Edited by: Chao Deng, University of Wollongong, Australia
Reviewed by: Hermona Soreq, The Hebrew University of Jerusalem, Israel; Katerina Zavitsanou, University New South Wales, Australia
*Correspondence: Elizabeth Scarr, Department of Psychiatry, Melbourne Brain Centre, The University of Melbourne, Kenneth Myer Building, Melbourne, VIC 3010, Australia. e-mail:
This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in other forums, provided the original authors and source are credited and subject to any copyright notices concerning any third-party graphics etc.
Acetylcholine has been implicated in both the pathophysiology and treatment of a number of psychiatric disorders, with most of the data related to its role and therapeutic potential focusing on schizophrenia. However, there is little thought given to the consequences of the documented changes in the cholinergic system and how they may affect the functioning of the brain. This review looks at the cholinergic system and its interactions with the intrinsic neurotransmitters glutamate and gamma-amino butyric acid as well as those with the projection neurotransmitters most implicated in the pathophysiologies of psychiatric disorders; dopamine and serotonin. In addition, with the recent focus on the role of factors normally associated with inflammation in the pathophysiologies of psychiatric disorders, links between the cholinergic system and these factors will also be examined. These interfaces are put into context, primarily for schizophrenia, by looking at the changes in each of these systems in the disorder and exploring, theoretically, whether the changes are interconnected with those seen in the cholinergic system. Thus, this review will provide a comprehensive overview of the connectivity between the cholinergic system and some of the major areas of research into the pathophysiologies of psychiatric disorders, resulting in a critical appraisal of the potential outcomes of a dysregulated central cholinergic system.
The central cholinergic system has been implicated in the pathophysiology of schizophrenia (Raedler et al.,
In the human central nervous system, the cholinergic system has evolved into a complex network with three principle components,(i) projections from nuclei of the basal forebrain; these include the medial septal nucleus, the nucleus basalis of Meynert, the vertical nucleus of the diagonal band and the horizontal limb of the diagonal band nucleus, which innervate the hippocampus, most cortical regions and some subcortical nuclei, (ii) the pedunculopontine-lateral dorsal tegmental projections from the brainstem to the thalamus, midbrain and other brainstem regions and (iii) interneurons in the striatum (most abundant) and the nucleus accumbens (Everitt and Robbins,
The cholinergic system has been proposed to contribute to the pathophysiology of schizophrenia as a result of either an imbalance between central cholinergic and dopaminergic systems (Tandon and Greden,
The perturbations of the central cholinergic system have been thoroughly reviewed previously (Raedler et al., The most reproduced finding is a widespread decrease in levels of muscarinic receptors in the brains of people with schizophrenia, this has been replicated in four separate postmortem collections (Mancama et al., Epibatidine binding, predominantly to the α4β2 nicotinic receptor, has been reported to be increased in people with schizophrenia (Martin-Ruiz et al., The most investigated nicotinic receptor is the α7 nicotinic receptor which is associated with a sensory gating deficit present in people with schizophrenia (Adler et al.,
The first indication that the cholinergic system was involved in the pathophysiology of mood disorders came from the development of depressive symptoms in people who had been exposed to cholinesterase inhibitors (Rowntree et al.,
For the purpose of this review, the term intrinsic has been used to describe neurotransmitters that predominantly act locally throughout the central nervous system, although they may have some neurons that project across different brain regions. These neurotransmitters include the excitatory amino acid glutamate and the inhibitory amino acid gamma-amino butyric acid (GABA).
