- Email: [email protected]
Ion channels, receptors, agonists and antagonists
Pharmacology
Ion channels, receptors, agonists and antagonists
(or excitation), whereas hyperpolarization (inhibition) of cellular function is brought about by the opposite effect (i.e. movement of negative charges into, or positive charges out of, the cell). The ubiquitous but diverse nature of ion channels is reflected in the wide range of important biological roles they perform, including neuronal excitation, muscle contraction and intracellular signalling. In simple terms, ion channels are classified according to the stimulus required to open, or ‘gate’ the channel. Many different stimuli exist, including hydrogen ions (H+), temperature and mechanical pressure. However, this review will limit discussion to those activated by neurotransmitters (ligand-gated channels), or to changes in local membrane potential (voltage-gated channels).
Cameron J Weir
Abstract Ligand-gated ion channels are large glycoproteins responsive to activation by the binding of a chemical (agonist) to a distinct site or sites within the channel complex. In most cases, neuro transmitters released from presynaptic neurones serve as the primary agonists, acting as part of a complex integrated process required for rapid (millisecond timescale) neuronal communication. Binding of the neurotransmitter to the protein induces a conformational change, resulting in the opening of the integral ion channel. The charge and the direction of ion flux through the activated channel will then determine the effect on cellular function (i.e. depolarization, or hyperpolarization). The majority of fast inhibitory or excitatory neurotransmission within the CNS is mediated by a superfamily of genetically related ligand-gated ion channels, including excitatory (nicotinic acetylcholine (nACh) and 5 hydroxytryptamine type 3 (5HT3)) and inhibitory (γ-aminobutyric acid receptor A (GABAA) and glycine) receptors (Figure 1). This cys-loop family of receptors mediate a variety of physiological functions aimed at finely tuning the balance of excitatory and inhibitory activity within the CNS. Each member of the cys-loop family contains a conserved loop of amino acids within the extracellular domain joined by a cysteine to cysteine bridge. However, glutamate-gated ion channels (e.g. N-methyl-d-aspartate (NMDA), α-amino-3-hydroxy-5-methyl4-isoxazole proprionic acid (AMPA) and kainate receptors) are structurally distinct from the cys-loop family and have a different evolutionary background. Most classes of ligand-gated ion channels are susceptible to modulation with drugs used in anaesthesia and intensive care medicine, including general anaesthetics, anxiolytics, anti-emetics and neuromuscular blockers.
This article describes the physiology of ion channels and their modulation by commonly used drugs in anaesthesia and intensive care. It should give the clinician a better understanding of ion channel physiology and the molecular mechanisms underpinning the actions of drugs used in an acute-care setting. The concept of efficient and selective transport of ions across ‘impermeable’ plasma membranes is introduced, together with the mechanisms responsible for electrochemical signalling within cells. The classification and composition of voltage-gated ion channels are described in the context of their contribution to action potential generation in excitable cells. Drug–receptor interaction of the four main classes of receptor (ligand-gated ion channels, G-protein-coupled, enzyme-linked and nuclear receptors) and the associated signal-transduction mechanisms that initiate the cascades of intracellular events to control cellular function are also discussed. Finally, the principles of drug–receptor interaction of agonists, antagonists and inverse agonists are discussed in relation to their affinity, efficacy and potency.
Keywords G-protein-coupled receptors; inverse agonists; ligand-gated ion channels; voltage-gated ion channels
Ion channels Electrical signalling across lipid membranes is essential for communication within and between cells. However, under normal conditions cell membranes are densely packed phospholipid structures and therefore act as impenetrable barriers to charged molecules. To overcome this problem, nature has developed the ‘machinery’ to selectively transport charged particles across ‘electrically-insulated’ membranes to facilitate the important physiological roles mediated by ionic communication. In some cases such roles can be carried out using energy-dependent transporters, but this is a relatively slow process because the magnitude of the response is limited to the maximal ‘turnover’ rate of the transporter. A much more rapid and efficient system uses large, membrane-bound glycoproteins, containing waterfilled pores (ion channels) and established electrochemical gradients between the extra- and intra-cellular compartments. The activation, or gating, of an ion channel results in the rapid, but selective, movement of ions across the plasma membrane. Movement of positively charged ions (cations) into, or negatively charged ions (anions) out of, the cell produces depolarization
Receptor topology: cys-loop ion channels are composed of five membrane-spanning subunits arranged around a central aqueous pore (Figure 2). Each class of receptor has its own pool of subunits. For example GABAA is composed of five subunits taken from a pool of 18 possible subtypes, including α1−6, β1−3, γ1−3, δ, ɛ, π and ρ1−3. Subunits are constructed from approximately 450 amino acids containing a large extracellular domain and four transmembrane domains. The large extracellular domain contains the agonist (e.g. acetylcholine, GABA, glycine) binding site for each channel. The second transmembrane (TM2) domain from each subunit faces into the centre of the protein to line the surface of the aqueous pore or channel. The physicochemical properties (e.g. size and charge) of the amino acids within the TM2 region together with their location within the channel provide the ideal conditions for selectively filtering cations or anions during channel opening. Glutamate-gated channels are
Cameron J Weir, PhD, FRCA, is Consultant Anaesthetist and Senior Lecturer at Ninewells Hospital and Medical School, Dundee.
ANAESTHESIA AND INTENSIVE CARE MEDICINE 8:10
437
© 2007 Elsevier Ltd. All rights reserved.
