Interoceptive signals have a powerful impact on the motivation and emotional learning of animals during stressful experiences. However, current insights into the organization of interoceptive pathways stem mainly from observation and manipulation of adults, and little is known regarding the functional development of viscerosensory signaling pathways. To address this, we have examined central neural activation patterns in rat pups after treatment with lithium chloride (LiCl), a malaise-inducing agent. Rat pups were injected intraperitoneally with 0.15 M LiCl or 0.15 M NaCl (2% body wt) on postnatal day (P)0, 7, 14, 21, or 28, perfused 60 to 90 min postinjection, and their brains assayed for Fos protein immunolabeling. Compared with saline treatment, LiCl increased Fos only slightly in the area postrema, nucleus of the solitary tract, and lateral parabrachial nucleus on P0. LiCl did not increase Fos above control levels in the central nucleus of the amygdala, bed nucleus of the stria terminalis (BNST), or paraventricular nucleus of the hypothalamus on P0 but did on P7 and later. Maximal Fos responses to LiCl were observed on P14 in all areas except the BNST, in which LiCl-induced Fos activation continued to increase through P28. These results indicate that central LiCl-sensitive interoceptive circuits in rats are not fully functional at birth, and show age-dependent increases in neural Fos responses to viscerosensory stimulation with LiCl.
- nucleus of the soiltary tract
in adult rats, intraperitoneal injection of the toxemic agent lithium chloride (LiCl) induces gastrointestinal malaise, promotes anorexia, supports conditioned aversion learning, and induces behavioral signs of anxiety (2, 4, 19, 20, 37–39). Administration of LiCl also activates the hypothalamo-pituitary axis (HPA) stress axis to elevate plasma corticosterone concentrations (33, 36).
Despite the richness of the literature on the behavioral and neural responses of adult rats to interoceptive stressors, such as LiCl, very little is known about the functional development of the neural circuits that support these responses. Smotherman and Robinson (34) reported that rats given a LiCl/flavor pairing in utero on embryonic day (E)17 and subsequently retested with the flavorant alone on E19 displayed more aversive-like motor behaviors compared with controls given saline/flavor pairings. Other studies also have reported that rats display LiCl-conditioned taste aversions or flavor avoidance (CFA) during tests of independent ingestion on postnatal day (P)1 (9), P5 and P9 (12), and P10 to P15 (5, 10). Beginning at P15, pups are able to retain CFA learning and expression for periods of many weeks after initial LiCl exposure (5, 32), although younger pups show deficits in this respect. The neural circuits that support the learning and expression of these conditioned behaviors in young pups are virtually unknown, and may be distinct from circuits that operate in adult rats.
In adult rats, responses to interoceptive stressors are mediated, in part, by viscerosensory signals relayed to the nucleus of the solitary tract (NTS) (7, 18, 40, 43, 44, 45). Injection of LiCl activates the NTS via vagal sensory afferents and direct stimulation of neurons within the chemosensory area postrema (AP) (2, 46). The resulting neural signals are then relayed to other areas in the brain stem, such as the parabrachial nucleus (PBN) and to forebrain regions, such as the paraventricular nucleus of the hypothalamus (PVN), the central nucleus of the amygdala (CeA), and the bed nucleus of the stria terminalis (BNST).
The purpose of the present study was to document potential age-related changes in neuronal expression of the immediate early gene product Fos in response to LiCl injection. On the basis of results from previous studies from our laboratory reporting postnatal changes in central neural connectivity and Fos responsiveness to interoceptive stimuli (23–25, 27), three hypotheses were tested. First, we expected that newborn rats would display adult-like patterns of LiCl-induced Fos expression in the caudal brain stem (i.e., NTS, AP, and PBN). Second, we expected that LiCl-induced activation of neurons within the hypothalamus and limbic forebrain (i.e., PVN, BNST, and CeA) would be minimal or absent in newborn rats. Finally, we expected asymptotic increases in forebrain Fos responsiveness to LiCl that would achieve adult-like levels by P14.
