Hypercapnia causes cellular oxidation and nitrosation in addition to acidosis: implications for CO2 chemoreceptor function and dysfunction

Cellular mechanisms of CO2 chemoreception are discussed and debated in terms of the stimuli produced during hypercapnic acidosis and their molecular targets: protons generated by the hydration of CO2 and dissociation of carbonic acid, which target membrane-bound proteins and lipids in brain stem neurons. The CO2 hydration reaction, however, is not the only reaction that CO2 undergoes that generates molecules capable of modifying proteins and lipids. Molecular CO2 also reacts with peroxynitrite (ONOO−), a reactive nitrogen species (RNS), which is produced from nitric oxide (•NO) and superoxide (•O2−). The CO2/ONOO− reaction, in turn, produces additional nitrosative and oxidative reactive intermediates. Furthermore, protons facilitate additional redox reactions that generate other reactive oxygen species (ROS). ROS/RNS generated by these redox reactions may act as additional stimuli of CO2 chemoreceptors since neurons in chemosensitive areas produce both •NO and •O2− and, therefore, ONOO−. Perturbing •NO, •O2−, and ONOO− activities in chemosensitive areas modulates cardiorespiration. Moreover, neurons in at least one chemosensitive area, the solitary complex, are stimulated by cellular oxidation. Together, these data raise the following two questions: 1) do pH and ROS/RNS work in tandem to stimulate CO2 chemoreceptors during hypercapnic acidosis; and 2) does nitrosative stress and oxidative stress contribute to CO2 chemoreceptor dysfunction? To begin considering these two issues and their implications for central chemoreception, this minireview has the following three goals: 1) summarize the nitrosative and oxidative reactions that occur during hypercapnic acidosis and isocapnic acidosis; 2) review the evidence that redox signaling occurs in chemosensitive areas; and 3) review the evidence that neurons in the solitary complex are stimulated by cellular oxidation.

Focal acidosis in the pre-Bötzinger complex area of awake goats induces a mild tachypnea

There are widespread chemosensitive areas in the brain with varying effects on breathing. In the awake goat, microdialyzing (MD) 50% CO2 at multiple sites within the medullary raphe increases pulmonary ventilation (V̇i), blood pressure, heart rate, and metabolic rate (V̇o2) ( 11 ), while MD in the rostral and caudal cerebellar fastigial nucleus has a stimulating and depressant effect, respectively, on these variables ( 17 ). In the anesthetized cat, the pre-Bötzinger complex (preBötzC), a hypothesized respiratory rhythm generator, increases phrenic nerve activity after an acetazolamide-induced acidosis ( 31 , 32 ). To gain insight into the effects of focal acidosis (FA) within the preBötzC during physiological conditions, we tested the hypothesis that FA in the preBötzC during wakefulness would stimulate breathing, by increasing respiratory frequency (f). Microtubules were bilaterally implanted into the preBötzC of 10 goats. Unilateral MD of mock cerebral spinal fluid equilibrated with 6.4% CO2 did not affect V̇i, tidal volume (Vt), or f. Unilateral MD of 25 and 50% CO2 significantly increased V̇i and f by 10% ( P < 0.05, n = 10, 17 trials), but Vt was unaffected. Bilateral MD of 6.4, 25, or 50% CO2 did not significantly affect V̇i, Vt, or f ( P > 0.05, n = 6, 6 trials). MD of 80% CO2 caused a 180% increase in f and severe disruptions in airflow ( n = 2). MD of any level of CO2 did not result in any significant changes in mean arterial blood pressure, heart rate, or V̇o2. Thus the data suggest that the preBötzC area is chemosensitive, but the responses to FA at this site are unique compared with other chemosensitive sites.

Intracellular pH response to hypercapnia in neurons from chemosensitive areas of the medulla

We investigated whether neurons in two chemosensitive areas of the medulla oblongata [nucleus of the solitary tract (NTS) and ventrolateral medulla (VLM)] respond to hypercapnia differently than neurons in two nonchemosensitive areas of the medulla oblongata [inferior olive (IO) and hypoglossal nucleus (Hyp)]. Medullary brain slices from preweanling Sprague-Dawley rats were loaded with 2',7'-bis(carboxyethyl)-5(6)-carboxyfluorescein, and intracellular pH (pHi) was followed in individual neurons at 37 degrees C with the use of a fluorescence imaging system. Most neurons from the NTS and VLM did not exhibit pHi recovery when CO2 was increased from 5 to 10% at constant extracellular HCO3- concentration [extracellular pH (pHo) decreased approximately 0.3 pH unit] (hypercapnic acidosis). However, when CO2 was increased from 5 to 10% at constant pHo (isohydric hypercapnia), pHi recovery was seen. In contrast, all neurons from the IO and Hyp exhibited pHi recovery during hypercapnic acidosis. All pHi recovery in the four areas studied was inhibited by 1 mM amiloride and unaffected by 0.5 mM 4,4'-diisothiocyanostilbene-2,2'-disulfonic acid. These data indicate that 1) pHi regulation differs between neurons in chemosensitive (NTS and VLM) and nonchemosensitive (IO and Hyp) areas of the medulla, 2) pHi recovery is due solely to Na+/H+ exchange in all four areas, and 3) Na+/H+ exchange is more sensitive to inhibition by extracellular acidosis in NTS and VLM neurons than in IO and Hyp neurons.

