Aversive Prenatal Stimulation:
Effects on Behavioral, Biochemical, and Somatic Ontogeny in the Rat SONYA K. SOBRIAN Department of Psychology Carleton University Ottawa, Ontario, Cuiiuda Electric foot shock administered to prcgnant rats altered the ontogeny of spontaneous motor activity in their pups. Prenatally stimulated (PMS) offspring were more active than controls on Days 1-10 but less active during the 3rd postpartum week. The age of peak activity, a major developmental landmark, occurred in PMS pups around 10 days of age; in controls maximum activity was not Seen until the 3rd week. This effect was independent of the gender of the offspring and the timing of the gestational stimulation. Its appearance in both cross-fostered and fostered pups indicated the prenatal origin of the effect. The maturation of spontaneous alternation behavior and several reflexes and the appearance of physical features were not affected by prenatal stimulation. Moreover, both PMS and control groups exhibited an agerelated increase in brain concentrations of norepinephrine, serotonin, and 5-hydroxyindoleacetic acid. These findings indicate that spontaneous motor activity is uniquely sensitive to PMS, and as far as can be determined here, PMS produces no generalized alteration in behavioral and physical ontogeny.
The developing organism is susceptible to changes in its prenatal environment especially, as in mammals, when this environment is provided by the mother. Agents in the environment which affect the mother are potentially capable of influencing the fetus through their effects on the pregnant female (Joffe, 1969). Various environmental manipulations of the pregnant rat such as handling, crowding, conditioned avoidance and electric foot shock, swimming, tilting, and open-field exposure, in addition to a variety of pharmacological agents, have been shown to influence the behavior of her offspring (Archer & Blackman, 1971; Joffe, 1969). These qualitatively different manipulations have generally increased offspring emotionality as indicated primarily by decreased ambulation and increased defecation in the open field (Ader & Belfer, 1962; Doyle & Yule, 1959; Hockman, 1961; Thompson, 1957; Thompson, Watson, & Charlesworth, 1962). Though Joffe (1965) found the acquisition of a 2-way conditioned avoidance habit to be facilitated in the offspring of female rats exposed to avoidance conditioning during pregnancy, the majority of studies have reported the Request reprints from: Dr. S. K. Sobrian, Dept. of Psychology, Green Hall, Princeton University, Princeton, New Jersey 08540, U.S.A. Received for publication 25 July 1975 Revised for publication 4 February 1976 Developmental Psychobiology, 10(1): 41-51 (19 77) 01977 by John Wiley & Sons, Inc.
41
42
SOBRIAN
performance to be inferior to controls in a learning situation (Havlena &. Werboff, 1963; Thompson & Quinby, 1964; Thompson et al., 1962; Young, 1963). With few exceptions (Ader & Deitchman, 1970; Ossenkopp, Kolteck, & Persinger, 1972) most experiments concerned with the exposure of pregnant rats to various treatments have emphasized changes in the adult behavior of the offspring. The immediate consequences of these interventions have been ignored. This experiment was therefore designed to examine the effects of manipulation of the prenatal environment on behavioral, biochemical, and somatic ontogeny under the supposition that the alterations in adult behavior seen in the offspring of prenatally treated mothers might be linked t o disruptions in developing behavioral processes and/or alterations in central nervous system (CNS) maturation. Spontaneous motor activity (SMA) and spontaneous alternation (SA) were used to assess disruptions in behavioral ontogeny following prenatal treatment-exposure of the pregnant female to electric foot shock-as both of these behaviors show distinct developmental patterns (Bronstein, Dworkin, & Bilder, 1974; Campbell, Lytle, & Fibiger, 1969; Campbell & Mabry, 1973; Douglas, Peterson, & Douglas, 1073; Egger, 1973a,b; Pappas, Peters, Sobrian, Blouin, & Drew, 1975; Sobrian, Weltman, & Pappas, 1975) and have been linked to the development of serotonergic, cholinergic, and noradrenergic systems in the brain (Egger, Livesey, & Dawson, 1973; Mabry & Campbell, 1973; Swonger & Rech, 1972; Trimbach, 1972). The development of reflexes, physical features, and body weight were evaluated t o determine the extent t o which somatic systems were affected. Endogenous levels of norepinphrine (NE), serotonin (5-HT), and 5-hydroxyindoleacetic acid (5-HIAA) were used as indices of CNS development. The influence of the timing of the stress as well as gender and the fostering of pups were considered as they can interact with prenatal treatments t o modify the direction and magnitude of the change in the offsprings’ behaviors (Archer & Blackman, 1971).