MATERIALS AND METHODS
All procedures were conducted in accordance with National Institutes of Health standards for the use of animals in experimental research, under the oversight of the University of Pittsburgh Institutional Animal Care and Use Committee. Pregnant female Sprague-Dawley rats (Harlan, Indianapolis, IN) were brought into our animal care facility between gestational days 13 and 16. The facility is temperature controlled (20–22°C) with a light-dark cycle of 12:12 h, with lights on at 0700. Pregnant dams were housed individually in plastic tubs filled with soft wood chip bedding. Tubs were checked each day to ensure accurate dating of parturition, designated as P0. All litters were culled to 10 pups on P0 or P1. Both males and females were used in all experiments to minimize the number of litters needed to carry out this research. Pups were weaned at P21 and separated into same-sex groups. Some pups used at P0 were obtained during the culling of litters. The remaining pups were used in these experiments as whole litters or subsets of whole litters at ages P0, P7, P14, P21, and P28.
Beginning at 0900 on the experimental day, rat pups were removed from their dam, weighed and marked individually, and the entire litter was brought to a separate room in a small plastic transport tub with bedding from their home cage and placed in an incubator (Animal Care Products, Bryan, TX). No more than 12 pups from two litters were treated and perfused on the same day. Any pups remaining in a given litter after treatment of their littermates were excluded from further study. Pups aged P0 to P14 were maintained at 33°C, while pups aged P21 to P28 were maintained at 24°C, with humidity set at 50–60%. All pups were given 1 h to acclimate to the incubator before any experimental manipulation.
After acclimation, pups were treated according to a randomized rotating schedule with 10-min intervals separating the treatment of each pup. Using a 28-gauge needle, pups were injected intraperitoneally with either aqueous 0.15 M NaCl (saline control group; Abbott Laboratories, North Chicago, IL, volume: 2% body wt) or aqueous 0.15 M LiCl (Sigma; volume: 2% body wt). Fresh LiCl solution was made before each experiment. Saline and LiCl solutions were syringe-filtered (0.45 μm) and equilibrated in a water bath to 37°C before injection. After injection, each pup was marked as to treatment using indelible marker and returned to its littermates in the incubator. Pups then were left undisturbed for 60–90 min before perfusion. This posttreatment survival time is similar to those used in our previous Fos studies examining the effects of interoceptive stressors in neonatal and adult rats (25, 27) and is based on evidence that nuclear Fos protein immunolabeling peaks ∼1 h after significant neural activation and persists for ∼2 h, or longer if neural activation continues (35).
Perfusion and tissue collection.
Pups were anesthetized with pentobarbital sodium (Nembutal, 10 mg/kg body wt ip; Abbott Laboratories) and transcardially perfused using aqueous 0.15 M NaCl for the first minute followed by 4% paraformaldehyde and 1% acrolein (wt:wt) in 0.1 M PBS (pH 7.4) for 10 min. Heads were removed using scissors and stored overnight in 4% paraformadehyde. On the following day, brains were removed from the skull, postfixed for an additional 24 h in 4% paraformaldehyde, and then transferred to 30% aqueous sucrose (wt:wt) for 48 h for cryoprotection. Brains were blocked, frozen, and sectioned coronally (50 μm) on a freezing-stage microtome. Sections from the upper cervical spinal cord through the rostral corpus callosum were collected serially in four adjacent sets and stored in cryopreservant solution at −20°C before processing for immunohistochemistry. One set of tissue sections per animal (i.e., representing brain regions sampled at 200-μm intervals) was used for cell counting.
Sections were removed from cryopreservant, rinsed in phosphate buffer, and treated sequentially in 0.1% sodium borohydride (wt/vol; Sigma, St. Louis, MO) and aqueous 0.075% hydrogen peroxide (vol/vol; Sigma) with intervening buffer rinses. As detailed below, sections from each rat were processed for dual immunoperoxidase visualization of Fos and dopamine β-hydroxylase (DbH). DbH neuron and fiber immunolabeling was used as a guide to help standardize the anatomical regions and subnuclei subjected to quantitative analyses.