Pharmacology
Ligand-gated ion channels GABA A
Glycine
Nicotinic acetylcholine
Glutamate (NMDA/AMPA)
5-HT3
Cl –
Cl –
Na + (Ca 2+)
Na + (Ca 2+)
Na + (Ca 2+)
Extracellular Cl – 125 mM Na + 145 mM Ca 2+ 2 mM K + 4 mM
Intracellular Cl – 5 mM Na + 12 mM Ca 2+ 0.1 mM K + 140 mM
K+
K+
Hyperpolarization (~inhibition)
K+
Depolarization (~excitation)
Ligand-gated ion channels (LGICs) mediate the majority of fast neurotransmission within the CNS. Fine tuning of CNS activity is achieved by modulating the activity of inhibitory and excitatory LGICs. Many clinically useful neuroactive drugs alter the activity of LGICs 5-HT, 5 hydroxytryptamine; NMDA, N-methyl-D-asparate; AMPA, α-amino-3-hydroxy-5-methyl-4-isoazole proprionic acid
Figure 1
composed of four subunits each with three transmembrane domains and a re-entrant loop that contributes to the lining of the channel pore.
this respect because it is sensitive to the modulatory effects of a range of chemically diverse agents, including benzodiazepines, barbiturates and general anaesthetics. The mechanisms underpinning the pharmacology of benzodiazepines have received much attention in recent years. It now seems that some of their clinical effects can be mapped not only to individual subunits, but also to single amino acids within them. Genetically engineered mice
Ligand-gated ion channel modulation: ligand-gated ion channels are susceptible to modulation by many clinically useful classes of drugs. The inhibitory GABAA is rather promiscuous in
NH2
Ligand-gated ion channel topology
Agonist S
COOH
1
2
3
4
Most ligand-gated ion channels are oligomeric proteins containing five subunits. Individual subunits are formed from a sequence of amino acids that fold into a large extracellular domain and four transmembrane (TM) spanning regions. The TM2 region faces into the ion channel to line its surface
Figure 2
ANAESTHESIA AND INTENSIVE CARE MEDICINE 8:10
438
© 2007 Elsevier Ltd. All rights reserved.
Pharmacology
Ca2+ channels – Ca2+ ions are ubiquitous, but essential mediators of many important cellular processes, including neurotransmitter release and intracellular signalling. Ca2+ concentrations are regulated by a variety of pathways, many of which are closely integrated to ensure precise control of free intracellular calcium concentrations. G-protein-coupled receptors (via cyclic AMP and inositol triphosphate (IP3)), ligand-gated channels (via NMDA receptors) and energy-dependent Ca2+ transporters (Na+/Ca2+ exchange and Ca2+ ATPase) also contribute towards the regulation of Ca2+ flux across plasma membranes. Voltagegated Ca2+ channels are present in most excitable cells. They are composed of five major subtypes (L, N, T, P/Q and R) based on their electrophysiological characteristics and on their sensitivity and selectivity to antagonists. L-type channels, incorporating the pore-forming Cav1.1–Cav1.4 subunits, seem to be the most clinically relevant because they mediate Ca2+ related events in smooth and cardiac muscle and their subtypes are selectively blocked, albeit to varying degrees, by dihydropyridine antag onists (e.g. nifedipine, verapamil and diltiazem). T-type channels (Cav3.1–Cav3.3) contribute to pacemaker activity in atrial and vascular smooth muscle contraction and can be selectively blocked by mibefradil (withdrawn due to interaction with other drugs that prolong the QT interval). To date, no clinically useful antagonists have been introduced for the remaining three subtypes (N, Cav2.2; P/Q, Cav2.2 and R, Cav2.3).
harbouring single amino acid mutations within GABAA receptor subunits are selectively resistant to some of the clinical properties of diazepam. For example, the anxiolytic effects of diazepam appear to be mediated by the α2-subunit, whereas the sedative effects are mediated by the α1-subunit. Similar research using intravenous general anaesthetics in mice demonstrates that the sedative and hypnotic effects of etomidate are mediated by the β2- and β3-subunits, respectively. Moreover, it seems that the cardiorespiratory depressant effects of some of the intravenous anaesthetics are also mediated by β2 and β3 subunit-containing GABAA receptors. Voltage-gated ion channels, as their name suggests, are activ ated by changes to the local membrane potential. Three clinically important channels selectively permeable to sodium (Na+), potassium (K+) and calcium ions (Ca2+) are described below. Na+ channels – under normal conditions the resting membrane potential of most of the excitable cells is maintained at approximately –70 mV by a constant outward leak of K+. Small positive shifts of the membrane potential, beyond a threshold of approximately –55 mV, result in the simultaneous opening of Na+ channels. This channel opening results in a rapid inward movement of Na+, resulting in membrane depolarization to approximately +40 mV. Nine different subtypes of Na+ channel have been identified on the basis of the nature of the pore-forming α-subunit (Nav1–Nav9), their sensitivity to the Na+ channel blocker tetrodotoxin and their different rates of channel inactivation. Most clinically relevant voltage-activated Na+ channel subtypes are located in neurones and cardiac and skeletal muscle. Na+ channels are susceptible to blockade by local anaesthetics, some anticonvulsants and some antidysrhythmic drugs. Local anaesthetics, in particular lidocaine, preferentially bind to and stabilize the inactivated channel conformation and so prolong the duration of the refractory period between action potentials. Unfortunately, the lack of subtype specificity of local anaesthetic drugs results in a low therapeutic index and a number of unwanted side effects, including negative inotropic activity and seizures. K+ channels are important regulators of cellular excitability because of their influence on the frequency and duration of the action potential. For example, certain voltage-gated K+ channel subtypes are activated in the early phase of the action potential and produce a slow, but sustained outward K+ current. This slow outward current counteracts the rapid inward Na+ current and brings the membrane potential back towards the resting membrane potential. Alteration of K+ channel function will, therefore, have a profound effect on action potential kinetics. For example, inherited K+ channel mutations contribute to a range of neurological and cardiac diseases, including the life-threatening long QT syndrome. K+ channel nomenclature is broadly based on the number of transmembrane domains and pores present within the protein. Channels containing six transmembrane domains and a single pore (6T1P) are particularly sensitive to changes in voltage and intracellular calcium concentrations, whereas channels composed of two transmembrane domains and a single pore (2T1P; inward rectifying) are regulated mainly by G-proteins and intracellular ATP levels. Two pore domain K+ channels (4T2P; outward rectifying) contribute to a resting K+ current and some subtypes (i.e. TREK and TASK) are particularly sensitive to potentiation by volatile anaesthetic agents.
