Behavioral Neuroscience 2000, Vol. 114, No. 5, 912-922
Copyright 2000 by the American Psychological Association, Inc. 0735-7044/00/$5,00 DOI: 10.1037//0735-7044.114.5.912
Predator Odor as an Unconditioned Fear Stimulus in Rats: Elicitation of Freezing by Trimethylthiazoline, a Component of Fox Feces Karin J. Wallace and Jeffrey B. Rosen University of Delaware Four experiments tested whether an odor from a rat predator can unconditionally elicit a fear response in rats. In a large chamber, rats displayed fear-related behaviors to trimethylthiazoline (TMT, a volatile compound isolated from fox feces), including avoidance and immobility, while showing less exploratory behavior. In a smaller chamber, TMT induced a species-typical fear response, freezing, whereas other odors did not. In addition, TMT systematically elicited more freezing as the amount of TMT increased. Moreover, there was no within-sessions or between-sessions habituation of freezing to TMT, nor did TMT promote contextual conditioning. The results indicate that the predator odor, TMT, can induce a fear-related behavioral response in rats that is controllable and quantifiable, suggesting that TMT-induced freezing may be a useful paradigm for a neurobehavioral system analysis of ecologically relevant, unconditioned fear.
do not emerge at all (R. J. Blanchard, Blanchard, & Hori, 1989). Instead, rats display behaviors involved in risk assessment or approach-avoidance conflict, such as orientation to the threatening object with a flattened back and stretched attention postures, sniffing from a distance, approaching the odor source for only short periods of time (R. J. Blanchard et al., 1989; Kaesermann, 1986), and spending more time in a sheltered environment (Zangrossi & File, 1992). These milder forms of defensive behavior may be due to the intensity of the stimulus, in which cat odor is less intense than a whole cat, and have been hypothesized to model human anxiety as opposed to fear (D. C. Blanchard & Blanchard, 1988; R. J. Blanchard et al., 1998). However, complicating this analysis is that the amount of odor is unknown and not controlled and that the specific compounds in the fur that elicit these behaviors are also unknown. The latter problem can confound interpretation, as seen in a comparison of mice responding to odor of cats on a vegetarian versus a carnivorous diet, in which mice displayed more avoidance of feces from meat-eating cats (Berton, Vogel, & Belzung, 1998). There are other predator odors, including specific components from fox feces and weasel secretions, that elicit avoidance responses similar to those seen with presentation of a cat or cat odors. One predator odor that has received attention is trimethylthiazoline (TMT), a component of fox feces. TMT is one of several sulfur-containing odors isolated from fox feces but is unique in its ability to induce stress and emotional responses from wild and laboratory-bred rats (Vernet-Maury, Polak, & Demael, 1984). Wild rats (rattus rattus) placed in a terrarium will avoid TMT (Vernet-Maury, Constant, & Chanel, 1992), and wild rats naive to fox avoid food in the presence of TMT as they do when in the presence of the scent from a known predator (mongoose; Burwash, Tobin, Woolhouse, & Sullivan, 1998). Laboratory rats are more vigilant in the presence of TMT than in the presence of nonpredator animal odors (e.g., from lion; Cattarelli & Chanel, 1979). Furthermore, rats exhibit analgesia in the presence of TMT (Hotsenpiller & Williams, 1997) similar to fear-associated analgesia induced by stressed rats or footshock (Fanselow, 1985; Fanselow
Fear is a functional behavioral response to a dangerous event (Rosen & Schulkin, 1998). Research on learned fear in rats, typically produced by Pavlovian conditioning (e.g., pairing neutral stimuli with aversive footshock), has been instrumental in advancing understanding of the neurobiology of fear (Davis, 1992; LeDoux, 1996). In contrast, unconditioned fear, such as fear of a predator, has not been studied in as much depth. Antipredator defense models provide considerable analytical advantages over footshock-based models of fear and defense for studying the determinants of defensive behavior and the neural and neurochemical systems that control them (D. C. Blanchard, 1997). Footshocks are not typically encountered by rats in natural environments but may represent a class of rare stimuli that are painful. When pain does occur, the painful stimuli may be associated with predator attack rather than with other more common features of the environment that signal a predator (D. C. Blanchard, 1997). In the wild, when a rat is attacked by a predator, it is usually fatal and too late for defensive maneuvers (Bolles, 1970). On the other hand, presentation of predators and stimuli associated with predators are "natural" threats and can serve as painless, controllable stimuli to study the behavior and biology of fear and defense (D. C. Blanchard, 1997). The odor from the fur of cats has been used as an unconditioned, ecologically relevant olfactory fear stimulus for rats (e.g., R. J. Blanchard & Blanchard, 1989; Zangrossi & File, 1992). However, when rats are exposed to this odor (a cloth rubbed on a cat's fur), typical fear behaviors (i.e., freezing) are displayed only briefly or
