«Distribution Agreement In presenting this thesis or dissertation as a partial fulfillment of the requirements for an advanced degree from Emory ...»
However, some of the most robust effects on amygdala function are caused by CRF, the stress neurohormone. CRF is released in the amygdala following acute stress and contributes to the behavioral aspects of the stress response, including anxiety-like behaviors (Merlo Pich et al., 1995; Gray and Bingaman, 1996; Sajdyk et al., 1999; Cook, 2004). Furthermore, CRF release in the BLA promotes aversive memory formation (Roozendaal et al., 2002). The effects of acute CRF exposure are explained, in part, by its impact on the function of BLA principal neurons, including increased excitability and reduced action potential slow afterhyperpolarization (Rainnie et al., 1992), as well as enhanced sensitivity to afferent stimulation (Ugolini et al., 2008). These effects on amygdala inputs and neuronal excitability, combined with attenuation of GABAergic transmission in the BLA, promote synaptic plasticity and thereby contribute to stress-induced facilitation of the learning of fearful associations (Rodriguez Manzanares et al., 2005).
The effects of acute and chronic stress can differ wildly, with chronic exposure to stress often found to be detrimental (Brunson et al., 2003). There are mixed reports of the effects of chronic stress on HPA axis function (Miller et al., 2007), suggesting the link between chronic stress and psychopathology may occur elsewhere in the brain. Following chronic stress, exaggerated responses to subsequent stressors and psychiatric disorder pathogenesis may be due to well-documented alterations to amygdala function (McEwen, 2007; Roozendaal et al., 2009).
Specifically, stress-induced release of CRF has been hypothesized to induce plasticity in the BLA that contributes to the persistent overexpression of anxiety (Shekhar et al., 2005). Chronic stress in rats increases dendritic arborization and spine density of BLA principal neurons (Vyas et al., 2002; Vyas et al., 2006; Cui et al., 2008). Chronic activation of CRF receptors in the BLA has a long-lasting (30 day) anxiogenic effect, with a long-term attenuation of GABAergic transmission in the BLA that causes hyper-excitability (Rainnie et al., 2004).
Even weak or acute activation of the stress response can cause deficits in amygdala function. Subthreshold activation of CRF receptors in the BLA, which does not cause acute behavioral effects, sensitizes animals to panicogenic agents (Sajdyk et al., 1999). Like CRF, subthreshold doses of GABA receptor antagonists in the BLA ‘prime’ anxiety-like behavior and physiological aspects of the stress response (Sanders et al., 1995). This suggests the long-term effects of CRF in the amygdala may act via alterations in GABAergic transmission. Acute stress may also promote anxiety through actions on the amygdala; for instance, long term enhancement of fear learning by acute stress exposure coincides with reduced expression of GABA receptor subunits in the BLA (Ponomarev et al., 2010).
1.2.3 GABA, a Key Regulator of BLA Function As mentioned above, GABAergic transmission in the BLA is sensitive to acute and chronic stress. This is relevant to the pathogenesis of anxiety for two reasons: 1) GABA is a key neurotransmitter in the amygdala that is linked to adult anxiety, and 2) GABA systems are classically regulated by brain development. For these reasons, GABAergic dysfunction is proposed to be a key mediator of neurodevelopmental disorders (Chattopadhyaya and Cristo, 2012; King et al., 2013).
GABAergic transmission has been well characterized in the mature BLA (Washburn and Moises, 1992; Martina et al., 2001). GABAergic transmission, which mediates all of the fast synaptic inhibition in the amygdala, occurs primarily from BLA interneurons. The BLA is a cortical-like structure, with a similar distribution of cell types as the neocortex and hippocampus – a mix of principal neurons and inhibitory, GABAergic interneurons (Carlsen and Heimer, 1988). Principal neurons of the BLA are excitatory, projection neurons that constitute approximately 80-85% of its total cell population and mediate all of the output of the nucleus.
The remainder is local-circuit interneurons that release GABA and inhibit the neighboring principal neurons (McDonald, 1985; Rainnie et al., 1993). GABAergic fibers in the BLA mainly originate from local interneurons, but also include extrinsic inhibitory inputs from nearby paracapsular intercalated cells, as well as the basal forebrain (Marowsky et al., 2005) (Mascagni and McDonald, 2009). BLA interneurons are highly interconnected, being targets of BLA principal neurons and extrinsic excitatory fibers; activation of excitatory inputs to the BLA elicits robust feed-forward inhibition that limits the response of the amygdala (Rainnie et al., 1991a).
