Gene expression in the anterior cingulate cortex and amygdala of adolescent marmoset monkeys following parental separations in infancy

Abstract

Early life adversities are risk factors for later mood and emotional disorders. Repeated separation of infant marmosets from their parents provides a validated primate model of depression vulnerability, producing in-vivo biochemical and behavioural effects indicative of persistently altered stress reactivity and mild anhedonia. Here we report the long-term effect (in adolescence) of this intervention on the expression of synaptophysin, GAP-43, VGluT1, VGAT, MAP-2, spinophilin, and 5-HT_1A and 5-HT_2A receptors, in the anterior cingulate cortex (ACC; supragenual and subgenual areas) and amygdala (lateral, basal and central nuclei). These genes and regions are implicated in the response to stress or in mood disorder. The profile of 5-HT_1A receptor binding in ACC was affected by early deprivation, notably in the subgenual region, with a decrease in deep laminae but an increase in superficial laminae. Following early deprivation, spinophilin mRNA was reduced in subgenual ACC. In the amygdala, no significant effects of the manipulation were seen, but expression of several transcripts was sexually dimorphic. There were correlations between expression of some transcripts and in-vivo measurements. The results show that early deprivation in a non-human primate has a selective long-term effect on expression of genes in the ACC, particularly the subgenual area. The results differ from those reported in the hippocampus of the same animals, indicating the presence of limbic region-specific long-term molecular responses to early life stress.

Introduction

Experimental and epidemiological evidence shows that early life stressors, such as parental separation, are major risk factors for mood and anxiety disorders (De Bellis et al.1999; Heim & Nemeroff, 2001). The preclinical data come mostly from studies of rodents, but recently some relevant non-human primate models have been developed (Gilmer & McKinney, 2003; Pryce et al.2005; Sanchez et al.2001). In particular, in the common marmoset (Callithrix jacchus), a small New World monkey, daily short periods of isolation from the family group during the first month of life (‘early deprivation’; ED) causes endocrine stress responses, and leads to biochemical, cardiovascular, and behavioural effects during juvenility and adolescence, indicative of increased basal activity and reactivity in stress systems, and mild anhedonia (Dettling et al.2002a, b, 2007; Pryce et al.2004a, b). These findings indicate that the ED protocol leads to a ‘pro-depressive’ state that persists until at least month 12 of life, which is adolescence in this species. Moreover, ED causes long-term changes in gene expression in the hippocampus, including effects on 5-HT_1A receptors (5-HT_1ARs) and markers of synaptic plasticity (Law et al.2008) that may reflect the behavioural consequences of ED and which show similarities to alterations reported in the hippocampi of people with mood disorder.

Along with the hippocampus, the anterior cingulate cortex (ACC; Devinsky et al.1995; Diorio et al.1993; Paus, 2001; Vogt, 2005) and the amygdala (Davidson, 2002; LeDoux, 2000; Murray, 2007) are brain regions strongly implicated in emotion, mood and the responses to stress. They are also involved centrally in the neuropathology and pathophysiology of a range of psychiatric and neurological conditions, especially mood and anxiety disorders (Drevets et al.1997; Ebert & Ebmeier, 1996; Harrison, 2002; Mayberg et al.1997; Nestler et al.2002). The clinical studies have contributed to the realization that the ACC is anatomically, functionally, and pathophysiologically, heterogeneous. One key distinction is that between the supragenual and subgenual ACC (Carmichael & Price, 1996; Gittins & Harrison, 2004; Palomero-Gallagher et al.2008; Vogt et al.1995). It is the subgenual area that is most implicated in mood disorder (Drevets et al.1997; Hajek et al.2008; Koo et al.2008; Mayberg, 1997; Öngür et al.1998a) and its treatment (Drevets et al.2002; Mayberg et al.2005).

