| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Autonomic Neuroscience Centre, Royal Free and University College Medical School, London, United Kingdom
ABSTRACT I. INTRODUCTION II. BACKGROUND A. Autonomic Neuromuscular Transmission B. Autonomic Ganglia C. Central Nervous System D. Purinergic Receptor Subtypes 1. P1 receptors 2. P2X receptors 3. P2Y receptors 4. Heteromultimeric receptors E. ATP Storage, Release, and Breakdown 1. ATP storage and release 2. ATP breakdown F. Plasticity of Purinergic Signaling III. ATP AS A COTRANSMITTER A. Sympathetic Nerves B. Parasympathetic Nerves C. Sensory-Motor Nerves D. Intrinsic Nerves in the Gut and Heart E. Peripheral Motor Nerves F. Nerves in the Brain and Spinal Cord IV. NEUROTRANSMISSION AND NEUROMODULATION IN AUTONOMIC GANGLIA A. Sympathetic Ganglia B. Adrenal Chromaffin Cells C. Parasympathetic Ganglia D. Enteric Ganglia V. SENSORY NEURONS A. Dorsal Root Ganglia B. Nodose Ganglia C. Trigeminal Ganglia D. Petrosal Ganglia E. Retinal Ganglia F. Sensory Nerve Fibers and Terminals 1. Lung 2. Gut 3. Carotid body 4. Heart 5. Skin, muscle, and joints 6. Inner ear 7. Nasal organ 8. Taste buds VI. NEUROTRANSMISSION AND NEUROMODULATION IN THE CENTRAL NERVOUS SYSTEM A. Cortex B. Hippocampus C. Cerebellum D. Basal Ganglia E. Midbrain F. Thalamus G. Habenula H. Behavioral Studies 1. Learning and memory 2. Sleep and arousal 3. Locomotion 4. Feeding 5. Mood and motivation VII. CENTRAL CONTROL OF AUTONOMIC FUNCTION A. Ventrolateral Medulla B. Trigeminal Mesencephalic Nucleus C. Area Postrema D. Locus Coeruleus E. Nucleus Tractus Solitarius F. Motor and Sensory Nuclei G. Hypothalamus H. Spinal Cord VIII. NEURON-GLIA INTERACTIONS A. P1 and P2 Receptors on Glial Cells B. Neuron-Astrocyte Interactions C. Interactions of Axons With Schwann Cells and Oligodendrocytes D. Interactions Between Neurons and Microglia IX. PURINERGIC NEUROEFFECTOR TRANSMISSION A. Exocrine Glands 1. Salivary glands 2. Exocrine pancreas 3. Lacrimal glands 4. Sweat glands 5. Mammary glands B. Endocrine and Neuroendocrine Cells 1. Pituitary 2. Thyroid 3. Thymus 4. Testis 5. Ovary 6. Endocrine pancreas 7. Pineal 8. Adrenal 9. Neuroendocrine cells C. Endothelial Cells D. Secretory Epithelial Cells in Visceral Organs 1. Airways 2. Kidney 3. Gut 4. Liver and gall bladder 5. Reproductive organs 6. Adipose tissue E. Immune Cells F. Bone Cells, Joints, and Keratinocytes G. Choroid Plexus H. Interstitial Cells of Cajal I. Sensory Epithelia 1. Eye 2. Ear 3. Nasal mucosa X. ONTOGENY AND PHYLOGENY OF PURINERGIC NEUROTRANSMISSION A. Pre- and Postnatal Development and Aging 1. Central nervous system 2. Ganglia 3. Retina 4. Skeletal neuromuscular junction 5. Gastrointestinal tract 6. Cardiovascular system 7. Lung 8. Urinary bladder 9. Inner ear 10. Vas deferens and seminal vesicles 11. Other organs B. Purinergic Neurotransmission in Invertebrates and Lower Vertebrates 1. Invertebrates A) COELENTERATES. B) PLATYHELMINTHS. C) ECHINODERMS. D) ANNELIDS. E) NEMATODES. F) MOLLUSCS. G) ARTHROPODS: CRUSTACEANS. H) ARTHROPODS: INSECTS. 2. Lower vertebrates A) ELASMOBRANCH FISH. B) TELEOST FISH. C) AMPHIBIANS. D) REPTILES. E) BIRDS. 3. Summary XI. NEUROPATHOLOGY A. Peripheral Nervous System 1. Diseases of the lower urinary tract 2. Penile erection 3. Heart failure 4. Hypertension 5. Diabetes 6. Gut and liver disorders 7. Diseases of kidney and ureter 8. Purinergic mechanosensory transduction and nociception A) URINARY BLADDER. B) URETER. C) GUT. D) UTERUS. E) TOOTH PULP. F) TONGUE. G) OLFACTORY EPITHELIUM. H) SKIN, MUSCLE, AND JOINTS. I) CANCER. 9. Respiratory diseases 10. Musculoskeletal disorders and arthritis B. Central Nervous System 1. Trauma 2. Neurodegenerative diseases A) ALZHEIMER'S DISEASE. B) PARKINSON'S DISEASE. C) HUNTINGTON'S DISEASE. D) MS. 3. Cerebral ischemia 4. Migraine 5. Neuropsychiatric disorders 6. Epileptic seizures 7. Cancer and encephalitis 8. Abnormalities in central control of peripheral function 9. Central purinergic pathways and neuropathic pain 10. Alcohol and drug addiction 11. Diseases of special senses A) EYE. B) EAR. C) NASAL ORGANS. XII. CONCLUDING COMMENTS AND FUTURE DIRECTIONS ACKNOWLEDGMENTS REFERENCES
 |
ABSTRACT |
|---|
 Top NextReferences |
|---|
|
I. INTRODUCTION |
|---|
|
TopPreviousNextReferences |
|---|
|
II. BACKGROUND |
|---|
|
TopPreviousNextReferences |
|---|
A. Autonomic Neuromuscular Transmission
