INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Exposure to hypoxia in newborn mammals induces a brisk increase in minute ventilation (E), mediated by increased carotid chemoreceptor afferent activity. Upon continued exposure to hypoxia, E gradually decreases to levels near or below those observed during breathing of room air (RA) (1, 2), giving rise to the characteristic "biphasic" hypoxic ventilatory response. This decrease in E is attributed to a central inhibitory process that may involve adenosine (3), -aminobutyric acid (4), and decreased release of nitric oxide (5). In contrast to newborn mammals, adult mammals commonly display an initially more robust increase in E, followed by milder ventilatory declines to levels that remain higher than those measured during breathing of RA. Although the postnatal maturational pattern of the hypoxic ventilatory response has long been recognized, the mechanisms underlying such maturation are currently unknown.

Several studies indicate that N-methyl-D-aspartate (NMDA) glutamate receptors play a pivotal role in brain structures underlying hypoxic chemosensitivity. NMDA glutamate receptors are extensively expressed throughout the adult brain, including brainstem regions traditionally associated with respiratory control, such as the nucleus tractus solitarii (nTS), the site of termination of the carotid chemoreceptor afferents (6). Systemic administration of the NMDA glutamate receptor antagonist ([+]-5 methyl-10,11-dihydro-5H-dibenzo[a,d]cyclopenten-5,10-imine maleate (dizocilpine maleate; MK-801), an NMDA receptor channel blocker, has been found to markedly attenuate both the hypoxic ventilatory response and the peripheral chemoreceptor response to sodium cyanide in conscious, adult rats (7). Similar findings have been reported in anesthetized dogs (8) and developing piglets (9). It has also been found that under hypoxic conditions, increases in glutamate concentrations occur within the nTS of conscious adult rats, and coincide with increases in E (10). Both carotid body denervation and application to the nTS of MK-801 substantially reduce the E responses to hypoxia, indicating that hypoxia-induced increases in nTS glutamate levels are necessary for expression of the hypoxic ventilatory response (10, 11). Thus, peripheral chemoreceptor-mediated hypoxic ventilatory responses are critically dependent on NMDA receptor activation.

NMDA glutamate receptor expression within the brainstem undergoes postnatal developmental changes. NMDA receptors are present in cardiorespiratory areas of the neonatal brainstem, including the pneumotaxic area; however, their density is lower than in the adult animal (12), and with advancing postnatal age there is also a progressive increase in the heterogeneity of NMDA receptors throughout the brain (13). However, potential relationships between postnatal development of NMDA receptors and the maturation of the ventilatory response to hypoxia are yet to be established.

Increases in expression of the protooncogene c-fos occur after synaptic activation, and are considered a reliable marker of activation for individual neurons (14). Hypoxic exposure in conscious adult rats is associated with enhanced c-fos expression within nTS neurons (15, 16), which is attenuated by pretreatment with MK-801 (15). These findings further support the hypothesis that nTS neurons expressing NMDA receptors play a major role in the hypoxic ventilatory response. However as mentioned earlier, the role of NMDA receptors in the hypoxic ventilatory response has not been examined in developing rat pups. Furthermore, the patterns of c-fos recruitment within nTS neurons during hypoxia in relation to NMDA glutamate receptor expression are unknown.

We therefore hypothesized that the maturation of the ventilatory response to hypoxia coincides with an increase in NMDA receptor expression and functional recruitment within the nTS of the developing rat. Accordingly, younger animals would have lower densities of NMDA receptors within the nTS and lesser c-fos activation in this region with hypoxia, and consequently the hypoxic ventilatory response would be minimally affected by administration of an NMDA receptor antagonist. Conversely, more mature rat pups would be expected to exhibit increased densities of NMDA receptors in nTS neurons, enhanced c-fos expression in the nTS, and increased attenuation of the ventilatory response to hypoxia upon treatment with an NMDA receptor blocker.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The experimental protocols used in the study were approved by the Institutional Animal Care and Use Committee of the State University of New York at Buffalo. Time-pregnant Sprague-Dawley rats were obtained from a commercial breeder (Harland-Sprague Dawley, Indianapolis, IN), and delivery times of the rat pups were recorded.