Central effects of endothelin on respiratory output during development

Both endothelin-1 protein and endothelin-1 specific binding sites have been identified in areas of the medulla oblongata involved in respiratory control. We examined whether endothelin acting centrally affects respiratory output during early postnatal life. We initially examined the effect of intracisternally administrated endothelin on respiratory output in 10 2- to 18-day-old piglets. Endothelin-1 administration at 50 nmol to 1 mumol caused respiratory inhibition. We subsequently examined whether this response is mediated through chemosensitive areas of the ventral medulla. Endothelin-1 was microinjected into specific ventral or dorsal medullary regions in 31 14- to 22-day-old piglets. Microinjection of endothelin-1 (10 fmol to 0.1 pmol) just above the hypoglossal roots, lateral to the pyramids, and within 1 mm from the surface (n = 24) attenuated respiratory output, and complete apnea occurred with 1 pmol in all animals. However, microinjection of endothelin-1 3 mm below the ventral surface (n = 5) and into the dorsal medulla (n = 3) had no inhibitory effect. Comparable doses of angiotensin II (n = 5) and norepinephrine (n = 5) microinjected into the endothelin-1 sensitive area also did not influence respiratory output. These effects of endothelin-1 were not altered by prior endothelin-B receptor blockade (IRL-1038) but could be reversed by endothelin-A receptor blockade (BQ-610). These results suggest that endothelin-1 release may cause ventilatory depression mediated through endothelin-A receptors located in the chemosensitive areas of the ventrolateral medulla.

Central effects of somatostatin and atrial natriuretic peptide on tracheal tone

The effects of somatostatin and atrial natriuretic peptide applied topically to the ventral surface of the medulla (VMS) on tracheal tone and phrenic nerve activity (Phr) were studied in chloralose-anesthetized and paralyzed cats artificially ventilated with 7% CO2 in O2. Surface application of drugs to the chemosensitive areas of the VMS significantly decreased tracheal tension measured by changes in pressure in a balloon placed in a bypassed segment of the trachea (Ptseg). Application of somatostatin (9 cats) caused a mean decrease in Ptseg from 17.3 +/- 1.8 (SE) to 4.3 +/- 1.4 cmH2O (P < 0.01) and a reduction in Phr from 24.9 +/- 3.4 to 10.3 +/- 3.4 units (P < 0.05). Like somatostatin, application of atrial natriuretic peptide to the VMS (5 cats) produced tracheal relaxation (Ptseg decreased from 19.3 +/- 2.6 to 9.9 +/- 1.3 cmH2O, P < 0.01), but in contrast there was an insignificant reduction in Phr (from 18.5 +/- 3.6 to 16.1 +/- 3.8 units, P > 0.05). When parasympathetic activity was abolished by atropine methylnitrate and tracheal tone was restored with 5-hydroxytryptamine, somatostatin and atrial natriuretic peptide applied on the VMS had no effect on tracheal pressure, suggesting that observed changes were not caused by direct action of peptides on tracheal smooth muscle via the bloodstream or by facilitation of inhibitory pathways. Both somatostatin and atrial natriuretic peptide applications were associated with a slight but significant decrease in arterial blood pressure. These data suggest that somatostatin and atrial natriuretic peptide acting on the chemosensitive structure of the VMS may play significant roles in modulating para-sympathetic outflow to airway smooth muscle.

Widespread sites of brain stem ventilatory chemoreceptors

We produced local tissue acidosis in various brain stem regions with 1-nl injections of acetazolamide (AZ) to locate the sites of central chemoreception. To determine whether the local acidosis resulted in a stimulation of breathing, we performed the experiment in chloralose-urethan anesthetized vagotomized carotid-denervated (cats) paralyzed servo-ventilated cats and rats and measured phrenic nerve activity (PNA) as the response index. Measurements of extracellular brain tissue pH by glass microelectrodes showed that AZ injections induced a change in pH at the injection center equivalent to that produced by an increase in end-tidal PCO2 of approximately 36 Torr and that the change in brain pH was limited to a tissue volume with a radius of < 350 microns. We found AZ injections sites that caused a significant increase in PNA to be located 1) within 800 microns of the ventrolateral medullary surface at locations within traditional rostral and caudal chemosensitive areas and the intermediate area, 2) within the vicinity of the nucleus tractus solitarii, and 3) within the vicinity of the locus coeruleus. Single AZ injections produced increases in PNA that were < or = 69% of the maximum value observed with an increase in end-tidal PCO2. We conclude that central chemoreceptors are distributed at many locations within the brain stem, all within 1.5 mm of the surface, and that stimulation of a small fraction of all central chemoreceptors can result in a large ventilatory response.