Primary and secondary antisera were diluted in 0.1 M phosphate buffer containing 1% normal donkey serum, 1% BSA and 0.3% Triton X-100. The Fos protein was localized using a rabbit antiserum (kindly provided by Dr. Philip Larsen, Denmark) raised against amino acids 4–17 of synthetic Fos protein (1:50,000). The specificity of this antiserum for Fos protein in rats has been reported (27). Tissue sections were incubated for 24 h at room temperature, rinsed, and then incubated in biotinylated donkey anti-rabbit IgG (1:500; Jackson Immunoresearch Laboratories, West Grove, PA) for 1 h. After rinsing, sections were treated with Elite Vectastain reagents (Vector Laboratories, Burlingame, CA), using nickel-intensified diaminobenzidine to produce a blue-black nuclear peroxidase label.
Fos-immunolabeled sections were subsequently processed for immunolocalization of DbH to identify noradrenergic (NA) brain stem neurons and their terminal fields within viscerosensory regions of the pons, hypothalamus, and limbic forebrain. DbH was visualized using a mouse monoclonal antibody (1:30,000; Chemicon, Temecula, CA). Sections were incubated overnight at room temperature, rinsed, and then incubated for 1 h in biotinylated donkey anti-mouse IgG (1:500; Jackson Immunoresearch Laboratories, West Grove, PA). After rinsing, sections were treated with Elite Vectastain reagents (Vector Laboratories, Burlingame, CA), using plain DAB to produce a brown cytoplasmic peroxidase label. Tissue sections were then rinsed and mounted onto Superfrost Plus microscope slides (Fisher Scientific, Pittsburgh, PA), dehydrated, and defatted by serial immersion in graded ethanols and xylene, and placed under coverslips with Cytoseal 60 (VWR, West Chester, PA).
Quantification and data analysis.
Cell counting was performed to document the extent of treatment-related Fos immunoreactivity in the AP, NTS, PBN, CeA, PVN, and BNST. Anatomical landmarks were defined with reference to the atlas of Paxinos and Watson (21), with sampling regions further standardized across animals based on the subregional distribution of DbH immunolabeling. Fos-positive neurons were defined by the appearance of blue-black nuclei, or blue-black punctae distributed throughout the nucleus, independent of intensity. DbH-positive neurons were defined by a brown-stained cell body with a clearly defined nucleus. All areas were sampled bilaterally at 200-μm intervals. The AP was sampled along its full rostrocaudal extent, typically 2–4 sections per animal. The NTS was sampled across the full rostrocaudal extent of its viscerosensory subregions, coextensive with the dorsal medullary A2/C2 NA cell groups. NTS subregional analyses were conducted for Fos-positive neurons caudal to the level of the AP (A2 cell group region; 4–10 sections per animal, depending on age), at the level of the AP (A2/C2 region, 2–4 sections per animal), and rostral to the level of the AP (C2 region, 2–6 sections per animal). Cell counting within the C2 region of the NTS was discontinued at the rostral level where the NTS moves laterally away from the wall of the fourth ventricle. Cell counting within the PBN was restricted to its external lateral subdivision at the level of the superior cerebellar peduncle (1–2 sections per animal). The CeA was sampled across its entire rostrocaudal extent (3–9 sections per animal). The BNST was sampled within sections located just rostral to, caudal to, and at the midline crossing of the anterior commissure, including the lateral and ventrolateral subdivisions (2–5 sections per animal). Because of the well-established links between early postnatal development and the emergence of HPA axis neuroendocrine responses to stressors, cell counting within the PVN was restricted to its medial parvicellular subdivision, immediately caudal to the optic chiasm and medial to the lateral magnocellular subdivision (1–3 sections per animal).
