Chronic exposure to intermittent hypoxia (CIH) has been used in animals to mimic the arterial hypoxemia that accompanies sleep apnea. Humans with sleep apnea and animals exposed to CIH have elevated blood pressures and augmented sympathetic nervous system responses to acute exposures to hypoxia. To test the hypothesis that exposure to CIH alters neurons within the nucleus of the solitary tract (NTS) that integrate arterial chemoreceptor afferent inputs, we measured whole cell currents induced by activation of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA) and N-methyl-d-aspartate (NMDA) receptors in enzymatically dispersed NTS neurons from normoxic (NORM) and CIH-exposed rats (alternating cycles of 3 min at 10% O2 followed by 3 min at 21% O2 between 8 AM and 4 PM for 7 days). To identify NTS neurons receiving carotid body afferent inputs the anterograde tracer 4- (4-(dihexadecylamino)styryl-N-methylpyridinum iodide (DiA) was placed onto the carotid body 1 wk before exposure to CIH. AMPA dose-response curves had similar EC50 but maximal responses increased in neurons isolated from DiA-labeled CIH (20.1 ± 0.8 μM, n = 9) compared with NORM (6.0 ± 0.3 μM, n = 8) rats. NMDA dose-response curves also had similar EC50 but maximal responses decreased in CIH (8.4 ± 0.4 μM, n = 8) compared with NORM (19.4 ± 0.6 μM, n = 9) rats. These results suggest reciprocal changes in the number and/or conductance characteristics of AMPA and NMDA receptors. Enhanced responses to AMPA receptor activation could contribute to enhanced chemoreflex responses observed in animals exposed to CIH and humans with sleep apnea.
- sleep apnea
sleep apnea is a significant health risk with ∼10% of the population affected by this disorder (44, 45). Whether central or peripheral in origin, the apnea results in arterial hypoxemia. Chronic exposure to intermittent hypoxia (CIH) during the nocturnal period is widely used in animals to mimic the arterial hypoxemia of sleep apnea (3, 10, 11, 16). Humans with sleep apnea and animals exposed to CIH have elevated blood pressure and sympathetic outflow (3, 6, 10, 11, 17, 18, 24). In both humans with sleep apnea (26, 38) and animals exposed to CIH (16), blood pressure and sympathetic nerve responses to acute hypoxia are enhanced. Therefore, sleep apnea in humans and CIH in animals appear to result in augmented arterial chemoreceptor activation of the sympathetic nervous system leading to a sustained increase in blood pressure (25).
Arterial chemoreceptors monitor levels of oxygen in the blood and initiate increases in sympathetic outflow during exposures to hypoxia (19). Arterial chemoreceptor afferent fibers make their first synapses within the central nervous system in the nucleus of the solitary tract (NTS) (22). Excitatory amino acid receptors, both the NMDA and non-NMDA subtypes, are the primary mediators of chemoreceptor afferent-evoked excitation of NTS neurons (42, 46, 47). We hypothesize that the enhanced sympathoexcitatory responses to acute hypoxia following exposure to CIH are mediated by enhanced responses of NTS neurons receiving arterial chemoreceptor inputs to excitatory amino acid agonists. To test this hypothesis, we examined the responses of NTS neurons to AMPA and NMDA after exposure to CIH.
Experiments were performed on 47 adult, male Sprague-Dawley rats (275–350 g, Charles River Laboratories, Wilmington, MA) housed two per cage in a fully-accredited (Association for Assessment and Accreditation of Laboratory Animal Care and United States Department of Agriculture) laboratory animal room with free access to food and water. Before being used for any procedures, rats were given at least 1 wk to acclimate. All experimental protocols were approved by the Institutional Animal Care and Use Committee.
Rats were anesthetized with medetomidine (0.5 mg/kg ip; Pfizer) and ketamine (75 mg/kg ip; Fort Dodge Laboratory) 1–2 wk prior to placement in the intermittent hypoxia chambers and under aseptic conditions, crystals of the anterograde fluorescent tracer DiA (Molecular Probes, Eugene OR) were applied to the carotid body region and sealed in place with Kwik-Sil (World Precision Instruments, Sarasota, FL) to visualize chemoreceptor synaptic terminals and neurons receiving these synaptic contacts, as we and others have previously described (20, 40). Anesthesia was terminated by atipamezole (1 mg/kg ip; Pfizer) at the conclusion of the surgical procedures. Postoperative analgesics (Nubaine, im) were available as needed.
