«Distribution Agreement In presenting this thesis or dissertation as a partial fulfillment of the requirements for an advanced degree from Emory ...»
Real-time PCR was performed using an Applied Biosystems 7500 Fast-Real Time PCR system (Applied Biosystems, Foster City, CA, USA). 2 µl of cDNA obtained from the isolated RNA were combined with Taqman probes specific for 18S rRNA (Accession No. X03205) or GABAA receptor α1 subunit (Accession No. NM_183326) and 1x Taqman universal PCR Master Mix (Applied Biosystems), and the reaction for each sample was performed in triplicate. The thermal cycling program consisted of: cycle 1 (20 min at 95° C) and cycles 2 through 40 (3 s at 95° C followed by 30 min at 60° C). Expression of the α1 subunit was normalized for all samples to expression of 18S rRNA. The 2−ΔΔCt method of relative quantification was used to calculate the fold change in expression of genes, and these values were used for statistical analysis (Livak and Schmittgen, 2001).
5.3.6 Behavioral Testing 18.104.22.168 Elevated Plus-Maze All behavioral tests were conducted during the light phase of a 12:12 hour light-dark cycle. The elevated plus-maze consisted of two open arms and two closed arms, each 50 x 11 cm and elevated to a height of 50 cm. Rats from ages P60-70 were placed for 5 min on the maze in a room illuminated by red lights suspended over the center of the maze. Videos of the session were recorded and analyzed offline. The durations of time spent in open and closed arms and maze center were measured using Topscan 2.00 (CleverSys Inc., Reston, VA, USA) with the following analysis parameters: animal size – 100 to 1000 px, max movement per frame - 200 px, animal detection threshold - contrast level 1. Center crossings and head dips (defined as extension of the head below the plane of the maze while the body is on an open arm) were counted manually by an experimenter blinded to animal group.
22.214.171.124 Sociability and Social Choice Test (SCT) The SCT was conducted using a previously established paradigm for social discrimination and social novelty preference in rodents (van der Kooij and Sandi, 2012). On testing day, in a room illuminated with red light, rats were first acclimated for 5 min to the testing apparatus. The apparatus, made of clear plexiglass, consists of 3 adjacent chambers of dimensions 50x25x29 cm (LxWxH). Circular portals, 11 cm in diameter, on the inner walls allow access to between the chambers, and rats could freely explore the box. The corner of each outer chamber contained a wire mesh cage, of dimensions 22x14x18 cm, which was empty during acclimation.
After acclimation, the test rat was removed from the apparatus and the 5 min sociability test was administered as follows: before the test rat was returned to the apparatus, into one of the wire mesh cages was placed an unfamiliar male conspecific of comparable age (C1), and into the other an inanimate object of comparable size to a P60 rat. The test rat was placed in the center chamber of the testing apparatus, and its behavior was video recorded via an overhead camera. Following the sociability test, the test rat was removed and another conspecific (C2) was placed into the cage formerly containing the inanimate object. For the SCT, the test rat was placed back in the testing apparatus and behavior was recorded for 5 min. The conspecific rats C1 and C2 were used for up to 4 rounds of testing (4 test rats), alternating between being the first and second rat presented. Time spent interacting with the object and conspecific rats were scored by hand, offline, from overhead videos by a blinded experimenter.
126.96.36.199 Isolation-Induced Ultrasonic Vocalizations (USVs) This test was based on the protocol described by Hofer et al. (2002). Pups were tested for USVs at P7 and P17. To minimize the separation stress to the pups, dams were left with the litters during habituation for 1 hour prior to testing. After habituation, pups were separated from the litter in series, and each was taken to another room and placed in a test cage with bedding for 5 min. The Sonotrack system (Metris, Kruisweg, Netherlands) was used to detect and analyze USVs, and the detector was mounted approximately 50 cm above the cage. The number of
vocalizations was obtained for each pup offline with the following analysis parameters:
discrimination factor: 10, width: 15 ms, and amplitude threshold: 22.13 dB. USVs with a mean frequency below 25kHz were excluded, and the detection ceiling was 90kHz.
5.3.7 Statistics For all statistical tests, significance was assigned based on α = 0.05. All data sets were tested for normality using the Shapiro–Wilk test (α = 0.05) and for homoscedasticity using Levene's test (α = 0.0001), implemented in Matlab. Comparisons of behavior in the EPM were made using Student’s t-tests in Prism 4 (GraphPad, LaJolla, CA). Measures of open arm times and dips in the EPM failed the Shapiro-Wilk test, and were therefore analyzed with MannWhitney U tests in Prism. For the social choice and interaction tests, time spent interacting with probe objects and conspecifics were analyzed with Two-way ANOVA, and preference indices were analyzed with a Student’s t-test. Ultrasonic vocalization frequency was analyzed at both time points using a Student’s t-test. The effects of age and stress on all electrophysiological properties were tested using Two-way ANOVAs in Prism and, if a significant main effect was observed, with Bonferroni post-tests. The following data were log-transformed before ANOVA to achieve homoscedasticity: body mass, input resistance, membrane time constant, AP half-width, 10-90% rise-time, 90-10% decay-time, rheobase, and current for maximum spike rate. Since only 5 neurons had a detectable fAHP at P14, this time point was excluded from analysis, allowing the data to meet the homoscedasticity assumption.