There was early recognition of atropine-resistant responses of the gastrointestinal tract to parasympathetic nerve stimulation (999, 1138, 1321). However, it was not until the early 1960s that autonomic transmission other than adrenergic and cholinergic was established. In 1963, electrical activity was recorded in the guinea pig taenia coli using the sucrose-gap technique, and after stimulation of the intramural nerves in the presence of adrenergic and cholinergic blocking agents, an inhibitory hyperpolarizing potential was observed (282, 283). The hyperpolarizing responses were blocked by tetrodotoxin (TTX), a neurotoxin that prevents the action potential in nerves without affecting the excitability of smooth muscle cells (233; Fig. 1A), indicating their neurogenic nature and establishing them as inhibitory junction potentials (IJPs) in response to NANC nerves. This work was extended by an analysis of the mechanical responses to NANC nerve stimulation of the taenia coli (284). NANC mechanical responses were also observed by Martinson and Muren in the cat stomach upon stimulation of the vagus nerve (1112), and NANC inhibitory innervation of the portal vein was also demonstrated (780).
|
By the end of the 1960s, evidence had accumulated for NANC nerves in the respiratory, cardiovascular, and urinogenital systems as well as in the gastrointestinal tract (239). The existence of NANC neurotransmission is now firmly established in a wide range of peripheral and central nerves, and fuller accounts of the development of this concept are available (see Refs. 248, 279, 288).
In the late 1960s, systematic studies were undertaken in an attempt to identify the transmitter utilized by the NANC nerves of the gut and urinary bladder. Several criteria, which must be satisfied before establishing a substance as a neurotransmitter (503), were considered (241, 285). First, a putative transmitter must be synthesized and stored within the nerve terminals from which it is released. Once released it must interact with specific postjunctional receptors, and the resultant nerve-mediated response must be mimicked by the exogenous application of the transmitter substance. Also, enzymes that inactivate the transmitter and/or uptake systems for the neurotransmitter or its derivatives must also be present, and finally, drugs that affect the nerve-mediated response must be shown to modify the response to exogenous transmitter in a similar manner.
Many substances were examined as putative transmitters in the NANC nerves of the gastrointestinal tract and bladder, but the substance that best satisfied the above criteria was the purine nucleotide ATP (285; Fig. 1, B and C). Nerves utilizing ATP as their principal transmitter were subsequently named "purinergic" (240), and a tentative model of storage, release, and inactivation of ATP for purinergic nerves was proposed (241; Fig. 2). Since then a great deal of evidence followed in support of the purinergic hypothesis (see Refs. 257, 260, 292, 488, 661, 1278, 1977), although there was considerable opposition to this idea in the first decade or two after it was put forward (see Refs. 242, 643). I believe that this was partly because biochemists felt that ATP was established as an intracellular energy source involved in various metabolic cycles and that such a ubiquitous molecule was unlikely to be involved in extracellular signaling. However, ATP was one of the biological molecules to first appear and, therefore, it is not surprising that it should have been used for extracellular, in addition to intracellular, purposes early in evolution (258). The fact that potent ectoATPases were described in most tissues in the early literature was also a strong indication for the extracellular actions of ATP. In more recent studies, transient clusters of receptors for ATP have been shown to accumulate on smooth muscle membranes opposite autonomic nerve varicosities in close contact with them (707, 1782).
|
An effect of ATP on autonomic ganglia was first reported in 1948 when Feldberg and Hebb (537) demonstrated that intra-arterial injection of ATP excited neurons in the cat superior cervical ganglia (SCG). Later work from de Groat's laboratory showed that in the cat vesical parasympathetic ganglia and rat SCG, purines inhibited synaptic transmission through adenosine receptors, but high concentrations of ATP depolarized and excited the postganglionic neurons (1699, 1700). The earliest intracellular recordings of the action of ATP on neurons were obtained in frog sympathetic ganglia (24, 1566). ATP produced a depolarization through a reduction in K+ conductance. ATP was shown to excite mammalian dorsal root ganglia (DRG) neurons and some neurons from the dorsal horn of the spinal cord (819, 966). These responses were associated with an increase in membrane conductance (see sects. VA and VIIH).