Method Animals Adult (225-250 g) nulliparous female Wistar rats (Rattus noruegicus) were mated to Wistar males (in the ratio of 3:2) 48 hr after arrival at the laboratory from Bio-Breeding Labs, Ottawa, Canada. Vaginal smears were not used to determine pregnancy as previous research has shown that this procedure 1) increases the number of false pregnancies and 2) decreases the number of live births (Sobrian, 1075). Females were left with the males for 5 days. After this time, the females were removed, randomly assigned t o 1 of 3 experimental conditions, and housed individually in polypropylene nesting cages (44 x 25 x 30 cm) with ad Eibitum water and Purina rat chow. They were maintained on a reversed 12-hr light:12-hr dark cycle with lights off at 0800 hours. Cages were inspected daily for births at 1000, 1400, and 2000 hours. Day 1 was defined as the day of birth and pups tested in the Day 1 condition were a maximum of 14 hours old. All subsequent ages were calculated on this basis. All litters were culled to 10 pups on Day 1. Pups were weaned at 25 days of age following activity testing and subsequently housed in groups of 4 of like-gender in wire mesh colony cages (38 x 35 x 18 cm).
PRENATAL INFLUENCES ON RAT ONTOGENY
43
Stimulation Procedure Fifteen unsignaled, inescapable electric foot shocks (.5 mA for .5 sec) administered during a 30-min test session served as the stimulation. The shock source was a constant current neon grid scrambler unit manufactured by the Carleton University Science Workshop. Shocks were programmed on a variable interval, 2-min schedule. The shock apparatus was a 6 3 x 27 x 26-cm Plexiglas compartment with a stainless steel grid floor. The compartment was housed in a sound-attenuated chamber and was cleaned and sprayed with chemical disinfectant (Lysol, Lehn & Fink Products, New Jersey) between each rat to mask odor trails. A fan, mounted on the outside of the chamber, provided ventilation and a mild masking noise. All females were treated between 1200 and 1600 hours in the 12-hr dark cycle.
Experimental Conditions Early Prenatal Stimulation (EPS): Females assigned to this group were exposed to electric foot shock for 7 consecutive days immediately following removal from the males, i.e., Days 6-12 post Day 1 of Mating. Late Prenatal Stimulation (LPS): Females assigned to this group were exposed to electric foot shock for 7 consecutive days beginning 15 days after mating, e.g., Days 16-22 post Day 1 of Mating. Control (C): These females were removed to nesting cages after mating and were not disturbed after this time. The duration of the mating period (5 days), the treatment period (7 days), and the interval between treatment of EPS and LPS groups were chosen to minimize overlap between treated groups. Females in the EPS group at the end of the treatment period could have been a minimum of 7 and a maximum of 12 days pregnant. Females in the LPS group at the beginning of their treatment could have been a minimum of 11 and a maximum of 16 days pregnant. Therefore, the overlap between the days of gestation during which EPS and LPS groups were stimulated was 1 day. Approximately 63% of the females bred gave birth to viable litters resulting in 17, 15, and 16 litters in the EPS, LPS, and C groups, respectively.
Fostering Procedure Postnatal maternal influences on offspring’s behavior can be separated from prenatal ones by the fostering of litters (Joffe, 1969). At birth (Day l), half of the C, EPS, and LPS litters were cross-fostered to non-stimulated primiparous Wistar females. The remainder of the EPS, LPS, and C litters were fostered to mothers within their respective treatment groups, thus holding the postnatal manipulations involved in introducing pups to a non-biological mother constant for all litters. Pups assigned to Day 1 activity testing were cross-fostered or fostered immediately after being found and allowed to nurse the foster mother at least once before testing. The remainder of the litters were fostered or cross-fostered within 12 hr of being found.
44
SOBRIAN
Offspring Testing Procedures All behavioral observations were made using a repeated measures design and pups were tested during the 12-hr dark cycle. Spontaneous Motor Activity. Six male and 6 female pups from each of the 3 fostered and cross-fostered treatment groups were tested at Days 1, 3, 6, 9, 12, 15, 18, 21, 25, and 30. All animals were tested at each age and each testing group consisted of the offspring from at least 3 litters. Activity was recorded during a 30-min test period with a field capacitance movement sensing device (Griffiths, Chapman, & Campbell, 1967) sufficiently sensitive to record large torsional body movements. The rats were placed in a cardboard box (containing wood shavings) which rested on the sensing coil. The dimensions of the box varied with age of the animal: Days I , 3, and 6, 7.5 x 7.5 x 5 cni; Days 13, a n d 15, 8.7 x 10.0 x 15 cm; Days 18, 21, 25, and 30, 11.0 ~ 9 . 