Perinatal hyperoxic exposure reconfigures the central respiratory ...

J Appl Physiol 116: 47–53, 2014. First published October 24, 2013; doi:10.1152/japplphysiol.00224.2013.

Perinatal hyperoxic exposure reconfigures the central respiratory network contributing to intolerance to anoxia in newborn rat pups Alexis M. Bierman,1 Clarke G. Tankersley,2 Christopher G. Wilson,3 Raul Chavez-Valdez,1 and Estelle B. Gauda1 1

Department of Pediatrics, Neonatology Research Laboratories, Johns Hopkins Medical Institutions, Baltimore, Maryland; Department of Environmental Health Science, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland; and 3Department of Pediatrics, Neonatology and Neurosciences, Case Western Reserve University, Cleveland, Ohio 2

Submitted 20 February 2013; accepted in final form 17 October 2013

Bierman AM, Tankersley CG, Wilson CG, Chavez-Valdez R, Gauda EB. Perinatal hyperoxic exposure reconfigures the central respiratory network contributing to intolerance to anoxia in newborn rat pups. J Appl Physiol 116: 47–53, 2014. First published October 24, 2013; doi:10.1152/japplphysiol.00224.2013.—Perinatal exposure to hyperoxia (30 – 60% O2) alters the respiratory control system via modulation of peripheral arterial chemoreceptor development and function. Furthermore, hyperoxic exposure during the first two postnatal weeks of life can alternatively modulate the different phases of the hypoxic ventilatory response. Given the effects of perinatal hyperoxia, the aims of our study were 1) to determine the effect on survival time in response to lethal anoxic stimuli in rat pups and 2) to characterize the output of the isolated central respiratory network in response to acute hypoxic stimuli. We hypothesized that perinatal hyperoxic exposure would modify the neonatal rat ventilatory response to anoxia by affecting a central component of the respiratory network in addition to the maturation of the carotid body chemoreceptors. We found that animals continuously exposed to 60% oxygen up to age 5 days after parturition (P5) have reduced breathing frequency at baseline and within the first 10 min of a fatal anoxic challenge. Hyperoxic rat pups also have a shortened time to last gasp in response to anoxia that is not associated with lung injury or inflammation. This study is the first to demonstrate that these in vivo findings correlate with reduced phrenic burst frequency from the isolated brainstem ex vivo. Thus hyperoxic exposure reduced the phrenic burst frequency at baseline and in response to ex vivo anoxia. Importantly, our data suggest that perinatal hyperoxia alters ventilation and the response to anoxia at P5 in part by altering the frequency of phrenic bursts generated by the central respiratory network. hypoxic ventilatory response; brainstem-spinal cord preparation; perinatal hyperoxia; anoxic exposure; chemoafferent input

response (HVR) consists of two distinct phases: Phase 1, the acute response, is charac˙ E) early after hypoxic terized by elevated minute ventilation (V exposure and is mostly mediated by the excitatory input from carotid body chemoreceptors; and Phase 2, the late response, is characterized by ventilatory depression primarily mediated by intrinsic and extrinsic inhibitory factors affecting neurons in the central respiratory network (7, 14). As the mammalian HVR matures, the Phase 1 response becomes greater and Phase 2 becomes less. In most mammalian species, the HVR is fully mature by ⬃2 wk of postnatal life (7).

THE MAMMALIAN HYPOXIC VENTILATORY

Address for reprint requests and other correspondence: E. B. Gauda, Neonatology Research Laboratories, Johns Hopkins Medical Institutions, Dept. of Pediatrics, Neonatology Research Labs, 600 N. Wolfe St., CMSC 6-104, Baltimore, MD 21287 (e-mail: [email protected]). http://www.jappl.org

Early environmental exposures, particularly the level of oxygen exposure during early postnatal development, can induce functional and morphological changes in the mammalian respiratory network, thereby altering the HVR (1, 3, 6). For example in newborn rat pups, sustained hyperoxic exposure for the first 2– 4 wk of postnatal development markedly decreases the HVR later in life (1–3, 6). This sustained exposure inhibits the development of carotid body chemoreceptors, thereby eliminating Phase 1 of the HVR (5, 14). On the other hand, hyperoxic exposure for the first 4 days of life attenuates Phase 2 of the HVR, leading to less hypoxic ventilatory depression and reductions in tidal volume (VT), breathing frequency (F), and ˙ E during eupneic breathing in room air (5). Because thus V sustained ventilation in response to severe hypoxia (i.e., anoxia) is an important survival mechanism, we aimed to know 1) the effect of perinatal hyperoxic exposure in response to anoxia in vivo and 2) the integrated output of the central respiratory network ex vivo. We hypothesized that hyperoxic exposure during early development modifies the central respiratory network leading to anoxic intolerance when assessed in vivo in freely moving rat pups and ex vivo using the brainstemspinal cord preparation. METHODS