Following the early studies of Feldberg and Sherwood (538), there were reports that showed that adenosine acted via adenylate cyclase to produce cAMP in cerebral cortex slices and that this was antagonized by the methylxanthines, theophylline and caffeine (1504). Electrically evoked release of nucleotides and nucleosides from both brain slices and synaptosomes prepared from cerebral cortex raised the possibility that they may participate in intercellular transmission (983, 1384). These in vitro experiments were extended to the intact cerebral cortex (1665). It was shown that iontophoretic application of adenosine and several adenine nucleotides depressed the excitability of cerebral cortical neurons including identified Betz cells; cAMP, adenine, and inosine were less effective, whereas ATP caused an initial excitation followed by depression (1347). Adenosine and ATP also depressed firing in cerebellar Purkinje cells (959). About the same time microiontophoretic application of adenine nucleotides was shown to depress the spontaneous firing of corticospinal and other unidentified cerebral cortical neurons, although ATP had an additional excitant action on some neurons (1344, 1648). In other studies, adenosine and adenine nucleotides were shown to have an inhibitory action on the N-wave (a postsynaptic potential) amplitude in neurons of guinea pig olfactory cortex slices, but not on postsynaptic potentials in superior colliculus (1274, 1518). Schubert and Kreutzberg (1521) showed that after injections of tritiated adenosine into the visual cortex of rabbits, it was taken up and converted to radioactive nucleotides, which subsequently appeared in the thalamocortical relay cells of the lateral geniculate nucleus, consistent with synaptic transmission. This was supported by similar experiments in the somatosensory cortex (1875).
ATP was shown to activate units of the emetic chemoreceptor trigger zone of the area postrema of cat brain (174). Premature arousal of squirrels from periods of hibernation was evoked by adenosine nucleotides, but not by other purine nucleotides, and it was suggested that this effect was due to their action on neurons in the central nervous system (CNS) (1744). The infusion of cAMP into the hypothalamus of fowl induced behavioral and electrophysiological sleep, whereas dibutyryl cAMP produced arousal (1107). Local or systemic administration of adenosine in normal animals produced electroencephalogram (EEG) and behavioral alterations of the hypnogenic type (714). Cornforde and Oldendorf (385) demonstrated two independent transport systems across the rat blood-brain barrier, one for adenine and the other for adenosine, guanosine, inosine, and uridine. High levels of 5'-nucleotidase were demonstrated histochemically in the substantia gelatinosa of mouse spinal cord (1672). Early studies of the actions of purines on the CNS were reviewed by Burnstock (245), and important papers about the excitatory actions of ATP on subpopulations of spinal dorsal horn neurons were published in the early 1980s (608, 819, 1490) and excitation of single sensory neurons in the rat caudal trigeminal nucleus by iontophoretically applied ATP (1486). Although most of the early emphasis was about the neuromodulatory roles of adenosine, it was later recognized that fast synaptic transmission involving ATP was widespread in the CNS (see sects. VI and VII).
Early observations of mentally ill patients suggested that purines may play a role in the brain of man (see also sect. XIB5). Thus adenine nucleotides were implicated in depressive illness (7, 708, 1189). In the hypothesis proposed for the mechanism of depression by Abdulla and McFarlane (7), the effect of adenine nucleotides on prostaglandin biosynthesis was implicated. Blood levels of ATP and/or adenosine and urinary cAMP excretion were significantly elevated in patients diagnosed as schizophrenic or in psychotic and neurotic depression (6, 219, 709, but see also Ref. 830). Inherited disorders of purine metabolism in the brain were related to psychomotor retardation, athetosis, and self-mutilation (Lesch-Nyhan syndrome) (133, 1020, 1534). Antidepressant drugs such as imipramine and amitriptyline potentiated the suppression of neuronal firing in rat cerebral cortex by adenosine (1342, 1648). It was claimed that depressive symptoms in patients relate to hypoxanthine levels in the cerebrospinal fluid (1245). Competitive interactions between adenosine and benzodiazepines in cerebral cortical neurons were reported, and evidence was presented to suggest that morphine releases ATP, and that after breakdown to adenosine, depresses neurotransmission in the cortex (1350). It was suggested that adenosine was involved in the initial phase of seizure-induced functional hyperemia in the cortex (1519).
The majority of studies of the extracellular actions of ATP have been concerned with the short-term events that occur in neurotransmission and in secretion. However, there is increasing awareness that purines and pyrimidines can have potent long-term (trophic) roles in cell proliferation and growth and in disease and cytotoxicity (see Ref. 4; Table 1). An example of synergism between purines and trophic factors comes from studies of the transplantation of the myenteric plexus into the brain (1695, 1696). In these studies, which were originally designed to explore enteric nerves as a possible source for replacement of missing messengers such as dopamine for Parkinson's disease, the myenteric plexus was shown to cause a marked proliferation of nerve fibers in the corpus striatum. An analysis, using coculture of striatal neurons with various elements of the myenteric plexus and enteric neurotransmitters, showed that a growth factor released by enteric glial cells works synergistically with nitric oxide (NO) and ATP (via adenosine) released from NANC inhibitory nerves to promote nerve regeneration (760).
