Schizophrenia Flashcards
Delay and Deniker
o The chance observation that chlorpromazine induced a calming effect initiated a remarkable period of efficacious therapy
o In France, chlorpromazine was being tried as an anaesthesia adjunct that might prevent shock or have beneficial properties
o Chosen as the least sedative of the antihistamines available at the time, it was tested in a wide variety of patients
o When tried in patients with SCZ, Delay and Deniker reported their results at a psychiatric meeting in Luxemburg and then very quickly in a journal article in 1952 describing spectacular results in acutely psychotic SCZ pts
Wright (2000)
Found a number of changes e.g. ventricular enlargement, decreased cortical, grey matter, hippocampal+thalamic volume + larger basal ganglia. Likewise, there is reduced blood flow to the globus pallidus and frontal lobes Overall, cerebral volume decreased by 3% but w careful sterotaxic counts it is seen that a higher density of neurons remains in this volume (no real change in number of neurons so brain is more densely populated - strange!)
Moncrieff and Leo (2010)
SR of MRI studies looking at the role of antipsychotic medication onstructural changes found decreased grey matter and increased ventricular volume associated w antipsychotic treatment; limited changes in drug-naïve pts 14 of 26 longitudinal studies showed a decline in global brain or grey matter volume or an increase in ventricular or CSF volume during the course of drug treatment some evidence points towards the possibility that antipsychotic drugs reduce the volume of brain matter and increase ventricular fluid volume, ANTIPSYCHOTICS MAY CONTRIBUTE TO THE GENESIS OF SOME OF THE ABNORMALITIES USUALLY ATTRIBUTED TO SCHIZOPHRENIA
Connel 1957
published a series of case reports showing amphetamines, which cause DA release, induce psychotic symptoms akin to SCZ positive symptoms
APs (indeed all currently licensed) reduce DA transmission and inhibit positive symptoms of DA in ~70% of pts
Angrist et al 1970
Looked at effects of amphetamine in 4 healthy subjects. Lead to hallucinations and delusions. Known to increase DA
amphetamine also increases NA and 5-HT, participants were experienced amphetamine users - might have neurophysiological/chemical changes in brain
Johnstone 1978
clinical trial compared AP drug alpha-flupenthixol vs placebo and against beta-flupenthixol (optical isomer which does not bind to D2R), which was more efficacious than control and isomer, demonstrating necessity of DR2 blockage
Seeman 1976
They used PET. Measured 3H haloperidol conc before/after addition of different APs. Plotted DAR affinity (ability to displace haloperidol) with clinical dose needed for efficacy, finding a strong correlation between efficacy of APs and affinity for D2 receptors
Was in vitro - does this apply in vivo?
Howes 2013
18F-DOPA (radiolabelled DOPA) in PET of AP-free SCZ, increased DOPA uptake in striatal neurons, w extent of uptake correlating w the severity of symptoms (key finding), and elevated in pts w prodromal psychotic symptoms (before onset of disease), and further elevated w onset of acute psychosis
Volkow 1999
Used 11[C]-raclopride which binds D2Rs to measure amphetamine-induced DA release, fewer radiotracer-bound Rs in SCZ, demonstrating greater DA release. DA release increment correlated to aggravated psychosis
11C raclopride has D3 binding. Amphetamine-induced DA release is not a physiologically similar situation to SCZ, and argued that effect could have been due to differences in amphetamine metabolism or its ability to enter the CNS - it is artificial - how much can you read into this? A
Abi-Dargham 2000
used SPECT with [123I]-IBZM to measure baseline dopamine occupancy of D2 receptors. o Schizophrenic patients and control subjects were administered alpha-methyl-para-tyrosine (TH inhibitor) to deplete endogenous dopamine. The decrease in receptor occupancy was significantly greater for schizophrenic patients (19% compared to 9%), indicating greater receptor occupancy at baseline, which supports the mesolimbic hyperdopaminergic theory of schizophrenia.
