**3.1 Acquisition of active avoidance and changes in startle reactivity**

There are several examples of shock-induced changes in various behavioral indexes of anxiety-like reactions outside of the shock-exposure context (Servatius et al., 1994; Servatius et al., 1995; Beck et al., 2002; Cordero et al., 2003; Beck & Servatius, 2005; Manion et al., 2007; Daviu et al., 2010; Manion et al., 2010), which may lead one to assume that stressors which cause pain have a particularly significant role in causing context-independent changes in general arousal. However, in the case of acquiring behavior that is conducive to active avoidance of shock, the acute role of shock exposure in early acquisition can be contrasted with the expression of stimulus control during asymptotic performance levels. This is an important distinction in any avoidance-based model of PTSD, since the clinical condition does not necessarily involve an acute increase in arousal (as defined by startle reactivity), but the development of avoidance does parallel the general worsening of symptoms (O'Donnell et al., 2007). The implication of this correlation is that other symptoms, such as hyperarousal, may come as a result of increasing stimulus control, as active avoidance coping strategies strengthen.

Acquisition of Active Avoidance Behavior as a Precursor

to Changes in General Arousal in an Animal Model of PTSD 81

Fig. 5. Exposure to avoidance learning was associated with an increase in startle sensitivity in both strains of male rats (main effect Avoidance, F (1, 28) = 4.2, p < .05. Otherwise, significant differences across the three stimulus intensities used (main effect Intensity F (2, 56) = 595.5, p < .001) were coupled with a marginally significant difference in the percentage of startles elicited across the 4 weeks of acquisition (Avoidance x Week interaction F (3, 84) = 2.4, p < .07. By the fourth week of avoidance acquisition, differences in startle sensitivity between the avoidance-trained group and the homecage control group were greatly

reduced. Although not statistically significant, signs of increased startle magnitude began to

2010; Beck et al., 2011), suggesting that they may exhibit greater changes in general arousal, as indexed by startle reactivity measures, with the experience of shock and/or the acquisition of the avoidant behavior. Using the same procedures as described for the male rats, female rats demonstrated similar patterns of initial startle reactivity: greater startle magnitudes in WKY rats but equivalent startle sensitivity across strain. During avoidance

become apparent in SD rats by the last week of acquisition.

Male SD and WKY rats were trained to acquire a lever-press, active avoidance behavior (or served as non-trained homecage controls), and were tested weekly to assess changes in startle sensitivity and responsivity. In concert with prior studies, male WKY rats demonstrated greater baseline startle magnitudes and equivalent startle sensitivity. Subsequently, acquisition of avoidance appeared to be equivalent between the strains. In addition, WKY rats reached greater asymptotic avoidance performance, as seen in previous studies (see Figure 4).

Fig. 4. Lever-press behavior to avoid intermittent footshock was reliably acquired in both strains of male rats (main effect Session, F (9, 135) = 25.0, p < .001 and Session x Trial interaction, F (171, 2394) = 1.4, p < .001), but WKY rats exhibited more avoidance responses per session in the later sessions of the acquisition phase. During extinction, differences between strains were not apparent across sessions, but they were significant within sessions (Strain x Trial interaction, F (19, 266) = 3.3, p < .001.

Increases in startle sensitivity and responsivity are expected early in acquisition if shock exposure causes changes in vigilance and arousal, but, if learning and performing an avoidance behavior causes anxiety, startle measures should be elevated during the later weeks of acquisition. Thus, as shown in Figure 5, there was a strain-independent elevation in startle sensitivity displayed by those being trained in avoidance behavior. This difference is evident from the beginning of acquisition, when animals are receiving the most of shocks, and dissipates by the end of acquisition. Conversely, startle responsivity, as demonstrated by relative increases in startle magnitudes above baseline, is largely unchanged early in acquisition, when enhanced startle sensitivity is greatest. Startle responsivity in avoidancetrained animals increased toward the end of acquisition, during the refinement of avoidance behavior when few (if any) shocks are being received. Therefore, as startle sensitivity differences dissipate (possibly a sign of normalizing vigilance), enhancements in responsivity appear (a possible sign of increased arousal).

Historically, female rats are known to generally acquire discrete lever-press avoidance, as well as other active avoidance behaviors, better than their male counterparts (Beatty & Beatty, 1970; Scouten et al., 1975; Van Oyen et al., 1981; Heinsbroek et al., 1983; Beck et al.,

Male SD and WKY rats were trained to acquire a lever-press, active avoidance behavior (or served as non-trained homecage controls), and were tested weekly to assess changes in startle sensitivity and responsivity. In concert with prior studies, male WKY rats demonstrated greater baseline startle magnitudes and equivalent startle sensitivity. Subsequently, acquisition of avoidance appeared to be equivalent between the strains. In addition, WKY rats reached greater asymptotic avoidance performance, as seen in previous

Fig. 4. Lever-press behavior to avoid intermittent footshock was reliably acquired in both strains of male rats (main effect Session, F (9, 135) = 25.0, p < .001 and Session x Trial interaction, F (171, 2394) = 1.4, p < .001), but WKY rats exhibited more avoidance responses per session in the later sessions of the acquisition phase. During extinction, differences between strains were not apparent across sessions, but they were significant within sessions