|
Implicit in the concept of purinergic neurotransmission is the existence of postjunctional purinergic receptors, and the potent actions of extracellular ATP on many different cell types also implicates membrane receptors. Purinergic receptors were first defined in 1976 (244), and 2 years later a basis for distinguishing two types of purinoceptor, identified as P1 and P2 (for adenosine and ATP/ADP, respectively), was proposed (246). At about the same time, two subtypes of the P1 (adenosine) receptor were recognized (1069, 1766), but it was not until 1985 that a proposal suggesting a pharmacological basis for distinguishing two types of P2 receptor (P2X and P2Y) was made (291). A year later, two further P2 purinoceptor subtypes were identified, namely, a P2T receptor selective for ADP on platelets and a P2Z receptor on macrophages (661). Further subtypes followed, perhaps the most important being the P2U receptor, which could recognize pyrimidines such as UTP as well as ATP (1265, 1536). In 1994 Mike Williams made the point at a meeting that a classification of P2 purinoceptors based on a "random walk through the alphabet" was not satisfactory, and Abbracchio and Burnstock (3), on the basis of studies of transduction mechanisms (486) and the cloning of nucleotide receptors (194, 1080, 1763, 1843), proposed that purinoceptors should belong to two major families: a P2X family of ligand-gated ion channel receptors and a P2Y family of G protein-coupled receptors. This nomenclature has been widely adopted, and currently seven P2X subunits and eight P2Y receptor subtypes are recognized, including receptors that are sensitive to pyrimidines as well as purines (see Refs. 163, 275, 521, 1392). Receptors for diadenosine polyphosphates have been described on C6 glioma cells and presynaptic terminals in rat midbrain, although they have yet to be cloned (444).
Four subtypes of P1 receptors have been cloned, namely, A1, A2A, A2B, and A3 (see Refs. 370, 423, 582, 1392). All P1 adenosine receptors couple to G proteins and, in common with other G protein-coupled receptors, they have seven putative transmembrane (TM) domains of hydrophobic amino acids, each believed to constitute an 
-helix of 
21–28 amino acids (see Fig. 3A). The NH2 terminus of the protein lies on the extracellular side, and the COOH terminus lies on the cytoplasmic side of the membrane. Typically, the extracellular loop between TM4 and TM5 and the cytoplasmic loop between TM5 and TM6 are extended. The intracellular segment of the receptor interacts with the appropriate G protein, with subsequent activation of the intracellular signal transduction mechanism. It is the residues within the transmembrane regions that are crucial for ligand binding and specificity and, with the exception of the distal (carboxyl) region of the second extracellular loop, the extracellular loops, the COOH terminus, and the NH2 terminus do not seem to be involved in ligand recognition (1275). Site-directed mutagenesis of the bovine A1 adenosine receptor suggests that conserved histidine residues in TM6 and TM7 are important in ligand binding. Specific agonists and antagonists are available for the P1 receptor subtypes (817; Table 2). For a specific review of P1 receptors in the nervous system, the reader is referred to Reference 1423.
|
|
Members of the existing family of ligand-gated nonselective cation channel P2X1–7 receptor subunits show a subunit topology of intracellular NH2 and COOH termini possessing consensus binding motifs for protein kinases; two transmembrane-spanning regions (TM1 and TM2), the first involved with channel gating and the second lining the ion pore; a large extracellular loop, with 10 conserved cysteine residues forming a series of disulfide bridges; hydrophobic H5 regions close to the pore vestibule, for possible receptor/channel modulation by cations (magnesium, calcium, zinc, copper, and proton ions); and an ATP-binding site, which may involve regions of the extracellular loop adjacent to TM1 and TM2 (see Fig. 3B). The P2X1–7 receptors show 30–50% sequence identity at the peptide level. The stoichiometry of P2X1–7 receptor subunits is thought to involve three subunits that form a stretched trimer (see Refs. 109, 891, 1155, 1240).
It has become apparent that the pharmacology of the recombinant P2X receptor subtypes expressed in oocytes or other cell types is often different from the pharmacology of P2X receptor-mediated responses in naturally occurring sites. This is partly because heteromultimers as well as homomultimers are involved in forming the trimer ion pores (see below). Spliced variants of P2X receptor subtypes might play a part (334, 1578). For example, a splice variant of the P2X4 receptor, while it is nonfunctional on its own, can potentiate the actions of ATP through the full-length P2X4 receptors (1721). Third, the presence in tissues of powerful ectoenzymes that rapidly break down purines and pyrimidines is not a factor when examining recombinant receptors, but is in vivo (1979).
P2X7 receptors are predominantly localized on immune cells and glia, where they mediate proinflammatory cytokine release, cell proliferation, and apoptosis. P2X7 receptors, in addition to small cation channels, upon prolonged exposure to high concentrations of agonist, large channels, or pores are activated that allow the passage of larger molecular weight molecules. The possible mechanisms underlying the transition from small to large channels have been considered (508, 621).
The P2X receptor family shows many pharmacological and operational differences (634; see Table 2). The kinetics of activation, inactivation, and deactivation also vary considerably among P2X receptors. Calcium permeability is high for some P2X subunits and Cl– permeability for others, properties that are functionally important. For more specific reviews of the molecular physiology of the P2X receptor, the reader is referred to Khakh et al. (891), North (1257), Egan et al. (508), Stojilkovic et al. (1645), and Roberts et al. (1429).