ANAESTHESIA AND INTENSIVE CARE MEDICINE 8:10
Voltage-gated ion channel topology (Figure 3): the α-subunit within voltage-gated Na+ and Ca2+ channels is composed of four identical regions (I–IV), each containing six transmembrane domains (S1–S6). The voltage-sensing region contains mainly positively charged amino acids and is located within the S4 segment. Quaternary folding of the protein produces a ring-shaped molecule, with the amino acids between the S5–S6 segments contributing to the internal channel pore. In the closed state a mobile intracellular loop joining regions III–IV serves as an electromechanical block to close, or inactivate, the channel. In contrast, K+ channels are primarily composed of single subunits containing a variable number of transmembrane and pore-forming regions. Some K+ channels probably exist as dimers (4TM family) or tetramers (2TM and 6TM families). To date, more than 70 different K+ channel subtypes have been identified. Receptors Paul Ehrlich (1854–1915) first proposed the concept of highly specific interactions between drugs and receptors: corpora non agunt nisi fixata (drugs do not act unless they are bound). However, before we discuss drug interaction, it is important to clarify the definition of a receptor. To a pharmacologist, a receptor is a protein that recognizes an endogenous chemical mediator such as a neurotransmitter, hormone or inflammatory molecule. When the mediator (agonist) binds to the receptor, a series of reactions takes place, ultimately leading to a change of function of the host cell. For example, the binding of GABA to GABAA inhibits neur onal function by inducing an inward flow of chloride ions (Cl−) through an integral ion channel. Often the term ‘receptor’ is used loosely to describe a target whose function is altered by an exogenous drug rather than an endogenous mediator. For example, voltage-gated Na+ channels have been described as local anaesthetic receptors, but the local anaesthetic drug does 439
© 2007 Elsevier Ltd. All rights reserved.
Pharmacology
Voltage-gated sodium and calcium ion channel topology ( subunits) I + + 1 2 3 4 5 + +
a
6
II
III
IV
+ + 1 2 3 4 5 + +
+ + 1 2 3 4 5 + +
+ + 1 2 3 4 5 + +
6
6
6
NH2 COOH Na +
II I b
Closed/active
II III
I
IV
III IV
Open
Courtesy of Professor JA Peters, University of Dundee
Figure 3 (a) Two-dimensional view; (b) three-dimensional view.
not mediate a natural physiological process. In this case, the Na+ channel may be more appropriately labelled a ‘drug target’.
receptor induces a series of conformational changes, resulting in an increased affinity of the GPCR for the G-protein. The G-protein binds to the intracellular loop of the receptor and simultaneously exchanges GDP for GTP, which triggers the dissociation of the G-protein into α- and βγ-subunits. The newly liberated α- and βγ-subunits are then free to activate the relevant effector enzymes or ion channels. The reaction is terminated after GTP is hydrolysed to GDP by the inherent GTPase activity of the α-subunit, and the α- and βγ-subunits re-assemble into their parent G-proteins. The beauty of GPCR signalling lies with the range of different Gprotein isoforms available for interaction with receptors/effector systems and the subsequent multi-stepped cascade of enzyme mediated events. Such a system allows for not only specificity, but also amplification of the receptor transduction signal. Second messengers • The principal targets for GPCRs include enzymes responsible for the generation of second messengers: adenylyl cyclase – cyclic AMP (cAMP); phospholipase C – inositol triphosphate (IP3) and diacylglycerol (DAG). In addition, voltage-gated ion channels (including K+, Ca2+ and Na+ channels) are directly modulated by the βγ-subunit of certain G-proteins. Adenylyl cyclase is a ubiquitous membrane bound enzyme that catalyses the conversion of ATP into cAMP. Most of the biological effects of cAMP are mediated by the activation of protein kinase A and subsequent phosphorylation events. However, direct modulation of hyperpolarization-activated, cyclic-nucleotide gated (HCN) ion channels is also well documented. Cyclic AMP levels are controlled by two Gprotein isoforms (Gs and Gi), which stimulate and inhibit adenylyl cyclase activity, respectively. In addition, cAMP is inactivated to 5′AMP by phosphodiesterase enzymes susceptible to inhibition by theophyllines. Cholera toxin produces hypersecretion within the
Receptor classification: in simple terms receptors can be classified into four categories (Figure 4). • Ligand-gated ion channels • G-protein-coupled receptors • Enzyme-linked receptors • Nuclear receptors G-protein-coupled receptors (GPCRs) form the largest single category of receptors that mediate physiological functions, ranging from olfaction to regulation of the autonomic nervous system. More than 1000 subtypes of GPCRs have been identified, including muscarinic acetylcholine, adreno- and chemokine receptors. As a consequence of their ubiquitous distribution and functional characteristics, GPCRs act as principal targets for approximately 60% of all prescribed drugs. Activation of GPCRs initiates a cascade of intracellular signalling mechanisms, a process which relies on intermediary G-proteins designed to link the signal between agonist binding and activation of target enzymes or channels (effectors). GPCRs are composed of approximately 400 amino acid residues folded into seven transmembrane domains (Figure 4). The agonist-binding sites are located within either the extracellular N-terminal domain (e.g. peptide receptors) or the transmembrane regions (e.g. receptors recognizing small amines). An intracellular loop of variable length between TM5 and TM6 provides an anchoring point for diffusible G-proteins. G-proteins are membrane-associated trimeric complexes composed of one α-subunit and a βγ-dimer. Before receptor activation, a single GDP molecule is coupled to the α-subunit. Agonist activation of the ANAESTHESIA AND INTENSIVE CARE MEDICINE 8:10
440
© 2007 Elsevier Ltd. All rights reserved.
Pharmacology
Receptor classification Ligand-gated channels
G-protein-coupled receptors
Enzyme-linked receptors
Nuclear receptors
N
Ions N βγ Enzyme
βγ Gαi
Gαs
GTP
C GTP
Second messenger
Depolarization/ hyperpolarization
Enzyme
Second messenger
Change in [Ca 2+] Protein kinase activity
C Phosphorylation
Gene transcription/protein synthesis
Timescale Fast (msecs) βγ α
G-protein (Gαi –inhibitory, Gαs – stimulatory)
Slow (hours) Drug
Figure 4
gastrointestinal tract by stimulating Gs, whereas pertussis toxin uncouples Gi and Go isoforms from their GPCRs. • Phospholipase C/IP3: the Gq isoform of G-proteins facilitates the production of two important second messengers from membrane phospholipids: IP3 and DAG. Gq activates phospholipase C, the enzyme responsible for the catalytic conversion of phosphat idylinositol 4,5-bisphosphate into IP3 and DAG. IP3 is a watersoluble agonist for a ligand-gated channel (IP3R), the principal mediator of GPCR-induced Ca2+ release from intracellular stores. In contrast, DAG is a membrane-bound second messenger, which activates one of ten different isoforms of protein kinase C. Inactivation of the inositol phosphate system is brought about by metabolism of IP3 and DAG by a combination of phosphorylation (kinases) and dephosphorylation (phosphatases).
