«Distribution Agreement In presenting this thesis or dissertation as a partial fulfillment of the requirements for an advanced degree from Emory ...»
The function of the HPA axis is not consistent throughout life, and developmental changes in the stress response may contribute to vulnerability to early-life stress. Normative increases in stress reactivity occur during childhood and adolescence, and may generally be adaptive for the novel experiences of this developmental period (Spear, 2009). For example, prepubertal rodents exhibit a prolonged response to acute stress compared to adults, corresponding with increased activation of the paraventricular nucleus of the hypothalamus (Romeo et al., 2006). However, this window of stress hyper-reactivity may intensify a propensity for excessive anxiety in at-risk individuals (Spear, 2009). In contrast, infancy seems to represent a period of reduced sensitivity to stress. In rats, the stress response reaches maturity at the age of weaning (Takahashi et al., 1991). In contrast, during infancy, in rodents there exists a period of reduced responsiveness to stressors, called the “stress hypo-responsive period” (SHRP) or “stress non-responsive period” (Sapolsky and Meaney, 1986). An analogous window of reduced stress reactivity is thought to be expressed in infant humans as well (Gunnar and Quevedo, 2007). Rat pups during this period have low basal and stress-induced cortisol levels (for review, see Levine, 2005). The SHRP is thought to involve active suppression of the stress response (Levine, 2001).
18.104.22.168 Mediators of Early Life Stress Sensitivity The potent effects of ELS may be explained by heightened sensitivity of the immature brain to chemical mediators of the stress response. The SHRP has been suggested to protect the developing brain from stress hormones (Sapolsky and Meaney, 1986; Francis et al., 1999b), and particularly stressful experiences can elicit a stress response during the SHRP and may thereby cause later life deficits (Levine, 1967; Ladd et al., 2004). The SHRP may be necessary because the effects of CRF are more potent in the immature brain – infants are far more sensitive than adults to the pro-convulsant effects of CRF, in part because they possess a much higher density of CRF receptors in the brain (Baram and Schultz, 1991; for review, see Korosi and Baram, 2008).
Expression of the CRF1 receptor peaks at 180% of adult values during infancy in the amygdala, a critical brain region for the expression and sensitivity to stress that is introduced in Section 1.2 (Avishai-Eliner et al., 1996). During development from conception through P21, transient overexpression of CRF in the amygdala and throughout the forebrain is anxiogenic later in life, suggesting early CRF sensitivity of the brain may be a key mediator of ELS effects (Kolber et al., 2010).
The effects of PS are mediated, at least in part, by exposure of the fetus to maternal glucocorticoids (Barbazanges et al., 1996; Talge et al., 2007; Field and Diego, 2008). However, the role of glucorcorticoids is still unclear, considering maternal cortisol responses to stress are reduced late in pregnancy, but early in pregnancy, maternal cortisol may be less accessible to the fetus (Talge et al., 2007). One explanation is that, because PS can reduce expression of 11βHSD2, the placental enzyme responsible for metabolizing cortisol and insulating the fetus (for review, see O'Donnell et al., 2009), chronic PS may weaken the placental barrier and subsequently influence the fetus late in gestation.
PS may also influence offspring through effects on maternal behavior, which has been shown to influence the development of the stress response in the offspring. Variations in maternal care are thought to reflect aspects of the environment, and as such can influence emotional behavior in the offspring (for review, see Zhang et al., 2006). In support of the notion that maternal behavior mediates some effects of PS, cross-fostering to control dams can attenuate some effects of late PS (Barlow et al., 1978; Maccari et al., 1995; however, see Del Cerro et al., 2010). Furthermore, fostering offspring of control dams to dams exposed to stress during pregnancy introduces some mild effects of PS. These findings suggest the effects of PS are mediated both directly at the time of stress and indirectly by maternal behavior after birth (Barlow et al., 1978).
1.1.3 ‘Critical Periods’ in Development as Windows of Vulnerability In order to understand how ELS contributes to anxiety disorder etiology, it is necessary to consider its effects in the context of ongoing development of the nervous system. Developmental changes in brain function have been hypothesized to increase an organism’s sensitivity to environmental influences (Casey et al., 2000). This notion is exemplified by ‘critical periods’ (CP), defined as specific developmental windows when ongoing changes align to produce heightened sensitivity to stimuli. Our understanding of CPs has been derived mostly from research in development of the visual system. During visual development, temporary occlusion of a single eye leads to permanent loss of acuity in that eye, which is achieved through reorganization of visual cortex to provide less cortical area for the processing of information from the occluded eye (Wiesel and Hubel, 1963). This effect can only be achieved during a specific developmental window, due to heightened sensitivity to the visual occlusion because of ongoing reorganization of synapses in the visual cortex that naturally occurs during said window (for review, see Hensch, 2005).
