Ser⁴⁷³ in PFC (F=4.65, p=0.041) and NAc (F=6.1, p=0.021) (Fig. 3c, d). The effect of NVHL on the Akt phosphorylation at Thr³⁰⁸ in NAc was due to changes in males (p=0.0077) rather than females (p=0.38). In contrast, in CPu the change in females was more evident (Fig. 3c). There were no differences in the total Akt level between the groups or sexes (Fig. 3b).

Akt substrates

GSK-3α,β is one of the best characterized substrates of Akt (Manning & Cantley, 2007) (Fig. 3a) and is known to be involved in mediating dopaminergic signalling in the brain (Beaulieu et al. 2007a, b). The GSK-3α(Ser²¹) (Fig. 4a, b) and GSK-3β(Ser⁹) (Fig. 4a, d) phosphorylation levels were uniformly higher in females than in males in all brain regions examined. NVHL significantly reduced the phosphorylation levels of GSK-3α in PFC (F=5.9, p=0.023) and CPu (F=4.46, p=0.045). PFC, but not CPu, also showed significant reduction in the GSK-3β phosphorylation across sexes (F=5.54, p=0.02). In CPu, only males displayed significant reduction in the phosphorylation of GSK-3β (p=0.031) (Fig. 4d). No differences were seen in the total concentration of GSK-3α,β (Fig. 4c).

Fig. 4.

The effect of sex and neonatal ventral hippocampal lesion (NVHL) on the phosphorylation of GSK-3 and mTOR. (a) Representative Western blots for phosphorylated GSK-3α and GSK-3βbin the PFC. Note a much high levels of both phosphorylated GSK-3 isoforms in females compared to males and reduced concentration of phospho-GSK-3 in neonatally lesioned animals of both sexes. Left five lanes (standards): lysate of HEK293 stimulated with insulin (60 nM) for 5 min (amount of total protein in μg loaded per lane is indicated for each blot). (b) Quantification of Western blot data for phospho-GSK-3α(Ser²¹). Single-line brackets indicate the difference between the sexes across control and lesion groups; double-line brackets indicate the differences between the control and lesioned groups across sexes. (c) Representative Western blots for total GSK-3α (upper panel) and GSK-3β (lower panel) in the PFC. Note similar expression of both isoforms in males and females and the lack of effect of thge lesion. (d) Quantification of Western blot data for phospho-GSK-3β(Ser⁹). Single-line brackets indicate the difference between the sexes across control and lesion groups; double-line brackets indicate the differences between the control and lesioned groups across sexes. * p<0.05 for respective control group by Student's t test. (e) Representative Western blots for phosphorylated at Ser²⁴⁴⁸ (upper panel) and total (lower panel) mTOR. Standards (left four lanes) are lysates of HEK293 stimulated with FGF (10 ng/ml) for 5 min (amount of total protein in μg loaded per lane is indicated for each blot). Note a high level of individual variability but no differences between males and females or control and lesioned animals. (f) Quantification of Western blot data for phospho-mTOR (Ser²⁴⁴⁸). No significant differences are detected between the sexes or control vs. lesioned animals.

Akt phosphorylates mTOR at Ser²⁴⁴⁸ (Yang & Guan, 2007) (Fig. 3a). We measured the level of mTOR (Ser²⁴⁴⁸) phosphorylation and found no significant difference between males and females and no effect of NVHL (Fig. 4e, f). There were also no differences in the total levels of mTOR (Fig. 4e).

Akt activation

We examined the phosphorylation of PDK1 at Ser²⁴¹, which is required for the kinase activation (Casamayor et al. 1999). However, in contrast to Akt, which is turned on by PDK1-dependent phosphorylation, the PDK1 activity is regulated largely by the substrate conformation and subcellular relocalization (Toker & Newton, 2000). We found no differences between sexes and no effect of NVHL on the level of the PDK1 phosphorylation at Ser²⁴¹ or on the total PDK1 concentration (Fig. 5) suggesting that Ser²⁴¹ phosphorylation is not rate-limiting for PDK1 activity in these conditions. The second regulatory residue in Akt, Ser⁴⁷³, is phosphorylated by mTOR in complex with Rictor (Yang & Guan, 2007). We found no differences in the concentration of Rictor between males and females in any of the brain regions examined. There was also no effect of NVHL on the concentration of Rictor in either sex (Fig. 5).

Fig. 5.

The effect of sex and neonatal ventral hippocampal lesion (NVHL) on the phosphorylation and expression of PDK1 and expression of Rictor. Representative Western blots for phospho-PDK1 (Ser²⁴¹) (upper panel), total PDK1 (middle panel), and Rictor (lower panel) in the PFC. Left five lanes on the phospho-PDK1 blot: lysate of HEK293 stimulated with FGF (10 ng/ml) for 5 min (amount of total protein in μg loaded per lane is indicated for each blot) used as standards. Note that phospho-PDK1 (Ser²⁴¹) runs as multiple bands. For total PDK1 and Rictor, lysates of normal HEK293 cells were used for standards (amount of total protein in μg loaded per lane is indicated for each blot). Note the lack of sex differences or NVHL effects.

