Ventilatory and carotid body responses to acute hypoxia in rats exposed to chronic hypoxia during the first and second postnatal weeks

Abstract

Chronic hypoxia (CH) during postnatal development causes a blunted hypoxic ventilatory response (HVR) in neonatal mammals. The magnitude of the HVR generally increases with age, so CH could blunt the HVR by delaying this process. Accordingly, we predicted that CH would have different effects on the respiratory control of neonatal rats if initiated at birth versus initiated later in postnatal development (i.e., after the HVR has had time to mature). Rats had blunted ventilatory and carotid body responses to hypoxia whether CH (12% O₂) occurred for the first postnatal week (P0 to P7) or second postnatal week (P7 to P14). However, if initiated at P0, CH also caused the HVR to retain the “biphasic” shape characteristic of newborn mammals; CH during the second postnatal week did not result in a biphasic HVR. CH from birth delayed the transition from a biphasic HVR to a sustained HVR until at least P9–11, but the HVR attained a sustained (albeit blunted) phenotype by P13–15. Since delayed maturation of the HVR did not completely explain the blunted HVR, we tested the alternative hypothesis that the blunted HVR was caused by an inflammatory response to CH. Daily administration of the anti-inflammatory drug ibuprofen (4 mg kg⁻¹, i.p.) did not alter the effects of CH on the HVR. Collectively, these data suggest that CH blunts the HVR in neonatal rats by impairing carotid body responses to hypoxia and by delaying (but not preventing) postnatal maturation of the biphasic HVR. The mechanisms underlying this plasticity require further investigation.

1. Introduction

Chronic exposures to low oxygen levels elicit plasticity in the control of breathing in humans and other mammals, profoundly altering eupneic ventilation and protective ventilatory chemoreflexes (Pamenter and Powell, 2016; Bavis and MacFarlane, 2017). Interestingly, the effects of chronic hypoxia (CH) on some aspects of respiratory control, notably the acute hypoxic ventilatory response (HVR), may be quite different in newborn and adult mammals. In adult mammals, for example, exposures to sustained normobaric or hypobaric hypoxia lasting from several hours to several weeks produce ventilatory acclimatization to hypoxia (VAH), a progressive increase in ventilation associated with an augmented HVR and persistent hyperventilation after acute return to normoxia (Dempsey et al., 2014; Pamenter and Powell, 2016). In contrast, similar durations of hypoxia from birth generally have the opposite effect on the HVR of newborn mammals (Bavis, 2005). Several studies have documented that children born at high altitude (Lahiri et al., 1976; Lahiri, 1981) or who experience hypoxemia secondary to cyanotic heart disease (Sorensen and Severinghaus, 1968; Edelman et al., 1970; Blesa et al., 1977) develop blunted HVR, suggesting that CH attenuates the HVR in newborns rather than enhancing it. This effect has been confirmed in animal experiments in which neonatal rats (Eden and Hanson, 1987a; Wyatt et al., 1995; Reeves and Gozal, 2006; Mayer et al., 2014; MacFarlane et al., 2016; Stryker et al., 2018), cats (Hanson et al., 1989a; Hanson et al., 1989b), and sheep (Sladek et al., 1993) exposed to CH (5–30 days in 10–15% O₂, depending on the study) were shown to have reduced HVR compared to individuals reared in normoxia.

Changes in the HVR after CH are attributed (at least in part) to altered function of the carotid bodies, the peripheral arterial chemoreceptors that monitor partial pressure of O₂ (Po₂) in the arterial blood and initiate the ventilatory response. Whereas CH increases carotid body O₂ sensitivity and the gain for the central integration of carotid chemoafferent input in adult mammals (Powell et al., 2000; Teppema and Dahan, 2010; Kumar and Prabhakar, 2012; Pamenter and Powell, 2016), carotid body responsiveness decreases when CH begins at birth. For example, carotid chemoreceptor responses to hypoxia recorded from the carotid sinus nerve (CSN) were reduced in juvenile rats (20–28 d of age; (Donnelly and Doyle, 1994)) and cats (12–23 d of age; (Hanson et al., 1989a)) born and reared in hypoxia. Moreover, glomus (type I) cells isolated from CH-reared neonatal rats exhibited smaller depolarization responses (Wyatt et al., 1995) and intracellular Ca²⁺ ([Ca²⁺]_i) responses (Sterni et al., 1999) to acute hypoxia, indicating that the diminished afferent nerve activity has a presynaptic origin.

Carotid body chemoreceptors exhibit a marked “resetting” of O₂ sensitivity in the early postnatal period, with the threshold Po₂ at which carotid body activity increases gradually shifting to higher values to accommodate the greater arterial Po₂ in the postnatal environment (Carroll and Kim, 2013). Accordingly, the magnitude of the carotid body and ventilatory responses to hypoxia increase over the first few postnatal days or weeks in newborn mammals (Bissonnette, 2000; Teppema and Dahan, 2010). It has been suggested that the postnatal rise in Po₂ is itself a key regulator of carotid body resetting (Carroll and Kim, 2013). Indeed, an earlier rise in Po₂ appears to cause premature resetting of carotid body chemoreceptors in fetal lambs (Blanco et al., 1988). It is possible, therefore that decreased carotid body O₂ sensitivity (and thus the blunted HVR) observed after chronic postnatal hypoxia represents delayed maturation of the carotid body chemoreceptors (Landauer et al., 1995; Sterni et al., 1999); in other words, by extending the relatively hypoxic in utero environment into the postnatal period, the carotid body resetting process may be slowed or completely prevented. If this hypothesis is correct, one might predict that CH would not attenuate the carotid body or ventilatory responses to hypoxia if the exposure is initiated in the postnatal period but after the carotid chemoreceptors have already reset. We are not aware of any studies that have tested this prediction with respect to carotid body function, although there is evidence that the steady-state HVR is blunted in rats exposed to CH beginning at 11 days of age (Mayer et al., 2014).

Changes to carotid body function may not be solely responsible for abnormal development of the HVR. Newborn mammals exhibit a strongly biphasic HVR in which an initial increase in ventilation is quickly followed by a decline in ventilation toward or below baseline levels. Whereas the initial ventilatory increase (i.e., the early phase) is dominated by the activation of the carotid bodies, the subsequent decline (i.e., the late phase) represents the balance between carotid body stimulation and inhibitory processes within the CNS (Bissonnette, 2000; Teppema and Dahan, 2010). As mammals mature, however, the HVR is characterized by a more sustained increase in ventilation due to diminished CNS inhibition. Like the resetting of carotid chemoreceptor O₂ sensitivity, there is some evidence that this transition is modulated by perinatal oxygenation. Specifically, chronic postnatal hyperoxia causes the transition from the biphasic HVR to the sustained HVR to occur earlier in neonatal rats (Bavis et al., 2010; Bavis et al., 2014). Similarly, the biphasic HVR may persist longer in rats and cats born into hypoxia (Eden and Hanson, 1987a; Hanson et al., 1989a; Joseph et al., 2000), although the persistence of the biphasic HVR during CH has not been evident in all rat studies (Reeves and Gozal, 2006) nor in sheep (Sladek et al., 1993). Moreover, whether a biphasic HVR would re-emerge if hypoxia is initiated later in the postnatal period (i.e., after already a sustained HVR phenotype) has not been studied.

