Context and strain-dependent behavioral response to stress

The open field test is a complex behavioral paradigm, whose various component measures have been widely used to measure emotionality [1], exploration [2], general activity or locomotion [3], fear [3, 4], and anxiety [5, 6] in rodents. The elevated plus maze has also been widely used in rodents as a test of fear, anxiety [7], and more recently, risk assessment in mice [8, 9] and rats [10, 11]. On a pharmacological level, the OFT and EPM have both been validated as tests of anxiety [7, 12].

Despite data supporting that OFT and EPM share areas of overlap in the behavioral traits they seek to measure and define, there are discrete but important differences between these two tests of anxiety. For instance, although activity in the OFT and open arm entries in the EPM share a genetic locus on chromosome 1, these same behaviors dissociate at the genetic locus found on chromosome 12 in mice [13]. Similarly, in rats, open arm entries in the EPM share a locus with rearing and activity in the OFT on chromosome 5, but not at other loci [14].

Factor analytic studies further illustrate that these tests measure different aspects of anxious behavior [15, 16]. For example, OFT measures of rearing, outer line crossings, and inner line crossings all load unto a factor called general activity, while EPM measures of time spent in, and entries into open arms load onto an independent factor called anxiety [3]. In another study, only activity measures of OFT and EPM loaded into the same factor in a large F2 intercross of Roman rat strains [17].

Both the OFT and the EPM are inherently stressful tests, but to a different degree. If differences in the tests' stressfulness explain the non-overlapping measures in the OFT and EPM, repetition of these tests could eliminate these behavioral differences, and prior stress could accentuate them. Behavioral stress-responses are known to be altered by prior stress exposure depending on the nature of the prior stress, the stressfulness of the behavioral test and the individual variation in stress-responsiveness [18, 19]. In this study, we compared the behavior of two inbred strains of rats, the Fisher 344 (F344) that is thought to be anxious and acutely hyper-reactive to stress [20‐22] and the Wistar Kyoto (WKY) rats purported to be an endogenous model of chronic stress state and depression [23‐30] in the OFT and the EPM.

This study posed the question whether strain differences in stress-reactivity lead to differential behavioral responses in two different tests of anxiety. The major findings of this study affirm this prediction. Restraint stress prior to the behavioral tests seems to have unmasked the differences: in the EPM, prior stress decreased anxiety-like behaviors only in F344s, while in the OFT, prior stress led to decreased activity in both strains. This decrease in activity was more pronounced in WKYs. These observations confirm the active coping style previously reported in F344s [31] and the passive coping style previously reported in WKYs [30‐32]. Repeated testing resulted in an increase in anxiety in both tests and for both strains. Measures of activity decreased in response to repeated testing in both tests, but only in OFT in response to stress.

Repetition of the behavioral tests resulted in increased anxiety-related behavior in both tests and both strains of animals. Similarly, previous reports of repeated EPM testing found a reduction in open arm exploration [33‐36], while others found no change from baseline [7, 37, 38]. Regarding what exactly occurs after repeated testing, no clear consensus is formed; whether familiarity causes a habituation or sensitization to fear via context-dependent learning, as it has been suggested previously [8, 39, 40], or whether repeated testing decreases the animal's novelty seeking behavior.

The differences between the F344 and WKY inbred strains in their anxiety-related behavioral response to stress are very prominent. Decreased anxiety is found in F344s by an increase in open arm exploration in the EPM. The unique profile of acute hyper-reactivity to stress seen in the F344 could be responsible for the stimulatory and anxiolytic effect of stress in this inbred strain. In contrast, Wistar Kyoto rats, with their purportedly chronic stress states responded minimally to prior acute stress in the EPM, but showed greater freezing than F344 rats in response to stress in the OFT. Thus, prior restraint stress in two strains previously shown to differ in stress reactivity, produced different behaviors in two different tests (OFT and EPM) long thought to measure some of the same aspects of behavior.

The paradoxical anxiolytic effects of stress on the F344s in the EPM, is confirmed by their grooming behavior in response to stress. Grooming is thought to be an indicator of stress perception and reactivity [41], and decreased grooming of F344s in the EPM post-stress, confirm their decreased stress perception or reactivity after stress. In contrast, grooming is dramatically increased in the second day tests in the stress group, suggesting that repeated testing induces sensitization to the test environment leading to learned avoidance as suggested before [39].

The reason for the differential effect of a prior stressor on F344s and WKYs in the EPM highlights their differences in reactivity to stress, mediated by the underlying genetic makeup of these two inbred strains. When stress is first applied, considerable excitation is observed, but if the stressor persists without successful coping, a longer period of behavioral depression follows with the animal behaving in a very passive fashion [42‐45]. It has been suggested that the WKY progresses quickly through the activation phase of stress to the second phase of behavioral inhibition and passivity [23]. It seems likely, therefore, that our finding – that the F344 responds to stress in the EPM by increasing exploration of the open arms – reflects the propensity for the F344 to remain in the activating phase of stress for a longer time than the WKY, which, shortly after being faced with a stressor, rapidly enters the phase of passive coping. Likewise, our finding of increased open arm exploration after stress could reflect an increased sensitivity of the F344 to the activating effects of stress, manifesting itself as stress-induced increases in risk taking behaviors.

The differences in the behavioral responses recognized by the OFT and EPM may also be inherent in the design of these two tests. The OFT is aversive mainly due to factors of novelty whereas the EPM is aversive due to novelty as well as the height of the maze. In addition, the plus maze seems to offer more choices to animals when compared to the OFT, with the center of the plus maze being used as a crossroads or choice point, from where animals initially engage in high levels of risk assessment (see Rodgers [46], for review). Thus, if an animal is particularly indecisive or ambivalent about which arm to enter in the process of exploring the maze, that animal will likely spend more time in the center portion of the plus maze. In our study, the WKYs spend significantly more time in the center of the plus maze while spending less time in the closed arms of the maze than the F344s. We propose that this pattern of behavior in the WKYs can be viewed as ambivalence or indecision. Pare [29], in comparing WKYs to F344s in a modified one-way avoidance procedure, also found that the WKYs were more ambivalent than the F344s, and suggested that this was due to behavioral inhibition elicited by stressors used during the task. Since ambivalence or the inability to make decisions is frequently observed in clinical depression, our finding that the WKYs show increased ambivalence as measured by time spent in center of the plus maze further adds to the large body of work suggesting that the WKYs are in a state of chronic stress and reflect many characteristics of an animal model of depression [24, 26, 29, 30, 47, 48].

The present study shows that repeated testing increases anxiety-related measures in both strains regardless of prior stress. However, prior restraint stress exaggerates the differences between OFT and EPM, with stress resulting in a general decrease in behavioral responses in the OFT in both strains, contrasted with a strain-dependent response to stress in the EPM. Restraint stress results in decreased anxiety-like behavior (increased exploration of open arms) in the F344, with no change in the WKY, drawing attention to the importance of genetic differences of inbred strains in determining behavioral responsiveness to stress. Our original assumption that WKYs model at least some aspects of depressed behavior is supported by our finding of greater ambivalence, as measured by increased time spent in center of the plus maze, in the WKY as compared to the F344. The results of this study suggest that differences in the stressfulness of these tests may contribute to the distinction between them; the OFT is a better measure of passive coping while the EPM is a more sensitive measure of active coping in response to stress.