Increases in startle sensitivity and responsivity are expected early in acquisition if shock exposure causes changes in vigilance and arousal, but, if learning and performing an avoidance behavior causes anxiety, startle measures should be elevated during the later weeks of acquisition. Thus, as shown in Figure 5, there was a strain-independent elevation in startle sensitivity displayed by those being trained in avoidance behavior. This difference is evident from the beginning of acquisition, when animals are receiving the most of shocks, and dissipates by the end of acquisition. Conversely, startle responsivity, as demonstrated by relative increases in startle magnitudes above baseline, is largely unchanged early in acquisition, when enhanced startle sensitivity is greatest. Startle responsivity in avoidancetrained animals increased toward the end of acquisition, during the refinement of avoidance behavior when few (if any) shocks are being received. Therefore, as startle sensitivity differences dissipate (possibly a sign of normalizing vigilance), enhancements in

Historically, female rats are known to generally acquire discrete lever-press avoidance, as well as other active avoidance behaviors, better than their male counterparts (Beatty & Beatty, 1970; Scouten et al., 1975; Van Oyen et al., 1981; Heinsbroek et al., 1983; Beck et al.,

(Strain x Trial interaction, F (19, 266) = 3.3, p < .001.

responsivity appear (a possible sign of increased arousal).

studies (see Figure 4).

Fig. 5. Exposure to avoidance learning was associated with an increase in startle sensitivity in both strains of male rats (main effect Avoidance, F (1, 28) = 4.2, p < .05. Otherwise, significant differences across the three stimulus intensities used (main effect Intensity F (2, 56) = 595.5, p < .001) were coupled with a marginally significant difference in the percentage of startles elicited across the 4 weeks of acquisition (Avoidance x Week interaction F (3, 84) = 2.4, p < .07. By the fourth week of avoidance acquisition, differences in startle sensitivity between the avoidance-trained group and the homecage control group were greatly reduced. Although not statistically significant, signs of increased startle magnitude began to become apparent in SD rats by the last week of acquisition.

2010; Beck et al., 2011), suggesting that they may exhibit greater changes in general arousal, as indexed by startle reactivity measures, with the experience of shock and/or the acquisition of the avoidant behavior. Using the same procedures as described for the male rats, female rats demonstrated similar patterns of initial startle reactivity: greater startle magnitudes in WKY rats but equivalent startle sensitivity across strain. During avoidance

Acquisition of Active Avoidance Behavior as a Precursor

to Changes in General Arousal in an Animal Model of PTSD 83

Fig. 7. Exposure to avoidance learning was associated with strain and intensity-dependent changes in startle sensitivity. Early acquisition showed elevations in the startles elicited at the lowest intensity. For SD females this continued into the second week and expanded to the middle intensity. Differences within the SD strain declined over the later two weeks, but, at the same time, a decrease in startle sensitivity became apparent in the female WKY rats. These impressions were confirmed by significant Strain x Avoidance, F (1, 28) = 5.2, p < .03, Strain x Intensity, F (2, 56) = 11.3, p < .001, and Avoidance x Week x Intensity, F (6, 168) = 3.0, p < .007, interactions. Startle magnitudes did not significantly differ between groups

caused reported increases in agitation towards the researchers and chronic increases in mean systolic and diastolic blood pressure in well-avoiding monkeys (Forsyth, 1969). In contrast, the reduction in startle sensitivity, in female WKY rats, is suggestive of an avoidance-induced reduction in general vigilance. Might this suggest female WKY rats

across the four weeks of acquisition.

training, female rats exhibited rapid acquisition of the lever-press avoidance response (see Figure 6). However, the effect avoidance had on arousal was different than the effects observed in male rats of these strains. As shown in Figure 7, startle sensitivity was transiently elevated in SD rats training in avoidance learning, but WKY rats exhibited much less elevations in sensitivity early in acquisition. In stark contrast, at the end of acquisition, WKY rats trained in avoidance were showing a reduction in startle sensitivity. Thus, while female SD rats demonstrate an enhancement in sensitivity similar to male rats, WKY rats demonstrate a unique decrease in reactivity compared to their untrained controls in the later phase of acquisition. Also, unlike the male rats, no consistent changes in startle responsivity were observed in SD or WKY females as was seen in male rats.

Fig. 6. Lever-press behavior to avoid intermittent footshock was reliably acquired in both strains of female rats (main effect Session, F (9, 135) = 33.4, p < .001 and Session x Trial interaction, F (171, 2394) = 1.5, p < .001). During extinction, differences between strains became evident within the first session and continued throughout the extinction phase of the experiment (main effects, Strain, F (1, 14) = 7.4, p < .05, Session, F (9, 135) = 13.7, p < .001, and Trial, F (19, 266) = 10.9, p < .001).