Ventilatory Responses to Hypoxia and Hypercapnia

For hypoxic and hypercapnic ventilatory studies, pups were randomly selected from every litter at postnatal Days 5, 10, and 15. These ages have been shown to be representative of maturational changes in the biphasic response to hypoxia (1).

Protocol

Ventilatory challenges with 10% O₂ (balance: N₂) or 5% CO₂ (balance: air) were initially performed in each rat pup after intraperitoneal administration of normal saline (control) at 0.01 ml/g body weight. Following a 30- to 45-min period of acclimatization, each pup was exposed to either 10% O₂ or 5% CO₂ for 30 min. Switching of gases was done by rapidly bleeding the premixed gas mixture into the recording chamber. Animals were then allowed to recover with their mothers in RA for at least 60 min, and were then injected intraperitoneally with MK-801 (RBI, Natick, MA) at a dose of 1 mg/kg in 0.01 ml/g body weight normal saline. Thirty minutes after MK-801 administration, ventilatory challenges were repeated.

Ventilatory Recordings

Respiratory measures were continuously recorded in the freely behaving, unrestrained animals placed in a previously calibrated 160-ml barometric chamber (Buxco Electronics, Troy, NY). Chamber temperature was maintained with an external heat source within the range of 29° C to 30° C, which corresponds to usual temperatures recorded in the rat nest. A calibration volume of 0.5 ml of air was repeatedly introduced into the chamber at a rate similar to the breathing frequency of the rat pup before and upon completion of recordings. At least 30 min before the start of each protocol, animals were allowed to acclimate to the chamber, through which humidified air was passed at a rate of 0.4 L/min with a precision-flow pump-reservoir system. Pressure changes within the chamber caused by changes in inspiratory and expiratory temperature (17) were measured with a high-gain differential pressure transducer. Analog signals were continuously digitized and analyzed on-line, using a microcomputer software program (Buxco Electronics). A rejection algorithm was included in the breath-by-breath analysis routine, and allowed the accurate rejection of motion-induced artifacts. Tidal volume (VT), respiratory frequency ( f ), and E were computed and stored for subsequent off-line analysis.

Because of the influence of body temperature on VT, VT was corrected according to the recommendations of Mortola and Frappell (18) for whole-body plethysmography, to account for body and chamber temperatures. To ensure undisturbed ventilation during recordings with the pups, rectal temperature was not measured during the ventilatory recordings, but was measured in a separate set of animals (postnatal age 8 d) during a 30-min period of acclimation followed by 30-min of exposure to hypoxia or hypercapnia. These measurements were used for the corrections in 5- and 10-d-old pups. Since thermoregulation in the rat is established by 15 d postnatally, corrections were not made in this age group.

Immunocytochemistry

To determine the developmental pattern of NMDA receptor expression and the topography of neuronal recruitment in response to hypoxia, double-label immunocytochemistry was performed for the NR1 subunit of the NMDA receptor and for the early gene product c-fos (19).