All experiments were approved by the Animal Care and Use Committee at the Johns Hopkins University School of Medicine and followed the Guide for the Care and Use of Laboratory Animals, 8th edition, 2011 provided by the National Institutes of Health. Whole body plethysmography. Sprague-Dawley rat pups were exposed to either hyperoxia (fraction of inspired O2 0.60, n ⫽ 15) or room air (normoxia, n ⫽ 15) for 2 days before birth (E18) and for 4 –5 days after parturition (P4 –5; 4 litters/group). Whole body plethysmography was used to measure postnatal ventilatory parameters in freely moving rat pups at P4 –5 (2, 26). Each animal was permitted to acclimate to the cylindrical chamber (400 cm3) for at least 30 min before ventilatory parameters were measured. The ambient temperature of the chamber was maintained within the known nesting temperature of animals at this age (34.4 ⫾ 0.2°C) (21) using a heating pad that was wrapped around the chamber. Ambient chamber temperature was continuously monitored using a type-T thermocouple (IT-18; AD Instruments, Colorado Springs, CO) and recorded using a PowerLab data acquisition module and LabChart software (AD Instruments). At a constant chamber volume, changes in pressure due to inspiration and expiration were measured using a differential pressure transducer (model 8510B-2; Endevco, San Juan Capistrano, CA) and recorded to identify breathing cycle timing. Following acclimation, 60 s of baseline breathing on room air was recorded. Thereafter, pups were exposed to anoxia for 75 s (5% CO2-95% N2). Breathing effort was recorded until the final gasp, and data were analyzed using a specialized ventilation analysis program designed with Igor Pro (XOP;

8750-7587/14 Copyright © 2014 the American Physiological Society

47

48

Perinatal Hyperoxia Contributes to Anoxic Intolerance

Wavemetrics, Portland, OR) to measure inspiratory time (Ti), expiratory time (Te), total breathing time (Tt), and breathing frequency (f). Assessment of lung inflammation. A separate group of animals were used to measure lung alveolar permeability and inflammation. After the 4 –5 day exposure, hyperoxic (n ⫽ 9) and room air (n ⫽ 10) exposed pups were given a lethal dose of isoflurane, bronchoalveolar lavage fluid (BALF) was retrieved, and lungs were processed for histology. Specifically, the left pulmonary vessels and bronchi were clamped with a micro serrefine (Fine Science Tools, Foster City, CA) to isolate the left lobe of the lung for histology. The remaining right lobes were lavaged with 0.5 ml four times with sterile 1% 1⫻ PBS (Thermo Fisher Scientific, USA). All cell counts were normalized to the volume BALF retrieved. The number of total leukocytes in BALF was counted using a hemocytometer after cells were resuspended in a 1-ml solution (1⫻ PBS, 1% FBS, and 1 mM EDTA). A differential cell count was done on BALF cells after cytospinning (Thermo Scientific, Asheville, NC) and staining with Diff-Quik (Seimens Healthcare Diagnostics, Newark, DE). Inflammatory cells were identified, counted, and normalized to the total leukocytes in BALF. We also assayed for alveolar permeability by measuring soluble protein concentrations in the BALF using the Bradford Bio-Rad protein assay. Measurements were performed in triplicate (8). Histology. For histology, the lung was inflated to 25 cmH2O with 10% formalin. The trachea was tied, and the formalin-inflated lung was placed in 40 ml of formalin for 48 h. The left lobe was cut into three equal horizontal sections and dehydrated with 70% ethanol. After embedding the tissue in paraffin, transverse lung sections (5 ␮m) were stained with hematoxylin and eosin. Images were acquired using a Nikon Eclipse 80i microscope at ⫻20 magnification. The entire sections were photographed, and a systematic random sampling was done to yield ⬃25 fields/section. Mean airspace chord length (Lm) was measured with sampling grid lines using Nikon NISElements AR 3.00 software. Lung morphology was assessed by identifying alveoli, alveolar ducts, and blood vessels, then quantitated using a method described by Knudsen et al. (19) based on point intersect counting of ⬃600 points per animal. Brainstem-spinal cord preparation. Two separate age-matched groups of pups were exposed to either hyperoxia (n ⫽ 8) or room air (n ⫽ 9) from E18 to P4 or P5. At P4 –5, pups from both groups were anesthetized (isoflurane) and then decapitated. The brainstem-spinal cord was dissected and isolated en bloc while submerged in artificial cerebral spinal fluid (aCSF; 124 mM NaCl, 3 mM KCl, 1.5 mM CaCl2, 1 mM MgSO4, 0.5 mM NaH2PO4, 25 mM NaHCO3, and 30 mM D-Glucose) at ⬃4°C (pH, 7.40) and bubbled with 95% O2-5% CO2, as previously reported (25). The brainstem was transected rostrally caudal to the pons (to eliminate inhibitory signals to the medulla), and the spinal cord was transected immediately rostral to the C5-C6 rootlet. The preparation, en bloc, was then pinned ventral side up to a plastic elastomer-covered dish and while continuously perfused (3– 4 ml/min) with carbogenated (95% O2-5% CO2) aCSF (pH 7.40) and warmed to ⬃27°C. Temperature was maintained using a bath-heating element and monitored via a T-pod with a thermocouple (AD Instruments) placed in the preparation bath. Extracellular phrenic nerve activity (fictive breathing) was recorded from ventral cervical rootlets 1– 4 (C1-C4) via a fire-polished glass suction electrode. Rootlet nerve activity was continuously recorded via a PowerLab data acquisition system (AD Instruments, 4 kilosamples/s). The signal was amplified (50,000-fold) using a high-impedance probe connected to an amplifier (P511 series, Grass Technologies), then integrated, fullwave rectified, band-pass filtered (10 –1,000 Hz), and averaged over 200 ms (Fig. 1). Once rhythmic activity was detected and bath conditions (temperature and flow) were stable for ⬃15 min, a 5- to 10-min recording was obtained under baseline conditions (aCSF bubbled with 95% O2-5% CO2) followed by 30 min of activity during exposure to anoxic solution (aCSF bubbled with 95% N2-5% CO2). The first 5 min of activity during anoxia perfusion was not included in