5.4 Results Electrophysiological data were collected from 407 BLA neurons from 78 male rats on postnatal days (P)10, 14, 17, 21, 28, and 60, representing 23 distinct litters (at least 4 litters at each time-point/group). Neurons were divided into different datasets based on the recording configuration and patch solution, and a total of 238 neurons were recorded in voltage-clamp mode with a Cesium-based patch solution, and 169 neurons were recorded in current-clamp mode with a Potassium-based patch solution. Neurons were recorded in both configurations from most animals included in the study. A total of 203 neurons were recorded from PS animals and 204 from controls. From 3 animals sacrificed for electrophysiology in both groups at each time point, a portion of the BLA was microdissected for RT-PCR. Another 3 animals for both groups per time point were perfused for immunohistochemistry. In addition 37 male rats from 10 litters (5 PS and 5 control litters) were tested in the elevated plus-maze for anxiety-like behavior. Nine adult rats from 2 litters were used for social novelty testing, and 19 rats from 4 litters were tested for ultrasonic vocalizations. There was no effect of PS on litter size (11.7 ± 0.4 pups per control litter vs. 11.4 ± 0.6 pups per PS litter, Student’s t-test, p = 0.66, n = 18), but body weight was significantly greater in PS offspring than controls (Figure 5.2; Two-way ANOVA, main effect of PS: F1,69 = 4.09; p 0.05). At P60, there was a 10.2% increase in body weight due to PS (408.9 ±
40.7 g for PS vs. 371.1 ± 32.7 g for control).
5.4.1 Reduced emotionality in adult and juvenile PS rats In light of previous findings on the effects of PS on emotional behavior in the offspring, specifically reports on contradictory findings of anxiogenic and anxiolytic effects (Estanislau and Morato, 2006; Richardson et al., 2006; Baker et al., 2008; Darnaudery and Maccari, 2008), we first tested the effect of our PS paradigm on offspring behavior. As illustrated in Figure 5.3, adult offspring of dams exposed to PS exhibited reduced anxiety-like behavior as measured in the elevated plus-maze. PS rats at P60 spent significantly more time than controls in the open arms of the plus-maze (Figure 5.3A; 96.5 ± 10.6 s vs. 54.8 ± 6.9 s for controls; Mann-Whitney U test, U32 = 60, ** p 0.01). PS rats also performed more head dips, indicative of reduced anxiety on the plus-maze (Figure 5.3B; 5.3 ± 0.6 vs. 3.3 ± 0.7 dips for controls; U32 = 85, * p 0.05). PS rats did not exhibit any differences in terms of open arm entries (Figure 5.3C; 8.81± 0.6 vs. 7.2 ± 0.7 for controls; U32 = 96.5, p = 0.104) or locomotor activity, with a comparable number of center crossings for each group (Figure 5.3D; 11.9 ± 0.8 vs. 10.2 ± 1.0 crossings for controls; Student’s t-test, t32 = 1.3, p 0.05).
We also measured isolation-induced ultrasonic vocalization calls (USVs, Figure 5.4) to assess emotional behavior in developing rats. This under-powered experiment suggests USVs were comparable at P7 across groups (Figure 5.4A; 428.4 ± 63.6 vs. 406.8 ± 49.8 calls by controls; t20 = 0.271, p 0.05; 10 ≤ N ≤ 12), but tended to occur less frequently in PS rats compared to controls at P17 (Figure 5.4B; 153.3 ± 38.4 vs. 96.8 ± 23.8 calls in controls; t15 = 1.319, p 0.05; 7 ≤ N ≤ 10). On average, pups made many more USVs at P7 (416.6) than at P17 (120.1).
