What happens to a neurotransmitter after it is released from the receiving dendrite?

Presynaptic Versus Postsynaptic Effects on Synaptic Transmission

General anesthetics have potent and specific effects on synaptic transmission, including presynaptic actions (by altering transmitter release) and postsynaptic actions (by altering the postsynaptic responses of neurons to specific transmitters). The relative contributions of presynaptic compared with postsynaptic anesthetic effects on synaptic transmission have been difficult to resolve, probably because the effects are transmitter- and synapse-specific. The net effect of anesthetics on synaptic transmission is determined by the relative magnitude and direction of both their presynaptic and postsynaptic effects. The general effects of inhaled anesthetics are to increase inhibitory synaptic transmission and to inhibit excitatory synaptic transmission (Fig. 19.9).

Excitatory synaptic excitation is generally decreased by volatile anesthetics (Fig. 19.10). Experiments in various slice preparations indicate that reduced excitation is primarily caused by presynaptic mechanisms.87,133,214,227-229 A postsynaptic mechanism is also involved because the response to directly applied glutamate is reduced to some degree.229-231 Volatile anesthetics have inconsistent effects on cloned AMPA or NMDA glutamate receptors, but they potentiate kainite receptors,115,129,232,233 consistent with a predominantly presynaptic mechanism for glutamatergic synapses. By contrast, the effects of the nonhalogenated inhaled anesthetics (xenon, nitrous oxide, cyclopropane) appear to be mediated primarily by inhibition of postsynaptic NMDA receptors (discussed earlier). Under some circumstances, such as in patients with defects in mitochondrial complex 1234 and in mice carrying mutations in mitochondrial complex 1,235 inhaled agents suppress glutamate release by interfering with the energy-intensive glutamate recycling pathways,236 thereby leading to extreme anesthetic sensitivity. Recent evidence from conditional knockout mice indicates this mechanism may contribute to various end points even in nonpathogenic states.237

Augmentation of GABAergic inhibition by most general anesthetics is mediated by both presynaptic and postsynapticmechanisms. Enhancement of postsynaptic and extrasynaptic GABAA receptors is well recognized.116 Volatile anesthetics increase spontaneous GABA release and inhibitory postsynaptic current (IPSC) frequency238-242—that is, their presynaptic effects at GABAergic terminals are distinct from those at glutamatergic synapses.

The mechanisms for the presynaptic effects of inhaled anesthetics, like those for their postsynaptic effects, are complex and involve multiple targets. Although a synapse-specific contribution of presynaptic Ca2+ channels is likely,243 presynaptic Na+ channels are more sensitive than the Ca2+ channels coupled to glutamate release. This finding is consistent with observations that the predominant Ca2+ channel coupled to neurotransmitter release at hippocampal glutamatergic synapses (P/Q-type) is insensitive to isoflurane.156 Other presynaptic mechanisms have been proposed, including actions on the vesicle fusion process, as demonstrated in the model organismCaenorhabditis elegans.244,245 However, isoflurane effects on exocytosis in rat hippocampal neurons occur primarily upstream of vesicle fusion.139,246

4.3 Control of dendritic arbor size by neurotransmission

Neural transmission from presynaptic inputs can increase or decrease dendrite growth and complexity. The effect of activity depends on the neuron type or developmental stage. Manipulations of neurotransmission can result in global, cell-wide adaptations or local adjustments to dendritic growth. One manipulation that globally affects dendritic arbor size is inhibition of GABAergic transmission, which also increases excitatory drive. In the optic tectum of Xenopus tadpoles, tectal dendrites (see example in Fig. 1) receive glutamatergic inputs from retinal ganglion cells and inhibitory inputs from local interneurons. Blockade of inhibitory GABA transmission by overexpression of a dominant-negative GABA-A receptor subunit causes dendrite arbors to expand (Shen, Da Silva, He, & Cline, 2009). Similar increases are observed in rodent hippocampal neurons following GABA-A receptor inhibition (Wayman et al., 2006), and in gerbil auditory neurons upon decreased glycinergic transmission (Sanes & Hafidi, 1996; Sanes, Markowitz, Bernstein, & Wardlow, 1992). In the mouse cortex, long-range GABAergic neurons that project to the upper cortical layer stimulate dendrite branching and synaptogenesis of pyramidal neurons (Chen & Kriegstein, 2015). In this example, the effects of blocking GABA release are limited to nearby apical dendrites receiving inputs because outgrowth of distantly located basal dendrites is not affected.