To date, few studies have examined the long-term molecular consequences of early life stress on the ACC or amygdala, and primate data are limited to a single study (Sabatini et al.2007). Our study had two related objectives. First, to discover whether ED in the marmoset has a long-term effect on gene expression in these regions, including a separate analysis of supra- and subgenual ACC. Second, to identify whether there are correlations between expression of the genes and the biochemical and behavioural effects of ED measured. We focused on three groups of genes, each of which has been implicated in the pathophysiology of mood disorder, and studied previously in the hippocampus of ED animals (Law et al.2008): serotonin (5-HT_1A and 5-HT_2A) receptors, and genes indexing presynaptic (synaptophysin, VGluT1, VGAT, GAP-43) and dendritic (MAP-2, spinophilin) functioning. To the extent that the ED model provides a model of depression vulnerability, the findings also have relevance to the role that early life stress may have in the pathogenesis of ACC or amygdala involvement in mood disorder.

Methods

The animals were bred, and all in-vivo studies conducted, at the Laboratory for Behavioural Neurobiology of the Swiss Federal Institute of Technology, Zurich, under experimental permit in accordance with the Swiss Animal Protection Act (1978). The brains were shipped to Oxford for study under licence from the Convention for International Trade in Endangered Species of Wild Fauna and Flora (CITES), administered by the Swiss Federal Office for Veterinary Affairs and the UK Department for Environment, Food and Rural Affairs.

The ED intervention

The ED intervention and its effects in vivo in infancy, juvenility and adolescence have been described fully (Dettling et al.2002a, b; Law et al.2008; Pryce et al.2004a, b). Briefly, the marmoset is characterized by monogamous breeding, dizygotic twins, and high levels of caregiving by both parents. In this study, each set of parents contributed, in random order, one pair of ED twins and one pair of control twins; one set of parents contributed one pair of ED twins only. The total sample was therefore 10 pairs of ED twins and nine pairs of control twins (Table 1). On post-natal days 2–28, ED infants were separated from their parents for 30–120 min each day, using variable durations and timings. ED was carried out consecutively within each twin pair, such that one infant remained with the parents at all times. Controls were briefly handled on the back of the carrying parent each day. After day 28, subjects remained with the family group, and there were no further interventions that differed between ED and control subjects. Subsequent behavioural testing, and the collection of physiological samples, were carried out in the home cage, except that at age 18–20 wk, animals were studied in six 60-min tests of isolation from the family in a novel physical environment.

Laboratory methods

Brains were snap-frozen, and stored at −80°C before being coronally sectioned on a cryostat at 10 µm thickness and prepared for in-situ hybridization or receptor autoradiography (Law et al.2008). Every 20th slide was taken for Nissl staining to identify anatomical landmarks.

Since marmoset cDNA sequences were not available, the proposed target region of each transcript was amplified from marmoset cDNA using RT–PCR with primers designed to the human cDNA sequence. RT–PCR products were then sequenced, and oligonucleotide probes designed that were 100% complementary to the marmoset cDNA sequence (data available on request). For in-situ hybridization, the frozen sections were thawed, fixed, acetylated and delipidated (Law et al.2003). Oligonucleotides were 3′ end-labelled with [³⁵S]dATP (1250 Ci/mmol; PerkinElmer, UK) in a 10:1 molar ratio using terminal deoxynucleotidyl transferase (Promega, UK). The experimental conditions and film exposure times for each probe were optimized in pilot studies; the definitive experiment for each transcript was performed in a single run. Sections were incubated overnight at 40–42°C with hybridization buffer containing 1.0×10⁶ cpm of labelled probe, as previously described (Law et al.2003). Post-incubation washes were carried out in 1×SSC at 55°C for 3×20 min and 1 h at room temperature. Experimental controls comprised: concurrent hybridization with sense strand probes, hybridization in the presence of 50-fold excess unlabelled probe, and ribonuclease (RNase A 200 µg/ml at 37°C for 20 min) pre-treatment. After air drying, slides were apposed to autoradiographic film (Kodak, USA) along with ¹⁴C microscales (Amersham Pharmacia Biotech, Sweden) for 3 wk (GAP-43 and 5-HT_1AR), 2 wk (synaptophysin, spinophilin and VGAT), or 1 wk (VGluT1 and MAP-2).