15 0 cm. ~ The entire apparatus was housed in a 64 x 64 x 57-cni sound-attenuated chamber. A fan mountcd outside the box provided ventilation and a low masking noise. A 100-W red floodlight suspended 55 cm above the movement sensor served as the light and heat source. The temperature was maintained at 35-36°C. A thin layer of fresh shavings was placed in the bottom of the box before testing. The device was calibrated t o a predetermined optimal sensitivity before each session such that a moderate body displacement (i.e., vertical or horizontal displacement or whole trunk rotation) triggered the adjustable voltage sensor. Each triggering was scored as 1 activity unit. Reflex and Physical Development. Immediately after activity testing, body weight was recorded. In addition, on Days 9, 12, 15, 18, and 21 eyelid dysjunction and incisor eruption were scored. On Days 9, 1 2 , 15, 18, and 21 the startle response motor response (a jerk of the head and extension of hind limbs) to a n auditory stimulus (snap of a mousetrap placed 15 cm above and behind the rat’s head)-and the free-fall righting response-the turning in mid-air and landing on all 4 paws after being dropped upside down from 35-cm-were also scored. Spontaneous Alternation. Alternation behavior was recorded a t 17,22, 27, 32, and 37 days of age. Six males and 6 females, not previously tested from at least 3 litters within each fostered and cross-fostered treatment condition, were tested repeatedly at all 5 ages. Testing was conducted in a gray plywood T-maze. A sectioned, hinged, Plexiglas lid permitted separate access to the start box, the main alley, and the 2 goal arms. The main alley (1 1.5 x 20.0 x 11.5 cm) was separated from the :;tart box ( 1 1.5 x 15.0 x 11.5 cm) by a black guillotine door. The goal arms ( 1 1.5 x 25.0 x 11.5 cm) had 9-cm diameter holes providing access to them and black plastic guillotine doors separating them from the alley. The maze was housed in a small, windowless room maintained at a constant temperature of 22°C. A shaded 25-W red bulb suspended 25 cm above the choice point of the maze provided the only illumination. Each alternation test consisted of 2 trials. A trial consisted of placing the animal in the start box for 10 sec; the guillotine door was then raised, and the rat given 5 niin in which to enter 1 of the goal arms with all 4 paws. The rat was then left in the chosen arm for 50 sec after which it was removed t o a holding cage for 60 sec. During
PRENATAL INFLUENCES ON RAT ONTOGENY
45
this intertrial-interval the maze was wiped with a solution of disinfectant (Micro-Quat, Economics Laboratory, Toronto) and sprayed with Lysol brand disinfectant to mask odor trails. Alternation behavior was scored twice daily. A 4-hr intertest-interval insured the independence of results from each session (Dember & Fowler, 1958).
Biochemical Determinations A fluorometric assay technique was used to determine the endogenous levels of serotonin (5-HT), 5-hydroxyindoleacetic acid (5-HIM), and norepinephrine (NE) in brain tissue. The assay technique was modified by Peters, Mazurkiewicz-Kwilecki, and Pappas (1974) from the methods of Miller, Cox, Snodgrass, and Maickel (1970) and Curzon and Green (1970). Rats that had not undergone behavioral testing were sacrificed by decapitation on Days 1, 12, or 25. The brains were quickly removed and dissected on an ice-cooled, saline-rinsed plate. Whole brains were first divided in half by a sagittal section along the midline suture. The cerebra and cerebella were removed from one half and the remaining tissue, which included diencephalic, mesencephalic, and rhombencephalic structures, was used for biochemical analysis. Tissue from 12 pups from each of the 6 groups was assayed at each age, with an equal number of males and females in each group.
Stat istical Analysis Analyses of variance, 2 x 2 x 3 x 10 (with 1 repeated measure) for Gender, Fostering, Treatment, and Age, were used t o analyze neonatal activity and body weight data. Conservative degrees of freedom were used to evaluate the significance of the F statistic computed for the repeated measure (Age) and all interactions containing this factor. The Newman-Keuls statistic was used for post-hoe comparisons of significant main effects. Simple main effects analyses were computed for significant interactions (Kirk, 1968). Lord (1953) indicated that nominal data which conform to certain distribution characteristics can be analyzed using an analysis of variance procedure without violating the assumptions of this statistic. Consequently, ‘(no’’ and “yes” scores obtained for reflex and physical features development, were assigned numerical values of “0” and “1 ,” respectively, and frequency distributions were determined. Data from each of the groups exhibited an L-shaped distribution, indicating that an analysis of variance was appropriate. A 2 x 2 x 3 x 5 analysis (with 1 repeated measure) for Gender, Fostering, Treatment, and Age was then computed. Spontaneous alternation data were first analyzed for Trial 1 position preference in order to determine the expected probability of alternation on subsequent trials for each treatment group (Dember & Fowler, 1958). This probability level was calculated using the formula I-( [percent responses to preferred side] + [percent responses to non-preferred side] ’). This value was then used in the x2 statistic to determine whether alternation behavior was statistically different from chance at a particular age. The within-subject design necessitated the use of a statistic other than x’ to determine the pattern of SA over the 5 test days. Therefore, SA data for each subject
’
46
SOBRlAN
for each day were given a score of “0,” “1 ,” or “2” (0 = no SA on a given day, 1 = SA on either the 1st or 2nd test, 2 = SA on both tests). Data were then analyzed using 3 2 x 3 x 5 split-plot factorial analysis of variance for Fostering, Treatment, and Age. Biochemical data were subjected t o a multivariate analysis of variance (Cooley & Lohnes, 1971 ; Kramer & Jensen, 1969). Univariate analysis of variance were performed on each variable only If the multivariate F statistic was significant (Hummel & Shgo, 1971).