Ventrolateral medullary lesions and fastigial cardiovascular response in beagles

These studies were designed to verify that the putative vasomotor center in the rostral ventrolateral medulla (RVLM) contained the outflow paths for the fastigial nucleus (FN) sympathoexcitatory cardiovascular response. If so, then lesions placed by radiofrequency heating (75 degrees C) or application of kainic acid (40 mM) pledgets would reduce or ablate the pressor-tachycardia response after electrical stimulation of FN. Anesthetized beagles (alpha-chloralose, 115 mg/kg) were used in this study to maintain consistency in medullary brain stem anatomy. A ventral brain stem approach exposed the three chemosensitive zones of Mitchell, Schlaefke, and Loeschcke. In other species the intermediate zone and a portion of the rostral zone underlie the Hokfelt C1 cell group and the putative RVLM vasomotor center. Heart rate, arterial and left ventricular (LV) pressures, and maximal rate of pressure development (LV dp/dt) increased 14-84% above control levels in response to stimulating FN 5-10 times the stimulation threshold. The cardiovascular response was abolished in four of six dogs that received bilateral radiofrequency lesions at a depth of 1-2 mm. In five of seven dogs that received kainate surface lesions, the response was substantially reduced but not abolished. These lesions were effective only in the RVLM, above the corresponding intermediate, but not adjacent rostral or caudal chemosensitive areas. The data support the hypotheses of others that an epinphrine-containing cell group in this region is a final common pathway of sympathoexcitation. Expression of the FN cardiovascular response is primarily mediated through this vasomotor region previously identified by others in the rat, cat, rabbit, and primate.(ABSTRACT TRUNCATED AT 250 WORDS)

Pirenzepine prevents diethyl pyrocarbonate inhibition of central CO2 sensitivity

Application by pledget of the M1-antimuscarinic receptor agent pirenzepine (40 mM) to the rostral chemosensitive areas of the ventrolateral medulla in anesthetized, paralyzed, vagotomized, glomectomized, and servoventilated cats inhibited the slope of the integrated phrenic response to CO2 by 32.5% (P less than 0.03) and the maximum value by 21.1% (P less than 0.01). Similar application of the imidazole-histidine blocking agent diethyl pyrocarbonate (DEPC) decreased the slope by 40.3% (P less than 0.01) and the maximum value by 29.3% (P less than 0.05). Both responses confirm previous results. DEPC treatment decreased the effectiveness of subsequent pirenzepine application such that although slope and maximum were further decreased, the values were not significantly different from those after DEPC. Pirenzepine treatment prevented any subsequent DEPC inhibitory effect. The results raise the possibility that the inhibitory effects of DEPC on CO2 chemosensitivity are via muscarinic receptors and that muscarinic receptor involvement in CO2 chemosensitivity requires the presence of imidazole-histidine. Analysis by scintillation counting of successive 100-micron sections of medulla after rostral area application of [3H]pirenzepine indicated that the pirenzepine and DEPC effects are most probably within 2.0 mm of the ventral surface as measured from the midline, well away from the dorsal and ventral respiratory group neurons.

Kainic acid on the rostral ventrolateral medulla inhibits phrenic output and CO2 sensitivity

We used the neurotoxin, kainic acid, which is known to stimulate neuronal cell bodies as opposed to axons of passage by binding to specific amino acid receptors to determine whether cells with such receptors have access to the ventrolateral medullary surface and are involved in central ventilatory chemosensitivity. Pledgets with 4.7 mM kainic acid were placed bilaterally on the rostral, intermediate, or caudal ventilatory chemosensitive areas for 1-2 min in chloralose-urethan-anesthetized, paralyzed, vagotomized, glomectomized, and servo-ventilated cats. Application of kainic acid on the caudal or intermediate areas produced no consistent significant effects on eucapnic phrenic output or on the slope or maximum value of the phrenic nerve response to increased end-tidal PCO2. Rostral area kainic acid produced immediate augmentation and then diminution of blood pressure and phrenic output. Apnea developed in six of nine cats by 40 min. In all five cats in which it could be tested, the slope of the CO2 response was clearly decreased. Of [3H]kainic acid applied to the rostral area, 88.4% was shown to be within 2 mm of the ventral surface. Comparison of surface application sites of this and other studies suggests that an area overlapping the border of the original rostral and intermediate areas allows access to neurons involved in the chemoreception process, which may also provide tonic facilitatory input to cardiorespiratory systems.