Respiratory Plasticity Induced by Chronic Hyperoxia in Juvenile and Adult Rats

Abstract

Chronic hyperoxia during early postnatal development depresses breathing when neonatal rats are returned to room air and causes long-lasting attenuation of the hypoxic ventilatory response (HVR). In contrast, little is known about the control of breathing of juvenile or adult mammals after chronic exposure to moderate hyperoxia later in life. Therefore, Sprague-Dawley rats were exposed to 60% O₂ for 7 days (juveniles) or for 4 and 14 days (adults) and ventilation was measured by whole-body plethysmography immediately after the exposure or following a longer period of recovery in room air. Hyperoxia-treated juvenile rats appeared to hypoventilate when returned to room air (11-13% lower ventilation and CO₂ convection requirement relative to age-matched controls), but chronic hyperoxia did not alter normoxic ventilation in adult rats. In contrast, pre-treatment with chronic hyperoxia augmented the HVR in both juvenile rats (+41%) and adult rats (+28–50%). The hypercapnic ventilatory response (7% CO₂) also tended to be augmented in adult rats after 14 days of hyperoxia, but this effect was not significant after accounting for variation in metabolic rate (i.e, CO₂ convection requirement). These findings confirm that chronic hyperoxia elicits age-specific respiratory plasticity in rats. These age-dependent differences are not caused by a lack of plasticity in adult-exposed rats; rather, there are qualitative differences in the plasticity that is expressed after chronic hyperoxia in neonates, juveniles, and adults as well as differences in its persistence.

1. Introduction

Inspired oxygen influences breathing across several time domains. Whereas acute and chronic hypoxia generally stimulate breathing in mature mammals (Teppema and Dahan, 2010; Pamenter and Powell, 2016; Bavis et al., 2023), ventilatory responses to hyperoxia are complex. It is well established that acute exposure to hyperoxia causes a transient decrease in ventilation as high arterial O₂ partial pressures (Po₂) silence peripheral chemoreceptors, but this ventilatory depression gradually transitions to hyperventilation if hyperoxia continues for several minutes (Bavis, 2020; Dean and Stavitzski, 2022). Relatively brief exposures to hyperoxia (minutes – hours) may also induce respiratory plasticity that alters the control of breathing even after the return to normoxia. For example, ventilation remains elevated for at least 15 minutes after return to normoxia in adult humans exposed to 30 minutes of isocapnic hyperoxia (30 – 75% O₂) (Becker et al., 1995; Becker et al., 1996). Moreover, pre-exposure to 10 – 15 minutes of 100% O₂ augments the acute hypoxic ventilatory response (HVR) in adult humans (Honda et al., 1996), rats (Gozal, 1998; Pokorski et al., 2005; Budzinska and Ilasz, 2008), and cats (Pokorski et al., 2004), while pre-exposure to 4 hours of 30% O₂ augments the HVR in goats (Herman et al., 2001).

Respiratory plasticity also occurs after chronic exposures to moderate hyperoxia lasting several days to weeks, at least in developing mammals. In rats, chronic exposure to 30 – 60% O₂ during a critical period of early postnatal development impairs carotid body development and may permanently blunt the acute HVR (Ling et al., 1996, 1997; Erickson et al., 1998; Bavis et al., 2002; Fuller et al., 2002; Bavis et al., 2013); there do not appear to be significant long-lasting effects on normoxic ventilation or the hypercapnic ventilatory response (HCVR) (reviewed in Bavis, 2020). Similar phenotypes have been reported in neonatal and adult humans that received supplemental O₂ following preterm birth (Katz-Salamon and Lagercrantz, 1994; Katz-Salamon et al., 1995; Bates et al., 2014; Bates et al., 2018), although a causal link to postnatal hyperoxemia has not been firmly established. In contrast, exposing adult rats to similar durations of moderate hyperoxia does not affect carotid body size or cause long-lasting attenuation of the HVR (Ling et al., 1996, 1997; Bavis et al., 2013). Indeed, after accounting for variation in metabolic rate, rats exposed as adults to one month of 60% O₂ may have somewhat enhanced hypoxic responses when studied after two months of recovery in room air (Ling et al., 1996).

Although a few studies have investigated whether four weeks of hyperoxia (60% O₂) during adulthood has long-lasting consequences on the control of breathing (Ling et al., 1996, 1997), almost nothing is known about the control of breathing of adult mammals immediately following chronic exposures to moderate hyperoxia (Bavis, 2020). This is an important gap in knowledge since there are several changes in breathing in neonatal rats immediately after chronic hyperoxia that either do not persist into adulthood or are apparent only after shorter exposures to hyperoxia. For example, neonatal rats hypoventilate immediately after return to normoxia, but this effect is more pronounced in younger neonates (e.g., 4 vs. 14 days of age) and ventilation returns to normal within 24 hours (Bavis et al., 2010; Bavis et al., 2014; Kavanagh et al., 2018). There is also evidence for excitatory CNS plasticity during the biphasic HVR of neonatal rats after 4 days of chronic hyperoxia that is no longer evident after 14 days. Notably, the early phase of the HVR is generally normal in 4-day-old rats reared in hyperoxia despite well-developed carotid body impairment, while there is good agreement between carotid body dysfunction and blunting of the HVR by 14 days of age (Bavis et al., 2013). Finally, while prior studies demonstrated distinct effects of chronic hyperoxia on the adult HVR depending on whether the exposure occurs during postnatal development or in adulthood (i.e., the HVR is blunted or not, respectively), it is unclear whether this reflects qualitative differences in the immediate expression of respiratory plasticity versus differences in the capacity for spontaneous recovery. In other words, the HVR could be blunted in adult rats immediately after chronic hyperoxia even if this plasticity is no longer evident several weeks later. Indeed, severe hyperoxia (i.e., several hours – days in 100% O₂ at normobaric or hyperbaric pressures) reversibly blunts the HVR of adult rats in a dose-dependent manner (Arieli et al., 1988; Liberzon et al., 1989; Waisman et al., 1992; Arieli, 1994). However, blunting of the HVR following extremely high Po₂ (≥ 1 ATA) may reflect overt O₂ toxicity (e.g., carotid body injury secondary to excessive reactive oxygen species production) rather than neural plasticity per se (Bavis and MacFarlane, 2017).

The present study investigated whether chronic exposure to moderate, normobaric hyperoxia (60% O₂) induces plasticity in the control of breathing in juvenile and adult rats. We hypothesized that pre-treatment with chronic hyperoxia might elicit changes in breathing similar to those previously reported for neonatal rats – i.e., hypopnea / hypoventilation in normoxia and blunted HVR – but that these effects would diminish over time once room air was restored. Accordingly, normoxic breathing (21% O₂ / 0% CO₂) and ventilatory responses to hypoxia and/or hypercapnia were measured in juvenile and adult rats immediately following the chronic hyperoxia exposure and after a period of recovery in room air.

2. Methods

2.1. Experimental animals and chronic hyperoxia exposures

Experiments were completed on juvenile and adult Sprague Dawley rats (Crl:SD) either born to late-gestation dams obtained from Charles River Laboratories (Wilmington, MA, USA) or obtained as adults from the same supplier. For chronic exposures, standard rat cages (2 rats per cage) were placed inside environmental chambers flushed with either room air or a mixture of room air and O₂ at sufficient flow rates to maintain the desired levels of O₂ (21% O₂ for Control or 60% O₂ for Hyperoxia) while minimizing accumulation of CO₂ (< 0.4% CO₂). Chambers were opened briefly every 3 – 4 d for routine cage cleaning and to remove individuals for study, as needed.