Resuit s Body Weight Neither the gender of the offspring nor the postnatal rearing conditions (fostered vs cross-fostered) affected the significant increase in body weight seen in all groups between 1 and 30 days of age ( F = 373.82, df= 1/60, p < .0001). A simple main effects analysis of the significant Age x Stimulation interaction ( F = 3.62, df = 2/30, p < .OS) indicated that C groups were significantly lighter ( p ’ s < .05) than both the EPS and LPS groups at 6 days of age (EPS: 12.7 k .49 g; LPS: 12.6 2 .59 g; C: 10.7 f .60 g). At 21 days of age, LPS groups were heavier (p < .05) than the other 2 groups (EPS: 32.9 .82 g; LPS: 34.1 f .49 g; C: 32.2 f .78 g). At 30 days of age, C pups were significantly heavier than EPS (p < .05) but not LPS offspring (EPS: 70.95 k 1.6 g; LPS: 75.6 5 1.8 g; C: 75.9 1.4g). _+
_+
Spontaneous Motor Activity The initial statistical examination of the data failed t o reveal any main or interactive effects due t o gender. Thus, this variable was collapsed in subsequent statistical analyses of SMA data. As both the Age x Stress x Fostering ( F = 3.70, @ = 2/60, p < .05) and the Age x Stress ( F = 13.85, df= 2/60, p < .OOl) interactions were significant, data from the 3 fostered and the 3 cross-fostered groups were analyzed separately to evaluate PMS effects o n the development of SMA. Fostered Pups. The analysis of variance on 1-30 day-old mean total activity data revealed a significant main effect of Age and an Age x Stress interaction ( F = 6.98, df = 1/33, p < .05; F = 3.36, df’= 2/33, p < .05, respectively; see Fig. 1). A simple main effects analysis of the interaction supplemented by Newman-Keuls tests indicated that the activity of pups in the EPS and LPS groups increased and remained above the C pups during the first 2 weeks postpartum (p < .05 and < .01, respectively) after which it gradually declined. I n contrast, the activity of the C offspring remained relatively unchanged between 3 and 15 days of age, after which time it rapidly increased and then sharply declined. Activity levels at 25 and 30 days of age were similar for all 3 groups. The results of trend analyses (Kirk, 1968) indicated that 2nd degree equations were appropriate and they were generated for each curve and the age at which maximum activity occurred was determined using the formula y’ = (-b/2a). The results of the curve-fitting procedure indicated that maximum activity occurred in the fostered EPS and LPS pups at 10.1 and 10.0 days of age, respectively, but not until 16.8 days of age in the fostered C pups.
PRENATAL INFLUENCES ON RAT ONTOGENY
90@
-
c-.E M - F LFS-F C-F
-
600
47
-
300
''
-1
d
12@0I-
c
EPS-CF +-dLPS-CF
*-+C-CF
I' I , 900.
600
300
.
1
1
1
1
I
I
3
6
9
I
I
12
15
' 18
a " " '
21
25
30
t Weaned
AGE
(DAYS)
Fig. 1. Mean activity scores during a 30-rnin test session for Fostered (upper panel) and Cross-fostered (lower panel) EPS, LPS, and C pups at 1-30 days of age. Data shown are collapsed across gender.
48
SOBRIAN
Cross-Fostered Pups. An analysis o f variance on neonatal activity data revealed a significant main effect of Age and the Age x Stress interaction ( F = 10.96, df= 1/33, p < .001; F = 17.406, df 2/33, p < .001, respectively). Examination of the interaction term indicated that the activity of the C pups was low during the first 2 weeks after birth, increasing sharply at 21 days of age, and declining between 21 and 25 days of age. Although the activity in both of the PMS groups was significantly (p < . O l ) elevated over C levels in the period between 1 and 15 days of age, peak activity appeared to occur a t slightly different ages in these 2 groups (Fig. 1). The results of trend analysis and curve-fitting procedures confirm this observation. The EPS and LPS PUPS exhibited maximum activity levels at 8.6 and 7.7 days of age, respectively-substantially earlier than the 19.3 days of age observed in the fostered C pups. At 25 and 30 days of age, activity levels were not significantly different. Cross-Fostering us Fostering. The effect of cross-fostering pups to non,-stimulated females on the development of SMA was evident only in a complex interaction involving Stress and Age ( F = 13.85, df = 2/60, p < .OOl). Cross-fostering delayed the age at which peak activity occurred in the C group (F=4.96, df = 1/22, p < .05). Cross-fostered and fostered EPS and LPS groups did not differ. Timing of Stimulation. The gestational period during which feiiiales were treated did not significantly influence the response of the offspring. Both early gestational stimulation and late gestational stimulation were equally effective in adtering the development of SMA in both cross-fostered ( F = 17.41, df = 2/33, p
Effects on Behavioral, Biochemical, and Somatic Ontogeny in the Rat SONYA K. SOBRIAN Department of Psychology Carleton University Ottawa, Ontario, Cuiiuda Electric foot shock administered to prcgnant rats altered the ontogeny of spontaneous motor activity in their pups. Prenatally stimulated (PMS) offspring were more active than controls on Days 1-10 but less active during the 3rd postpartum week. The age of peak activity, a major developmental landmark, occurred in PMS pups around 10 days of age; in controls maximum activity was not Seen until the 3rd week. This effect was independent of the gender of the offspring and the timing of the gestational stimulation. Its appearance in both cross-fostered and fostered pups indicated the prenatal origin of the effect. The maturation of spontaneous alternation behavior and several reflexes and the appearance of physical features were not affected by prenatal stimulation. Moreover, both PMS and control groups exhibited an agerelated increase in brain concentrations of norepinephrine, serotonin, and 5-hydroxyindoleacetic acid. These findings indicate that spontaneous motor activity is uniquely sensitive to PMS, and as far as can be determined here, PMS produces no generalized alteration in behavioral and physical ontogeny.
