«Distribution Agreement In presenting this thesis or dissertation as a partial fulfillment of the requirements for an advanced degree from Emory ...»
1.4 Impact of Early Life Experience on Amygdala Development It is now possible to describe the effects of ELS on the amygdala, using the normative development of the amygdala as a context. As discussed in Section 1.3.1, the complex and interdependent processes of brain development make single time points of little value for understanding outcomes. Specific changes to amygdala function caused by ELS are hypothesized to be most impactful during CPs in development. It is necessary to consider the direct effects of ELS during those windows, not in adulthood, to understand the trajectory over which anxiety disorders can be instantiated and thereby prevented. To this end, the various effects of ELS on the amygdala will be outlined, with greater preference placed on those spanning developmental CPs (Section 1.4.1) but some attention paid to those with limited developmental perspective (Section 1.4.2). Finally, environmental factors that have been shown to promote resiliency will be presented in Section 1.4.3, to contrast the effects of ELS and identify targets for intervention in anxiety disorder pathogenesis.
1.4.1 Early Life Stress and the Trajectory of Amygdala Development As discussed above, amygdala dysfunction following ELS has been suggested to contribute to the pathogenesis of psychiatric disorders (Heim and Nemeroff, 2002; Zhang et al., 2004; Sadler et al., 2011; Blackford and Pine, 2012; Callaghan et al., 2013). The specific effects of ELS on the amygdala and the contribution of those effects to disease risk are complex; not only the timing of exposure to ELS, but also the timing of measurement, influences the effects on the biology of the amygdala (Tottenham and Sheridan, 2009). Therefore, this section will outline the current state of knowledge regarding the effects of ELS on the trajectory of amygdala development and summarize what conclusions can be drawn from this sparse and sometimes contradictory literature.
Studies investigating the trajectory of amygdala development following ELS in terms of volume have identified age-specific effects that emphasize the importance of considering trajectories. A longitudinal study of human families found that maternal cortisol early in pregnancy positively correlated with amygdala volume in offspring at 7 years old, which partly mediated the association between maternal cortisol and affective problems in the children (Buss et al., 2012). Further study of human development found adverse caregiving increased amygdala volumes during childhood (Mehta et al., 2009; Tottenham et al., 2010). Parallel findings in rodents support the increase of developing amygdala volume following ELS: PS increased the volume, neuron density, and glia density of the rat BLA specifically in adolescence (at P25 but not P7, P45, or P60; Kraszpulski et al., 2006). Interestingly, childhood trauma reduces amygdala volume in adults with psychiatric disorders (Driessen et al., 2000; Schmahl et al., 2003), however a number of studies have failed to replicate this finding (Bremner et al., 1997; Stein et al., 1997;
De Bellis et al., 1999; Andersen et al., 2008). One explanation for the discrepant findings in children and adults is that ELS increases amygdala volume in childhood, possibly via excessive or accelerated growth, but robust synaptic pruning or cell death lead to reduced volumes by adulthood. Thus, increased amygdala volume in childhood may contribute to the pathogenesis of psychiatric disorders or serve as a marker of risk, but further developmental processes during adolescence contribute to the final outcome.
One particular study supports the idea that adolescent development contributes to amygdala outcomes. Stress exposure during adolescence (P27-29) was shown to impact emotional behavior in adulthood, including enhancing cued fear conditioning, impairing avoidance learning, and reducing exploratory behavior. In addition, the stress exposure enhanced expression in the BLA, within 4 days and persisting into adulthood, of a neuronal cell adhesion molecule, L1. Expression of L1 normally wanes into adulthood, but adolescent stress opposes this effect. L1 is involved in several neurodevelopmental processes including neurite outgrowth, axon guidance, and cell adhesion, so persistent expression in adulthood may contribute to alterations of neuron density and volume of the BLA (Tsoory et al., 2010). The effects of juvenile stress and ELS may interact to produce the complex trajectory suggested to contribute to anxiety disorder pathogenesis. Furthermore, ELS may enhance stress reactivity in adolescence, increasing the likelihood juvenile stress alters subsequent amygdala development.