These data suggest there are different aspects to an avoidance-induced change in vigilance and/or general arousal (as reflected by an increase in startle sensitivity and responsivity, respectively). Increases in startle sensitivity generally occurred proximal to experiences with periodic foot-shock during the early phase of acquisition (when rats are slowly transitioning from a majority of escape responses to an increasing number of avoidance responses). This pattern appears consistent, albeit to varying degrees, across both sexes of each strain, and suggests experience with a painful stimulus is increasing vigilance in those animals. However, strain and sex differences become evident as avoidance responses occur in a greater majority. Of note are the changes in startle responsivity in male SD rats and startle sensitivity in female WKY rats. These groups exhibited divergent changes, with male SD rats exhibiting enhanced startle responsivity and female WKY rats exhibiting decreased startle sensitivity. The enhancement in startle responsivity, as a possible index of general arousal, observed in male SD rats is akin to the described usage of avoidance decades ago, where daily sessions of non-cued avoidance (i.e. Sidman avoidance) over several months

training, female rats exhibited rapid acquisition of the lever-press avoidance response (see Figure 6). However, the effect avoidance had on arousal was different than the effects observed in male rats of these strains. As shown in Figure 7, startle sensitivity was transiently elevated in SD rats training in avoidance learning, but WKY rats exhibited much less elevations in sensitivity early in acquisition. In stark contrast, at the end of acquisition, WKY rats trained in avoidance were showing a reduction in startle sensitivity. Thus, while female SD rats demonstrate an enhancement in sensitivity similar to male rats, WKY rats demonstrate a unique decrease in reactivity compared to their untrained controls in the later phase of acquisition. Also, unlike the male rats, no consistent changes in startle responsivity

Fig. 6. Lever-press behavior to avoid intermittent footshock was reliably acquired in both strains of female rats (main effect Session, F (9, 135) = 33.4, p < .001 and Session x Trial interaction, F (171, 2394) = 1.5, p < .001). During extinction, differences between strains became evident within the first session and continued throughout the extinction phase of the experiment (main effects, Strain, F (1, 14) = 7.4, p < .05, Session, F (9, 135) = 13.7, p < .001,

These data suggest there are different aspects to an avoidance-induced change in vigilance and/or general arousal (as reflected by an increase in startle sensitivity and responsivity, respectively). Increases in startle sensitivity generally occurred proximal to experiences with periodic foot-shock during the early phase of acquisition (when rats are slowly transitioning from a majority of escape responses to an increasing number of avoidance responses). This pattern appears consistent, albeit to varying degrees, across both sexes of each strain, and suggests experience with a painful stimulus is increasing vigilance in those animals. However, strain and sex differences become evident as avoidance responses occur in a greater majority. Of note are the changes in startle responsivity in male SD rats and startle sensitivity in female WKY rats. These groups exhibited divergent changes, with male SD rats exhibiting enhanced startle responsivity and female WKY rats exhibiting decreased startle sensitivity. The enhancement in startle responsivity, as a possible index of general arousal, observed in male SD rats is akin to the described usage of avoidance decades ago, where daily sessions of non-cued avoidance (i.e. Sidman avoidance) over several months

were observed in SD or WKY females as was seen in male rats.

and Trial, F (19, 266) = 10.9, p < .001).

Fig. 7. Exposure to avoidance learning was associated with strain and intensity-dependent changes in startle sensitivity. Early acquisition showed elevations in the startles elicited at the lowest intensity. For SD females this continued into the second week and expanded to the middle intensity. Differences within the SD strain declined over the later two weeks, but, at the same time, a decrease in startle sensitivity became apparent in the female WKY rats. These impressions were confirmed by significant Strain x Avoidance, F (1, 28) = 5.2, p < .03, Strain x Intensity, F (2, 56) = 11.3, p < .001, and Avoidance x Week x Intensity, F (6, 168) = 3.0, p < .007, interactions. Startle magnitudes did not significantly differ between groups across the four weeks of acquisition.

caused reported increases in agitation towards the researchers and chronic increases in mean systolic and diastolic blood pressure in well-avoiding monkeys (Forsyth, 1969). In contrast, the reduction in startle sensitivity, in female WKY rats, is suggestive of an avoidance-induced reduction in general vigilance. Might this suggest female WKY rats

Acquisition of Active Avoidance Behavior as a Precursor

F (2, 56) = 10.5, p < .001).

responsivity compared to SD rats.

to Changes in General Arousal in an Animal Model of PTSD 85

Fig. 8. During the extinction phase, avoidance-trained male SD and WKY rats did not show significantly different startle sensitivity measures compared to their same-strain controls. However, differences in startle magnitude continued into the extinction phase for those male SD rats that had been trained in avoidance behavior. In addition, in the midst of extinction sessions, startle magnitudes of male WKY rats were elevated compared to their homecage control counterparts (main effects of Avoidance, F (1, 28) = 4.1, p < .05 and Week,

decreases in startle sensitivity seen in WKY female rats did not persist. Innate differences in startle behavior, however, remained. WKY rats of both sexes demonstrated higher startle

Since these between-group differences in startle responsivity are not observed during the subsequent month where they remained in the home-cage, these results suggest there is a connection between displaying some level of active-avoidance behavior, on a regular basis, and persistent changes in vigilance or general arousal. For males, differences were observed

adapt better to a stressful environment? If so, would these changes in vigilance and arousal be maintained in the absence of the actual threat? On the other hand, is the reduction in startle responsivity evidence for a difference in the underlying associations made during the acquisition of the avoidance behavior? Continued assessment of startle sensitivity and responsivity during the extinction phase should provide further evidence for or against these interpretations.