Glutamate is the most abundant excitatory neurotransmitter in the human central nervous system, the effects of which are mediated via two classes of receptors; ionotropic[N-methyl-D-aspartate (NMDA), 2-amino-3-(3-hydroxy-5-methyl-isoxazol-4-yl)propanoic acid (AMPA), and kainate receptors] and metabotropic (mGluR1−8) receptors (Traynelis et al.,
Magnetic resonance spectroscopy studies have reported elevated glutamate levels in the hippocampus and prefrontal cortex of patients with schizophrenia (van Elst et al.,
While small decreases in AMPA receptor radio ligand binding are reported in CA2 of the hippocampus (Gao et al.,
With regards to the kainate receptor, a reduction in radioligand binding density and a reduction in GluR5 mRNA expression have been reported in the prefrontal cortex from people with schizophrenia (Scarr et al.,
There is increasing awareness of the potential for targeting metabotropic glutamate receptors as modulators of glutamate release, ionotropic receptor response, and glutamatergic signal transduction, in the treatment of schizophrenia (Vinson and Conn,
Acetylcholine has been shown to modulate glutamatergic excitatory postsynaptic potentials in several brain regions (Li and Pan,
In the hippocampus, M1 and M3 have been shown to potentiate kainate receptor currents, increasing mossy fiber axon excitability. This modulation is subunit dependant, for example; muscarinic receptor activation potentiates heteromeric GluR6/KA1 and GluR6/KA2 receptors, but not homomeric GluR6 receptors (Benueniste et al.,
Glutamatergic signaling has been shown to modulate acetylcholine release, predominantly via the ionotropic receptors. For instance, cortical microinjections of the NMDA receptor antagonist 3-(2-Carboxypiperazin-4-yl)propyl-1-phosphonic acid (CCP) increased acetylcholine release in the nucleus accumbens, an effect blocked by local perfusions of both CCP and the AMPA receptor antagonist 6,7-Dinitroquinoxaline-2,3-dione (DNQX) (Del Arco et al.,
The respective modulation of glutamate and acetylcholine release by cholinergic and glutamatergic pathways respectively depend on the co-expression of appropriate receptors within neurons and their synaptic connections. Microdialysis of AMPA into rat cortex facilitated acetylcholine release in the parietal and prefrontal cortices, an effect attenuated by DNQX (Nelson et al.,
Studies have also shown that ventral tegmental presynaptic metabotropic glutamate and muscarinic receptors preferentially inhibit the NMDA mediated component of synaptic transmission (Zheng and Johnson,
The dual role of the cholinergic system, activating and inhibiting glutamatergic signaling, presents challenges in predicting the effects of (i) the M1 deficits associated with and (ii) the NMDA receptor hypofunction predicted in schizophrenia. However, animal studies have shown that inhibitory avoidance memory consolidation can be repressed by co-administration of muscarinic and NMDA antagonists to the ventral tegmentum, at doses that were ineffective when used alone (Mahmoodi et al.,
Importantly, the processes governing acetylcholine and glutamate release in turn regulate and are regulated by additional neurotransmitters. For example, stimulating nicotinic receptors reduces AMPA -evoked synaptosomal dopamine overflow (Grilli et al.,
GABA is the major central inhibitory neurotransmitter, in mammals 25–50% of central synapses utilize GABA (Petroff and Rothman,
There is strong evidence to support the theory that schizophrenia is associated with deficits in GABAergic neurotransmission [see (Blum and Mann,
Glutamic acid decarboxylase (GAD) 67 is essential for GABA synthesis and is used as a marker for GABAergic cells. Cortical expression of mRNA for both GAD67 and the GABA transporter, GAT1, are reported to be decreased in tissue from people with schizophrenia (Volk et al.,
The striatum is the major input structure of the basal ganglia and has been implicated in the pathophysiology of schizophrenia (Lester et al.,
Nicotinic receptors, on the other hand, appear to facilitate GABA release (Lena et al.,
To obtain insight into GABA-acetylcholine interactions, a number of studies investigated the effects of GABA agonists, such as; muscimol, progabide, SL75102, δ-aminovaleric acid, and 2-pyrrolidone, on acetylcholine levels. In a number of brain regions, low doses of GABA agonists increased acetylcholine levels (Scatton and Bartholini,
The number of striatal cholinergic interneurons has been shown to be decreased in people with schizophrenia (Holt et al.,
The systems considered in this section are neurotransmitter systems whose neurons arise from discreet brain structures and project to distal regions of the brain, affecting the activity of the intrinsic neurotransmitters in those regions. The choice of projection systems to be included in this review was driven, in part, by the known pathophysiologies of schizophrenia, and therefore focuses on the dopaminergic and serotonergic systems.