enzyme-linked receptors containing intrinsic guanylyl cyclase activity to modulate the activity of the cardiovascular system. Nuclear receptors form a small number of important targets regulating the activity of many diverse agents, including steroids and fat-soluble vitamins. In contrast to the receptors described previously, the so-called nuclear receptors are soluble proteins found either in the cytoplasm (Class I) or within the cell nucleus (Class II) (Figure 4). Agonist activation of Class I targets (e.g. by steroids or prolactin) results in the formation of dimeric receptor complexes, which translocate into the nucleus and bind to ‘hormone-response elements’ within DNA to regulate gene transcription. Class II receptors reside within the cell nucleus and regulate lipid and drug metabolizing enzymes. Both classes of receptor contain highly conserved structural motifs (zinc fingers) to assist with recognition and binding of the receptor to hormone response elements of DNA.
Enzyme-linked receptors are composed of approximately 1000 amino acid residues arranged into a large extracellular domain, a single transmembrane region and a variable-length intracellular domain (Figure 4). Enzyme-linked receptors mediate the actions of various proteins involved in tissue repair, defence and growth (e.g. insulin, inflammatory cytokines and growth factors). One of the defining features of this class of receptor is the presence of specific enzymes, usually kinases, within the intracellular domain. On ligand binding, many receptors undergo dimerization (coup ling of two receptors into one large complex) which, in many cases, facilitates the ‘autophosphorylation’ of tyrosine, serine or threonine residues on each receptor. The phosphorylated residues can then attract specific SH2 (Src homology) proteins and enzymes to alter cellular function through changes in gene transcription. An alternative mechanism adopted by atrial natriuretic factor uses
ANAESTHESIA AND INTENSIVE CARE MEDICINE 8:10
Agonists An agonist is a chemical that binds to and activates a receptor to alter the function of its host cell. Agonists possess both affinity and efficacy. In simple terms, the ability of an agonist (or drug) to bind to the receptor is termed affinity and is defined as the ratio of the binding rate (k+1) and dissociation rate (k−1), that is, affinity (kA )=k − 1/k+1 , and is often measured experimentally using radioactive binding techniques. In contrast, efficacy describes the ability of the drug to activate the receptor once it has bound. Full agonists are capable of producing a maximal response or maximal efficacy (this may occur when only a fraction of receptors are occupied, hence the concept of ‘spare’ receptors), whereas 441
© 2007 Elsevier Ltd. All rights reserved.
Pharmacology
artial agonists are not capable of producing a full response even p in the presence of high concentrations of agonist (Figure 5):
Competitive reversible and irreversible receptor antagonism 100
where: A is the agonist (or drug A), R is the receptor, AR is the bound receptor, AR* is the activated receptor, and β and α are the receptor activation and deactivation constants, respectively. In the clinical setting measurement of affinity and efficacy is problematic because a drug might produce its effects by altering a combination of physiological parameters. For example, epinephrine increases blood pressure by increasing the rate and force of contraction of the heart (β-adrenoreceptors) but also causes profound vasoconstriction (α-adrenoreceptors). Therefore, the term potency defines the concentration of a drug required to produce a measurable clinical response, usually 50% of the maximal possible response (effective concentration; EC50). Highly potent drugs will have smaller EC50s compared with less potent drugs (Figure 5).
Response (% maximum)
Increasing Efficacy
Response (% maximum)
Decreasing 0.1
1
10
100
1000
Log agonist concentration Agonists: a and b are full agonists; a is a more potent agonist than b, but has equal efficacy; c has a lower efficacy (partial agonist) than either a or b; a and c are equipotent; c is more potent than b
Figure 5
ANAESTHESIA AND INTENSIVE CARE MEDICINE 8:10
100
1000
Inverse agonists Inverse agonists, as their name suggests, produce the opposite effects of agonists. This newly recognized phenomenon is possible only if the receptor is active in the absence of agonist (i.e. it has constitutive activity). In this situation a competitive antagonist on its own will have no effect on constitutive activity (because no agonist is present), but an inverse agonist will produce a concentration-dependent reduction in receptor activity (i.e. it has negative efficacy). Examples of receptors displaying constitutive activity include certain subtypes of GPCRs, GABAergic and cannabinoid receptors. ◆
0 0.01
10
effects can be overcome by increasing the concentration of competing agonist (e.g. atracurium-induced neuromuscular blockade can be overcome by increasing acetylcholine concentrations with neostigmine, although clinically this is probably achieved by activating spare receptors). Graphically, this is represented by a parallel and rightward shift of the concentration–response curve (Figure 6). In contrast, irreversible competitive antagonists covalently bind to the receptor, and blockade cannot be overcome by increas ing the agonist concentration (i.e. the effect is insurmountable). Consequently, the maximal effect of the agonist is reduced in the presence of irreversible antagonist (Figure 6). Clinically useful examples of irreversible competitive antagonists include aspirin and omeprazole acting on an enzyme and a proton pump, respectively. The term non-competitive antagonism is reserved for drugs that do not compete at the agonist binding site, but prevent signal transduction within the receptor complex. The effects of non-competitive antagonists are also insurmountable.
c 20
1
Figure 6
60
40
0.1
The binding of a reversible competitive antagonist can be overcome by increasing agonist concentration (agonist curve shifts to the right but efficacy remains unchanged) whereas an irreversible antagonist decreases the efficacy of the agonist
Decreasing
b
Agonist + reversible competitive antagonist
Log agonist concentration
100
a
40
0.01
Agonist concentration–response curves
80
60
0
where: B is the antagonist, R is the receptor, BR is the bound receptor. Reversible competitive antagonism is surmountable. In other words, when an antagonist binds to the agonist-binding site its
Potency
Agonist + irreversible competitive antagonist
20
Antagonists Competitive receptor antagonists are classed as either reversible or irreversible. Competitive antagonists bind to the receptor and therefore possess affinity, but they are unable to produce a response and thus lack efficacy. Furthermore, they prevent the agonist from binding and so block its ability to activate the receptor:
Increasing
Agonist alone
80
442
© 2007 Elsevier Ltd. All rights reserved.