#### **3.2 Extinction of active avoidance, removal of avoidance context and persistent changes in startle reactivity**

As shown in Figure 4, unlike previous studies, there was minimal difference in the extinction rates of male SD and WKY rats; still the WKY rats extinguished slightly slower. Nonetheless, the resultant effects on the indexes of startle were rather clear. As with the end of the acquisition phase, startle sensitivity did not appreciably change during the extinction phase; however, differences in startle responsivity grew in appearance (see Figure 8). Differences first observed at the end of acquisition in the male SD rats continued to be present during the first two weeks of extinction. Interestingly, by the end of the extinction phase, avoidance-trained WKY rats were also exhibiting greater startle responsivity than their non-trained counterparts. This suggests the arousal displayed by male SD rats is contingent upon emitting a certain level of avoidance behavior, even in the absence of shock. In contrast, the male WKY rats show differences as they extinguish the avoidant behavior. This may reflect an increase in arousal due to the slow abandonment of the avoidant behavior. Following this logic, the male WKY rats could have perceived that their behavior does not control the presence or absence of the shock anymore; the result is an increase in general arousal during extinction.

The female rats exhibited a substantial difference in their rates of extinction (see Figure 6), with female WKY rats extinguishing much slower than female SD rats. This may be attributable to differences in how the females of these strains extinguish the response, since both groups exhibited very similar acquisition rates and attained a similar asymptotic level of responding. However, under similar avoidance learning conditions we have not observed such a difference (see Figure 3). The one difference in procedure from our previous experiments is the addition of the weekly startle tests. This may be an example of the vigilance/arousal test influencing performance on subsequent avoidance acquisition/extinction session days. The results of the startle test showed that intra-strain differences in the startle sensitivity continued from the end of avoidance acquisition. As shown in Figure 9, female WKY rats showed reduced responding, which eventually normalized through the extinction period. Female SD rats still showed signs of increased startle sensitivity early in the extinction phase. These data suggest that elevated vigilance continues to be expressed in female SD rats for a period of time even in the absence of shock. Conversely, the decreased vigilance in female WKY rats, attributed to the reduction in shock exposure, continues, as the shocks remain absent in extinction, then eventually normalize.

Following extinction training, all rats were allowed to remain in their home cages for a number of weeks, with startle measures taken every week, as during acquisition and extinction. These sessions allowed for the measurement of long-term changes in arousal and vigilance following avoidance training. In male rats, the elevations in startle responsivity returned to baseline. Thus changes in arousal that were evident during avoidance training did not persist following cessation of training. The same was seen in female rats, where the

adapt better to a stressful environment? If so, would these changes in vigilance and arousal be maintained in the absence of the actual threat? On the other hand, is the reduction in startle responsivity evidence for a difference in the underlying associations made during the acquisition of the avoidance behavior? Continued assessment of startle sensitivity and responsivity during the extinction phase should provide further evidence for or against

**3.2 Extinction of active avoidance, removal of avoidance context and persistent** 

As shown in Figure 4, unlike previous studies, there was minimal difference in the extinction rates of male SD and WKY rats; still the WKY rats extinguished slightly slower. Nonetheless, the resultant effects on the indexes of startle were rather clear. As with the end of the acquisition phase, startle sensitivity did not appreciably change during the extinction phase; however, differences in startle responsivity grew in appearance (see Figure 8). Differences first observed at the end of acquisition in the male SD rats continued to be present during the first two weeks of extinction. Interestingly, by the end of the extinction phase, avoidance-trained WKY rats were also exhibiting greater startle responsivity than their non-trained counterparts. This suggests the arousal displayed by male SD rats is contingent upon emitting a certain level of avoidance behavior, even in the absence of shock. In contrast, the male WKY rats show differences as they extinguish the avoidant behavior. This may reflect an increase in arousal due to the slow abandonment of the avoidant behavior. Following this logic, the male WKY rats could have perceived that their behavior does not control the presence or absence of the shock anymore; the result is an

The female rats exhibited a substantial difference in their rates of extinction (see Figure 6), with female WKY rats extinguishing much slower than female SD rats. This may be attributable to differences in how the females of these strains extinguish the response, since both groups exhibited very similar acquisition rates and attained a similar asymptotic level of responding. However, under similar avoidance learning conditions we have not observed such a difference (see Figure 3). The one difference in procedure from our previous experiments is the addition of the weekly startle tests. This may be an example of the vigilance/arousal test influencing performance on subsequent avoidance acquisition/extinction session days. The results of the startle test showed that intra-strain differences in the startle sensitivity continued from the end of avoidance acquisition. As shown in Figure 9, female WKY rats showed reduced responding, which eventually normalized through the extinction period. Female SD rats still showed signs of increased startle sensitivity early in the extinction phase. These data suggest that elevated vigilance continues to be expressed in female SD rats for a period of time even in the absence of shock. Conversely, the decreased vigilance in female WKY rats, attributed to the reduction in shock exposure, continues, as the shocks remain absent in extinction, then eventually

Following extinction training, all rats were allowed to remain in their home cages for a number of weeks, with startle measures taken every week, as during acquisition and extinction. These sessions allowed for the measurement of long-term changes in arousal and vigilance following avoidance training. In male rats, the elevations in startle responsivity returned to baseline. Thus changes in arousal that were evident during avoidance training did not persist following cessation of training. The same was seen in female rats, where the

these interpretations.

normalize.

**changes in startle reactivity** 

increase in general arousal during extinction.

