Chemoreceptors, Breathing and pH

The regulation of acid–base balance depends on renal mechanisms that determine net excretion of acid or base and breathing mechanisms that alter the partial pressure of carbon dioxide (PCO ₂ ) in body tissues. The level of ventilation (breathing) is regulated, in part, by pH-sensitive receptors called chemoreceptors, which are present peripherally at the carotid body and centrally within the hindbrain. Their physiology is the subject of this chapter. Arterial, and tissue, PCO ₂ is very sensitive to changes in ventilation, which, in turn, is very sensitive to changes in pH. Metabolic acidosis stimulates ventilation lowering PCO ₂ , which minimizes the acidosis, a classic feed-back control system. Chemoreception is surprisingly complex with an emerging understanding of the interdependence of peripheral and central sites and of the functions of multiple central sites.

Key Words

carbon dioxide, brainstem, bicarbonate, cerebrospinal fluid, acid–base balance, control of breathing, chemoreception

Introduction

Our thinking on mammalian acid–base regulation focuses on pH. Intracellular pH, which is of great importance in the maintenance of normal protein function, is regulated by membrane ion transporters, intracellular buffers, and by changes in cell metabolism. The success of intracellular pH regulation is dependent on the extracellular pH being constrained in its variation. Renal processes can alter extracellular proton and bicarbonate balance by changes in ion excretion. In this chapter we focus on how breathing can affect the partial pressure of carbon dioxide in extracellular fluid, here the partial pressure of carbon dioxide in arterial blood, PaCO ₂ . The traditional approach utilizes the straightforward relationships shown in Eqs 1 and 2 .

{CO}_{2} + H_{2} O = H_{2} {CO}_{3}^{-} = H^{+} + {HCO}_{3}^{-}

${CO}_{2} + H_{2} O = H_{2} {CO}_{3}^{-} = H^{+} + {HCO}_{3}^{-}$

pH = {pK}^{-} + \log ([{HCO}_{3}^{-}] / [S \times {PCO}_{2}])

$pH = {pK}^{-} + \log ([{HCO}_{3}^{-}] / [S \times {PCO}_{2}])$

where S =the solubility of CO ₂ in blood and pK ^– is a combined dissociation constant.

Note that pH can be affected by primary changes in either bicarbonate or PCO ₂ . Arterial bicarbonate changes by a small amount as CO ₂ is changed due to the presence of blood buffers. But by far the most important changes in arterial bicarbonate occur either as the result of a primary metabolic disturbance, which changes the ionic composition of blood thereby affecting bicarbonate, or as the result of the kidney, which can also change the ionic composition of blood thereby affecting bicarbonate.

CO ₂

Ventilation and CO ₂

Arterial blood pH is the clinically relevant extracellular variable because changes in breathing alter the PaCO ₂ directly. The PaCO ₂ is determined by the ratio of CO ₂ production and alveolar ventilation. For a constant metabolic rate (and CO ₂ production), an increase in alveolar ventilation will lower the PaCO ₂ and, conversely, a decrease in alveolar ventilation will elevate the PaCO ₂ . Both events have acid–base consequences. Hyperventilation eliminates CO ₂ faster than it is being produced and, as a result, pH increases; hypoventilation eliminates CO ₂ more slowly than it is being produced and, as a result, pH decreases. The changes are quantitatively speaking not small. If alveolar ventilation doubles, the normal PaCO ₂ of 40 mm Hg will quickly attain a new value of 20 mm Hg. Arterial pH will quickly change from the normal value of 7.40 to 7.70. Conversely, if alveolar ventilation is reduced by half, arterial PCO ₂ will quickly attain a new value of 80 mm Hg. Arterial pH will change from the normal value of 7.4 to 7.10. Changes in alveolar ventilation quickly and dramatically affect PaCO ₂ and pH. A 10% increase in alveolar ventilation will decrease PaCO ₂ by 4 mm Hg and increase arterial pH from 7.40 to 7.45 all within seconds. This change in arterial pH is determined by the decrease in PCO ₂ and by the effectiveness of blood buffers, which include most importantly the red blood cell hemoglobin concentration. In the absence of any protein buffers, the pH change would be much greater. If the hyperventilation and hypocapnia are sustained, there are secondary renal adjustments that act to excrete more bicarbonate and to lessen the initial alkalosis. These events take minutes to hours and are governed by the lowered PCO ₂ . Changes in PCO ₂ , whether decrease or increase, are quickly reflected in all blood and tissue compartments due to the diffusability of CO ₂ .

