Anxiety, fear, panic: An approach to assessing the defensive behavior system across the predatory imminence continuum

Comprehensive ethological analysis of fear expression in rats using DeepLabCut and SimBA machine learning model

Introduction

Defensive responses to threat-associated cues are commonly evaluated using conditioned freezing or suppression of operant responding. However, rats display a broad range of behaviors and shift their defensive behaviors based on immediacy of threats and context. This study aimed to systematically quantify the defensive behaviors that are triggered in response to threat-associated cues and assess whether they can accurately be identified using DeepLabCut in conjunction with SimBA.

Methods

We evaluated behavioral responses to fear using the auditory fear conditioning paradigm. Observable behaviors triggered by threat-associated cues were manually scored using Ethovision XT. Subsequently, we investigated the effects of diazepam (0, 0.3, or 1 mg/kg), administered intraperitoneally before fear memory testing, to assess its anxiolytic impact on these behaviors. We then developed a DeepLabCut + SimBA workflow for ethological analysis employing a series of machine learning models. The accuracy of behavior classifications generated by this pipeline was evaluated by comparing its output scores to the manually annotated scores.

Results

Our findings show that, besides conditioned suppression and freezing, rats exhibit heightened risk assessment behaviors, including sniffing, rearing, free-air whisking, and head scanning. We observed that diazepam dose-dependently mitigates these risk-assessment behaviors in both sexes, suggesting a good predictive validity of our readouts. With adequate amount of training data (approximately > 30,000 frames containing such behavior), DeepLabCut + SimBA workflow yields high accuracy with a reasonable transferability to classify well-represented behaviors in a different experimental condition. We also found that maintaining the same condition between training and evaluation data sets is recommended while developing DeepLabCut + SimBA workflow to achieve the highest accuracy.

Discussion

Our findings suggest that an ethological analysis can be used to assess fear learning. With the application of DeepLabCut and SimBA, this approach provides an alternative method to decode ongoing defensive behaviors in both male and female rats for further investigation of fear-related neurobiological underpinnings.

Pavlovian or associative sensitization and its biological significance

Pavlovian conditioning is typically distinguished from sensitization but a Pavlovian conditional stimulus (CS) also results in sensitization. A Pavlovian CS can sensitize responding to a probe stimulus that is related to the unconditional stimulus (US) or to the US itself. Pavlovian sensitization has been studied in the defensive, sexual, and feeding systems. In Pavlovian sensitization, the focus is not on a conditional response (CR) directly elicited by the CS but on the response mode that is activated by the CS. Activation of a response mode increases the probability of particular responses and also increases reactivity to various stimuli. Pavlovian sensitization reflects this increased stimulus reactivity. Pavlovian sensitization helps uncover successful learning in situations where a conventional CR does not occur. Pavlovian sensitization also encourages broadening our conceptions of Pavlovian conditioning to include changes in afferent processes. Implications for biological fitness and for basic and translational research are discussed.

Central amygdala CRF+ neurons promote heightened threat reactivity following early life adversity in mice

Failure to appropriately predict and titrate reactivity to threat is a core feature of fear and anxiety-related disorders and is common following early life adversity (ELA). A population of neurons in the lateral central amygdala (CeAL) expressing corticotropin releasing factor (CRF) have been proposed to be key in processing threat of different intensities to mediate active fear expression. Here, we use in vivo fiber photometry to show that ELA results in sex-specific changes in the activity of CeAL CRF+ neurons, yielding divergent mechanisms underlying the augmented startle in ELA mice, a translationally relevant behavior indicative of heightened threat reactivity and hypervigilance. Further, chemogenic inhibition of CeAL CRF+ neurons selectively diminishes startle and produces a long-lasting suppression of threat reactivity. These findings identify a mechanism for sex-differences in susceptibility for anxiety following ELA and have broad implications for understanding the neural circuitry that encodes and gates the behavioral expression of fear.