Autoradiographic analysis of 5-HT_1AR binding-site densities was carried out using [³H]WAY 100,635 (Burnet et al.1997). Briefly, sections were thawed and pre-incubated at room temperature in 50 mm Tris–HCl buffer (pH 7.4) for 30 min. Sections were then incubated in 50 mm Tris–HCl containing 3 nm [³H]WAY 100,635 for 2 h at room temperature. Non-specific binding was determined by incubation of adjacent sections with 10 µm 5-HT. Slides were washed with the same buffer for 2×4 min at 4°C. Sections were air dried and apposed to Biomax film for 3 wk. Autoradiography for 5-HT_2AR binding sites was carried out with 2 nm [³H]ketanserin (Eastwood et al.2001; Pazos et al.1985); non-specific binding was determined using 10 µm mianserin, and film exposure was 6 wk.

Image and data analysis

Densitometric measurements, converted to nCi/g, were taken over the supragenual and subgenual ACC, and over the amygdala (basal, lateral and central nuclei, referred to as BNA, LNA and CNA respectively), with reference to an atlas (Stephan et al.1980). Within ACC, the average signal across the depth of the grey matter was measured, except for [³H]WAY 100,635 binding, for which separate readings were taken over superficial and deep laminae, reflecting its markedly inhomogeneous localization, with an absence of significant binding in middle layers (Burnet et al.1997; Hall et al.1997). Because of limited tissue availability (or low level of expression of some transcripts), not all areas could be measured in every experiment. In each region, signal was measured over the duplicate or triplicate sections and the mean value used for statistical analysis. Readings were converted to nCi/g using co-exposed ¹⁴C microscales. All experiments and analyses were conducted blind to treatment group.

Data were examined for normality using the Kolmogorov–Smirnov one-sample test. The effect of ED was measured in each region by ANOVA, with group (ED, control) and sex (male, female) as between-subjects factors. For [³H]WAY 100,635 binding in the ACC, repeated-measures ANOVA was used with compartment (superficial, deep) as the within-subjects factor; the ratio of binding between the two compartments was also measured. Sex was included in all ANOVAs because ED×sex interactions were seen in the hippocampus of these animals (Law et al.2008).

To explore whether the parameters measured here were related to three biochemical and three behavioural indices which were affected by the ED intervention (see Table 1) we used Spearman correlations. As a partial control against multiple testing, we set α=0.02, and followed up each significant correlation by examining whether it persisted after partialling for the effect of ED; we also inspected whether a similar correlation was present (at least as a trend) in control and ED groups considered separately.

Results

Distribution of transcripts and receptor binding sites

Figure 1a shows a representative coronal section of the marmoset brain that includes the subgenual ACC, with the regions wherein supragenual and subgenual measurements were taken. The distribution of VGluT1 (Fig. 1 b), VGAT (Fig. 1 c), GAP-43 (Fig. 1c), synaptophysin (Fig. 1e), spinophilin (Fig. 1f) mRNAs are shown, as is [³H]WAY 100,635 binding (Fig. 1g) and [³H]ketanserin binding (Fig. 1h). Note in Fig. 1g, [³H]WAY 100,635 binding showed the anticipated laminar distribution in the supragenual ACC, with a band of prominent signal over the superficial lamina of the grey matter, with a separate, weaker band in the deep laminae. In the subgenual ACC, however, this profile was reversed (Fig. 1g). A representative section through the amygdala is also shown (Fig. 1i).

The experimental controls for each probe showed minimal background signal over the ACC and amygdala. In addition to the sequence verification of each probe target, and the fact that each mRNA had the expected distribution based on other primate and rodent studies, this indicates the specificity of signal for each transcript. Similarly, the distribution of radioligand binding to 5-HT_1ARs and 5-HT_2ARs was as anticipated, and showed no detectable labelling after cold displacement.