A separate group of rat pups (ages 2 d, 5 d, 10 d, and 20 d) and adult rats were studied after 3-h exposures to either RA or hypoxia (10% O₂, balance N₂). For all pups in this group, exposures were conducted in the presence of the mother, to prevent separation stress and while maintaining environmental temperatures at approximately 28° to 30° C. Five animals were studied per experimental group. At the end of the exposures, rats were anesthetized with pentobarbital (50 mg/kg intraperitoneally) and perfused transcardially with 40 to 200 ml of phosphate-buffered saline (PBS) at ambient temperature, and then with 2.5% paraformaldehyde in cold PBS containing 5% sucrose, pH 7.4. The brain was removed immediately from the skull of each pup after perfusion, and was placed overnight in a fixative containing 1% paraformaldehyde in PBS and 30% sucrose at 4° C. Coronal sections (20 to 30 µm thick) were cut on a freezing microtome and were divided into two series. One series was stained for Nissl substance with thionine, and the other was processed immunocytochemically. Sections were washed extensively in PBS, and incubated for 1 h in 0.4% Triton X-100 in PBS containing 1.5% normal goat serum (Vector Laboratories, Burlingame, CA). Sections were then incubated with anti-c-fos antibody (sc-52, 1:10,000 dilution; Santa Cruz Biotechnology, Santa Cruz, CA), and with an antibody to the NMDA receptor subunit NR1, raised against a synthetic peptide, LQNQKDTVLPRRAIEREEGQLQLCSRHRE, corresponding to the C-terminus of the rat NMDA receptor NR1 subunit (AB1516, 1:2,500; Chemicon, Temecula, CA) (20). The sections were then washed extensively in PBS, incubated in biotinylated antirabbit IgG (Vector) diluted in 0.4% Triton X-100 in PBS for 1 h, washed three times in PBS, incubated for 1 h in avidin-biotinylated horseradish peroxidase (Vectastain Elite kit; Vector), diluted in 0.4% Triton X-100 in PBS, rinsed three times in Tris (pH 7.6), and incubated for variable intervals, until appropriate staining was achieved, in 50 mg% diaminobenzidine tetrahydrochloride (Sigma, St. Louis, MO) and 0.005% H₂O₂ (Sigma) diluted in Tris pH 7.6. The reaction was stopped in PBS, and the sections were mounted from sodium acetate solution onto slides coated with gelatin chrom-alum. The resulting double-labeled neurons were easy to identify because c-fos was localized to the nucleus, whereas the NMDA receptor marker was found in the cytoplasm or plasma membrane. Control experiments were done to determine whether the primary or secondary antibodies produced false-positive results.

Sections were assessed with a light microscope, and the distribution and number of cells containing c-fos, NR1, or c-fos-NR1 immunoreactivities was indicated on camera lucida drawings and maps of the nTS, using the atlases by Paxinos and Watson (21) and Paxinos and colleagues (22). The cytoarchitectural boundaries of the various rostral medullary nuclei were defined by superimposing the adjacent thionine-stained sections on the camera lucida drawings. We primarily analyzed the rostrocaudal levels of the dorsal aspect of the adult rat medulla oblongata encompassed by the region from 12.1 to 14.5 mm caudal to the bregma, for each of the treatments.

Data Analysis

Values are reported as mean ± SEM. For ventilatory challenges, early and late responses were assessed as the average values in the first and last 3-min periods of each 30-min challenge, whereas baseline ventilation was defined as the average of values in the 3 min immediately preceding gas switching. Although the initial 3 min of a hypoxic challenge period may not always correspond to the peak E response in a particular animal, values of E, f, and VT during this period are primarily representative of the peripheral chemoreceptor-mediated component, with little contamination from central sources, and were therefore selected for comparative analyses. Differences in data among the various age groups or within each age group for the saline and MK-801 treatments were compared through analysis of variance (ANOVA) (two-way ANOVA for repeated measures) and the Newman-Keuls post hoc test. For immunocytochemical analyses, data were tabulated and responses were compared by using two-way ANOVA (postnatal age and hypoxia) followed by the Newman- Keuls post hoc test. A value of p < 0.05 was considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effects of MK-801 on Resting Ventilation

Ventilation during breathing of RA (baseline) increased progressively with advancing postnatal age (p < 0.001), due to increases in VT (p < 0.001; Table 1). Values of f were similar in 5-d and 10-d-old pups, but by 15 d f was 15% lower than at the younger ages (p < 0.01). MK-801 administration was without effect on baseline E, VT, or f in 5-d and 10-d-old pups. However, in 15-d-old pups treated with MK-801, f was increased after treatment, resulting in an increase in E (p < 0.001; Table 1).

Effects of MK-801 on the Hypoxic Ventilatory Response

The minute-by-minute biphasic response to hypoxia in unblocked animals is shown in Figure 1. An age-dependent increase in the magnitude of the early hypoxic ventilatory response was observed (Figure 1) between the ages of 10 d and 15 d (p < 0.001). Following administration of MK-801, the early hypoxic ventilatory response was unaltered in 5-d and 10-d-old pups (Figures 2 and 3). However, the large, early E increase observed in 15-d-old pups was attenuated after administration of MK-801, resulting in an early response similar to that observed in the younger age groups (Figures 2 and 3). In all age groups, the early hypoxic ventilatory response before treatment with MK-801 was associated with increases in f (p < 0.05), whereas VT remained unchanged (p = NS). Following MK-801, the f response to hypoxia was diminished at all ages (p < 0.001).