Metabotropic P2Y receptors (P2Y1, P2Y2, P2Y4, P2Y6, P2Y11, P2Y12, P2Y13, and P2Y14) are characterized by a subunit topology of an extracellular NH2 terminus and intracellular COOH terminus, the latter possessing consensus binding motifs for protein kinases; seven transmembrane-spanning regions, which help to form the ligand-docking pocket; a high level of sequence homology between some transmembrane-spanning regions, particularly TM3, TM6, and TM7; and a structural diversity of intracellular loops and COOH terminus among P2Y subtypes, so influencing the degree of coupling with Gq/11, Gs, Gi, and Gi/o proteins (5; see Fig. 3C). Each P2Y receptor binds to a single heterotrimeric G protein (Gq/11 for P2Y1,2,4,6), although P2Y11 can couple to both Gq/11, and Gs, whereas P2Y12 and P2Y13 couple to Gi and P2Y14 to Gi/o. Many cells express multiple P2Y subtypes (see Refs. 5, 1800). P2Y receptors show a low level of sequence homology at the peptide level (19–55% identical) and, consequently, show significant differences in their pharmacological and operational profiles. Some P2Y receptors are activated principally by nucleoside diphosphates (P2Y1,6,12), while others are activated mainly by nucleoside triphosphates (P2Y2,4). Some P2Y receptors are activated by both purine and pyrimidine nucleotides (P2Y2,4,6), and others by purine nucleotides alone (P2Y1,11,12). In response to nucleotide activation, recombinant P2Y receptors either activate phospholipase C (PLC) and release intracellular calcium or affect adenylyl cyclase and alter cAMP levels. In recent years P2Y G protein-coupled receptors in neurons have been found to modulate the activity of voltage-gated ion channels in the cell membrane through certain actions of activated G proteins. For example, P2Y receptor subtypes that act via Gi/o proteins can involve N-type Ca2+ channels, while the M-current K+ channel can be inhibited through the activation of Gq/11-linked P2Y receptor subtypes (5). There is little evidence to indicate that P2Y5, P2Y9, and P2Y10 sequences are nucleotide receptors or affect intracellular signaling cascades and consequently have been dropped from International Union of Pharmacology (IUPHAR) P2Y receptor nomenclature and have been termed "orphans."
2-Methylthio ADP (2-MeSADP) is a potent agonist of mammalian P2Y1 receptors and N6-methyl-2'-deoxyadenosine 3',5'-bisphosphate (MRS2179), MRS2269 and MRS2286 have been identified as selective antagonists (221). At P2Y2 and P2Y4 receptors in the rat, ATP and UTP are equipotent, but the two receptors can be distinguished with antagonists, that is, suramin blocks P2Y2, while Reactive blue 2 blocks P2Y4 receptors (165, 1862). P2Y6 is UDP selective, while P2Y7 turned out to be a leukotriene receptor (1936). P2Y8 is a receptor cloned from frog embryos, where all the nucleotides are equipotent (164), but no mammalian homolog has been identified to date, apart from a report of P2Y8 mRNA in undifferentiated HL60 cells (13). The P2Y12 receptor found on platelets was not cloned until more recently (754), although it has only 19% homology with the other P2Y receptor subtypes. It seems likely to represent one of a subgroup of P2Y receptors, including P2Y13 and P2Y14, for which transduction is entirely through adenylate cyclase (2). It has been suggested that P2Y receptors can be subdivided into two subgroups, namely, one that includes P2Y1,2,4,6,11, the other includes P2Y12,13,14, largely on the basis of structural and phylogenetic criteria (see Ref. 5).
For a recent review of P2Y receptor molecular biology, pharmacology, cell distribution, and physiology, the reader should refer to Reference 5. The structure and properties of current P2Y receptor subtypes and the current status of P2 receptor subtype agonists and antagonists are summarized in Table 2.
The pharmacology of purinergic signaling is complicated because P2X receptor subunits can combine to form either homomultimers or heteromultimers (see Refs. 1257, 1800). Heteromultimers are clearly established for P2X2/3 receptors in nodose ganglia (1027, 1388), P2X4/6 receptors in CNS neurons (1006), P2X1/5 receptors in some blood vessels (699, 1718), and P2X2/6 receptors in the brain stem (916). P2X7 receptors do not form heteromultimers, and P2X6 receptors will not form a functional homomultimer without extensive glycosylation (1717).
P2Y receptor subtypes can also form heteromeric complexes (5, 1206), and most recently, adenosine A1 receptors have been shown to form a heteromeric complex with P2Y1 receptors (see Refs. 1381, 1941). Dopamine D1 and adenosine A1 receptors have also been shown to form functionally interactive heteromeric complexes (645).
E. ATP Storage, Release, and Breakdown
The cytoplasm of most neurons contains 2–5mM ATP, and higher concentrations of ATP (up to 100 mM) are stored in synaptic vesicles. Synaptic vesicles also contain other nucleotides such as ADP, AMP, Ap4A, Ap5A, and GTP, but at lower concentrations. Sperm, tumor cells, and the epithelial cells of the lens of the eye have exceptionally high intracellular levels of ATP and granules in adrenal chromaffin cells, Merkel cells, platelets, and pancreatic insulin-containing 
-cells also contain significant amounts of ATP (see Refs. 1260, 1627). A recent paper has shown that ATP transport into brain synaptic vesicles can be distinguished from other neurotransmitter transport systems in terms of its mechanism and energy requirements (1947).