Dopaminergic cells are found almost exclusively in the substantia nigra (SN) and ventral tegmental area (VTA), forming four major dopaminergic pathways in the mammalian brain, these are the (i) mesolimbic, (ii) mesocortical, (iii) nigrostriatal, and (iv) tuberoinfundibular pathways (Albanese et al.,
There are two types of G-protein coupled dopamine receptors, which are widely distributed centrally; D1-like receptors (D1 and D5), which couple to Gs proteins and stimulate cAMP production and D2-like receptors (D2,3, and 4), which couple to Gi/o proteins and either have no effect on or inhibit cAMP (Schetz,
In both Lewy Body dementia and Alzheimer's disease, where there is a loss of cholinergic neurons, patients have a loss of cognitive function and neuropsychiatric symptoms. Although both groups have similar levels of delusions, anxiety, and depression, patients with mild Lewy Body dementia have more visual and auditory hallucinations than patients with Alzheimer's disease (Auning et al.,
The dopaminergic system has long been considered a major component of schizophrenia pathophysiology (Carlsson,
The classic hypothesis for schizophrenia proposed that hyperactivity of dopaminergic transmission was responsible for the positive symptoms, however, the awareness of enduring negative symptoms and cognitive deficits, with their resistance to D2 antagonism, led to a reformulation of this hypothesis. Functional imaging studies suggested that altered functionality of the prefrontal cortex [PFC; see (Knable and Weinberger,
Blocking the D2 receptor reduces positive symptoms in people with schizophrenia (Carlsson,
The apparent inconsistencies between dopaminergic systems has been resolved by studies showing reciprocal and opposite regulation between the cortical and subcortical systems (Pycock et al.,
The striatum is densely innervated by tonically active cholinergic interneurons (Butcher and Woolf,
All muscarinic receptors are expressed in the striatum, suggesting all have the potential to modulate dopamine release (Weiner et al.,
A more direct approach to delineating the muscarinic-dopaminergic interactions, came from studies on M1-5 receptor deficient mice (Hamilton et al.,
The initial association between nicotine addiction and dopaminergic striatal signaling suggested the existence of a nicotinic-dopaminergic interaction (see Corrigall,
Systemic nicotine has been shown to increase dopamine release in the mesolimbic (Imperato et al.,
Subsequent experiments demonstrated that striatal nicotinic control of dopamine release is mediated predominantly by receptors containing the β2 subunit (Zhou et al.,
Whilst it is apparent that nicotine can stimulate dopamine release (Marshall et al.,
Early studies proposed that dopamine inhibits acetylcholine transmission, with the development of more specific ligands and newer techniques, it is now apparent that blockade of D1 receptors reduces acetylcholine release whilst activation stimulates release (Bertorelli and Consolo,
The lack of consensus regarding the status of dopaminergic receptors in schizophrenia makes it difficult to speculate as to whether they may impact on cholinergic function. Conversely, there is strong evidence to suggest that nicotine stimulates dopamine release, with receptors containing a β2 subunit playing a significant role (Zhou et al.,
The central serotonergic system is widespread, innervating nearly all brain regions [see (Hornung,
The large family of serotonergic receptors gives the neurotransmitter an even greater functional capacity than is conferred by the diffuse serotonergic projections. In this review 5-ht5a, and 5-ht1e receptors are not considered because of their lack of a robust signal in native tissue. Of the remaining 11 receptors, most are metabotropic; 5-HT4, 6&7 receptors canonically couple to Gs, increasing levels of cAMP; the 5-HT1 receptors canonically couple with Gi/o and reduce levels of cAMP whilst the 5-HT2 receptors canonically couple to Gq/11 and increase inositol phosphate hydrolysis. The 5-HT3 receptor is a pentameric ligand-gated cation channel; 5-HT3A subunits can form functional homomeric receptors whilst the 5-HT3B, C, D &E subunits form functional heteromeric receptors with 5-HT3A subunits (Barnes et al.,
The first suggestion that serotonin might play a role in the pathophysiology of schizophrenia arose from the observation that lysergic acid diethylamide (LSD), a serotonergic agonist, caused psychoses which were proposed to have similarities to the positive symptoms of schizophrenia (Wooley and Shaw,
A number of studies looked at the major serotonin metabolite, 5-hydroxyindoleacetic acid, in cerebrospinal fluid from people with schizophrenia; the results are inconclusive, with reports of increases, decreases, and no change (see Abi-Dargham et al.,
Projections from the medial septal nucleus and the diagonal band of Broca innervate the raphe nuclei (Kalén and Wiklund,
Acetylcholine has been reported to stimulate serotonin release in the caudate via nicotinic (Becquet et al.,
The basal forebrain receives afferents from numerous systems, including serotonergic fibers from the dorsal raphe (Semba et al.,