The cellular and molecular mechanisms linking CH to changes in respiratory control are not fully understood, but recent studies indicate that inflammatory processes influence respiratory plasticity in adult animals exposed to hypoxia. Sustained or intermittent hypoxia can activate microglia in the brainstem and increase the expression of pro-inflammatory molecules in peripheral and central tissues critical to the control of breathing (reviewed in (Hocker et al., 2017; Pena-Ortega, 2019)). In adults, CH-induced inflammation contributes to the peripheral and central mechanisms underlying VAH (Liu et al., 2011b; Hocker et al., 2017). CH increases the numbers of macrophages and the expression of cytokines (IL-6, IL-1β, TNFα) and chemokines (MCP-1) within the carotid body, while administration of anti-inflammatory drugs (ibuprofen, dexamethasone) diminish these effects and block the expected increase in carotid body O₂ sensitivity (Liu et al., 2009; Liu et al., 2013). Similarly, chronic administration of ibuprofen prevents the increase in hypoxic ventilation characteristic of VAH in adult rats (Popa et al., 2011) and humans (Basaran et al., 2016). Importantly, while these data point toward excitatory effects of neuroinflammation on respiratory control, inflammatory processes may also have inhibitory effects on respiratory control and respiratory plasticity (e.g., (Herlenius, 2011; Gauda et al., 2013; Hocker et al., 2017)). In neonatal rats, for example, lipopolysaccharide (LPS) exposure elicits a systemic inflammatory response that is associated with abnormal carotid body structure, decreased carotid body responses to hypoxia, and decreased ventilatory responses to changing O₂ levels (Master et al., 2016). Moreover, CH in neonatal rats (beginning at 11 days of age) increases the activation of microglia in the brainstem, and this contributes to the blunted HVR observed in these animals (MacFarlane et al., 2016). Therefore, it is possible that CH-induced activation of inflammatory signaling pathways also contribute to the blunted HVR of rats born and raised in hypoxia.

The present study addressed two overarching hypotheses. First, we tested the hypothesis that chronic postnatal hypoxia attenuates the HVR of neonatal rats by delaying the maturation of its peripheral (early phase) and central (late phase) components. If CH delays normal postnatal maturation, we predicted that hypoxia would have different effects on carotid body function and the biphasic HVR if initiated at birth versus initiated at a later postnatal age (i.e., after the carotid body and HVR attain their mature phenotypes). Further, we predicted that prolonging the period of hypoxia from birth would continue to delay the transition from the biphasic HVR to the sustained HVR. Second, we tested the hypothesis that inflammatory signaling processes contribute to the blunted HVR in neonatal rats reared in CH. Specifically, we predicted that chronic ibuprofen treatment would prevent the effects of CH on the acute HVR of neonatal rats.

2. Methods

All experimental protocols were approved by the Bates College Institutional Animal Care and Use Committee.

2.1. Chronic hypoxia treatment

Timed-pregnant Sprague-Dawley rats were obtained from a commercial supplier (Charles River Laboratories). At the appropriate ages for each set of experiments (see below), each litter was housed in a standard cage that was placed into a clear, acrylic chamber flushed with mixtures of room air and N₂ at flow rates sufficient to maintain approximately 21% O₂ (Control) or 12% O₂ (Chronic Hypoxia, or CH) and less than 0.4% CO₂; pups were housed with their own mothers throughout the exposure period. Rats were exposed to a 12-h light : 12-h dark cycle and were provided food and water ad libitum. The environmental chambers were opened briefly for routine cage-cleaning every few days and to remove individuals for study as needed.

In Experiment #1, nine litters of rats (4 Control, 5 CH) were placed into the environmental chambers on the day of birth (P0) approximately 6–12 hours after parturition and eight litters (4 Control, 4 CH) were placed into the chambers at 7 days of age (P7). Individuals were maintained in the chambers for one week until the pups were studied at 7 or 14 days of age (i.e., P7 and P14), respectively. In a follow up study, eight additional litters (4 Control, 4 CH) were placed into the environmental chambers on the day prior to birth (i.e, embryonic day 21, E21) and maintained in the chambers until the pups were studied at 13–15 days of age (P13–15).

In Experiment #2, eight litters (4 Control, 4 CH) were placed into the environmental chambers on P0 approximately 6–12 hours after parturition as described above. Within each litter, half of the pups were selected to receive daily injections of ibuprofen and the remainder received daily injections of saline (see below). Pups received their first injections prior to being placed into the environmental chambers, and individual litters were removed from the chambers each day (<15 min) to perform subsequent injections until studied at 9–11 days of age (P9–11). Since pups received their final daily injection on the day of the ventilation measurements (see below), we wondered whether circulating ibuprofen might itself alter the control of breathing. Accordingly, the acute effects of saline and ibuprofen on breathing were studied using two additional litters of rats that had been reared in room air (i.e., outside of the environmental chamber) through 9–10 days of age (P9–10).

2.2. Ibuprofen treatment

Ibuprofen sodium salt (#I1892, Sigma) was prepared in sterile saline to a concentration of 0.44 mg ml⁻¹ (i.e., ~ 0.4 mg ibuprofen ml⁻¹ after accounting for the molecular weight of sodium). The ibuprofen solution was administered intraperitoneally at 10 μl g⁻¹ body weight to achieve a final ibuprofen dose of 4 mg kg⁻¹. For the CH study, pups received their initial injections just prior to being placed into the environmental chamber on P0 and then received one injection daily (between 08:00 and 09:30) until studied at P9–11; the final injection was delivered on the morning of the ventilation measurements. The dosage and schedule of administration of ibuprofen (4 mg kg⁻¹ d⁻¹, i.p.) was selected to match those shown to be effective in adult rats exposed to CH (e.g., (Liu et al., 2009; Liu et al., 2011a; Popa et al., 2011; Liu et al., 2013)). A vehicle control group (i.e., siblings of the pups receiving ibuprofen) received equivalent volumes of sterile saline.

Since ibuprofen was administered daily to neonatal rats during the CH exposure up to and including the day of the ventilation measurements, it was necessary to first rule out any acute effects of ibuprofen on the control of breathing. Therefore, a separate group of P9–10 rats received a single injection of ibuprofen (4 mg kg⁻¹, i.p.) or saline approximately one hour prior to ventilation measurements.

2.3. Ventilation measurements

Ventilation measurements were made using a head-body plethysmograph. The plethysmograph chamber consisted of a cylindrical body compartment (3.1 cm inner diameter) and an overlapping head compartment (4.75 cm inner diameter; 70 ml). The head and body compartments were separated by a flexible collar constructed of layers of latex film and Parafilm (Pechiney Plastic Packaging Co.) through which the pup’s head would protrude; a layer of petroleum jelly was applied around the pup’s neck to ensure an airtight seal. A continuous flow of air was delivered to the head chamber at a flow rate of 1.5 L min⁻¹ using a mass flow controller (MFC-4; Sable Systems) and valves (840 series; Sierra Instruments). Meanwhile, a pneumotach and differential pressure transducer (MLT1L and ML141, respectively; ADInstruments) were used to monitor airflow in and out of the body compartment caused by the animal’s breathing movements; the pneumotach was calibrated with air injections prior to each experiment. Output from the pressure transducer was recorded at a sampling rate of 1000 Hz, passed through a high pass filter (0.1 Hz), and integrated to obtain respiratory volumes (PowerLab 8SP and LabChart 7 software with the spirometry extension; ADInstruments). The plethysmograph chamber was placed into a temperature-controlled incubator to maintain ambient temperature at 32–34°C; air temperature within the body compartment was continuously monitored (IT-18 thermocouple; Physitemp Instruments).

Rat pups were weighed and sealed into the plethysmograph chamber within 30 min after being removed from the environmental chambers. Following a 10–15 min adjustment period (21% O₂, balance N₂), baseline ventilation was recorded for approximately 5–10 min. Gas concentrations were then switched to 12% O₂ (balance N₂) for 15 min to assess the HVR. Respiratory parameters (tidal volume, respiratory frequency, and minute ventilation) were calculated from 40–60 s of the record during the baseline period, and from 10–15 s of the record during minutes 1, 5, 10, and 15 of hypoxia to characterize the shape of the ventilatory response.