Release of ATP from exercising human forearm muscle was reported by Tom Forrester and colleagues (569) and from the perfused heart during coronary vasodilation in response to hypoxia was reported by Paddle and Burnstock (1288). However, until recently, it was usually assumed that the only source of extracellular ATP acting on purinoceptors was damaged or dying cells, but it is now recognized that ATP release from healthy cells is a physiological mechanism (see Refs. 160, 487, 1004, 1529). ATP is released from both peripheral and central neurons (285, 755, 881, 1302, 1652, 1855, 1857, 1889), but also from many nonneuronal cell types during mechanical deformation in response to shear stress, stretch, or osmotic swelling, as well as hypoxia and stimulation by various agents (160, 179).
ATP release by mechanical distortion of urothelial cells during distension of the bladder was first demonstrated by Ferguson and colleagues in 1997 (544) and later by Vlaskovska et al. (1797). ATP release by distension was demonstrated from urothelial cells in ureter (932) and from mucosal epithelial cells of the colorectum (1895). It is also released from osteoblasts (1441); astrocytes (373); epithelial cells in the tongue (1444), lung (56, 466), and kidney (127); keratinocytes in the skin; and glomus cells in the carotid body.
There is an active debate about the precise transport mechanism(s) involved in ATP release. There is compelling evidence for exocytotic vesicular release of ATP from nerves, but for ATP release from nonneuronal cells, various transport mechanisms have been proposed, including ATP-binding cassette (ABC) transporters, connexin, or pannexin hemichannels or possibly plasmalemmal voltage-dependent anion and P2X7 receptor channels, as well as vesicular release (see Refs. 160, 417, 439, 1004, 1475, 1529, 1631). Perhaps surprisingly, evidence was presented that the release of ATP from urothelial cells in the ureter is also largely vesicular, since monensin and brefeldin A, which interfere with vesicular formation and trafficking, inhibited distension-evoked ATP release, but not gadolinium, an inhibitor of stretch-activated channels, or glibenclamide, an inhibitor of members of the ABC protein family (932). Exocytotic vesicular release of ATP from endothelial cells (160, 261), osteoblasts (1441), fibroblasts (179), and astrocytes (373, 1169) has also been reported. There is increased release of ATP from endothelial cells during acute inflammation (159).
ATP released from nerves, or by autocrine and paracrine mechanisms from nonneuronal cells, is involved in a wide spectrum of physiological and pathophysiological activities, including synaptic transmission and modulation, pain and touch perception, vasomotor effects, platelet aggregation, endothelial cell release of vasorelaxants, immune defense, epithelial ion and water transport, as well as cell proliferation, migration, differentiation, and death.
Local probes for real-time measurement of ATP release in biological tissues have been developed recently (583, 1061, 1203, 1329).
ATP released from cells is regulated by a number of proteins that have their catalytic site on the outer side of the plasma membrane (121, 976, 1978). A recent review focuses on ectonucleotidases in the nervous system (1980). Extracellular nucleotides can be hydrolyzed by nonspecific enzymes, such as glycoslyphospatidylinositol-anchored ectoalkaline phosphatases and ecto-5'-nucleotidases or ectonucleotidases with more distinct characteristics that are now classified into two families (see Fig. 3D). E-NTPDases (CD39) family are ecto-nucleoside triphosphate diphosphohydrolases that hydrolyze nucleoside 5'-tri- and diphosphates. Another family of enzymes, E-NPP (with 3 subtypes), are ecto-nucleotide pyrophosphatase/phosphodiesterases with a broad substrate specificity. These can hydrolyze phosphodiester bonds of nucleotides and nucleic acids and pyrophosphatase bonds of nucleotides and nucleotide sugars, e.g., cleavage of ATP to AMP and PPi and conversion of cAMP to AMP. Some of these ectonucleotidases have distinct patterns of distribution in different cell types and are regulated during physiological and pathophysiological processes, probably in association with purine and pyrimidine signaling. The catalytic site of ectonucleotidases faces the extracellular medium, but some isoforms can be cleaved or released in a soluble form, in which case they can be regarded as ectonucleotidases. There are also other ectoenzymes that contribute to levels of extracellular nucleotides, such as interconversion of nucleotides by ecto-nucleoside diphosphokinase (ecto-NDPK) and ectoadenylate cyclase, possibly a production of ATP by F0-F1 ATP synthase, and use of ATP as a phosphate donor for ecto-protein kinase reactions. Ectoenzymes can act to regulate synaptic activity, controlling ATP and adenosine levels, depending on the synaptic plasticity developed in both physiological and pathophysiological conditions (171, 657). Excellent reviews are available summarizing the current status of extracellular ATP breakdown (see Refs. 1978, 1980; see also the recent Special Issue of "Purinergic Signaling" devoted to Ecto-nucleotidases, volume 2, number 2, 2006).
