Search

Social buffering is broadly defined as an individual’s ability to suppress the physiological, behavioral, and/or emotional consequences of adverse events in another individual. A particularly potent and vital form of social buffering is caregiver buffering, which protects the developing infant brain from the deleterious effects of stress. Rodent work has provided valuable information about the behavioral, endocrine, and neurobiological mechanisms of maternal regulation of threat and how those mechanisms may be altered by early life experiences. The ability of infant rats to acquire Pavlovian odor-shock associations generally emerges when they are about ten days old, but is under tight regulation by the mother. Previous studies have shown that when infant rats underwent an odor-shock Pavlovian threat learning experience in the presence of an anesthetized mother, they did not avoid the conditioned odor when tested, unlike pups conditioned without maternal presence. This behavioral effect was, in part, mediated by the mother’s ability to suppress the infant rat’s stress response and amygdala reactivity during the threat learning experience. Follow-up studies have shown that disruption of the relationship between mother and infant can affect the ability of the mother to regulate fear in her infants. In this thesis, I address several remaining questions about the functions and underlying mechanisms of maternal buffering in infant rats. In Chapter 1, I briefly review the existing human and rat literature on caregiver regulation of stress, threat learning, and neural activity and review the trajectory of infant rat development. In Chapter 2, I demonstrate that the effect of maternal presence during a threat learning experience can be observed using a robustly studied defense response – threat conditioned-induced freezing – and that female infant rats may be more susceptible to maternal buffering of freezing than male infant rats. In Chapter 3, I examine the functional networks engaged by infant rats conditioned with and without maternal presence and apply graph theory to analyze patterns of immediate early gene expression. Overall functional connectivity was significantly increased in pups conditioned with maternal presence. A graph theoretical analysis revealed that the network engaged by pups conditioned with maternal presence was more integrated and lacked distinctive hubs; in contrast, the network engaged by pups conditioned without maternal presence was more segregated and had distinct hubs: the lateral amygdaloid nucleus, dorsolateral part, basolateral amygdaloid nucleus, anterior part, and medial amygdaloid nucleus, anterior dorsal part. In Chapter 4, I examine the ontogeny of threat learning and stress responsivity in a vulnerable phenotype prone to high-anxiety like behavior in adulthood, and then question whether maternal presence and fibroblast growth factor 2 are capable of regulating threat learning in these animals. In Chapter 5, I summarize results from each chapter and discuss future directions. The findings outlined in this thesis provide an important next step in characterizing maternal buffering and illuminate exciting topics for future inquiry.

Anxiety and fear-based disorders are among the most common mental health disorders globally, however, current pharmacological and psychological interventions for anxiety are often associated with poor long-term treatment outcomes. A greater understanding of the neural mechanisms and circuits underlying fear acquisition and inhibition is crucial for the improvement of current interventions and the development of new and more selective treatments. This thesis investigated the causal role of glutamatergic basolateral amygdala (BLA) principal neurons and their output pathways in fear learning, extinction learning, and fear expression using a rat model of Pavlovian fear conditioning and extinction. First, it was shown that two major BLA output pathways, to the central nucleus of the amygdala (CeA) and to the nucleus accumbens (Acb), are anatomically intermingled and formed from largely distinct subpopulations of neurons. Next, it was shown that optogenetic inhibition of the global population of BLA glutamatergic neurons at the time of a footshock unconditioned stimulus (US) during fear acquisition impaired fear learning, whereas this inhibition at the time of US omission during extinction training augmented extinction learning. The next set of experiments showed how these effects related to the distinct BLA output pathways. Silencing the BLA to CeA pathway recapitulated the global BLA population findings, showing that photoinhibition at US onset in acquisition impaired fear learning, while photoinhibition at US omission in extinction augmented extinction learning. Silencing this pathway at CS onset had no effect on expression. In contrast, silencing the BLA to Acb pathway at US onset augmented fear learning, whereas silencing this pathway at US omission in extinction and at CS onset had no effect on extinction learning and fear expression, respectively. Taken together, these findings confirm that BLA principal neurons mediate fear learning due to their activity during the US and show for the first time that presentation of the US in fear learning initiates two functionally opposing processes via distinct BLA output pathways.