•

Bierman AM et al.

the analysis due to contamination of the signal from bubbles generated in the chamber from changing the oxygen content of the superfusate. The integrated rootlet activity was later analyzed with MATLAB software (Mathworks, Natick MA) with windowed detection of amplitude and rise time to determine burst onset/offset, peak amplitude, and area of burst. Representative tracing demonstrating how timing variables were analyzed is shown in Fig. 1. Figure 1F displays burst frequency (burst F ⫽ 1/interburst interval), total respiratory cycle time (Fig. 1Fi), burst duration (Fig. 1Fii), and interburst interval (Fig. 1Fiii). Statistical analysis. Ventilation and fictive breathing variables were compared between control and hyperoxic groups. Two-way repeated measures ANOVA was used for longitudinal analysis of the ventilatory parameters determining significance of within-subject (baseline vs. anoxia exposure) and between-subject (control vs. hyperoxia) effects using Prism software (GraphPad, La Jolla, CA). Only breathing data from baseline to the first 20 min were used for plethysmography ventilation analysis, due to significant unbalanced mortality thereafter. Bonferroni correction was used for comparison by pairs. Significance was established at P ⬍ 0.05. Survival to anoxia was evaluated using a Kaplan-Meier survival analysis (Prism, GraphPad). All reported results in figures are expressed as mean ⫾ SE. RESULTS

Hyperoxia does not alter body weight in P4 –5 rats. Perinatal exposure to hyperoxia did not alter the body weight in rat pups for any of the litters used in this study. For each experimental group the body weight (means ⫾ SE) in normoxic vs. hyperoxic exposed animals was: 10.06 ⫾ 0.71 vs. 10.80 ⫾ 0.51 g (P ⫽ 0.90) for animals used for ventilation experiments; 10.63 ⫾ 0.41 vs. 11.97 ⫾ 0.84 g (P ⫽ 0.13) for animals used for lung inflammation and histological observations; and 9.26 ⫾ 0.71 vs. 9.31 ⫾ 0.68 g (P ⫽ 0.41) for animals used for ex vivo (electrophysiology) brainstem-spinal cord experiments. Ventilatory parameters (in vivo) following perinatal hyperoxic exposure at baseline and during anoxia until last gasp. Representative breathing traces for normoxic and hyperoxic animals during baseline eupneic breathing are shown in Fig. 2, A and B (left). These traces show a slower breathing frequency at baseline for the hyperoxic animal relative to control. Figure 2 also displays representative breathing traces for control and hyperoxic animals during anoxia until last gasp [Fig. 2, A and B (right)]. Compared with baseline traces, the traces for both animals during anoxia show a drastic increase in amplitude and reduction in frequency. The representative hyperoxic animal has a shortened time to last gasp during anoxia compared with the control animal. The timing variables represented in Fig. 2 are quantified for 15 animals per group in Fig. 3. The mean breathing frequency at baseline is ⬃20% lower in hyperoxic vs. control animals (P ⬍ 0.05). Both Ti and Te components of breathing (Fig. 3, A and B) contribute to the lower baseline F in hyperoxic animals (P ⬍ 0.03), although Ti and Te did not differ between the two groups during eupneic breathing. Perinatal hyperoxic exposure decreases ventilatory parameters and shortens time to last gasp during an anoxic challenge. Hyperoxic exposure significantly altered ventilatory responses to anoxia in pups at P4 –5. Within 5 min following anoxia exposure, both normoxic- and hyperoxic-exposed pups responded with a gasping frequency that was 96% slower than the frequency during eupneic breathing (Fig. 3C, P ⬍ 0.001).

