Calcium signalling and psychiatric disease: bipolar disorder and schizophrenia
- First Online:
- Cite this article as:
- Berridge, M.J. Cell Tissue Res (2014) 357: 477. doi:10.1007/s00441-014-1806-z
- 870 Views
Neurons have highly developed Ca2+ signalling systems responsible for regulating many neural functions such as the generation of brain rhythms, information processing and the changes in synaptic plasticity that underpins learning and memory. The signalling mechanisms that regulate neuronal excitability are particularly important for processes such as sensory perception, cognition and consciousness. The Ca2+ signalling pathway is a key component of the mechanisms responsible for regulating neuronal excitability, information processing and cognition. Alterations in gene transcription are particularly important as they result in subtle alterations in the neuronal signalling mechanisms that have been implicated in many neural diseases. In particular, dysregulation of the Ca2+ signalling pathway has been implicated in the development of some of the major psychiatric diseases such as bipolar disorder (BPD) and schizophrenia.
KeywordsCalciumInositol trisphosphateSchizophreniaBipolar diseaseReactive oxygen species
Calcium signalling has a major role in regulating various neuronal processes such as transmitter release from presynaptic endings and several postsynaptic processes including the regulation of neuronal excitability and the changes in synaptic plasticity responsible for learning and memory. Many of the major neural diseases have been linked to alterations in many of these Ca2+-sensitive neural processes. In Alzheimer’s disease, for example, the characteristic progression of this disease may arise through dysregulation of Ca2+ signalling pathways (Lidow 2003; Berridge 2012a, 2012b, 2014). In the case of psychiatric diseases, such as bipolar disorder (BPD) and schizophrenia, evidence is increasing that subtle alterations occur in specific Ca2+ signalling pathways (Berridge 2012a, 2012b; Giegling et al. 2010). Such alterations seem to result in a phenotypic switch in the excitatory or inhibitory neurons, which has a major impact on the mechanisms that generate the neural rhythms that are essential for brain function. Indeed, increasing emphasis is being laid on an integrative analysis of neural circuitry as a basis for understanding schizophrenia (Lisman et al. 2008; Lewis and Sweet 2009; Lisman 2012). The regulation of excitability, which depends on the tonic excitatory drive, might be of particular importance as it is responsible for controlling the neural brain rhythms that are essential for both information processing and cognition (Berridge 2012b, 2014). In this review, I will argue that alterations in the signalling mechanisms responsible for Ca2+-dependent neuronal gene transcription and the generation of the tonic excitatory drive might play important roles in psychiatric diseases such as BPD and schizophrenia.
Neuronal basis of brain rhythms
The different oscillatory modes that occur during the sleep/wake cycle are regulated by an interaction between the ascending arousal system and the sleep-inducing system (Fig. 1). The onset of sleep is initiated by neurons located in the preoptic area of the brain, such as the ventrolateral preoptic (VLPO) neurons that release gamma-aminobutyric acid (GABA) to inhibit the activity of the neurons that constitute the arousal system. Waking behaviour and consciousness is controlled by a large number of neurons that constitute the ascending arousal system (Pace-Schott and Hobson 2002; Saper et al. 2005; Schwartz and Roth 2008; Datta 2010; Edlow et al. 2012). This ascending arousal system consists in an array of neurons distributed in various brain regions such as orexinergic neurons in the lateral hypothalamus; cholinergic neurons in the basal forebrain the pedunculopontine tegmental nucleus and the laterodorsal tegmental nucleus; serotonergic neurons in the dorsal raphe; histaminergic neurons in the tuberomammillary nucleus; noradrenergic neurons in the locus coeruleus (LC) and dopaminergic neurons in the ventral tegmental area (Pace-Schott and Hobson 2002; Datta 2010; Schwartz and Roth 2008). The neurotransmitters released by this arousal system act on neural circuits to regulate the sleep/wake cycle. The orexinergic neurons that release orexin are master regulators that act to stimulate the other arousal neurons (Alexandre et al. 2013). Recent optogenetic experiments have highlighted the way that various arousal neurons act to control the sleep/wake cycle (Touriño et al. 2013). For example, the optical activation of the orexin neurons during sleep greatly enhance the probability of awakening. Activation of the noradrenergic LC neurons is even more effective in that it causes immediate wakening. The central role of orexinergic neurons is also evident in the way that they integrate the action of a number of other sleep regulatory factors such as ghrelin, leptin, glucose and adenosine (Sakurai 2007). The sleep/wake cycle is further controlled by the circadian clock, which might account for the observation that psychiatric illnesses seem to be linked to disturbances in circadian rhythms and sleep (Harvey 2008; Jagannath et al. 2013; McClung 2013).