Ion channels, receptors, agonists and antagonists
(or excitation), whereas hyperpolarization (inhibition) of cellular function is brought about by the opposite effect (i.e. movement of negative charges into, or positive charges out of, the cell). The ubiquitous but diverse nature of ion channels is reflected in the wide range of important biological roles they perform, including neuronal excitation, muscle contraction and intracellular signalling. In simple terms, ion channels are classified according to the stimulus required to open, or ‘gate’ the channel. Many different stimuli exist, including hydrogen ions (H+), temperature and mechanical pressure. However, this review will limit discussion to those activated by neurotransmitters (ligand-gated channels), or to changes in local membrane potential (voltage-gated channels).
Cameron J Weir
Abstract Ligand-gated ion channels are large glycoproteins responsive to activation by the binding of a chemical (agonist) to a distinct site or sites within the channel complex. In most cases, neuro transmitters released from presynaptic neurones serve as the primary agonists, acting as part of a complex integrated process required for rapid (millisecond timescale) neuronal communication. Binding of the neurotransmitter to the protein induces a conformational change, resulting in the opening of the integral ion channel. The charge and the direction of ion flux through the activated channel will then determine the effect on cellular function (i.e. depolarization, or hyperpolarization). The majority of fast inhibitory or excitatory neurotransmission within the CNS is mediated by a superfamily of genetically related ligand-gated ion channels, including excitatory (nicotinic acetylcholine (nACh) and 5 hydroxytryptamine type 3 (5HT3)) and inhibitory (γ-aminobutyric acid receptor A (GABAA) and glycine) receptors (Figure 1). This cys-loop family of receptors mediate a variety of physiological functions aimed at finely tuning the balance of excitatory and inhibitory activity within the CNS. Each member of the cys-loop family contains a conserved loop of amino acids within the extracellular domain joined by a cysteine to cysteine bridge. However, glutamate-gated ion channels (e.g. N-methyl-d-aspartate (NMDA), α-amino-3-hydroxy-5-methyl4-isoxazole proprionic acid (AMPA) and kainate receptors) are structurally distinct from the cys-loop family and have a different evolutionary background. Most classes of ligand-gated ion channels are susceptible to modulation with drugs used in anaesthesia and intensive care medicine, including general anaesthetics, anxiolytics, anti-emetics and neuromuscular blockers.
This article describes the physiology of ion channels and their modulation by commonly used drugs in anaesthesia and intensive care. It should give the clinician a better understanding of ion channel physiology and the molecular mechanisms underpinning the actions of drugs used in an acute-care setting. The concept of efficient and selective transport of ions across ‘impermeable’ plasma membranes is introduced, together with the mechanisms responsible for electrochemical signalling within cells. The classification and composition of voltage-gated ion channels are described in the context of their contribution to action potential generation in excitable cells. Drug–receptor interaction of the four main classes of receptor (ligand-gated ion channels, G-protein-coupled, enzyme-linked and nuclear receptors) and the associated signal-transduction mechanisms that initiate the cascades of intracellular events to control cellular function are also discussed. Finally, the principles of drug–receptor interaction of agonists, antagonists and inverse agonists are discussed in relation to their affinity, efficacy and potency.
Keywords G-protein-coupled receptors; inverse agonists; ligand-gated ion channels; voltage-gated ion channels
Ion channels Electrical signalling across lipid membranes is essential for communication within and between cells. However, under normal conditions cell membranes are densely packed phospholipid structures and therefore act as impenetrable barriers to charged molecules. To overcome this problem, nature has developed the ‘machinery’ to selectively transport charged particles across ‘electrically-insulated’ membranes to facilitate the important physiological roles mediated by ionic communication. In some cases such roles can be carried out using energy-dependent transporters, but this is a relatively slow process because the magnitude of the response is limited to the maximal ‘turnover’ rate of the transporter. A much more rapid and efficient system uses large, membrane-bound glycoproteins, containing waterfilled pores (ion channels) and established electrochemical gradients between the extra- and intra-cellular compartments. The activation, or gating, of an ion channel results in the rapid, but selective, movement of ions across the plasma membrane. Movement of positively charged ions (cations) into, or negatively charged ions (anions) out of, the cell produces depolarization
Receptor topology: cys-loop ion channels are composed of five membrane-spanning subunits arranged around a central aqueous pore (Figure 2). Each class of receptor has its own pool of subunits. For example GABAA is composed of five subunits taken from a pool of 18 possible subtypes, including α1−6, β1−3, γ1−3, δ, ɛ, π and ρ1−3. Subunits are constructed from approximately 450 amino acids containing a large extracellular domain and four transmembrane domains. The large extracellular domain contains the agonist (e.g. acetylcholine, GABA, glycine) binding site for each channel. The second transmembrane (TM2) domain from each subunit faces into the centre of the protein to line the surface of the aqueous pore or channel. The physicochemical properties (e.g. size and charge) of the amino acids within the TM2 region together with their location within the channel provide the ideal conditions for selectively filtering cations or anions during channel opening. Glutamate-gated channels are
Cameron J Weir, PhD, FRCA, is Consultant Anaesthetist and Senior Lecturer at Ninewells Hospital and Medical School, Dundee.
ANAESTHESIA AND INTENSIVE CARE MEDICINE 8:10
437
© 2007 Elsevier Ltd. All rights reserved.