J Appl Physiol • doi:10.1152/japplphysiol.00224.2013 • www.jappl.org


A

•

49

Bierman AM et al.

VENTRAL BRAINSTEM & SPINAL CORD

Facial n. Glossopharyngeal n. Vagal n.

C

HIGH IMPEDANCE AMPLIFIER (P511-AC, Grass) • amp = 50,000x • band pass = 10 – 1000 Hz

CCA

Hypoglossal n.

BA

D

Cervical rootlets

POWER LAB (AD Instruments)

B

E

VENTRAL CERVICAL ROOTLET ACTIVITY

F

ZOOMED VIEW

C1-4 ii

i

iii

∫C1-4 1 min

10 s

Fig. 1. Fictive breathing activity measurement using the brainstem-spinal cord en bloc preparation. A: the ventral brainstem and spinal cord were dissected en bloc from 4 –5 days after partuition (P4 –5) rats. Ventral medullary rootlets are labeled VII, IX, X, XII. CCA, caudal cerebellar artery; BA, basilar artery. B: spontaneous rhythmic activity (fictive breathing) recorded from one of the C1–C4 rootlets via suction electrode. Activity was amplified and filtered via a high-impedance amplifier (C) and recorded with a PowerLab data acquisition system (D). E: raw recordings of 10 min of cervical rootlet activity (C1– 4) with the integrated signal below (兰C1– 4) averaged over 200 ms. F: zoomed view of 1 min of activity noting timing variables on the integrated signal below as follows: i, total burst cycle; ii, burst duration; and iii, interburst interval.

However, by 10 min of anoxia, hyperoxic-exposed animals had a much longer time between gasps compared with controls (Fig. 3B, P ⫽ 0.029). This contributed to a greater decrease in F by 10 min of anoxia (Fig. 3C, P ⫽ 0.044). These data suggest that hyperoxic

exposure from E18 to P4 impairs gasping within the first 10 min of an anoxic challenge via a greater time between gasps. Further evaluation with Kaplan-Meier survival analysis displayed in Fig. 4 shows that the last gasp occurred by 25 min of last gasp

Control

A

Eupneic tidal breathing

Hyperoxic

B

+5 min anoxia

Anoxia until last gasp last gasp

10 units 5 secs

10 units 5 mins

Fig. 2. Eupneic tidal breathing and survival measured via plethysmography with and without anoxia until last gasp. Breathing traces from control (A) and hyperoxic (B) P4 –5 rats during eupneic tidal breathing (left) followed by anoxia (right) that was measured until last gasp. Baseline eupneic breathing traces (left) only represent 10 s of data. Both anoxic traces (right) display the typical biphasic appearance of the hypoxic ventilatory response at P4 –5 and a prolonged survival in the control pup (A) vs. the hyperoxic pup (B). The control trace extends ⬎25 min. These traces are representative of results found in this study where hyperoxic animals have a slower breathing frequency at baseline (eupneic breathing, left) and display a shortened time to last gasp with anoxia compared with control animals (right). J Appl Physiol • doi:10.1152/japplphysiol.00224.2013 • www.jappl.org

50


•

Bierman AM et al.

hyperoxia did not alter protein content (43 ⫾ 11 vs. 40 ⫾ 10 ␮g/ml), normoxic control vs. hyperoxia, respectively. Thus hyperoxic exposure from E18 to P4 –5 did not induce significant lung inflammation or injury. Perinatal hyperoxic exposure modified the ex vivo spontaneous rhythmic activity (fictive breathing) generated in the isolated brainstem-spinal cord. The rhythmic phrenic bursts from C1–C4 rootlets were measured to determine the effects of hyperoxia on the central respiratory network. The representative traces of phrenic rhythmic bursting in Fig. 5, A and B (left panels) depicts the changes observed at baseline in Fig. 5, C and D. Hyperoxic exposure reduced the mean baseline respiratory burst frequency by ⬃40% (P ⫽ 0.01, Fig. 5C). This was mediated mainly through an increase in interburst interval (frequency, Fig. 5D; refer to Fig. 1 for explanation of measured values). Following 5 min of exposure to anoxic superfusate, spontaneous rhythmic activity from both hyperoxic and control preparations were reduced (P ⬍ 0.01) with decreased bursting frequencies compared with baseline activities (Fig. 5C). These data are consistent with the baseline frequency observations in vivo (Fig. 3C). Compared with controls, preparations from hyperoxic-exposed animals maintained a lower gasp frequency throughout the 30 min ex vivo challenge due to a greater increase (P ⬍ 0.01) in interburst intervals in hyperoxic preparations (P ⬍ 0.01 vs. controls; Fig. 5D). In summary, hyperoxic exposure reduced the phrenic bursting characteristics of the neonatal brainstem at baseline and in response to anoxia. DISCUSSION