GABA determines the excitability of the amygdala and thereby regulates emotional behavior and learning (Shekhar et al., 2003; Ehrlich et al., 2009). A balance of excitatory and inhibitory synaptic transmission is critical for the function of local circuits in the brain (Shu et al., 2003). Activation of amygdala principal neurons is thought to bidirectionally influence anxiety states and aversive learning; GABAergic transmission in the amygdala is suggested to alleviate anxiety by reducing excitability of the nucleus as a whole. To this end, blocking GABA receptors in the BLA promotes fearful and anxious behaviors (for review, see Quirk and Gehlert, 2003), whereas enhancing GABA function attenuates these behaviors (Davis et al., 1994; Sanders and Shekhar, 1995). Furthermore, ablation of a subset of BLA interneurons reduces sociability (Truitt et al., 2007). GABA receptor activation blocks signal propagation throughout the amygdala (Wang et al., 2001), and loss of GABAergic transmission in the amygdala results in generalization of fearful associations (Shaban et al., 2006). Acute stress induces GABA release in the amygdala (Cook, 2004), but chronic stress attenuates GABA release caused by subsequent stressors, suggesting reductions in amygdala GABAergic transmission may contribute to stress allostasis (Reznikov et al., 2008). GABAergic transmission in the amygdala is, for all these reasons, a classic target of anxiolytic drugs, including benzodiazepines and barbiturates (Sandford et al., 2000).
Described further in Section 220.127.116.11, the amygdala plays a well-defined role in a form of associative fear learning called classical or Pavlovian fear conditioning. In classical fear conditioning, a benign (conditioned) stimulus is repeatedly presented along with a noxious (unconditioned) stimulus, and the conditioned stimulus comes to independently elicit a fear response. Synaptic input representing the conditioned and unconditioned stimuli converge in the BLA, where synaptic plasticity is thought to enable the conditioned stimulus to elicit the same response as the unconditioned stimulus (Fendt and Fanselow, 1999; Davis, 2000; LeDoux, 2007).
GABA receptors exert control over classical fear conditioning, including the acquisition and expression of memories for the fear association (Ehrlich et al., 2009).
Not only does BLA inhibition regulate fear learning, but it is also essential for the learned suppression of conditioned fear known as ‘extinction.’ Learning to extinguish fearful associations is particularly relevant for anxiety disorders, which involve excessive fear responses to benign stimuli that must be curbed by treatment (Coles and Heimberg, 2002; Lissek et al., 2005; Rainnie and Ressler, 2009). In extinction training, the formerly conditioned CS is repeatedly presented without the US, and over time the fearful response is curbed as the CS loses predictive value.
This learning depends on the formation of a new memory in the BLA that inhibits the fearful memory (Rescorla, 2001; Barad et al., 2006; Bouton et al., 2006; Myers and Davis, 2007), and requires strengthening of GABAergic synapses in the amygdala (Lin et al., 2009). Furthermore, infusion of a GABA receptor agonist in the BLA enhances extinction (Akirav et al., 2006). Fear extinction enhances expression in the amygdala of genes related to GABAergic function, and fear conditioning causes a downregulation of GABA-related genes, suggesting the amygdala GABA system can be bidirectionally modulated to control the balance of fear and safety learning (Heldt and Ressler, 2007).
18.104.22.168 Parvalbumin Interneurons Organize BLA Activity The most common class of interneuron found in the amygdala, cerebral cortex, and hippocampus is identified by its expression of the calcium-binding protein, parvalbumin (McDonald and Pearson, 1989; Baimbridge et al., 1992; Kemppainen and Pitkanen, 2000;
McDonald and Mascagni, 2001; Schwaller et al., 2002). Parvalbumin-expressing (PV+) interneurons comprise a large portion of basket cells, which form periosmatic baskets on principal neurons that provide robust inhibition at the soma (Somogyi et al., 1983; Bartos and Elgueta, 2012). In the BLA, these neurons have a variety of electrophysiological properties and include both burst-firing and stutter-firing interneurons (Rainnie et al., 2006). PV+ neurons in the BLA and elsewhere in the brain innervate hundreds of neighboring neurons and can form multiple synapses on each of those neurons (Tamas et al., 1997; Wang et al., 2002; Rainnie et al., 2006).