Local modulation of dendritic outgrowth by neurotransmission shapes the arbors of nucleus laminaris neurons (NL) in the chick auditory brainstem. The bi-tufted NL neurons have two dendritic arbors that are similar in size and complexity, with one receiving excitatory inputs from the ipsilateral ear and the other from the contralateral ear. Denervation of one NL dendrite tuft, by transection of axonal inputs or application of tetrodotoxin (TTX), causes the denervated dendrite arbor to shrink (Deitch & Rubel, 1984; Wang & Rubel, 2012). The effect on dendrite size is not due to loss in afferent contact but rather due to loss of neurotransmission, because ablation of the cochlea two synapses away causes similar shrinkage. Differences in input strengths also differentially affect arbor growth: stimulated dendrites increased in arbor size and total branch length while unstimulated dendrites diminished in size (Sorensen & Rubel, 2006, 2011). This adaptive property allows dendrite arbor size to rapidly adjust to functional changes in presynaptic connectivity, thus ensuring optimal sound localization.

Basic Electrophysiology of Neurotransmission

Fig. 12.5 illustrates the results of the classic depolarizing action of acetylcholine on end-plate receptors. Normally, the pore of the channel is closed by approximation of the cylinders (i.e., subunits). When an agonist occupies both α-subunit sites, the protein molecule undergoes a conformational change with a twisting movement along the central axis of the receptor that results in the opening of the central channel through which ions can flow along a concentration gradient. When the central channel is open, sodium and calcium flow from the outside of the cell to the inside and potassium flows from the inside to the outside. The channel in the tube is large enough to accommodate many cations and electrically neutral molecules, but it excludes anions (e.g., chloride). The current transported by the ions depolarizes the adjacent membrane. The net current is depolarizing and creates the end-plate potential that stimulates the muscle to contract. In this instance, downward-going (i.e., depolarizing) current can be recorded bythe patch-clamp electrophysiologic technique previously described (seeFig. 12.4).

The pulse stops when the channel closes by a reversed mechanical conformation (see earlier discussion), which is typically initiated when one or both agonist molecules detach from the receptor. In the activated, open state, the current that passes through each open channel is minuscule, only a few picoamperes (approximately 104 ions/ms). However, each burst of acetylcholine from the nerve normally opens approximately 500,000 channels simultaneously, and the total current is more than adequate to produce depolarization of the end plate and contraction of muscle. Opening of a channel causes conversion of chemical signals from a nerve to the flow of current on the muscle disease to cause end-plate potentials, thereby leading to muscle contraction. The end-plate potential has been viewed as a graded event that may be reduced in magnitude or extended in time by drugs, but, in reality, the end-plate potential is the summation of many all-or-nothing events simultaneously occurring at myriad ion channels. It is these tiny events that are affected by drugs.

Receptors that do not have two molecules of agonist (e.g., acetylcholine) bound remain closed. Both α-subunits must be simultaneously occupied by agonist; if only one of them is occupied, then the channel remains closed (seeFig. 12.5). This is the basis for preventing depolarization by antagonists. NDMRs act by binding to either or both α-subunits and thus preventing acetylcholine from binding and opening the channel. This interaction between agonists and antagonists is competitive, and the outcome—transmission or block—depends on the relative concentrations and binding characteristics of the drugs involved (see section on “Drug Effects on Postjunctional Receptors”).

Individual channels are also capable of a wide variety of conformational states.17,57 They may stay open or remain closed and thereby affect total current flow across the membrane, but they can do more. They may open for a longer or shorter time than normal, open or close more gradually than usual, open briefly and repeatedly (i.e., flickering), or pass fewer or more ions per opening than they usually do. Their function is also influenced by drugs, changes in fluidity of the membrane, temperature, electrolyte balance in the milieu, and other physical and chemical factors.38,39 Receptor channels are dynamic structures that are capable of a wide variety of interactions with drugs and of entering a wide variety of current-passing states. All these influences on channel activity are ultimately reflected in the strength or weakness of neuromuscular transmission and contraction of a muscle.