To test whether the anxiolytic effect of PS reflects reduced emotionality in the adult offspring, we measured sociability and social novelty preference (Figure 5.5). These experiments were also under-powered, but suggest sociability may be reduced in adult offspring of stressed dams. Both PS and control rats tended to spend more time with a novel conspecific than a novel object, but the social preference was weaker in PS rats (Figure 5.5A, C; Control total time interacting with conspecific: 101.8 ± 27.2 s, with object: 20.9 ± 3.1 s, sociability preference index: 0.64 ± 0.06; PS total time interacting with conspecific: 60.6 ± 7.6 s, with object: 22.1 ± 5.4 s, sociability preference index: 0.49 ± 0.11; Two-way ANOVA, interaction effect: F1,14 = 0.90, p 0.05; 3 ≤ N ≤ 5). When control rats were presented with a novel conspecific as well as the now familiar conspecific, they spent more time on average with the novel conspecific (Figure 5.5B, C; Control total time interacting with novel conspecific:51.2 ± 11.4 s, with familiar: 34.4 ± 3.2 s,
social novelty preference index: 0.15 ± 0.15; PS total time interacting with novel conspecific:
24.9 ± 9.3 s, with familiar:22.4 ± 10.6 s, social novelty preference index: 0.15 ± 0.18). Consistent with the sociability test, PS rats spent less time with either conspecific, but did not exhibit any difference from controls in their social novelty preference (Figure 5.5C, Two-way ANOVA, main effect of PS: F1,14 = 4.04, p = 0.06). Together, these behavioral findings suggest the altered emotionality in PS offspring emerges early in development and persists into adulthood.
5.4.2 PS altered intrinsic properties of BLA principal neurons during development Having observed altered emotional behavior in PS offspring early in postnatal development, we assessed the effect of PS on the developmental trajectory of BLA principal neuron electrophysiology. Figure 5.6 illustrates the typical developmental reduction in input resistance and membrane time constant, showing a nearly three-fold reduction in both measures between P10 and P28 (Ehrlich et al., 2012). We found no effect of PS on input resistance (Figure
5.6A; Two-way ANOVA: F1,144 = 1.65, p 0.05) or membrane time constant (Figure 5.6B; F1,142 = 0.07, p 0.05). However, we did find an effect of PS on resting membrane potential (RMP, Figure 5.6C), such that RMP was more hyperpolarized in PS animals across all ages, typically between 1 and 2 mV (F1,148 = 4.63, p 0.05). As expected, there was a significant effect of age, with a RMP in control animals of -55.3 ± 1.0 mV at P10 that became hyperpolarized by P17, reaching the mature value of -62.9 ± 0.8 (Two-way ANOVA, main effect of age: F5,148 = 17.3, p 0.0001).
The effects of PS on features of action potentials (APs) of BLA principal neurons are depicted in Figure 5.7, as we have previously shown APs change profoundly with age in this population (Ehrlich et al., 2012). AP threshold changed with age until around P21, as expected, but PS altered its developmental trajectory, reducing the threshold by 1-2 mV from P10 to P21 (Figure 5.7A, B; Two-way ANOVA: F1,146 = 5.85, p 0.05). AP amplitude increased with age, from around 55 mV at P10 to 70 mV at P21-60, and was greater in PS animals (for example, at P17: 71.4 ± 1.8 mV in PS animals vs. 61.8 ± 2.5 mV in controls), corresponding with the more hyperpolarized threshold (Figure 5.7C; Two-way ANOVA, main effect of stress: F1,149 = 4.62, p 0.05).
We also investigated the fast afterhyperpolarization (fAHP) of the AP, which we have previously reported to emerge during development of BLA principal neurons (Ehrlich et al., 2012). In neurons expressing a fAHP, its peak voltage deflection became more hyperpolarized with age until P28, from around -38 mV at P10 to below -42 mV at P28 and P60 (Figure 5.7A, D). Beginning at P21 the fAHP peak was more hyperpolarized in neurons from PS animals relative to controls, and was approximately 2 mV more hyperpolarized in neurons from PS animals in adulthood (not significant, Two-way ANOVA, main effect of PS: F1,66 = 2.34, p 0.05).
Unlike AP threshold and amplitude, AP kinetics did not change due to PS (Figure 5.7EG). Half-width decreased approximately 40% from around 1.25 ms at P10 to around 0.75 ms at P28 (Figure 5.7E; Two-way ANOVA, main effect of age: F5,149 = 42.18, p 0.0001), but was comparable in PS and control animals (main effect of PS: F1,149 = 3.10, p 0.05). AP 10-90% rise time also decreased with age, around 25% from P10 to P21 (Figure 5.7F; Two-way ANOVA, main effect of age: F5,150 = 25.97, p 0.0001) but exhibited no effect of PS (F1,150 = 2.02, p 0.05). Similarly, AP 90-10% decay time decreased more than two-fold with age, from approximately 1.8 ms at P10 to approximately 0.7 ms at P28 and P60 (Figure 5.7G; Two-way ANOVA, main effect of age: F5,149 = 53.28, p 0.0001). AP decay times tended to be faster at P28 and P60 in PS animals, but the effect was not significant (main effect of PS: F1,149 = 3.09, p 0.05).