Tissue CO ₂

Tissue PCO ₂ levels are determined by the arterial value, the tissue rate of CO ₂ production, and the amount of tissue blood flow. Changes in blood flow and cell metabolism can affect cell PCO ₂ and pH for any given arterial PCO ₂ . For example, the brain has blood vessels that determine blood flow resistance and are very sensitive to changes in CO ₂ . An increase in PaCO ₂ would tend to directly cause an increase in tissue and cell PCO ₂ and a decrease in their pH. The strong vasodilatory action of high CO ₂ on cerebral vessels would decrease resistance and allow blood flow to increase, which would clear more CO ₂ and indirectly minimize the increase in tissue and cell PCO ₂ . Tissue and cell PCO ₂ and pH are determined by blood flow for any PaCO ₂ value and by metabolic rate.

CO ₂ and Ventilation

A change in pH is detected by a physiological process called chemoreception. The pH sensors are present in the peripheral blood at the bifurcation of the carotid artery, the carotid body, and within the brainstem at multiple locations. Both the carotid body and the brainstem respond to a fall in pH, for example, by increased stimulation of breathing, which lowers PaCO ₂ and tends to minimize the initial acidosis. This feedback control of the alveolar ventilation level, and hence the PaCO ₂ , by chemoreceptor-detected changes in pH is the essence of this chapter. The body can detect small pH changes at many sites and quickly bring about a change in breathing that acts to correct the initial perturbation. It is an appropriate response to correct pH. This system detects arterial pH at the carotid body and at the central chemoreceptors; it likely detects a pH value somewhere between the arterial and brain interstitial fluid pH.

Peripheral Chemoreceptors

Location of Carotid Body

The carotid body is well located to detect changes in arterial pH and PCO ₂ . In fact, this location at the bifurcation of the carotid body rapidly detects small changes in PaCO ₂ that reflect minor variations in the normal level of alveolar ventilation, and thus serves admirably as a feedback control detector site for the maintenance of a normal level of alveolar ventilation. This site is also a useful one for the detection of pH changes that reflect abnormal physiology, but for this purpose it is difficult to construct an argument that makes this anatomical location of special utility. In fact, a chemoreceptor site closer to the tissue location of altered metabolism in a metabolic acid–base disorder might theoretically be of greater use. For example, mixed venous CO ₂ receptors have long been sought for in order to explain the tight link between increased metabolism in muscular exercise and alveolar ventilation. Here venous CO ₂ levels rise; but arterial PCO ₂ remains normal or decreases slightly as alveolar ventilation changes to match the increase in metabolic rate. Sporadically, the discovery of mixed venous chemoreceptors has been reported, but none of these have stood the test of time. Specific central chemoreceptor locations in the brain might well reflect tissue and cell PCO ₂ and pH, that is, tissue chemoreceptors may be present.

Carotid Body Function

The carotid body is a fascinating tissue. It is quite small, and is difficult to find by gross anatomical dissection, but has a large metabolic rate and a high perfusion. This tissue is arguably the only and certainly the major detector of low O ₂ levels. Hypoxia strongly excites the carotid body with powerful stimulatory effects on breathing and arousal. It is the detector for the hypoxia emergency warning system, and also detects changes in PCO ₂ , including values that may drop below normal. In non-rapid eye movement sleep, apneas that occur within seconds of a transient hyperventilation have been attributed to hypocapnia sensed at the carotid body. Thus, in sleep there is a tonic nervous activity from the carotid body to the brain that maintains a normal level of ventilation. Transient diminution of this activity by brief hypocapnia can lead to apnea, the cessation of breathing. There is evidence as well for tonic carotid afferent activity that is important in the maintenance of appropriate levels of ventilation in wakefulness. Surgical removal of the carotid bodies in experimental animals and in humans results in a stable new steady state of hypoventilation. The animal then maintains a “normal” PaCO ₂ that is a few mmHg higher than in animals with intact peripheral chemoreceptors. This state is maintained despite the presence of central chemoreceptors. It seems that normal ventilation requires the tonic afferent input from peripheral chemoreceptors. In 1938, Heymans received the Nobel Prize for the discovery of the carotid body. He proposed that the carotid body was the major detector of CO ₂ , as well as for hypoxia.