Assumes the magnitude of DA depletion is the same between SCZ and ctrl
Howes et al. 2012
meta-analysis of 44 in vivo studies with 600 pts and 600 controls looking at striatal dopamine with PET “Around a doubling of synaptic DA, and of D2 R occupancy in pts compared to controls They found that there was a highly sig elevation in presynaptic dopaminergic function in SCZ (dopamine synthesis capacity, basline DA levels and dopamine release). These are not targeted by current drug treatments, which primarily act to block D2 Rs - fail to target these abnormalities. Need future drugs to focus on the control ofpresynaptic DA synthesis and release capacity”
Abi-Dargham 2002
PET study measuring selective D1 receptor availability with labelled D1 antagonist. confirmed D1R increased throughout cortex in SCZ, only statistically sig difference was in dorsolateral PFC; thought to be increased as compensatory phenomenon due to DA deficit. May be that this upregulation is not functional due to lack of stimulation by endogenous DA and so PFC-mediated deficits (underlying negative + cognitive symptoms)
Slifstein et al. 2015
PET has only recently become sensitive enough to detect PFC DA alterations. This is the first in vivo example of this. They looked a 20 schizophrenics and 20 controls. They did PET imaging using 11[C]-GLB457 radiotracer which binds to both D2 and D3 receptors. They gave a bolus of amphetamine to each group (for DA release). Demonstrate that DA release is significantly reduced in DL PFC of SCZ.
- Egan 2001
They looked at COMT (enzyme involved in DA degradation) – specifically a common functional polymorphism, that accounts for 4-fold variation in enzyme activity and dopamine catabolism. Compared the presence of this to performance on Wisconsin Card Sorting test of executive function in 175 SCZ, 219 unaffected siblings, 55 controls. Found the genotype explained 4% of variance of WCS score (p=0.001); low activity Met allele predicted enhanced cognitive performance. Then examined COMT genotype on physiological PFC response using fMRI and Met allele predicted more efficient physiological response in PFC. They did a family-based association analysis and found that there was significant transmission of the Val allele to SCZ offspring (not the met allele) o So overall more COMT activity = more prefrontal DA catabolism = worse WCS score (impaired prefrontal cognition) and reduced PFC signalling (= means association with)
Kellendonk et al 2006
- generated transgenic mouse model in which D2R expression could be selectively upregulated in the striatum, mimicking DA hyperactivity
- Mice had decreased DA turnover in the PFC, and performed poorly in PFC-dependent cognitive tasks that persisted even long after the transgene had been turned off
- Also get D1R upregulation in these mice, as PET studies in humans SCZs show you do, but limited D1R antagonist trials show lack of efficacy or symptoms worsening
- Idea that excessive DA during development somehow causes cortical DA insufficiency and cognitive deficit, hence D2R antagonist in early adulthood have no effect on cognitive or negative symptoms
Limitations
- No direct link to cog deficits (could have used opto/chemgen to activate/inhibit). Also going from one transgenic mouse model to this is the core pathology of SCZ - bit of a leap.
Carlsson et al 2006
Came up with DA deficit hypothesis
Kapur 2003
Came up w aberrant salience theory
Olney and Farber 1995
put forward first NMDAR dysfunction hypothesis of SCZ, showing that animals given NMDAR antagonists developed neurotoxic changes in cortical brain regions, which they suggested were similar to reduction in brain volume seen in pts w SCZ. AMPAR antagonists could block downstream effects of NMDAR antagonist on neurotoxicity, hypothesising that glutamate release might underlie neurotoxic effects
Moghaddam et al. 1997
confirmed this w microdialysis of healthy rats w subanaesthetic doses of ketamine, which they found increased extracellular glutamate in the PFC and found that this also impaired cognitive functions and produced motor sterotypy. This increase in extracellular glutamate had functional significance because blockade of AMPA receptors reduced these motoric and cognitive detriments of NMDAR blockade. Therefore, NMDA antagonsits appeared to increase glutamate release at some synapses which then abnormally increased glutamate neurotransmission at non-NMDAR, in particular AMPA receptors. These results suggests that the behavioural conseqeucnes of NMDAR deficiency is not due to a generalised glutamate hypofunction but dysregulation of glutamate neurotransmission that may involve NMDAR hypofunciton and excessive activity of non-NMDA receptors
Moghaddam 98
Group II mGluRs are autoreceptors that inhibit glutamate release. They gave group II agonist prior to PCP (NMDA antagonist). Microdialysis = increase glutamate in PFC and motor sterotypy and decrease in cog performance. All reversed by the agonist and confirmed to not be reduced by a reduction in DA
General animal model stuff (stereotypy not scz), furthermore, although in 2007 clinical trial mGlu2/3 agonists were shown to improve both positive and negative symtpoms, future trials have failed to show effect - maybe its to do with different subcategories of scz?