As shown in the BLA and elsewhere, the diverse projections of PV+ interneurons are useful for coordinating the activity of groups of principal neurons (Chapter 6; Cobb et al., 1995;
Miles et al., 1996; Ryan et al., 2012). PV+ basket cells control rhythms in neural networks, organizing their target neurons into oscillations that are critical for network function (for review, see Freund and Katona, 2007). In the hippocampus, PV+ neurons contribute to gamma oscillations in the hippocampus and thereby support memory function (Bartos et al., 2002; Fuchs et al., 2007; Sohal et al., 2009). In the BLA and throughout the brain, PV+ neurons form a syncytium with dendritic and axonal gap junctions, coordinating their activity and supporting their role in oscillation production (Muller et al., 2005). Supporting the critical role of PV+ interneurons in the BLA circuit, acute stress and a variety of anxiogenic agents cause activation of parvalbumin neurons in the rat BLA as measured by immediate early genes (Reznikov et al., 2008; Hale et al., 2010).
While parvalbumin is typically considered solely a neuronal marker, its expression does influence the function of interneurons. Parvalbumin functions as a so-called “slow-onset” calcium buffer in synaptic terminals, and genetic deletion of parvalbumin enhances short-term facilitation of synapses, strengthening repetitive GABAergic release events (Caillard et al., 2000;
Vreugdenhil et al., 2003; Collin et al., 2005; Orduz et al., 2013). However, in neurons with low parvalbumin concentrations, it has no effect on synaptic transmission (Eggermann and Jonas, 2012). Furthermore, the concentration of parvalbumin in individual neurons and its variability across neurons differs by brain region (ibid.). Increases in parvalbumin concentration also reduce the excitability of fast-spiking interneurons (Bischop et al., 2012; Orduz et al., 2013). Together these data support a role for the expression of parvalbumin protein in reducing GABAergic transmission through synaptic depression and reducing interneuron excitability. In line with this logic, an inbred rat strain with high emotionality and anxiety has greater density of PV+ neurons in the BLA (Yilmazer-Hanke et al., 2002).
22.214.171.124 GABA Promotes Network Oscillations Observed During Fear As mentioned above, PV+ interneurons play an important role in organizing neurons into network oscillations. These neurons exemplify the complex role of GABA in the brain; this neurotransmitter not only serves to dampen the activity of amygdala neurons, but can also promote BLA activation through the organization of neuron firing (Chapter 6; Ryan et al., 2012).
Coordinated inhibitory input across multiple neurons is a common mechanism to synchronize their action potential firing and generate network oscillations (Soltesz and Deschenes, 1993;
Buzsaki, 1997; Penttonen et al., 1998; Pouille and Scanziani, 2001; Person and Perkel, 2005;
Sohal et al., 2006; Szucs et al., 2009). Optogenetic silencing of PV+ neurons in awake, behaving mice influences the phase of ongoing theta (6-10 Hz) oscillations, supporting a role for these neurons in coordinating slow network oscillations (Royer et al., 2012).
Recent evidence suggests oscillatory activity of neurons in the BLA plays a key role in regulating affect in awake, behaving animals (for review, see Pape and Pare, 2010). More specifically, it is now evident that the amygdala, hippocampus, and prefrontal cortex produce coordinated high delta / low theta (4-5 Hz) oscillations during acquisition (Madsen and Rainnie,
2009) and retrieval (Sangha et al., 2009) of learned fear, which then diminish over the course of subsequent extinction learning. The importance of network oscillations for BLA function emphasizes the critical role PV+ interneurons play in amygdala function. As discussed in the following section, the developmental emergence of PV+ interneurons is one of many indicators of the changing function of the amygdala throughout postnatal life.
1.3 Maturation of Amygdala Physiology and Function Towards the goal of understanding the etiology of anxiety disorders and interpreting the effects of ELS within a developmental framework, this section describes the current state of knowledge regarding amygdala development – in terms of function, physiology, morphology, connectivity, and synaptic transmission. Section 1.3.1 contains an argument for the role of amygdala development in shaping emotional outcomes. Then in Sections 1.3.2 and 1.3.3, respectively, the current knowledge is presented regarding developmental changes to the function of the primate and rodent amygdala. To facilitate comparisons across species, it is useful to consider the following developmental time points: a baby rat has comparable cortical maturity to a newborn human around postnatal day (P)12, which coincides with opening of the eyes and the onset of the capacity to hear; rats are weaned around P21, compared to around 6 months of age for rhesus macaques and humans; rats reach sexual maturity between P40 and P60, compared to 4-5 years old for macaques; rat life expectancy is around 3 years, while macaques typically live less than 40 years (Cork and Walker, 1993; Prather et al., 2001; Coe and Shirtcliff, 2004; Quinn, 2005; Tritsch and Bergles, 2010). Here, we will focus on postnatal periods including infancy and early adolescence, when the BLA undergoes rapid and drastic maturation.