2.4. Single-unit carotid chemoafferent activity

Shortly after the completion of the ventilation measurements in Experiment #1, some rat pups were deeply anesthetized with CO₂ and decapitated. The carotid bifurcation and petrosalnodose ganglion complex were removed en bloc and placed in an ice-cold saline solution (containing in mM: 125 NaCl, 5 KCl, 2 CaCl₂, 1 Na₂HPO₄, 1 MgSO₄, 24 NaHCO₃, and 5 dextrose) bubbled with a 95% O₂ / 5% CO₂ gas mixture. The tissue was exposed to a solution of 0.10–0.15% collagenase (collagenase P; Roche Diagnostics) and 0.02–0.025% protease (type XIV; Sigma-Aldrich) in the oxygenated saline for 30 min at 37°C and excess tissue was then removed until only the carotid body, CSN, glossopharyngeal nerve, and petrosal ganglion remained. The carotid body-petrosal ganglion complex was then transferred to a chamber (RC-22C; 140 μL; Warner Instruments) on the stage of an inverted microscope (Eclipse TE-300; Nikon) and superfused with warm (~37°C) saline solution bubbled with a 21% O₂ / 5% CO₂ gas mixture. Flow rates through the perfusion chamber were approximately 3 mL min⁻¹.

Single-unit nerve activity was recorded from soma of chemoafferent neurons using a suction electrode placed on the petrosal ganglion. Electrical signals were amplified 2,000–5,000×(EX-1; Dagan Instruments), passband-filtered (0.1–2 kHz), digitized (10 kHz sample rate), and recorded to a computer (Powerlab 8/30 and LabChart 7 software; AD Instruments). Chemoafferent neurons were identified under baseline conditions (21% O₂ /5% CO₂) based on the appearance of an orthodromic action potential after stimulating the carotid body (~200 μA × 0.05 ms pulse duration) (Isostim A320; World Precision Instruments). Spontaneous activity was recorded for 2 min under baseline conditions followed by 2 min exposures to acute hypoxia (10% or 5% O₂ / 5% CO₂ / balance N₂); the order of the hypoxic exposures was randomized and they were separated by a 5 min recovery period in 21% O₂ / 5% CO₂. One recording was made for each carotid body-petrosal ganglion complex. Single-unit chemoreceptor activity was discriminated off-line (Spike Histogram Module v.2; ADInstruments). The number of action potentials was counted in 1 s bins throughout the protocol, and the peak activity in each condition was determined from a 3 s moving average of action potential frequency. The number of action potentials was also counted for the final minute of baseline to determine the average baseline discharge rate.

2.5. Carotid body volume

Shortly after the completion of the ventilation measurements in Experiment #2, some rat pups were euthanized with CO₂ and decapitated for dissection of the carotid bifurcations. The bifurcations were fixed with 4% paraformaldehyde in 0.1 M phosphate buffered saline (PBS; pH 7.4) for at least 1 h, cyoprotected in 30% sucrose in PBS (24 h at 4°C), and embedded in Tissue-Tek O.C.T. compound (Sakura Finetek). Embedded samples were flash frozen on dry ice and stored at −80°C until sectioned at 12 μm with a cryostatic microtome, mounted on slides, and stained with hematoxylin and eosin. Carotid body sections were imaged with a 2 megapixel camera attached to an Eclipse 80i microscope (Nikon). The area of each carotid body section was then calculated using Image J software (version 1.43u; National Institutes of Health). Carotid body volume was estimated by summing the product of section area and section thickness for all sections through the carotid body.

2.6. Hematocrit and hemoglobin concentration

Shortly after the completion of the ventilation measurements in Experiment #2, some rat pups were euthanized with CO₂ and decapitated for blood collection. Blood was sampled from the severed neck vessels. For hematocrit, blood was collected into heparinized hematocrit tubes (HematoClad; Drummond) in duplicate and centrifuged for 5 min (ZIPocrit; LW Scientific). The hematocrit was then estimated from the average ratio of packed red cell volume to the total sample volume. Hemoglobin concentration was determined from separate blood samples from the same individuals using an Hb 201⁺ analyzer (HemoCue).

2.7. Statistical analysis

Changes in ventilation over time and changes in carotid chemoafferent activity across levels of hypoxia were compared among hypoxia treatment groups (i.e., CH vs. Control) using two-way repeated measures ANOVA (factor 1: treatment, factor 2: time or level of O₂) followed by Tukey’s post hoc tests (SigmaStat 4.0; Systat Software). Where necessary, data were transformed to meet the equal variance and/or normality assumptions for parametric tests prior to running the ANOVA. Additional two-way ANOVA tests (without repeated measures) were used to determine whether sex or ibuprofen influenced the effect of CH on respiratory variables, carotid body volume, and/or hematological parameters (factor 1: treatment, factor 2: sex or drug) followed by Tukey’s post hoc tests where appropriate (Prism 8; GraphPad Software); sex did not have a significant effect on any of the respiratory variables studied (see section 3.2.3 Sex effects on ventilation), so data for males and females were pooled for all other analyses to increase power. Unpaired t-tests (with Welch’s correction where necessary) were used when comparing only two groups (Prism 8; GraphPad Software). Significance was determined at P≤0.05 for all statistical tests. Values are presented as mean ± SEM.

3. Results

3.1. Body mass

Rat pups exposed to CH were smaller than age-matched control pups reared in normoxia (Table 1). CH rats exposed for the first postnatal week (P0 to P7) weighed 27% less than their Control counterparts (P<0.001). Similar durations of hypoxia during the second postnatal week (P7–P14) reduced body mass as well, but to a lesser extent (approximately 12% smaller; P<0.001). Longer exposures to postnatal hypoxia (i.e., 9–15 days beginning at or just prior to birth) reduced the average body mass by more than 40% (Table 1).

Table 1.

Body mass and baseline respiratory characteristics for neonatal rats after chronic exposure to 21% O₂ (Control) or 12% O₂ (CH).

Period of Exposure	Age	Treatment	n	Mass (g)	Frequency (breaths min−1)	Tidal Volume (ml)	Mass-Specific Tidal Volume (ml 100g−1)	Ventilation (ml min−1)	Mass-Specific Ventilation (ml min−1 100g−1)
Experiment #1
P0 to P7	P7	Control	35	17.2 ± 0.3	147.7 ± 3.2	0.11 ± 0.003	0.65 ± 0.01	16.3 ± 0.4	95.4 ± 2.5
		CH	43	12.6 ± 0.2 *	149.4 ± 2.4	0.11 ± 0.003	0.92 ± 0.02 *	17.0 ± 0.5	136.0 ± 4.0 *
P7 to P14	P14	Control	37	30.7 ± 0.2	121.0 ± 3.2	0.23 ± 0.01	0.75 ± 0.02	27.8 ± 1.2	90.9 ± 4.0
		CH	27	27.1 ± 0.5 *	123.4 ± 2.4	0.24 ± 0.01	0.90 ± 0.02 *	30.0 ± 1.0	111.3 ± 3.8 *
E21 to P13–15	P13–15	Control	19	29.7 ± 0.7	125.8 ± 3.0	0.20 ± 0.01	0.67 ± 0.02	24.7 ± 0.7	83.6 ± 2.1
		CH	17	17.4 ± 0.7 *	173.1 ± 12.0*	0.15 ± 0.01 *	0.84 ± 0.05 *	23.9 ± 1.2	138.8 ± 6.9 *
Experiment #2
P0 to P9–11	P9–11	Control + Saline	13	21.7 ± 0.5	151.7 ± 3.7	0.14 ± 0.01	0.63 ± 0.03	20.6 ± 0.6	94.0 ± 3.0
		Control + Ib	14	21.1 ± 0.6	145.5 ± 4.2	0.13 ± 0.01	0.62 ± 0.02	18.9 ± 0.8	90.0 ± 3.6
		CH + Saline	14	11.5 ± 0.6 §	172.7 ± 5.4 §	0.11 ± 0.01 §	1.02 ± 0.06 §	19.4 ± 0.8	172.8 ± 8.3 §
		CH + Ib	16	12.0 ± 0.7 §	175.2 ± 5.1 §	0.12 ± 0.01 §	0.99 ± 0.04 §	20.2 ± 1.0	171.5 ± 7.7 §

Values are mean ± SEM. For experiment #1,

^*P<0.05 vs. Control within the same age group. For experiment #2,

^§P<0.05 vs. Control (i.e., main effect for CH treatment); Ib, Ibuprofen.