Publisher Summary

Neurotransmission is the fundamental process that drives information transfer between neurons and their targets. It regulates both excitatory and inhibitory functions in the central nervous system (CNS), underlies sensory processing, and regulates autonomic and motor outputs in species ranging from small invertebrates to highly evolved mammals. Modulation of synaptic transmission is believed to drive cognitive processes such as learning and memory. Neurotransmission occurs at specialized regions between neurons and their targets, called the synapse. The synapse is a highly specialized contact between a presynaptic and a postsynaptic cell built to transmit information with high fidelity. Synaptic transmission is mediated by repeated cycles of exocytosis of neurotransmitters followed by endocytosis of synaptic vesicles (SVs) at nerve terminals. This chapter reviews neuronal exocytosis and the recent advances in the molecular and cell biological understanding of this process. Neuronal exocytosis is the final step in a cycle that leads to information transfer across synapses. The vesicle cycle is highly regulated via essential vesicular and plasma membrane proteins that mediate the various steps including neurotransmitter loading into vesicles, docking, priming, calcium sensing, exocytosis, and recycling. Despite the progress in understanding the roles of individual proteins using physiological, biochemical, and genetic methods, much remains unknown regarding the sequence and kinetics of protein–protein interactions that drive vesicle recycling.

Abnormalities of neurotransmission

Since the early 1970s, a variety of hypotheses have suggested that HE is caused by disordered neurotransmission. Although early hypotheses related to putative false neurotransmitters were disproved, there is still effort in this direction.

As a result of the false neurotransmitter hypothesis, it was shown that the ratio of plasma amino acids (valine + leucine + isoleucine) to (phenylalanine + tyrosine) was abnormal in encephalopathic patients, leading to the development of branched chain amino acid (BCAA) solutions designed to normalize this ratio, which are now commercially available. A meta-analysis of studies analyzing the effects oforal or intravenous application of BCAA came to the conclusion that BCAAs have a beneficial effect upon HE, but not upon mortality in patients with liver cirrhosis (Gluud et al., 2017). Substantial effort has been focused on potential abnormalities of the GABA–benzodiazepine complex. Initial attention was directed at GABA itself. However, early reports that GABA concentrations were elevated in patients with encephalopathy have been disproved. Still, a number of anecdotal reports have described dramatic improvements in patients after they were given flumazenil—a benzodiazepine antagonist; very low concentrations of benzodiazepines and their metabolites may be found in blood and CSF of patients with encephalopathy. In controlled studies, patients given flumazenil are more likely to improve than those given placebo. It is unclear whether benzodiazepine displacement is the mechanism because these patients do not usually have clinically significant blood levels of benzodiazepines.

More recent theories have linked the presence of increased expression of peripheral types of benzodiazepine receptors (currently called translocator protein [TSPO]) to HE. These receptors are found on mitochondrial membranes and are implicated in intermediary metabolism and neurosteroid synthesis. Hyperammonemia causes an increase in TSPO and thereby stimulates the production of neurosteroids such as allopregnanolone, which activates GABA and benzodiazepine receptor sites of the GABA-A receptor, resulting in an increase in GABA-ergic tone in the brain.

In addition, there are significant alterations in cerebral serotonin and dopamine metabolism and a reduction in postsynaptic glutamate receptors of theN-methyl-d-aspartate type. Thus, there is a substantial interest in the potential role of neurotransmitters in the pathogenesis of HE. As of yet, there is no unifying hypothesis and no rational therapeutic approach based on altering neurotransmission.