Pharmacology
Ligand-gated ion channels GABA A
Glycine
Nicotinic acetylcholine
Glutamate (NMDA/AMPA)
5-HT3
Cl –
Cl –
Na + (Ca 2+)
Na + (Ca 2+)
Na + (Ca 2+)
Extracellular Cl – 125 mM Na + 145 mM Ca 2+ 2 mM K + 4 mM
Intracellular Cl – 5 mM Na + 12 mM Ca 2+ 0.1 mM K + 140 mM
K+
K+
Hyperpolarization (~inhibition)
K+
Depolarization (~excitation)
Ligand-gated ion channels (LGICs) mediate the majority of fast neurotransmission within the CNS. Fine tuning of CNS activity is achieved by modulating the activity of inhibitory and excitatory LGICs. Many clinically useful neuroactive drugs alter the activity of LGICs 5-HT, 5 hydroxytryptamine; NMDA, N-methyl-D-asparate; AMPA, α-amino-3-hydroxy-5-methyl-4-isoazole proprionic acid
Figure 1
composed of four subunits each with three transmembrane domains and a re-entrant loop that contributes to the lining of the channel pore.
this respect because it is sensitive to the modulatory effects of a range of chemically diverse agents, including benzodiazepines, barbiturates and general anaesthetics. The mechanisms underpinning the pharmacology of benzodiazepines have received much attention in recent years. It now seems that some of their clinical effects can be mapped not only to individual subunits, but also to single amino acids within them. Genetically engineered mice
Ligand-gated ion channel modulation: ligand-gated ion channels are susceptible to modulation by many clinically useful classes of drugs. The inhibitory GABAA is rather promiscuous in
NH2
Ligand-gated ion channel topology
Agonist S
COOH
1
2
3
4
Most ligand-gated ion channels are oligomeric proteins containing five subunits. Individual subunits are formed from a sequence of amino acids that fold into a large extracellular domain and four transmembrane (TM) spanning regions. The TM2 region faces into the ion channel to line its surface
Figure 2
ANAESTHESIA AND INTENSIVE CARE MEDICINE 8:10
438
© 2007 Elsevier Ltd. All rights reserved.
Pharmacology
Ca2+ channels – Ca2+ ions are ubiquitous, but essential mediators of many important cellular processes, including neurotransmitter release and intracellular signalling. Ca2+ concentrations are regulated by a variety of pathways, many of which are closely integrated to ensure precise control of free intracellular calcium concentrations. G-protein-coupled receptors (via cyclic AMP and inositol triphosphate (IP3)), ligand-gated channels (via NMDA receptors) and energy-dependent Ca2+ transporters (Na+/Ca2+ exchange and Ca2+ ATPase) also contribute towards the regulation of Ca2+ flux across plasma membranes. Voltagegated Ca2+ channels are present in most excitable cells. They are composed of five major subtypes (L, N, T, P/Q and R) based on their electrophysiological characteristics and on their sensitivity and selectivity to antagonists. L-type channels, incorporating the pore-forming Cav1.1–Cav1.4 subunits, seem to be the most clinically relevant because they mediate Ca2+ related events in smooth and cardiac muscle and their subtypes are selectively blocked, albeit to varying degrees, by dihydropyridine antag onists (e.g. nifedipine, verapamil and diltiazem). T-type channels (Cav3.1–Cav3.3) contribute to pacemaker activity in atrial and vascular smooth muscle contraction and can be selectively blocked by mibefradil (withdrawn due to interaction with other drugs that prolong the QT interval). To date, no clinically useful antagonists have been introduced for the remaining three subtypes (N, Cav2.2; P/Q, Cav2.2 and R, Cav2.3).
harbouring single amino acid mutations within GABAA receptor subunits are selectively resistant to some of the clinical properties of diazepam. For example, the anxiolytic effects of diazepam appear to be mediated by the α2-subunit, whereas the sedative effects are mediated by the α1-subunit. Similar research using intravenous general anaesthetics in mice demonstrates that the sedative and hypnotic effects of etomidate are mediated by the β2- and β3-subunits, respectively. Moreover, it seems that the cardiorespiratory depressant effects of some of the intravenous anaesthetics are also mediated by β2 and β3 subunit-containing GABAA receptors. Voltage-gated ion channels, as their name suggests, are activ ated by changes to the local membrane potential. Three clinically important channels selectively permeable to sodium (Na+), potassium (K+) and calcium ions (Ca2+) are described below. Na+ channels – under normal conditions the resting membrane potential of most of the excitable cells is maintained at approximately –70 mV by a constant outward leak of K+. Small positive shifts of the membrane potential, beyond a threshold of approximately –55 mV, result in the simultaneous opening of Na+ channels. This channel opening results in a rapid inward movement of Na+, resulting in membrane depolarization to approximately +40 mV. Nine different subtypes of Na+ channel have been identified on the basis of the nature of the pore-forming α-subunit (Nav1–Nav9), their sensitivity to the Na+ channel blocker tetrodotoxin and their different rates of channel inactivation. Most clinically relevant voltage-activated Na+ channel subtypes are located in neurones and cardiac and skeletal muscle. Na+ channels are susceptible to blockade by local anaesthetics, some anticonvulsants and some antidysrhythmic drugs. Local anaesthetics, in particular lidocaine, preferentially bind to and stabilize the inactivated channel conformation and so prolong the duration of the refractory period between action potentials. Unfortunately, the lack of subtype specificity of local anaesthetic drugs results in a low therapeutic index and a number of unwanted side effects, including negative inotropic activity and seizures. K+ channels are important regulators of cellular excitability because of their influence on the frequency and duration of the action potential. For example, certain voltage-gated K+ channel subtypes are activated in the early phase of the action potential and produce a slow, but sustained outward K+ current. This slow outward current counteracts the rapid inward Na+ current and brings the membrane potential back towards the resting membrane potential. Alteration of K+ channel function will, therefore, have a profound effect on action potential kinetics. For example, inherited K+ channel mutations contribute to a range of neurological and cardiac diseases, including the life-threatening long QT syndrome. K+ channel nomenclature is broadly based on the number of transmembrane domains and pores present within the protein. Channels containing six transmembrane domains and a single pore (6T1P) are particularly sensitive to changes in voltage and intracellular calcium concentrations, whereas channels composed of two transmembrane domains and a single pore (2T1P; inward rectifying) are regulated mainly by G-proteins and intracellular ATP levels. Two pore domain K+ channels (4T2P; outward rectifying) contribute to a resting K+ current and some subtypes (i.e. TREK and TASK) are particularly sensitive to potentiation by volatile anaesthetic agents.