3.2. Experiment #1: Ventilation and carotid body activity after chronic postnatal hypoxia

3.2.1. Ventilation after exposures lasting from P0 to P7

At baseline (21% O₂), absolute values of respiratory frequency, tidal volume, and ventilation (i.e., not normalized to body mass) were similar between CH and Control rats (P=0.65, 0.52, and 0.31, respectively) (Table 1). Given the lower body mass of CH rats, however, mass-specific tidal volume and ventilation were increased by 42% and 43%, respectively, in CH rats compared to Controls (both P<0.001) (Table 1 and Fig. 1B – C, left panels).

Fig. 1.

Hypoxic ventilatory response of P7 rats reared in 21% O₂ (Control; n=35) or 12% O₂ (CH; n=43) from P0 to P7. (A) Respiratory frequency, (B) tidal volume, and (C) minute ventilation were measured in 21% O₂ (baseline, BL) and during 15 min of 12% O₂ and are expressed in raw units (left panels) and as a percentage increase from baseline (right panels); values are mean ± SEM. * P<0.05 vs. Control at the same time point and ^# P<0.05 vs. 1^st minute of hypoxia within the same treatment group.

CH affected the changes in ventilation that occurred during acute 12% O₂ challenges (treatment × time, P<0.001) (Fig. 1C, left panel). Respiratory frequency increased to a similar extent in both treatment groups during acute hypoxia, but only Control rats increased their tidal volumes. As a result, the increase in mass-specific ventilation during the first minute of hypoxia was smaller in CH rats than in Control rats (Δ = 31.8 ± 2.8 vs. 42.4 ± 2.2 ml min⁻¹ 100g⁻¹; P=0.004). Although mass-specific ventilation remained greater in CH rats during the first minute of hypoxia due to the greater baseline values (Fig. 1C, left), the relative increase in ventilation was substantially smaller in CH rats (25.2 ± 2.7 vs. 45.8 ± 2.7% increase from baseline; P<0.001) (Fig. 1C, right).

After the first minute of hypoxia, respiratory frequency and tidal volume remained fairly stable and well above baseline for the duration of the hypoxic challenge in Control rats (Figs. 1A and 1B). In contrast, respiratory frequency and tidal volume declined in CH rats. Specifically, respiratory frequency was significantly reduced at minutes 10 and 15 relative to the 1^st minute of hypoxia and tidal volume was reduced at minutes 5, 10, and 15 (all P<0.05). As a result, CH rats exhibited a strongly biphasic HVR with ventilation falling slightly below baseline levels by the 15^th minute of hypoxia (Fig. 1C). The change in mass-specific ventilation between the 1^st and 15^th minutes of hypoxia was calculated to quantify this biphasic shape, and this analysis confirmed that the ventilatory decline was significantly greater in CH rats than in Control rats (Δ = −42.3 ± 3.8 vs. −5.0 ± 3.3 ml min⁻¹ 100g⁻¹; P<0.001). As a result of the greater ventilatory decline in CH rats, differences in the magnitude of the HVR became more pronounced over the course of the hypoxic challenge (−4.7 ± 3.2 vs. 41.3 ± 3.9% change from baseline at minute 15 in CH and Control rats, respectively; P<0.001) (Fig. 1C, right).

3.2.2. Ventilation after exposures lasting from P7 to P14

Similar to rats exposed to CH during the first postnatal week, CH during the second postnatal week affected baseline breathing in P14 rats (Table 1 and Fig. 2). Absolute values of respiratory frequency, tidal volume, and ventilation were similar between CH and Control rats in 21% O₂ (P= 0.56, 0.07, and 0.19, respectively) (Table 1). Given their lower body mass, however, mass-specific tidal volume and ventilation were increased by 20% and 22%, respectively, in CH rats at P14 (both P<0.001) (Table 1 and Fig. 2B – C, left panels).

Fig. 2.

Hypoxic ventilatory response of P14 rats reared in 21% O₂ (Control; n=37) or 12% O₂ (CH; n=27) from P7 to P14. (A) Respiratory frequency, (B) tidal volume, and (C) minute ventilation were measured in 21% O₂ (baseline, BL) and during 15 min of 12% O₂ and are expressed in raw units (left panels) and as a percentage increase from baseline (right panels); values are mean ± SEM. There was no significant treatment × time interaction in panel A, so ^‡ P<0.05 vs. 1^st minute of hypoxia for both treatment groups. There was a significant treatment × time interaction in panels B and C, so * P<0.05 vs. Control at the same time point and ^# P<0.05 vs. 1^st minute of hypoxia within the same treatment group.

CH treatment during the second postnatal week also influenced how mass-specific ventilation changed during acute exposure to 12% O₂ (treatment × time, P=0.023) (Fig. 2C, left). Due to a smaller increase in tidal volume, CH rats had a smaller increase in minute ventilation during the first minute of acute hypoxia (Δ = 48.0 ± 7.6 vs. 76.5 ± 4.5 ml min⁻¹ 100g⁻¹; P=0.002). This blunting is more evident when the ventilatory response is expressed relative to baseline (Fig. 2C, right), with the initial increase in ventilation being significantly less in CH rats than Control rats (46.7 ± 8.0 vs. 88.9 ± 6.2% increase from baseline; P<0.001).

When considering the late phase of the HVR (i.e., minutes 5–15), the basic shape of the HVR was similar in CH and Control rats: there was a modest ventilatory decline during the 10^th and 15^th minutes of hypoxia, but ventilation remained well above baseline in both treatment groups (Fig. 2C). The decline in mass-specific ventilation between the 1^st and 15^th minutes of hypoxia was not significantly different between treatment groups (Δ = −13.4 ± 7.9 vs. −23.3 ± 6.2 ml min⁻¹ 100g⁻¹ in CH and Control rats, respectively; P=0.32). If anything, CH rats tended to show a less biphasic HVR than Control rats given that post hoc tests detected more differences in tidal volume and minute ventilation over time in Control rats than in CH rats (Fig. 2). Overall, the HVR of CH rats was attenuated relative to Control rats throughout the duration of the acute hypoxia challenge (Fig. 2C, right).

3.2.3. Sex effects on ventilation

In the preceding sections, respiratory data for female and male rats were pooled. In a previous study, however, it was shown that long-term effects of chronic postnatal hypoxia on the HVR vary by sex (Bavis et al., 2004); only male rats exhibited a blunted HVR when studied after approximately two months of recovery in room air in that study. Therefore, the data from sections 3.2.1 and 3.2.2 were re-examined to investigate potential sex differences; these experiments were intentionally designed to have large enough sample sizes for both sexes to permit this comparison. Sex did not influence the effects of chronic postnatal hypoxia on the acute HVR at P7 (sex, P=0.58 and treatment × sex, P=0.10) or at P14 (sex, P=0.10 and treatment × sex, P=0.82) (Fig. 3). As expected, this analysis confirmed that postnatal hypoxia decreased the HVR whether delivered in the first or second postnatal week after controlling for sex (treatment, P<0.001 for both).