Tonic excitatory drive
The transmitters that are released from the arousal neurons act on receptors that are coupled to various signalling pathways to adjust the level of neuronal excitability responsible for regulating the transition of the various brain rhythms that characterise the sleep/wake cycle (Berridge 2012b, 2014). As the level of neuronal excitation can remain constant for considerable periods to maintain diverse behavioural states, the membrane depolarisation maintained by these signalling pathways will be referred to as the tonic excitatory drive. This variable tonic excitatory drive, which controls the hierarchy of rhythms with the lowest frequencies occurring during sleep that can then be switched to the higher frequency rhythms of the wake state, functions much like a rhythm rheostat in that it provides a continuously variable degree of excitation, much like an electrical resistor used to regulate current (Fig. 1). The membrane depolarisation responsible for the tonic excitatory drive is induced by various signalling mechanisms that either activate inward Na+ currents or close outward K+ currents (Berridge 2014). For example, hydrolysis of the phospholipid phosphatidylinositol 4,5-bisphosphate (PtdIns4,5P2) has two effects. First, it closes the KV7.2/KV7.3 channels responsible for the M current (IM; Suh and Hille 2002). Switching off this M current depolarises the membrane to increase neuronal activity. Secondly, hydrolysis of PtdIns4,5P2 to form inositol 1,4,5-trisphosphate (InsP3) releases Ca2+ that stimulates the Ca2+-activated non-selective cation (CAN) channel. The CAN channel can also be activated by Ca2+ entering through the voltage-operated Ca2+ (VOC) channel. Interestingly, mutations of the CACNA1C and CACNB2 genes, which encode the α1C and β2 subunits of such CaV1.2 L-type Ca2+ channels have been linked to psychiatric disorders such as BPD and schizophrenia (Ferreira et al. 2008; Sklar et al. 2008; Giegling et al. 2010; Cross-Disorder Group of the Psychiatric Genomics Consortium 2013). Noradrenaline and dopamine act through the cyclic AMP signalling pathway to enhance the activity of the hyperpolarising-activated cyclic nucleotide-gated (HCN) channel responsible for the depolarising Ih current.
An important feature of this tonic excitatory drive is that it is applied equally to both the excitatory and inhibitory neurons; this stimulation has to be finely balanced for proper brain function. I will argue later that alterations in this excitation-inhibition (E-I) balance may occur in psychiatric diseases such as BPD in which excessive excitation may be responsible for the manic phase, whereas depression may result from excessive inhibition. Another important consequence of these alterations in neuronal activity is their impact on neuronal gene transcription, which is likely to play a major role in both BPD and schizophrenia.
Neuronal gene transcription
The CREB/CBP complex has emerged as the prototypical transcriptional factor that is activated during neuronal activity. The activity of CREB is also tightly regulated by protein acetylation of both histones and various transcription factors. The acetylation of histones by histone acetyltransferases (HATs) such as p300 serves to remodel chromatin such that it becomes more accessible to transcription factors and cofactors (Fig. 2). This open structure of chromatin is reversed by histone deacetylases (HDACs). The HDACs, such as HDAC4, suppress nuclear gene transcription by deacetylating histones leading to chromatin condensation. The same Ca2+ signal that activates gene transcription also phosphorylates HDAC, which is then exported from the nucleus. Valproate, which is one of the drugs used to treat BPD, inhibits HDAC and thus acts to enhance gene transcription as discussed later.
Other transcription factors that are activated include serum response factor (SRF), ETS-like transcription factor-1 (Elk-1) and the nuclear factor of activated T cells (NFAT). Alterations in many of the genes that are regulated by CREB have been implicated in psychiatric diseases as described below.
Inositol depletion hypothesis of BP
The inositol depletion hypothesis (Berridge et al. 1982, 1989), which was developed initially to explain how Li+ acts to control BPD, has now been extended to provide new insights into the nature of the changes in neuronal activity that underlie this disorder. The new concept is that changes in the activity of either the excitatory or inhibitory neurons result in subtle alterations in the neuronal circuits that control behaviour. The basic idea is that the periodic switching between depression and mania, which is a characteristic feature of BPD (Salvadore et al. 2010), is caused by an alteration in the E-I balance described earlier. During the generation of brain rhythms, the excitatory and inhibitory neurons have to be activated equally (Fig. 1). In effect, the tonic excitatory drive that determines the pacemaker potential for each neuronal type must be balanced. The idea that BPD is caused by such an E-I imbalance is supported by the observation that many of the transmitters (e.g., acetylcholine [ACh], dopamine and 5-hydroxytryptamine) that have been linked to this disease activate the tonic excitatory drive (Fig. 1) by using a variety of signalling mechanisms (Berridge 2012b, 2014). For example, ACh acts through M1 receptors to stimulate PtdIns4,5P2 hydrolysis, which contributes to depolarisation by decreasing the M current, while promoting the Ca2+-activated non-selective cation current (as described earlier). Dopamine, which acts through the cyclic AMP signalling pathway to activate the HCN current, also has a marked effect on mood. In addition, variants in the DAT1 gene, which codes for the dopamine transporter (DAT1), increase the risk of BPD (Greenwood et al. 2006). Drugs such as haloperidol that decreases dopaminergic transmission are antimanic. Finally, some of the most effective mood-stabilising drugs, such as lithium (Li+) and valproate (Eickholt et al. 2005), act by modulating downstream intracellular signalling pathways that generate the tonic excitatory drive that regulates the membrane excitability that controls mood (Fig. 3). An imbalance in the activity of these signalling pathways that regulate the tonic drive may thus be one of the causes of BPD.