Rats were divided into two groups: normoxic controls (NORM; n = 30) and rats exposed to CIH (n = 23). Hypoxic rats were housed in hypoxia chambers with ad libitum food and water access and were exposed to 8 h of CIH, using the paradigm described by Fletcher and colleagues (10, 11). Chambers were flushed with 100% N2 until a fractional O2 concentration of 8–10% was reached and maintained for 3 min. After that, the chambers were flushed with room air (21% O2) for 3 min. Exposures to CIH occurred during the light period (08:00–16:00) for 7 days. The animals were not exposed to CIH during the remainder of the light period (16:00–21:00) and during the dark period (21:00–07:00). NORM animals were kept in the same types of chambers as CIH-exposed rats, and gas switching occurred on the same time frame as in CIH-exposed rats. However, in NORM rats, gas switching was between sources of room air so that NORM rats always inspired room air (21% O2).
Dispersion of caudal NTS neurons.
Experiments were performed 7 days after placement in the intermittent hypoxia chambers. Rats were anesthetized with isoflurane, and the brain stem was rapidly removed and placed in ice-cold Krebs buffer. The brain stem was cut into 440-μm-thick transverse sections using a vibratome, and sections were incubated in PIPES buffer with trypsin (Sigma type XI, 3–5 mg/10 ml) for 55 min at ∼34°C. After enzyme treatment, slices were rinsed three times with incubation solution and maintained at room temperature (22–25°C) in a continuously oxygenated jar with PIPES buffer. Slices were placed on a glass slide, and the caudal NTS region was identified and excised by using a scalpel blade. The excised NTS region was gently triturated in a DMEM buffer using a series of fire-polished pipettes. Following trituration, an aliquot of the DMEM/neuron suspension was placed in the perfusion bath and allowed to settle for 5 min, after which superfusion of the cells was begun with a normal physiological solution.
Whole cell patch-clamp recordings were performed on acutely dissociated NTS neurons in an external solution consisting of (in mM) 140 NaCl; 33 d-glucose; 10 HEPES; 3 KCl; 2 CaCl2; and 2.4 MgCl2 (pH 7.4 with NaOH; osmolality 315–320 mosM). For NMDA protocols, MgCl2 was omitted. Electrodes were formed from thin-walled quartz glass with a laser-based puller (Sutter Instrument, Novato, CA) to resistances of 3–7 MΩ. Electrodes were filled with (in mM): 150 CsCl; 0.1 NaGTP; 4 NaATP; and 12 HEPES (pH 7.2 with CsOH; osmolality 280–300 mosM). Neurons were viewed under appropriate wavelength fluorescence, and cells exhibiting DiA-labeled somatic appositions were studied, as well as cells that did not exhibit DiA somatic appositions. Cells were voltage clamped at −60 mV by using an Axopatch 200B amplifier (Axon Instruments). Currents were expressed as current density (pA/pF) to normalize for cells of different sizes. Currents were low-pass-filtered at 1 kHz, sampled at 10 kHz, and analyzed using the pCLAMP version 8.1 software suite (Axon Instruments). If leak current was more than 50 pA, data were not included in the analysis. Prior to each drug application, a membrane test (Axon Instruments) was performed to ensure that access resistance did not change over time or during different experimental conditions.
We repeated the drug application protocol described in our previous publications (40, 41). AMPA-containing solutions consisted of (in mM) 132 NaCl; 33 d-glucose; 10 TEA·HCl; 10 HEPES; 3 KCl; 2 CaCl2; and 0.5 μM TTX. NMDA-containing solutions consisted of (in mM): 140 NaCl; 33 d-glucose; 10 HEPES; 3 KCl; 2 CaCl2; 0.5 μM glycine; and 0.5 μM TTX. AMPA and NMDA were applied by gravity flow to isolated cells via a two-barrel square glass pipette using a fast-step perfusion apparatus (Warner Instruments) so that rise time for the drug responses were <100 ms. Each pipette was attached via tubing to separate valves connecting each pipette to up to six (for AMPA) or seven (for NMDA) solutions. The first pipette was positioned adjacent and as close as possible to the target cell. A steady stream of AMPA-free or NMDA-free TTX containing 0.5 μM bath solution abolished action potential discharge and protected the neuron from leakage from the second downstream drug-containing pipette. The second pipette was connected to six or seven different concentrations of AMPA or NMDA, respectively. To apply AMPA or NMDA, the pipette assembly was rapidly moved, repositioning the TTX-containing pipette upstream to the cell and the AMPA or NMDA-ejecting pipette adjacent to the cell. The pipette assembly was returned to the starting position after a user-defined interval. Dose-response relationships were obtained during sequential, 5-s applications of 1, 3, 10, 30, 100, and 300 μM AMPA or 1, 3, 10, 30, 100, 300, and 1,000 μM NMDA with at least 2-min intervals between all drug applications. Preliminary experiments indicated that after exposure to drug, neurons returned to control values within 2 min. Time-control experiments indicated that neuron properties and drug responses were stable over the period of recording, and if neuronal properties and drug responses changed during the course of the recording, the neuron was excluded from analysis.