5-HT1A agonists were also shown to facilitate acetylcholine release in the cortex (Bianchi et al.,
The most reproduced finding for the cholinergic system in schizophrenia is a decrease in central muscarinic receptors, in particular the M1, in people with the disorder [see (Scarr and Dean,
Another reproducible finding in schizophrenia is decreased expression of nicotinic receptors (Martin-Ruiz et al.,
The most widely reproduced finding for the serotonergic system in schizophrenia is the decrease in cortical 5-HT2A receptors (see Dean,
In summary, although the data that best fits with the current knowledge regarding the pathophysiology of schizophrenia is that reduced levels of muscarinic receptors contribute to a reduced inhibition of the serotonergic system and a subsequent decrease in post-synaptic serotonergic receptors, as seen with 5-HT2A and 5-HT7 receptors, this is overly simplistic given that these are not the only post-synaptic serotonergic receptors; the data regarding 5-HT1A receptors in schizophrenia is inconclusive but there are no reports of decreased levels in schizophrenia. Studies that look at all components of both systems in a single cohort are required to reach definitive conclusions about the interactions of the two systems in schizophrenia.
Historically, the focus of the pathophysiology of psychiatric disorders has been on markers of neurotransmission. There is, however, growing data suggesting that molecules traditionally associated with a response to inflammation or infection are abnormally expressed in people with psychiatric disorders (see Potvin et al.,
A significant body of literature suggests there are physiologically relevant interactions between the cholinergic system (neuronal and non-neuronal) and inflammatory/immune pathways in the periphery (Bencherif et al.,
The hypothesis that the cholinergic system is involved in modulating central inflammatory and/or immune system is perhaps best tested at the systems level. With this regard, it is significant that microglial cells, which are widely viewed as resident macrophages in the central nervous system, express α7 nicotinic receptors, activation of which attenuates the pro-inflammatory responses in cultured microglial cells (Carnevale et al.,
Additional evidence suggests that the ability of the cholinergic system to modulate the activity of microglia may be multifaceted; donepezil, a reversible, non-competitive cholinesterase inhibitor, has been shown to attenuate microglial production of nitric oxide and TNF, possibly by inhibiting the canonical inflammatory NF-κB signaling (Hwang et al.,
As with all biological systems, there seems to be the potential for a two-way interaction between the inflammatory/immune systems and the nicotinic system. For example, it has been shown that IL-1β and TNF can alter nicotinic receptor sub-unit assembly (Gahring et al.,
The physiological outcomes of the cholinergic system are usually determined by the balance between nicotinic and muscarinic receptors (Decker and McGaugh,
Somewhat surprisingly, it has been shown that activating central, but not peripheral, M2 receptors modulates levels of TNF in serum (Pavlov et al.,
Given the clear relationship between the cholinergic and the inflammation/immune systems it remains to conceptualize a mechanism by which this can occur centrally. There are a number of options, one of which is that these changes are the result of interactions between the M2 and α7 receptors and inflammation/immune pathways within microglia. Significantly, it has been shown that carbachol, a pan-muscarinic receptor agonist, caused a rapid influx of calcium into microglia (Zhang et al.,
To briefly summarise the potential interactions that might occur in psychiatric disorders, in schizophrenia where decreases in M1 are widely reported, these could result in reduced kainate function, which in turn could contribute to a glutamatergic hypofunction. The reduced α7 nicotinic capacity reported to exist in schizophrenia would result in reduced GABA efflux, with the potential to cause increased levels of postsynaptic GABAergic, such as GABAA receptors. An expected consequence of the increased levels of β2 containing nicotinic receptors and the decreased levels of M1 and/or M4 receptors is an increase in dopamine release, potentially contributing to the imbalance in dopaminergic systems proposed to exist in schizophrenia. With regards to the serotonergic system, the decreases in M1/M4 receptors seen in schizophrenia could cause an increase in serotonin release, which would cause the downregulation of postsynaptic receptors, including 5-HT2A. Conversely, if the small global increase in the 5-HT1A receptors is substantiated, this could affect the cholinergic system causing increased cholinergic release, a consequence of which might be the downregulation of postsynaptic cholinergic receptors, including the M1 and M4. Finally, it is possible that the dysregulation of molecules traditionally associated with inflammation/immune responses in psychiatric disorders centers around disrupted interactions between the central cholinergic system, mediated by M2 and α7 receptors and microglia.