Four sets of experiments were completed; the number of rats in each set are provided in the relevant tables and figure legends. The primary objective of experiment # 1 was to assess the effects of pre-treatment with chronic hyperoxia on hypoxic ventilatory and metabolic responses in juvenile rats. Accordingly, juvenile rats were placed into the environmental chambers for 7 d beginning between 27 and 30 d of age (i.e., after the critical period for hyperoxia-induced developmental plasticity is expected to have closed (Bavis et al., 2002; Bavis et al., 2013)). Hypoxic responses were measured immediately following this exposure and again after rats were returned to room air for 7 d. The observed changes in the hypoxic responses of juvenile rats were unexpected, prompting additional experiments to characterize the effects of chronic hyperoxia on respiratory control in adult rats; experiment #2 focused on the effect on adult hypoxic responses, experiment #3 focused on adult hypercapnic responses, and experiment #4 focused on normoxic ventilation. Prior work in neonatal rats indicated that the effects of chronic hyperoxia may change over the course of the exposure, with some effects being more evident after a 4-d exposure to hyperoxia and others being more evident after a 14-d exposure (Bavis et al., 2010; Bavis et al., 2014). Therefore, in experiments #2 – 4, adult rats (assumed >2.5 months of age based on body mass) were placed into the environmental chambers for a total of 14 d, with breathing and metabolism measurements taking place on the fourth and fourteenth days. Rats were then returned to room air and measurements were repeated after 27 – 31 d (experiment #3) or 4 d (experiment #4) of recovery.

Throughout the study, rats were provided with food and water ad libitum and maintained on a 12:12-h light-dark cycle. All procedures were approved by the Institutional Animal Care and Use Committees at Bates College.

2.2. Surgical preparation

In rats studied as adults (experiments #2 – 4), body temperature (T_b) was measured during the ventilation measurements using temperature-sensitive E-mitter transponders (G2 or PTD 4000; Starr Life Sciences, Oakmont, PA, USA) or RFID microchip transponders (UCT-2112; Unified Information Devices (UID), Lake Villa, IL, USA). E-mitter transponders were surgically implanted into the abdominal cavity at least one week prior to initiating chronic hyperoxia. Rats were anesthetized with isoflurane (~2.5% isoflurane, balance O₂) and transponders were placed into the abdominal cavity through a ventral, midline incision. Carprofen (5 mg kg⁻¹, s.c.; Pfizer Animal Health, Exton, PA, USA) was administered as an analgesic. For implantation of microchip transponders, rats were briefly anesthetized (~2.5% isoflurane, balance O₂) and the transponders were injected through a sterile needle. Although the intention was to place all microchip transponders intraperitoneally, some were discovered to be subcutaneous at the end of the experiment; however, there was no significant difference in average T_b between rats with intraperitoneal versus subcutaneous microchip transponders or between microchip transponders and E-mitter transponders (data not shown).

2.3. Ventilation measurements in juvenile rats (Experiment #1)

Breathing was measured in juvenile rats using a custom-built, whole-body plethysmograph. Individuals were placed into a clear acrylic chamber (14 cm i.d., ~2.3 L). Gas mixtures were directed through the chamber at 1.5 L min⁻¹ (STPD) using a mass flow controller (MFC-4; Sable Systems, Las Vegas, NV, USA) and valves (series 840; Sierra Instruments, Monterey, CA, USA). To ensure that air within the chamber was saturated with water vapor, the incurrent gas mixture was bubbled through distilled water just prior to entering the chamber and distilled water (~ 250 ml) covered the floor beneath a platform on which the rat rested. When the chamber was sealed, respiratory-related pressure fluctuations within the chamber were measured with a differential pressure transducer (DP45; Validyne Engineering, Northridge, CA, USA); the system was calibrated prior to individual trials by repeated 0.1 ml air injections. Pressure fluctuations and chamber temperature (21-23°C; IT-18; Physitemp Instruments, Clifton, NJ, USA) were recorded to a computer (PowerLab 8SP and LabChart 7; ADInstruments, Colorado Springs, CO, USA). Body temperature (T_b) was measured using a rectal probe (RET-3 probe and BAT-12 microprobe thermometer; Physitemp, Clifton, NJ, USA) approximately 30 min after sealing the rat into the chamber (for baseline calculations) and again at the end of the hypoxia exposure (for hypoxia calculations). These measurements were used to calculate tidal volume (Vt; ml (BTPS); (Drorbaugh and Fenn, 1955)), respiratory frequency (fr; breaths min⁻¹), and their product, minute ventilation ($\stackrel{.}{\mathrm{V}}I$; ml min⁻¹); volumes were then normalized to body weight.

All measurements were made during the light portion of the light-dark cycle. Individuals were weighed and placed into the chamber flushed with 21% O₂ (balance N₂) for at least one hour; the chamber was opened briefly for T_b measurements midway through this adjustment period (see above). When the rat was resting quietly, the chamber was briefly sealed (approximately 60 – 90 s) by clamping the inlet and outlet ports to record respiratory-related pressure fluctuations. The objective was to obtain 45-60 s of data free from movement artifacts and irregular breaths (e.g., sniffs, sighs), and measurements were repeated if necessary. Individuals then were exposed to 12% O₂ (balance N₂) for 20 min before re-sealing the plethysmograph chamber to assess breathing.

2.4. Ventilation measurements in adult rats (Experiments #2 – 4)

Breathing was measured in adult rats using a commercially available whole-body barometric plethysmograph system (Data Sciences International (DSI), St. Paul, MN, USA). Gas flow through the plethysmograph chamber (P/N 601-1427-001; DSI) was set at 3 L min⁻¹ (STPD) using a gas-mixing mass flow controller (MFC-4; Sable Systems) and valves (series 840; Sierra Instruments, Monterey, CA, USA). This flow was monitored with an integrated pneumotach and a differential pressure transducer (Buxco TRD5700; DSI) that was calibrated prior to each animal with known gas flow rates (series 4100 mass flow meter; TSI, Shoreview, MN, USA). Respiratory-related oscillations in air flow, temperature and humidity inside the chamber (temp/humidity probe, P/N 601-2249-001; DSI), and body temperature (see below) were then used to calculate Vt, fr, and $\stackrel{.}{\mathrm{V}}I$ (Ponemah software version 5.20; DSI) in one-minute bins every fifth minute. Movement artifacts and irregular breaths (e.g., sniffs, sighs) were manually removed if not already excluded by the software filters; if there were fewer than 20 usable breaths in a bin, the preceding or subsequent minute was used instead.

Individual rats were weighed and sealed into the plethysmograph chamber during the light portion of the light:dark cycle and exposed to 21% O₂ (balance N₂) for at least one hour. In experiment #2, breathing was measured during the final 15 min of this baseline period and then during a 15-min exposure to 12% O₂. In experiment #3, breathing was measured during the final 15 min of the baseline period (21% O₂, 0% CO₂, balance N₂) and then during a 20-min exposure to 7% CO₂ (21% O₂, balance N₂) to assess the hypercapnic ventilatory response. Once measurements were completed after 14 d of chronic hyperoxia, individuals were returned to room air and re-studied after approximately 4 weeks of recovery (27 – 31 d). During these trials, breathing was measured during the final 15 min of the baseline period (21% O₂ / 0% CO₂) and then during 15-min exposures to hypercapnia (21% O₂ / 7% CO₂), recovery (21% O₂ / 0% CO₂), and hypoxia (12% O₂ / 0% CO₂); all individuals received test gases in the same order. In experiment #4, normoxic breathing was measured on the same individuals four times: just prior to being placed into the Control or Hyperoxia treatments (Pre), after 4 and 14 d of the treatment, and 4 d after return to room air (Recovery). During each of these trials, individuals were exposed to 21% O₂ (balance N₂) for a total of 90 min and breathing was recorded every fifth minute during the final 60 min of this period.