The exact nature and cause of this E-I imbalance is not clear but a possible mechanism can be inferred based on the mode of action of Li+, which is one of the most effective drugs for treating BPD. One of the proposed targets for Li+ is inositol monophosphatase, which hydrolyses inositol 1-phosphate (InsP1) to inositol (Fig. 3; Allison and Stewart 1971; Berridge et al. 1982, 1989). The inhibitory effect of Li+ depends on an uncompetitive mode of action, which is unusual in that the Li+-binding site appears very briefly when the target enzyme is bound to its substrate. An interesting consequence of this unusual inhibitory action is that Li+ acts as a homeostatic drug in that it exerts little inhibitory action when the enzyme is operating normally but becomes increasingly effective the more abnormally active the signalling system becomes (Gee et al. 1988). On the basis of this evidence, we can reasonably suggest that the enhanced activity of either the excitatory or the inhibitory neurons may be driven by an overactive phosphoinositide signalling pathway that would enhance Ca2+ signalling through two mechanisms. First, Ca2+ entry increases, because an increase in the precursor PtdIns4,5P2 is known to enhance the activity of L-type Ca2+ channels (Suh et al. 2013). Secondly, excessive formation of InsP3 will result in enhanced Ca2+ release from internal stores leading to larger elevations of Ca2+ with a subsequent increase in the tonic excitatory drive (Fig. 3). Such a mechanism explains why Li+ is effective in treating both mania and depression. If the E-I imbalance is such that the tonic excitatory drive to the excitatory neurons is over-active, then this could account for mania, whereas an overactive inhibitory system would explain depression.
The basis of the inositol depletion hypothesis (Berridge et al. 1982) is that Li+ reduces the supply of inositol that is maintaining the PtdIns4,5P2 levels responsible for the excessive generation of InsP3 and the resulting abnormal increase in Ca2+ signalling described earlier. The operation of the phosphoinositide signalling pathway is highly dependent on the constant recycling of inositol (Fig. 3). Once PtdIns4,5P2 has been hydrolysed to InsP3, the latter is then sequentially dephosphorylated to InsP1 that is then hydrolysed to inositol by the Li+-sensitive inositol monophosphatase (IMPase). By inhibiting the IMPase, Li+ effectively chokes off the supply of inositol to reduce the resynthesise of the PtdIns4,5P2 required to maintain the InsP3/Ca2+ signalling pathway. Valproate, which is another potent mood-stabilising drug, may also act to reduce the supply of inositol through two separate mecanisms. First, it may act to inhibit the inositol synthase responsible for the de novo synthesis of inositol from glucose-6-phosphate (Agam et al. 2002; Shaltiel et al. 2004). Secondly, valproate has been shown to reduce the levels of PtdIns4P and PtdIns4,5P2 (Chang et al. 2012, 2014). Such a reduction in the level of the phosphoinositides would have a profound effect on a number of signalling mechanisms that regulate neuronal excitability. As described earlier, PtdIns4,5P2 regulates the activity of the KV7.2/KV7.3 channels responsible for the M current (IM; Suh and Hille 2002). High levels of PtdIns4,5P2 would also enhance the formation of inositol 1,4,5-trisphosphate (InsP3), which releases Ca2+ to stimulate the Ca2+-activated CAN channel. The ability of valproate to reduce the levels of PtdIns4,5P2 is thus consistent with the inositol depletion hypothesis, as it will also result in a reduction in the phosphoinositide-dependent signalling pathways. The ability of Li+ and valproate to increase synapse formation between cultured hippocampal neurons can be explained by this reduction in the supply of membrane phosphoinositides (Kim and Thayer 2009). Support for the inositol depletion hypothesis has come from experiments in which inositol has been found to reverse the behavioural effects of Li+ (Kofman and Belmaker 1990; Belmaker et al. 1996; Agam et al. 2009). Likewise, the behavioural effects of ebselen, which is a “safe lithium mimetic”, can also be reversed by inositol (Singh et al. 2013). All this evidence supports the notion that an abnormal elevation in the InsP3/Ca2+ signalling pathway and the consequent increase in the tonic excitatory drive in either the excitatory or inhibitory neurons is responsible for the onset of BPD.
Several observations indicate that the resting and activated levels of Ca2+ are elevated in BPD (Warsh et al. 2004). The neuronal Ca2+ sensor-1 (NCS-1), which is known to be elevated in the prefrontal cortex in both BPD and schizophrenia (Koh et al. 2003), is known to enhance the activity of the InsP3 receptors (Schlecker et al. 2006) and this would contribute to an increase in the intracellular level of Ca2+ (Fig. 3). Another factor that can influence the InsP3/Ca2+ signalling pathway is an increase in reactive oxygen species (ROS), some of which may arise from the inflammatory responses that have been described in BPD (Najjr et al. 2013). The deleterious effects of ROS may be increased by a decline in the levels of the antioxidant glutathione (GSH; Dean et al. 2009). ROS can contribute to the hyperactivity of the Ca2+ signalling system because the activity of the InsP3Rs is markedly sensitised following their oxidation by ROS (Missiaen et al. 1991; Thrower et al. 1996). Such hyperactivity of the InsP3/Ca2+ signalling pathway would act to enhance the tonic excitatory drive and could thus explain the disturbed sleep patterns that are associated with the IL-6-induced neuroinflammation that occurs in BPD (Ritter et al. 2013). Reduction of ROS activity by using antioxidants such as GSH and N-acetylcysteine (NAC) offer a novel pharmacological approach to controlling BPD and other psychiatric conditions (Berk et al. 2008, 2013).