Fig. 8. During the extinction phase, avoidance-trained male SD and WKY rats did not show significantly different startle sensitivity measures compared to their same-strain controls. However, differences in startle magnitude continued into the extinction phase for those male SD rats that had been trained in avoidance behavior. In addition, in the midst of extinction sessions, startle magnitudes of male WKY rats were elevated compared to their homecage control counterparts (main effects of Avoidance, F (1, 28) = 4.1, p < .05 and Week, F (2, 56) = 10.5, p < .001).

decreases in startle sensitivity seen in WKY female rats did not persist. Innate differences in startle behavior, however, remained. WKY rats of both sexes demonstrated higher startle responsivity compared to SD rats.

Since these between-group differences in startle responsivity are not observed during the subsequent month where they remained in the home-cage, these results suggest there is a connection between displaying some level of active-avoidance behavior, on a regular basis, and persistent changes in vigilance or general arousal. For males, differences were observed

Acquisition of Active Avoidance Behavior as a Precursor

with high levels of active avoidance behavior.

**3.3 Active avoidance and general physiology** 

constitute a PTSD diagnosis.

weight (especially in WKY rats).

to Changes in General Arousal in an Animal Model of PTSD 87

example, since the males learned to manipulate the environment to avoid noxious stimuli, maybe defensive reactions are enhanced; the animal is more "at-the-ready". Increases in vigilance may cause the female SD rats to perceive their environment in a more apprehensive manner. Interestingly, inescapable shock causes a transient reduction in SD startle responsivity without affecting startle sensitivity (Beck et al., 2002; Beck & Servatius, 2005). With this in mind, it was surprising to see reductions in startle sensitivity in the female WKY rats, although we should acknowledge it was long after the type of shock experienced during inescapable shock that reduces female SD startle responsivity. We could speculate that this lack of sensitivity to the startle pulses in the WKY females is the counterpoint to "increased vigilance", being post-stress "numbness", but we do not have any other data to substantiate such a claim. Additional testing across other modalities may give us a better idea of the scope of this reduction in reactivity that appears to be correlated

These data provide an interesting example of how startle reactivity can be enhanced by prior exposure to an escapable and avoidable stressor. Moreover, as was observed following inescapable stress (Servatius et al., 1995; Beck et al., 2002; Manion et al., 2007; Manion et al., 2010), the presentation of enhanced startle reactivity in male rats did not occur proximal to any period of significant shock exposure. This finding is important for 2 reasons. First, it shows that inescapable and uncontrollable stress is not necessary to increase startle reactivity, and yet, the appearance of the startle enhancement is still delayed following predictable and controllable shocks. Second, these features are suggestive that a mechanism not specifically triggered by the shock is causing: 1) general arousal to increase over time in male rats; 2) vigilance to remain elevated in female SD rats; and 3) vigilance to be reduced in female WKY rats (possibly a transient numbing effect). It is important to also note the differences in these patterns across subject groups can be translated into different symptoms associated with PTSD: arousal, vigilance, and numbness. These group differences may help us understand why different individuals present with certain symptoms yet, in total, still

Reactivity to stressors may be further characterized by their effect on the general physiology of the rats. Growth, as measured by changes in bodyweight from the beginning of the experiment, provides a measure to investigate the effects of avoidance on general physiology, and to see if correlations exist between these changes and startle reactivity. In male SD rats, growth was suppressed during acquisition, but recovered by the termination of extinction training (see Figure 10). In WKY rats, the exact opposite effect was observed. Differences in bodyweight slowly emerged across training, with decreased growth in avoidance trained rats becoming evident in extinction and further developing following the end of training. Thus while male SD rats demonstrate a decrease in growth proximal to shock administration, WKY rats demonstrate a continual decrement in growth that developed after the removal of the shock. In females, SD rats demonstrated a similar pattern to male WKY rats, with differences in growth emerging over training. In female WKY rats, no differences in growth are seen at any point during training. Thus it appears that changes in bodyweight, often thought to reflect physiological reactions to stress, do not mimic all of the changes observed in startle sensitivity and responsivity across all groups; although, changes in startle responsivity in male rats do somewhat follow periods of lower body

Fig. 9. Exposure to avoidance learning was associated with strain and intensity-dependent changes in startle sensitivity in female rats that persisted into the extinction phase. For SD rats startle sensitivity was still elevated during the first week of extinction. In a similar fashion, female WKY rats this continued to show a reduction in startle sensitivity into the initial week of extinction. These impressions were confirmed by significant Strain x Avoidance, F (1, 28) = 4.6, p < .04, Strain x Intensity, F (2, 56) = 6.5, p < .002, and Week x Intensity, F (4, 112) = 12.6, p < .001, interactions. There were no significant differences during the extinction phase measures of startle responsivity attributable to prior avoidance learning.

during extinction (in both strains) suggesting that the behavior itself (not actual response to shocks) may be sufficient to induce a state of general arousal. Herein lies the possible connection to the growth of general PTSD symptoms with the trajectory of avoidance symptoms (O'Donnell et al., 2007). Similarly, the persistent elevation in vigilance observed in female SD rats can also fit the description of PTSD. It may be that avoidance learning causes changes in the brain systems underlying fundamental defensive behaviors. For

