Fear Conditioning, Freezing

views updated

Fear Conditioning, Freezing

When threatened, some animals simply freeze. This defensive behavior has been studied in greatest detail in rats. Freezing behavior has been one index of fear in Pavlovian conditioning experiments investigating fear learning and memory. This article describes the adaptive value of freezing, its use in Pavlovian fear conditioning, and its neural substrates.

Freezing Behavior

Animals have developed numerous behaviors that help them avoid threats and danger. For example, when a deer mouse encounters a predator such as a snake, it may attempt to run away. Or it might try to fend off the predator by biting or fighting with it. When threatened by predators or dominant conspecifics, rats, mice, squirrels, and other prey species also freeze, often in a crouch, utterly devoid of movement except for breathing. Freezing is not passive, however; it is a coordinated, protective defense against danger.

Why should a rat freeze and not run away when threatened? The utility of freezing can be understood by observing the response of a cat to a ball suspended on a piece of string. The cat will likely ignore the ball if it is stationary. In contrast, if you wiggle the string, the cat becomes vigilant and vigorously pounces or paws the ball. This pouncing is predatory hunting behavior. The moving ball elicits this behavior for two reasons: the cat's visual system is very good at detecting movement, and movement triggers predatory attack. Freezing evolved to counter the predator's sensitivity to movement. Prey species freeze because many predators see stationary objects poorly, and they are less likely to detect immobile prey. Hence, in the vicinity of a predator, a still rat is a safer rat.

Animals typically spend more time freezing as the perceived danger increases. Therefore, freezing is a useful index of fear. For example, a rat will spend more time freezing in a chamber in which it received five aversive electric shocks than in a chamber in which it received one.

Freezing is best measured with time sampling. In this procedure, at regular intervals of a few seconds, a trained observer makes a judgment as to whether the animal is freezing or not at a given instant. The percentage of instantaneous samples scored as freezing provides a probability estimate of the behavior that is appropriate for statistical analysis. This sampling procedure is accurate, reliable, and reproducible.

Pavlovian Fear Conditioning

Fear is rapidly learned and measured in the laboratory with a procedure called Pavlovian fear conditioning. This method has become a standard means of exploring the behavioral processes and neural mechanisms of learning. In a typical Pavlovian fear-conditioning procedure, a rat is placed in a chamber where it is presented with a tone that is followed by a brief but mildly aversive foot shock. Later, during a test session, the rat is reexposed to either the chamber or the tone. During this reexposure the rat will display fear. With this preparation, the tone and the chamber serve as conditional stimuli (CS). They were originally neutral signals that did not predict danger, but after they were paired with an unconditional stimulus (US)—the foot shock—the animal responds fearfully to the CS. Such a response to the CS is called a conditional response (CR); freezing is an example of a CR. The shock-paired stimuli trigger CRs, which are measures of learning in Pavlovian experiments.

Brain Circuit of Fear Conditioning and Freezing

Scientists use several experimental techniques to characterize brain circuits. Lesion studies seek to damage a brain structure with surgical techniques. In this procedure a scientist may inject a neurotoxin into a specific target region to kill the cells there. After the animal recovers from surgery, it is tested to determine what effect the lesion has on its behavior. Infusion studies seek to temporarily alter function in a target brain structure through the injection of chemicals through surgically implanted injectors mounted to the animal's skull. Drugs can be infused before, during, or after a behavioral treatment. This technique has the advantage of being temporary and reversible. Electrophysiological studies seek to measure the electrical properties of targeted neurons associated with specific behaviors such as freezing.

Rat, humans, and other mammals share fundamentally similar brain circuits that underlie fear. Indeed, behavioral neuroscientists have described a set of interconnected brain regions that constitute a "fear circuit" that is similar in humans, rats, mice, rabbits, and monkeys. In all of these species a structure called the amygdala is a prominent component of the fear circuit

The amygdala is composed of a set of interconnected clusters of neurons called nuclei that lie in the interior portion of the temporal lobe. The Pavlovian fear-conditioning paradigm has shown that three nuclei within the amygdala make major contributions to fear behavior: the lateral (LA), basal (BA), and central nuclei (CEA). The LA and BA nuclei make up the frontotemporal amygdala (FTA). This part of the amygdala communicates most closely with the frontal and temporal lobes of the brain, and it is important for fear learning.