The developing organism is susceptible to changes in its prenatal environment especially, as in mammals, when this environment is provided by the mother. Agents in the environment which affect the mother are potentially capable of influencing the fetus through their effects on the pregnant female (Joffe, 1969). Various environmental manipulations of the pregnant rat such as handling, crowding, conditioned avoidance and electric foot shock, swimming, tilting, and open-field exposure, in addition to a variety of pharmacological agents, have been shown to influence the behavior of her offspring (Archer & Blackman, 1971; Joffe, 1969). These qualitatively different manipulations have generally increased offspring emotionality as indicated primarily by decreased ambulation and increased defecation in the open field (Ader & Belfer, 1962; Doyle & Yule, 1959; Hockman, 1961; Thompson, 1957; Thompson, Watson, & Charlesworth, 1962). Though Joffe (1965) found the acquisition of a 2-way conditioned avoidance habit to be facilitated in the offspring of female rats exposed to avoidance conditioning during pregnancy, the majority of studies have reported the Request reprints from: Dr. S. K. Sobrian, Dept. of Psychology, Green Hall, Princeton University, Princeton, New Jersey 08540, U.S.A. Received for publication 25 July 1975 Revised for publication 4 February 1976 Developmental Psychobiology, 10(1): 41-51 (19 77) 01977 by John Wiley & Sons, Inc.
41
42
SOBRIAN
performance to be inferior to controls in a learning situation (Havlena &. Werboff, 1963; Thompson & Quinby, 1964; Thompson et al., 1962; Young, 1963). With few exceptions (Ader & Deitchman, 1970; Ossenkopp, Kolteck, & Persinger, 1972) most experiments concerned with the exposure of pregnant rats to various treatments have emphasized changes in the adult behavior of the offspring. The immediate consequences of these interventions have been ignored. This experiment was therefore designed to examine the effects of manipulation of the prenatal environment on behavioral, biochemical, and somatic ontogeny under the supposition that the alterations in adult behavior seen in the offspring of prenatally treated mothers might be linked t o disruptions in developing behavioral processes and/or alterations in central nervous system (CNS) maturation. Spontaneous motor activity (SMA) and spontaneous alternation (SA) were used to assess disruptions in behavioral ontogeny following prenatal treatment-exposure of the pregnant female to electric foot shock-as both of these behaviors show distinct developmental patterns (Bronstein, Dworkin, & Bilder, 1974; Campbell, Lytle, & Fibiger, 1969; Campbell & Mabry, 1973; Douglas, Peterson, & Douglas, 1073; Egger, 1973a,b; Pappas, Peters, Sobrian, Blouin, & Drew, 1975; Sobrian, Weltman, & Pappas, 1975) and have been linked to the development of serotonergic, cholinergic, and noradrenergic systems in the brain (Egger, Livesey, & Dawson, 1973; Mabry & Campbell, 1973; Swonger & Rech, 1972; Trimbach, 1972). The development of reflexes, physical features, and body weight were evaluated t o determine the extent t o which somatic systems were affected. Endogenous levels of norepinphrine (NE), serotonin (5-HT), and 5-hydroxyindoleacetic acid (5-HIAA) were used as indices of CNS development. The influence of the timing of the stress as well as gender and the fostering of pups were considered as they can interact with prenatal treatments t o modify the direction and magnitude of the change in the offsprings’ behaviors (Archer & Blackman, 1971).