Abstract

Chemical neurotransmission requires the coordination of pre- and postsynaptic events across the nervous system. Following action potential generation, large quantities of neurotransmitters are released into the synaptic compartment. For most neurotransmitters, the response they elicit within the postsynaptic cell occurs through two general mechanisms: a fast mechanism mediated by ionotropic receptors and a slow mechanism mediated by metabotropic receptors. These receptors, present at pre- and postsynaptic sites, are what dictate the actions of the neurotransmitter. Receptor location, stability in the plasma membrane, their subunit composition, and posttranslational modifications, all dynamically regulate the response due to neurotransmitter binding throughout development. This chapter reviews the basic steps and concepts involved in neurotransmission to lay a foundation for understanding and interpreting the actions of well-known neurotransmitters. We also include a section on the role of glia in the regulation of chemical neurotransmission in the brain.

6.4 Obsessive–Compulsive Disorder

NMDARMN is ubiquitous and involved in many fundamental functions of CNS including cognition, rewarding, motor function, etc. Its modulation can certainly offer beneficial outcomes in the circuitries involving these functions. In hindsight, although NMDA enhancement treatment is particularly relevant to schizophrenia given that the NMDAR antagonists generate “schizophrenia-like” symptoms, it is not surprising that the treatment is also beneficial for a variety of CNS disorders. In fact, the efficacy of NMDA treatment is not limited to schizophrenia and depression. The efficacy had also been shown in improving the symptoms of obsessive–compulsive disorder (OCD) by sarcosine treatment (Wu, Tang, Lane, Tsai, & Tsai, 2011), which is consistent with the involvement of glutamatergic neurotransmission in the circuitry of OCD (Pittenger, 2015).

OCD is a common psychiatric disorder, affecting 2–3% of the population. There is an alteration of the glutamate receptor-mediated neurotransmission in OCD; glutamate levels estimated by magnetic resonant spectroscopy are significantly elevated in the caudate but significantly reduced in the anterior cingulate cortex in drug-naïve OCD patients (Rosenberg et al., 2000). NMDAR antagonists such as AP5, ketamine, and phencyclidine cause a pathological increase of glutamate (Liu & Moghaddam, 1995), which is reversible by either NMDAR agonists, such as glycine, or nonspecific glutamate inhibitors, such as lamotrigine, in both animal and human preclinical studies (Anand et al., 2000). Genetic association studies of OCD have identified two susceptibility genes that are vital for glutamatergic neurotransmission: a glutamate transporter gene, SLC1A1 (Wendland et al., 2009), and the NR2B subunit gene (Arnold et al., 2004). Two recent transgenic animal models of alteration of NMDA function demonstrate compulsive behavior: SAPAP3 knockout mouse, which has striatum-specific alternation of NMDAR subunit composition (Welch et al., 2007), and G72/G30, a DAAO regulator, transgenic mouse (Otte et al., 2009). In addition, the NMDAR antagonist MK-801 exacerbates repetitive climbing and leaping behavior in a transgenic D1CT-7 mouse model of comorbid Tourette's syndrome and OCD (McGrath, Campbell, Parks, & Burton, 2000). Therefore, potentiation of NMDA function may correct the maladaptive NMDARMN that underlies OCD behaviors. On the other hand, memantine, but not the AMPA antagonist and riluzole, significantly inhibits murine marble-burying behavior, a potential animal model for OCD (Egashira et al., 2008). Infusion of ketamine renders a much better treatment response rate that can persist for at least 1 week (Rodriguez et al., 2013). Overall, this suggests that both agonists and antagonists of NMDAR can improve the symptoms of OCD, similar to depression.

Diverse NMDAR agonists and antagonists have differential regionally and temporally effects on the fronto-subcortical circuitry (FSC), given that: (1) NMDARs are composed of different subunits and differentially expressed both regionally and temporally during development (Monyer et al., 1992); and (2) alternative composition of the NMDAR subunits results in functional diversity of the ion channel (Chapman, Keefe, & Wilcox, 2003). Polymorphism of NR2B subunit gene GRIN2B has been involved with OCD, and the SAPAP3 gene-deleted mouse has a decreased NR2A/NR2B ratio in the striatum with significantly reduced field EPSCs and expresses OCD-related phenotype (Welch et al., 2007). Synaptic processing of excitatory input is also different in the ventromedial vs dorsolateral striatum (Chapman et al., 2003). Given the molecular, anatomical, developmental, and physiology complexities of NMDARMN, the “direct” and “indirect” pathways unbalanced hypothesis may explain the clinical and preclinical reports that both NMDAR agonists and uncompetitive NMDAR antagonists are efficacious for OCD.