Effect of ED on expression of pre-synaptic protein genes

VGluT1 mRNA in the ACC did not show a main effect of ED or sex, or an interaction between them (Table 2). In the BNA, there was a main effect of sex [F(1, 14)=6.74, p=0.021], with VGluT1 mRNA signal greater in males than females (values±s.d.) (405±13 vs. 338±26 nCi/g; t=−2.59, d.f.=16, p=0.022). There was also an ED×sex interaction [F(1, 14)=5.42, p=0.035]. The latter reflected a trend for BNA VGluT1 mRNA to be increased by ED in females and decreased in males, but neither post-hoc test was significant (p=0.35 and p=0.15 respectively). In the CNA, there was a main effect of sex on VGluT1 mRNA [F(1, 14)=5.88, p=0.029], again reflecting higher signal in males than females (487±22 vs. 402±22 nCi/g; t=−2.61, d.f.=16, p=0.019) but no effect of ED or ED×sex interaction.

VGAT mRNA and GAP-43 mRNA showed no effect of ED, sex, or interaction between them (all p>0.1) in any region measured (Tables 2 and 3).

Synaptophysin mRNA did not differ between control and ED groups in ACC (Table 2) or amygdala (Table 3). However, in the LNA, there was an effect of sex [F(1, 14)=5.97, p=0.021], with expression being higher in females than in males (356±42 vs. 251±15 nCi/g; t=2.37, p=0.047). A similar trend was seen in the CNA [F(1, 14)=4.07, p=0.063].

Effect of ED on expression of post-synaptic protein genes

MAP-2 mRNA was not affected by ED, nor did it differ between sexes or show ED×sex interactions in either supragenual or subgenual ACC (Table 2); note that the subgenual ACC analyses were underpowered in that due to lack of tissue, only four control animals were available. In the amygdala, there were no effects of ED (Table 3) or ED×sex interactions on MAP-2 mRNA (all p>0.25), but there was a main effect of sex (p⩽0.05) in each nucleus, with MAP-2 mRNA being higher in males than females: in LNA (615±82 vs. 392±58 nCi/g; t=2.21, p=0.044); in BNA (552±78 vs. 323±45 nCi/g; t=2.55, p=0.027); in CNA (538±81 vs. 321±48) nCi/g; t=2.31, p=0.05).

Spinophilin mRNA in the supragenual ACC did not show main effects of ED or sex, or an interaction between them. In subgenual ACC there was a main effect of intervention [F(1, 11)=7.98, p=0.017], reflecting a reduction of spinophilin mRNA after ED (Table 2), with no effect of, nor interaction with, sex.

Effect of ED on binding and expression of 5-HT_1ARs and 5-HT_2ARs

For 5-HT_1AR binding, in the supragenual ACC, there was a main effect of laminar compartment [F(1, 15)=169.4, p<0.001], reflecting the higher signal in superficial laminae, and an ED×laminar compartment interaction [F(1, 15)=6.76, p=0.020]. The effect of ED on [³H]WAY 100,635 binding in supragenual ACC was not significant in either compartment considered separately (Fig. 2a), but the superficial:deep ratio of binding was decreased after ED [F(1, 15)=8.67, p=0.010; Fig. 2c, left panel]. In the subgenual ACC there was a main effect of laminar compartment [F(1, 15)=16.2, p=0.001], in this instance arising from the higher level of binding in the deep laminae. There was an ED×laminar compartment interaction within subgenual ACC [F(1, 15)=8.15, p=0.012]; in superficial laminae [³H]WAY 100,635 binding was increased following ED [Fig. 2 b, left panel; F(1, 15)=6.58, p=0.022], whereas in deep laminae [³H]WAY 100,635 binding was reduced after ED (Fig. 2 b, right panel; F(1, 15)=5.23, p=0.037]. The ratio of binding in subgenual ACC was increased after ED [Fig. 2c, right panel; F(1, 15)=7.03, p=0.018]. [³H]WAY 100,635 binding was unaffected by ED in the BNA (Table 3). There were no main effects of sex, or ED×sex interactions on [³H]WAY 100,635 binding in any region.