Method Animals Adult (225-250 g) nulliparous female Wistar rats (Rattus noruegicus) were mated to Wistar males (in the ratio of 3:2) 48 hr after arrival at the laboratory from Bio-Breeding Labs, Ottawa, Canada. Vaginal smears were not used to determine pregnancy as previous research has shown that this procedure 1) increases the number of false pregnancies and 2) decreases the number of live births (Sobrian, 1075). Females were left with the males for 5 days. After this time, the females were removed, randomly assigned t o 1 of 3 experimental conditions, and housed individually in polypropylene nesting cages (44 x 25 x 30 cm) with ad Eibitum water and Purina rat chow. They were maintained on a reversed 12-hr light:12-hr dark cycle with lights off at 0800 hours. Cages were inspected daily for births at 1000, 1400, and 2000 hours. Day 1 was defined as the day of birth and pups tested in the Day 1 condition were a maximum of 14 hours old. All subsequent ages were calculated on this basis. All litters were culled to 10 pups on Day 1. Pups were weaned at 25 days of age following activity testing and subsequently housed in groups of 4 of like-gender in wire mesh colony cages (38 x 35 x 18 cm).
PRENATAL INFLUENCES ON RAT ONTOGENY
43
Stimulation Procedure Fifteen unsignaled, inescapable electric foot shocks (.5 mA for .5 sec) administered during a 30-min test session served as the stimulation. The shock source was a constant current neon grid scrambler unit manufactured by the Carleton University Science Workshop. Shocks were programmed on a variable interval, 2-min schedule. The shock apparatus was a 6 3 x 27 x 26-cm Plexiglas compartment with a stainless steel grid floor. The compartment was housed in a sound-attenuated chamber and was cleaned and sprayed with chemical disinfectant (Lysol, Lehn & Fink Products, New Jersey) between each rat to mask odor trails. A fan, mounted on the outside of the chamber, provided ventilation and a mild masking noise. All females were treated between 1200 and 1600 hours in the 12-hr dark cycle.
Experimental Conditions Early Prenatal Stimulation (EPS): Females assigned to this group were exposed to electric foot shock for 7 consecutive days immediately following removal from the males, i.e., Days 6-12 post Day 1 of Mating. Late Prenatal Stimulation (LPS): Females assigned to this group were exposed to electric foot shock for 7 consecutive days beginning 15 days after mating, e.g., Days 16-22 post Day 1 of Mating. Control (C): These females were removed to nesting cages after mating and were not disturbed after this time. The duration of the mating period (5 days), the treatment period (7 days), and the interval between treatment of EPS and LPS groups were chosen to minimize overlap between treated groups. Females in the EPS group at the end of the treatment period could have been a minimum of 7 and a maximum of 12 days pregnant. Females in the LPS group at the beginning of their treatment could have been a minimum of 11 and a maximum of 16 days pregnant. Therefore, the overlap between the days of gestation during which EPS and LPS groups were stimulated was 1 day. Approximately 63% of the females bred gave birth to viable litters resulting in 17, 15, and 16 litters in the EPS, LPS, and C groups, respectively.
Fostering Procedure Postnatal maternal influences on offspring’s behavior can be separated from prenatal ones by the fostering of litters (Joffe, 1969). At birth (Day l), half of the C, EPS, and LPS litters were cross-fostered to non-stimulated primiparous Wistar females. The remainder of the EPS, LPS, and C litters were fostered to mothers within their respective treatment groups, thus holding the postnatal manipulations involved in introducing pups to a non-biological mother constant for all litters. Pups assigned to Day 1 activity testing were cross-fostered or fostered immediately after being found and allowed to nurse the foster mother at least once before testing. The remainder of the litters were fostered or cross-fostered within 12 hr of being found.
44
SOBRIAN
Offspring Testing Procedures All behavioral observations were made using a repeated measures design and pups were tested during the 12-hr dark cycle. Spontaneous Motor Activity. Six male and 6 female pups from each of the 3 fostered and cross-fostered treatment groups were tested at Days 1, 3, 6, 9, 12, 15, 18, 21, 25, and 30. All animals were tested at each age and each testing group consisted of the offspring from at least 3 litters. Activity was recorded during a 30-min test period with a field capacitance movement sensing device (Griffiths, Chapman, & Campbell, 1967) sufficiently sensitive to record large torsional body movements. The rats were placed in a cardboard box (containing wood shavings) which rested on the sensing coil. The dimensions of the box varied with age of the animal: Days I , 3, and 6, 7.5 x 7.5 x 5 cni; Days 13, a n d 15, 8.7 x 10.0 x 15 cm; Days 18, 21, 25, and 30, 11.0 ~ 9 . 