Karin J. Wallace and Jeffrey B. Rosen, Department of Psychology, University of Delaware. This study was supported by National Science Foundation Grant IBN9904623. Correspondence concerning this article should be addressed to Jeffrey B. Rosen, Depat~nent of Psychology, 220 Wolf Hall, University of Delaware, Newark, Delaware 19716. Electronic mail may be sent to
[email protected]. 912
UNCONDITIONED FREEZING TO A FOX ODOR & Tighe, 1988). These data imply that T M T may be an effective natural fear stimulus for rats. However, as with cat odors (R. J. Blanchard et al., 1989), fear responses typically measured in tests o f conditioned fear (e.g., freezing) are not easily evoked by T M T in the contexts in w h i c h animals have been tested (Vernet-Maury et al., 1984). It is possible that because the odor is only a signal for the presence o f a predator, it does not elicit sufficient fear to induce freezing. On the other hand, the testing conditions used to date may not have been optimal for inducing freezing with predator odors. In long runways or large areas in which T M T has been tested, animals were able to stay away from the odor source and effectively create sufficient distance from the odor to reduce its intensity to nonthreatening levels. In the current series o f experiments, the hypothesis that T M T is an unconditioned fear stimulus for rats was examined. Freezing was the prototypical fear response measured in most o f the experiments. Several aspects o f T M T - i n d u c e d fear were examined. W e tested w h e t h e r T M T elicited freezing and other behavioral responses in a m e d i u m - s i z e d c h a m b e r where the rat could m o v e away from the source o f T M T and in a confined space where the rat could not escape the odor source. In addition, to test whether the intensity o f T M T affected the amount o f freezing, we generated a response curve for increasing amounts o f TMT. Furthermore, w h e t h e r the unconditioned freezing response habituated to T M T after repeated exposure and whether it could be used as an unconditioned stimulus to support contextual fear conditioning were also tested. Experiment
1
In the first study we observed behavior in response to T M T and other odors in a typical laboratory shoebox cage w h e r e the rat was able to roam. W e observed and measured avoidance o f the odor and odor-induced immobility as well as nondefensive behaviors such as rearing and grooming. In addition, the risk-assessment behaviors o f exploring and s t r e t c h - s n i f f at the odor source were measured.
Me~od Subjects. Nine male Sprague-Dawley rats (Harlan, Indianapolis, IN) weighing 225-250 g at the beginning of the experiment were used. The subjects were housed individually in hanging wire cages. They were maintained on a 12-hour light-dark cycle, and food and water were continuously available. All experimental procedures were approved by the University of Delaware Institutional Animal Care and Use Committee. Apparatus. Testing was conducted in a standard, clear Plexiglas shoebox cage measuring 35 × 24 × 21 cm with woodchip bedding on the floor of the cage. A Plexiglas lid with holes to allow air circulation and tape marking three equally sized areas of 11.3 × 24 × 21 cm was placed on top of the cage during the experiment (Area 1 contained odor, Area 2 was in the middle, and Area 3 was farthest from the odor). The odor was presented by pipetting it onto a 6 × 6 cm piece of Kimwipe taped to a 14.0 × 10.5 cm piece of Plexiglas that was propped up against the endwall. The cage was placed in a fumehood during experimentation so the volatile odors would not spread through the experimental room. A new cage was used for each rat so no odor except for the odor of the woodchips would be present in the chamber. Odor remaining in the fumehood was allowed to dissipate before the next subject was run. The experiment was videotaped for later analysis. The camera was located directly above the test chamber.