These actions of Li+ and valproate suggest that BPD may be caused by the hyperactivity of phosphoinositide signalling with enhanced activity of both the InsP3/Ca2+ and DAG/protein kinase C (PKC) pathways. This is consistent with the finding that single nucleotide polymorphisms (SNPs) in the diacylglycerol kinase (DGKH) gene, which metabolise DAG, is a significant risk factor for BPD (Baum et al. 2008). Drugs that inhibit PKC may have considerable potential for treating BPD (Zarate and Manji 2009). Likewise, SNPs of the Bcl-2 gene, which increase the risk of developing BPD, are associated with elevated basal Ca2+ levels and enhanced InsP3-mediated cytosolic Ca2+ release (Machado-Vieira et al. 2010). With regard to the InsP3/Ca2+ signalling pathway, the abnormal elevation in Ca2+ could contribute to BPD in a number of ways. It will increase membrane excitability and this may distort the neural components of the circuits that control mood. Changes in Ca2+ will also affect gene transcription through its action on CREB (Fig. 2) and it can also act through proline-rich tyrosine kinase 2 (Pyk2) to phosphorylate and activate glycogen synthase kinase-3β (GSK-3β; Sayas et al. 2006), which is an important aspect of the neurogenesis hypothesis of BPD.
Neurogenesis hypothesis of BD
Increasing evidence links changes in mood to dysfunctional signalling mechanisms that control neurogenesis (Machado-Vieira et al. 2009; Martinowich et al. 2009; Chiu et al. 2013). Neurogenesis is responsible for the generation, differentiation and survival of new neurons necessary to maintain the neural circuitry in the adult brain (Quiroz et al. 2010). Much interest is now focused on the idea that BPD may be caused by a decline in the action of neurotrophins, such as brain-derived neurotrophic factor (BDNF), which regulate neurogenesis (Quiroz et al. 2010; Chiu et al. 2013). The risk of BPD has been linked to SNPs in the gene that codes for BDNF (Liu et al. 2008). Patients with depression have low serum levels of BDNF. Despite all this evidence linking BPD to defects in BDNF signalling, less information is available on what causes the initial decrease in BDNF signalling.
Much attention is now being directed to the possibility that the decline in BDNF results from a decrease in its expression. The transcription factor CREB, which is controlled by a number of signalling pathways, appears to be a key regulator of BDNF expression (Fig. 2). The decline in BDNF seems to be associated with the depressive phase of BPD, as no evidence exists for such a decline during mania. Furthermore, a decline in the expression of BDNF is also a feature of major depressive disorder (Liu et al. 2008). The inositol depletion hypothesis developed earlier suggests that depression arises through over-active inhibitory neurons reducing the activity of the excitatory neurons and this may provide an explanation for the association between a decline in BDNF and depression. Since the majority of the neurons in the brain are excitatory neurons, they are likely to be responsible for producing and releasing most of the BDNF. A defect in BDNF formation may thus have an impact on neurogenesis by decreasing the proliferation of neuronal progenitor cells. The decline in BDNF release could be explained if the activity of the excitatory neurons is abnormally reduced by the overactive GABAergic inhibitory neurons during the depressive phase of BPD. A significant consequence of a decrease in the activity of the excitatory neurons will be a reduction in the opening of the CaV1.2 L-type Ca2+ channel that plays such a central role in the Ca2+-dependent activation of neuronal gene transcription and particularly that regulated by CREB (Fig. 2; Flavell and Greenberg 2008).
Mutations of the CACNA1C and CACNB2 gene, which code for the α1C and β2 subunits of the CaV1.2 channel, respectively, have been linked to psychiatric disorders such as BPD (Ferreira et al. 2008; Sklar et al. 2008; Cross-Disorder Group of the Psychiatric Genomics Consortium 2013). Alterations in the gating properties of this channel, which are sensitive to the membrane levels of PtdIns4,5P2 (Suh et al. 2013), which seem to be altered in BPD, are likely to have a profound effect on neural function given that the channel has a central role in the control of both tonic excitatory drive and neural gene transcription (as described earlier, see Fig. 2; Moosmang et al. 2005; Flavell and Greenberg 2008; Lacinova et al. 2008). Such an action is consistent with observations that alterations in the activity of the brain circuits, which function in emotional and memory processing in regions such as the prefrontal cortex and hippocampus, have been described in BPD (Bigos et al. 2010; Erk et al 2010).
Another important regulator of neuronal gene transcription and neurogenesis is GSK-3β, which inhibits both CREB and the transcription factor β-catenin (Rowe et al. 2007; Hur and Zhou 2010) (Fig. 2). β-Catenin is part of the Wnt signalling pathway and may be particularly important in neurogenesis. The finding that Li+ is an inhibitor of GSK-3β (Li and Jope 2010) suggests that one of the causes of BPD may be overactive GSK-3β, which suppresses the activity of CREB and β-catenin. Such an interpretation is consistent with the fact that mutations in DISC1 (disrupted in schizophrenia-1), which reduce its inhibitory effect on GSK-3β and is one of the prominent genes mutated in both schizophrenia and BPD, results in a decrease in neurogenesis (Mao et al. 2009). Some of the enhanced activity of GSK-3β may be related to the elevation in Ca2+ described earlier, because Ca2+ can activate Pyk2 (Fig. 3), which phosphorylates and activates GSK-3β (Sayas et al. 2006). An increase in the activity of Pyk2 can also stimulate long-term depression (Hsin et al. 2010), which is responsible for memory erasure and thus could help to explain the cognitive decline that occurs in both BPD and schizophrenia.