Intracellular polyamines, such as spermine, have been shown to modulate current flow through both NMDA and AMPA receptors (28). Therefore, current-voltage (I-V) relationships were obtained by using electrodes that contained spermine (100 μM) and electrodes that did not contain spermine.
Dose-response curves were fit using where I is the AMPA- or NMDA-induced current, Imax is the maximal AMPA or NMDA current, Imin is the minimal AMPA or NMDA current, A is the concentration of AMPA or NMDA, and n is the Hill slope. Curve fits were obtained by a nonlinear regression using a four-parameter logistic function. Statistical significance of differences between dose-response curves were determined using two-way ANOVA with the Holim-Sidak test used for post hoc comparisons. Percentile data were analyzed by Student's t-test following logarithmic transformation. All values are expressed as means ± SE, and significance was accepted at P < 0.05.
The membrane capacitance of neurons provides an approximate index of cell size, and there was no difference in neurons obtained from normoxic (9.77 ± 0.97 pF, n = 45) compared with hypoxic (12.02 ± 1.07 pF, n = 34) (P = 0.126) rats.
Responses to AMPA.
Sequential increases in the concentration of AMPA increased the steady-state level of inward current (Fig. 1A). Curve fitting of the dose-response curves for cells exhibiting DiA labeling revealed no difference in the EC50 of cells isolated from NORM rats (13.4 ± 2.9 μM, n = 9) compared with cells isolated from CIH rats (19.6 ± 2.3 μM, n = 8) (P = 0.07) (Fig. 2). However, the maximum AMPA-evoked current was significantly greater in cells isolated from CIH rats (20.1 ± 0.8 pA/pF) compared with NORM (6.0 ± 0.3 pA/pF) (P < 0.001).
Similar findings were obtained in NTS neurons that did not exhibit DiA labeling. Curve fitting of the dose-response curves for such cells revealed no difference in the EC50 of cells isolated from NORM rats (13.0 ± 2.4 μM, n = 26) compared with cells isolated from CIH rats (12.6 ± 0.6 μM, n = 30) (P = 0.86) (Fig. 2). However, the maximum AMPA-evoked current was significantly greater in cells isolated from CIH rats (15.8 ± 0.2 pA/pF) compared with NORM (5.7 ± 0.3 pA/pF) (P < 0.001).
I-V relationships for AMPA-evoked currents were linear in neurons from NORM and CIH rats (Fig. 3) and inclusion of spermine in the patch pipette (100 μM) did not change the shape of the I-V relationship (n = 7). The slope of the I-V curve was less in NTS neurons isolated from NORM [0.05 ± 0.01 (pA/pf)/mV, n = 15] compared with CIH rats [0.33 ± 0.01 (pA/pf)/mV, n = 11] (P = 0.001). AMPA-evoked currents (30 μM, n = 6) were blocked by the non-NMDA receptor antagonist 6-cyano-7-nitroquinoxaline-2,3-dione (100 μM).
Application of cyclothiazide (100 μM) to block AMPA receptor desensitization markedly enhanced the amplitude of AMPA-evoked currents in NTS neurons isolated from NORM (318 ± 37%, n = 5) and CIH (268 ± 92%, n = 8) rats (Fig. 4). The percentage increase in AMPA-evoked current by cyclothiazide was not different compared with neurons from NORM and CIH rats (P = 0.71).
Responses to NMDA.
Sequential increases in the concentration of NMDA increased the steady-state level of inward current (Fig. 5). Curve fitting of the dose-response curves for cells exhibiting DiA labeling revealed no difference in the EC50 of cells isolated from NORM rats (75.9 ± 7.3 μM, n = 8) compared with cells isolated from CIH rats (55.0 ± 8.0 μM, n = 9) (P = 0.08) (Fig. 6). However, the maximum NMDA-evoked current was significantly reduced in cells isolated from CIH rats (8.4 ± 0.4 pA/pF) compared with NORM (19.4 ± 0.6 pA/pF) (P < 0.001).
