Steroid hormones activate nuclear receptors, which as transcription factors, can exert powerful effects on neurodevelopmental processes to create permanent changes in neural connectivity and synaptic function. While it is well established that steroid hormones influence the development of brain regions that regulate adult reproductive behavior or neuroendocrine function, evidence indicates that steroid hormones can also profoundly alter connectivity and function in regions traditionally associated with complex cognitive function (reviewed in Juraska, 1991) such as the hippocampus and neocortex. Manipulations of gonadal hormones have been shown to affect connectivity in these regions both acutely in adulthood (e.g. Gould et al., 1990; Wooley et al., 1990; Reyna-Neyra et al., 2002; Choi et al., 2003; Kinsley et al., 2006; Chen et al., 2009; Camacho-Arroyo et al., 2011) and during early development (Munoz-Cueto et al., 1990, Munoz-Cueto & Ruiz-Marcos, 1994; Venkatesan & Kritzer, 1999; Isgor & Sengelaub, 2003; Nuñez et al., 2003; Grobin et al., 2006; Martìnez-Cerdeño et al., 2006). Steroid hormones are thought to play an organizational role in development of the mammalian cortex by activating nuclear receptors that are expressed during critical periods of development (Munoz-Cueto et al., 1990; Lopez & Wagner, 2009). Specifically, steroid hormones alter connectivity patterns (Venkatesan & Kritzer, 1999; Grobin et al., 2006), synapse formation (Munoz-Cueto & Ruiz-Marcos, 1994), and cell number (Reid & Juraska, 1992). Still, little is known about the complex role that steroid hormones may play in the development of the neocortex and cortically-mediated behaviors.
The Progesterone Receptor (PR) expression is especially abundant in the developing rat neocortex (Hagirhara et al., 1992; Kato et al., 1993; Lopez & Wagner, 2009). In the rat, PR expression is first detected at embryonic day 18 (E18) in the subplate, a developmentally critical structure (Grossberg & Seitz, 2003; Kanold & Shatz, 2006). PR expression in the subplate peaks on postnatal day 2 (P2) and is absent by P14 (Jahagirdar & Wagner, 2009; Lopez & Wagner, 2009; Quadros et al., 2007), a developmental period corresponding with the arrival of thalamic afferents to the cortex (Kichula & Huntley, 2008). Beginning at P3, PR is expressed in layer V, followed by layers II/III by P6, with peak expression in these layers occurring between P7 and P10 (Quadros et al., 2007; Lopez & Wagner, 2009). Subsequently, PR expression gradually declines and is virtually undetectable by P27. The neurons of layer V and II/III are primarily excitatory pyramidal cells with intracortical callosal and subcortical projections. PR is expressed in neurons, as it is predominately co-expressed with the neuronal marker, MAP-2, and is not colocalized with glial-specific markers (Lopez & Wagner, 2009). PR is not expressed in GABAergic cells, but is expressed in cells with distinct pyramidal cell morphology as indicated by MAP2-ir (unpubl. obs.) These observations suggest that PR activity within subplate neurons and excitatory pyramidal cells during perinatal development may be critical for the normal establishment and maturation of cortical connectivity in rodents.
Consistent with this idea, the administration of progesterone to pregnant rats early in gestation increased dendritic branching and spine density in somatosensory pyramidal cells (Menzies et al., 1982) and neonatal administration of exogenous progesterone improved performance on learning and memory tasks in adulthood (Snyder & Hull, 1980). Given this, it is possible that perturbations in progesterone levels and/or in PR expression or activity could result in detrimental effects on the proper development of neocortical connectivity, which could in turn, lead to deficits in cortically-dependent functions. To directly address this question, we turned to the PR knockout (PRKO) mouse, which lack functional expression of both PRA and PRB isoforms. Because preliminary observations suggest that PR is predominately expressed in the somatosensory cortex in perinatal mice, we tested the hypothesis that functional expression of PR is required for the normal maturation of sensorimotor function. We first conducted immunocytochemistry for PR to determine the extent and precise timing of PR expression in the neonatal mouse somatosensory cortex. We then examined an array of early behavioral milestones that measure basic somatosensory-dependent motor reflexes and responses and that serve as indicators of early nervous system development (Glynn et al., 2007). To our knowledge, this study represents the first comprehensive characterization of neurological and behavioral development in a steroid receptor knockout mouse model.