Sex and region difference in the expression of arrestins and GRKs

No significant differences in the concentration of arrestin2 or 3 and in GRK2 or 6 (Fig. 6a) were detected in males and females or following NVHL in any of the brain structures examined. The expression of GRK3 was significantly higher in males than in females in PFC and CPu, but not in NAc [F(1, 24)=30.54 and 9.67, p<0.0001 and p=0.005, respectively] regardless of the lesion (Fig. 6a, b). There was no effect of lesion or significant sex × lesion interaction in any of the forebrain structures. Similarly to GRK3, the expression of GRK5 in PFC and NAc, but not in CPu, was significantly higher in males than in females [F(1, 24)=6.97 and 5.23, p=−0.014 and p=0.031, respectively] (Fig. 6a, b). In NAc, NVHL significantly elevated the expression of GRK5 in females to the male level, but no changes were seen in males (Fig. 6b).

Fig. 6.

The differences between males and females in the expression of G-protein-coupled receptor kinases (GRKs). (a) Representative Western blots for arrestin and GRK isoforms in the PFC. Left five lanes show purified recombinant arrestin and GRK proteins of known concentration (amount loaded per lane in ng is indicated for each isoform). Note even expression of arrestins, GRK2 and GRK6 in males and females and in control and lesioned rats. Also note much higher levels of GRK3 and GRK5 in males compared to females. (b) Quantification of Western blot data for GRK3 (upper panel) and GRK5 (lower panel) isoforms. The data were analysed by two-way ANOVA with group (control vs. lesion) and sex as main factors. Brackets indicate the difference between the sexes across control and lesion groups; * p<0.05 for respective control group by Student's t test.

Discussion

The aim of present study was to examine alterations in the signalling pathways induced by NVHL in adult animals that may be important in schizophrenia. The neonatal destruction of the ventral hippocampus induces behavioural deficits resembling those seen in schizophrenia (Tseng et al. 2009). Molecular mechanisms responsible for the NVHL-induced pathology remain poorly understood. PFC seems to be the predominant site of functional deficits induced by NVHL (Tseng et al. 2006, 2008), although abnormalities in NAc and CPu have also been noted (Flores et al. 2005; Laplante et al. 2005). Adult rats with NVHL exhibit enhanced sensitivity to dopaminergic and glutamatergic stimuli as well as to stimuli known to induce DA release in the mesolimbic system, such as stress and exposure to novel environment (Tseng et al. 2007, 2009). Recently, adult supersensitivity of animals with NVHL to α₁-adrenergic and muscarinic stimulation has been reported (Kamath et al. 2008; Laplante et al. 2005). These alterations suggest that NVHL induces specific modification of signalling pathways in the affected brain regions responsible for enhanced sensitivity to specific stimuli.

Here we demonstrate, in agreement with previous reports (Tseng et al. 2009), that NVHL significantly increased spontaneous locomotor activity in post-pubertal rats. Interestingly, females displayed the overall higher novelty-induced locomotor activity, but showed similar increase following NVHL. In our study, the novelty and MK-801 effects were similar in males and females, whereas the apomorphine-induced hyperlocomotion was evident in male rats only. Previous studies showed increased locomotion produced by novelty and amphetamine in adult male as well as female NVHL rats (Black et al. 1998; Silva-Gomez et al. 2003). The reason for this disagreement may be the drug used (amphetamine vs. apomorphine) or differences in oestrous cycle of examined female rats. Another explanation may be the difference in the way locomotion was measured: we recorded locomotion in a confined area as opposed to the open-field-type apparatus commonly used. It is possible that the experimental setting devoid of stress/anxiety component helped to reveal sex differences. Spontaneous locomotion was shown to be independent of hormonal fluctuations, whereas DA-induced locomotion is affected by the oestrous cycle (Kazandjian et al. 1987; Steiner et al. 1980). In any case, the hyperactivity in NVHL males develops earlier and is more pronounced than in females (Black et al. 1998). The sexually dimorphic behavioural and physiological effects of NVHL may derive from the sexual dimorphism of the dopaminergic system. Male and female rats exhibit differences in the anatomy of the dopaminergic system (Kritze & Creutz, 2008), sensitivity to dopaminergic drugs (Zhou et al. 2009), and responsiveness of the dopaminergic system to stress (Sullivan et al. 2009). However, gender-specific molecular mechanisms involved in the NVHL pathology remain largely unexplored.