OCD is another example that the bivalent NMDA treatments—enhancement or inhibition—can improve the symptoms involved. This is consistent with the hypothesis that the imbalanced glutamatergic tone, rather than being too high or too low, in FSC is associated with the behavioral manifestations of OCD (Saxena, Brody, Schwartz, & Baxter, 1998). Therefore, enhancing NMDARMN in FSC to achieve a balanced tone between direct and indirect pathways can be beneficial for patients with OCD. For example, sarcosine may exert its therapeutic effect through modulation of the imbalanced “direct” vs “indirect” pathways in FSC implicated in OCD (Rosenberg et al., 2000; Saxena et al., 1998; Wendland et al., 2009).

In 3 case reports and 1 small open-label trial, memantine, a weak noncompetitive NMDAR antagonist, proved to be efficacious as add-on treatment to resistant OCD (Hezel, Beattie, & Stewart, 2009; Hosenbocus & Chahal, 2013). Following a new paradigm using d-cycloserine, a partial agonist acting on the NMDAR co-agonist site, to facilitate exposure therapy for anxiety disorders (Rothbaum, 2008), 2 trials on OCD found advantage of adjunctive d-cycloserine treatment over a placebo (Andersson et al., 2015; Wilhelm et al., 2008).

d-Cycloserine is critically implicated in fear learning and fear extinction in both rodents and humans. Specifically, NMDA antagonists block fear extinction in rodents, whereas NMDA agonists enhance fear extinction. Long-term administration of different classes of antidepressants downregulate NMDAR subunits as well as the co-agonist site to which d-cycloserine binds, while imipramine abolishes the normal d-cycloserine-induced facilitation of fear extinction in rats, and the magnitude of d-cycloserine-facilitated virtual reality therapy for posttraumatic stress disorder is smaller in patients receiving psychotropic medications such as SRIs.

Consistent with the interaction of d-cycloserine with the antidepressants, d-cycloserine does not augment the effects of cognitive behavior therapy (CBT) but shows a significant interaction with the antidepressant medication. This suggests that antidepressants may be interchangeable with d-cycloserine to block its facilitating effect on fear extinction (Andersson et al., 2015). Therefore, the use of d-cycloserine may be a promising strategy to improve OCD symptoms, but should be limited only to antidepressant-free patients. Interestingly, SRI and the NMDA agent may act on the same neuronal substrate to improve the symptoms of OCD. Pharmacotherapy with SRIs or exposure-based psychotherapy improves only 40–60% of patients. Therefore, it is important to explore whether there is a treatment path that can address the different subpopulations that respond to SRI (CBT) vs NMDA agents.

In our sarcosine trial for patients with OCD, the effect of sarcosine occurs with doses lower than that for the patients with schizophrenia (Lane et al., 2005; Tsai, Lane, et al., 2004). The patients receive an average sarcosine dose of 1520 ± 549 mg/day. The mean initial and final scores of Yale Brown Obsessive Compulsive (Y-BOCS) and Hamilton Anxiety Scales decrease significantly over time. It is interesting that drug-naïve subjects respond more favorably than subjects who have been exposed to SRI treatment before; 4 (50%) drug-naïve subjects are rated as responders (symptom reduction range: 46.7–69.2%). The decrease in Y-BOCS scores is also greater in the drug-naïve than in the nonnaïve groups. Our findings are consistent with a mutually exclusive effect between SRIs and the NMDA agents (Andersson et al., 2015). In addition, the response is quick in the first two weeks by sarcosine treatment. Five of the 8 final responders met criteria of response within 2–4 weeks of sarcosine treatment, which is quicker than the onset of therapeutic response with SRIs.