Moreover, critical components of the memory established during fear conditioning are located in the FTA. First, the FTA is connected to auditory, visual, olfactory, and brain regions that govern pain sensation. Thus, sensory information of the CS and pain information of the US converge in the FTA. Second, Pavlovian fear conditioning enhances the response of cells in the FTA that respond to tone CSs. Third, damage to the FTA produces a pronounced and often total loss of many Pavlovian fear responses such as freezing. Fourth, chemical inactivation of this structure disrupts fear learning. Thus, the FTA is critical for learning fear, and it may be the locus of the establishment of fear memory.

The CEA may be the output of the amygdala. It is closely tied with the striatum and specializes in modulating motor outflow. The FTA projects to the CEA, which in turn projects to a variety of structure that include the periaqueductal gray (PAG), the reticular formation, and the lateral hypothalamus. Damage to the CEA disrupts the expression of a wide range of defensive behaviors, including freezing.

The periaqueductal gray, or PAG, is highly interconnected with the CEA. This region coordinates defensive behaviors. Chemical or electrical stimulation of the dorsal-lateral PAG (dlPAG) triggers bursts of forward movements resembling flight. Damaging this region disrupts flightlike behavior. Consequently, the dlPAG seems to coordinate defensive reactions like flight. In contrast, chemical or electrical stimulation of the ventral PAG (vPAG) triggers freezing, and damage to this structure disrupts it. Interconnections between these areas may allow the animal to rapidly shift between freezing and flight.

The hippocampus is also plays a role in the expression of some types of fear behavior. When a rat is trained with a tone-shock pairing inside a chamber, the animals will later display fear of both the tone and chamber—the context. However, if the animal's dorsal hippocampus is damaged after training, it will freeze in response to the tone but not the context. Presumably the hippocampus plays a role in recognition of the chamber. An important aspect of context fear is that the effect of the hippocampal damage grades with time. That is, if the hippocampus is damaged one day after the training session, the amount of contextual fear disruption is large; if the damage occurs two months after the training session, the amount of contextual fear disruption is small. This property of contextual fear is similar to the retrograde amnesia seen in human patients who have hippocampal damage: these patients cannot recall events immediately before their brain damage, but they can recall events from years before.


Freezing, a defensive response common in rodents and other prey animals, has been employed as an index of fear in Pavlovian conditioning experiments that investigate fear learning and memory. The amygdala plays a role in fear responses in all mammalian species; contextual fear shares properties with human declarative memory. Thus, freezing behavior and Pavlovian fear conditioning provide a useful animal model of learning and memory.



Edmunds, M. (1974). Defence in animals: A survey of antipredator defences. Burnt Mill, England: Longman.

Fanselow, M. S. (1994). Neural organization of the defensive behavior system responsible for fear. Psychonomic Bulletin & Review 1, 429-438.

—— (2000). Contextual fear, gestalt memories, and the hippocampus. Behavioural Brain Research 110, 73-81.

Fanselow, M. S., and LeDoux, J. E. (1999). Why we think plasticity underlying Pavlovian fear conditioning occurs in the basolateral amygdala. Neuron 23, 229-232.

Fendt, M., and Fanselow, M. S. (1999). The neuroanatomical and neurochemical basis of conditioned fear. Neuroscience and Biobehavioral Reviews 23, 743-760.

Squire, L. R., Clark, R. E., and Knowlton, B. J. (2001). Retrograde amnesia. Hippocampus 11, 50-55.

Swanson, L. W., and Petrovich, G. D. (1998). What is the amygdala? Trends in Neurosciences 21, 323-331.

Bill P.Godsil

Michael S.Fanselow