Body temperature (T_b) is required for Vt calculations in whole-body plethysmography, so T_b was recorded continuously from implanted transponders during breathing measurements using a receiver placed beneath the plethysmograph chamber (E-mitter transponders: ER-4000 and VitalView 4.1 software, Starr Life Sciences; microchip transponders: URC-500HP and Matrix software, UID). In male rats, microchip transponders placed intraperitoneally occasionally migrated into the scrotum between implantation and the breathing measurements resulting in erroneously low T_b values. For several individuals (2 Control and 2 Hyperoxia in experiment #3), T_b was instead determined using a rectal probe (RET-3 probe and BAT-12 microprobe thermometer; Physitemp Instruments). Rectal T_b was measured immediately prior to sealing the individual into the chamber (baseline) and at the end of the final test gas; since we did not open the plethysmograph chamber during intermediate conditions, we did not calculate breathing parameters for hypercapnia for these individuals for the trials in which individuals were exposed to both hypercapnia and hypoxia (i.e., after one month of recovery in room air). Individuals were excluded from the study if transponders malfunctioned and rectal T_b was not measured.

2.5. Metabolic rate measurements

The plethysmograph chamber doubled as a respirometer chamber for the purpose of assessing metabolic rate alongside breathing measurements in experiments #1 – 3; metabolic rate was not measured in experiment #4. Fractional concentrations of O₂ and CO₂ in the air entering (FiO₂ and FiCO₂) and exiting (FeO₂ and FeCO₂) the chamber were measured with a calibrated gas analyzer (ML206; AD Instruments); air was dried (Direrite, W. A. Hammond Drierite Co., Xenia, OH, USA) before passing through the gas analyzer. In experiment #1, FeO₂ and FeCO₂ were recorded for 30-60 s just prior to sealing the plethysmograph chamber for breathing measurements and FiO₂ and FiCO₂ were recorded while the chamber was sealed. In experiments # 2 and 3, immediately following the final time point for baseline, hypoxia, or hypercapnia breathing measurements, a small port on the side of the DSI plethysmograph chamber was opened to draw gas into the gas analyzer; these samples (30-60 s) were taken as FeO₂ and FeCO₂ under the assumption that gases inside the chamber were well mixed. FiO₂ and FiCO₂ were then measured by subsampling gas directly from the incurrent line before switching to the next condition (hypoxia or hypercapnia) or removing the rat from the chamber.

O₂ and CO₂ measurements were then used with the incurrent gas flow rate to calculate O₂ consumption ($\stackrel{.}{\mathrm{V}}{O}_2$) and CO₂ production ($\stackrel{.}{\mathrm{V}}{\text{CO}}_2$) with the equations from Frappell et al. (1992).

2.6. Calculations and statistical analysis

In experiments #2 – 4, where breathing parameters were determined in 1-min bins every fifth minute, T_b, Vt, fr, and $\stackrel{.}{\mathrm{V}}I$ were calculated as the average of two consecutive bins. In experiments # 2 and 3, the average of the final two bins from each test condition was used. For experiment #4, the moving average of two bins was calculated over the 60-min test for each individual and the averages (T_b, Vt, fr, and $\stackrel{.}{\mathrm{V}}I$) corresponding to the lowest value for $\stackrel{.}{\mathrm{V}}I$ were selected; the moving average of three bins yielded qualitatively similar results (data not shown). Respiratory exchange ratio (RER) was calculated as $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$ / $\stackrel{.}{\mathrm{V}}{O}_2$. The air convection requirement was calculated as $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{O}_2$ and $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$; for experiments # 2 and 3, $\stackrel{.}{\mathrm{V}}I$ was taken from the final time point (i.e., immediately preceding FeO₂ and FeCO₂ measurements) for these calculations.

Body mass, T_b, and baseline breathing and metabolism parameters were compared between treatment groups and sexes using two-way ANOVA (experiments #1 – 3). In experiment #4, where measurements prior to commencing the treatment (i.e., “Pre” measurements) meant that each individual effectively served as their own control, data for females and males were pooled due to relatively small sample sizes. Accordingly, body mass, T_b, and baseline breathing and metabolism parameters were compared between treatments over time (Pre, 4 d, 14 d, and Recovery) by two-way repeated measures ANOVA; Tukey’s multiple comparison test was used where significant Treatment × Time interactions were detected, and Dunnett’s multiple comparison test was used to compare values to the “Pre” value where only significant effects of Time were detected.

In juvenile rats studied immediately after 7 d of hyperoxia, preliminary analysis revealed that sex had no effect on ventilatory or metabolic responses to hypoxia, so data for female and male rats were pooled and hypoxic responses were compared between treatment groups by two-way repeated measures ANOVA with one-factor repetition (FiO₂); Fisher’s LSD post hoc test was used where significant Treatment × FiO₂ interactions were detected. For most other data sets, ventilatory and metabolic responses to hypoxia and/or hypercapnia were compared between treatment groups and between sexes by three-way repeated measures ANOVA with one-factor repetition (FiO₂ or FiCO₂). However, three-way mixed-effects models were used rather than repeated measures ANOVA where data were incomplete for some individuals (i.e., $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$ and $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$ responses to hypoxia in experiment #2 and ventilatory and metabolic responses measured after approximately one month of recovery in experiment #3). Where significant interactions between factors were detected, Šídák's multiple comparisons tests were used to compare means between groups differing by only one factor.

All statistical tests were run using Prism 10 (GraphPad Software, Boston, MA, USA). P-values from the ANOVA or mixed-effects models are presented in the text along with the corresponding main effect (e.g., “Treatment” or “Sex”) or interaction (e.g., “Treatment × Sex”) term. Statistical significance was determined at P<0.05. Values in the text are reported as mean ± SEM.

3. Results

3.1. Chronic hyperoxia in juvenile rats (Experiment #1)

3.1.1. Body mass

Chronic hyperoxia did not affect the body mass of juvenile rats (Table 1). Hyperoxia and Control rats had similar weights immediately following the 7-d exposure and after 7 d of recovery in room air (Treatment and Treatment × Sex, all P>0.05). As expected, female rats weighed less than male rats at both time points (Sex, both P<0.001).

Table 1.

Baseline (21% O₂) ventilatory and metabolic measurements for juvenile rats immediately following 7 d of chronic hyperoxia and after 7 d of recovery in room air (Experiment #1).

	Female		Male		Statistical comparisons a
	Control	Hyperoxia	Control	Hyperoxia	Treatment (comparison, P)	Sex (comparison, P)
Immediately post-exposure:
n	8	8	10	10
Age (d)	36±0.3	36±0.2	35±0.1	35±0.2
Mass (g)	121±5	127±4	146±4	147±2	n.s.	F<M, <0.001
fr (breaths min−1)	131.1±9.2	127.9±9.1	132.5±6.2	138.4±7.4	n.s.	n.s.
Vt (ml 100g−1)	0.82±0.02	0.71±0.03	0.76±0.03	0.67±0.02	H<C, <0.001	n.s.
$\stackrel{.}{\mathbf{V}}I$ (ml min−1 100g−1)	106.1±5.4	88.9±4.1	101.5±6.2	91.8±3.1	H<C, 0.010	n.s.
$\stackrel{.}{\mathbf{V}}{O}_2$ (ml min−1 100g−1)	3.17±0.21	3.05±0.13	3.12±0.13	3.37±0.25	n.s.	n.s.
$\stackrel{.}{\mathbf{V}}{\mathrm{CO}}_2$ (ml min−1 100g−1)	3.06±0.24	2.80±0.10	2.91±0.12	3.03±0.12	n.s.	n.s.
RER	0.96±0.03	0.92±0.03	0.93±0.02	0.91±0.03	n.s.	n.s.
$\stackrel{.}{\mathbf{V}}I$ / $\stackrel{.}{\mathbf{V}}{O}_2$	33.9±1.1	29.5±1.6	32.6±1.7	27.9±1.3	H<C, 0.005	n.s.
$\stackrel{.}{\mathbf{V}}I$ / $\stackrel{.}{\mathbf{V}}{\mathrm{CO}}_2$	35.4±1.5	31.9±1.4	35.0±2.0	30.5±0.8	H<C, 0.014	n.s.
Tb (°C)	37.5±0.1	37.6±0.1	37.4±0.1	37.4±0.1	n.s.	n.s.
After 7 d in room air:
n	8	8	10	10
Age (d)	43±0.3	43±0.2	42±0.1	42±0.2
Mass (g)	157±6	158±5	199±5	204±4	n.s.	F<M, <0.001
fr (breaths min−1)	115.3±4.9	124.6±4.3	126.5±4.9	127.5±4.4	n.s.	n.s.
Vt (ml 100g−1)	0.82±0.03	0.72±0.02	0.68±0.02	0.62±0.03	H<C, 0.004	F>M, <0.001
$\stackrel{.}{\mathbf{V}}I$ (ml min−1 100g−1)	95.2±6.5	89.7±3.2	85.9±4.1	78.8±2.6	n.s.	F>M, 0.023
$\stackrel{.}{\mathbf{V}}{O}_2$ (ml min−1 100g−1)	3.32±0.14	2.99±0.11	2.86±0.11	3.04±0.14	n.s.	n.s.
$\stackrel{.}{\mathbf{V}}{\mathrm{CO}}_2$ (ml min−1 100g−1)	3.06±0.10	2.81±0.11	2.71±0.12	2.73±0.11	n.s.	n.s.
RER	0.93±0.02	0.94±0.02	0.95±0.01	0.90±0.01	n.s.	n.s.
$\stackrel{.}{\mathbf{V}}I$ / $\stackrel{.}{\mathbf{V}}{O}_2$	28.8±1.9	30.2±1.4	30.4±1.8	26.3±1.2	n.s.	n.s.
$\stackrel{.}{\mathbf{V}}I$ / $\stackrel{.}{\mathbf{V}}{\mathrm{CO}}_2$	31.1±2.0	32.0±1.2	32.0±1.7	29.2±1.2	n.s.	n.s.
Tb (°C)	37.7±0.1	37.7±0.2	37.4±0.1	37.3±0.1	n.s.	F>M, 0.005