The role of CREB activation in neurogenesis is largely dependent on its role in regulating the expression of both BDNF and Bcl-2, which will promote proliferation and prevent apoptosis respectively (Fig. 3). SNPs of the Bcl-2 gene, which increase the risk of developing BPD, are associated with elevated basal Ca2+ levels and enhanced InsP3-mediated cytosolic Ca2+ release (Machado-Vieira et al. 2010). Both Li+ and valproate can markedly enhance the level of Bcl-2 (Chen et al. 1999). These observations are of particular importance, because one of the actions of Bcl-2 is to reversibly inhibit the ability of InsP3 to open the InsP3 receptor channel to release Ca2+ from the endoplasmic reticulum (Rong and Distelhorst 2008). The low level of Bcl-2 will enhance apoptosis and will also reduce its inhibitory effect on InsP3-induced Ca2+ release, which may explain the increase in both resting and activated levels of Ca2+ that are a characteristic feature of BPD (Warsh et al. 2004).
Schizophrenia, which is a severe psychiatric condition, is characterised by both positive (hallucinations and paranoia) and negative (poor attention, decline in social interactions and lack of motivation) symptoms (Ross et al. 2006). Subtle changes in the brain rhythms described earlier, which are driven by the tonic excitatory drive and are responsible for processes such as perception, consciousness and memory, may be responsible for many of these symptoms. The high frequency gamma rhythms (Fig. 1), which fire in a sustained and synchronous manner, are the ones that are impaired in schizophrenia (Lee et al. 2003; Lisman 2012; Uhlhaas and Singer 2010; Uhlaas 2013). The inhibitory GABAergic interneurons and the excitatory pyramidal glutamatergic neurons are the main components of the network oscillator that generates this gamma rhythm (Lewis and Sweet 2009; Curley and Lewis 2012). Many of the susceptibility genes and pharmacological inducers that have been linked to schizophrenia seem to be associated with a phenotypic remodelling of the GABAergic inhibitory neurons resulting in a decline in the ability of their NMDA receptors (NMDARs) to respond to glutamate. This hypofunction of the NMDARs interferes with their role in the network oscillator resulting in changes in the gamma and theta rhythms that are a feature of schizophrenia (Gonzalez-Burgos and Lewis 2012). This defect in the ability of the NMDARs on the inhibitory interneurons to respond to glutamate is the basis of the NMDAR hypofunction hypothesis of schizophrenia (Coyle 2006; Kantrowitz and Javitt 2010; Nakazawa et al. 2012; Snyder and Gao 2013). The hypothesis is supported by the finding that the administration of NMDAR antagonists such as ketamine and phencyclidine can induce schizophrenic symptoms in healthy adults (Javitt and Zukin 1991; Umbricht et al. 2000).
The gene for neuregulin-1 (NRG1) acts through the ErbB3 signalling pathway to maintain the GABAergic phenotype (Ross et al. 2006; Buonanno 2010). NRG1 controls the expression of various NMDAR signalling functions such as NMDAR itself, the scaffolding protein PSD95 that links the NMDAR to neuronal nitric oxide synthase (nNOS) and the nicotinic acetylcholine receptor (nAChR; Fig. 4). ACh can facilitate the release of glutamate at presynaptic endings by acting through ionotropic nAChRs (Sharma and Vijayaraghavan 2003). In patients with schizophrenia, the expression of nAChRs, which is regulated by this NRG1/ErbB pathway, is reduced and this could contribute to the hypofunction of the glutamatergic signalling pathway. Mutations in NRG1 decrease the activity of its protein and may thus contribute to the hypofunction of the NMDAR and a remodelling of the GABAergic phenotype. A reduction in the activity of the NRG1/ErbB3 signalling pathway also inhibits the Src kinase, which is a regulator of the NMDAR (Pitcher et al. 2011).
DAOA, DAO and serine racemase
Polymorphisms in G72, which codes for DAOA, a protein that acts normally to modulate the metabolism of D-serine, have been implicated in susceptibility to schizophrenia and BPD (Chumakov et al. 2002; Shinkai et al. 2007; Labrie and Roder 2010). Mutations of this gene enhance the activity of the DAO that degrades D-serine. In addition, there are indications that mutations in DAO, which metabolises D-serine (Corvin et al. 2007; Shinkai et al. 2007) and in the serine racemase, which converts L-serine to D-serine (Morita et al. 2007), are also linked to schizophrenia. The consequences of all these mutations would account for the lower levels of D-serine that have been observed in patients with schizophrenia (Hashimoto et al. 2005; Bendikov et al. 2007). Such a decline would result in NMDAR hypofunction, because D-serine is an agonist at the glycine modulation site on the NMDAR (Fig. 4).
CACNA1C and CACNB2
Mutations of the CACNA1C and CACNB2 genes, which code for the α1C and β2 subunits of the CaV1.2 channel, respectively, have been linked to psychiatric disorders such as BPD and schizophrenia (Ferreira et al. 2008; Sklar et al. 2008; Giegling et al. 2010; Cross-Disorder Group of the Psychiatric Genomics Consortium 2013). A gain-of-function mutation of CACNA1C has also been linked to Timothy syndrome, which is a multisystem disorder characterised by cardiac arrhythmia, developmental abnormalities such as syndactyly, immune deficiency and behavioural abnormalities that resemble autism (Splawski et al. 2004). Just how these mutations influence the function of the L-type CaV1.2 channel in BPD and schizophrenia is unclear. However, any alteration in the gating properties of this channel is likely to have a profound effect on neural function given that it has a central role in the control of neural gene transcription (as described earlier, see Fig. 2) and it also regulates both excitability and synaptic plasticity (Moosmang et al. 2005; Lacinova et al. 2008). Such an action is consistent with observations that alterations in the activity of the brain circuits, which function in emotional and memory processing in regions such as the prefrontal cortex and hippocampus, have been described in BPD and schizophrenia (Bigos et al. 2010; Erk et al. 2010).