View larger version (41K): [in this window] [in a new window]   Figure 1.   Minute-by-minute ventilatory responses (% change from baseline) to 30 min of hypoxia (10% O2; top panels) and hypercapnia (5% CO2; bottom panels) in rat pups aged 5 d, 10 d, and 15 d.

View larger version (19K): [in this window] [in a new window]   Figure 2.   E (% change from baseline) in three individual rat pups (aged 5 d, 10 d, and 15 d) in response to 30 min of hypoxia (10% O2) before (open squares) and after (closed squares) systemic MK-801 administration.

View larger version (27K): [in this window] [in a new window]   Figure 3.   Average early (2 to 4 min) and late (28 to 30 min) changes (% of baseline) in E, VT, and f in response to hypoxia (10% O2) before (5 d [open circles]; 10 d [open squares]; 15 d [open triangles]) and after (5 d [closed circles]; 10 d [closed squares]; 15 d [closed triangles]) systemic MK-801 administration.

During the 30-min period of the hypoxic challenge, E decreased from the early peak at all postnatal ages. In 5-d and 10-d-old pups E decreased to levels similar to those in RA (Figure 1), whereas in 15-d-old pups the late hypoxic E remained above baseline RA levels (p < 0.001), similar to the adult response (7). Interestingly, the decline in E from the early phase to the late phase was approximately 30% in all age groups. Therefore, because of the larger initial increase in E in the older animals, the 30% decline in ventilation resulted in a level of ventilation during the late phase that remained above baseline values. Following MK-801 administration, the late phase of the ventilatory response to hypoxia was unaltered in 5-d and 10-d-old pups (p = NS). However, in 15-d-old animals MK-801 administration resulted in a late-phase E that was not different from that observed in the younger animals.

Effects of MK-801 on the Hypercapnic Ventilatory Response

To examine whether the effect of MK-801 administration was specific to the hypoxic ventilatory response, we measured the ventilatory response to 5% CO₂ in a separate group of pups (Figure 1). Upon exposure to hypercapnia, E increased by ~ 25% within the first 3 min in all age groups (p < 0.001). By 10 min of CO₂ exposure, E was ~ 40 to 50% greater than its levels in RA in all pups, and remained at this level for the remainder of the exposure period (p < 0.001). Initially, both VT and f increased in response to hypercapnia (Figure 4); however, over the 30-min challenge period, the increased E was sustained by continued increases in VT (p < 0.001).

View larger version (27K): [in this window] [in a new window]   Figure 4.   Average early (2 to 4 min) and late (28 to 30 min) changes (% of baseline) in E, VT, and f in response to hypercapnia (5% CO2) before (5 d [open circles]; 10 d [open squares]; 15 d [open triangles]) and after (5 d [closed circles]; 10 d [closed squares]; 15 d [closed triangles]) systemic MK-801 administration.

After administration of MK-801, the ventilatory response to CO₂ remained intact at all ages, with E reaching levels similar to those observed during control conditions (Figure 4). The increase in E was attributable solely to increases in VT (p < 0.001), since f was not altered at any time during hypercapnia after MK-801 administration (p = NS; Figure 4). These findings indicate that MK-801 administration did not alter the hypercapnic ventilatory response at any postnatal age.

Immunocytochemistry

Under normoxic conditions, low levels of c-fos expression were apparent in the nTS and were similar at all postnatal ages. Indeed, c-fos immunoreactivity was found in 3.5 ± 1.5%, 3.9 ± 1.3%, 5.2 ± 1.7%, 4.9 ± 2.0%, and 6.7 ± 1.8% of cells counted (Figure 5) for 2-d, 5-d, 10-d, 20-d, and adult animals, respectively (p = NS). In contrast, NR1 exhibited site-specific developmental patterns of expression. Indeed, increasing numbers of neurons staining for NR1 were identified in the nTS with advancing age (Figure 5B), whereas the opposite (i.e., decreasing expression of NR1) occurred in the hypoglossal nucleus (Figure 5D).