Values are mean±SEM. fr, respiratory frequency; Vt, tidal volume; $\stackrel{.}{\mathrm{V}}I$, minute ventilation; $\stackrel{.}{\mathrm{V}}{O}_2$, O₂ consumption; $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$, CO₂ production; RER, respiratory exchange ratio; $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{O}_2$, O₂ convection requirement; $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$, CO₂ convection requirement; T_b, body temperature.

^aNo significant Treatment × Sex interactions were detected (all P>0.05), so only main effects from the ANOVA model are reported. n.s., nonsignificant; F, Female; M, Male; C, Control; H, Hyperoxia.

3.1.2. Ventilatory and metabolic responses to acute hypoxia

Hyperoxia rats appeared to hypoventilate immediately following the 7-d exposure (Table 1). Baseline (21% O₂, balance N₂) Vt was lower in Hyperoxia rats (Treatment, P<0.001), resulting in approximately 11-13% lower $\stackrel{.}{\mathrm{V}}I$ and $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$ (Hyperoxia, P=0.010 and 0.014, respectively). When acutely exposed to hypoxia, Hyperoxia and Control rats increased $\stackrel{.}{\mathrm{V}}I$ and reduced $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$, resulting in an overall increase in $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$ (Fig. 1). However, the increases in $\stackrel{.}{\mathrm{V}}I$ and $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$ were greater in Hyperoxia rats (Treatment × FiO₂, P=0.018 and 0.038, respectively). Although the absolute values of $\stackrel{.}{\mathrm{V}}I$ and $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$ did not differ between treatment groups while breathing 12% O₂, their changes from baseline were 41% and 16% greater, respectively, in Hyperoxia rats due to the lower starting values (Δ $\stackrel{.}{\mathrm{V}}I$ = 34.0±4.2 vs. 24.1±3.6 ml min⁻¹ 100g⁻¹; Δ $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$ = 23.7±1.9 vs. 20.5±2.2). The greater hypoxic response was primarily explained by a greater Vt response in Hyperoxia rats, although the fr response also tended to be greater (Appendix A, Fig. S1). Chronic hyperoxia did not alter the metabolic response to hypoxia (Treatment × FiO₂, P=0.944) (Fig. 1B).

Fig. 1.

Ventilatory and metabolic responses to acute hypoxia in juvenile rats after exposure to 7 d of 60% O₂ (Hyperoxia; n=18) or 21% O₂ (Control; n=18). Values (mean±SEM) are shown for (A) minute ventilation ($\stackrel{.}{\mathrm{V}}I$), (B) CO₂ production ($\stackrel{.}{\mathrm{V}}{\text{CO}}_2$), and (C) CO₂ convection requirement ($\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$) at baseline (21% O₂) and during hypoxia (12% O₂). P-values are provided for the Treatment (Tr) × FiO₂ interaction from the ANOVA models; if the interaction was not significant, the main effect for Treatment is also provided. * P<0.05 vs. Control at the same FiO₂ in the post hoc analysis.

3.1.3. Recovery of hypoxic responses

Rats were restudied 7 d after return to room air. Baseline $\stackrel{.}{\mathrm{V}}I$ and $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$ had returned to normal in Hyperoxia rats (Treatment and Treatment × Sex, all P>0.05), but Vt continued to be a bit lower than in Controls (Treatment, P=0.004) (Table 1). Although sex did not affect normoxic breathing or metabolic rate when measured the prior week (Table 1), females had significantly greater Vt (Sex, P<0.001) and $\stackrel{.}{\mathrm{V}}I$ (Sex, P=0.023, respectively). Therefore, sex was included in the statistical model when assessing hypoxic responses after 7 d of recovery. Juvenile male rats tended to have a greater $\stackrel{.}{\mathrm{V}}I$ response than females independent of the treatment group (Sex × FiO₂, P=0.038) (Fig. 2A). After accounting for sex, the HVR was similar between Hyperoxia and Control rats (Treatment × FiO₂, P=0.075 and 0.313 for $\stackrel{.}{\mathrm{V}}I$ and $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$, respectively) (Fig. 2). Corresponding data for Vt and fr responses are provided in Appendix A (Fig. S2).

Fig. 2.

Ventilatory and metabolic responses to acute hypoxia after 7 d of recovery in room air for juvenile rats previously exposed to 7 d of either 60% O₂ (Hyperoxia) or 21% O₂ (Control). Values (mean±SEM) are shown for (A) minute ventilation ($\stackrel{.}{\mathrm{V}}I$), (B) CO₂ production ($\stackrel{.}{\mathrm{V}}{\text{CO}}_2$), and (C) CO₂ convection requirement ($\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$) at baseline (21% O₂) and during hypoxia (12% O₂). Sample sizes (female, male) are 18 Hyperoxia (8, 10) and 18 Control (8, 10). P-values are provided for the Treatment (Tr) × FiO₂ and Sex × FiO₂ interactions from the ANOVA models; if either of these interactions was not significant, the main effect for Treatment or Sex is also provided. All Tr × Sex and Tr × Sex × FiO₂ interactions were nonsignificant.

3.2. Chronic hyperoxia in adult rats

The augmented HVR of juvenile rats after pre-treatment with chronic hyperoxia (section 3.1.2) was opposite to the hypothesized effect. To determine whether this plasticity was unique to juvenile rats or might also be expressed in adults, several groups of adult rats were exposed to 60% O₂ for 4 d and 14 d.

3.2.1. Body mass

Chronic hyperoxia did not affect the body mass of adult rats. Hyperoxia rats weighed approximately the same as Control rats immediately following 4 d and 14 d of exposure to 60% O₂ in experiments # 2 and 3 (Treatment and Treatment × Sex, all P>0.05) (Table 2). As expected, female rats weighed less than male rats at both time points (Sex, both P<0.001). In experiment #4, the same individuals were weighed prior to being placed into the environmental chambers for the chronic exposures, after 4 d and 14 d of exposure to 60% O₂, and after 4 d of recovery in room air. Although individuals gained mass over the course of the 18 days of the experiment (Time, P=0.009), chronic hyperoxia did not alter this growth (Treatment, P=0.819 and Treatment × Time, P=0.648) (Table 3).