ANAESTHESIA AND INTENSIVE CARE MEDICINE 8:10
Voltage-gated ion channel topology (Figure 3): the α-subunit within voltage-gated Na+ and Ca2+ channels is composed of four identical regions (I–IV), each containing six transmembrane domains (S1–S6). The voltage-sensing region contains mainly positively charged amino acids and is located within the S4 segment. Quaternary folding of the protein produces a ring-shaped molecule, with the amino acids between the S5–S6 segments contributing to the internal channel pore. In the closed state a mobile intracellular loop joining regions III–IV serves as an electromechanical block to close, or inactivate, the channel. In contrast, K+ channels are primarily composed of single subunits containing a variable number of transmembrane and pore-forming regions. Some K+ channels probably exist as dimers (4TM family) or tetramers (2TM and 6TM families). To date, more than 70 different K+ channel subtypes have been identified. Receptors Paul Ehrlich (1854–1915) first proposed the concept of highly specific interactions between drugs and receptors: corpora non agunt nisi fixata (drugs do not act unless they are bound). However, before we discuss drug interaction, it is important to clarify the definition of a receptor. To a pharmacologist, a receptor is a protein that recognizes an endogenous chemical mediator such as a neurotransmitter, hormone or inflammatory molecule. When the mediator (agonist) binds to the receptor, a series of reactions takes place, ultimately leading to a change of function of the host cell. For example, the binding of GABA to GABAA inhibits neur onal function by inducing an inward flow of chloride ions (Cl−) through an integral ion channel. Often the term ‘receptor’ is used loosely to describe a target whose function is altered by an exogenous drug rather than an endogenous mediator. For example, voltage-gated Na+ channels have been described as local anaesthetic receptors, but the local anaesthetic drug does 439
© 2007 Elsevier Ltd. All rights reserved.
Pharmacology
Voltage-gated sodium and calcium ion channel topology ( subunits) I + + 1 2 3 4 5 + +
a
6
II
III
IV
+ + 1 2 3 4 5 + +
+ + 1 2 3 4 5 + +
+ + 1 2 3 4 5 + +
6
6
6
NH2 COOH Na +
II I b
Closed/active
II III
I
IV
III IV
Open
Courtesy of Professor JA Peters, University of Dundee
Figure 3 (a) Two-dimensional view; (b) three-dimensional view.
not mediate a natural physiological process. In this case, the Na+ channel may be more appropriately labelled a ‘drug target’.
receptor induces a series of conformational changes, resulting in an increased affinity of the GPCR for the G-protein. The G-protein binds to the intracellular loop of the receptor and simultaneously exchanges GDP for GTP, which triggers the dissociation of the G-protein into α- and βγ-subunits. The newly liberated α- and βγ-subunits are then free to activate the relevant effector enzymes or ion channels. The reaction is terminated after GTP is hydrolysed to GDP by the inherent GTPase activity of the α-subunit, and the α- and βγ-subunits re-assemble into their parent G-proteins. The beauty of GPCR signalling lies with the range of different Gprotein isoforms available for interaction with receptors/effector systems and the subsequent multi-stepped cascade of enzyme mediated events. Such a system allows for not only specificity, but also amplification of the receptor transduction signal. Second messengers • The principal targets for GPCRs include enzymes responsible for the generation of second messengers: adenylyl cyclase – cyclic AMP (cAMP); phospholipase C – inositol triphosphate (IP3) and diacylglycerol (DAG). In addition, voltage-gated ion channels (including K+, Ca2+ and Na+ channels) are directly modulated by the βγ-subunit of certain G-proteins. Adenylyl cyclase is a ubiquitous membrane bound enzyme that catalyses the conversion of ATP into cAMP. Most of the biological effects of cAMP are mediated by the activation of protein kinase A and subsequent phosphorylation events. However, direct modulation of hyperpolarization-activated, cyclic-nucleotide gated (HCN) ion channels is also well documented. Cyclic AMP levels are controlled by two Gprotein isoforms (Gs and Gi), which stimulate and inhibit adenylyl cyclase activity, respectively. In addition, cAMP is inactivated to 5′AMP by phosphodiesterase enzymes susceptible to inhibition by theophyllines. Cholera toxin produces hypersecretion within the
Receptor classification: in simple terms receptors can be classified into four categories (Figure 4). • Ligand-gated ion channels • G-protein-coupled receptors • Enzyme-linked receptors • Nuclear receptors G-protein-coupled receptors (GPCRs) form the largest single category of receptors that mediate physiological functions, ranging from olfaction to regulation of the autonomic nervous system. More than 1000 subtypes of GPCRs have been identified, including muscarinic acetylcholine, adreno- and chemokine receptors. As a consequence of their ubiquitous distribution and functional characteristics, GPCRs act as principal targets for approximately 60% of all prescribed drugs. Activation of GPCRs initiates a cascade of intracellular signalling mechanisms, a process which relies on intermediary G-proteins designed to link the signal between agonist binding and activation of target enzymes or channels (effectors). GPCRs are composed of approximately 400 amino acid residues folded into seven transmembrane domains (Figure 4). The agonist-binding sites are located within either the extracellular N-terminal domain (e.g. peptide receptors) or the transmembrane regions (e.g. receptors recognizing small amines). An intracellular loop of variable length between TM5 and TM6 provides an anchoring point for diffusible G-proteins. G-proteins are membrane-associated trimeric complexes composed of one α-subunit and a βγ-dimer. Before receptor activation, a single GDP molecule is coupled to the α-subunit. Agonist activation of the ANAESTHESIA AND INTENSIVE CARE MEDICINE 8:10
440
© 2007 Elsevier Ltd. All rights reserved.
Pharmacology
Receptor classification Ligand-gated channels
G-protein-coupled receptors
Enzyme-linked receptors
Nuclear receptors
N
Ions N βγ Enzyme
βγ Gαi
Gαs
GTP
C GTP
Second messenger
Depolarization/ hyperpolarization
Enzyme
Second messenger
Change in [Ca 2+] Protein kinase activity
C Phosphorylation
Gene transcription/protein synthesis
Timescale Fast (msecs) βγ α
G-protein (Gαi –inhibitory, Gαs – stimulatory)
Slow (hours) Drug
Figure 4
gastrointestinal tract by stimulating Gs, whereas pertussis toxin uncouples Gi and Go isoforms from their GPCRs. • Phospholipase C/IP3: the Gq isoform of G-proteins facilitates the production of two important second messengers from membrane phospholipids: IP3 and DAG. Gq activates phospholipase C, the enzyme responsible for the catalytic conversion of phosphat idylinositol 4,5-bisphosphate into IP3 and DAG. IP3 is a watersoluble agonist for a ligand-gated channel (IP3R), the principal mediator of GPCR-induced Ca2+ release from intracellular stores. In contrast, DAG is a membrane-bound second messenger, which activates one of ten different isoforms of protein kinase C. Inactivation of the inositol phosphate system is brought about by metabolism of IP3 and DAG by a combination of phosphorylation (kinases) and dephosphorylation (phosphatases).