Similar findings were obtained in NTS neurons that did not exhibit DiA labeling. Curve fitting of the dose-response curves for such cells revealed no difference in the EC50 of cells isolated from NORM rats (67.6 ± 11.2 μM, n = 23) compared with cells isolated from CIH rats (70.3 ± 6.5 μM, n = 15) (P = 0.86) (Fig. 6). However, the maximum NMDA-evoked current was reduced in cells isolated from CIH rats (8.0 ± 0.2 pA/pF) compared with NORM (15.2 ± 0.7 pA/pF) (P < 0.001).
I-V relationships for NMDA-evoked currents studied in Mg2+-free solution were linear in neurons from NORM and CIH rats (Fig. 7). The slope of the I-V curve was less in neurons isolated from CIH [0.14 ± 0.01 (pA/pf)/mV, n = 11] compared with NORM rats [0.33 ± 0.02 (pA/pf)/mV, n = 13] (P = 0.001).
Exposure to CIH is used to mimic the arterial hypoxemia that accompanies sleep apnea. CIH in rats results in elevated basal levels of sympathetic outflow and blood pressure, as well as enhanced responses to acute exposures to hypoxia (3, 10, 11, 16). These same changes have been reported in humans with sleep apnea (6, 17, 18, 24–26).
It has been proposed that enhanced arterial chemoreflex activity can lead to persistent increases in sympathetic outflow, which could contribute to the hypertension associated with CIH and/or sleep apnea. The carotid body chemoreceptors could contribute to persistent activation of sympathetic outflow following CIH, as enhanced arterial chemoreceptor responses to hypoxia have been observed following comparable exposures to CIH rats (16, 33) and in humans with sleep apnea (25, 26). Enhanced chemoreceptor responses to acute hypoxia have been reported after exposure to CIH (29–31). Since the transmitter released by chemoreceptor afferents within the NTS is assumed to be glutamate (42, 46, 47), our results suggest that central alterations in glutamatergic transmission within the NTS could also contribute to enhanced chemoreflex responses.
The results of the present study indicate that following 7 days of exposure to CIH, the responses of NTS neurons to activation of AMPA receptors are enhanced within the voltage range of −30 mV to −120 mV. These changes occur with no obvious alteration in the level of desensitization of AMPA receptors at rest. In addition, following exposure to CIH, NTS neuronal responses to activation of NMDA receptors are reduced in the voltage range of −20 to −120 mV. The responses to AMPA and NMDA in the subpopulation of NTS neurons with DiA-labeled presumptive boutons were not different from those observed in non-DiA-labeled neurons, suggesting that the observed changes in NTS neurons were independent of whether the neuron received a peripheral arterial chemoreceptor input.
The lack of rectification in the I-V relationship during application of AMPA suggests that the receptors contain the GluR2 subunit (36) and that this does not change as a result of exposure to CIH. Intracellular polyamines, such as spermine, are responsible for inward rectification of calcium-permeable AMPA receptors (28). The lack of rectification observed in the present studies was not due to dialysis of endogenous polyamines, as addition of spermine to the electrode solution did not alter the I-V relationship. GluR1 and GluR2/3 AMPA receptor subunits have been localized within the NTS (2, 13, 21, 35, 43); however, systemic administration of AMPA receptor antagonists has little effect on the hypoxic ventilatory response under normal conditions (13, 43). The limitations of interpretation of the effects of systemic administration of antagonists notwithstanding, this suggests that following exposure to CIH, AMPA-mediated neurotransmission within the NTS may play a more prominent role than previously considered.
The functional significance of these changes in agonist-evoked currents are best considered in the context of the voltage range over which the changes are evident. A neuron with a resting membrane potential of −60 to −80 mV will respond to application of AMPA with significantly enhanced current responses. While we have shown reduced NMDA-evoked currents in Mg2+-free solution at hyperpolarized voltage levels, it is possible that these changes will have minimal physiological consequence, as in normal, Mg2+-containing solution one does not observe NMDA-evoked currents until potentials of less than ∼0 mV are reached. The I-V relationship suggests that at depolarized potentials, NMDA receptors on NTS neurons from CIH rats conduct less current than neurons from normoxic rats; however, it is not possible to obtain dose-response curves at depolarized potentials to directly test this.
One consequence of the reciprocal changes observed in AMPA- vs. NMDA-induced currents could be that during activation of arterial chemoreceptors, AMPA receptor-induced current flow provides the predominant depolarizing drive to the cell. NMDA-induced current is highly sensitive to reduction/oxidation modulation, with oxidation decreasing current flow (9), while AMPA receptors are not very sensitive to reduction/oxidation modulation (1). Prabhakar and colleagues (29, 31) have shown that exposure to CIH induces oxidative stress which increases the sensitivity of the arterial chemoreceptors to hypoxia and augments the respiratory responses to hypoxia. Similar oxidative stress could occur within the NTS following exposure to CIH, and this could underlie the changes observed in NMDA receptor function. An in vivo analysis of NTS neuronal responses to chemoreceptor activation in rats exposed to CIH would provide insights into the functional consequences of enhanced AMPA and reduced NMDA induced currents.