Fig. 3. Hyperoxic exposure alters baseline breathing and the responses to anoxia in vivo. The following ventilatory variables were measured with plethysmography and averaged for 5 min intervals: inspiratory time (A), expiratory time (B), and breathing frequency (C). Data are means ⫾ SE for control (open squares) and hyperoxic (closed squares) rats. Baseline (Base, normoxia) and the anoxic challenge (to the right of the dotted line, from 5 to 20 min) are shown. *P ⬍ 0.05 (repeated measures ANOVA), significant within (time) and between (hyperoxia exposure) subject effects. The number of surviving animals during the anoxic challenge is listed below the x-axis of C.

anoxia in all pups exposed to hyperoxia, whereas the last gasp occurred by 41 min in all normoxic-exposed animals. Overall, hyperoxic pups had a time to last gasp that was 7 min shorter than normoxic pups, 17 vs. 24 min, respectively (log rank, P ⫽ 0.0006). Thus hyperoxic and normoxic groups had median survival times of 25 and 30 min, respectively (Fig. 4). In summary, hyperoxic pups ceased gasping about three times sooner during anoxia compared with animals raised in normoxia. Effect of hyperoxic exposure on the lung. To understand how perinatal hyperoxic exposure shapes survival and ventilatory responses to anoxia in vivo, we determined whether our paradigm of perinatal hyperoxic exposure could induce lung inflammation and injury. Thus BALF was analyzed for inflammatory cell profile and protein content in age-matched animals. Relative to normoxia, hyperoxic exposure did not significantly change the total number (expressed in thousands) of leukocytes (94 ⫾ 12 vs. 95 ⫾ 15), neutrophils (3 ⫾ 0.62 vs. 4 ⫾ 0.88), or macrophages (86 ⫾ 12 vs. 88 ⫾ 14) retrieved in BALF. Also,

The effects of perinatal hyperoxia on the ventilatory and carotid sinus nerve responses to hypoxia are well characterized in rodents (10, 11, 14). Here we show that hyperoxic exposure from E18 to P4 –5 in neonatal rats alters eupneic breathing and severely impairs survival to a lethal anoxic challenge. These effects were not associated with significant hyperoxic-induced lung injury or inflammation. Our study is the first to report that in vivo perinatal hyperoxic exposure also reduces the integrated phrenic activity of the superfused isolated brainstemspinal cord preparation at baseline and during anoxia. In addition to altering the structure and function of carotid body chemoreceptors (3, 4, 6, 10), these data suggest that perinatal hyperoxia also modifies the central respiratory rhythm gener-

Fig. 4. Hyperoxic exposure shortens the time to last gasp during anoxia in vivo. Kaplan-Meier curves show the survival responses to fatal anoxic challenge for control (n ⫽ 15) and hyperoxic (n ⫽ 15) rats. Hyperoxic pups stop gasping ⬃3 times sooner during anoxia compared with normoxic pups (log rank test; P ⫽ 0.0006).



•

Bierman AM et al.

51

Fig. 5. Hyperoxic exposure reduces the phrenic burst frequency via an increase in interburst interval in the neonatal (P4 –5) rat isolated brainstem-spinal cord preparation. Representative baseline (artificial cerebral spinal fluid: 95% O2-5% CO2) activity measurements (raw: C4, and integrated: 兰C4) from the C4 rootlet are zoomed over 1 min (left) for control (A) and hyperoxic pups (B). The zoomed control recording (A, left) has a frequency of 7 bursts/min while the hyperoxic trace (B, left) has a frequency of 5 bursts/min. The gray arrow indicates 30 min of the ex vivo anoxic challenge (5% CO295% N2). Fictive breathing activity was recorded from the C1-C4 rootlets from control (open circles, n ⫽ 9) and hyperoxic (closed circles, n ⫽ 8) rats, and timing variables were averaged over 5-min intervals for baseline followed by 30 min of ex vivo anoxia (C and D). Values are means ⫾ SE for burst frequency (C) and interburst interval (D); *P ⬍ 0.05. The effect of time is significant (P ⬍ 0.0001; repeated measures ANOVA and Bonferroni correction).