Fig. 3.

Influence of sex on the HVR of (A) P7 rats and (B) P14 rats exposed to 21% O₂ (Control) or 12% O₂ (CH) for seven days (i.e., from P0 to P7 or from P7 to P14, respectively). Data for the early phase of the HVR (i.e., 1^st minute of hypoxia) from Figures 1C and 2C are replotted by sex; values are mean ± SEM. Sample sizes (female, male) are Control: 16, 19 and CH: 25, 18 in panel A and Control: 19, 18 and CH:14, 13 in panel B. ^§ P<0.05 vs. Control (i.e., main effect for hypoxia treatment).

3.2.4. Carotid chemoafferent responses to hypoxia

To assess carotid body involvement in the attenuated HVR after one week of postnatal hypoxia, single-unit carotid chemoafferent activity was recorded in vitro under normoxic baseline conditions (21% O₂) and in two levels of hypoxia (10% and 5% O₂). Average baseline activity measured over a 1 min period was significantly greater for CH rats than for Control rats at P7 (2.4 ± 0.4 vs. 0.9 ± 0.2 impulses s⁻¹, respectively; P=0.002) but not at P14 (2.8 ± 0.6 vs. 1.4 ± 0.4 impulses s⁻¹; P=0.08). A similar trend was observed for the peak activity under baseline conditions (i.e., peak activity calculated from the 3 s moving average of action potential frequency), but these trends did not reach statistical significance (Fig. 4).

Fig. 4.

Single-unit chemoafferent activity recorded in vitro from carotid bodies of (A) P7 rats and (B) P14 rats exposed to 21% O₂ (Control) or 12% O₂ (CH) for seven days (i.e., from P0 to P7 or from P7 to P14, respectively). Peak discharge frequency is reported under baseline conditions (21% O₂) and at two levels of hypoxia (10% and 5% O₂); values are mean ± SEM. Sample sizes are 16 Control, 17 CH in panel A and 15 Control, 17 CH in panel B. * P<0.05 vs. Control at the same O₂ level.

At both ages, carotid chemoafferent response to hypoxia were reduced after CH (treatment × O₂ level, both P<0.001) (Fig. 4). For rats treated with CH during the first postnatal week (P0 to P7), the peak activity in hypoxia was significantly reduced at both 10% O₂ and 5% O₂ relative to Controls (both P<0.001) (Fig. 4A). The differences between groups are more pronounced when hypoxic responses are expressed as a change from baseline (Δ activity = 2.3 ± 0.6 vs. 10.5 ± 1.7 impulses s⁻¹ at 10% O₂ and 3.9 ± 1.1 vs. 14.0 ± 1.1 impulses s⁻¹ at 5% O₂ for CH and Control rats, respectively; both P<0.001). For rats treated with CH during the second postnatal week (P7 to P14), the peak activity in hypoxia was reduced only at 5% O₂ (P<0.001) (Fig. 4B); a similar trend at 10% O₂ was not statistically significant (P=0.06). After accounting for variation in baseline activity, however, chemoafferent responses to hypoxia were attenuated in P14 rats at both levels of hypoxia (Δ activity = 3.4 ± 0.7 vs. 8.6 ± 1.5 impulses s⁻¹ at 10% O₂ and 3.9 ± 0.9 vs. 15.2 ± 2.0 impulses s⁻¹ at 5% O₂ for CH and Control rats; P=0.004 and <0.001, respectively).

3.2.5. Ventilation after exposures lasting from E21 to P13–15

Since one week of hypoxia from birth prolonged the period during which rat pups expressed a biphasic HVR, an additional group of rats was studied to determine whether longer hypoxic exposures could perpetuate the biphasic HVR indefinitely; exposures began at E21 in these experiments to prevent even a brief rise in inspired O₂ during the early postnatal period. CH from E21 to P13–15 affected the respiratory pattern exhibited by rat pups at baseline (21% O₂) (Table 1 and Fig. 5). Tidal volume (not normalized to body mass) was 25% smaller in CH rats (P<0.001), but this was offset by a 38% increase in baseline respiratory frequency (P<0.001); accordingly, there was no net change in ventilation when measured in absolute units (P=0.55) (Table 1). The decrease in tidal volume in CH rats was less than expected based on their decrease in body size, however, so CH rats still exhibited greater mass-specific tidal volume (+25%, P=0.004) and ventilation (+66%, P<0.001) than age-matched Control rats (Table 1 and Fig. 5B – C).

Fig. 5.

Hypoxic ventilatory response of P13–15 rats reared in 21% O₂ (Control; n=19) or 12% O₂ (CH; n=17) from E21 until studied. (A) Respiratory frequency, (B) tidal volume, and (C) minute ventilation were measured in 21% O₂ (baseline, BL) and during 15 min of 12% O₂ and are expressed in raw units (left panels) and as a percentage increase from baseline (right panels); values are mean ± SEM. * P<0.05 vs. Control at the same time point. Note that there were no significant differences between the 1^st minute of hypoxia and any of the subsequent time points during hypoxia within either treatment group (i.e., all P>0.05).

Changes in mass-specific ventilation during acute 12% O₂ were attenuated in P13–15 rats after CH (treatment × time, P<0.001) (Fig. 5C, left). The expected increase in tidal volume was completely abolished in CH rats (Fig. 5B). In contrast to CH rats treated with only one week of hypoxia (Figs. 1A and 2A), these rats also exhibited a smaller increase in respiratory frequency during the first minute of the hypoxic challenge compared to age-matched Controls (9.1 ± 2.7 vs. 31.3 ± 5.5% increase from baseline, P=0.001) (Fig. 5A, right). Accordingly, CH rats exhibited a smaller increase in mass-specific minute ventilation during the first minute of acute hypoxia (Δ = 13.5 ± 3.0 vs. 50.6 ± 5.6 ml min⁻¹ 100g⁻¹; P<0.001) (Fig. 5C, left). Similarly, when the ventilatory response is expressed relative to baseline (Fig. 5C, right), the initial increase in ventilation was significantly less in CH rats than Control rats (10.5 ± 2.5 vs. 61.4 ± 7.0% increase from baseline; P<0.001).

After the first minute of hypoxia, respiratory frequency and tidal volume remained fairly stable for the duration of the hypoxic challenge in both CH and Control rats (Fig. 1A and 1B). Likewise, no differences were detected between the 1^st minute of hypoxia and any subsequent time point in hypoxia for minute ventilation for either treatment group (all P>0.05). Therefore, the change in mass-specific ventilation between the 1^st and 15^th minutes of hypoxia (i.e., ventilatory decline) also was not significantly different between CH and Control rats (Δ = 1.0 ± 8.3 vs. −10.3 ± 6.9 ml min⁻¹ 100g⁻¹, respectively; P=0.30).

3.3. Experiment #2: Chronic postnatal hypoxia and ibuprofen

3.3.1. Acute effects of ibuprofen on ventilation

Since rats that were treated daily with ibuprofen received their final dose on the morning of the ventilation measurements, a separate group of P9–10 rats was studied first to confirm that acute ibuprofen administration does not affect the HVR; 10 rats received saline and 11 rats received ibuprofen (4 mg kg⁻¹, i.p.) approximately one hour prior to the measurements. Baseline (21% O₂) mass-specific ventilation was unaffected by acute ibuprofen (84.9 ± 3.2 vs. 79.5 ± 4.0 ml min⁻¹ 100g⁻¹ in saline controls; P=0.31). The HVR was also similar between ibuprofen and saline groups whether assessed during the 1^st minute of hypoxia (96.3 ± 6.9 vs. 87.9 ± 5.8% increase from baseline; P=0.36) or the 15^th minute of hypoxia (67.9 ± 7.8 vs. 62.8 ± 11.6% increase from baseline; P=0.71).