5-HT_1AR mRNA was not significantly affected by ED in either ACC area (Table 2) or in the amygdala (Table 3), nor were there main effects of sex or ED×sex interactions.

[³H]ketanserin binding and 5-HT_2AR mRNA were unaffected by ED in ACC (Table 2). There was a trend for binding to be lower, but this did not get close to significance, in part because of the smaller number of subjects available for this analysis (Table 2). 5-HT_2AR mRNA was also unaltered after ED in the LNA and CNA (Table 3).

Correlations between gene expression and in-vivo measures

Two correlations between gene expression and behaviour met our criteria for significance (see Materials and Methods section). Impulsivity, measured in a palatable-reward reaching task, was increased in ED juveniles compared to controls (Pryce et al.2004a), and impulsivity scores correlated inversely with synaptophysin mRNA in the subgenual ACC (r=−0.757, n=13, p=0.003; partialling for effect of ED: r=−0.812, d.f.=10, p=0.001), with similar trends in control (r=−0.975, n=5, p=0.005) and ED (r=−0.527, n=8, p=0.18) groups (Fig. 3a). Second, motivation for palatable reward, measured on a progressive ratio (PR) schedule of reinforcement, was reduced in ED adolescents (Pryce et al.2004b), and scores for rewards obtained correlated with VGluT1 mRNA in BNA (r=0.718, n=14, p=0.004), a relationship seen in control (r=0.750, n=7, p=0.052) and ED (r=0.775, n=7, p=0.041) groups considered separately (Fig. 3 b), and which remained significant when partialling for the effect of ED (r=0.742, d.f.=11, p=0.004). A similar but weaker relationship was seen for VGluT1 mRNA in subgenual ACC (not shown).

Discussion

The marmoset ED paradigm provides one of the first well-validated primate models of depression vulnerability. It produces not only behavioural and biochemical effects indicative of long-term effects on stress reactivity and mild anhedonia (Dettling et al.2002a, b, 2007; Pryce et al.2004a, b), but also leads to several long-term changes in hippocampal gene expression that are indicative of an effect of ED on synaptic plasticity and that are similar to findings reported in subjects with mood disorders (Law et al.2008; see also Manji et al.2001; Nestler et al.2002). Here we report a companion study in two other limbic brain regions implicated in stress responses and emotional regulation, the ACC and the amygdala. The changes were less striking than those in the hippocampus, being limited to a change in the distribution of 5-HT_1AR binding, and in expression of MAP-2, in the ACC. We also identified two correlations between gene expression and behavioural measures (impulsivity with synaptophysin mRNA in subgenual ACC, and reward-seeking behaviour with VGluT1 mRNA in BNA), and a sexual dimorphic expression of three transcripts (synaptophysin, VGluT1 and MAP-2) in the amygdala.

Cortical 5-HT_1AR distribution is inhomogeneous across the depth of the grey matter. In other primate species, the highest level of mRNA expression (Burnet et al.1995), immunoreactivity (DeFelipe et al.2001; Palchaudhuri & Flügge, 2005), and ligand binding (Burnet et al.1996, 1997; Hall et al.1997), is in the superficial (supragranular) laminae with a lesser band in the deep (infragranular) laminae. The marmoset corresponds to this profile, although interestingly, in the subgenual ACC, the pattern is reversed, with maximal binding in the deep layers. At a cellular level, cortical 5-HT_1ARs are mainly expressed by pyramidal neurons (Burnet et al.1995; Palchaudhuri & Flügge, 2005; Santana et al.2004) and most studies agree they are localized to the axon initial segment, with much lower levels elsewhere, e.g. in dendrites (Cruz et al.2004; DeFelipe et al.2001; but see Kia et al.1996). We measured 5-HT_1AR expression, and binding, since the serotonin system and these receptors are implicated in stress, depression and anxiety, and they are decreased in the hippocampus after ED (Law et al.2008) and some other stressors (Flügge, 1995; Preece et al.2004). Following ED, we found no overall effect on 5-HT_1AR binding-site density in any region, but in ACC there was a complex shift in binding between the superficial and deep laminae (Fig. 2). This pattern was marked in the subgenual ACC (Fig. 2 b), wherein ED led to a significant increase in binding in superficial layers, and a decrease in deep layers, as well as a significant change in the ratio between them. The opposite redistribution (from superficial to deep) occurred in the supragenual ACC (Fig. 2a). There are precedents for alterations in receptor binding sites of this kind. For example, there is an altered relative laminar distribution of cortical 5-HT_1AR binding sites in schizophrenia (Burnet et al.1996; Simpson et al.1996), of ACC neurokinin-1 receptors in major depression (Burnet & Harrison, 2000), and of cerebellar 5-HT_2ARs in schizophrenia (Eastwood et al.2001).

