Respiratory Control

Central and Peripheral Mechanisms

Dexter F. Speck, Michael S. Dekin, W. Robert Revelette, Donald A. Frazier

• Publisher: The University Press of Kentucky
• Year: 2021
• ISBN:
  9780813181776
• Topic: Medicine
• Subtopic: Physiology

Understanding of the respiratory control system has been greatly improved by technological and methodological advances. This volume integrates results from many perspectives, brings together diverse approaches to the investigations, and represents important additions to the field of neural control of breathing.

Topics include membrane properties of respiratory neurons, in vitro studies of respiratory control, chemical neuroanatomy, central integration of respiratory afferents, modulation of respiratory pattern by peripheral afferents, respiratory chemoreception, development of respiratory control, behavioral control of breathing, and human ventilatory control.

Forty-seven experts in the field report research and discuss novel issues facing future investigations in this collection of papers from an international conference of nearly two hundred leading scientists held in October 1990. This research is of vital importance to respiratory physiologists and those in neurosciences and neurobiology who work with integrative sensory and motor systems and is pertinent to both basic and clinical investigations.

Respiratory Control is destined to be widely cited because of the strength of the contributors and the dearth of similar works.

PART I

Membrane Properties of Respiratory Neurons

1

An Overview

Michael S. Dekin

A major goal of respiratory neurophysiologists has been to characterize the membrane properties of respiratory neurons and associate these properties with models of the respiratory central pattern generator (respiratory CPG). Much of this effort has been directed towards describing intrinsic cellular mechanisms underlying rhythm generation such as endogenous pacemaker activity and postinhibitory rebound phenomena. More recently we have also begun to appreciate the role of membrane properties in nonrhythmgenerating functions such as pattern formation. In this regard, it is now recognized that many respiratory neurons do not act as high fidelity followers of their synaptic inputs. Rather, these neurons possess an impressive repertoire of intrinsic mechanisms for integrating their synaptic inputs (8). These integrative capabilities arise largely as a consequence of the kinetic and neurochemical properties of specific membrane channels found in respiratory neurons. In this brief overview, the experimental strategies used to study membrane channels will be discussed with particular attention given to their strengths and weaknesses. In addition, the types of information gained from such studies will be considered in terms of what they can tell us about the respiratory CPG.

EXPERIMENTAL APPROACHES

Over the past twenty years, it has become apparent that neurons possess many types of membrane channels in addition to the classic sodium and potassium channels responsible for generating the action potential (11). These additional membrane channels include several classes of voltage-dependent potassium (2) and calcium channels (14), as well as channels such as the m-channel (1) which exhibit both chemical and voltage-dependent gating properties. Another important observation is that not all neurons necessarily possess each of the known channel types. A major emphasis of contemporary neurobiology, therefore, has been to determine the unique complement of membrane channels found in specific types of neurons and the integrative capabilities these channels impart. These integrative capabilities define the limits of what specific classes of neurons can and cannot do. A complete understanding of how neuronal circuits such as the respiratory CPG are maintained and modulated will only be realized when such information is available for each type of cell in the network.

In general, membrane channel studies are done using “reduced” in vitro preparations so that sophisticated biophysical recording techniques such as voltage clamp can be performed. Mechanical stability for intracellular (or patch) recordings is essential. In addition, these studies require that the particular type of channel being investigated be isolated from other channels in the membrane by utilizing ion substitutions and pharmacological agents. Thus, the reduced preparation must provide not only mechanical stability but also complete control over the extracellular fluid medium. Preparations meeting these criteria include long- and short-term cell cultures, brain slices, and perfused brain preparations. Each comes with a particular set of strengths and weaknesses. For instance, do neurons in culture exhibit the same properties as those in the fully differentiated brain? What are the detrimental effects of cutting neuronal processes during the brain slicing procedure? To what extent does tissue hypoxia compromise the activity of neurons in perfused brain (and brain slice) preparations? In many cases, these criticisms are muted by the fact that similar membrane properties can often be observed in both reduced and whole animal preparations. It is only in the reduced preparation, however, that a quantitative analysis of channel properties can be done.

One criticism which is not easily dismissed is that the loss of synaptic connectivity in reduced preparations, and hence circuit integrity, makes comparisons with intact networks such as the respiratory CPG difficult. The real danger here is that membrane properties might be studied which are not normally expressed by the intact circuit. This is particularly true for studies in cell cultures, and to a lesser extent in brain slices. As will be shown below, assigning a potential function to a given membrane channel without knowing the properties of other channels in the membrane or the synaptic inputs to the cell could easily lead to spurious conclusions. For these reasons, it is always important to interpret data obtained from reduced preparations with caution.

Another major consideration in using reduced preparations such as cell culture or brain slices is verifying that the recorded neurons are actually part of the neuronal network being investigated. For highly organized areas of the brain such as the hippocampus and cerebellum, anatomical localization within the brain slice (or microdissected area for cell culture) is sufficient to ensure that this is the case. For brainstem neurons which make up the respiratory CPG, however, such anatomical delineation is rarely possible. An alternative approach has been to identify unique classes of neurons within respiratory areas of the brain by combining electrophysiological recordings and intracellular labeling with dyes to study cell morphology. Further refinement of the localization process has then been done using tract tracing techniques. This method of identifying putative respiratory neurons has been used in brain slices to successfully study several parts of the respiratory circuit including the dorsal respiratory group (7), the ventral respiratory group (10), and the hypoglossal nucleus (15). It is likely only a matter of time before a similar approach is taken for studying respiratory neurons in cell culture.