enzyme-linked receptors containing intrinsic guanylyl cyclase activity to modulate the activity of the cardiovascular system. Nuclear receptors form a small number of important targets regulating the activity of many diverse agents, including steroids and fat-soluble vitamins. In contrast to the receptors described previously, the so-called nuclear receptors are soluble proteins found either in the cytoplasm (Class I) or within the cell nucleus (Class II) (Figure 4). Agonist activation of Class I targets (e.g. by steroids or prolactin) results in the formation of dimeric receptor complexes, which translocate into the nucleus and bind to ‘hormone-response elements’ within DNA to regulate gene transcription. Class II receptors reside within the cell nucleus and regulate lipid and drug metabolizing enzymes. Both classes of receptor contain highly conserved structural motifs (zinc fingers) to assist with recognition and binding of the receptor to hormone response elements of DNA.
Enzyme-linked receptors are composed of approximately 1000 amino acid residues arranged into a large extracellular domain, a single transmembrane region and a variable-length intracellular domain (Figure 4). Enzyme-linked receptors mediate the actions of various proteins involved in tissue repair, defence and growth (e.g. insulin, inflammatory cytokines and growth factors). One of the defining features of this class of receptor is the presence of specific enzymes, usually kinases, within the intracellular domain. On ligand binding, many receptors undergo dimerization (coup ling of two receptors into one large complex) which, in many cases, facilitates the ‘autophosphorylation’ of tyrosine, serine or threonine residues on each receptor. The phosphorylated residues can then attract specific SH2 (Src homology) proteins and enzymes to alter cellular function through changes in gene transcription. An alternative mechanism adopted by atrial natriuretic factor uses
ANAESTHESIA AND INTENSIVE CARE MEDICINE 8:10
Agonists An agonist is a chemical that binds to and activates a receptor to alter the function of its host cell. Agonists possess both affinity and efficacy. In simple terms, the ability of an agonist (or drug) to bind to the receptor is termed affinity and is defined as the ratio of the binding rate (k+1) and dissociation rate (k−1), that is, affinity (kA )=k − 1/k+1 , and is often measured experimentally using radioactive binding techniques. In contrast, efficacy describes the ability of the drug to activate the receptor once it has bound. Full agonists are capable of producing a maximal response or maximal efficacy (this may occur when only a fraction of receptors are occupied, hence the concept of ‘spare’ receptors), whereas 441
© 2007 Elsevier Ltd. All rights reserved.
Pharmacology
artial agonists are not capable of producing a full response even p in the presence of high concentrations of agonist (Figure 5):
Competitive reversible and irreversible receptor antagonism 100
where: A is the agonist (or drug A), R is the receptor, AR is the bound receptor, AR* is the activated receptor, and β and α are the receptor activation and deactivation constants, respectively. In the clinical setting measurement of affinity and efficacy is problematic because a drug might produce its effects by altering a combination of physiological parameters. For example, epinephrine increases blood pressure by increasing the rate and force of contraction of the heart (β-adrenoreceptors) but also causes profound vasoconstriction (α-adrenoreceptors). Therefore, the term potency defines the concentration of a drug required to produce a measurable clinical response, usually 50% of the maximal possible response (effective concentration; EC50). Highly potent drugs will have smaller EC50s compared with less potent drugs (Figure 5).
Response (% maximum)
Increasing Efficacy
Response (% maximum)
Decreasing 0.1
1
10
100
1000
Log agonist concentration Agonists: a and b are full agonists; a is a more potent agonist than b, but has equal efficacy; c has a lower efficacy (partial agonist) than either a or b; a and c are equipotent; c is more potent than b
Figure 5
ANAESTHESIA AND INTENSIVE CARE MEDICINE 8:10
100
1000
Inverse agonists Inverse agonists, as their name suggests, produce the opposite effects of agonists. This newly recognized phenomenon is possible only if the receptor is active in the absence of agonist (i.e. it has constitutive activity). In this situation a competitive antagonist on its own will have no effect on constitutive activity (because no agonist is present), but an inverse agonist will produce a concentration-dependent reduction in receptor activity (i.e. it has negative efficacy). Examples of receptors displaying constitutive activity include certain subtypes of GPCRs, GABAergic and cannabinoid receptors. ◆
0 0.01
10
effects can be overcome by increasing the concentration of competing agonist (e.g. atracurium-induced neuromuscular blockade can be overcome by increasing acetylcholine concentrations with neostigmine, although clinically this is probably achieved by activating spare receptors). Graphically, this is represented by a parallel and rightward shift of the concentration–response curve (Figure 6). In contrast, irreversible competitive antagonists covalently bind to the receptor, and blockade cannot be overcome by increas ing the agonist concentration (i.e. the effect is insurmountable). Consequently, the maximal effect of the agonist is reduced in the presence of irreversible antagonist (Figure 6). Clinically useful examples of irreversible competitive antagonists include aspirin and omeprazole acting on an enzyme and a proton pump, respectively. The term non-competitive antagonism is reserved for drugs that do not compete at the agonist binding site, but prevent signal transduction within the receptor complex. The effects of non-competitive antagonists are also insurmountable.
c 20
1
Figure 6
60
40
0.1
The binding of a reversible competitive antagonist can be overcome by increasing agonist concentration (agonist curve shifts to the right but efficacy remains unchanged) whereas an irreversible antagonist decreases the efficacy of the agonist
Decreasing
b
Agonist + reversible competitive antagonist
Log agonist concentration
100
a
40
0.01
Agonist concentration–response curves
80
60
0
where: B is the antagonist, R is the receptor, BR is the bound receptor. Reversible competitive antagonism is surmountable. In other words, when an antagonist binds to the agonist-binding site its
Potency
Agonist + irreversible competitive antagonist
20
Antagonists Competitive receptor antagonists are classed as either reversible or irreversible. Competitive antagonists bind to the receptor and therefore possess affinity, but they are unable to produce a response and thus lack efficacy. Furthermore, they prevent the agonist from binding and so block its ability to activate the receptor:
Increasing
Agonist alone
80
442
© 2007 Elsevier Ltd. All rights reserved.