The volume of each sampled brain region within each animal was determined using a light microscope with integrated software package (Simple PCI version 6.0). In each section, the region of interest was manually (digitally) outlined, and its area was measured. To compute an estimated volume of the total region represented in sampled tissue sections, each value was multiplied by 50 (section thickness 50 μm) and then by 4 (each animal was sampled using a single serial set out of 4 taken). As expected, the volume of each brain region sampled increased over development (see results). We assumed that these increases in regional volume are due primarily to increased neuropil (i.e., fibers, neurites, and glia) and/or increases in cell size and are not due to postnatal increases in neuron number (see results for verification of this assumption). If the brain region in question grows larger, but the number of neurons expressing Fos remains unchanged, there will be more sections sampled at later ages and an apparent net decrease in the average number of cells counted per section. If the number of neurons expressing Fos increases along with age at or below the rate of growth in the region of interest, then the average number of neurons per section will appear to be constant. To avoid confounding cell count and brain growth, cell count data were collected and analyzed in terms of the total number of Fos-positive neurons identified, across all sections. Cell count data are presented as group means ± SE, except as noted. Data were first analyzed to determine the effects of age, pup sex, and treatment on Fos labeling, including activation of DbH-positive neurons in the brain stem, using a three-factor ANOVA. Interaction effects between age, sex, and treatment were also examined. Results were considered statistically significant when P < 0.05.
Subsequent to ANOVA analysis, linear contrast analysis was used to detect age-related trends in regional Fos responses to LiCl treatment (8, 28). This approach has higher statistical power and provides a more focused test of experimental predictions than ordinary null hypothesis-based post hoc testing. We predicted that brain stem regions (i.e., NTS, AP, and PBN) would be activated maximally (i.e., in an adult-like manner) after LiCl treatment on P0. We also tested two alternative hypotheses regarding age-related trends in Fos immunoreactivity in the CeA, PVN, and BNST. For this purpose, contrast weights were derived for three models: adult-like trend, linear trend, and asymptotic trend. To test for the adult-like trend, we predicted that maximal (i.e., adult-like) values of Fos labeling after LiCl treatment, and minimal values after saline treatment, would be present beginning on P0 and persist across all 28 postnatal days. To test the linear trend model, minimum levels of Fos immunoreactivity were predicted for both saline- and LiCl-treated rats on P0, with a gradual and linear increase in Fos responsiveness until P28, at which time Fos immunoreactivity in LiCl treated rats would peak. To test for an asymptotic trend in Fos immunoreactivity over development, minimum levels of Fos activation were predicted for both saline and LiCl treated rats on P0, with a gradual and linear increase until P14 when Fos immunoreactivity in the LiCl treated rats would peak, with no further change at later ages. For the latter model, P14 was chosen as the asymptotic point of development because it marks the end of the so-called stress hyporesponsive period (reviewed in Refs. 16 and 41) and also because ascending NA viscerosensensory pathways and reciprocal preautonomic outputs to the viscera undergo substantial modification during the first 2 wk postnatal in rats (23, 24). The quality of fit of a particular trend to the data was judged by the size of the correlation between the data and the hypothetical trend (Rcontrast). Statistical significance was accepted when P < 0.05, an indication of a significant correlation between the model predictions and the data. Comparing R2 values between these different models (Rcomparison) provides a direct index as to which trend best fits the data (8).
Body weight and brain growth.
Variations in the timing of birth and individual differences in developmental patterns make the use of postnatal age as a proxy for developmental completion somewhat problematic, particularly when comparing data across laboratories. In addition, variable increases in the volume of different brain regions over development can complicate the interpretation of cell count data (see materials and methods). Table 1 shows changes in animal body weight and calculated volumes of brain regions sampled in this study over the course of development. Postnatal increases in calculated regional brain volumes ranged from less than twofold (PVN) to approximately fivefold (NTS; Table 1). Although these tissue volume increases are significant, they do not account for the postnatal increases in the number of Fos-positive neurons counted within each brain region in rats after LiCl treatment compared with Fos counts in age-matched controls, as described below.
Fos and DbH colocalization in the NTS.
The calculated volume of the NTS sampled per rat increased almost fivefold over the first 28 postnatal days (Table 1). Most of the increase in regional volume occurred by P14. Treatment with LiCl only slightly increased the number of Fos-positive NTS neurons compared with Fos counts in saline-treated controls on P0 (Figs. 1 and 2A). The number of Fos-positive NTS neurons continued to rise in LiCl-treated animals until peaking on P14 (Fig. 2A). LiCl-induced Fos immunoreactivity decreased slightly after P14 but remained level from P21 to P28. Age and treatment, but not sex, had significant effects on Fos immunoreactivity. The age-times-treatment interaction was also statistically significant. Summary statistics for all effects are provided in Table 2. Subregional analysis revealed similar overall patterns of Fos immunoreactivity in the caudal, AP-level, and rostral NTS (Table 3), with Fos-positive cell counts peaking at P14 and decreasing slightly thereafter. At each age examined, LiCl-induced elevation of Fos was greater in the AP-level and rostral NTS than in the caudal NTS (Table 3).