A number of animal studies have suggested the effects of PS are mediated by changes in stress sensitivity at the level of the amygdala. Exposure throughout pregnancy to glucocorticoids increases expression of glucocorticoid receptors in the BLA of adult offspring, and also reduces exploratory behavior (Welberg et al., 2001). Changes in the CRF system may also contribute, as late prenatal stress (from E14-21) increases CRF content in and release from the amygdala in adulthood (Cratty et al., 1995). The effects of PS may be directly mediated by exposure of the fetus to stress hormones. In support of this argument, glucocorticoid receptor expression is increased in the BLA of adult offspring following prenatal inhibition of the placental barrier enzyme for maternal stress hormones, 11β-hydroxysteroid dehydrogenase 2 (Welberg et al., 2000). However, because these studies only measured effects in adulthood, they do not preclude that increased stress sensitivity in childhood or adolescence contribute to the deficits observed, or whether changes in amygdala reactivity to stress may emerge downstream of other effects on the brain. In fact, ELS may perturb stress reactivity of the developing BLA through effects on serotonin release, which provides negative feedback on the stress response (for review, see Joels and Baram, 2009). ELS from P16 to P20 reduces the concentration in the amygdala of serotonin and its main metabolite (Matsui et al., 2010). Further supporting the need to consider developmental timing of stress and its effects, the same stressor applied from P11 to P15 has the opposite effect on these molecules in the amygdala (ibid.) 22.214.171.124 ELS Causes Precocious Amygdala Development Many effects of ELS on the amygdala can be simplified as accelerated maturation. For instance, exposure to ELS causes precocious activation of the amygdala to fearful faces.
Typically, adults exhibit greater amygdala activation to fearful faces than neutral faces, while children do not (Davis and Whalen, 2001; Thomas et al., 2001a). However, children exposed to ELS, exhibit the adult phenotype: greater amygdala activation to fearful than neutral faces (Maheu et al., 2010; Tottenham et al., 2011). The hypothesized precocious development of the BLA following ELS is supported by studies of CPs in the development of two amygdaladependent behaviors: the emergence of avoidance learning and the switch in fear extinction mechanism from erasure to suppression.
As described in Section 126.96.36.199, the rodent amygdala is not activated in response to footshocks at P8, and stimuli that are aversively conditioned at this age elicit an approach response (Sullivan et al., 2000). ELS from P1 to P7 caused precocious expression of aversive learning and amygdala activation at P8, an effect blocked by glucocorticoid receptor antagonists at the time of training (Moriceau et al., 2009). This study suggests ELS accelerates development of the stress response, which causes precocious activation of the amygdala. It is important to note that changes to the amygdala itself may act upstream of the altered stress response. Acute stress later in development can also cause precocious fear learning, in this case the expression of fearpotentiation of startle responses at P20 instead of P22, suggesting ELS may generally act to accelerate amygdala activation and fear learning (Yap and Richardson, 2007).
ELS also leads to precocious expression of the mature form of extinction learning. As described in Section 188.8.131.52, extinction of conditioned fear at P17 leads to fear erasure, but extinction at P24 consists of suppression of conditioned fear that allows for fear renewal and reinstatement (Kim and Richardson, 2007, 2008). However, precocious expression at P17 of mature extinction, allowing for fear renewal and reinstatement, is caused by ELS (Callaghan and Richardson, 2011). In this case, ELS consists of separation of the pup from the mother from P2In addition, maternally separated rats exhibit much longer retention of fear that is conditioned at P17 (Callaghan and Richardson, 2012). Interestingly, recent work suggests the precocious fear learning and extinction caused by chronic ELS can be mimicked by a stressor lasting only 24 hours (Cowan et al., 2013).
Precocious activation of the amygdala following ELS may be due to accelerated integration into the fear circuit. Another form of ELS, early weaning at P14, causes anxiety-like behavior as early as P21 and accelerates myelination of the BLA (Ono et al., 2008). We propose strengthening the early connections of the BLA comes at the expense of the connections that form late in development, including those with the frontal cortices. In other words, ELS shifts amygdala connectivity towards brain regions it interacts with early, at the expense of those connections that typically emerge late. Human imaging studies support this hypothesis; ELS diminishes the integrity of fiber bundles connecting the amygdala and PFC (Eluvathingal et al., 2006; Govindan et al., 2010). This hypothesis fits with the model for diminished top-down control of the amygdala in adolescence contributing to psychopathology (Correll et al., 2005;
Casey et al., 2010).