Central Chemoreceptors

There is clear evidence for receptors within the brain that detect changes in PCO ₂ or pH and bring about a change in ventilation. The ventilatory response to changes in pH as mediated by central chemoreceptors is very sensitive. Figure 56.1 shows data obtained from unanesthetized goats in a series of classic experiments. Alveolar ventilation, VA, is plotted against the pH in cerebrospinal fluid. The shaded area presents data obtained during chronic metabolic acidosis and alkalosis maintained for days in the goats as they breathed air or inhaled CO ₂ from 0 to 10%. Note that in these steady-state conditions of chronic metabolic acid–base disorders, the response of alveolar ventilation to changes in cerebral pH is very sensitive—it doubles for a cerebrospinal fluid (CSF) pH change from 7.33 to 7.28. The X symbols in Fig. 56.1 will be discussed below.

Figure 56.1, The alveolar ventilation response to chronic acid–base disorders and to inhaled CO 2 is shown versus the pH measured in cerebrospinal fluid in the shaded area, which presents ±SEM (standard error of measurement) of 81 separate measurements obtained in five goats. The crosses derive from experiments in which the goats had ventriculo-cisternal perfusion of cerebrospinal fluid with differing bicarbonate concentrations. Alveolar ventilation in these cases is plotted versus the pH calculated for a site that lies three-quarters of the distance along the concentration gradient between ventricular cerebrospinal fluid and blood.

History: Ventrolateral Medulla

The presence of central chemoreceptors was first suggested by the continued presence of a ventilatory response to an increase in PaCO ₂ after surgical removal of the peripheral chemoreceptors and solidified by the presence of a ventilatory response to an acid load applied directly into the cerebral ventricles. Direct application of small pieces of cotton, soaked in an acidic solution, to various brainstem surfaces further localized the site of central chemoreception to the surface of the ventral lateral medulla, a site that still captures the imagination of investigators. All of these studies were performed under surgical anesthesia, which has a powerful depressant effect on the sensitivity of the ventilatory response to CO ₂ , and required very acidic stimuli. Further, the blood supply to the medulla arises from vessels on the ventral surface, which could have easily carried the stimulus to deeper structures. Cooling or coagulation of this area decreased ventilatory output dramatically in anesthetized animals.

The effects of anesthesia on chemosensitivity cannot be overemphasized. For example, Akilesh et al. found that in rats, the change in ventilation breathing 7% CO ₂ compared to breathing air is decreased by about 70% with the introduction of anesthesia. There are many similar examples of how much anesthesia affects the chemoreceptor response sensitivity. Nevertheless, these early studies demonstrated that chemoreception is present in the medulla, and it is accessible from the ventral medullary surface.

Widespread Central Chemoreception

More recently, a series of studies have led to the proposal that central chemoreception is a phenomenon that is widely present in the hindbrain. Before proceeding, a few clarifying definitions are warranted. A functional definition for chemoreception refers to a ventilatory response to a change in CO ₂ /pH. Chemosensitivity or chemodetection refers to the response of a described unit, say a particular type of neuron, to changes in CO ₂ /pH. Chemosensitivity and chemodetection do not necessarily translate to chemoreception, that is, the presence of a type of neuron with a known sensitivity in vitro to changes in CO ₂ /pH does not necessarily mean that the neuron is important in a chemoreceptor response, one that involves a change in ventilation in vivo .

A number of experimental approaches support the concept that central chemoreception is a widely distributed function within the hindbrain. Experiments using expression of the early gene, c-fos, following exposure to elevated CO ₂ , described the presence of activated neurons at locations near the ventral lateral medullary surface, as proposed in early studies, but also deeper in the brainstem at other locations including the locus ceruleus, nucleus tractus solitarius, medullary raphe, rostral aspect of the ventral respiratory group, and fastigial nucleus of the cerebellum. Studies of neurons in slice preparations of the medulla that included the deeper nucleus tractus solitarious, locus ceruleus, and medullary raphe have all described neurons that were excited by CO ₂ . These experimental approaches did not demonstrate chemoreception, however. The neurons expressing c-fos could have been “downstream” to the actual chemodetector cells, and the neurons in the slice preparations were not connected to a ventilatory output.