Given the cellular and subcellular localization of 5-HT_1ARs summarized above, we hypothesize that the shifts in [³H]WAY 100,635 binding-site distribution after ED represent a differential regulation of these receptors on the axon initial segments of superficial (laminae II–III) vs. deep (laminae V–VI) pyramidal neurons. This may result from differences in serotonergic innervation (see DeFelipe et al.2001), perhaps in turn related to the alterations in serotonergic system and hypothalamo-pituitary axis function produced by ED (Dettling et al.2002a, 2007; Pryce et al.2004b) and other early life stressors in primates (Ichise et al.2006; Rilling et al.2001). Pyramidal neuron 5-HT_1ARs are hyperpolarizing (Arenada & Andrade, 1991) and play a major role in inhibitory regulation of these cells (Cruz et al.2004; DeFelipe et al.2001) and hence the receptor redistribution may impact functionally, by changing ACC output characteristics (Czyrak et al.2003). The efferent cortical and subcortical connections of the ACC vary between supra- and subgenual areas and between superficial and deep laminae (Carmichael & Price, 1996; Freedman et al.2000; Johansen-Berg et al.2008; Öngür et al.1998b; Vogt et al.1995, 2005), and thus any such functional implications will be anatomically complex. For example, the subgenual ACC findings (Fig. 2 b) suggest, simplistically, that pyramidal neurons projecting to other cortical regions (i.e. those in laminae II–III) are subject to increased inhibition, whereas those projecting to subcortical sites (in laminae V–VI) may be disinhibited. The fact that ACC 5-HT_1AR mRNA was unchanged after ED indicates that the alterations in binding-site distribution are not secondary to altered gene expression, but probably reflect changes in receptor trafficking.

Spinophilin expression was decreased in the subgenual ACC after ED. Spinophilin is a protein concentrated in dendritic spines (Allen et al.1997), and its reduced expression after ED implies that there is a change in the function, or the structure (or both) of dendritic spines on neurons within this area (Feng et al.2000). The fact that there were no alterations in MAP-2, a dendritic marker (Johnson & Jope, 1992; Pei et al.1998) or in density of 5-HT_2AR binding sites [a receptor localized mainly to dendritic shafts (Jakab & Goldman-Rakic, 1998)], suggests that whatever abnormality is being indexed by reduced spinophilin expression is limited to the spines rather than affecting the dendritic tree as a whole (Eastwood et al.2007; Law et al.2004a, b). Since dendritic spines receive most excitatory synapses (whereas inhibitory and monoaminergic synapses more often terminate elsewhere on the neuron), the decrease may indicate a particular decrement of glutamatergic innervation to subgenual neurons. This issue merits further study, including direct assessment of dendritic morphology and spine density using Golgi stains.

In the amygdala, unlike the ACC (and hippocampus), ED had no overall effect on any of the transcripts or binding sites. It is possible that ED does not produce long-term effects on amygdala gene expression. However, it is more likely that there are alterations which do not include the candidate genes we measured. Notably, a recent microarray study of the amygdala in 3-month-old rhesus macaques who had experienced maternal separation in infancy found only 20/8405 transcripts (0.24%) were differentially expressed, none of which were the genes studied here (Sabatini et al.2007). Their most striking finding was for GUCY1A3, a soluble nitric oxide receptor, and it would be of interest to measure that transcript in the ED model.