ator and pattern formation network. These modifications could include alterations in connectivity and synaptic weights within the network or the level of activation or inhibition of modulatory influences. Hyperoxic exposure decreases anoxic tolerance in P4 –5 rat pups. A fundamental survival strategy in mammals is the ability to alter ventilation to meet oxygen and metabolic needs of cells and tissues (3, 9, 12, 16). Thus the respiratory network is dynamic and robust and consists of 1) sensors, 2) controllers, and 3) effectors. Carotid chemoreceptors are an important source of sensory input to the central respiratory controller. Although input from these receptors is not essential for the initiation of breathing at the time of birth, stimulatory input during early postnatal development appears to stabilize the respiratory network, and chemodenervation in newborn animals causes respiratory failure and increases the mortality risk for several weeks following the surgery (17, 18) and as reviewed by Gauda et al. (15). Our study examines the effect of hyperoxic exposure on ventilation and the response to a lethal hypoxic or anoxic challenge during carotid body development. Hyperoxic exposure during this period markedly reduces carotid sinus nerve activity (3, 5, 15) and thus input to the developing central respiratory network. We found that hyperoxic exposure reduced baseline ventilatory breathing frequency in P4 –5 rat pups. These results are consistent with other studies in neonatal pups where perinatal hyperoxic exposure reduces eupneic ventilation within the first week of life, but not in adult rats (⬎2 wk) or mice with similar paradigms of exposure (2, 5).

Chronic perinatal hyperoxic exposure impairs the postnatal development of oxygen sensitivity in adult mammals. This occurs via hindered development of the carotid body chemoreceptors (1, 14). The HVR following perinatal hyperoxic exposure is well characterized in the rat, mouse, and other vertebrates and resembles the HVR observed in the youngest animals (age ⬍P5) raised in normoxia (21% O2) (1, 3, 4, 11, 27). Thus perinatal hyperoxic exposure reduces the stimulatory HVR in adults (Phase 1), and this change can be lifelong depending on the length and degree of exposure (3, 6, 7). In this study, perinatal hyperoxia from E18 to P4 –5 reduces Phase 1 HVR and hastens the time to last gasp in response to an anoxic challenge. While perinatal hyperoxia impairs carotid body structure and hypoxic chemosensitivity of the carotid sinus nerve (15, 17, 18), it may also contribute to decreased tolerance to anoxia in hyperoxic animals as shown in this study. Previous studies suggest that perinatal hyperoxia appears to attenuate the function and development of the carotid body chemoreceptors with no evidence of an effect on the central controller in adult animals (3, 6). Phrenic nerve responses to electrical stimulation of the carotid sinus nerve in adult animals previously exposed to perinatal hyperoxia for the first 4 wk of postnatal development are not different from control animals (3, 6). Our results suggest that there may be a period during which the central respiratory circuit is affected by perinatal hyperoxic exposure, since we show that in vivo exposure alters ex vivo responses in the brainstem-spinal cord preparation in animals at P4 –5.


52


Hyperoxic exposure depressed the HVR (Phase 1) to anoxia in vivo and ex vivo via a reduction in frequency. While the intact hyperoxic exposed animals expired in less than 30 min of the anoxic challenge in vivo, the ex vivo anoxic challenge in our isolated brainstem preparation was not fatal until well beyond this time period. These effects found in the brainstem in the absence of afferent input suggest that perinatal hyperoxic exposure modulates the central respiratory system in the medulla at this age in rat pups. The pre-Bötzinger complex (PBC) is the primary inspiratory rhythm-generating kernel located in the ventral medulla (13, 22). Configuration of the PBC is influenced by postnatal age. Ex vivo electrophysiology studies using transverse-lateral slices and en bloc brainstem-spinal cord preparations (as we have done here) have helped to identify and characterize the rhythmic activity that originates from this region. In these ex vivo preparations taken from animals within the first week of life, the immature phenotype is described as eupneic activity or bursting activity in preparations superfused with physiological solutions equilibrated with 95% O2-5% CO2. Regardless of whether one considers the PBC to be the source of eupneic or gasp-like activity, the PBC provides a timing signal that drives the rest of the respiratory column early in life. The developmental changes in the PBC that occur over the first two postnatal weeks may result in reconfiguration of the inspiratory rhythm-generating circuitry to depend more heavily on inhibitory feedback to stabilize inspiratory and expiratory phases (13, 22). The PBC may have intrinsic hypoxic chemosensitivity which modulates phrenic motor output. Direct focal hypoxic exposure to the PBC results in several gasp-like excitatory phrenic responses ranging from augmented tonic firing to high amplitude, short duration bursts (13, 23) This observation suggest that a sub-population of the PBC neurons directly senses oxygen tension and this may facilitate changes in rhythm generation to rapidly increase ventilation compensating for hypoxia. Thus, during cerebral hypoxia the pacemaker activity changes to “gasping activity,” defined by a faster rise to burst, a shorter burst duration and slower burst frequency (13, 23). In the first week of life, the tonic respiratory burst phenotype changes from a longer and more frequent burst denoted as the immature phenotype into a slower and shorter burst phenotype that resembles the gasping activity observed when immature brainstems are challenged with hypoxia (24). This developmental change in activity is thought to be due to an uncoupling between motor outputs and the PBC (20, 22). During hypoxia, most of the respiratory neurons hyperpolarize, while smaller populations of neurons depolarize or are not affected (pacemaker neurons) thereby maintaining normal bursting behavior and typical membrane potential (20, 24). Here, we suggest that hyperoxic exposure reduced the total ex vivo baseline activity as well as the phrenic response to ex vivo anoxia via a greater reduction in burst frequency. Thus these results suggest that hyperoxic exposure may induce uncoupling between the pacemaker PBC and motor output from the medulla. Hyperoxic exposure may also reconfigure the network by affecting intranetwork properties such as cellular and molecular processes which modify connectivity and synaptic weights within the network, and may also affect other systems that directly influence respiratory network activity.