It is possible to extend this concept of CPs to the study of emotional development, raising the hope of identifying developmental windows and corresponding processes that are most vulnerable to risk-factors for psychiatric disease. CPs in the development of emotional behaviors have been suggested to reflect CPs for underlying emotional brain circuits (Machado and Bachevalier, 2003). Human studies are beginning to provide support for considering CPs as windows of vulnerability for the effects of ELS (Andersen et al., 2008). However, interpreting the impact of ELS on the development of healthy and maladaptive emotion requires consideration of normative brain development, with specific focus on windows of heightened plasticity and the changing brain circuits that may exhibit CPs for stress sensitivity. To this end, Section 1.2 introduces the basolateral amygdala, a brain region that is implicated in the pathophysiology of anxiety disorders, undergoes protracted and marked development, and is sensitive to the effects of stress across the lifespan.
1.2 Juvenile and Adult Anxiety Disorders Involve Dysfunction of the Amygdala The amygdala is a temporal lobe structure critically involved in emotional processing and behavior. The involvement of the amygdala in social and emotional cognition was first described through gross experimental and specific lesions, (Rosvold et al., 1954; Pellegrino, 1968; Kling et al., 1970; Adolphs et al., 1994; Adolphs et al., 1998; Aggleton and Saunders, 2000). We now know the amygdala functions to assign emotional valence to sensory input, including social stimuli (Adolphs et al., 2000; Davis and Whalen, 2001; Adolphs, 2010) and plays a key role in both the experience and learning of fear and safety (Goddard, 1964; Davis, 2000; LeDoux, 2000).
The amygdala contributes not only to the perception and processing of negative stimuli, but we now know it is involved in positive emotional responses including reward (Breiter et al., 1996;
Maren, 2003; Hennenlotter et al., 2005; Costafreda et al., 2008).
The amygdala is comprised of many distinct nuclei that act together in a well described circuit, together known as the extended amygdala. Aligned with its role in sensory processing, the amygdala receives input from a number of primary and higher-order sensory regions, the majority of which enters at its basolateral complex, which includes the basolateral nucleus of the amygdala (BLA; McDonald, 1998). The BLA is critical for the production of appropriate emotional responses and the processing of emotional memories (Davis et al., 2003; LeDoux, 2007; Pape and Pare, 2010; Stuber et al., 2011). The BLA not only receives input from a wide variety of brain regions, but also has widespread output: it innervates amygdala output nuclei like the central nucleus (CeA) and the bed nucleus of the stria terminalis (BNST), which mediate fearful and anxious behaviors (for review, see Walker and Davis, 2008); innervates sensory areas to influence sensory perception and memory (Pessoa and Adolphs, 2010; Chavez et al., 2013; Chen et al., 2013); and innervates the nucleus accumbens to regulate the reward system (Ambroggi et al., 2008; Stuber et al., 2011).
1.2.1 Implicating Amygdala Dysfunction in Anxiety Disorders The amygdala contributes to the pathophysiology of both adult and juvenile anxiety disorders. The amygdala as a whole was first implicated in anxiety because lesions in humans have anxiolytic effects (Narabayashi et al., 1963). A variety of studies have, more recently, implicated amygdala dysfunction, particularly that of the BLA, in the pathophysiology of anxiety disorders (Davis et al., 1994; Quirk and Gehlert, 2003; Rainnie et al., 2004; Boyle, 2013).
Furthermore, changes to amygdala function following traumatic events may underlie disorders like PTSD (Bremner, 2007; Rainnie and Ressler, 2009), consistent with the allostatic load model (McEwen, 2007).