Table 2.

Baseline (21% O₂ / 0% CO₂) ventilatory and metabolic measurements for adult rats immediately following 4 d and 14 d of chronic hyperoxia (Experiments #2 and 3 combined).

	Female		Male		Statistical comparisons a
	Control	Hyperoxia	Control	Hyperoxia	Treatment (comparison, P)	Sex (comparison, P)
After 4 d in hyperoxia:
n	26	23	18	20
Mass (g)	240±7	234±4	372±13	386±14	n.s.	F<M, <0.001
Tb (°C)	37.2±0.1	37.0±0.1	37.3±0.1	37.1±0.1	H<C, 0.047	n.s.
fr (breaths min−1)	65.3±2.1	63.4±2.2	74.7±2.1	75.3±2.9	n.s.	F<M, <0.001
Vt (ml 100g−1)	0.49±0.01	0.50±0.02	0.35±0.01	0.35±0.01	n.s.	F>M, <0.001
$\stackrel{.}{\mathbf{V}}I$ (ml min−1 100g−1)	31.3±0.8	31.8±1.4	26.0±0.8	25.9±0.9	n.s.	F>M, <0.001
n b	17	14	13	14
$\stackrel{.}{\mathbf{V}}{O}_2$ (ml min−1 100g−1)	1.84±0.11	1.73±0.0.09	1.59±0.07	1.67±0.10	n.s.	n.s.
$\stackrel{.}{\mathbf{V}}{\mathrm{CO}}_2$ (ml min−1 100g−1)	1.48±0.08	1.43±0.08	1.40±0.06	1.51±0.08	n.s.	n.s.
RER	0.82±0.02	0.83±0.02	0.88±0.02	0.91±0.02	n.s.	F<M, <0.001
$\stackrel{.}{\mathbf{V}}I$ / $\stackrel{.}{\mathbf{V}}{O}_2$	17.4±1.1	18.9±1.3	17.2±1.1	16.9±1.2	n.s.	n.s.
$\stackrel{.}{\mathbf{V}}I$ / $\stackrel{.}{\mathbf{V}}{\mathrm{CO}}_2$	21.4±1.2	22.5±1.3	19.5±1.1	18.5±1.2	n.s.	F>M, 0.017
After 14 d in hyperoxia:
n	26	23	19	21
Mass (g)	249±7	243±4	382±13	393±12	n.s.	F<M, <0.001
Tb (°C)	37.3±0.1	37.2±0.2	37.4±0.1	36.8±0.1	H<C, 0.008	n.s.
fr (breaths min−1)	68.0±2.5	57.5±1.7	70.4±2.2	68.2±2.7	H<C, 0.008	F<M, 0.007
Vt (ml 100g−1)	0.50±0.01	0.55±0.02	0.37±0.01	0.40±0.03	H>C, 0.027	F>M, <0.001
$\stackrel{.}{\mathbf{V}}I$ (ml min−1 100g−1)	33.2±1.3	31.0±1.2	25.8±0.6	26.7±1.2	n.s.	F>M, <0.001
n b	18	16	15	19
$\stackrel{.}{\mathbf{V}}{O}_2$ (ml min−1 100g−1)	1.66±0.10	1.87±0.10	1.65±0.11	1.63±0.08	n.s.	n.s.
$\stackrel{.}{\mathbf{V}}{\mathrm{CO}}_2$ (ml min−1 100g−1)	1.46±0.07	1.65±0.10	1.43±0.10	1.42±0.07	n.s.	n.s.
RER	0.89±0.02	0.88±0.02	0.88±0.02	0.87±0.01	n.s.	n.s.
$\stackrel{.}{\mathbf{V}}I$ / $\stackrel{.}{\mathbf{V}}{O}_2$	19.7±1.1	18.2±1.0	16.2±1.1	17.3±0.9	n.s.	F>M, 0.038
$\stackrel{.}{\mathbf{V}}I$ / $\stackrel{.}{\mathbf{V}}{\mathrm{CO}}_2$	21.9±0.9	20.6±1.1	18.3±0.9	19.9±1.1	n.s.	F>M, 0.038

Values are mean±SEM. T_b, body temperature; fr, respiratory frequency; Vt, tidal volume; $\stackrel{.}{\mathrm{V}}I$, minute ventilation; $\stackrel{.}{\mathrm{V}}{O}_2$, O₂ consumption; $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$, CO₂ production; RER, respiratory exchange ratio; $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{O}_2$, O₂ convection requirement; $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$, CO₂ convection requirement.

^aNo significant Treatment × Sex interactions were detected (all P>0.05), so only main effects from the ANOVA model are reported. n.s., nonsignificant; F, Female; M, Male; C, Control; H, Hyperoxia.

^bMetabolism was not assessed for every individual, so sample sizes differed between breathing and metabolism measurements.

Table 3.

Normoxic (21% O₂) ventilatory measurements for adult rats before hyperoxia (Pre), immediately following 4 d and 14 d of chronic hyperoxia, and after 4 d of recovery in room air (Experiment #4).

	Control				Hyperoxia				Treatment × Time? b
	Pre	4 d	14 d	Recovery	Pre	4 d	14 d	Recovery	P
n a	11				10
Age (d)	118±5	122±5	132±5	136±5	119±5	123±5	133±5	137±5
Mass (g)	411±39	412±38	417±38 †	421±38 †	399±37	401±37	407±36 †	405±36 †	n.s.
Tb (°C)	37.0±0.1	36.8±0.1	37.0±0.1	36.8±0.1	36.7±0.05 *	36.7±0.1	36.8±0.1	37.1±0.1	0.006
fr (breat hs min−1)	57.9±2.9	58.1±3.3	60.3±3.8	63.5±3.5 †	57.7±3.1	62.8±3.3	56.2±2.7	65.0±4.3 †	n.s.
Vt (ml 100g−1)	0.54±0.03	0.55±0.04	0.55±0.04	0.52±0.03	0.52±0.02	0.50±0.03	0.55±0.03	0.57±0.02	0.002
$\stackrel{.}{\mathbf{V}}I$ (ml min−1 100g−1)	31.1±0.4	32.1±1.2	32.6±0.9	32.5±1.0	30.5±0.6	31.2±1.1	31.8±1.2	37.5±1.9 * #	0.003

Values are mean±SEM. T_b, body temperature; fr, respiratory frequency; Vt, tidal volume; $\stackrel{.}{\mathrm{V}}I$, minute ventilation.

^aData for female and male rats are pooled within treatment groups (Control: 4 female, 7 male; Hyperoxia: 4 female, 6 male).

^bWhere the Treatment × Time interaction is significant

^*P<0.05 vs. Control at the same time point and

^#P<0.05 vs. “Pre” within the same treatment group. Where there are only main effects for Time (i.e., body mass and fr)

^†P<0.05 vs. “Pre” for both treatment groups.

3.2.2. Baseline ventilation and metabolic rate (Experiments # 2 and 3)

Chronic hyperoxia during adulthood had little effect on baseline (21% O₂ / 0% CO₂) breathing or metabolic rate (Table 2 for experiments #2 and 3 combined; see also Figures 3 and 4 for baseline data separated by experiment). No differences in breathing or metabolic rate were detected after 4 d of hyperoxia (Treatment and Treatment × Sex, all P>0.05), although Hyperoxia rats exhibited a modest (~ 0.2°C) reduction in T_b (Table 2). After 14 d in 60% O₂, Hyperoxia rats breathed with a slower, deeper pattern than Controls (9% lower fr, 9% greater Vt; both P<0.05), but this did not result in differences in $\stackrel{.}{\mathrm{V}}I$ (Treatment, P=0.586), $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{O}_2$ (Treatment, P=0.861), or $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$ (Treatment, P=0.884) (Table 2). T_b continued to be ~ 0.2°C lower in Hyperoxia rats after 14 d of exposure.