F. Plasticity of Purinergic Signaling
There was early recognition that the expression of purinergic cotransmitters and of receptors in the autonomic nervous system shows marked plasticity during development and ageing, in the nerves that remain following trauma or surgery, under the influence of hormones and in various disease situations (see Refs. 4, 254, 292). For plasticity of purinergic signaling in development and aging, see section XA1, and in disease, see section XI. Examples of neuronal plasticity occurring in healthy adults during pregnancy or following surgical interventions in visceral organs and the CNS follow.
Plasticity of purinergic signaling has been observed in the urinary bladder. Suppressed bladder contractility during pregnancy is associated with decreased muscarinic receptor density, while the affinity of purinergic receptors for ATP is increased (1716). The amplitude of NANC transmission in detrusor strips from mature female rats was diminished in ovariectomized animals (509). Pregnancy substantially increases the purinergic components of the response of the rabbit bladder to field stimulation (1022). In contrast, there was a decrease in excitatory junction potentials (EJPs), probably mediated by ATP, in guinea pig uterine artery (1686).
Capsaicin treatment of newborn rats leads to selective degeneration of some sensory nerve fibers. In a study of rat bladder in 3-mo-old rats treated at birth with capsaicin, contractions evoked by electrical field stimulation were significantly larger than those of control (vehicle-treated) animals, an effect which preferentially involves the cholinergic component of the response, although there was some increase, too, in the purinergic component (1975). However, since contractions in response to exogenous carbachol or ATP were not significantly different, this suggested that the changes involve prejunctional mechanisms. Capsaicin treatment, causing selective sensory denervation of the rat ureter, leads to increased sympathetic innervation (1496), perhaps involving an increase in release of both NE and ATP.
The urinary bladder of the rat, deprived of its motor innervation, increases severalfold in weight in response to distension; this increase in weight is due to both hyperplasia and hypertrophy of the smooth muscle (514). Since it is now known that distension of the bladder leads to substantial release of ATP from urothelial cells (see sect. IIE1) and ATP is known to have proliferative actions (4), purines seem likely to participate in the trophic changes that occur in the bladder. Incorporation of bowel tissue into the bladder wall has been used to increase bladder capacity and/or decrease bladder pressure; after 4–12 wk, the contractile response of the transplanted rabbit intestine underwent a partial change in the response to nerve stimulation and ATP from that of intestine towards that of detrusor (116).
Following chemical sympathectomy produced by long-term treatment with guanethidine and subsequent loss of the cotransmitter ATP in the vas deferens and spleen, there is an increase in density of P2X receptor sites, although there was a decrease in receptor affinity (1966).
Chronic food restriction alters P2Y1 receptor mRNA expression in the nucleus accumbens of the rat (969). Cerebellar lesion upregulates P2X1 and P2X2 receptors in the precerebellar nuclei of the rat, perhaps related to the survival of injured neurons (563). In vitro studies of organotypic cultures and in vivo experiments on hippocampus from gerbils subjected to bilateral common carotid occlusion showed that P2X2 and P2X4 receptors were upregulated by glucose/oxygen deprivation (321, 322). It was speculated that the changes in P2X receptor expression might be associated with ischemic cell death. There are data indicating a trophic role for ATP in the hippocampus (1584). It was shown that ATP and its slowly hydrolyzable analogs strongly inhibited neurite outgrowth and also inhibited aggregation of hippocampal neurons; it was suggested that the results indicate that extracellular ATP may be involved in synaptic plasticity through modulation of neural cell adhesion molecule (NCAM)-mediated adhesion and neurite outgrowth. Utilization of green fluorescent protein (GFP)-tagged P2X2 receptors on embryonic hippocampal neurons has led to the claim that ATP application can lead to changes in dendritic morphology and receptor distribution (891).
Propagation of intercellular Ca2+ waves between astrocytes depends on the diffusion of signaling molecules through gap junction channels (see sect. VIIIA). Deletion of the main gap junction protein connexin-43 (Cx43) by homologous recombination results in a switch in mode of intercellular Ca2+ wave propagation to a purinoceptor-dependent mechanism. This compensatory mechanism in Cx43 knockout mice for intercellular Ca2+ wave propagation is related to a switch from P2Y1 to a UTP-sensitive P2Y4 receptor in spinal cord astrocytes (1654). Trophic effects of purines on neurons and glial cells have been reviewed (4, 1226, 1398).
Following chronic constriction injury to the sciatic nerve, the number of P2X3 receptor-positive small- and medium-diameter neurons increased in DRG, compared with sham-operated animals (1261, 1734). In addition, spinal cord immunoreactivity increased on the side ipsilateral to the ligated nerve, consistent with upregulation of purinergic receptors on presynaptic terminals of the primary sensory nerves. A decrease in P2X3 immunoreactivity and function in DRG of rats occurs after spinal nerve ligation (854). Changes in gene expression of multiple subtypes of P2X receptors on DRG neurons (L5) after spinal nerve ligation have been reported (904). After nerve injury, the mRNA for P2X5 receptors was increased, those for the P2X3 and P2X6 receptors were decreased, and those for P2X2 and P2X4 receptors were unchanged. However, immunostaining for receptor protein showed an increase from 23 to 73% P2X2 receptor-positive DRG neurons after nerve ligation. Two days following unilateral section of the cervical vagus nerve, there was a dramatic ipsilateral increase in P2X1, P2X2, and P2X4 receptor immunoreactivity in the cell soma of vagal efferent neurons in the dorsal vagal motor nucleus, but not in the nucleus ambiguous (72). Following surgical sympathectomy, 28% of the spontaneously active afferent fibers in sciatic nerve responded to ATP, compared with none in intact rats (343). After nerve injury, P2X4 receptor expression increased strikingly in hyperactive microglia, but not in neurons or astrocytes, in the ipsilateral spinal cord; this appears to be associated with tactile allodynia (1731 and see sect. XIB9).