•

Bierman AM et al.

The design of this study does not allow us to delineate the precise mechanism of the hyperoxic-associated respiratory effects in the PBC or other respiratory-related neurons in the central respiratory network. In conclusion, this correlative study is the first to examine exposure to hyperoxia with deficiency of ventilatory timing and anoxic intolerance observed in vivo (at the youngest age, P4 –5). These results are further supported by using an ex vivo approach to determine an essential component of the central respiratory network associated with these deficiencies. We suggest that hyperoxic exposure during E18 to P4 –5 can hasten the uncoupling between the PBC and gasping centers of the medulla that subsequently activate respiratory motor output. The effect of hyperoxia in the first week of life has a central effect and this contributes to the reduced ventilatory frequency observed at baseline in vivo. Additionally, these hyperoxic-induced central nervous system effects impair the ventilatory response to anoxia in vivo, ultimately contributing to a hastened time to last gasp. Our findings have relevance to the human infant. Perinatal hyperoxic exposure during a critical window of development may reconfigure the respiratory network, decreasing the infant’s ability to sustain gasping responses during severe hypoxia, and thereby decreasing the possibility of effective autoresuscitation. ACKNOWLEDGMENTS The authors acknowledge the technical support of Catherine Mayer and Paulina Getsy for electrophysiology methods using the brainstem-spinal cord preparation at the Department of Neurosciences, Case Western University, and Ms. Ariel Mason for her technical assistance with the preparation at Johns Hopkins. The authors also acknowledge the technical support of Jason Kirkness for the ventilation analysis program designed with Igor for plethysmography breathing analysis developed at the Department of Pulmonary and Critical Care Medicine, Johns Hopkins Medical Institutions. GRANTS This work, in part, was supported by NIH Grant R25 DA021630. DISCLOSURES No conflicts of interest, financial or otherwise, are declared by the author(s). AUTHOR CONTRIBUTIONS Author contributions: A.M.B., C.G.T., C.G.W., R.C.-V., and E.B.G. conception and design of research; A.M.B., C.G.W., and E.B.G. performed experiments; A.M.B. and R.C.-V. analyzed data; A.M.B., C.G.T., C.G.W., R.C.-V., and E.B.G. interpreted results of experiments; A.M.B. prepared figures; A.M.B. drafted manuscript; A.M.B., C.G.T., C.G.W., R.C.-V., and E.B.G. edited and revised manuscript; A.M.B., C.G.T., C.G.W., R.C.-V., and E.B.G. approved final version of manuscript. REFERENCES 1. Bavis RW. Developmental plasticity of the hypoxic ventilatory response after perinatal hyperoxia and hypoxia. Respir Physiol Neurobiol 149: 287–299, 2005. 2. Bavis RW, Dmitrieff EF, Young KM, Piro SE. Hypoxic ventilatory response of adult rats and mice after developmental hyperoxia. Respir Physiol Neurobiol 177: 342–346, 2011. 3. Bavis RW, Mitchell GS. Long-term effects of the perinatal environment on respiratory control. J Appl Physiol 104: 1220 –1229, 2008. 4. Bavis RW, Simons JC. Developmental hyperoxia attenuates the hypoxic ventilatory response in Japanese quail (Coturnix japonica). Respir Physiol Neurobiol 164: 411–418, 2008. 5. Bavis RW, Young KM, Barry KJ, Boller MR, Kim E, Klein PM, Ovrutsky AR, Rampersad DA. Chronic hyperoxia alters the early and late phases of the hypoxic ventilatory response in neonatal rats. J Appl Physiol 109: 796 –803, 2010.