Fig. 3.

Ventilatory and metabolic responses to acute hypoxia in adult rats after 4 d and 14 d of 60% O₂ (Hyperoxia) or 21% O₂ (Control). Values (mean±SEM) are shown for (A, B) minute ventilation ($\stackrel{.}{\mathrm{V}}I$), (C, D) CO₂ production ($\stackrel{.}{\mathrm{V}}{\text{CO}}_2$), and (E, F) CO₂ convection requirement ($\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$) at baseline (21% O₂) and during acute exposure to 12% O₂. Sample sizes (female, male) are 27 Hyperoxia (15, 12) and 28 Control (17, 11) in panels A and B, 12 Hyperoxia (6, 6) and 14 Control (8, 6) in panels C and E, and 18 Hyperoxia (8, 10) and 16 Control (9, 7) in panels D and F. P-values are provided for the Treatment (Tr) × FiO₂ and Sex × FiO₂ interactions from the ANOVA or mixed-effects models; if either of these interactions was not significant, the main effect for Treatment or Sex is also provided. All Tr × Sex and Tr × Sex × FiO₂ interactions were nonsignificant. * P<0.05 vs. Control at the same FiO₂ in the post hoc analysis.

Fig. 4.

Ventilatory and metabolic responses to acute hypercapnia in adult rats after 4 d and 14 d of 60% O₂ (Hyperoxia) or 21% O₂ (Control). Values (mean±SEM) are shown for (A, B) minute ventilation ($\stackrel{.}{\mathrm{V}}I$), (C, D) CO₂ production ($\stackrel{.}{\mathrm{V}}{\text{CO}}_2$), and (E, F) CO₂ convection requirement ($\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$) at baseline (21% O₂ / 0% CO₂) and hypercapnia (21% O₂ / 7% CO₂). Sample sizes (female, male) are 16 Hyperoxia (8, 8) and 16 Control (9, 7) in panels A, C, and E, and 17 Hyperoxia (8, 9) and 17 Control (9, 8) in panels B, D, and F. P-values are provided for the Treatment (Tr) × FiCO₂ and Sex × FiCO₂ interactions from the ANOVA models; if either of these interactions was not significant, the main effect for Treatment or Sex is also provided. All Tr × Sex and Tr × Sex × FiO₂ interactions were nonsignificant.

3.2.3. Ventilatory and metabolic responses to acute hypoxia (Experiment #2)

Chronic hyperoxia augmented the HVR after 4 d, and this effect remained after 14 d of exposure (Fig. 3). Hyperoxia rats increased $\stackrel{.}{\mathrm{V}}I$ more during acute exposure to 12% O₂ than Control rats at both time points (Treatment × FiO₂, P=0.007 and 0.001, respectively) (Fig. 3A and 3B). Although the effects initially appeared to be greater in females, this impression was not supported statistically (Treatment × Sex and Treatment × Sex × FiO₂, all P>0.05). When averaged across the sexes, the HVR was 28% greater in Hyperoxia rats than in Control rats after 4 d of exposure (Δ $\stackrel{.}{\mathrm{V}}I$ = 20.4±1.2 vs. 15.9±1.0 ml min⁻¹ 100g⁻¹). After 14 d, the HVR was 50% greater in Hyperoxia rats (Δ $\stackrel{.}{\mathrm{V}}I$ = 19.3±1.3 vs. 12.9±1.3 ml min⁻¹ 100g⁻¹). Despite the greater $\stackrel{.}{\mathrm{V}}I$ response to hypoxia, hypoxic $\stackrel{.}{\mathrm{V}}I$ was only 8-9% greater than in Controls when averaged across sexes. Indeed, post hoc analysis only detected a greater hypoxic $\stackrel{.}{\mathrm{V}}I$ in female Hyperoxia rats at 4 d (Fig. 3A). A greater increase in fr appeared to contribute to the augmented HVR at 14 d, but the relative contributions of Vt and fr responses were otherwise variable (Appendix A, Fig. S3).

Metabolic responses to hypoxia did not differ between treatment groups or sexes (Fig. 3C and 3D), so the $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$ response to acute hypoxia was also greater in Hyperoxia rats than in Controls (Treatment × FiO₂, P=0.005 and 0.020 after 4 d and 14 d, respectively) (Fig. 3E and 3F). When averaged across the sexes, the increase in $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$ was 63% greater in Hyperoxia rats than in Control rats after 4 d of exposure (Δ $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$ = 17.4±1.5 vs. 10.7±1.5). After 14 d, the increase in $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$ was 56% greater in Hyperoxia rats (Δ $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$ = 13.6±1.6 vs. 8.7±1.5).

3.2.4. Ventilatory and metabolic responses to acute hypercapnia (Experiment #3)

To determine whether the effects of chronic hyperoxia were unique to the HVR, hypercapnic ventilatory responses (HCVR) were assessed in a separate group of adult rats. The ventilatory and metabolic responses to 7% CO₂ were not affected by 4 d of chronic hyperoxia (Fig. 4A, 4C, 4E). Both the increases in $\stackrel{.}{\mathrm{V}}I$ and $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$ were similar between Hyperoxia and Control rats (Treatment × FiCO₂, P=0.179 and 0.277, respectively). In contrast, 14 d of chronic hyperoxia increased the $\stackrel{.}{\mathrm{V}}I$ response in Hyperoxia rats (Treatment × FiCO₂, P=0.004) (Fig. 4B). This was true for both sexes (Treatment × Sex, P=0.658 and Treatment × Sex × FiCO₂, P=0.643), with the HCVR being approximately 21% greater in Hyperoxia rats (Δ $\stackrel{.}{\mathrm{V}}I$ = 61.5±2.5 vs. 50.8±2.2 ml min⁻¹ 100g⁻¹). Although post hoc analysis did not detect a significant difference in $\stackrel{.}{\mathrm{V}}I$ between Hyperoxia and Control rats while they were breathing 7% CO₂, hypercapnic $\stackrel{.}{\mathrm{V}}I$ tended to be 15% greater in Hyperoxia rats when pooled across sexes. This effect reflected a greater increase in Vt for Hyperoxia rats during hypercapnia (Appendix A, Fig. S4). However, the HCVR was no longer different between treatment groups after accounting for variation in metabolic rate (i.e., $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$; Treatment × FiCO₂, P=0.809) (Fig. 4F).

3.2.5. Recovery of hypoxic and hypercapnic responses (Experiment #3)

To determine whether the ventilatory effects of chronic hyperoxia persisted, hypoxic and hypercapnic responses were re-studied in Hyperoxia and Control rats from experiment #3 after approximately one month (27 – 31 d) in room air. No differences were detected between treatment groups in terms of baseline $\stackrel{.}{\mathrm{V}}I$, $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$, or $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$, or for their responses to 7% CO₂ or 12% O₂ (Treatment and Treatment × Test Gas, all P>0.05) (Fig. 5). This was true for both sexes (Treatment × Sex and Treatment × Sex × Test Gas, all P>0.05). Corresponding data for Vt and fr responses are provided in Appendix A (Fig. S5).

Fig. 5.

Ventilatory and metabolic responses to acute hypercapnia and acute hypoxia after 27 – 31 d of recovery in room air for adult rats previously exposed to 14 d of either 60% O₂ (Hyperoxia) or 21% O₂ (Control). Values (mean±SEM) are shown for (A) minute ventilation ($\stackrel{.}{\mathrm{V}}I$), (B) CO₂ production ($\stackrel{.}{\mathrm{V}}{\text{CO}}_2$), and CO₂ convection requirement ($\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$) at baseline (21% O₂), during hypercapnia (7% CO₂), and during hypoxia (12% O₂). Sample sizes are 18 Hyperoxia (8 female, 10 male) and 17 Control (9 female, 8 male); T_b data during 7% CO₂ were unavailable for one male rat from each treatment group, so the sample sizes are lower for those measurements. P-values are provided for the Treatment (Tr) × Test gas and Sex × Test gas interactions from the mixed-effects models; if either of these interactions was not significant, the main effect for Treatment or Test gas is also provided. All Tr × Sex and Tr × Sex × Test gas interactions were nonsignificant.