3.3.2. Ventilation after chronic hypoxia and ibuprofen

Daily treatment with ibuprofen had no discernable effect on the magnitude or shape of the acute HVR of P9–11 rat pups reared in hypoxia (Fig. 6A). Mass-specific ventilation was increased in CH rats when returned to normoxia (treatment, P<0.001), but ibuprofen did not alter this effect (drug, P=0.63 and treatment × drug, P=0.78) (Table 1 and Fig. 6B). Ibuprofen did have an effect on the early phase of the HVR (treatment × drug, P=0.02), but this was driven by changes in the response of Control rats (Fig. 6C). Specifically, Control rats that received ibuprofen increased ventilation more in the first minute of hypoxia than their saline-injected counterparts (65.2 ± 5.0 vs. 48.1 ± 5.9% increase from baseline, P=0.04), but no such difference was detected between ibuprofen and saline within CH rats (P=0.94). Overall, the early phase of the HVR was significantly reduced by CH whether rats received saline (P<0.001) or ibuprofen (P<0.001) (Fig. 6C). Finally, ibuprofen had no effect on the ventilatory decline during the late phase of the HVR (Fig. 6D). The decrease in mass-specific ventilation between the 1^st and 15^th minutes of hypoxia (i.e., ventilatory decline) was greater in CH rats than in Control rats at P9–11 (treatment, P<0.001), but the magnitude of the ventilatory decline was not affected by ibuprofen (drug, P=0.84 and treatment × drug, P=0.98).

Fig. 6.

Influence of ibuprofen on the hypoxic ventilatory response (HVR) of P9–11 rats exposed to 21% O₂ (Control) or 12% O₂ (CH) from P0 until studied. Individuals received daily injections of saline (13 Control, 14 CH) or 4 mg kg⁻¹ ibuprofen (14 Control, 16 CH). Panel A shows ventilation measured in 21% O₂ (baseline, BL) and during 15 min of 12% O₂. Panels B-C highlight specific attributes of the HVR that were subjected to statistical comparisons, including (B) baseline ventilation, (C) the increase in ventilation during the 1^st minute of hypoxia, and (D) the ventilatory decline between the 1^st and 15^th minutes of hypoxia. Values are mean ± SEM. Where there was a significant treatment × drug interaction (i.e., panel C), * P<0.05 vs. Control in the same drug treatment and ^† P<0.05 vs. Saline within the same treatment group. Otherwise (i.e., panels B and D), ^§ P<0.05 vs. Control (i.e., main effect for CH treatment).

3.3.3. Carotid body volume after chronic hypoxia and ibuprofen

Chronic postnatal hypoxia increased carotid body size (Fig. 7). Mass-specific carotid body volume of CH rats was more than triple that of Control rats at P9–11 (i.e., a 223% increase in size when pooled across drug treatments; treatment, P<0.001). Ibuprofen did not alter this effect (drug, P=0.74 and treatment × drug, P=0.75).

Fig. 7.

Influence of chronic hypoxia and ibuprofen on the carotid body of P9–11 rats. Rats were exposed to 21% O₂ (Control) or 12% O₂ (CH) from P0 until studied. Individuals received daily injections of saline (6 Control, 5 CH) or 4 mg kg⁻¹ ibuprofen (6 Control, 4 CH). Values are mean ± SEM. ^§ P<0.05 vs. Control (i.e., main effect for CH treatment).

3.3.4. Hematocrit and hemoglobin concentration after chronic hypoxia and ibuprofen

CH increased the hematocrit and hemoglobin concentration in the blood of P9–11 rats (treatment, both P<0.001) (Table 2). Ibuprofen did not alter this effect for hematocrit (drug, P=0.92 and treatment × drug, P=0.51) or hemoglobin concentration (drug, P=0.98 and treatment × drug, P=0.71).

Table 2.

Blood characteristics for neonatal rats after chronic exposures to 21% O₂ (Control) or 12% O₂ (CH) from birth to P9–11 with or without daily ibuprofen treatment (4 mg kg⁻¹, i.p.).

	Control		CH
	Saline	Ibuprofen	Saline	Ibuprofen
n	9	9	10	10
Hematocrit (%)	34.6 ± 0.8	33.9 ± 1.1	47.7 ± 0.9 §	48.3 ± 0.9 §
Hemoglobin (g dl−1)	8.3 ± 0.2	8.4 ± 0.2	10.6 ± 0.3 §	10.5 ± 0.3 §

Values are mean ± SEM.

^§P<0.05 vs. Control (i.e., main effect for hypoxia treatment).

4. Discussion

Neonatal rats exposed to CH had blunted HVR, consistent with previous studies on rats (Eden and Hanson, 1987a; Wyatt et al., 1995; Reeves and Gozal, 2006; Mayer et al., 2014), cats (Hanson et al., 1989a; Hanson et al., 1989b), and sheep (Sladek et al., 1993) exposed to hypoxia during development. This effect was evident whether CH was initiated at birth or at 7 days of age, but the proximate mechanisms responsible for the blunting depended on when CH began. In rats exposed to hypoxia from P0 to P7, the decreased HVR reflected both a smaller initial increase in ventilation (attributable at least in part to diminished carotid body O₂ sensitivity) and a more pronounced hypoxic ventilatory depression in the late phase of the HVR. In contrast, when the exposure occurred from P7 to P14, the decreased HVR was entirely due to a smaller initial increase in ventilation; ventilatory decline during the late phase of the HVR was not altered by CH in this age group. The latter result is consistent with the hypothesis that CH delays maturation of the late phase of the HVR: CH preserved the biphasic HVR in the younger rats, but this effect was absent in older rats that had already passed this developmental milestone. Hypoxia does not prevent maturation entirely since rats reared continuously in hypoxia eventually made the transition to a sustained (albeit blunted) HVR.

The respiratory phenotype of neonatal rats reared in CH shared some similarities with adult rats undergoing comparable CH exposures, but there were also distinct differences. Among the similarities, neonatal rats exhibited greater mass-specific ventilation after their return to normoxia. Persistent hyperpnea / hyperventilation in normoxia is also a hallmark of ventilatory acclimatization in adult mammals (Teppema and Dahan, 2010), including in adult rats exposed to normobaric or hypobaric CH (e.g., (Olson and Dempsey, 1978; Popa et al., 2011)). However, mass-specific ventilation is typically increased at all inspired Po₂ in adult mammals after CH and the acute HVR (i.e., the slope of the response) tends to be increased as well (Teppema and Dahan, 2010; Pamenter and Powell, 2016); these changes are obvious after one week of CH in adult rats (Olson and Dempsey, 1978; Popa et al., 2011). In contrast, hypoxic ventilation was not increased after one week of hypoxia in neonatal rats and was only significantly greater at one time point after two weeks of hypoxia (Fig. 4, left). Indeed, the acute HVR was decreased in neonatal rats after either one or two weeks of hypoxia. Moreover, ibuprofen treatment did not prevent CH from altering the HVR in neonatal rats whereas it blocks the effects of CH on hypoxic ventilation in adult rats (Popa et al., 2011) and humans (Basaran et al., 2016). Therefore, CH appears to have fundamentally different effects on the HVR of neonatal and adult rats.