In contrast to the lack of effect of ED in the amygdala, there were several sex differences therein: synaptophysin mRNA was higher in females, whereas VGluT1 and MAP-2 mRNAs were higher in males. The synaptophysin mRNA finding is also seen in the marmoset hippocampus, but MAP-2 and VGluT1 mRNAs are not sexually dimorphic in that region (Law et al.2008). Sex differences in gene expression have been reported in various brain regions and species (Cahill, 2006; Cosgrove et al.2007; Rinn & Snyder, 2005), and might contribute to the roles that the amygdala plays in sex-specific social behaviours (Madeira & Lieberman, 1995; Simerly, 2002). To our knowledge there are no prior studies investigating sex differences in these transcripts in the amygdala. In that MAP-2 is a dendritic marker (see above), higher MAP-2 expression may reflect a greater dendritic arborization in the male amygdala, as is observed in the rat medial amygdala (Cooke et al.2007; Nishizuka & Arai, 1981). Similarly, elevated VGluT1 mRNA in males may reflect a greater activity or density of glutamate synapses (Wilson et al.2005) in the male amygdala, which would be compatible with ultrastructural findings in rats (Cooke & Woolley, 2005).

We observed two molecular-behaviour correlations: synaptophysin mRNA (putatively, an index of synaptic density and/or function; see Harrison & Eastwood, 2001) in subgenual ACC correlated inversely with impulsive behaviour, whereas higher VGluT1 mRNA (putatively, a marker of enhanced glutamate neurotransmission; Wilson et al.2005) in the amygdala correlated with greater motivation for reward. The measure of impulsive behaviour was increased in ED juveniles compared to control juveniles (Pryce et al.2004a) whereas reward motivation was decreased in ED adolescents compared to control adolescents (Pryce et al.2004b). The two transcripts involved were unaltered by ED, and the correlations were present in both ED and control groups; as such, they represent molecular correlates of individual variation in the two behaviours. Since synaptophysin and VGluT1 are indexing, albeit indirectly, neural connectivity and/or activity, the correlations may be seen as analogous to those between indices of brain function/activity in human subjects and various behaviours (e.g. Hariri et al.2006; Smolka et al.2005).

Many further parameters remain to be determined in this primate model of depression vulnerability. First, stereological assessment of ACC and amygdala volumes and their cellular composition, to establish whether these are affected after ED, as has been reported in mood disorder (Bowley et al.2002; Chana et al.2003; Cotter et al.2001; Hajek et al.2008; Harrison, 2002; Öngür et al.1998a; Sheline et al.1998). Second, investigation of other molecular markers of different cell types and biochemical processes, as a step towards delineating mechanisms that mediate between the ED intervention and the long-term in-vivo effects and the alterations in gene expression observed here. Additional experimental studies of this kind have the potential to reveal how early life stress leads to vulnerability to mood and emotional disorders, and in turn may offer the prospect of improved and even preventative therapies.

Acknowledgements

The study was funded by the Wellcome Trust, with additional support from the National Science Foundation, Switzerland (Project grant 3167791.02) and National Center for Competence in Research: Swiss Etiological Study of Adjustment and Mental Health (grant no. 51A240-104890). We thank Phil Burnet, Andrea Dettling, Helen Gordon-Andrews and Mary Walker for their expert contributions.

Statement of Interest

In the past three years P.J.H. has received unrestricted educational grants from GlaxoSmithKline (GSK), and honoraria for educational lectures or chairing scientific meetings from Astra Zeneca, Bristol–Myers Squibb, GSK, Janssen, Lilly, Merck, Sanofi and Servier, and has been a scientific advisor to Curidium, Janssen, Merck, and Wyeth.

References