TYPES OF INFORMATION NEEDED

Biophysical studies on membrane channels can provide several types of information. Ionic mechanisms, gating properties, inactivation properties, and modulation by neurochemicals are particularly important in describing how a neuronal network functions. Knowledge of the ionic species permeating a membrane channel may suggest functions for the channel in addition to its immediate action on the membrane potential level. For example, activation of calcium channels not only depolarizes the cell but also provides a mechanism for increasing intracellular calcium levels which can then act as a second messenger for specific enzymes and other channels. Specific ionic mechanisms may also suggest pharmacological agents which can be used to alter the channel’s activity. The mechanism of channel gating is important for knowing both when and how the channel will be activated. This is especially important in determining if the channel is activated alone or in combination with other types of membrane channels (4, 6). Likewise, the mechanism of inactivating or turning off the channel may place specific constraints on channel activity and even lead to “history”-dependent phenomena which can alter circuit function (5, 6, 8, also see below).

Membrane channel properties (excluding ionic selectivity) are also subject to modulation by endogenous neurotransmitters and neuropeptides. For example, m-channels which are selective for potassium ions are often open at or near resting membrane potentials (1) and have been observed in some respiratory neurons (3). Depolarization causes further opening of these channels. Muscarinic agonists close these channels and this results in a membrane depolarization at resting membrane potentials. In both hippocampal (16) and respiratory neurons (see Champagnat et al., this volume), the neuropeptide somatostatin has been shown to have the opposite effect: that is, it causes further activation of these channels and membrane hyperpolarization. Neurochemical modulation of membrane channels provides a mechanism for altering the integrative capabilities of neurons and by doing so, the ongoing activity of neuronal circuits.

INTERPRETING MEMBRANE PROPERTIES IN THE CONTEXT OF THE WHOLE NEURON

As stated above, a major pitfall in interpreting the role of a particular membrane channel is that the ensemble activity of the neuron must be considered. This ensemble activity arises from both the other types of channels present in the membrane and extrinsic synaptic inputs. This can be illustrated by the following example. Bulbospinal neurons in the dorsal respiratory group (DRG) of the guinea pig possess a large number of different membrane channel types (6, 8). Two of these are the A-channel and a low-voltage-activated (LVA) calcium channel. Activation of A-channels can cause transient membrane hyperpolarization mediated by potassium ions while the LVA calcium channel can both depolarize the membrane potential and contribute to increases in intracellular calcium levels. Together, these channels impart unique integrative capabilities to these neurons (6, 8). In one class of DRG neurons, called type I cells, the voltage dependence of activation of the A-channels and LVA calcium channels is similar; both are turned on by depolarization to membrane potential levels more positive than –60 mV (4). In another class of DRG neurons, called type II, this is not the case. Here the A-channels are activated at membrane potential levels more positive than –50 mV, while the LVA calcium channels are only activated by depolarizations to membrane potential levels more positive than –40 mV (4).

The voltage-dependent gating properties described above will determine when and how the A-channels and calcium channels will be expressed. Normally, DRG neurons receive an alternating pattern of synaptic excitation during inspiration and inhibition during expiration (13). In type I neurons, inspiratory phase depolarization will cause the simultaneous activation of both the outward A-channels and inward calcium channels. Thus, the opposing currents carried by these channels negate their individual effects on membrane potential. Nonetheless, the activation of the LVA calcium channels causes an increase in intracellular calcium and subsequent expression of an outward calcium-activated potassium channel. This calcium-activated potassium channel is responsible for both spike frequency adaptation (6) and a postburst hyperpolarization which may contribute to the initial stages of expiration (12). Thus, the role of the inward calcium current in this cell type appears to be limited to providing a source of intracellular calcium rather than being a mechanism for depolarizing the membrane potential.

In type II neurons, the A-channels will initially be activated alone and the outward current carried by them can compete effectively with the depolarizing drive. As a result, the membrane potential can be kept at a relatively hyperpolarized level, which prevents activation of the calcium channels and causes a long delay (up to several hundred milliseconds) between the onset of depolarization and the beginning of spike activity (5, 6). As the A-channels inactivate, their ability to control the membrane potential will gradually wane, leading to the delayed depolarization of the membrane potential and activation of the LVA calcium channels. Once activated, the inward calcium current will also contribute to the overall depolarizing drive and allow repetitive spike activity to begin. As was the case for type I neurons, the role of these two channel types in controlling the membrane excitability of type II neurons is dependent upon their individual properties as well as their interaction(s).

Finally, both the A-channels and LVA calcium channels display voltage-dependent inactivation during depolarization. Hyperpolarization of the membrane potential level for periods up to several hundred msec is required to remove this inactivation. This property of A-channels and LVA calcium channels makes their expression history dependent. That is, these channels would depend upon expiratory phase inhibition to remove their inactivation so that they can be expressed during subsequent inspiratory phase depolarization. The amount of inactivation removed can be varied by altering the size and duration of the membrane hyperpolarization. As a consequence, the numbers of A-channels and LVA calcium channels expressed during inspiratory phase depolarization can be modulated by the previous expiratory phase inhibition. An example of this history-dependent effect is shown in figure 1.1 for the LVA calcium current in type I neurons activated during a step depolarization to –15 mV. The calcium current in these cells consisted of two components: the LVA current, which completely inactivated with time, and a high-voltage-activated (HVA) current, which did not inactivate. As the pre-depolarization membrane potential level was made more negative, the amplitude of the LVA current increased while the HVA current was not affected. The complete removal of inactivation from the LVA current required pre-depolarization membrane potential levels more negative than –65 mV.

SUMMARY

Our knowledge of the role of membrane channels in determining the activity o...