Mice used in this study were of a mixed background (C129SvEv X C57Bl), and were either homozygous for an insertional mutation in the PR gene (PRKO), were homozygous for the wildtype PR gene (WT) or heterozygous (HZ). PRKO mice were created by the insertion of a neomycin resistance gene and a lacZ reporter gene into Exon 1, thereby preventing PR transcription and inhibiting functional mRNA expression for both PR isoforms A and B (Lydon et al., 1995). Animals were housed on a reverse 12-h light, 12-h dark cycle at a constant temperature of 25 ± 2°C, with food and water available ad libitum. All pups used for testing were generated by crossing HZ males with HZ females. Pregnant females were allowed to deliver their pups normally. Animals were marked on the abdomen or tail with permanent marker for the identification of individual animals from day to day. Male and female pups were tested daily from the day of birth (P1) through P14 and were sacrificed after testing on P14. All animal procedures used in this study were approved by the Institutional Animal Care and Use Committee at the University at Albany.
Determination of Genotype
Following behavioral testing on P14, animals were euthanized with sodium pentobarbital (60 mg/kg, i.p.). Tail tissue was collected and total genomic DNA was purified and amplified using the Qiagen DNeasy Blood and Tissue kit. PCR was performed on purified DNA using three separate primers for PR and Neomycin (PR: 5′-TAGACAGTGTCTTAGACTCGTTGTTG-3′; 5′-GATGGGCACATGGATGAAAC-3′; Neomycin: 5′- GCATGCTCCAGACTGCCTTGGGAAA-3′). Using gel electrophoresis, genotype was determined by the molecular weight of the extracted PCR product. Genotyping revealed that of the 38 animals tested, 10 were KO (5 male), 9 were WT (5 male), and 19 were HZ (11 male). Members of each genotype were distributed across litters such that no one litter contained more than 3 WT or KO pups, or more than 4 HZ animals. Additionally, each genotype was represented in at least 5 of the 7 litters tested.
Tissue Collection and Immunocytochemistry
Tissue from additional WT male and female pups was collected at P2 and P7, and processed for immunocytochemistry for PR. For each age group, three male and three female WT pups were collected from a total of eight different litters. Brains were fixed in 5% acrolein for 6 hours, followed by three-day incubation in 30% sucrose. Coronal sections were cut at 50 um on a freezing microtome, preserved in cryoprotectant and processed for immunocytochemistry for PR. For identification of cortical layers, alternate sections were stained with Cresyl Violet, as described in Lopez & Wagner (2009).
All incubations were conducted at room temperature unless specified otherwise. Sections were rinsed 3 × 5 min in 0.05 M tris-buffered saline (TBS; pH 7.6), then incubated in 1% sodium borohydride in TBS for 10 min. Following five rinses in TBS (5 min each), tissue was blocked in TBS containing 20% normal goat serum, 1% hydrogen peroxide, and 1% bovine serum albumin for 30 min. Sections were then incubated in TBS containing 0.3% triton-X-100, 2% NGS, and PR antiserum (1:1000; rabbit-polyclonal, DAKO) for 72 hours. After primary incubation, tissue was rinsed 3 × 5 min in TBS, then incubated in TBS containing 5 ug/mL biotinylated goat, anti-rabbit IgG secondary antibody for 90 min (Vector Labs), followed by two rinses in TBS-containing 2% NGS and 0.3% triton-X-100 and two rinses in TBS alone (5 min each). Sections were then incubated in TBS containing avidin–biotin complex reagent (Vectastain Elite kit, Vector Labs). After three rinses in TBS, sections were incubated in TBS-containing 0.05% diaminobenzadine, 0.75 nM nickel ammonium sulfate, 0.15% Beta-D Glucose, 0.04% ammonium chloride, and 0.001% glucose oxidase for 10 minutes. Sections were then rinsed five times (5 min each) in TBS, mounted on gelatin-coated slides and coverslipped with Permount (Fisher Scientific).