Fig. 9. Exposure to avoidance learning was associated with strain and intensity-dependent changes in startle sensitivity in female rats that persisted into the extinction phase. For SD rats startle sensitivity was still elevated during the first week of extinction. In a similar fashion, female WKY rats this continued to show a reduction in startle sensitivity into the initial week of extinction. These impressions were confirmed by significant Strain x Avoidance, F (1, 28) = 4.6, p < .04, Strain x Intensity, F (2, 56) = 6.5, p < .002, and Week x Intensity, F (4, 112) = 12.6, p < .001, interactions. There were no significant differences during the extinction phase measures of startle responsivity attributable to prior avoidance

during extinction (in both strains) suggesting that the behavior itself (not actual response to shocks) may be sufficient to induce a state of general arousal. Herein lies the possible connection to the growth of general PTSD symptoms with the trajectory of avoidance symptoms (O'Donnell et al., 2007). Similarly, the persistent elevation in vigilance observed in female SD rats can also fit the description of PTSD. It may be that avoidance learning causes changes in the brain systems underlying fundamental defensive behaviors. For

learning.

example, since the males learned to manipulate the environment to avoid noxious stimuli, maybe defensive reactions are enhanced; the animal is more "at-the-ready". Increases in vigilance may cause the female SD rats to perceive their environment in a more apprehensive manner. Interestingly, inescapable shock causes a transient reduction in SD startle responsivity without affecting startle sensitivity (Beck et al., 2002; Beck & Servatius, 2005). With this in mind, it was surprising to see reductions in startle sensitivity in the female WKY rats, although we should acknowledge it was long after the type of shock experienced during inescapable shock that reduces female SD startle responsivity. We could speculate that this lack of sensitivity to the startle pulses in the WKY females is the counterpoint to "increased vigilance", being post-stress "numbness", but we do not have any other data to substantiate such a claim. Additional testing across other modalities may give us a better idea of the scope of this reduction in reactivity that appears to be correlated with high levels of active avoidance behavior.

These data provide an interesting example of how startle reactivity can be enhanced by prior exposure to an escapable and avoidable stressor. Moreover, as was observed following inescapable stress (Servatius et al., 1995; Beck et al., 2002; Manion et al., 2007; Manion et al., 2010), the presentation of enhanced startle reactivity in male rats did not occur proximal to any period of significant shock exposure. This finding is important for 2 reasons. First, it shows that inescapable and uncontrollable stress is not necessary to increase startle reactivity, and yet, the appearance of the startle enhancement is still delayed following predictable and controllable shocks. Second, these features are suggestive that a mechanism not specifically triggered by the shock is causing: 1) general arousal to increase over time in male rats; 2) vigilance to remain elevated in female SD rats; and 3) vigilance to be reduced in female WKY rats (possibly a transient numbing effect). It is important to also note the differences in these patterns across subject groups can be translated into different symptoms associated with PTSD: arousal, vigilance, and numbness. These group differences may help us understand why different individuals present with certain symptoms yet, in total, still constitute a PTSD diagnosis.

#### **3.3 Active avoidance and general physiology**

Reactivity to stressors may be further characterized by their effect on the general physiology of the rats. Growth, as measured by changes in bodyweight from the beginning of the experiment, provides a measure to investigate the effects of avoidance on general physiology, and to see if correlations exist between these changes and startle reactivity. In male SD rats, growth was suppressed during acquisition, but recovered by the termination of extinction training (see Figure 10). In WKY rats, the exact opposite effect was observed. Differences in bodyweight slowly emerged across training, with decreased growth in avoidance trained rats becoming evident in extinction and further developing following the end of training. Thus while male SD rats demonstrate a decrease in growth proximal to shock administration, WKY rats demonstrate a continual decrement in growth that developed after the removal of the shock. In females, SD rats demonstrated a similar pattern to male WKY rats, with differences in growth emerging over training. In female WKY rats, no differences in growth are seen at any point during training. Thus it appears that changes in bodyweight, often thought to reflect physiological reactions to stress, do not mimic all of the changes observed in startle sensitivity and responsivity across all groups; although, changes in startle responsivity in male rats do somewhat follow periods of lower body weight (especially in WKY rats).

Acquisition of Active Avoidance Behavior as a Precursor

and Session x Trial, F (171, 2565) = 1.4, p < .01 interactions).

to Changes in General Arousal in an Animal Model of PTSD 89

Fig. 11. Avoidance performance of SD and WKY rats. Strain designations are in the figure legend. WKY rats acquire the avoidance response faster and reach higher asymptotic performance than SD rats (main effect, Strain, F (1, 15) = 9.2, p < .01 and Session x Trial interaction, F (171, 2565) = 1.3, p < .001. WKY rats also extinguished the response less than SD rats (main effect, Strain, F (1, 15) = 7.0, p < .01, Strain x Trial, F (19, 285) = 2.1, p < .01

effect is not strain dependent and dissipates as the rats are exposed to fewer shocks. With respect to startle magnitude, WKY rats exhibited higher startle magnitudes than SD rats (see Figure 12). In SD rats, no differences in relative startle magnitudes were observed between avoidance and yoked rats, but in WKY rats, elevations in startle responsivity developed late