A series of experiments have utilized the approach of examining the ventilatory response to a small region of focal acidosis produced at various sites within the brainstem. While studying the role of carbonic anhydrase in cerebral pH regulation, Coates et al. noted that focal application of acetazolamide, an inhibitor of carbonic anhydrase, resulted in focal acidosis. They then used tiny 1-nl injections of acetazolamide to produce very focal acidosis in the brainstem of anesthetized cats and rats. The presence of a ventilatory response following such injections indicated the presence of chemoreception at that site. Central chemoreception was present at many locations, including the retrotrapezoid nucleus (RTN) just below the ventral medullary surface (a possible site for the older surface chemoreceptors), nucleus tractus solitarious, locus ceruleus, midline medullary raphe, rostral aspect of the ventral respiratory group, and fastigial nucleus in the core of the cerebellum (the same sites as described by the c-fos studies). Focal acidosis produced by the 1-nl injections of acetazolamide resulted in an increase in fictive ventilatory output in these anesthetized animals, which provided functional evidence for multiple central chemoreceptor locations. Brain pH measurements showed the tissue pH change to be similar to that observed with a 20-mm Hg increase in arterial PCO ₂ in anesthetized animals. A second approach to induce a focal acidosis in various regions of the brainstem utilizes reverse microdialysis with a CO ₂ laden artificial cerebrospinal fluid (aCSF) , which reduced brain tissue pH by an amount like that observed with an increase in arterial PCO ₂ of 5-6 mm Hg, a small stimulus intensity . This approach demonstrated ventilatory responses in the sites shown in Fig. 56.2 .

Figure 56.2, A simplified, schematic view of chemoreceptor sites that, when stimulated, can increase ventilation in wakefulness. Solid lines identify connections known to be present in wakefulness. LH, lateral hypothalamus; LC, locus ceruleus; NTS, nucleus tractus solitarius; CB, carotid body; PBC, pre-Bötzinger complex; rVRG, rostral ventral respiratory group; MR, medullary raphe; VLM, ventrolateral medulla.

Putative Chemoreceptor Cell Type

What cell type is responsible for the detection of CO ₂ or pH within the brainstem chemoreceptor sites? This seemingly simple question has proven difficult to answer because there are numerous possibilities. While it is difficult to study the response of a given cell type to acidic stimulation in a brainstem slice preparation or in cultured neurons, it is even more difficult to interpret in vivo findings of chemoreception. Nonetheless, there is in vitro and in vivo complementary evidence for involvement of serotonergic, noradrenergic, and glutamatergic neurons. This evidence is discussed below.

Serotonergic Neurons

Serotonergic neurons are chemosensitive when studied in slices taken from young rats or in neurons taken from newborn rats and grown in culture for several days. These neurons are in the midline medullary raphe in close proximity to large penetrating blood vessels. The blood supply of the medulla arises from vessels on the ventral surface that penetrate deep into the tissue supplying neurons through to the dorsal aspect of the medulla. A serotonergic neuron in culture will increase its firing rate with a mild CO ₂ stimulus, e.g., from 1 to 3 Hz. If expressed as the percent change in firing rate, these neurons are quite sensitive in this reduced preparation. This sensitivity in vitro , response to mild stimulus intensities, and anatomical proximity to large vessels has been interpreted to indicate that serotonergic neurons are major central chemoreceptors. In vivo data support this hypothesis. Serotonergic neurons can be specifically destroyed by injection of a cell toxin, saporin, which has been conjugated to an antibody to the serotonin transport protein (anti-SERT-SAP). The antibody recognizes the serotonergic cell, the conjugate is internalized, the saporin is released, and, over a period of days, the cell is killed. Such injections reduce the number of serotonergic neurons in the medullary raphe region of the rat by 28%, and also reduce the ventilatory response to CO ₂ by 15 and 18%, respectively, in wakefulness and in NREM sleep. Serotonergic neurons are clearly involved in chemoreception. These data do not prove that they are chemodetector neurons in vivo .