3.2.6. Normoxic ventilation (Experiment #4)

As described above, chronic hyperoxia appeared to affect baseline breathing in juvenile rats (Table 1) but not in adult rats (Table 2). An additional group of adult rats was studied over a longer period of time (60 min) in normoxia (21% O₂, balance N₂) to confirm the latter result. Normoxic $\stackrel{.}{\mathrm{V}}I$ did not change in Hyperoxia rats over the course of the exposure when compared to their own pre-exposure values or to Control rats studied in parallel, but $\stackrel{.}{\mathrm{V}}I$ did increase significantly after return to room air (Treatment × Time, P=0.003) (Table 3 and Fig. 6). Specifically, when re-studied after 4 d in room air, $\stackrel{.}{\mathrm{V}}I$ in Hyperoxia rats was 23% greater than their pre-exposure values (P=0.014) and 15% greater than in Control rats at the same time point (P=0.039). Females and males were pooled for this analysis given the small sample sizes for females (only four per treatment), but the pattern was similar for both female and male rats [$\stackrel{.}{\mathrm{V}}I$ after recovery (Hyperoxia vs. Control): 39.1±2.4 vs. 34.7±1.0 ml min⁻¹ 100g⁻¹ for females and 36.5±2.9 vs. 31.3±1.2 ml min⁻¹ 100g⁻¹ for males]. In both treatment groups, fr increased over the course of the study (Time, P=0.012) such that fr was greater during the recovery period than during the pre-exposure measurements (P=0.031) (Table 3). Changes in Vt were less predictable, but Vt tended to increase between the pre-exposure and recovery time points in Hyperoxia rats while it tended to decrease in Control rats (Treatment × Time, P=0.002) (Table 3). Accordingly, the increase in $\stackrel{.}{\mathrm{V}}I$ in Hyperoxia rats after 4 d in room air appears to be the product of modest increases in both fr and VT.

Fig. 6.

Normoxic ventilation (mean±SEM) for adult rats exposed to 60% O₂ (Hyperoxia; n=10) or 21% O₂ (Control; n=11). Minute ventilation ($\stackrel{.}{\mathrm{V}}I$) was measured in 21% O₂ on four occasions: before placing rats into the environmental chambers (Pre), after 4 d and 14 d of exposure, and after 4 d of recovery in room air (Recovery); the same individuals were studied at each time point. * P<0.05 vs. Control at the same time point; # P<0.05 vs. “Pre” within the same treatment group.

4. Discussion

Chronic exposure to moderate hyperoxia (60% O₂) affected multiple aspects of the control of breathing in juvenile and adult rats. First, juvenile rats exhibited significant hypopnea and hypoventilation compared to age-matched controls when acutely returned to room air, consistent with the phenotypes previously reported for neonatal rats exposed to comparable durations of hyperoxia during the first two weeks of life (e.g., Bavis et al., 2010; Bavis et al., 2014; Kavanagh et al., 2018). This was not the case for adults in which chronic hyperoxia had no effect on baseline $\stackrel{.}{\mathrm{V}}I$, $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{O}_2$, or $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$. However, there was an unexpected increase in normoxic $\stackrel{.}{\mathrm{V}}I$ in adult rats after 4 d of recovery in room air. Second, pre-treatment with chronic hyperoxia altered the ventilatory chemoreflexes of juvenile and adult rats, but these effects diverged from those reported for neonates. Specifically, chronic hyperoxia augmented the acute HVR in both juvenile and adult rats and tended to also augment the HCVR in adults; the HCVR was not studied in juveniles. In contrast, chronic hyperoxia causes a progressive blunting of the HVR in neonatal rats that lasts into adulthood while having minimal effect on the HCVR (Ling et al., 1996; Bavis et al., 2010; Bavis et al., 2017). Plasticity in normoxic ventilation, the HVR, and the HCVR were all reversible in juvenile and adult rats, unlike the long-lasting attenuation of the HVR after chronic postnatal hyperoxia (Ling et al., 1996; Fuller et al., 2002; Bavis, 2020). Taken together, these findings confirm that chronic hyperoxia elicits age-specific respiratory plasticity in rats. Importantly, these age-dependent differences are not simply caused by a lack of plasticity in adult-exposed rats; rather, the present studied revealed that there are qualitative differences in the plasticity that is expressed after chronic hyperoxia in neonates, juveniles, and adults as well as differences in its persistence.

4.1. Plasticity in normoxic ventilation

Hypopnea and hypoventilation have been observed consistently in neonatal rats returned to room air after chronic hyperoxia (e.g., Bavis et al., 2010; Roeser et al., 2011; Bavis et al., 2014; Kavanagh et al., 2018), so it is not surprising to see a similar phenotype in juvenile rats. This reduction in normoxic ventilation might be explained in part by reduced tonic drive from peripheral chemoreceptors since developmental hyperoxia decreases the number of O₂-sensitive glomus cells in the carotid body, decreases the number of chemoafferent neurons in the carotid sinus nerve, and decreases basal activity in the remaining neurons (Erickson et al., 1998; Donnelly et al., 2005; Bavis et al., 2011; Bavis et al., 2013). However, carotid body growth is no longer susceptible to chronic hyperoxia by 4 weeks of age (Bavis et al., 2013), so a reduced number of glomus cells seems unlikely to explain hypoventilation in juvenile rats. Alternatively, studies using in vitro brainstem-spinal cord preparations also point to a central component since fictive breathing frequency is substantially lower in neonatal rats after chronic hyperoxia (Bavis et al., 2014; Bierman et al., 2014). Although carotid body hypoplasia appears permanent after developmental hyperoxia (Fuller et al., 2002; Bavis et al., 2013), normoxic ventilation recovers within 24 h in neonatal rats (Bavis et al., 2014) and within a week, if not sooner, in juvenile rats (Table 1), suggesting that the other mechanisms underlying this plasticity are reversible.

In contrast to neonatal and juvenile rats, chronic hyperoxia had minimal influence on normoxic ventilation in adult rats; baseline $\stackrel{.}{\mathrm{V}}I$, $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{O}_2$, and $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$ were similar to those of control rats in experiments #2 and 3 and this result was confirmed after a more extensive examination of normoxic $\stackrel{.}{\mathrm{V}}I$ in experiment #4. The consistent absence of a hypopneic response to chronic hyperoxia in adult rats is striking. One potential explanation is that chronic hyperoxia has less of an impact on the peripheral drive to breathe in adults since their carotid bodies are more mature at the time of exposure. However, postnatal maturation of glomus cell O₂ sensitivity is largely completed within the first two postnatal weeks in rats (Bavis et al., 2011; Carroll and Kim, 2013), so one might have expected similar results in month-old, juvenile rats and adult rats; this was not the case. This could point to age-dependent sensitivity of central components of respiratory drive (e.g., the pre-Bötzinger Complex) to chronic hyperoxia or to the emergence of compensatory plasticity in adult rats. To the latter point, there was an unexpected increase in $\stackrel{.}{\mathrm{V}}I$ observed 4 d after adult rats were returned to room air. It is possible that adult rats express excitatory plasticity that initially offsets the depressive effect of chronic hyperoxia. If the excitatory plasticity resolves more slowly once normoxia is restored, the balance could shift towards excitation and produce a transient hyperpnea (i.e., overcompensation). In any event, the mechanism underlying this late-appearing hyperpnea warrants further investigation.