Acute stress yields a sex-dependent facilitation of signaled active avoidance in rats

Post-traumatic stress disorder (PTSD) is a debilitating disorder characterized by excessive fear, hypervigilance, and avoidance of thoughts, situations or reminders of the trauma. Among these symptoms, relatively little is known about the etiology of pathological avoidance. Here we sought to determine whether acute stress influences avoidant behavior in adult male and female rats. We used a stress procedure (unsignaled footshock) that is known to induce long-term sensitization of fear and potentiate aversive learning. Rats were submitted to the stress procedure and, one week later, underwent two-way signaled active avoidance conditioning (SAA). In this task, rats learn to prevent an aversive outcome (shock) by performing a shuttling response when exposed to a warning signal (tone). We found that acute stress significantly enhanced SAA acquisition rate in females, but not males. Female rats exhibited significantly greater avoidance responding on the first day of training relative to controls, reaching similar levels of performance by the second day. Males that underwent the stress procedure showed similar rates of acquisition to controls but exhibited resistance to extinction. This was manifest as both elevated avoidance and intertrial responding across extinction days relative to non-stressed controls, an effect that was not observed in females. In a second experiment, acute stress sensitized footshock unconditioned responses in males, not females. However, males and females exhibited similar levels of stress-enhanced fear learning (SEFL), which was expressed as sensitized freezing to a shock-paired context. Together, these results reveal that acute stress facilitates SAA performance in both male and female rats, though the nature of this effect is different in the two sexes. We did not observe sex differences in SEFL, suggesting that the stress-induced sex difference in performance was selective for instrumental avoidance. Future work will elucidate the neurobiological mechanisms underlying the differential effect of stress on instrumental avoidance in male and female rats.

Line- and sex-dependent effects of juvenile stress on contextual fear- and anxiety-related behavior in high- and low-alcohol-preferring mouse lines

Juvenile stress (JS) is a known risk factor for the development of alcohol use disorder (AUD) and post-traumatic stress disorder (PTSD), both of which are frequently co-morbid. Data suggest there may be common, genetically-influenced biological responses to stress that contribute to the development of both AUD and PTSD. The present study investigated the impact of JS on contextual fear learning and extinction, as well as corticosterone (CORT) responses before and after JS, before and after contextual fear conditioning (CFC), and after fear extinction in male and female high-alcohol-preferring (HAP2) and low-alcohol-preferring (LAP2) mouse lines. We also measured unconditioned anxiety-related behavior in the light-dark-transition test before CFC. HAP2 and LAP2 mice did not differ in fear acquisition, but HAP2 mice showed faster fear extinction compared to LAP2 mice. No effects of JS were seen in HAP2 mice, whereas in LAP2 mice, JS reduced fear acquisition in males and facilitated fear extinction in females. Females showed greater fear-related behavior relative to males, regardless of subgroup. HAP2 males demonstrated more anxiolytic-like responses than LAP2 males and LAP2 females demonstrated more anxiolytic-like responses than LAP2 males in the light-dark transition test. HAP2 and LAP2 mice did not differ in CORT during the juvenile stage; however, adult LAP2 mice showed greater CORT levels than HAP2 mice at baseline and after CFC and extinction testing. These findings build upon prior work in these unique mouse lines that differ in genetic propensity toward alcohol preference and provide new information regarding contextual fear learning and extinction mechanisms theorized to contribute to co-morbid AUD and PTSD.

Estimating foraging behavior in rodents using a modified paradigm measuring threat imminence dynamics

Animals need to respond to threats to avoid danger and approach rewards. In nature, these responses did not evolve alone but are always accompanied by motivational conflict. A semi-naturalistic threat imminence continuum model models the approach-avoidance conflict and is able to integrate multiple behaviors into a single paradigm. However, its comprehensive application is hampered by the lack of a detailed protocol and data about some fundamental factors including sex, age, and motivational level. Here, we modified a previously established paradigm measuring threat imminence continuum dynamics, involving modifications of training and testing protocols, and utilization of commercial materials combined with open science codes, making it easier to replicate. We demonstrate that foraging behavior is modulated by age, hunger level, and sex. This paradigm can be used to study foraging behaviors in animals in a more naturalistic manner with relevance to human approach-avoid conflicts and associated psychopathologies.