4.1. Attenuated carotid body responses to hypoxia

The early phase of the HVR is primarily attributed to activation of the carotid body chemoreceptors while the late phase of the HVR reflects a balance between carotid body stimulation and CNS neuromodulation (Bissonnette, 2000; Simakajornboon and Kuptanon, 2005; Teppema and Dahan, 2010). Carotid bodies isolated from chronically hypoxic neonatal rats exhibited reduced O₂ sensitivity in the present study, both in terms of the absolute activity during hypoxia and the slope of the relationship between CSN activity and O₂ level (Fig. 4). Accordingly, diminished carotid body responses to hypoxia should play a central role in the smaller increase in ventilation evident during the first minute of hypoxia, although additional plasticity in the CNS and/or respiratory mechanics cannot be ruled out. This is the opposite of what happens in adult mammals in which CH augments carotid body O₂ sensitivity, ultimately enhancing the HVR (Teppema and Dahan, 2010; Kumar and Prabhakar, 2012; Pamenter and Powell, 2016). The mechanisms underlying the decreased carotid body responses were not assessed here, but previous studies indicate that postnatal CH impairs O₂ sensing at the glomus cell (Wyatt et al., 1995; Sterni et al., 1999) and augments the turnover of dopamine (an inhibitory neuromodulator) in the carotid body (Hertzberg et al., 1992).

Birth is associated with a dramatic rise in arterial Po₂ and it has been proposed that this rise in Po₂ is a key factor in initiating the normal postnatal resetting of carotid body O₂ sensitivity (Carroll and Kim, 2013). It follows that CH might delay the resetting process by perpetuating the low arterial Po₂ levels encountered in utero, and this in turn could explain the diminished carotid body responses in these animals (Sladek et al., 1993; Landauer et al., 1995; Sterni et al., 1999; Carroll and Kim, 2013). This is an attractive hypothesis since artificially increasing arterial Po₂ in fetal sheep has the opposite effect of causing the resetting process to occur earlier (Blanco et al., 1988). If chronic postnatal hypoxia simply delays carotid body maturation, however, one might predict that hypoxia would only attenuate carotid body responses during the early postnatal period before resetting has occurred. On the contrary, one week of hypoxia had virtually identical effects on single-unit carotid chemoafferent responses regardless of whether it was initiated at birth or at P7. Although some studies suggest that carotid body chemoreceptor resetting is not complete until the end of the second postnatal week in rats (Carroll and Kim, 2013), that is not the case for the strain of rats used in the present study. In previous experiments with Sprague-Dawley rats from the same supplier, chemoafferent responses to hypoxia increased over the first postnatal week but showed no further increase during the second postnatal week (Bavis et al., 2011). The data from the present study confirmed these results: the peak responses to hypoxia were the same at P7 and P14 in control rats (compare panels A and B in Fig. 4). Therefore, this carotid body plasticity must involve more than prolongation of the immature state.

Carotid body size increased several-fold during CH, but the functional significance of this hypertrophy is unclear. CH simulates the proliferation of glomus cells in neonatal (Wang and Bisgard, 2005) and adult (Wang and Bisgard, 2002; Wang et al., 2008; Kumar and Prabhakar, 2012) mammals. Glomus cell hyperplasia might increase carotid body responsiveness since more neurotransmitter would be expected to be released near afferent nerve endings, so this seems unlikely to explain the attenuated carotid body responses observed here. However, part of the carotid body hypertrophy is mediated by vasodilation and vascular remodeling (Wang and Bisgard, 2002; Kumar and Prabhakar, 2012). More blood flow (relative to cellular metabolism) could increase local Po₂ and attenuate glomus cell responses in the intact animal, but changes in blood flow cannot explain the blunted carotid body responses in the present study since they were measured using an in vitro, superfused preparation.

4.2. Delayed maturation of the biphasic HVR

Rats reared in 12% O₂ from birth until either P7 (Fig. 1) or P9–11 (Fig. 6) exhibited a strongly biphasic HVR, with the mean ventilation falling below baseline levels by the 10^th minute of hypoxia. In contrast, the HVR generally transitions to a more sustained increase in ventilation around P7 in rats (e.g., (Eden and Hanson, 1987b; Bavis et al., 2010) and present study). This result is consistent with the hypothesis that CH from birth delays the transition from a biphasic HVR to a sustained HVR as suggested by previous studies in rats (Eden and Hanson, 1987a; Joseph et al., 2000) and cats (Hanson et al., 1989a). It should be noted, however, that not all studies have observed this effect during postnatal CH. Reeves and Gozal (2006) exposed rats to 10% O₂ from birth until they were studied at P5, P10, P15, and P30, and at every age the change in ventilation between the early (peak) and late phases of the HVR (i.e., the magnitude of the ventilatory decline) appeared similar to that of rats reared in room air. The reason for the difference between their results and those of other studies (including the present one) is not known.

If chronic hypoxia delays normal maturation of the biphasic HVR, this could be an example of physiological heterokairy (i.e., plasticity in the rate or timing of development) (Spicer and Burggren, 2003). A key question is whether the mechanisms responsible for the biphasic HVR in neonatal rats and kittens exposed to CH are the same as those underlying the biphasic HVR in newborns. In support of this, CH did not cause rats to develop a more biphasic HVR when the chronic exposure began at P7 (i.e., after the HVR had already transitioned to a more sustained phenotype), extending and strengthening the conclusion that hypoxia slows the normal developmental sequence. The late phase of the HVR reflects a balance between the actions of excitatory neurochemicals (e.g., glutamate, nitric oxide) and inhibitory neurochemicals (e.g., platelet-derived growth factor (PDGF), adenosine, GABA) within the brainstem (Bissonnette, 2000; Simakajornboon and Kuptanon, 2005); inhibitory pathways dominate in newborns leading to ventilatory decline and a biphasic HVR. It is possible, for example, that CH prolongs the dominance of inhibitory chemicals (e.g., delaying the downregulation of their receptors), but this awaits further study. Interestingly, chronic postnatal hyperoxia has the opposite effect on the development of the HVR. Specifically, chronic hyperoxia causes the transition to a sustained HVR to occur at an earlier age, and this appears to be associated with a premature upregulation of glutamatergic signaling and downregulation of PDGF signaling (Bavis et al., 2010; Bavis et al., 2014).

Neonatal rats attained a more sustained HVR by P14 despite continuous exposure to CH, indicating that the transition between the biphasic and sustained HVR phenotype was delayed by several days but was not entirely prevented. In contrast, Eden and Hanson (1987a) concluded that CH (13–15% O₂) delayed maturation of the biphasic HVR in rats for at least 5–10 weeks (i.e., for as long as CH was sustained). Some caution must be used when comparing their study to the present one. First of all, Eden and Hanson used different levels of inspired O₂ as “baseline” during ventilation measurements for their treatment groups: control animals experienced a drop in inspired O₂ from 21% O₂ to either 12 or 8% O₂ for measurements of the HVR while the CH rats experienced a much smaller drop from their “baseline” of 13–15% O₂ to 12 or 8% O₂. The use of different baselines complicates comparisons of the relative changes in ventilation between treatment groups. Second, although ventilation decreased below baseline during the late phase of the HVR in rats exposed to CH at P5 (versus remaining slightly above baseline in control animals), the actual decrease in ventilation between the early and late phases of the HVR appears to have been much smaller than in the control rats; again, this could reflect differences in the baseline conditions (i.e., 21% O₂ vs. 13–15% O₂) if CH rats were already in a state of ventilatory depression during the baseline period. Finally, the authors reported that a biphasic HVR was still evident in response to 8% O₂ (but not in response to 12% O₂) in 5–10 weeks old rats exposed continuously to hypoxia. While this result is consistent with delayed maturation of the HVR, no age-matched control group was studied (Eden and Hanson, 1987a), and even adult mammals exhibit some ventilatory roll-off during moderate-to-severe hypoxia (Teppema and Dahan, 2010). Importantly, Eden and Hanson (1987a) did not observe a biphasic HVR in their CH rats at P14 when tested at either 12% O₂ or 8% O₂, consistent with the present study (Fig. 5). Similarly, Joseph et al. (2000) studied the development of rats born and raised at high altitude (3,600 m). These investigators only measured ventilation at the end of the hypoxic challenge (i.e., late phase of the HVR), but they observed ventilatory depression in the youngest rats which indicates the presence of a biphasic HVR. The rats reared in hypoxia were eventually capable of sustaining an increase in ventilation during acute hypoxia, but this occurred one week later than in rats reared as sea level (Joseph et al., 2000). Therefore, most data indicate that chronic hypoxia will not perpetuate the biphasic HVR indefinitely; rather, hypoxia may simply slow the rate at which the developmental program progresses.