Perinatal Hyperoxia Contributes to Anoxic Intolerance 6. Bisgard GE, Olson EB, Wang ZY, Bavis RW, Fuller DD, Mitchell GS. Adult carotid chemoafferent responses to hypoxia after 1, 2, and 4 wk of postnatal hyperoxia. J Appl Physiol 95: 946 –952, 2003. 7. Bissonnette JM. Mechanisms regulating hypoxic respiratory depression during fetal and postnatal life. Am J Physiol Regul Integr Comp Physiol 278: R1391–R1400, 2000. 8. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72: 248 –254, 1976. 9. Caruana-Montaldo B. The control of breathing in clinical practice. Chest 117: 205–225, 2000. 10. Chavez-Valdez R, Mason A, Nunes AR, Northington FJ, Tankersley C, Ahlawat R, Johnson SM, Gauda EB. Effect of hyperoxic exposure during early development on neurotrophin expression in the carotid body and nucleus tractus solitarii. J Appl Physiol 112: 1762–1772, 2012. 11. Dmitrieff EF, Wilson JT, Dunmire KB, Bavis RW. Chronic hyperoxia alters the expression of neurotrophic factors in the carotid body of neonatal rats. Respir Physiol Neurobiol 175: 220 –227, 2011. 12. Dubois C, Houchi H, Naassila M, Daoust M, Pierrefiche O. Blunted response to low oxygen of rat respiratory network after perinatal ethanol exposure: involvement of inhibitory control. J Physiol 586: 1413–1427, 2008. 13. Feldman JL, Del Negro CA, Gray PA. Understanding the rhythm of breathing: so near, yet so far. Annu Rev Physiol 75: 423–452, 2012. 14. Gauda EB, Carroll JL, Donnelly DF. Developmental maturation of chemosensitivity to hypoxia of peripheral arterial chemoreceptors: invited article. Adv Exp Med Biol 648: 243–255, 2009. 15. Gauda EB, Shirahata M, Mason A, Pichard LE, Kostuk EW, ChavezValdez R. Inflammation in the carotid body during development and its contribution to apnea of prematurity. Respir Physiol Neurobiol 185: 120 –131, 2013. 16. Gresham K, Boyer B, Mayer C, Foglyano R, Martin R, Wilson CG. Airway inflammation and central respiratory control: results from in

17. 18. 19.

20. 21. 22. 23. 24. 25. 26. 27.

•

Bierman AM et al.

53

vivo and in vitro neonatal rat. Respir Physiol Neurobiol 178: 414 –421, 2011. Hofer MA. Sleep-wake state organization in infant rats with episodic respiratory disturbance following sinoaortic denervation. Sleep 8: 40 –48, 1985. Hofer MA. Role of carotid sinus and aortic nerves in respiratory control of infant rats. Am J Physiol Regul Integr Comp Physiol 251: R811–R817, 1986. Knudsen L, Weibel ER, Gundersen HJG, Weinstein FV, Ochs M. Assessment of air space size characteristics by intercept (chord) measurement: an accurate and efficient stereological approach. J Appl Physiol 108: 412–421, 2010. Peña F. Neuronal network properties underlying the generation of gasping. Clin Exp Pharmacol Physiol 36: 1218 –1228, 2009. Serra A, Brozoski D, Hedin N, Franciosi R, Forster HV. Mortality after carotid body denervation in rats. J Appl Physiol 91: 1298 –1306, 2001. Smith JC, Ellenberger HH, Ballanyi K, Richter DW, Feldman JL. Pre-Bötzinger complex: a brainstem region that may generate respiratory rhythm in mammals. Science 254: 726 –729, 1991. Solomon IC, Edelman NH, Neubauer JA. Pre-Bötzinger complex functions as a central hypoxia chemosensor for respiration in vivo. J Neurophysiol 83: 2854 –2868, 2000. St John WM. Noeud vital for breathing in the brainstem: gasping—yes, eupnoea— doubtful. Philos Trans R Soc Lond B Biol Sci 364: 2625–2633, 2009. Suzue T. Respiratory rhythm generation in the in vitro brainstem-spinal cord preparation of the neonatal rat. J Physiol 354: 173–183, 1984. Tankersley CG, Fitzgerald RS, Levitt RC, Mitzner WA, Ewert SL, Kleeberger SR. Genetic control of differential baseline breathing pattern. J Appl Physiol 82: 874 –881, 1997. Vulesevic B, Perry SF. Developmental plasticity of ventilatory control in zebrafish, Danio rerio. Respir Physiol Neurobiol 154: 396 –405, 2006.