Pharmacological activation of the amygdala, but not single prolonged footshock-induced acute stress, interferes with cue-induced motivation toward food rewards in rats

In the face of threats, animals adapt their behaviors to cope with the situation. Under such circumstances, irrelevant behaviors are usually suppressed. In this study, we examined whether food-seeking motivation would decrease under activation of the amygdala, an important nucleus in the regulation of stress response in the central nervous system, or after a physical acute stress session. In Experiment 1, we pharmacologically activated the basolateral nucleus (BLA) or the central nucleus of the amygdala (CeA) before a cue-induced reinstatement test in rats. Our results showed that activation of the BLA or the CeA abolished cue-induced motivation toward food rewards, while locomotor activity and free food intake were not affected. In Experiments 2 and 3, we further assessed anxiety and despair levels, as well as cue-induced reinstatement, after a single prolonged footshock-induced acute stress in rats. Behaviorally, acute stress did not affect anxiety level, despair level, or cue-induced motivation toward food rewards. Physiologically, there was no difference in cellular activities of the amygdala immediately after acute stress. To conclude, our results suggested that pharmacological activation of the amygdala decreased cue-induced motivation toward food reward. However, physiological acute stress did not immediately interfere with the negative emotions, motivation, or amygdala activities of the animals.

Beyond Nature Versus Nurture: the Emergence of Emotion

Affective science is stuck in a version of the nature-versus-nurture debate, with theorists arguing whether emotions are evolved adaptations or psychological constructions. We do not see these as mutually exclusive options. Many adaptive behaviors that humans have evolved to be good at, such as walking, emerge during development - not according to a genetically dictated program, but through interactions between the affordances of the body, brain, and environment. We suggest emotions are the same. As developing humans acquire increasingly complex goals and learn optimal strategies for pursuing those goals, they are inevitably pulled to particular brain-body-behavior states that maximize outcomes and self-reinforce via positive feedback loops. We call these recurring, self-organized states emotions. Emotions display many of the hallmark features of self-organized attractor states, such as hysteresis (prior events influence the current state), degeneracy (many configurations of the underlying variables can produce the same global state), and stability. Because most bodily, neural, and environmental affordances are shared by all humans - we all have cardiovascular systems, cerebral cortices, and caregivers who raised us - similar emotion states emerge in all of us. This perspective helps reconcile ideas that, at first glance, seem contradictory, such as emotion universality and neural degeneracy.

Nature and nurture: comparing mouse behavior in classic versus revised anxiety-like and social behavioral assays in genetically or environmentally defined groups

Widely used rodent anxiety assays like the elevated plus maze (EPM) and the open field test (OFT) are often conflated with rodents' natural preference for dark over light environments or protected over open spaces. The EPM and OFT have been used for many decades, yet have also been criticized by generations of behavioral scientists. Several years ago, two revised anxiety assays were designed to improve upon the "classic" tests by excluding the possibility to avoid or escape aversive areas of each maze. The 3-D radial arm maze (3DR) and the 3-D open field test (3Doft) each consist of an open space connected to ambiguous paths toward uncertain escape. This introduces continual motivational conflict, thereby increasing external validity as an anxiety model. But despite this improvement, the revised assays have not caught on. One issue may be that studies to date have not directly compared classic and revised assays in the same animals. To remedy this, we contrasted behavior in a battery of assays (EPM, OFT, 3DR, 3Doft, and a sociability test) in mice defined either genetically by isogenic strain, or environmentally by postnatal experience. Findings indicate that the optimal assay to assess anxiety-like behavior may depend upon grouping variable (e.g. genetic versus environment). We argue that the 3DR may be the most ecologically valid of the anxiety assays tested, while the OFT and 3Doft provided the least useful information. Finally, exposure to multiple assays significantly affected sociability measures, raising concerns for designing and interpreting batteries of behavioral tests in mice.

Sub-chronic administration of fluoxetine does not alter prey-capture or predator avoidance behaviors in adult South African clawed frogs (Xenopus laevis)