It is evident from these brief overviews that a dysfunctional central cholinergic system can have far reaching consequences. A common theme in considering these interactions is that the regulatory mechanisms are two-way systems, often with a third implicated as an intermediary. Thus, even considering the interactions between two systems is overly simplistic, suggesting that a whole systems approach is necessary to fully understand the relationships between central systems that become unstable in psychiatric disorders.
In this review, we identified the most commonly replicated changes in neurochemical markers associated with psychiatric disorders and interpreted them in the light of basic research elucidating interactions between the cholinergic and other central neurotransmitter systems. Acetylcholine was chosen as the pivotal transmitter system because of the extent of its innervations and because it is a target of choice for many drug development strategies aimed at novel therapies for psychiatric disorders. For example, acetylcholinesterase inhibitor use has expanded from their initial role of improving cognitive impairment in dementias (Hollander et al.,
One caveat of this review is that most of the data related to the neurochemical changes in psychiatric disorders has arisen from postmortem studies. Therefore, we cannot ascertain which of the chemical changes occurred first, hindering our attempts to construct a theory around these changes. Even if we could look at all of the markers detailed in this review in the same cohort of living people, whilst we would be able to confirm or disprove some of the proposed interactions, we still would not be able to determine the cause and effect relationship.
Furthermore, given the emphasis on the neurodevelopmental aspect of many of these disorders (Sigurdsson et al.,
The development of new technologies and our increasing understanding of the processes involved in the translation from gene sequence to active product also offer a number of new approaches that can be utilized to improve our knowledge regarding the interactions between central transmitter systems. For example, the relatively new field of optogenetics—where light can be used to activate specific neurons—offers great scope to activate specific receptors in tissue of interest and identify the consequences of that activation. This approach will be of particular use in determining which receptors are involved in the cross-talk between transmitter systems, thereby circumventing the problems associated with using drugs that, although they have a high affinity for a particular receptor often have the capacity to stimulate or inhibit the actions of other receptors.
What was once a “simple” process of a gene being transcribed into RNA which was then translated into the corresponding protein is gradually being unraveled to reveal a far more complex series of events than previously imagined. We now know that factors such as gene methylation and histone modification (epigenetics) can determine whether or not a gene can be transcribed. Assuming the RNA is generated, the next step in the process can also be regulated, this time by microRNAs (miRNAs) which have the ability to block the translation of mRNA into proteins. Therefore, these factors also have to be taken into account when considering the interactions between central neurotransmitters, particularly since both epigenetics and miRNAs have been implicated in psychiatric disorders. For example, does the activation of one system affect the prevalence or type of epigenetic markers on the genes that encode components other systems? Will such changes in turn affect the fundamental regulation of expression for that gene? Does transmitter X affect the expression of particular miRNAs? If so, which of the myriad of theoretical interactions ascribed to each miRNA actually occur physiologically and of those, which are relevant to the process under investigation? The involvement of both miRNAs (Dwivedi,
Andrew Stuart Gibbons, Madhara Udawela, and Jaclyn Neoreport no competing interests. The following authors have previously received remuneration: Elizabeth Scarrreceived honorarium from Astra-Zeneca and travel support from GlaxoSmithKline (GSK). Brian Deanreceived travel support from GSK, honorarium from Pfizer, Eli Lilly, and Merck Sharp and Dohme (MSD).
This work was supported by the National Medical and Health Research Council (Senior Research Fellowship #APP1002240 to Brian Dean), the Australian Research Council (Future Fellowship FT100100689 to Elizabeth Scarr) as well as Operational Infrastructure Support (OIS) from the Victorian State Government.