Qualitative observations on the abundance and distribution of PRir cells in specific layers of somatosensory cortex were made in male and female animals from P2 and P7. In accordance with previous findings in the rat (Lopez & Wagner, 2009), qualitative assessment revealed no observable sex differences in PR expression or distribution. One representative section was chosen from an animal at each age group and line drawings of the distribution of individual PRir nuclei in subplate, layer V and layers II/III of the somatosensory cortex were prepared as follows. Digital images of sections were captured with a Model 1.3.0 SPOT digital camera (Diagnostic Instruments) attached to a Nikon E600 microscope using a 20x objective. Visualization of alternate Crystal Violate-stained sections was used to differentiate distinct cortical layers (Fig. 1). Somatosensory cortex was identified as described in Lopez & Wagner (2009), using a rodent brain atlas for reference points (Paxinos & Watson, 1998). Using Adobe Photoshop, line drawings were superimposed over photomicrographs of select somatosensory cortex sections that contained PRir. Black dots were used to indicate the presence of an individual cell nucleus that contained PR.
Behavioral Testing Procedures
All behavioral testing was performed between 1300 and 1700 hr under red lighting conditions, by an experimenter blind to genotype. On the day of birth (P1), testing commenced several hours after the pups had been fed, as measured by the appearance of the milk band. Testing was conducted daily from P1 through P14 unless otherwise indicated. The majority of the tests performed were adapted from Glynn et al. (2007) in their evaluation of sensory and motor development in complexin1 knockout mice. The tasks utilized in this study were designed to test basic developmental milestones (eye-opening and ear-twitching), sensory reflex behavior (righting, rooting, clasping, cliff aversion and negative geotaxis), and motor behavior (crawling, activity level, homing). The order of the test administration was as listed below and was the same for each animal. All animals were weighed after testing each day. The entire testing procedure lasted approximately 10 minutes.
The outer ear was gently stroked with a cotton swab and the presence or absence of a twitch of the ear was noted. This test was performed daily from P1 to P14.
Pups were examined daily and the first day both eyes were fully open was recorded. Once an individual animal's eyes were opened, examination for this measure was not continued.
Pups were placed in a plastic box (open on top) with absorbent filter paper marked with grid lines covering the floor. Crawling was defined as lifting the front part of the body and moving using both forelimbs and hindlimbs. From P1 to P4, the presence or absence of crawling was recorded. Beginning at P5, animals were allowed to crawl freely in the test box for 3 min, and the number of grid lines crossed was documented, along with the number of entries into the center squares. Once animals are fully mobile, the number of entries into center squares can be used as an assessment of anxiety.
From P1 to P14, pups were held a few inches above the table surface and a toothpick was placed under the hind legs. The pups ability or inability to grab the toothpick with the hind feet for at least 1 second was recorded.
Pups were placed on their backs on a warm paper towel, and the latency to right themselves back onto the stomach was recorded (30 s maximum latency). This task was performed daily from P1 to P14.
Pups were positioned with their front paws slightly extended over the edge of a “cliff” 1.5 inches high (an inverted plastic weigh boat). The latency to back up, away from the edge, was recorded with a 30 second maximum latency. The experimenter prevented pups from falling if they failed to retreat from the edge.
Pups were placed on the bottom of an inverted, clean mouse cage positioned at a 45-degree angle. The pups were placed with their head facing down the incline. The latency for the pup to turn around and begin crawling up the incline was recorded, with a maximum latency of 30 seconds. This test was conducted daily from P5 to P14.
Pups were placed in the center of an empty cage. One end of the cage contained soiled bedding from the pup's home cage while the other side contained fresh bedding. The latency to reach the bedding from the home cage was recorded. Pups were given a maximum latency of 180 seconds to reach the home bedding.
Both sides of the pups face were gently stroked with a cotton swab and the presence or absence of head turning in the direction of the stroke was recorded. This task was conducted from P1 to P14 in all animals.
Quantitative behavioral data was calculated using two-way repeated measures ANOVA with genotype as a between-subjects variable and day of testing as a within-subjects variable. Proportional data (the presence or absence of a particular behavior) was compared between groups using the Chi-square test. As no sex differences were observed on any measure, data from males and females was pooled.