The amygdala, along with other brain structures involved in emotional regulation like the prefrontal and cingulate cortices and the hippocampus, is implicated in the pathophysiology a variety of juvenile affective disorders (Mana et al., 2010). Functional imaging studies indicate that children and adolescents with anxiety disorders exhibit increased amygdala activation, consistent with findings in adults. Excessive activation of the amygdala is also observed in healthy juveniles at high risk for anxiety, including children of parents with anxiety disorders, suggesting amygdala hyper-reactivity is not only a symptom but a risk-factor for anxiety disorders (for review, see Blackford and Pine, 2012). Anxious children show exaggerated amygdala responses to fearful faces, while children with major depression have blunted amygdala responses to the same stimuli. The degree of amygdala hyper-activation positively correlates with the severity of anxiety symptoms (Thomas et al., 2001b).
Hyper-activation of the amygdala in adolescence may be due to limited top-down control of its activity by frontal cortices. In its role as a hub, the BLA innervates the amygdala output regions, the CeA and BNST, and can regulate the activity of the amygdala as a whole (Figure 1.1). Via inputs to the BLA, frontal cortices including the medial, ventrolateral, and ventromedial prefrontal cortices (mPFC, vlPFC, and vmPFC) and the anterior cingulate cortex regulate amygdala function and emotional reactivity (Ochsner and Gross, 2005; Eippert et al., 2007;
Goldin et al., 2009; Sotres-Bayon and Quirk, 2010; Schulze et al., 2011). An imbalance in the activity of the amygdala and prefrontal cortex during adolescence, brought about by relatively delayed cortical development (Van Eden and Uylings, 1985; Bourgeois et al., 1994), may underlie the heightened emotionality and susceptibility to psychiatric disease onset during adolescence (Drevets, 2003; Yurgelun-Todd, 2007; Casey et al., 2010; Somerville et al., 2010).
Further supporting the late emergence of BLA-PFC interactions, projections from the BLA to the mPFC and cingulate cortex continue to mature and increase in density during adolescence (Cunningham et al., 2002, 2008).
Indeed, a number of studies show deficits in PFC function, particularly in terms of its regulation of the amygdala, in juveniles with anxiety disorders. Negative functional connectivity of the amygdala and vlPFC, indicative of cortical suppression of amygdala activity, is diminished in the youths with generalized anxiety disorder and correlates with symptom severity and amygdala response to angry faces (Monk et al., 2008). Adolescents with higher trait anxiety exhibit less habituation of amygdala responses in repeated emotional tests, also correlating with diminished functional connectivity of the amygdala and vlPFC (Hare et al., 2008).
Diminished top-down regulation of the amygdala, caused by excessive stress, may promote the emergence of anxiety disorders. Chronic stress reduces the influence of vmPFC afferents on amygdala sensitivity to emotional stimuli (Correll et al., 2005). Importantly, children previously exposed to socioemotional deprivation as orphans exhibit diminished integrity of the fiber bundle connecting the amygdala and PFC, the uncinate fasciculus (Eluvathingal et al., 2006;
Govindan et al., 2010). These studies suggest chronic stress and ELS can cause deficits in the regulation of the amygdala by frontal cortices.
As explained in Section 1.2.2, the amygdala plays a role initiating the stress response, and in turn its function and physiology can be altered by stress. Towards understanding how stress exposure early in life alters emotional development, Section 1.3 covers the normative development of the amygdala and Section 1.4 details how that development is altered by ELS.
1.2.2 The BLA in the Responses to Acute and Chronic Stress The amygdala helps coordinate many aspects of the stress response and is situated to do so via widespread connectivity throughout the brain (Goldstein et al., 1996; Herman et al., 2003;
Jankord and Herman, 2008; Dedovic et al., 2009). The BLA sends projections to the CeA and BNST, which in turn target the hypothalamus and brainstem nuclei to activate the stress response (for review, see Walker and Davis, 2008). The CeA and BNST are thought to be involved in phasic and sustained fear responses, respectively, with the latter more akin to anxiety (ibid.). The amygdala may contribute, more specifically, to initiating the response to psychological, rather than physical, stressors (for review, see Herman and Cullinan, 1997). Traumatic experiences correspond with reductions in amygdala volume and decreased amygdala reactivity, suggesting effectors of the stress response can, in turn, alter amygdala function (for review, see Shin et al., 2006).
Many chemical mediators of the stress response influence the function of neurons in the amygdala (Rodrigues et al., 2009). In vitro application of corticosterone reduces inhibitory synaptic transmission, mediated by the neurotransmitter γ-aminobutyric acid (GABA), in the BLA (Duvarci and Pare, 2007). Norepinephrine, a mediator of the stress response, suppresses GABAergic transmission in the amygdala and promotes synaptic plasticity (Tully et al., 2007).