15 0 cm. ~ The entire apparatus was housed in a 64 x 64 x 57-cni sound-attenuated chamber. A fan mountcd outside the box provided ventilation and a low masking noise. A 100-W red floodlight suspended 55 cm above the movement sensor served as the light and heat source. The temperature was maintained at 35-36°C. A thin layer of fresh shavings was placed in the bottom of the box before testing. The device was calibrated t o a predetermined optimal sensitivity before each session such that a moderate body displacement (i.e., vertical or horizontal displacement or whole trunk rotation) triggered the adjustable voltage sensor. Each triggering was scored as 1 activity unit. Reflex and Physical Development. Immediately after activity testing, body weight was recorded. In addition, on Days 9, 12, 15, 18, and 21 eyelid dysjunction and incisor eruption were scored. On Days 9, 1 2 , 15, 18, and 21 the startle response motor response (a jerk of the head and extension of hind limbs) to a n auditory stimulus (snap of a mousetrap placed 15 cm above and behind the rat’s head)-and the free-fall righting response-the turning in mid-air and landing on all 4 paws after being dropped upside down from 35-cm-were also scored. Spontaneous Alternation. Alternation behavior was recorded a t 17,22, 27, 32, and 37 days of age. Six males and 6 females, not previously tested from at least 3 litters within each fostered and cross-fostered treatment condition, were tested repeatedly at all 5 ages. Testing was conducted in a gray plywood T-maze. A sectioned, hinged, Plexiglas lid permitted separate access to the start box, the main alley, and the 2 goal arms. The main alley (1 1.5 x 20.0 x 11.5 cm) was separated from the :;tart box ( 1 1.5 x 15.0 x 11.5 cm) by a black guillotine door. The goal arms ( 1 1.5 x 25.0 x 11.5 cm) had 9-cm diameter holes providing access to them and black plastic guillotine doors separating them from the alley. The maze was housed in a small, windowless room maintained at a constant temperature of 22°C. A shaded 25-W red bulb suspended 25 cm above the choice point of the maze provided the only illumination. Each alternation test consisted of 2 trials. A trial consisted of placing the animal in the start box for 10 sec; the guillotine door was then raised, and the rat given 5 niin in which to enter 1 of the goal arms with all 4 paws. The rat was then left in the chosen arm for 50 sec after which it was removed t o a holding cage for 60 sec. During
PRENATAL INFLUENCES ON RAT ONTOGENY
45
this intertrial-interval the maze was wiped with a solution of disinfectant (Micro-Quat, Economics Laboratory, Toronto) and sprayed with Lysol brand disinfectant to mask odor trails. Alternation behavior was scored twice daily. A 4-hr intertest-interval insured the independence of results from each session (Dember & Fowler, 1958).
Biochemical Determinations A fluorometric assay technique was used to determine the endogenous levels of serotonin (5-HT), 5-hydroxyindoleacetic acid (5-HIM), and norepinephrine (NE) in brain tissue. The assay technique was modified by Peters, Mazurkiewicz-Kwilecki, and Pappas (1974) from the methods of Miller, Cox, Snodgrass, and Maickel (1970) and Curzon and Green (1970). Rats that had not undergone behavioral testing were sacrificed by decapitation on Days 1, 12, or 25. The brains were quickly removed and dissected on an ice-cooled, saline-rinsed plate. Whole brains were first divided in half by a sagittal section along the midline suture. The cerebra and cerebella were removed from one half and the remaining tissue, which included diencephalic, mesencephalic, and rhombencephalic structures, was used for biochemical analysis. Tissue from 12 pups from each of the 6 groups was assayed at each age, with an equal number of males and females in each group.
Stat istical Analysis Analyses of variance, 2 x 2 x 3 x 10 (with 1 repeated measure) for Gender, Fostering, Treatment, and Age, were used t o analyze neonatal activity and body weight data. Conservative degrees of freedom were used to evaluate the significance of the F statistic computed for the repeated measure (Age) and all interactions containing this factor. The Newman-Keuls statistic was used for post-hoe comparisons of significant main effects. Simple main effects analyses were computed for significant interactions (Kirk, 1968). Lord (1953) indicated that nominal data which conform to certain distribution characteristics can be analyzed using an analysis of variance procedure without violating the assumptions of this statistic. Consequently, ‘(no’’ and “yes” scores obtained for reflex and physical features development, were assigned numerical values of “0” and “1 ,” respectively, and frequency distributions were determined. Data from each of the groups exhibited an L-shaped distribution, indicating that an analysis of variance was appropriate. A 2 x 2 x 3 x 5 analysis (with 1 repeated measure) for Gender, Fostering, Treatment, and Age was then computed. Spontaneous alternation data were first analyzed for Trial 1 position preference in order to determine the expected probability of alternation on subsequent trials for each treatment group (Dember & Fowler, 1958). This probability level was calculated using the formula I-( [percent responses to preferred side] + [percent responses to non-preferred side] ’). This value was then used in the x2 statistic to determine whether alternation behavior was statistically different from chance at a particular age. The within-subject design necessitated the use of a statistic other than x’ to determine the pattern of SA over the 5 test days. Therefore, SA data for each subject
’
46
SOBRlAN
for each day were given a score of “0,” “1 ,” or “2” (0 = no SA on a given day, 1 = SA on either the 1st or 2nd test, 2 = SA on both tests). Data were then analyzed using 3 2 x 3 x 5 split-plot factorial analysis of variance for Fostering, Treatment, and Age. Biochemical data were subjected t o a multivariate analysis of variance (Cooley & Lohnes, 1971 ; Kramer & Jensen, 1969). Univariate analysis of variance were performed on each variable only If the multivariate F statistic was significant (Hummel & Shgo, 1971).