Distribution of PR in Neonatal Somatosensory Cortex
The ontogeny of PR expression in the neonatal mouse cortex, shown in Figure 1, coincides with previous analyses in the rat (Quadros et al., 2007; Lopez & Wagner, 2009). At P2, PR expression was clearly visible in both the cortical subplate and the pyramidal cell layer V. By P7, PR expression was virtually absent from the subplate, but was strongly expressed in layers V and II/III.
Early Developmental Milestones
There were no statistical differences between genotypes on the day of eye-opening or the incidence of ear-twitching. These tests have been reliably used as developmental markers for normal nervous system maturation. Although ear twitching was only present in one of the seven litters tested and was not correlated with genotype, all animals in all three treatment groups developed fully opened eyes by P13 and 14.
There was a significant main effect of age on body weight (F2,13 = 473.175; p < 0.001). However, there was no significant main effect of genotype on body weight (F2,22 = 0.448) and no significant interaction [Fig. 2(A)], although there appears to be a trend in which PRKO and HZ animals weigh slightly more than WT animals after P6.
After P5, when most animals were crawling, activity level was measured by number of grid line crossings. There was a significant main effect of age on the number of grid crossings (F2,9= 21.26, p < 0.001), but there was no significant main effect of genotype (F2,18= 0.28) [Fig. 2(B)]. The number of central entries also changed significantly with age (F2,9= 5.15, p < 0.001) with a peak at P9, but again, did not differ between genotypes (F2,18= 0.75) [Fig. 2(B) inset].
Crawling behavior developed in all animals by the end of the first postnatal week and continued throughout the entire period of testing. There was no significant effect of genotype (p > 0.05) on any day of testing [Fig. 2(C)].
The presence of the grasping reflex was highly variable over the first several days but virtually all animals displayed the reflex by P5 and retained it through P14. Chi-square analyses revealed no significant effect of genotype (p > 0.05) throughout the period of testing [Fig. 2(D)].
All animals were able to right themselves within 30 seconds on the first or second postnatal day. There was a main effect of age on latency to right (F2,13 = 27.816; p < 0.001), but there was no significant main effect of genotype (F2,26 = 0.98) and no significant interaction [Fig. 3(A)].
The cliff aversion reflex was present in all animals beginning on the first or second postnatal day. There was a significant main effect of age on latency to retreat (F2,9 = 24.277; p < 0.001), but there was no significant main effect of genotype (F2,18 = 0.908) and no significant interaction [Fig. 3(B)].
There was a significant main effect of age on the latency to climb back up the incline (F2,5= 14.03, p < 0.001), but there was no significant main effect of genotype (F2,10= 0.38) and no significant interaction [Fig. 3(C)].
There was a significant main effect of age on the latency to approach home bedding (F2,9= 24.08, p < 0.001) but no significant main effect of genotype (F2,18= 0.77) and no significant interaction [Fig. 3(D)].
Animals began to demonstrate the rooting reflex as early as P1, and by P11, no animals displayed this behavior. Chi-square analysis revealed that there was a significant effect of genotype for the presence of the rooting reflex on postnatal days 8 (χ2 = 8.34, p < 0.05) and 9 (χ2 = 8.34, p < 0.05) (Fig. 4) with a greater proportion of PRKO and HZ mice losing the rooting reflex earlier than WT mice.
PR is transiently expressed in the subplate and subsequently in layers V and II/III within the mouse somatosensory cortex during a developmental window in which cortical connectivity is established and a variety of sensorimotor developmental milestones emerge. Qualitative analysis revealed equal abundance and distribution of PR immunoreactivity in both males and females. The lack of sex difference in PR expression is consistent with the findings of Lopez and Wagner (2009) demonstrating no sex differences in the expression of PR mRNA in multiple regions of developing rat cortex. However, previous studies in mouse using steroid hormone autoradiography have reported higher levels of progestin binding in the cortex of females compared to males (Shughrue et al., 1991). This apparent discrepancy is most likely explained by differences in species or in methodology, as progestin binding, PR immunoreactivity and PR mRNA each reflect different aspects of PR expression.