4.2. Plasticity in the hypoxic ventilatory response

Chronic hyperoxia augmented the HVR in both juvenile and adult rats, with the acute change in $\stackrel{.}{\mathrm{V}}I$ (i.e., $\stackrel{.}{\mathrm{V}}I$ in hypoxia - $\stackrel{.}{\mathrm{V}}I$ in normoxia) increasing by approximately 28 – 50% depending on the age group and duration of the hyperoxic exposure. In contrast, chronic postnatal hyperoxia tends to blunt the HVR of neonatal rats. Newborn mammals typically exhibit a biphasic HVR in which an initial increase in $\stackrel{.}{\mathrm{V}}I$ (primarily reflecting carotid body stimulation) is followed by a gradual decline in $\stackrel{.}{\mathrm{V}}I$ back toward or below baseline levels (primarily reflecting metabolic depression and inhibitory signaling in the brainstem) (Bissonnette, 2000); the HVR becomes less biphasic over the course of postnatal development as the secondary ventilatory decline diminishes. The early phase of the HVR is generally unaffected by 4 d of 60% O₂ while longer durations of hyperoxia blunt the early HVR through progressive impairment of carotid body function (Bavis et al., 2010; Bavis et al., 2013). However, rats reared in hyperoxia do not exhibit the typical biphasic HVR (i.e., little or no hypoxic ventilatory decline), so the late phase of the HVR is actually greater in hyperoxia-treated rats than age-matched controls during the first postnatal week; this pattern reverses during the second postnatal week as the late phase of the HVR matures in control rats (Bavis et al., 2010) and the HVR remains blunted into adulthood (Ling et al., 1996). Importantly, the augmented HVR during the late phase of the HVR in the youngest neonatal rats is associated with premature changes in excitatory and inhibitory pathways that govern normal postnatal maturation of the biphasic HVR (reviewed in Bavis, 2020), so it is unclear that these same mechanisms would underly the augmented HVR observed in juvenile and adult rats when these pathways are already in their mature state. This conclusion is further supported by the observation that 14 d of hyperoxia continues to augment the HVR in adult rats but not in neonates (Bavis et al., 2010).

The augmented HVR in juvenile and adult rats after chronic 60% O₂ exposure may instead share mechanisms with the potentiation of the HVR following shorter (minutes – hours) exposures to hyperoxia in adult humans (Honda et al., 1996), rats (Gozal, 1998; Pokorski et al., 2005; Budzinska and Ilasz, 2008), cats (Pokorski et al., 2004), and goats (Herman et al., 2001); a similar increase in the HVR has been observed after ~24 h of 60% O₂ in 7 – 8 d old rats (Roeser et al., 2011). Prior exposure to hyperoxia is proposed to increase nitric oxide production within the brainstem, ultimately enhancing the release of the excitatory neurotransmitter glutamate during subsequent exposure to hypoxia (Gozal, 1998; Pokorski et al., 2005). Increased production of reactive oxygen species may also contribute to this plasticity (Budzinska and Ilasz, 2008), but antioxidant treatment did not prevent the potentiation of the HVR after ~24 h of hyperoxia in neonatal rats (Roeser et al., 2011).

There is a positive interaction between hypoxia and hypercapnia at the peripheral chemoreceptors such that higher background Pco₂ levels will increase the HVR for a given decrease in arterial Po₂ (reviewed in Teppema and Dahan, 2010). Although blood gases were not measured during the present study, it is likely that juvenile rats had somewhat higher baseline arterial Pco₂ after chronic hyperoxia since they were hypoventilating in normoxia (i.e., lower $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{O}_2$ and $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$). While this could contribute to the augmented HVR of juvenile rats after chronic hyperoxia, it probably is not the primary mechanism since adult rats also exhibited an augmented HVR after chronic hyperoxia without any changes in baseline $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{O}_2$ and $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$. Likewise, potentiation of the HVR in adult rats after a 10-min pre-exposure to 100% O₂ occurs without any changes in baseline blood gases (Gozal, 1998).

Augmentation of the HVR after chronic hyperoxia may be a short-lasting form of respiratory plasticity. In the present study, the HVR was tested in juvenile and adult rats more than an hour after being removed from hyperoxia (i.e., 10-15 min in room air prior to being sealed into the plethysmograph chamber and an additional 60 min of normoxia during the baseline period). However, the augmented HVR after chronic hyperoxia was no longer evident after one week (juveniles) or after approximately four weeks (adults) in room air. Shorter durations of recovery in room air (i.e., hours – days) were not studied, so it is possible that the HVR recovered relatively quickly. For example, potentiation of the HVR in adult rats after a 10-min pre-exposure to 100% O₂ is no longer evident 60 – 90 min later (Gozal, 1998). Importantly, the absence of long-lasting changes in the HVR of juvenile and adult rats is largely consistent with prior studies showing that chronic hyperoxia does not blunt the HVR if exposures occur after a critical period in development (approximately the first two postnatal weeks in rats) (Ling et al., 1996, 1997; Bavis et al., 2002; Bavis et al., 2013). Although Ling et al. (1996) reported a greater $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$ response to hypoxia in rats two months after they were exposed 60% O₂ as adults, this was largely driven by decreases in $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$ and they did not observe any changes in the $\stackrel{.}{\mathrm{V}}I$ response to hypoxia. Moreover, Ling et al. (1996) suggested that their adult-treated rats experienced lower arterial Po₂ during the hypoxic tests due to gas exchange impairment after prolonged hyperoxia (4 weeks in 60% O₂), potentially overestimating the hypoxic responses. In contrast, exposure to only 1 – 2 weeks of 60% O₂ does not appear to affect baseline or hypoxic arterial Po₂ in rats (Bavis et al., 2003; Bavis et al., 2007).

4.3. Plasticity in the hypercapnic ventilatory response

Developmental hyperoxia has little immediate or lasting effect on the hypercapnic responses of rats (Ling et al., 1996; Bavis et al., 2017), but little was previously known about the effects of short or chronic exposures to hyperoxia on the HCVR of adult rats. Therefore, the HCVR was also investigated to determine whether chronic hyperoxia elicits plasticity specific to the HVR versus a more global excitation of the ventilatory chemoreflexes. Chronic hyperoxia caused a modest increase in the $\stackrel{.}{\mathrm{V}}I$ response to 7% CO₂ in adult rats, reminiscent of the augmented HVR, but only after the 14-d exposure. Moreover, this effect was no longer evident when accounting for variation in metabolic rate, so it is unclear whether this reflects plasticity in the control of breathing per se. These differences in time course and $\stackrel{.}{\mathrm{V}}I$ / $\stackrel{.}{\mathrm{V}}{\text{CO}}_2$ responses suggest that chronic hyperoxia may have independent effects on hypoxic and hypercapnic sensing and/or central integration of chemoreceptor activity in adult rats.

5. Conclusion

Reversible augmentation of the HVR, and possibly HCVR, after chronic hyperoxia in juvenile and adult rats contrasts with the long-lasting blunting of the HVR after chronic hyperoxia during early, postnatal development. These findings confirm the unique character of developmental plasticity elicited by chronic hyperoxia and highlight the importance of studying respiratory plasticity across the age spectrum in humans and animal models. Although hyperoxia is rare in nature, repeated and extended bouts of hyperoxia / hyperoxemia are not uncommon during O₂ therapy and mechanical ventilation in clinical settings (e.g., Palmer et al., 2019; da Silva et al., 2024). Accordingly, the present findings suggest that liberal use of supplemental O₂ could have age-dependent effects on normoxic ventilation and protective ventilatory chemoreflexes. Augmented O₂ sensitivity could lead to unstable breathing for juveniles and adults in the short term and this possibility warrants further investigation in humans. However, unlike perinatal hyperoxia, these effects are unlikely to be permanent.

Highlights.

• Juvenile rats, but not adults, hypoventilate in room air after chronic hyperoxia.

• Chronic hyperoxia augments the hypoxic ventilatory response (HVR) of juvenile rats.

• Chronic hyperoxia augments the HVR and perhaps hypercapnic response of adult rats.

• Respiratory control recovers within weeks (maybe sooner) after return to room air.

• Hyperoxia-induced plasticity differs between juvenile/adult rats and neonates.