The interaction effect between age and treatment was explored using linear contrast analysis (8, 28). For LiCl-induced activation of Fos in the NTS, both asymptotic and adult-like trends explained a significant proportion of the variance (Rcontrast = 0.866, P < 0.001 and Rcontrast = 0.809, P < 0.001, respectively). Although both trends explained significantly more variance than the linear trend model, they did not differ significantly from each other (Rcomparison = 0.159, P = 0.163).
The number of DbH-immunoreactive neurons counted within the NTS did not change over development (Table 4), evidence that the increased volume of NTS measured from P0 through P28 was not accompanied by an increased number of NA neurons within the NTS. Total Fos counts within the NTS were only slightly elevated by LiCl compared with saline treatment on P0, and there was no significant effect of treatment on the proportion of DbH-positive medullary neurons expressing Fos at this early age (Table 3). LiCl-induced activation of Fos expression by DbH-immunoreactive neurons increased steadily over development (Fig. 3).
Subregional analysis of Fos activation in DbH-positive NA neurons in the caudal, AP-level, and rostral viscerosensory NTS mirrored the age-related pattern of Fos activation in the NTS as a whole. In each rostrocaudal level of the NTS, age and treatment, but not sex, had significant effects on Fos expression (Tables 2 and 3). The only exception was for the AP-level NTS, in which ANOVA did not reveal a significant effect of age on LiCl-induced Fos activation in DbH-positive neurons; however, the age-times-treatment interaction was significant at all three rostrocaudal NTS levels (Table 2). LiCl-induced Fos expression in DbH-immunoreactive NTS neurons increased steadily over development. By P28, 80% of all DbH-positive neurons counted within the NTS colocalized nuclear Fos immunolabeling after LiCl treatment (Table 3). LiCl-induced Fos expression within the rostral subregion of the NTS reached a plateau at P14, at which time 40–50% of DbH neurons colocalized Fos immunolabeling. As no prior hypotheses were made regarding the subregional distribution of Fos or its colocalization with DbH immunolabeling in the NTS, no additional contrast analyses were performed on these data.
Fos immunoreactivity in AP.
AP regional volume increased by 205% over the first 28 postnatal days (Table 1). As with the NTS, most of the growth occurred by P14. Treatment with LiCl only slightly increased Fos immunoreactivity in the AP on P0 (Figs. 1 and 2B), with further increases evident through P14. AP Fos responsiveness to LiCl subsequently decreased after P14 but remained higher than Fos responses to saline at each age examined (Fig. 2B). Age and treatment, but not sex, had significant effects on Fos activation in the AP (Table 2). The age-times-treatment interaction on AP Fos expression also was significant.
The age-times-treatment interaction effect was explored by using linear contrast analysis. For LiCl-induced activation of Fos in the AP, both asymptotic and adult-like trends explained a significant proportion of the variance (Rcontrast = 0.801, P < 0.001 and Rcontrast = 0.781, P < 0.001, respectively). Although both trends explained significantly more variance than the linear trend model, they did not differ significantly from each other (Rcomparison = 0.213, P = 0.083).
Fos immunoreactivity in PBN.
The parabrachial nucleus increased 228% in sampled volume over the first 28 postnatal days (Table 1). Compared with Fos labeling in saline-treated controls, LiCl treatment increased Fos immunoreactivity in the lateral PBN as early as P0 (Figs. 2C and 4). Peak LiCl-induced Fos immunoreactivity was seen at P14, with a decrease and plateau thereafter (Fig. 2C). Age and treatment, but not sex, had significant effects on Fos immunolabeling in the PBN (Table 2). The age-times-treatment interaction also was significant.
The age-times-treatment interaction effect was explored using linear contrast analysis. For LiCl-induced activation of Fos in the PBN, the asymptotic trend explained a significant proportion of the variance (Rcontrast = 0.801, P < 0.001). The asymptotic trend explained significantly more variance than both the linear trend model (Rcomparison = 0.839, P < 0.001), and the adult-like trend model (Rcomparison = 0.675, P < 0.001).