One can inhibit serotonergic neurons reversibly by stimulation of serotonin 1A receptors, which are primarily inhibitory autoreceptors, by direct application via reverse microdialysis of 8-OH-DPAT ((R)-(+)-8-hydroxy-2(di-n-propylamino)tetralin). When this is done in the medullary raphe region of the unanesthetized rat, the CO ₂ response is reduced. Dialysis of 10 mM 8-OH-DPAT reduces the CO ₂ response by ~20%, while dialysis of 30 mM 8-OH-DPAT reduces it by 40% (Taylor et al., unpublished observations). When DPAT is dialyzed in the medullary raphe region of the unanesthetized newborn piglet, the CO ₂ response is again reduced, but only in piglets older than seven days. In younger piglets, dialysis of DPAT increases the CO ₂ response. Serotonergic neurons are clearly involved in chemoreception but not in the early postnatal period. Again, these data do not prove that they are chemodetector neurons in vivo .

Finally, daily focal administration of the SERT inhibitor, fluoxetine, over 21 days in the unanesthetized rat, which should make more serotonin available in the medullary raphe region, results in an enhanced ventilatory response to CO ₂ .

Serotonergic neurons cell bodies are located in specific sites within the brainstem and have very widespread connections within the higher brain and spinal cord. They are most active in terms of firing rate in wakefulness and less so in NREM sleep, and are almost quiescent in REM sleep. Their main function remains incompletely understood; the concept that they may be chemodetectors important in ventilatory chemoreception is somewhat paradoxical.

In sum, these in vitro data showing the sensitivity of serotonergic neurons to CO ₂ but not proving their involvement in chemoreception, and these in vivo data showing that serotonergic neurons are involved in chemoreception but not demonstrating that they are directly chemosensitive, present a constellation of findings most easily interpreted as involving serotonergic neurons as chemodetectors. It is likely that these serotonergic neurons also modulate other aspects in regulating breathing and, in respect to chemoreception, they may act as modulators as well as detectors.

Glutamatergic Neurons

Glutamatergic neurons in the region of the retrotrapezoid nucleus (RTN) have also been proposed as the central chemosensitive neurons. The RTN, at first glance, is a rather unimpressive group of small, difficult-to-find neurons that lie in a small sliver of space between the ventral border of the facial nucleus and the ventral surface of the medulla. This group of neurons was discovered at about the same time by two quite different experimental approaches. In one, viral retrograde tracers injected into the phrenic motor nucleus demonstrated the presence of second-order neurons at this site; in the other, very small injections in the RTN of an excitatory amino acid toxin, kainic acid, stopped breathing and substantially reduced the ventilatory response to CO ₂ in anesthetized cats and rats. A series of studies then showed in unanesthetized rats that:

1.
Focal stimulation of the RTN by reverse microdialysis of a CO ₂ -rich artificial cerebrospinal fluid stimulated breathing, which indicated the presence of chemoreception in the RTN.
2.
Dialysis of the GABA-A receptor agonist muscimol decreased the ventilatory response to CO ₂ , supporting the presence of chemoreception in the RTN.
3.
Dialysis of the GABA-A receptor antagonist, bicuculline, stimulated breathing, which indicated the presence of a tonic GABAergic inhibition in the RTN.
4.
Stimulation of neuronal activity in the RTN by CO ₂ , which supported the presence of chemoreception in the RTN.
5.
Unilateral excitatory amino acid toxin-induced lesions reduced the ventilatory response to CO ₂ , but had no effect on baseline breathing or on the response to hypoxia supporting the presence of chemoreception in the RTN.
6.
Bilateral lesions of RTN neurons that express the neurokinin-1 receptor by injection of a cell-specific toxin, substance P (the ligand for the neurokinin-1 receptor) conjugated to the cell toxin saporin, decreased the ventilatory response to CO ₂ and induced hypoventilation, which supported the presence of chemoreception in the RTN.

Recent studies have identified the putative chemoreceptor neuron within the RTN as glutamatergic. In anesthetized rats, RTN neuronal activity was shown to be sensitive to CO ₂ delivered to the whole rat, even when other aspects of the respiratory control network were inhibited by systemic administration of a broad-spectrum glutamate receptor antagonist or by an opioid receptor antagonist. Further, the excitable neurons were shown by juxtacellular labeling to express the RNA message for the glutamate transport protein, VGLUT2, thereby identifying the neurons as glutamatergic. These authors also found that nearby serotonergic neurons were not responsive to CO ₂ , but they did not test serotonergic neurons at other sites where they are plentiful. Studies in brainstem slices also showed these glutamatergic neurons to be CO ₂ sensitive. The RTN is a central chemoreceptor site, and glutamatergic neurons in the RTN certainly seem to be involved in chemoreception and may be chemodetector neurons at this site.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here