Animals will halt foraging efforts and engage defensive behaviors in response to predator cues. Some researchers have proposed that the switch from appetitive to avoidance behavior resembles anxiety, but most work on this has been performed in a limited number of animal models, primarily zebrafish and rodents. We used adult South African clawed frogs (Xenopus laevis) to determine if the canonical anxiolytic fluoxetine alters predator-induced changes in appetitive and avoidance behavior in a laboratory-based trade-off task that mimics foraging/predator avoidance tradeoffs in the wild. We hypothesized that sub-chronic fluoxetine treatment (20 d) would not affect baseline behavior but would reverse predator-induced changes in food intake, appetitive and avoidance behavior, and the abundance of anxiety related gene transcripts in the optic tectum, a brain area central to ecological decision making in frogs. We found that fluoxetine significantly reduced baseline locomotion compared to vehicle-treated animals. Fluoxetine had no effect on appetitive and avoidance behaviors that were sensitive to predator cues in this assay and did not alter any of the anxiety-related transcripts in the tectum. We conclude that while peripheral sub-chronic administration of fluoxetine significantly reduces locomotion, it does not modify predator-induced changes in approach and avoidance behaviors in this assay. Our findings are not consistent with visual predator cues causing state anxiety in adult frogs.

Selective pressure on a saccharin intake phenotype and its correlates: a replication study

The Occidental High- and Low-Saccharin rats (respectively, HiS and LoS lines) were selectively bred for decades to examine mechanisms and correlates of a saccharin intake phenotype. Observed line differences ranged from taste and eating to drug self-administration and defensive behavior, paralleling human research on relationships between gustation, personality, and psychopathology. The original lines were terminated in 2019, and replicate lines (HiS-R and LoS-R) were selectively bred for 5 generations to test for reproducible, rapid selection for the phenotype and its correlates. The line differences chosen for replication included intake of tastants (saccharin, sugars, quinine-adulterated sucrose, sodium chloride, and ethanol) and foods (cheese, peas, Spam, and chocolate) and several noningestive behaviors (deprivation-induced hyperactivity, acoustic startle, and open field behavior). The HiS-R and LoS-R lines diverged on intake of saccharin, disaccharides, quinine-adulterated sucrose, sodium chloride, and complex foods, and open field behavior. Differences from the original lines also were observed. Reasons for and implications of the pattern of replication and lack thereof in 5 generations are discussed.

Negative valence systems: sustained threat and the predatory imminence continuum

This review describes the relationship between the National Institute of Mental Health (U.S.A.) Research Domain Criteria (RDoC) Negative Valence System related to responses to threat and the Predatory Imminence Continuum model of antipredator defensive behavior. While the original RDoC constructs of Potential Threat (anxiety) and Acute Threat (fear) fit well with the pre-encounter and post-encounter defense modes of the predatory imminence model, the Sustained Threat construct does not. Early research on the bed nuclei of the stria terminalis (BST) suggested that when fear responding needed to be sustained for a prolonged duration this region was important. However, follow-up studies indicated that the BST becomes critical not because responses needed to be sustained but rather when the stimuli triggering fear were more difficult to learn about, particularly when aversive stimuli were difficult to accurately predict. Instead, it is argued that the BST and the hippocampus act to expand the range of conditions that can trigger post-encounter defense (Acute Threat). It is further suggested that sustained threat refers to situations where the predatory imminence continuum becomes distorted causing defensive behavior to intrude into times when organisms should be engaging in other adaptive behaviors. Stress is seen as something that can cause a long-term disturbance of the continuum and this disturbance is a state of sustained threat.

The elegant complexity of fear in non-human animals

Activation of the fear system is adaptive, and protects individuals from impending harm; yet, exacerbation of the fear system is at the source of anxiety-related disorders. Here, we briefly review the 'why' and 'how' of fear, with an emphasis on models that encapsulate the elegant complexity of rodents' behavioral responding in the face of impending harm, and its relevance to developing treatment interventions.

Adaptation of Threat Responses Within the Negative Valence Framework

External threats are a major source of our experience of negatively valanced emotion. As a threat becomes closer and more real, our specific behavior patterns and our experiences of negative affect change in response to the perceived imminence of threat. Recognizing this, the National Institute of Mental Health's Research Domain Criteria (RDoC) Negative Valence system is largely based around different levels of threat imminence. This perspective describes the correspondence between the RDoC Negative Valence System and a particular neurobiological/neuroecological model of reactions to threat, the Predatory Imminence Continuum (PIC) Theory. Using the COVID-19 pandemic as an illustration, we describe both adaptive and maladaptive behavior patterns from this perspective to illustrate how behavior in response to a crisis may get shaped. We end with suggestions on how further consideration of the PIC suggests potential modifications of the negative valence systems RDoC.