View larger version (50K): [in this window] [in a new window]   Figure 5.   Percentages (mean ± SD) of c-fos (A), NR1 (B), and NR1-c-fos positive cells in the nTS (C), and of NR1 in the hypoglossal nucleus (D) in rat pups aged 2 d, 5 d, 10 d, and 20 d, and in adult hypoxic rats (n = 5 for each age group). Age-dependent increases in c-fos (p < 0.004), NR1 (p < 0.001), and NR1-c-fos double-labeled cell densities (p < 0.001) occurred in the nTS. In contrast, age-dependent decreases in NR1 expression occurred in the hypoglossal nucleus (D) (p < 0.001).

Hypoxic exposures were associated with increased c-fos expression at all postnatal ages in the nTS (Figure 6A). However, a marked increase in c-fos nuclear staining occurred in  10-d-old animals as compared with younger animals (Figure 6A). Similarly, the densities of c-fos-NR1 double-labeled neurons after hypoxic exposures increased progressively with maturation (Figure 6C).

View larger version (120K): [in this window] [in a new window]   Figure 6.   (Upper panels) Photomicrographs of coronal sections through the caudal ventrolateral medulla at the level of the area postrema illustrating NR1 and c-fos immunoreactivity after exposure to RA (A) and 3 h of hypoxia (B) in 20-d-old pups. Marked c-fos enhancements within the nTS are apparent with hypoxia. nTS = nucleus tractus solitarius; X = dorsal motor nucleus of the vagus; XII = hypoglossal nucleus; cc = central canal. The scale bar is shown in every image in the right hand bottom corner. (Lower panels) Higher magnifications of the nTS in 2-d, 5-d, 10-d, and 20-d-old rat pups following hypoxia. Please note the relative paucity of NR1 and c-fos in the youngest pups and the increased immunostaining for both NR1 and c-fos in the older animals.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study examined the maturation of the ventilatory response to hypoxia in developing rats, and investigated the role of NMDA receptor mechanisms in this response. Systemic administration of the NMDA receptor channel blocker MK-801 was without effect on ventilation in young (5-d and 10-d-old) animals under all conditions studied (RA, hypoxia, and hypercapnia). In contrast, in 15-d-old animals MK-801 was associated with higher values of f during breathing of RA than in the control condition, resulting in greater levels of E. In response to hypoxia, MK-801 abolished both the increased magnitude of early E and the ability to sustain E above its baseline value in RA, to the extent that the hypoxic ventilatory response became indistinguishable from that of the younger animals. The study also showed that significant developmental changes in NMDA NR1 receptor expression occur within the nTS and hypoglossal nucleus, and that with advancing postnatal age, hypoxia induces increases in c-fos immunoreactivity. Furthermore, in accord with the physiologic data in the study, the relative proportion of cells showing double labeling for NR1 and c-fos was enhanced with maturation, thereby suggesting that NMDA glutamate receptors in the nTS play an essential role in development of the mature expression of the hypoxic ventilatory response.

From birth to 2 wk of age, developmental changes were observed in resting ventilation. Baseline E, VT, and respiratory drive increased by ~ 250%, whereas f decreased over this period, similar to findings in previous studies (1, 5). NMDA receptor blockade in the younger animals was without effect during breathing of RA, whereas in 15-d-old pups MK-801 was associated with increases in f, resulting in higher levels of ventilation. Similar MK-801-induced effects on f have been observed in conscious adult rats (7), suggesting that NMDA receptor mechanisms involved in respiratory timing in the conscious rat become functional between 10 d and 15 d postnatally.