Besides oxidative stress, what other potential stimuli might initiate the changes in amino acid receptor responses we observed? Enhanced discharge of arterial chemoreceptors at rest and in response to hypoxia following exposures to CIH could increase levels of excitatory amino acids (23), GABA (39), angiotensin (34), or other neuromodulators/hormones stimulated by arterial chemoreflexes. Therefore, the changes in excitatory amino acid receptor characteristics could be hormonally mediated. In addition, NTS neurons have been shown to directly respond to hypoxia (27).
The observed changes could reflect a change in the number of binding sites for AMPA and NMDA. In addition, the phosphorylation status of the receptors could alter the amount of current conducted. Changes within the NTS in signal transduction systems that can mediate amino acid receptor phosphorylation have been reported in rats following acute exposures to hypoxia (8, 12, 14, 15), and these systems, and/or others, might enhance AMPA-mediated responses in NTS neurons following CIH. Phosphorylation can increase the affinity and conductance of AMPA receptors (7, 37). Phosphorylation can also alter responses by increasing receptor trafficking resulting in insertion of AMPA receptors in the membrane (4, 5). Our data are more consistent with changes in receptor number or receptor subunit composition that lead to increased conductance, as responses at each dose of AMPA were increased with no changes in EC50, or the affinity, of the receptor. Increases in NMDA receptor subunit expression within the NTS have been reported following similar exposures to CIH (32) and are not consistent with the reduced NMDA receptor responses observed in the present study. However, the study examined receptor expression in a fairly large area of the dorsal hindbrain, and the result may therefore not be representative of what occurs in a subpopulation of neurons within the area. Analysis of the role of these potential factors in the changes we report would be a useful future direction.
The fact that the changes in AMPA- and NMDA-evoked responses were observed in NTS neurons that did not exhibit DiA labeling suggests that the changes are not dependent upon whether or not the neuron receives an arterial chemoreceptor afferent input. The ubiquitous changes in AMPA and NMDA responses in both DiA- and non-DiA-labeled neurons suggest that the NTS neuronal response to all inputs should be enhanced, providing the inputs to activate AMPA receptors. For example, NTS neurons could exhibit enhanced responses to arterial baroreceptor, cardiopulmonary, and gastrointestinal afferent inputs or descending excitatory inputs from the hypothalamus. Analysis of other reflex responses in rats exposed to CIH could answer this question.
However, it is possible that the changes in excitatory amino acid responses we report are only functionally significant if the alterations in neuronal responses are coupled with enhanced afferent input. This would confer functional specificity to those neurons receiving arterial chemoreceptor afferent inputs. Additional changes might occur in other ligand-gated or voltage-gated channels that further influence the ultimate response of a neuron to a depolarizing input. As one example, we have reported that the sensitivity of NTS neurons to activation of GABAA receptors is reduced following chronic exposures to sustained hypoxia (40). Since the response of many NTS neurons to activation of arterial chemoreceptor inputs is a depolarization followed by GABAA receptor-mediated hyperpolarization (22), in rats chronically exposed to sustained hypoxia the reduced GABAergic inhibition could lead to an enhanced NTS neuronal response to chemoreceptor activation. The extent to which exposures to CIH alter GABAergic inhibition of NTS neurons, or any other determinant of neuronal function, could influence the ultimate response of the neuron. The present report should be considered a first step on the road to documenting adaptations of central neurons following exposures to CIH. Attempts to try and completely define neuronal responses on the basis of one observed alteration may be misleading.
In conclusion, exposure to CIH for 7 days alters the responses of NTS neurons to exogenous application of the inotropic excitatory amino acid agonists AMPA and NMDA. The changes occur in, but were likely not restricted to, NTS neurons receiving arterial chemoreceptor afferent inputs. If these changes are reflected in synaptic inputs to NTS neurons then this observation could provide an explanation for the enhanced chemoreceptor reflexes observed after chronic exposures to intermittent hypoxia (16, 26, 38).
This work was supported by National Heart, Lung, and Blood Institute HL-41894. Patrícia M. de Paula was supported by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior Grant CAPES-1644-03-3, Brazil and is a postdoctoral fellow.
The authors gratefully acknowledge the expert technical assistance of Jaci A. Castania, Myrna Herrera-Rosales, and Olga Tolstykh.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society