Fig. 12. Differences in startle magnitude emerged between the avoidance and yoked-shock

group during the later phase of acquisition in male WKY rats. This impression was confirmed by a significant Strain x Week interaction, F (3, 90) = 6.1, p < .001 and a marginally significant Avoidance x Week interaction, F (3, 90) = 2.3, p < .07. Unlike the previous experiment, male SD rats did not attain a level of avoidance performance proximal to that of the WKY rats, and failed to develop enhanced startle responsivity over training (left). The differences in startle magnitude between avoidance and yoke-control male WKY rats continued during the extinction phase of the experiment (right). This impression was

confirmed by a significant Strain x Week interaction, F (2, 60) = 7.5, p < .001 and a

behavior and their yoked controls in the WKY strain.

marginally significant Avoidance x Week interaction, F (2,60) = 2.5, p < .08. Two of the three weekly startle tests found higher startle magnitudes between those trained in avoidance

Fig. 10. Relative body weight of male and female SD and WKY rats. Data are averaged into phases: A-Acquisition, E-Extinction, and H-Home Cage. Relative body weights were determined by dividing average weight during a phase by the weight during the pretraining startle session. Male SD rats gain more relative weight than male WKY rats. Male SD rats that underwent avoidance demonstrated suppressed growth during acquisition and extinction. Male WKY rats demonstrated a diverging growth pattern; rats that underwent avoidance learning gained less weight than their home cage controls, an effect that did not emerge until extinction and persisted into the home cage phase. Female WKY rats gained more relative weight than female SD rats. Female SD rats that underwent avoidance demonstrated suppressed growth during extinction and the home cage phase relative to home cage controls. These differences in growth were not observed in female WKY rats. \*p< .05

#### **3.4 Active avoidance, shock exposure and changes in startle reactivity**

The confounding variables of shock exposure controllability and exposure to shock limited our ability to make firm conclusions regarding what caused the enhanced startle responses to occur in avoidance-trained rats. Hence, we designed a follow-up study that substituted an additional control group for our baseline comparison of avoidance-trained rats. This control group was placed in the training boxes at the same time the others were being trained to avoid the shock. The rats in this new condition were each paired to a rat in the avoidancetraining condition such that when an avoidance rat was shocked, so was the paired control (yoked condition). Thus, the yoked rats in this experiment heard, saw, and felt the same stimuli as their avoidance learning paired counterparts, but the lever in their chambers was disabled. For these yoked controls, they may learn the predictive relationship between the stimuli and the shocks, but they will not learn or experience any perceived control over the occurrence of shock.

As shown in Figure 11, the rats of both strains exhibited a clear acquisition of a lever-press avoidance response. WKY rats exhibited a much higher level of asymptotic performance than did SD rats. Furthermore, the subsequent extinction of the response was less apparent in WKY rats compared to SD rats. The analysis of startle sensitivity found that both strains of rats trained in avoidance learning increased their sensitivity to the acoustic startle pulses the day following the first training session, as did the yoked controls from baseline. This confirms that the short-term change in startle sensitivity is a product of shock exposure, not the acquisition process involved in learning to escape or avoid the shocks. Moreover, the

Fig. 10. Relative body weight of male and female SD and WKY rats. Data are averaged into phases: A-Acquisition, E-Extinction, and H-Home Cage. Relative body weights were determined by dividing average weight during a phase by the weight during the pretraining startle session. Male SD rats gain more relative weight than male WKY rats. Male SD rats that underwent avoidance demonstrated suppressed growth during acquisition and extinction. Male WKY rats demonstrated a diverging growth pattern; rats that underwent avoidance learning gained less weight than their home cage controls, an effect that did not emerge until extinction and persisted into the home cage phase. Female WKY rats gained more relative weight than female SD rats. Female SD rats that underwent avoidance demonstrated suppressed growth during extinction and the home cage phase relative to home cage controls. These differences in growth were not observed in female WKY rats.

**3.4 Active avoidance, shock exposure and changes in startle reactivity** 

The confounding variables of shock exposure controllability and exposure to shock limited our ability to make firm conclusions regarding what caused the enhanced startle responses to occur in avoidance-trained rats. Hence, we designed a follow-up study that substituted an additional control group for our baseline comparison of avoidance-trained rats. This control group was placed in the training boxes at the same time the others were being trained to avoid the shock. The rats in this new condition were each paired to a rat in the avoidancetraining condition such that when an avoidance rat was shocked, so was the paired control (yoked condition). Thus, the yoked rats in this experiment heard, saw, and felt the same stimuli as their avoidance learning paired counterparts, but the lever in their chambers was disabled. For these yoked controls, they may learn the predictive relationship between the stimuli and the shocks, but they will not learn or experience any perceived control over the

As shown in Figure 11, the rats of both strains exhibited a clear acquisition of a lever-press avoidance response. WKY rats exhibited a much higher level of asymptotic performance than did SD rats. Furthermore, the subsequent extinction of the response was less apparent in WKY rats compared to SD rats. The analysis of startle sensitivity found that both strains of rats trained in avoidance learning increased their sensitivity to the acoustic startle pulses the day following the first training session, as did the yoked controls from baseline. This confirms that the short-term change in startle sensitivity is a product of shock exposure, not the acquisition process involved in learning to escape or avoid the shocks. Moreover, the

\*p< .05

occurrence of shock.