An increase in the level of ERK protein and mRNA in the thalamus, but not in other brain regions, in schizophrenia has been reported (Kyosseva, 2004). However, it remains unclear whether these data reflect the increased MEK-ERK activation in schizophrenia or the elevated expression is actually not accompanied by excessive activation, since no phosphorylation measurements were performed. Due to long post-mortem delay, it is uncertain whether it is possible to reliably estimate the activity of the MEK-ERK pathway in human post-mortem samples. Therefore, animal models are of particular value for examining the functional role of signalling pathways where the activity of signalling molecules is regulated mostly by phosphorylation. The NVHL-induced reduction in the activity of the MEK-ERK pathways found in the present study argues that schizophrenia might be associated with a suppression of MEK-ERK activity. The fact that clinically effective antipsychotic drugs such as clozapine increase ERK1/2 activation in the PFC (Ahmed et al. 2008; Browning et al. 2005) indicates that elevated activity of the ERK pathway may be beneficial in schizophrenia. Furthermore, lower vulnerability of females in comparison to males to behavioural effects of NVHL correlates with the elevated basal activity of the MEK-ERK pathway coupled with less consistent reduction of this activity by NVHL. Taken together, these data provide evidence that schizophrenia pathology may be associated with reduced MEK-ERK activity.

The functional significance of the Akt-GSK-3 signalling pathway in schizophrenia is highlighted by human post-mortem studies that demonstrated altered levels of these signalling proteins in patients with schizophrenia at post-mortem, although there are notable inconsistencies in published reports. Thus, the concentrations of Akt, phospho-Akt (Ser⁴⁷³), and phospho-GSK-3β (Ser⁹) were found to decrease in the frontal cortex and hippocampus (Amar et al. 2008; Emamian et al. 2004; Zhao et al. 2006). These data suggest an increase in the GSK-3 activity in schizophrenia. In contrast, other studies found reduced levels of GSK-3β protein (Beasley et al. 2001; Kozlovsky et al. 2000) coupled with reduced GSK-3 activity (Kozlovsky et al. 2001) and unchanged Akt concentration (Amar et al. 2008; Ide et al. 2006) in the frontal cortex of patients with schizophrenia. Our data on decreased levels of Akt and GSK-3 phosphorylation in NVHL animals are in agreement with the human studies suggesting increased GSK-3 activity in schizophrenia. The elevated locomotor activity of NVHL animals may be due to enhanced GSK-3 activity (seen as reduced levels of Akt-GSK-3 phosphorylation compared to sham-operated animals). Previous studies demonstrated that enhanced activity of GSK-3 caused by overexpression of GSK-3 in the brain results in an increased locomotion (Prickaerts et al. 2006), whereas locomotion is attenuated in animals treated with GSK-3 inhibitors (Beaulieu et al. 2004; Gould et al. 2004) and in GSK-3 knockout mice (Beaulieu et al. 2004). Typical and atypical antipsychotics have been shown to inhibit GSK-3 activity (increased the level of phosphorylated GSK-3) in rat brain (Alimohamad et al. 2005; Emamian et al. 2004; Li et al. 2007), although opposite results have also been reported (Kozlovsky et al. 2006). Interestingly, GSK-3 as a substrate of Akt seems to be specifically involved, since the level of Akt-dependent phosphorylation of another substrate, mTOR, was unaffected by NVHL and not significantly different between males and females. Therefore, the available evidence supports the role of elevated GSK-3 activity in schizophrenia. Moreover, it is conceivable that reduced basal GSK-3 activity in females may confer resistance to some behavioural manifestations of schizophrenia. The critical importance of the Akt-GSK-3 pathway in schizophrenia is further underscored by the recent evidence that the pathway is linked to the activity of DISC1, the key susceptibility genes in schizophrenia (Kim et al. 2009; Mao et al. 2009). Furthermore, since DISC1-dependent regulation of Akt signalling plays a role in brain development (Mao et al. 2009), this may be one of the molecular mechanisms involved in the effects of NVHL and sex on the Akt signalling.

We found that females have significantly lower concentrations of GRK3 and GRK5 isoforms, whereas the concentration of GRK2 and GRK6 isoforms and arrestins were the same in both sexes. We did not detect any effect of NVHL on the expression of arrestins or GRKs in adult males. However, in females, the level of GRK5 in the NAc, which was normally significantly lower than that in males, was increased by NVHL to the male level. The role of the GRK/arrestin proteins in schizophrenia pathology is poorly defined. We recently found reduced expression of several GRK/arrestin isoforms in the cortical regions of schizophrenia patients, with GRK3 and GRK5 being specifically affected in younger and GRK6 in older patients with schizophrenia (Bychkov et al. 2007b). Furthermore, chronic treatment with haloperidol and clozapine induced drug-specific alterations in the concentration of arrestin and GRK isoforms in key brain regions (Ahmed et al. 2008). These data suggest that GRKs may be involved in the pathology of schizophrenia. The present finding of sexually dimorphic expression of GRK3 and GRK5 isoforms suggest these proteins may play a role in sex differences in the basal activity of signalling pathways and in differential vulnerability to the neonatal insult.

Basal differences in the concentration of signalling molecules and activity of the signalling pathways between males and females seen in the present study could presage differential responsiveness, quantitatively or qualitatively, of males and females to a combination of genes and life events that trigger schizophrenia. They could also play a decisive role in differential responsiveness of sexes to psychotropic drugs. Cataloguing and understanding sex differences in the basal and evoked activity of the signalling pathways may shed light on the pathophysiology of schizophrenia and lead to the design of highly effective antipsychotic drugs.