Animal Models Of Anxiety

All preclinical animal models of anxiety involve exposing animals (usually rats or mice) to environmental stimuli that disrupt their normal pattern of behaviour (Table 19.2). Obviously, it can never be confirmed that animals are actually experiencing the equivalent of human anxiety and so the validity of all preclinical models rests largely on confirming that the change in behaviour is prevented by drugs that have established anti-anxiety effects in humans.

An influential theory proposed by Gray (1987) suggests that environmental stimuli that induce anxiety in both rats and humans fall into three major groups, all of which present 'threats' to the individual. One of these is 'novelty' in which the subject's innate

Table 19.2 Environmental stimuli that induce changes in behaviour which are prevented by anti-anxiety drugs

'Ethological models' Elevated plus-maze Social interaction test Light/dark shuttle-box

Isolation-induced ultrasonic vocalisation (rodent pups) Models dependent on conditioned cues

Fear-potentiated startle reflex (conditioned fear) Four-plate test (conditioned fear) Geller-Seifter test ('signals' of conflict) Vogel conflict test ('signals' of conflict) Frustrative non-reward ('signals' of non-reward)

tendency to explore ('approach') novel stimuli is opposed by a tendency to avoid them; it is the conflict between approach and avoidance that gives rise to anxiety. The two other forms of anxiety-inducing ('anxiogenic') stimuli are those that are normally regarded as neutral but, as a result of innate factors (genetic programming) or the subject's previous experiences (associative learning), are interpreted as a signal for a stimulus that the subject would normally avoid. The signal can either warn that behaviour which is reinforced by reward will also be punished (e.g. by a footshock to rats; 'conflict') or warn that the reward will not materialise ('non-reward').

Testing the effects of drugs on animals' behavioural response to novel environmental stimuli offers the major advantage that it relies on evaluating changes in their innate behaviour (so-called 'ethological' models). Contrasting with this, testing the effects of drugs on animals' behavioural response to an environmental 'signal' requires extensive prior training (see below). The important point about this approach is that it is the 'signal' or 'threat' of the aversive event that triggers anxiety, rather than the aversive event itself. By analogy, environmental conditions that are associated with a threat of attack will provoke anxiety in humans whereas an actual attack triggers an (appropriate) acute stress response that recruits the 'fight or flight' (stress) reaction. In the following sections, specific behavioural models used to study anxiety and the effects of anti-anxiety drugs are described.

EVALUATING DRUG EFFECTS ON INNATE BEHAVIOUR (ETHOLOGICAL MODELS)

Most of these models evaluate the effects of drugs on the behaviour of animals when they are exposed to a novel environment. Novelty normally reduces animals' exploratory activity but established anti-anxiety drugs consistently increase exploration of, and approaches to, the novel stimulus and reduce the neophobic ('avoidance') reaction. There are several examples of tests based on this principle (Table 19.2) but two that are widely used are the 'plus-maze' and the 'social interaction' tests.

The plus-maze

This consists of a raised platform with four narrow arms, two of which have walls ('closed arms') and two which do not ('open arms') (Fig. 19.1). When placed on the

Figure 19.1 The elevated plus-maze. {Top) The apparatus is arranged with two open arms, two closed arms and a central zone, raised above the ground. Animals are placed in the central zone (usually facing an open arm) and their movements scored for: number of entries to the open and closed arms and the percentage time spent in the open arms. (Bottom) Chronic administration (5 days) of the anti-anxiety drug, chlordiazepoxide, increases the percentage time spent on the open arms to approximately 50% of the total. (Figure kindly provided by S. E. File)

Figure 19.1 The elevated plus-maze. {Top) The apparatus is arranged with two open arms, two closed arms and a central zone, raised above the ground. Animals are placed in the central zone (usually facing an open arm) and their movements scored for: number of entries to the open and closed arms and the percentage time spent in the open arms. (Bottom) Chronic administration (5 days) of the anti-anxiety drug, chlordiazepoxide, increases the percentage time spent on the open arms to approximately 50% of the total. (Figure kindly provided by S. E. File)

apparatus for the first time, animals explore all zones of the maze but spend most time (approximately 75%) in, and make most entries to, the closed arms. Pretreatment with an anti-anxiety drug increases exploration of the open arms so that approximately equal times are spent on the open and closed arms of the maze. When interpreting results from this test, it is important to establish that any drug effects are independent of non-specific effects on the animals' overall locomotor activity (i.e. their ability to make appropriate movements), particularly since many anti-anxiety drugs are highly sedative when given acutely. Detailed insight into some of the many assumptions and refinements of the use of the plus-maze is to be found in Rodgers and Dalvi (1997).

Social interaction test

In this test, it is the interaction (sniffing, grooming, etc.) between two rats in a test arena that is scored. Social interaction is dependent on the familiarity of the animals with the test arena (social interaction is reduced in an unfamiliar arena) and the intensity of illumination (social interaction is reduced in bright light). The reduction in social interaction under aversive conditions (unfamiliar arena and bright light) is prevented by pretreatment with anti-anxiety drugs (File and Hyde 1979). However, it is again important to establish that any drug effects are directed specifically at the behavioural response to the test environment, rather than overall locomotor activity.

MODELS THAT REQUIRE CONDITIONING

There are several models that depend on monitoring changes in animals' behaviour when they are exposed to conditioned threatening cues. One of these, the fear-potentiated startle reflex, rests on the development of an exaggerated startle on presentation of the conditioned cue. Although this response is prevented by anti-anxiety drugs, there is considerable debate over whether 'fear' is the same as 'anxiety'.

A model that fits better with Gray's criteria is the Geller-Seifter ('conflict') test. This is named after the two scientists who developed it and is still often used to screen putative anti-anxiety drugs (Geller, Kulak and Seifter 1962). Briefly, animals are trained to associate the pressing of a lever with a food reward ('operant' or 'instrumental conditioning'). After reaching a stable response on the lever, the rats are then trained to realise that when a (normally) neutral stimulus is presented, such as a buzzer or a light, they will experience a mild footshock, as well as receive the reward, when they press on the lever. This invokes a classic approach/avoidance conflict and animals invariably respond on the lever less frequently when the conditioned cue is presented (the 'punished phase'). An important distinction is that their response on the lever in the absence of the signal (the 'unpunished' phase) is unaffected. Anti-anxiety drugs abolish the inhibition of responding during the punished phase but do not affect unpunished responding (Fig. 19.2). Obviously, a drug that increases punished responding could be increasing animals' overall activity (as with amphetamine) but this can be excluded if it has no effect on lever responses during the unpunished phase. A drug-induced reduction in the discomfort caused by the footshock (as is achieved with analgesics) or amnesia (i.e. under the influence of the test drug the animal forgets that the cue warns of a footshock) must be ruled out also.

There are many variations of this model, a commonly used example being the Vogel licking (conflict) test. This evaluates the effects of drugs on the punished phase of drinking from a water spout (Vogel et al. 1980) which has the advantage that the animals do not have to be trained to initiate the behavioural response (drinking). However, the increase in baseline fluid intake induced by some anti-anxiety drugs, in the absence of any anxiogenic stimuli, can be a confounding factor.

Was this article helpful?

0 0

Post a comment