Sarcosine also significantly improved OCD symptoms in 27% (3 of 11) of the add-on group patients. NMDAR antagonists have direct or indirect effects on monoamine systems by blocking NMDAR located on glutamatergic and GABAergic neurons (Egashira et al., 2008). Alternatively, GlyT-1 inhibitors may not only restore glutamatergic influence on raphe serotonergic neurons, leading to normalized GABA/glutamate balance in the cerebral cortex and overall inhibition on prefrontal neuronal circuitry, but also directly exert their effect on the pathways in the FSC.

Taken together, we find that the treatment effects of SRI and NMDAR agents are not additive, but almost “mutually exclusive.” It is possible that the therapeutic effect of SRI or sarcosine for OCD converges on FSC by diminishing ventromedial basal ganglia activity relative to that in the dorsolateral system, or reducing glutamatergic hyperactivities in the frontal cortex. Either SRI or NMDAR agents alone may reach a therapeutic ceiling, and combination treatment cannot bring more improvement such as in the SRI nonnaïve group. This can be understood provided that: (1) the cortico-raphe glutamatergic and raphe-cortical serotonergic projections may form a loop by which excitatory input signals are converted into inhibitory output projecting back to the cerebral cortex; (2) chronic administration of SRI leads to adaptive expression of NMDAR subunits and region-specific change of NMDARMN in CNS; and (3) serotonin may exert dual actions by stimulating 5-HT 2A receptors on GABA interneurons and 5-HT1A receptors on glutamatergic neurons in the prefrontal cortex, thus indirectly inhibiting the primary glutamatergic output to the ventral striatum.

Neurotransmission in the nervous system is initiated at presynaptic terminals by fusion of synaptic vesicles with the plasma membrane and subsequent exocytic release of chemical transmitters. There are multiple methods to detect neurotransmitter release from nerve terminals. Most commonly employed methods monitor actions of released chemical substances on postsynaptic receptors or artificial substrates such as carbon fibers. These methods are closest to the physiological setting because they have a rapid time resolution and they measure the action of the endogenous neurotransmitters rather than the signals emitted by exogenous probes. However, postsynaptic receptors only indirectly report neurotransmitter release in a form modified by the properties of receptors, which are often nonlinear detectors of released substances. In the past decade, in addition to electrophysiological and biochemical methods, several fluorescence imaging modalities have been introduced which report synaptic vesicle fusion, endocytosis, and recycling. These methods either take advantage of styryl dyes that can be loaded into recycling vesicles or exogenous expression of synaptic vesicle proteins tagged with a pH-sensitive green fluorescent protein variant at regions facing the vesicle lumen. This article provides an overview of these methods, with emphasis on their relative strengths and weaknesses, and it discusses the types of information one can obtain from them.

Decreased Neural and Neuromuscular Transmission

Neural transmission to the respiratory muscles may be interrupted in phrenic nerve or spinal cord transection. Transmission may also be impaired in phrenic nerve injury (thermal, hypoxic, or traction injury during cardiac surgery52), demyelinating diseases, immunologic conditions (e.g., Guillain-Barré syndrome,50,53 multiple sclerosis) or toxin-induced disorders (e.g., diphtheria), or in diseases affecting the lower motoneurons, whether infectious (e.g., poliomyelitis) or degenerative (e.g., amyotrophic lateral sclerosis). Various other neuropathies could also be included, but are rare enough not to merit special mention in this context.

Neuromuscular transmission, in turn, may be impaired by toxins (e.g., botulism that inhibits presynaptic acetylcholine release), an episode of myasthenia gravis, and drugs (organophosphate poisoning, aminoglycosides, and, especially, neuromuscular blockers).50,54 Critical illness polyneuropathy is a syndrome of prolonged muscle weakness or paralysis typically manifesting as failure to wean from mechanical ventilation.50,55 The cause is unknown, although sepsis, multiple organ failure, shock, hypoxia, medications, and prolonged use of neuromuscular blocking agents have been implicated.56 Electrophysiologic studies reveal abnormalities primarily characterized by axonal degeneration.50,56,57 Although critical illness polyneuropathy is usually improved in parallel with the underlying disease, weaning is difficult, and the mortality rate is high in these patients.50,56