Decreased receptivity following ELS of the BLA to late-developing inputs could theoretically be achieved through precocious closing of a CP of plasticity. As explained in Section 184.108.40.206, developmental CPs can be closed due to activation of GABAA receptors, specifically those containing the α1 subunit (Huntsman et al., 1994; Hensch et al., 1998; Fagiolini et al., 2004). In addition, CP closure coincides with the emergence of PNNs on PV+ interneurons (Pizzorusso et al., 2002; Hensch, 2005; Dityatev et al., 2007; Nowicka et al., 2009), which themselves preferentially innervate α1-containing synapses (Nusser et al., 1996; Fritschy et al., 1998; Pawelzik et al., 1999; Thomson et al., 2000; Nyiri et al., 2001; Klausberger et al., 2002). In the following section, the potential link between ELS and altered CP timing is supported with evidence for experience-dependent regulation of GABAergic function and CP timing.
220.127.116.11 Experience-Dependent Changes to Critical Period Timing ELS is likely to alter the development of BLA inhibition, because early sensory experience is known to influence the development of GABAergic transmission and neurons in brain regions involved in processing that information in the adult. For example, early auditory experience influences the development of GABAergic synapse strength and the localization and kinetics of GABAA receptors (for review, see Sanes and Kotak, 2011). Specifically, hearing loss during a developmental CP prevents the typical, age-dependent increase in the strength of inhibitory synapses and density of GABAA receptors and the loss of short-term synaptic depression (ibid.). Furthermore, deprivation during CPs in a variety of sensory modalities reduces the content of GABA and its synthesizing enzymes (Hendry and Jones, 1986; Benevento et al., 1995; Jiao et al., 2006) and the density of GABAergic neurons and synapses in sensory cortex (Gabbott and Stewart, 1987; Micheva and Beaulieu, 1995). Sensory experience triggers balancing of levels of inhibition and excitation in the cortex, which leads to improved circuit performance (Dorrn et al., 2010).
The effects of sensory experience on CP expression also include changes to PV + interneurons and PNNs. Sensory deprivation delays the emergence of parvalbumin expression and the formation of perisomatic baskets on pyramidal neurons (Chattopadhyaya et al., 2004; Jiao et al., 2006). In songbirds, song learning corresponds with increased PNN expression on PV+ interneurons in a brain area involved in said learning. Furthermore, song deprivation reduces the expression of both parvalbumin and PNNs (Balmer et al., 2009).
Early stressful experience may trigger changes in BLA inhibition by altering neuronal activity in the developing amygdala. Numerous studies have identified a direct impact of synaptic transmission on GABAergic circuit development, with blockade of GABAA receptor activation mimicking the effects of sensory deprivation (for reviews, see Akerman and Cline, 2007; Katagiri et al., 2007; Huang, 2009). For instance, blockade of activity in developing cultured networks leads to the loss of GABA and GABAergic neurons (Ramakers et al., 1994; de Lima et al., 2004).
Reducing neurotransmission during development also delays the emergence of parvalbumin (Patz et al., 2004) and reduces PNN expression (Dityatev et al., 2007). Genetic knockdown of GABA synthesizing enzymes interferes with the formation of perisomatic basket synapses, which are preferentially innervated by PV+ interneurons (Chattopadhyaya et al., 2007). Promoting GABAergic transmission by blocking reuptake or applying receptor agonists rescued this deficit, further implicating GABAergic transmission in promoting GABAergic synapse formation and maintenance. GABAergic transmission in the embryonic brain promotes survival of PV+ interneurons (Luk and Sadikot, 2001). GABA is excitatory early in development not only for principal neurons, but also for incipient PV+ interneurons, suggesting GABA release may promote parvalbumin neuron survival through direct synaptic effects on these neurons (Sauer and Bartos, 2010). In support of this notion, the maturation of GABA expression depends on activityinduced calcium influx (Spitzer et al., 1993).