4.3. Role of inflammation in the control of breathing after chronic hypoxia

Several lines of evidence link inflammation to the expression of VAH in adult rats. CH increases the numbers and/or activation of immune cells and the expression of inflammatory signaling molecules (e.g., cytokines) in the carotid body and the nucleus tractus solitarius (NTS), a key brainstem region for the central integration of chemoafferent activity, over a similar time course as respiratory changes, and treatments that reduce these biomarkers (e.g., ibuprofen, dexamethasone, minocycline) also block some aspects of VAH in rats and humans (Liu et al., 2009; Popa et al., 2011; Liu et al., 2013; Basaran et al., 2016; Stokes et al., 2017). Although we did not measure cytokines or look for other signs of inflammation in the present study, it has previously been demonstrated that CH also increases cytokine expression in neonatal rats (e.g., (Lam et al., 2008; Radom-Aizik et al., 2013)). Despite this, the same dose of ibuprofen previously used to block CH-induced increases in carotid body O₂ sensitivity and hypoxic ventilation in adult rats (Liu et al., 2009; Liu et al., 2011a; Popa et al., 2011; Liu et al., 2013) had no effect on the HVR of neonatal rats reared in CH. Thus, differences in how inflammation interacts with the respiratory control system at different developmental stages may explain the opposite effects of CH on the HVR of neonatal and adult rats.

Although we failed to demonstrate a role for inflammation in the blunted HVR after postnatal CH exposure using ibuprofen, we cannot rule out a role for inflammation in this plasticity based solely on this negative result. Ibuprofen is a cyclooxygenase (COX) inhibitor that acts as an anti-inflammatory agent by blocking prostaglandin synthesis, and it is not clear that blocking this pathway will prevent all inflammatory processes that might contribute to respiratory neuroplasticity. For example, CH also activates microglia and astrocytes in the brainstem of neonatal and adult rats (MacFarlane et al., 2016; Stokes et al., 2017). Minocylcine, an antibiotic that inhibits glial activation, reduces VAH in adult rats (Stokes et al., 2017). Likewise, minocycline treatment prevented CH from blunting the HVR of neonatal rats (MacFarlane et al., 2016). In the latter study, however, the authors only observed a blunted HVR in rats exposed to CH from P11 to P15 (MacFarlane et al., 2016); equivalent durations of hypoxia (11% O₂) did not affect the HVR if the exposure occurred from P1 to P5 or from P21 to P25 (Mayer et al., 2014). In contrast, we observed blunted HVR in rats exposed to CH during the first or second postnatal week and this blunting was associated with weak carotid body responses in both age groups. Therefore, it remains to be determined what role, if any, glial activation in the CNS plays in the blunted HVR caused by CH during the first postnatal week.

4.4. Body mass and normoxic ventilation after chronic hypoxia

In addition to changes to the HVR, rats exposed to postnatal CH exhibited markedly greater mass-specific ventilation in normoxia as previously reported by others (e.g., (Mortola et al., 1986; Okubo and Mortola, 1988; Reeves and Gozal, 2006)). Changes in body size must be considered here since CH rats also weighed less than age-matched controls. Indeed, the absolute values for tidal volume and minute ventilation (i.e., not normalized to body mass) were similar to those of controls after seven days of CH during either the first or second postnatal week. The similarity of these values in CH and control rats may have been a coincidence, however, since others have observed increases in tidal volume and ventilation in both absolute and mass-specific units after CH (Mortola et al., 1986). Tidal volume and ventilation decreased after longer exposures to CH (≥ 9 d), but they decreased less than body mass which again resulted in an increase in mass-specific ventilation. Since CH also decreases lung size in developing rats (e.g., (Mortola et al., 1986; Mortola et al., 1990; Sant’Anna and Mortola, 2003)), baseline tidal volumes were likely larger relative to lung volume as well; relatively larger tidal volumes should enhance alveolar ventilation since respiratory frequency either stayed the same or increased. Thus, the change in mass-specific ventilation may be more reflective of altered respiratory control in this situation than the change (or lack of change) in absolute ventilation.

CH appears to depress growth in rats through direct effects of hypoxia on the pups rather than through maternal effects (Mortola et al., 1990), likely by decreasing the pups’ appetite, metabolic rate (hypoxic hypometabolism), and body temperature (hypoxic hypothermia) during the exposure period (Mortola et al., 1986; Mortola et al., 1990; Frappell and Mortola, 1994). Ultimately, the slow growth results in lower skeletal mass, smaller internal organs (e.g., brain, lung, heart, kidney, liver, thymus), and a shorter body length (Mortola et al., 1990; Sant’Anna and Mortola, 2003). Depressed growth of metabolically active tissues may be beneficial since it decreases the demand for oxygen during the CH exposure. Conversely, the hyperpnea observed after return to normoxia may reflect increased metabolic demands during a period of catch-up growth (Sant’Anna and Mortola, 2003). That said, others have shown that the neonatal and adult rats hyperventilate when studied shortly after return to normoxia (i.e., the increase in ventilation exceeds any increase in metabolic rate) (e.g., (Olson and Dempsey, 1978; Frappell and Mortola, 1994; Popa et al., 2011). Thus, persistent hyperventilation may also reveal underlying neural plasticity. In the present study, baseline carotid chemoafferent activity tended to be greater after CH, consistent with increased peripheral respiratory drive. CNS contributions remain to be investigated in neonates, but both peripheral and central mechanisms have been implicated in normoxic hyperventilation during VAH in adult mammals (Dempsey et al., 2014).

4.5. Implications

The present study demonstrates that chronic hypoxia from birth attenuates carotid body responses to hypoxia and delays the maturation of the biphasic HVR, ultimately resulting in a blunted HVR. This has implications for human infants that experience chronic hypoxemia in the neonatal period (e.g., high altitude, preterm birth, chronic lung disease, congenital heart defects). Specifically, abnormal carotid body function may increase the frequency and/or duration of apneas, impair arousal responses, and contribute to sudden infant death (Darnall, 2013; Gauda et al., 2013; MacFarlane et al., 2013; Porzionato et al., 2013). Importantly, hypoxic ventilatory depression associated with prolonged expression of the biphasic HVR may exacerbate these effects during a particularly vulnerable period of postnatal development (Wong-Riley et al., 2019). To our knowledge, no studies have investigated the effects of chronic hypoxemia on the maturation of the biphasic HVR in human infants; most studies of the control of breathing in newborns utilize very brief hypoxic or hyperoxic challenges to assess carotid body function (and thus provide data relevant only to the early phase of the HVR). The potential for oxygen levels to influence the rate of development of protective ventilatory chemoreflexes (i.e., heterokairy) warrants further investigation.

Highlights.

• Rats were exposed to chronic hypoxia (CH) for the first or second postnatal week.

• CH blunted the hypoxic ventilatory response (HVR) in both age groups.

• The blunted HVR was associated with diminished carotid body O₂ sensitivity.

• CH during the first (but not second) week delayed maturation of the biphasic HVR.

• The anti-inflammatory drug ibuprofen did not block this respiratory plasticity.