A polymorphism of the MIR137 gene, which controls the expression of genes such as CACNA1C (see above), CSMD1 (CUB and Sushi multiple domains 1) and TCF4 (transcription factor 4), has been implicated in schizophrenia (Guella et al. 2013). The miR-137 may also control the expression of a number of other schizophrenia risk genes such as ErbB4, the GABA receptor gene (GABRA1), the NMDA receptor 2A subunit gene (GRIN2A), the metabotropic glutamate receptor 5 gene (GRM5), the GSK-3β gene (GSK3B), the neuregulin 2 gene (NRG2) and the 5-hydroxytryptamine receptor 2 gene (HTR2C), all of which have important roles in regulating many neuronal cell functions (Wright et al. 2013).
The gene for the CaM-binding protein NRGN is one of the highest ranking susceptibility genes for schizophrenia (Lisman 2012). NRGN, which is particularly enriched in dendritic spines, has been implicated in spatial learning and hippocampal plasticity. It functions to regulate CaM levels in postsynaptic regions. In response to an increase in Ca2+, the NRGN/CaM complex is disrupted and the released CaM is then free to carry out its many postsynaptic functions such as those that control gene transcription (Fig. 2). One of the consequences of the NRGN mutation appears to be an alteration in the function of the anterior cingulate cortex (Ohi et al. 2012; Krug et al. 2013), which controls both the encoding and retrieval of memory and is known to be hyper-activated in schizophrenia. Mutations in NRGN may thus prevent its participation in the Ca2+-induced activation of the neuronal gene transcription necessary to maintain the phenotypic stability of the GABAergic phenotype of the inhibitory interneurons.
BDNF and TrkB
The function of the inhibitory interneurons, particularly during early development, is strongly influenced by BDNF. Both the expression of BDNF and its receptor TrkB are reduced in schizophrenia. The signalling pathways induced by TrkB receptors control the transcription factor CREB, which is the same transcription factor that is activated by Ca2+ to maintain the GABAergic phenotype as described earlier (Fig. 4). The TrkB receptors also activate the phosphatidyl inositol 3-kinase signalling pathway, which regulates the activity of GSK-3β, which in turn controls transcription factors such as β-catenin and CREB and has an important role in driving numerous neuronal functions including development (Lovestone et al. 2007; Hur and Zhou 2010). The activity of GSK-3β is also regulated by the protein called disrupted in schizophrenia (DISC1) as described below.
DISC1 is one of the most highly associated susceptibility genes for schizophrenia (Li and Jope 2010). A chromosomal translocation between chromosome 11 and chromosome 1 (in which DISC1 is located) results in the truncation of DISC1 and the subsequent loss of function seems to be responsible for the disruption of both brain structure and function. These widespread changes can be explained on the basis that DISC1 has many binding partners responsible for carrying out many physiological processes. For example, the mutated form of DISC1 is incapable of inhibiting GSK-3β resulting in a more pronounced inhibition of transcription thus contributing to a decline of the GABAergic phenotype and inhibitory interneuron function. In addition, mutations of DISC1 alter the transport of GABA-containing vesicles thus reducing the release of GABA and can also alter the activity of the phosphodiesterase 4B (PDE4B) that hydrolyses the cyclic AMP that will have an impact on the tonic excitatory drive (Fig. 4).
VIPR2 receptors may contribute to schizophrenia because duplications of the VIPR2 gene confers a significant risk for schizophrenia (Vacic et al. 2011). Vasoactive intestinal peptide (VIP) normally acts on VIPR2 to enhance the cyclic AMP signalling pathway to increase the tonic drive. An increase in the cyclic AMP signalling pathway and the tonic excitatory drive may contribute to schizophrenia by altering the gamma rhythms.
ROS and GSH
Hypofunction of the NMDAR has been linked to a change in a redox signalling mechanism, which is used normally as a mechanism to modulate the activity of this channel (Nakazawa et al. 2012). The fast spiking inhibitory interneurons in the cortex are particularly sensitive to the redox state of the brain. The NMDAR channel, which consists in NR1 and NR2A subunits, is physically linked to nNOS in the post-synaptic density through the scaffolding protein PSD95 (Fig. 4). nNOS is thus ideally positioned to respond to the pulses of Ca2+ entering through the NMDAR to generate the NO that then interacts rapidly with superoxide (O2–•) to form peroxynitrite (ONOO-), which is much more reactive than the two parent molecules. The ONOO- then nitrosylates the NR2A subunit at Cys-399 to inhibit the channel open probability to reduce the NMDAR gating of Ca2+ as part of a negative feedback loop to protect against excessive Ca2+ entry. An abnormal increase in the oxidation state of these inhibitory neurons may be responsible for the onset of schizophrenia (Behrens and Sejnowski 2009; Do et al. 2009).