The exploration of how learning alters neural coding to guide behaviors remains fundamental to neuroscience. At the most basic level, the ability for organisms to flexibly adapt to changing environments and situations is paramount to biological success and often manifests in behavioral responses controlled by neural activity. For example, organisms must modify their behavior to defensive responses in the face of biological threat. Neural circuitry is involved in coordinating an initial defensive behavioral response but must undergo reorganization in order to reliably employ defensive responses in subsequent encounters based on a cue that signals imminent danger. One such form of this learning is associative fear learning, in which an organism learns to associate an initially neutral stimulus, which by itself has no biological relevance, with an innately fear-inducing stimulus. After temporally pairing the two stimuli, organisms learn that the initially neutral stimulus predicts the fear-inducing stimulus such that encountering the former induces similar behavioral responses as the latter." "Interestingly, associative fear learning is often not specific and organisms display fear responses to completely neutral stimuli that have never been associated with the innately fear-inducing stimulus. This aberrant expression of fear is referred to as generalization and is a hallmark of many psychological disorders such as anxiety and post-traumatic stress disorder. The underlying cause of fear generalization remains unknown; however, there are two predominant theories. One theory is fear generalization arises as a better-safe-than-sorry strategy, wherein sensory information remains perceptually segregated between fear predictive and neutral stimuli but the organism responds to both the same way despite the perceptual information. The second theory argues that learning alters sensory processing to make the stimuli more difficult to perceptually distinguish, which cascades into behaviorally treating them as the same. The olfactory bulb (OB) represents an ideal model system for studying the extent to which fear learning alters sensory coding in ways that support failure of perceptual discrimination due to its odotopic organization that creates unique “maps” of odor representation for each experience odor. Therefore, we used both wide-field and 2-photon (2P) calcium imaging of the OBs of awake mice in combination with classical olfactory fear conditioning to investigate how learning changes sensory representation of the conditioned odor, the initially neutral odor specifically paired with fear-inducing foot shock, as well as neutral odors never paired with shock. This allowed direct testing of the extent to which fear learning reorganizes olfactory processing in a manner that supports the failure to discriminate hypothesis of fear generalization. Both wide-field and 2P imaging revealed enhanced odor-evoked responses following fear learning that likely signal increased salience of incoming sensory information. Furthermore, the responses evoked by neutral odors became more similar to those evoked by the conditioned odor at both the population level (wide-field) and in a specific subset of OB output cells (2P), indicating neutral odors were more difficult to distinguish from the conditioned odor. Importantly, the enhanced odor-evoked responses were not attributable to behavioral state change nor top-down influence from the amygdala, the area widely believed to be involved in aberrant fear. Together, this evidence supported the failure of perceptual discrimination at early stages of sensory process as the underlying mechanism of fear generalization. We additionally investigated the role of neuromodulators in basic olfactory fear learning through in vivo pharmacology, optogenetics, and relative gene expression analysis. Manipulation of acetylcholine in the OB during olfactory fear learning established cholinergic neurotransmission can enhance the strength of learned fear and that signaling through muscarinic receptors is required for the formation of olfactory fear. Gene expression analysis revealed that several neurotransmitter receptors are downregulated 4 hours following odor-shock pairing. The majority of the downregulated genes were associated with OB inhibition suggesting that fear learning, behavioral fear generalization, and altered OB coding may arise from decreased inhibition in the OB. Altogether these results characterize neural correlates of fear generalization in the olfactory bulb at both the population and single cell level and demonstrate the importance of neuromodulation in fear learning, supporting the idea that fear generalization may initially arise from altered processing of incoming information in sensory areas. These results also highlight the importance of investigating the mechanisms of learning-induced sensory processing alterations as they relate to behavior in understanding fear generalization. This will bring insights into basic processes of learning and has the immense potential to translate to the treatment of disorders of fear generalization.

This thesis investigated the role of secure attachment priming on fear learning. Across a series of five studies, we examined the extent to which attachment primes, as compared to positive-mood matched control primes, could impact fear responses and fear learning. Specifically, we examined the influence of attachment priming on the acquisition, consolidation and extinction of fear memories. Study 1 subliminally presented primes prior to an acoustic startle probe and measured startle reflex and skin conductance as an index of autonomic responding. This study found that the attachment prime inhibited the startle but not skin conductance response after a 4-second delay. Study 2a delivered a visual imagery attachment prime prior to fear acquisition and found that relative to imagery of a positive situation, the attachment prime slowed the capacity for participants to learn to acquire fear, as indexed by fear-potentiated startle. Study 2b examined the role of the above priming methods on the unconditioned response to the aversive shock stimulus and found that the attachment prime did not influence the autonomic fear response to the shock. Study 3 investigated the role of attachment priming on fear consolidation by administering the prime following fear conditioning and examining fear recall the following day. The attachment group demonstrated lower levels of fear-potentiated startle in the fear recall test, indicating reduced fear consolidation. Finally, Study 4 delivered the prime prior to undergoing extinction training. The attachment prime enhanced long-term extinction memory in that it reduced the relapse of fear in a reinstatement test the following day. Taken together, the results presented illustrate novel ways in which fear memories can be modified by attachment priming prior to or following their development, and how safety memories can be strengthened to allow for the long-term resolution of fear memories.