PR expression temporally coincides with ongoing developmental processes including dendritic branching and spinogenesis, the arrival of afferent axons and synaptogenesis (Juraska & Fifkova, 1979; Miller & Peters, 1981). As PR is a powerful transcription factor, it is reasonable to hypothesize that PR activity may exert an influence on these processes, and in turn, regulate somatosensory-dependent behaviors. However, neonatal PRKO mice and their HZ counterparts were developmentally typical in almost all measures. Although each of the behaviors observed was highly dependent on age, consistent with previous reports in mice (Glynn et al., 2007), there was no significant effect of genotype on the majority of milestones, although there was a trend for PRKO and HZ animals to have slightly higher body weights than WT after P6. PRKO mice did not differ in the day of eye opening or the emergence of the righting reflex, negative geotaxis, cliff aversion, motor activity level, homing behavior or the proportion of animals crawling or grasping.
In contrast, the loss of the rooting reflex was significantly influenced by genotype. The rooting reflex is a rapid turning of the head in the direction of a tactile stimulus on one side of the face and is critical for normal feeding behavior in neonates. The disappearance of the rooting reflex during postnatal life reflects the emergence of cortical control over an involuntary brainstem reflex and is therefore, an indication of cortical sensorimotor integration (Sherrington, 1917; Ohkawa et al., 1997; Okamoto et al., 2006). A significantly greater percentage of the WT animals maintained the rooting reflex on P8 and P9 compared to PRKO and HZ mice, suggesting that two functional copies of the PR gene are required for the normal developmental loss of this reflex. This suggests that PR plays a role in sensorimotor development.
While a delay in the loss of the rooting reflex can indicate improper or delayed cortical development, it is more difficult to interpret a premature disappearance of the reflex. The expression of PR within the rodent subplate coincides with the arrival of thalamocortical afferents in this layer and the subsequent “waiting period” (Molnar and Blakemore, 1995; Molnar et al., 1998) in which thalamic axons presumably receive chemotropic signals from subplate neurons regarding when to leave the subplate and innervate their correct final target in more superficial lamina. One hypothesis generated by the current study is that PR expression within the subplate may be critical to the proper expression of these chemotropic molecules. In the absence or aberrant expression of these signals, perhaps afferents do not “wait” properly and innervate the cortex prematurely, thus prematurely inhibiting the rooting reflex and possibility leading to improper connectivity that could manifest in further behavioral deficits later in life.
The lack of genotype effects on the majority of developmental milestones could have several potential explanations. The most obvious, of course, is that the expression of PR is not involved in the development of these specific behaviors. Also, while it is possible that PR may not play a critical role in cortical development, this seems somewhat unlikely given the timing of PR expression within pyramidal projection cells in rats (Lopez & Wagner, 2009) and preliminary findings from rats demonstrating that postnatal administration of the PR antagonist RU486 decreases levels of MAP2-ir and the density of dopaminergic innervation of the medial prefrontal cortex (Willing et al., 2010). It is possible that the role of PR in cortical development differs between mice and rats, as PR expression in rat cortex can be found in all functional regions including frontal, motor, somatosensory, auditory, visual cortices (Quadros et al., 2007; Lopez & Wagner, 2009) whereas in mice robust PR expression appears to be localized to somatosensory cortex. Additionally, the prospect that compensatory mechanisms, in the absence of PR gene expression, occurs in the developing cortex, cannot be ruled out, as is the case with negative findings in all knockout models.
The most parsimonious interpretation of the present findings is that the sensorimotor effects of PR knockout may not manifest until later in development. PR is expressed in the rat cortex in an anatomically-specific pattern through the first 4 weeks of life (Quadros et al., 2007; Jahagirdar & Wagner, 2009; Lopez & Wagner, 2009). It is likely that if PR expression during development plays a role in development of the cortex, genotype differences could emerge once cortical maturity is completed (e.g., following adolescence) or in more complex cognitive behavioral tasks rather than fundamental developmental reflexes.
In conclusion, effects of genotype on the loss of the rooting reflex suggest that PR expression may play a role in normal cortical maturation and connectivity even within the first week of life. While the development of most behaviors measured were not altered in PRKO mice, the present findings provide a broad behavioral and developmental characterization of the common transgenic mouse background strain, C57BL6/129SvEv.