Fos immunoreactivity in CeA.
The volume of the CeA increased ∼280% over the first 28 days of postnatal life (Table 1). Treatment with LiCl did not significantly affect CeA Fos immunoreactivity compared with saline controls at P0 (Figs. 2D and 5). The number of Fos-positive CeA neurons increased rapidly and linearly in LiCl-treated animals until reaching a peak at P14. LiCl-induced Fos activation subsequently declined at P21 and P28 below P14 peak levels (Fig. 2D). Age and treatment had significant effects on Fos-positive cell counts (Table 2). There was a nearly significant effect of pup sex on CeA Fos counts, with males having slightly higher Fos counts compared with females. The age-times-treatment interaction was also significant (Table 2).
The age-times-treatment interaction effect was explored using linear contrast analysis. For LiCl-induced activation of Fos in the CeA, the asymptotic trend explained a significant proportion of the variance (Rcontrast = 0.912, P < 0.001). The asymptotic trend explained significantly more variance than both the linear trend model (Rcomparison = 0.904, P < 0.001), and the adult-like trend model (Rcomparison = 0.485 P < 0.001).
Fos immunoreactivity in BNST.
The total volume of the BNST increased ∼352% over the first 28 days postnatal (Table 1). BNST Fos counts after LiCl treatment were not different from counts after saline treatment at P0 (Figs. 2E and 6). In contrast to the P14 peaks in Fos expression observed in other regions described above, LiCl-induced Fos immunoreactivity continued to increase within the BNST across the full 28 days of postnatal development examined (Fig. 2). Fos immunoreactivity within the BNST also increased over development in saline-treated control subjects. Age and treatment, but not sex, had significant effects on Fos-positive cell counts in the BNST. The age-times-treatment interaction also was significant.
The age-times-treatment interaction effect was explored using linear contrast analysis. For LiCl-induced activation of Fos in the BNST, the asymptotic trend explained a significant proportion of the variance (Rcontrast = 0.908, P < 0.001). The asymptotic trend explained significantly more variance than both the linear trend model (Rcomparison = 0.689, P < 0.001) and the adult-like trend model (Rcomparison = 0.595 P < 0.001).
Fos immunoreactivity in medial parvicellular PVN.
The sampled volume of the medial parvicellular PVN (mpPVN) increased by 178% over the first 28 days postnatal (Table 1). Treatment with LiCl did not induce higher levels of Fos within the mpPVN compared with Fos after saline treatment at P0. At P7, treatment-induced Fos expression increased after both LiCl and after saline, although counts were slightly higher in the LiCl-treated group (Figs. 2F). LiCl treatment more markedly increased Fos expression in the mpPVN compared with saline-treated controls beginning at P14 (Fig. 7). LiCl-induced Fos expression remained stable from P14 through P28 (Fig. 2F). Fos activation after saline injection continued to increase through P21 but declined by P28. Age and treatment, but not pup sex, had significant effects on Fos expression in the PVN (Table 2). The age-times-treatment interaction also was significant.
The age-times-treatment interaction effect was explored using linear contrast analysis. For LiCl-induced activation of Fos in the PVN, the asymptotic trend explained a significant proportion of the variance (Rcontrast = 0.855, P < 0.001). The asymptotic trend explained significantly more variance than both the linear trend model (Rcomparison = 0.765, P < 0.001), and the adult-like trend model (Rcomparison = 0.613 P < 0.001).