Fear learning is an adaptive process that ensures the protection and survival of the organism; however, it can turn maladaptive when fear responses are chronically unregulated or overgeneralized to safe stimuli and contexts. Individual differences in personality traits such as trait anxiety have been linked to differences in fear learning and fear regulation. The study of these differences can shed light into the learning mechanisms that may be associated with vulnerability to anxiety and inform prevention and treatment interventions for anxiety. In this thesis, three studies are presented that examine the relationship between individual differences in trait anxiety and conditioned fear responding, as well as responses to aversive unconditioned events. Study 1 examines whether individual differences in trait anxiety are related to variations in fear generalization from a conditioned compound to its unreinforced components and from a conditioned singleton to an unreinforced compound containing this singleton. Study 1 found evidence for both types of fear generalization (as indexed by pupillary dilation and US expectancy ratings) occurring early during learning. However, fear generalization was not modulated by trait anxiety levels. Study 2 investigated whether individual differences in trait anxiety modulated fear conditioned responses during acquisition and extinction when each was conducted over two different contexts, and during extinction recall, conducted 24 hours after extinction. Here, no impact of trait anxiety was observed upon fear conditioned responses at acquisition when it was conducted over two different contexts. Meanwhile, at extinction conducted across two different contexts, low trait anxiety was associated with decreased differential conditioned fear (as indexed by pupillary dilation), while high trait anxiety was associated with increased fear renewal with contextual changes during extinction. At extinction recall, meanwhile, no effects of anxiety and no differential fear recovery were observed. Study 3 examined the impact of trait anxiety on responses to aversive controllable and uncontrollable stimuli, and whether prior experience of control and anxiety impacted responses to subsequent aversive uncontrollable stimuli. Here, although experience of control was associated with a reduced pupillary response to aversive stimulation relative to experience of uncontrollability, anxiety had no effect in modulating these responses. In turn, prior experience of control was associated with decreased pupil responses to aversive uncontrollable stimuli in the high anxious, but not in the low anxious subjects. Interactions of anxiety and the experimental variables were observed in Study 2 and Study 3 only. These results suggested deficits in extinction learning (Study 2) and increased potentiation of aversive responding when primed by an aversive, uncontrollable experience (Study 3), in the high anxious subjects. Failure to observe anxiety effects on fear generalization in Study 1 may have been due to fast learning of the CS-US contingencies, which had a 100% reinforcement rate, and rapid extinction of fear generalization.

Due to difficulties in elucidating neurobiological aspects of psychological disorders, the National Institute of Mental Health (NIMH) created the Research Domain Criteria (RDoC), which encourages novel conceptualizations of the relationship between neurobiological circuitry and clinical difficulties. This approach is markedly different from the Diagnostic and Statistical Manual of Mental Disorders (DSM) based approach that has dominated clinical research to date. Thus, RDoC necessitates exploration of novel experimental and statistical approaches. Fear learning paradigms represent a promising methodology for elucidating connections between acute threat (“fear”) circuitry and fear-related clinical difficulties. However, traditional analytical approaches rely on central tendency statistics, which are tethered to a priori categories and assume homogeneity within groups. Growth Mixture Modeling (GMM) methods such as Latent Class Growth Analysis (LCGA) may be uniquely suited for examining fear learning phenotypes. However, just three extant studies have applied GMM to fear learning and only one did so in a human population. Thus, the degree to which classes identified in known studies represent characteristics of the general population and to which GMM methodology is applicable across populations and paradigms is unclear. This preliminary study applied LCGA to a fear learning lab study in an attempt to identify heterogeneity in fear learning patterns based on a posteriori classification. The findings of this investigation may inform efforts to move toward a trans-diagnostic conceptualization of fear learning. Consistent with the goals laid out in RDoC, explication of fear learning phenotypes may eventually provide critical information needed to spur innovation in psychotherapeutic and psychopharmacological treatment.

Trauma and exposure to extreme stressors greatly increases a person’s vulnerability to developing mental illnesses like post-traumatic stress disorder (PTSD). Patients with PTSD often have impaired fear and safety learning, and despite the fact that women are more than twice as likely to develop PTSD, much of the research on this disorder has relied on the use of male subjects. This paper will review potential contributors to the sex differences seen in PTSD and fear-related learning. Our group has found that female rats show greater fear discrimination abilities than their male counterparts, but show no difference in levels of safety learning. Analysis of specialized extracellular matrix structures called perineuronal nets (PNNs) revealed that females displayed a much higher density of PNNs in the basolateral amygdala (BLA) than males, but not in the prefrontal cortex (PFC).

The exposure to stress has been associated with increased depressive and anxiety symptoms, yet not all individuals respond negatively to the experience of stress. Recent rodent social defeat models demonstrate similar individual differences in response to social stress. In particular, mice subjected to chronic social defeat have been characterized as being either “susceptible” or “resilient” by the level of social interaction following social defeat. Susceptibility is associated with lasting social avoidance as well as increased anxiety-like behavior, and depressive-like symptoms. Resilient animals, however, do not show social avoidance or increased depressive-like symptoms, but retain increased anxiety-like behavior. Thus, it is unclear what “resilience” as measured by social interaction represents in terms of an overall behavioral and physiological phenotype. Here, we use an acute social defeat procedure, which produces distinct behavioral phenotypes in social interaction with no apparent changes in anxiety-like behavior. Susceptible mice display lasting social avoidance, whereas resilient mice display normal social interaction. Susceptible mice also displayed deficits in fear extinction retention but had normal within-session extinction. Paradoxically, resilience was associated with enhanced fear expression, and severe deficits in fear extinction and extinction retention beyond that observed in susceptible mice. These effects in resilient mice were only apparent after the experience of social stress and were not due to impaired behavioral flexibility. These data suggest that mechanisms controlling resilience to acute social defeat as characterized by social interaction leave animals vulnerable to maladaptive fear behavior.