Resuit s Body Weight Neither the gender of the offspring nor the postnatal rearing conditions (fostered vs cross-fostered) affected the significant increase in body weight seen in all groups between 1 and 30 days of age ( F = 373.82, df= 1/60, p < .0001). A simple main effects analysis of the significant Age x Stimulation interaction ( F = 3.62, df = 2/30, p < .OS) indicated that C groups were significantly lighter ( p ’ s < .05) than both the EPS and LPS groups at 6 days of age (EPS: 12.7 k .49 g; LPS: 12.6 2 .59 g; C: 10.7 f .60 g). At 21 days of age, LPS groups were heavier (p < .05) than the other 2 groups (EPS: 32.9 .82 g; LPS: 34.1 f .49 g; C: 32.2 f .78 g). At 30 days of age, C pups were significantly heavier than EPS (p < .05) but not LPS offspring (EPS: 70.95 k 1.6 g; LPS: 75.6 5 1.8 g; C: 75.9 1.4g). _+
_+
Spontaneous Motor Activity The initial statistical examination of the data failed t o reveal any main or interactive effects due t o gender. Thus, this variable was collapsed in subsequent statistical analyses of SMA data. As both the Age x Stress x Fostering ( F = 3.70, @ = 2/60, p < .05) and the Age x Stress ( F = 13.85, df= 2/60, p < .OOl) interactions were significant, data from the 3 fostered and the 3 cross-fostered groups were analyzed separately to evaluate PMS effects o n the development of SMA. Fostered Pups. The analysis of variance on 1-30 day-old mean total activity data revealed a significant main effect of Age and an Age x Stress interaction ( F = 6.98, df = 1/33, p < .05; F = 3.36, df’= 2/33, p < .05, respectively; see Fig. 1). A simple main effects analysis of the interaction supplemented by Newman-Keuls tests indicated that the activity of pups in the EPS and LPS groups increased and remained above the C pups during the first 2 weeks postpartum (p < .05 and < .01, respectively) after which it gradually declined. I n contrast, the activity of the C offspring remained relatively unchanged between 3 and 15 days of age, after which time it rapidly increased and then sharply declined. Activity levels at 25 and 30 days of age were similar for all 3 groups. The results of trend analyses (Kirk, 1968) indicated that 2nd degree equations were appropriate and they were generated for each curve and the age at which maximum activity occurred was determined using the formula y’ = (-b/2a). The results of the curve-fitting procedure indicated that maximum activity occurred in the fostered EPS and LPS pups at 10.1 and 10.0 days of age, respectively, but not until 16.8 days of age in the fostered C pups.
PRENATAL INFLUENCES ON RAT ONTOGENY
90@
-
c-.E M - F LFS-F C-F
-
600
47
-
300
''
-1
d
12@0I-
c
EPS-CF +-dLPS-CF
*-+C-CF
I' I , 900.
600
300
.
1
1
1
1
I
I
3
6
9
I
I
12
15
' 18
a " " '
21
25
30
t Weaned
AGE
(DAYS)
Fig. 1. Mean activity scores during a 30-rnin test session for Fostered (upper panel) and Cross-fostered (lower panel) EPS, LPS, and C pups at 1-30 days of age. Data shown are collapsed across gender.
48
SOBRIAN
Cross-Fostered Pups. An analysis o f variance on neonatal activity data revealed a significant main effect of Age and the Age x Stress interaction ( F = 10.96, df= 1/33, p < .001; F = 17.406, df 2/33, p < .001, respectively). Examination of the interaction term indicated that the activity of the C pups was low during the first 2 weeks after birth, increasing sharply at 21 days of age, and declining between 21 and 25 days of age. Although the activity in both of the PMS groups was significantly (p < . O l ) elevated over C levels in the period between 1 and 15 days of age, peak activity appeared to occur a t slightly different ages in these 2 groups (Fig. 1). The results of trend analysis and curve-fitting procedures confirm this observation. The EPS and LPS PUPS exhibited maximum activity levels at 8.6 and 7.7 days of age, respectively-substantially earlier than the 19.3 days of age observed in the fostered C pups. At 25 and 30 days of age, activity levels were not significantly different. Cross-Fostering us Fostering. The effect of cross-fostering pups to non,-stimulated females on the development of SMA was evident only in a complex interaction involving Stress and Age ( F = 13.85, df = 2/60, p < .OOl). Cross-fostering delayed the age at which peak activity occurred in the C group (F=4.96, df = 1/22, p < .05). Cross-fostered and fostered EPS and LPS groups did not differ. Timing of Stimulation. The gestational period during which feiiiales were treated did not significantly influence the response of the offspring. Both early gestational stimulation and late gestational stimulation were equally effective in adtering the development of SMA in both cross-fostered ( F = 17.41, df = 2/33, p