913
Procedure. All rats were handled for about 5 min a day in the testing room for approximately 1 week before testing. Subjects were transported for handling in the same type of cages as they were tested in, so the testing cage was not novel. On the day of testing each rat was placed in the chamber with a piece of Plexiglas containing no odor propped up against the endwall. The rat was videotaped for 5 min in this condition. The odorless Plexiglas was removed following the 5-min period and replaced by a piece of Plexiglas containing an odor for 20 min. Odors tested were TMT, a component of fox feces (molecular weight: 129.2; Phero Tech, Delta, British Columbia, Canada); butyric acid, an ester in butter with a rancid odor unpleasant to humans (molecular weight: 88.1; ICN Biomedicals, Aurora, Ohio); isoamyl acetate, odor of bananas and pears (molecular weight: 130.2; ICN Biomedicals); and the neutral odor of the test cage. Odors were not diluted and were used neat. Each subject was presented with 300 nmol of each odor in the same order: TMT (19.4 ~1), neutral odor, butyric acid (13.2 p~l), and isoamyl acetate (19.6 p,1). Avoidance, immobility, rearing, grooming, and stretch-sniff were measured for 5 min before the odor presentation and during the 20 min of odor presentation. An additional measure of exploring the odor source within 30 s of its being placed in the chamber was also scored. The behaviors were measured later by viewing the videotapes. The experimenter (Karin J, Wallace) rating the behaviors was aware of the odor presented in this and all subsequent experiments. The amount of time spent in each of the three areas was used as a measure of avoidance. More time spent in Area 3 was considered more avoidance. Immobility was considered a lack of movement and measured as amount of time spent immobile. Rearing was counted when the rat stood on its hind limbs and was not directly sniffing the odor source. Grooming was recorded as the amount of time spent cleaning. Stretch-sniff was counted when a rat had a stretched back with hind limbs splayed while sniffing in the direction of the odor; this was measured as number of times the animal stretch-sniffed. An additional measure of initial exploration of the odor source was scored if the rat entered Area 1 and sniffed at the Plexiglas within 30 s of its being placed into the chamber. Statistical analysis of avoidance involved a 4 (type of odor) × 3 (area of box) × 4 (5-min time bin) three-way analysis of variance (ANOVA). Significant effects were followed by individual contrasts. Analysis of immobility, grooming, and rearing involved 4 (type of odor) × 4 (5-min time bin) two-way within-subjects ANOVAs. Rats did not perform measurable levels of stretch-sniff after the first 5 min of odor presentation. Therefore, stretch-sniff in only the 5 min before the odor and in the first 5 min of odor presentation was analyzed. A one-way (type of odor) withinsubjects ANOVA was performed on these data. For all the above behaviors, a one-way (type of odor) ANOVA was performed to determine whether the preodor exposure was equivalent for each odor. As appropriate, significant results were followed by mean comparison contrasts. Exploration of the odor was analyzed by a Cochran Q test for related samples, followed by binomial tests. Statistical significance was set at p < .05 for all tests. Results In the 5 min before odor presentation, there was a difference in which area the rats spent time, F(2, 16) = 6.8, p < .007. Post hoc comparisons s h o w e d that rats spent more time in Area 1 (area containing the odor source) and Area 3 (area farthest from the odor source) than in Area 2 (middle area). There was a tendency for subjects to spend more time in Area 1 than Area 3, but this was not significant ( p = .07). This preference was likely due to the novel Plexiglas object in that area. In the analysis with odor present, there was no main effect o f time, F(18, 126) = 1.55, ns, indicating that the amount o f time spent in each area did not change throughout the 20-rain session. There was, however, a main effect o f the
914
WALLACE AND ROSEN acetate in the amount o f time spent in the three areas, F(4, 20) < 1, ns. With all three odors the majority o f time was spent in Area 3. Measuring immobility, we found no differences in the 5 min before presentation o f each odor, F(3, 24) = 1.6, ns. However, during the odor presentation there was more immobility to T M T than to the other odors (see Figure 1B). There was a main effect o f odor, F(3, 21) = 5.6, p < .006, that was due to less m o v e m e n t to T M T than to neutral odor, butyric acid, or isoamyl acetate. There was also a main effect o f time, F(3, 21) = 2.5, p < .01, that was due to significantly more immobility 1 0 - 2 0 min after the presentation o f the odors than within the first 5 min o f odor presentation. A n Odor × Time interaction, F(9, 63) = 2.5, p < .01, and post hoc
amount o f time spent in the three areas o f the chamber, F(2, 14) = 22.76, p < .0001, such that, overall, rats spent more time in Area 3. More importantly, there was a significant interaction effect o f Odor × Area, F(6, 42) = 6.74, p < .0001. This is shown in Figure 1 A . Time spent in Area 1 was significantly greater with the neutral odor than the other odors ( p < .001), whereas time spent in Area 3 was greater with the other odors than with the neutral odor ( p < .0001). While in Area 1, rats sniffed the odor source, and the subjects with the neutral odor manipulated the object containing the odor. W e conducted an A N O V A without the neutral odor to compare differences b e t w e e n the three odor groups. There were no differences b e t w e e n TMT, butyric acid, and isoamyl
[] N e u ~ [] Butyric Acid I I~ Isoamyl Acetate ] • TMT ]
A 2so T
I
B
[] Butyric Acid
160 + ~ .
~-~
.
[~
~ ~ 100 ~
- .~ -k
.
•
140 + ~ 120 -~
iiii: !
.'XX .....
[] Neutral
180
r
.r
~
4114
o