One cause of the enhanced oxidative state in schizophrenia is associated with an increase in neuroinflammation, which is elevated in a number of psychiatric diseases (Najjr et al. 2013). There appears to be a link between maternal viral infection during gestation and the incidence of schizophrenia in the offspring. Dysregulation of the redox signalling pathway may explain this developmental origin of schizophrenia. In rat models that reproduce this phenomenon, the developmental defects induced by such inflammatory responses are caused by activation of redox signalling. During inflammation, an increase occurs in interleukin-6 (IL-6), which has a prominent role in activating the redox signalling pathway (Behrens and Sejnowski 2009). IL-6 acts through the JAK/STAT (Janus tyrosine kinase/signal transducers andactivators of transcription) signalling pathway to increase the expression of Nox2, which is one of the NADPH oxidases responsible for generating O2–• (Fig. 4). Both the O2–• and the hydrogen peroxide (H2O2) formed from O2–• by superoxide dismutase, react with nitric oxide (NO) to form the ONOO- that then nitrosylates the NR2A subunit to reduce Ca2+ as described above. This reduction in Ca2+ signalling may have a direct bearing on the cognitive defects in schizophrenia. In addition, the reduction in Ca2+ signalling can explain the phenotypic remodelling of the GABAergic phenotype as described earlier.
The role of enhanced oxidation in driving these phenotypic alterations of the inhibitory interneurons is also consistent with studies of the changes in the antioxidant mechanisms that have been described in neurodegenerative disorders (Hardingham and Bading 2010) including schizophrenia. The denitrosylation reaction, which functions to reverse the nitrosylation reaction that reduces the activity of the NMDAR, is carried out by two main antioxidants: glutathione (GSH) and thioredoxin (TRX; Fig. 4). In schizophrenia, gene polymorphisms have been described in the enzymes responsible for the synthesis of GSH such as the glutamate cysteine ligase (GCL) catalytic subunit and a GCL modifier subunit that combine to form the GCL responsible for one of the steps of GSH synthesis (Berk et al. 2008; Dean et al. 2009). Such defects in GSH synthesis may explain the decreased GSH levels found in schizophrenia. Antioxidants such as GSH and NAC, which act to reduce ROS activity, offer a novel pharmacological approach to controlling schizophrenia and other psychiatric conditions such as BPD (Berk et al. 2008, 2013).
In summary, an increase in redox signalling plays a significant role in remodelling the GABAergic neurons such that their inhibitory role is reduced resulting in the alterations in the gamma rhythms that are responsible for schizophrenia. Modulation of the tonic excitatory drive of GABAergic interneurons in schizophrenia may also contribute to the changes in gamma rhythms.
Remodelling of the GAGAergic phenotype in schizophrenia
The repetitive Ca2+ signals generated by the NMDARs and the signalling pathway activated by BDNF converge on the transcription factor CREB (Fig. 4), which is responsible for maintaining the phenotype of these GABAergic inhibitory interneurons. CREB acts to increase the expression of a number of the signalling components that define the GABAergic phenotype. Hypofunction of the NMDARs or a reduction in BDNF signalling reduces the expression of glutamic acid decarboxylase 67 (GAD67; Figs. 2, 4), which is one of the key components of the phenotype in that it synthesises the inhibitory transmitter GABA (Gonzalez-Burgos and Lewis 2008). This GABA is packaged into vesicles that are transported to the presynaptic terminals where they are released as part of the network oscillatory mechanism that generates the gamma rhythms. A reduction in the level of GAD67 and the resulting decreased availability of GABA is one of the most characteristic features of schizophrenia. The transport of GABA-containing vesicles down the axon to the presynaptic ending is mediated by the dynein motor, which travels down microtubules (Fig. 4). The protein of DISC1, which is one of the genes mutated in schizophrenia (as described above), disrupts the transport of GABA vesicles.
Some of the phenotypic changes that occur in these interneurons seem to be compensatory responses caused by this primary defect in the expression of GAD67 (Gonzalez-Burgos and Lewis 2008). This may explain the decline in the level of the Ca2+ buffer parvalbumin (PV), which modulates the Ca2+-dependent release of GABA (Fig. 4). This release of GABA is controlled by a brief pulse of Ca2+, which is rapidly buffered by PV to reduce the process of facilitation that depends on the build-up of Ca2+ during synaptic activity. The decline in PV levels in schizophrenia will enhance facilitation and may be a compensatory response to the reduction in GABA levels caused by the decrease in GAD67 (Gonzalez-Burgos and Lewis 2008). Another compensatory mechanism depends on a decrease in the expression of the GABA membrane transporter 1 (GAT1), which is located in the presynaptic ending where it functions to return GABA back into the interneuron (Gonzalez-Burgos and Lewis 2008). A reduction in this removal mechanism will help to enhance the activity of the reduced amount of GABA being released in schizophrenia.
In summary, hypofunction of the NMDARs and the resulting reduction of Ca2+ signalling sets in motion a complex series of transcriptional and compensatory events that remodel the GABAergic phenotype. A reduction in the release of the inhibitory neurotransmitter GABA may explain the gamma rhythm changes that occur in schizophrenia. Alterations in rhythmical activity may also arise through modulation of the tonic excitatory drive of these GABAergic interneurons. The identification of the reason for the decrease in NMDAR function may help our understanding of the biochemical basis for schizophrenia. Excessive ROS signalling appears to be one of the primary causes of NMDAR hypofunction in schizophrenia.