Results from this study demonstrate a gradual emergence of central neural Fos responses to LiCl treatment over early postnatal life in rats. Brain stem regions involved in processing interoceptive stimuli, that is, AP, NTS, and PBN, showed surprisingly small increases in Fos activation after LiCl on P0, whereas LiCl-induced Fos expression in hypothalamic and limbic forebrain regions was elevated above control levels on P7 and later, but not on P0. For both NTS and AP, postnatal changes in Fos expression after LiCl injection were consistent with both the asymptotic growth and adult-like trend models (Table 2). Fos expression in the PBN was best fit by the asymptotic trend model. Fos expression in the CeA, BNST, and PVN after LiCl injection was best fit by the asymptotic trend model. LiCl-induced Fos expression in all of the brain regions examined increased during the first week postnatal (i.e., between P0 and P7). Interestingly, Fos responsiveness to LiCl treatment peaked on P14 in all brain regions examined, with the exception of the BNST. The utility of the asymptotic model in explaining forebrain responses to LiCl injection likely results from the gradual assembly of viscerosensory circuits during postnatal life (e.g., 23). The data did not allow us to distinguish between the adult-like and asymptotic trends for the NTS and AP. It appears that these areas are less sensitive to stimulation by LiCl in early postnatal life.
The significance of the apparently transient increased neural sensitivity to LiCl treatment on P14 is unclear. Similar findings have been reported in studies investigating responses to hypoxia in developing rats, in which Fos activation peaks around P10–P15 and subsequently declines (3) and in Fos and ACTH secretory responses to systemic LPS, which peak around P12 (6). Neural development includes both circuit assembly and regression (i.e., functional pruning), and our results may reflect both processes. There are reports of several bouts of postnatal synaptogenesis in the NTS between P12 and P14 (42). Moreover, the descending projections from PVN to NTS appear to become functionally mature during this time frame (24), as do reciprocal ascending NA projections from NTS to PVN (23).
Converging evidence indicates that the gradual emergence of Fos immunoreactivity in response to LiCl treatment observed in this study is the result of postnatal changes in hindbrain neural sensitivity to LiCl and/or to changes in signaling from the hindbrain to higher-level neural circuit components and is not due to changes in the capacity of individual neurons to express Fos when activated. For example, intraperitoneal injection of octapeptide (CCK) in 2-day-old rats activates robust Fos expression within the AP and NTS but not in the PVN, and the lack of PVN Fos expression is accompanied by a lack of pituitary hormone secretion after CCK treatment (25). Conversely, subcutaneous injection of hypertonic saline in 2-day-old rats provokes high levels of hypothalamic Fos immunoreactivity and significant pituitary hormone secretion, but little or no activation of NTS neurons (27). Other studies of developing animals also have shown that alterations in Fos expression in response to experimental manipulation correlate with changes in physiological and behavioral responses to the manipulation (13, 17). Thus, the present results demonstrating age-related increases in Fos responsiveness to LiCl are best explained by age-related changes in sensitivity to LiCl and/or changes in functional connectivity among the relevant circuit components. Subdiaphragmatic vagal sensory inputs to the NTS are already well established during embryonic development (26), and by the time of birth, the chemoarchitecture and cytoarchitecture of the NTS is remarkably adult-like (48). The surprisingly modest NTS and AP Fos responses to LiCl treatment on P0 is evidence that these neurons are relatively insensitive to LiCl in newborn rats, and this relative insensitivity cannot be discounted as a contributing factor to the correspondingly modest or absent Fos responsiveness of higher brain regions on P0.
In adult rats, conditioned learning in response to LiCl exposure depends on the action of chemosensitive neurons situated in the AP (15, 20), on signals relayed from the AP and lateral PBN (1, 22) to the CeA (29, 31, 47), and on portions of insular cortex (14, 30). Our results compare with others that show that conditioned responses to flavor/LiCl pairings in newborn rats (9, 12, 34) occur in the absence of adultlike circuit functionality as assessed by LiCl-induced Fos expression. Although pups as young as embryonic day 17 appear capable of conditioned learning in response to LiCl (34), it is not clear how the unconditioned LiCl stimulus is detected and neurally represented at this age.
Viscerosensory circuits play a critical role in behavioral, emotional, and physiological responses to environmental challenge. In spite of this, few studies of the development of neural responses to viscerosensory stressors can be found in the literature (but see Refs. 3, 6, 11) The present results add to this growing literature, and future studies will focus on how perturbations to normal development in early infancy alter the functional assembly of viscerosensory circuits in the rat.
This work was supported by National Institutes of Health Grants T32- MH-18273 (to T. J. Koehnle) and MH 59911 (to T. J. Koehnle and L. Rinaman).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society