The hypoxic ventilatory response also underwent significant changes with advancing postnatal age. At all ages, ventilation in response to hypoxia followed a biphasic pattern, with a brisk early ventilatory increase followed by a reduction in ventilation. Between postnatal Days 10 and 15, the early ventilatory response to hypoxia doubled. However, the ventilatory decline from the peak early phase was ~30% in all age groups. Therefore, it appears that the development of a late-phase response in which ventilation remains above resting levels is determined by the magnitude of the early-phase response. Indeed, the large early-phase ventilatory response in 15-d-old pups altered the late-phase response from one of marked hypoxic ventilatory depression to one characteristic of a mature mammal, in which ventilation remains above baseline levels during the late phase of the response. Augmentation of both f and VT responses to hypoxia accounted for the increases in E during both the early and late phases. This maturational change in the hypoxic ventilatory response is consistent with previous findings in developing mammals (1, 5).

The hypoxic ventilatory response was unaffected by administration of MK-801 to 5-d and 10-d-old pups. In contrast, NMDA receptor antagonism in the 15-d-old group resulted in a change from a mature pattern of hypoxic ventilatory response to an immature pattern similar to that observed in the younger animals. This dramatic alteration in the pattern of the hypoxic ventilatory response suggests that the addition of NMDA receptor-dependent mechanisms is critical for expression of the mature hypoxic ventilatory response. This finding is consistent with previous observations in both conscious (7, 10) and anesthetized (23) adult rats, which demonstrated critical dependency on NMDA receptor activation for full expression of the hypoxic ventilatory response. Interestingly, this dependence on NMDA receptor function for the hypoxic ventilatory response has also been observed in 6-d-old piglets (9), suggesting that in this animal, this particular aspect of respiratory control may be more mature at an earlier postnatal age than in the developing rat. Indeed, examination of the data for the unblocked hypoxic ventilatory response reported by Lin and coworkers (9) indicates that the mature ventilatory pattern (i.e., a late-phase response that remains above baseline levels) is already present in the 6-d-old piglet. Therefore, the findings of Lin and coworkers (9) are consistent with those of the present study in that expression of the mature hypoxic ventilatory response depends on NMDA receptor activation.

Hypercapnic exposure elicited increases in ventilation and VT at all ages, and was unaffected by MK-801 administration. Thus, the mature ventilatory response to CO₂ appears to be present by the fifth postnatal day in developing rats, and does not appear to require NMDA receptor activation for its full expression. We are unaware of other studies of the role of NMDA receptors in the hypercapnic ventilatory response of developing mammals. However, findings in previous studies with conscious adult rats (7, 10) and anesthetized cats (24) are consistent with our present findings, which show that the ventilatory response to CO₂ is unaltered by administration of an NMDA receptor antagonist.

The first central synapses for primary afferents originating from arterial chemoreceptors are within the nTS (6). Neurons within the nTS contain substantial numbers of both NMDA and non-NMDA glutamate receptors (20, 25). Evidence from adult mammals indicates that the early ventilatory response to hypoxia is primarily glutamatergic, and that this response is mediated by NMDA receptors (7, 10, 26). More recently, work in our laboratory has clearly shown that NMDA glutamate receptors are the primary ionotropic glutamate receptors underlying neuronal activation within the nTS during hypoxia in adult rats (27).

The patterns of glutamate receptor expression in particular neuronal populations within the caudal brainstem show marked postnatal developmental changes. In an earlier study, Salès and colleagues demonstrated decreased NMDA receptor binding within brainstem regions considered to mediate cardiovascular and respiratory functions in the neonate as compared with binding in the adult animal (12). Furthermore, Rao and associates reported dramatic increases in glutamatergic neurotransmission within the nTS during the first month of life (28). In the present study we used a polyclonal antibody to the NR1 subunit of the NMDA glutamate receptor that recognizes all of the known splice variants of the NR1 subunit (20). Since the NR1 subunit is viewed as a mandatory component of all functional NMDA receptors, immunocytochemical assessment for NR1 will identify and localize NMDA receptors in rat brain tissue (20), permitting assessment of developmental changes in overall NMDA glutamate receptor expression in the dorsocaudal brainstem. The presence of the NR1 subunit of the NMDA receptor increased in nTS neurons but decreased in neurons of the hypoglossal nucleus during the first 3 wk of development. Thus, regions within the brainstem appear to undergo different maturational changes in NMDA receptor expression, rather than displaying a homogeneous and monotonic increase in NMDA receptor expression over time.