Modulation of the tonic excitatory drive of GABAergic interneurons in schizophrenia
The tonic excitatory drive regulates the activity of the neuronal oscillator that generates brain rhythms (Fig. 1). This tonic drive is present in both the inhibitory GABAergic and excitatory glutamatergic neurons. In the case of the GABAergic interneurons, transmitters such as ACh, dopamine, glutamate, vasoactive intestinal peptide (VIP) and oxytocin, which use cyclic AMP or InsP3/Ca2+ signalling pathways to regulate this tonic excitatory drive, have been implicated in schizophrenia (Fig. 4). Considerable interest is now focused on the metabotropic glutamate receptor 5 (mGluR5), which is closely associated with the NMDAR through scaffolding proteins such as Homer, SHANK, guanylate-kinase-associated protein and PSD95. Alterations in mGluR5 may contribute to schizophrenia (Matosin and Newell 2013) and mutations in SHANK have been linked to various neuropsychiatric diseases including schizophrenia (Gauthier et al. 2010; Guilmatre et al. 2014). Furthermore, a deficit in ACh signalling mediated through muscarinic receptors has also been implicated in schizophrenia (Scarr and Dean 2008). A decrease in oxytocin, which may act through the InsP3/Ca2+ signalling pathway to regulate the tonic excitatory drive (Fig. 4), has been implicated in schizophrenia and other mental disorders such as autism (Meyer-Lindenberg et al. 2011). Schizophrenic-like symptoms also occur in response to drugs such as the amphetamines that release dopamine, whereas drugs that inhibit the D2 receptor are antipsychotic. These pharmacological actions thus predict that schizophrenic symptoms arise through excessive activation of the D2 receptor resulting in a reduction in the level of cyclic AMP and a decrease in the tonic drive. This inhibition of cyclic AMP signalling alters the input resistance by reducing the activity of the HCN1 channels that provides an inward Na+ current (Fig. 4). Since the activity of HCN1 is regulated by cyclic AMP, this could also explain the way that alterations in the activity of the dopamine D2 and VIPR2 receptors may contribute to schizophrenia. It is somewhat surprising to find that the duplication of VIPR2, which confers a significant risk for schizophrenia (Vacic et al. 2011), has the opposite effect on cyclic AMP levels to those just described for dopamine. VIP acting on VIPR2 enhances the cyclic AMP signalling pathway to increase the tonic drive. Changes in the cyclic AMP signalling pathway thus appear to be able either to enhance or to reduce the tonic excitatory drive and this would have repercussions for the generation of the gamma rhythms resulting in schizophrenia. Such contradictory effects on the tonic excitatory drive can be reconciled by the finding that gamma rhythms may be increased or reduced in the various schizophrenia syndromes (Lee et al. 2003).
An important role for cyclic AMP is evident from the finding that alterations in the activity of PDE4B, which hydrolyses cyclic AMP (Fig. 4), have been identified in schizophrenia. SNPs associated with the gene that codes for PDE4B have been described in schizophrenia. In addition, DISC1, the gene of which is mutated in schizophrenia, has also been shown to interact with PDE4B to alter the metabolism of cyclic AMP. Changes in the level of cyclic AMP will have an impact on the tonic excitatory drive and will thus contribute to the gamma rhythm alterations that have been described in schizophrenia.
Subtle alterations in the sophisticated Ca2+ signalling systems that regulate many neuronal activities have been implicated in psychiatric diseases such as BPD and schizophrenia. What is intriguing about these two diseases is that, even though they display different behavioural changes, the underlining signalling and cellular alterations have much in common. For example, an alteration in the mechanisms underpinning the tonic excitatory drive responsible for the brain rhythms that define the sleep/wake cycle is a feature of both diseases. The ascending arousal system activates the tonic excitatory drive regulating the activity of neuronal rhythms. I have argued that an E-I imbalance disturbs this rhythm-generating system resulting in severe behavioural defects. For example, the mood switching between depression and mania that characterises BPD might be explained by such an E-I imbalance. Depression may arise from a long-term increase in the activity of the GABAergic inhibitory interneurons that suppress the excitatory neurons. Conversely, excessive activity of the excitatory neurons account for the manic phase. In the case of schizophrenia, the hypofunction of the NMDA receptor during early development results in a phenotypic switch in the GABAergic inhibitory interneurons resulting in a decline in their participation in the neuronal rhythms.
In both BPD and schizophrenia, the phenotypic switching that distorts the neuronal rhythms seems to depend on alterations in the Ca2+-dependent gene transcription of essential signalling components. In the hypothesis developed earlier, the depressive phase of BPD is proposed to result from an excessive suppression of excitatory neurons by over-active GABergic inhibitory neurons. This reduction in the active excitatory neurons may reduce the Ca2+ signals necessary to maintain the expression of neutrophic factors such as BDNF, which regulates neurogenesis. Likewise, the hypofunction of the NMDA receptors that provide the regular pulses of Ca2+ that maintain the phenotypic stability of the GABAergic interneurons may be responsible for schizophrenia.
Although phenotypic switching is well established in both BPD and schizophrenia, the underlying signalling defects leading to the changes in gene transcription are still a matter of conjecture. Mounting evidence indicates that an increase in oxidative metabolism, which is enhanced by the neural inflammation that occurs in these psychiatric diseases, contributes to the onset of these diseases. For example, the hypofunction of the NMDAR in schizophrenia is enhanced by an increase in ROS formation. The effectiveness of Li+ in treating BPD suggests that an increase in the activity of the InsP3/Ca2+ signalling pathway causes this disorder. Another candidate is GSK-3β and an intriguing speculation is that the increased activity of this enzyme is linked to the increase in Ca2+ signalling through the Ca2+-sensitive protein kinase Pyk2.