The interactions of tyrosine kinase and P2Y2 receptor signaling pathways provide a paradigm for the regulation of neuronal differentiation and suggest a role for P2Y2 as a morphogen receptor that potentiates neurotrophin signaling in neuronal development and regeneration (64). P2Y2 receptors, which mediate the mitogenic effects of extracellular nucleotides on vascular smooth muscle, are upregulated in the synthetic phenotype in the neointima after balloon angioplasty (763).
|
III. ATP AS A COTRANSMITTER |
|---|
|
TopPreviousNextReferences |
|---|
In keeping with the concept of purinergic cotransmission, there was early recognition that ATP and adenosine modulated prejunctional inhibition of ACh release from the skeletal neuromuscular junction (647, 1424) and of NE release from peripheral sympathetic nerves in a wide variety of tissues, including rabbit kidney, canine adipose tissue, guinea pig vas deferens (367, 724), and rabbit central ear artery, saphenous vein, portal vein, and pulmonary artery (1651, 1780). Prejunctional modulation of ACh release from peripheral cholinergic nerves by purines was observed in the isolated guinea pig ileum and the myenteric plexus longitudinal muscle preparation (1170, 1507, 1795). Clear evidence for this was presented by De Mey et al. (436) who showed that the prejunctional actions of purine nucleotides in canine saphenous vein were mediated largely by adenosine following the rapid breakdown of ATP, since slowly degradable analogs of ATP were ineffective. More recently, evidence has been presented for a prejunctional modulatory action by ATP itself, as well as adenosine, in the iris (589), rat vas deferens (1802), and tail artery (1560).
Purine nucleotides and nucleosides can also act on postjunctional receptors to modulate cholinergic and adrenergic neurotransmission. Purines were reported to increase ACh receptor activity in various preparations, including the rat diaphragm muscle (528), frog skeletal muscle (23), and rabbit iris sphincter (695). The interactions are Ca2+ dependent and may involve interaction with the allosteric site of the receptor-ion channel complex. Purine nucleotides and nucleosides were shown to interact with NE postjunctionally in vitro in the guinea pig seminal vesicles (1204), rabbit kidney (724), guinea pig and mouse vas deferens (751, 1589, 1877), rabbit mesenteric artery (965), and rat mesenteric bed (1391). These neuromodulatory actions of purines have been extensively reviewed (1319, 1420, 1635).
It was recognized early that ATP was costored with catecholamines in adrenal medullary chromaffin cells (150, 741). Subsequently, ATP was shown to be coreleased with epinephrine from chromaffin cells (313, 468). The 1976 cotransmitter hypothesis included the suggestion that NE and ATP might be cotransmitters in sympathetic nerves, following the earlier demonstration that ATP was contained together with NE in sympathetic nerve terminals in a molar ratio estimated to be from 7:1 to 12:1, NE:ATP (627, 988, 1522, 1642). The first evidence for sympathetic cotransmission involving ATP together with NE came from studies of the taenia coli (1652). It was shown that stimulation of periarterial sympathetic nerves led to release of tritium from guinea pig taenia coli preincubated in [3H]adenosine (which is taken up and converted largely to [3H]ATP) and that the release of both tritium and NE was blocked by guanethidine. Soon after, the possibility that ATP might be coreleased with NE in chemical transmission from the hypogastric nerve to the seminal vesicle of the guinea pig was raised and that the substantial residual NANC responses of the cat nictitating membrane following depletion of NE by reserpine might be due to the release of ATP remaining in sympathetic nerves (998, 1204).
The most extensive evidence for sympathetic cotransmission, however, came from studies of the vas deferens, initially by Westfall and colleagues (536, 1853). Although it was not realized at the time, when EJPs were first recorded in smooth muscle cells of the vas deferens in response to stimulation of sympathetic nerves (289, 290; Fig. 1D), they were due to ATP rather than to NE. We were puzzled at the time that EJPs were not abolished by adrenoceptor antagonists; however, since they were abolished by the sympathetic neuron blocking agents bretylium and guanethidine (which are drugs that prevent nerve-mediated release of transmitter), we were correct in assuming they were produced by transmitter released from sympathetic nerves. Subsequent studies showed that EJPs are blocked by the ATP receptor antagonists arylazido aminopropionly-ATP (ANAPP3) and suramin and also following selective desensitization of the ATP receptor with the stable analog of ATP, ,-methylene ATP (,-meATP) (872, 1593), but not by depletion of NE with reserpine (919, 1596). Furthermore, injection of ATP mimicked the EJP, whereas NE did not (1595). Direct evidence for concomitant release of ATP with NE and neuropeptide Y (NPY) from sympathetic nerves supplying the vas deferens was later