Hypoxic exposure was associated with increased c-fos expression in the nTS at all postnatal ages, but more particularly so in animals 10 d old, suggesting that with maturation, the nTS assumes an increasingly preponderant role in the processing of neural afferent input from peripheral chemoreceptor afferents. The location and magnitude of enhancements in c-fos immunoreactivity in the developing and adult animals in our study are in close agreement with those previously reported by White and associates (29), Haxhiu and colleagues (15), and Teppema and colleagues (16) after similar hypoxic exposures.

The fundamental question we attempted to answer in this study was whether the developmental characteristics of the hypoxic ventilatory response followed a recognizable pattern linked to NMDA glutamate receptor expression and function. Developmental increases occurred in the densities of c-fos- NR1 double-labeled neurons during hypoxia, thereby indicating that with advancing maturation, larger populations of NMDA glutamate receptor-positive cells within the nTS are recruited by hypoxic stimuli. In other words, an increasing dependency on NMDA glutamate receptors emerges over time, such that immunocytochemical findings closely parallel and predict the changes in physiologic responses to hypoxia elicited by administration of the noncompetitive NMDA glutamate receptor antagonist MK-801. Thus, transition from an immature to a more mature hypoxic response requires full expression and functional connectivity of NMDA receptor-bearing neuronal populations within the nTS, and possibly also in other brainstem regions not examined in the present study.

Although identification of the specific composition of the NMDA receptors expressed during development was beyond the scope of the present study, the physiologic data obtained in the study are consistent with the developmental changes reported in NR2 subunit populations. NR2D appears to be the main NR2 subunit expressed in the brainstem during the first 2 wk of life (30, 31). The proposed role of the NR1/NR2D receptor heteromer is one of postsynaptic synchronization of asynchronous presynaptic activity. This has been postulated on the basis of the receptor's extremely slow offset decay of glutamate-induced currents (31). With progressive brain maturation, synchronization of pre- and postsynaptic activities may be "fine-tuned" by a switch to NMDA receptors containing subunits other than NR2D (e.g., NR2A-C), thereby providing current characteristics of a more precise mode of operation (32). This switch may involve an increased expression of the NR2A subunit, since NR2A subunits are coexpressed with NR2D subunits in a reciprocal fashion (33). Therefore, as numbers of the NR2D receptor begin to decline within the brainstem at around postnatal Days 10 to 14 (31, 33), a concomitant increase in NR2A is thought to occur, representing the change in subtype expression from embryonic type to mature type. The maturation of the hypoxic ventilatory response supports an enhancement in excitatory neurotransmission between postnatal Days 10 and 15, the period during which NR2D populations are decreasing in the brainstem and NR2A populations are increasing.

It should also be emphasized that hypoxia is a systemic stimulus with widespread effects on the behavior of multiple systems. For example, hypoxia is associated with changes in heart rate and arterial blood pressure, which could have elicited the activation of nTS NMDA neurons involved in the baroreflex pathway and subsequent induction of c-fos in these neurons (34). However, the histochemical analyses and experimental paradigms employed in our study do not distinguish and identify neuronal populations underlying ventilatory contributions from those involved in cardiovascular regulation. Such assessments are of clear importance and remain to be determined.

In summary, we have shown that developmental changes in the hypoxic ventilatory response parallel maturational changes in the expression of NMDA receptor populations within the nTS, such that a close functional and anatomic relationship between the two becomes apparent. Thus, a mature ventilatory response to hypoxia will occur only with full expression and synaptic connectivity of an NMDA receptor-based neuronal network. In this context, NMDA receptor contributions to the chemical control of breathing appear to be limited to the mature hypoxic ventilatory response, since NMDA receptor antagonism has little if any effect on immature hypoxic ventilatory responses or ventilatory responses to hypercapnia at any age. The neurotransmitter(s) underlying such immature hypoxic ventilatory responses remain to be determined.