Fig. 11. Avoidance performance of SD and WKY rats. Strain designations are in the figure legend. WKY rats acquire the avoidance response faster and reach higher asymptotic performance than SD rats (main effect, Strain, F (1, 15) = 9.2, p < .01 and Session x Trial interaction, F (171, 2565) = 1.3, p < .001. WKY rats also extinguished the response less than SD rats (main effect, Strain, F (1, 15) = 7.0, p < .01, Strain x Trial, F (19, 285) = 2.1, p < .01 and Session x Trial, F (171, 2565) = 1.4, p < .01 interactions).

effect is not strain dependent and dissipates as the rats are exposed to fewer shocks. With respect to startle magnitude, WKY rats exhibited higher startle magnitudes than SD rats (see Figure 12). In SD rats, no differences in relative startle magnitudes were observed between avoidance and yoked rats, but in WKY rats, elevations in startle responsivity developed late

Fig. 12. Differences in startle magnitude emerged between the avoidance and yoked-shock group during the later phase of acquisition in male WKY rats. This impression was confirmed by a significant Strain x Week interaction, F (3, 90) = 6.1, p < .001 and a marginally significant Avoidance x Week interaction, F (3, 90) = 2.3, p < .07. Unlike the previous experiment, male SD rats did not attain a level of avoidance performance proximal to that of the WKY rats, and failed to develop enhanced startle responsivity over training (left). The differences in startle magnitude between avoidance and yoke-control male WKY rats continued during the extinction phase of the experiment (right). This impression was confirmed by a significant Strain x Week interaction, F (2, 60) = 7.5, p < .001 and a marginally significant Avoidance x Week interaction, F (2,60) = 2.5, p < .08. Two of the three weekly startle tests found higher startle magnitudes between those trained in avoidance behavior and their yoked controls in the WKY strain.

Acquisition of Active Avoidance Behavior as a Precursor

resistance to cease that response (maintaining control).

**5. Acknowledgements** 

**6. References** 

to Changes in General Arousal in an Animal Model of PTSD 91

strain and sex differences in learning processes that are involved in forming predictive associations under conditions where some level of stress is involved (Wood & Shors, 1998; Ricart et al., 2011a; Ricart et al., 2011b; Beck et al., 2011). When the requirement to cope is brought to the fore, these inherent differences in learning processes can be seen both in their rate to acquire an avoidance coping response (i.e. gaining control) and any subsequent

The problems associated with anxiety disorders, such as PTSD, are multifaceted and variable. In part, this is because different individuals perceive and cope with stressors differently. Active-avoidance behavior, using a lever-press, is rarely uniform within a group of animals, and, as evidenced from the data from our lab; the resulting effects on startle responses can be variable in when they emerge over time. Yet, the variability caused by having a subject-controlled manipulation of stressor exposure is important for understanding the disorder. Controllability may selectively influence certain individuals in a manner that causes increases in general arousal whereas others are not so affected. Moreover, when we consider vulnerability factors, such as demonstrated by the WKY rats (behavioral inhibition and higher baseline startle responses), individual differences in coping with stressor and different rates of acquisition of avoidant strategies should occur – as in the human condition. Gaining an understanding of the relationship between different symptoms, as demonstrated here between avoidance and arousal, will provide us with the knowledge to broaden our expectations for how different populations may develop the symptoms of PTSD. As shown here, our data suggest that the acute experience of pain is not sufficient to immediately increase startle responsivity, and it may not be a good marker for tracking the development of PTSD. Our data suggests that the avoidance process is already well-acquired when this other symptom becomes evident. As has been suggested from the clinical literature, increased expression of avoidance may be a very good marker for tracking the development of the disorder (Karamustafalioglu et al., 2006; O'Donnell et al., 2007). Further, breaking the adoption and utilization of avoidance strategies may lead to a reduction in general arousal (at least in males); therefore, some non-pharmacological therapeutic approaches (e.g. cognitive-behavioral therapy) may have beneficial effects on these two core features of PTSD. Additional research is required to better understand and track the developmental course of symptom expression in different subpopulations (e.g. women) such that our animal model systems can be better tailored to reflect the cascade of changes occurring in those people, especially for those at risk for developing the symptoms of PTSD.

The presented work was supported by grants from the Department of Veterans Affairs Office of Biomedical Laboratory Research & Development (1l01BX000218) and the Department of Defense Psychological Health/Traumatic Brain Injury Research Program of the Office of the Congressionally Directed Medical Research Programs (W81XWH-08-2-0657) to KDB. The opinions expressed are those of the authors and do not reflect the official position of the U.S.

American Psychiatric Association (2000). Diagnostic and Statistical Manual of Mental Disorders. (4th-TR ed.) Washington, DC: American Psychiatric Association.

Army, U.S. Department of Defense, or U.S. Department of Veterans Affairs.

in acquisition in both avoidance and yoked rats. These differences persisted into extinction and following the termination of training. Because each yoked rat experienced the same stimuli as an avoidance-trained rat, they should have learned the predictability of the stimuli. However, what is not learned is that they have any controllability. The poorer learning by the avoidance-trained SD rats in the yoked-avoidance experiment would conform to this idea. Because they did not acquire the response to the same level as those in the earlier experiment, this may have led them to also fail to show any difference in startle responsivity from their yoked controls as training progressed. Observed differences in avoidance-trained WKY rats suggest they too may have similar increased arousal with the adoption and use of active avoidance behavior.