Krystal 1994
19pts. Given either placebo or 2 doses of K (0.1&0.5mg/kg). Produced behaviours similar to + and - symptoms (e.g. increases in hallucinations nand impaired in WCST so frontal dysfunction)
Recruitment through adverstising (homogenous population), not causal, ketamine is short-term but scz develops over a long period of time - different mechanisms?
mGlu2/3 agonists
Patil 2007 - phase II clinical trial - mGlu2/3 agonists were shown to improve both positive and negative symtpoms, future phase III trials have failed to show effect - maybe its to do with different subcategories of scz? Was only assessed for 4 weeks - AP drugs would need longer.Liu 2012 later showed that this agonist was only effective in pts with a specific SNP in the 5-HT2A receptor - although this second one was candidate gene approach (picked genes to investigate depending on their association with the pharmacological action) - maybe there is even more subcategorisation?
Mohn 1999
Generated mice expressing only 5% of the normal levels of NR1 subunit of NMDA - survive to adulthood and display behavioural abnormalities, including increased motor activity and deficits in social and sexual interactions, similar to those seen in pharmcologically-induced animal models. These can be ameliorated through treatment with APs (cloz and halo). No changes in DA with microdialysis (only glutamate and not DA) Supports model where reduced NMDAR activity causes SCZ-like behaviour
Maybe microdialysis at the time was not specific enough to pick up the D, could have used FSCV. Havent seen NR1 in GWAS. Used a homologus recombination technique which could lead to ectopic recombination in other places. Did global NDMA hypofuction, rather than in specific circuits. The mice also showed autism-related behaviooural deficits (maybe not representative of scz)
- Belforte et al. 2010
Built on previous studies (Mohn 99 in particular) .Transgenic mice using Cre-loxP, in which the essential NR1 subunit of the NMDAR was eliminated in 40-50% of cortical and hippocampal INs in early postnatal development.
Direct SCZ-related symptoms emerged after adolescence (including 1 novelty-induced hyperlocomotion, 2 mating and nest-building deficits, and 3 anhedonia-like and 4 anxiety-like behaviours). Also showed reduction in GAD67 - potential reduction in GABA synthesis - reduction in GABAergic inhibition - cortical disinhibition
not affected à neurodevelopmental hypothesis. Idea that early postnatal inhibition of NMDAR activity could contribute to pathophysiology of SCZ. Also no GAD67 reduction. Early postnatal NMDAR hypofunction in GABAergic IN is needed for SCZ-like behaviour
Limitations
1) Lack of consensus on altered subunit levels in post-mortem brain studies, so hard to ascertain construct validity (also in genetics), 2) based on expression of behavioural analogues for SCZ in humans, 3) which subtupe of GABA IN? 3) do both cortical and hippocampal INs play a role? (both were targetted), 4) only 50% of INs were even targetted
Maher 2012
showed this using optogenetics (light-sensitive channelrhodopsin-2 cation channel) to study glutamatergic release in light-stimulated wild-type and DISC1 knockdown neurons in a rodent model. o They found a significantly decreased glutamate release probability in DISC1-/- neurons, indicating the importance of DISC1 in presynaptic glutamate release
Tsai and Lin 2010
meta-analysis of NMDAR-modifying drugs. Looked at about 800 subjects from 26 studies. Showed these significantly improves positive + negative and cognitive symptoms when administered in addition to existing AP treatment
Bartha et al 1997
Proton magnetic resonance spectroscopy of 10 SCZ and 10 controls – found sig increased glutamine levels in medial PFC of SCZ compared to controls
Bustillo et al. 2014
MRS. found that chronic SCZ pts have elevated glutamine levels in anterior cingulate cortex compared to healthy controls. The levels of glutamine also positively correlated with the severity of positive symptoms
Cav- 1H-MRS doesn’t distinguish IC vs EC - cant see if glutamate has actually been released. Cant tell if this is a cause or a consequence of the disease
Flames et al 2004
“o In E13.5 medial ganglionic eminence (MGE)/cortex in vitro co-cultures, neurons from the MGE migrated towards the cortex (what happens in development) o However, MGE neurons electroporated (increase the permeability of the cell membrane with an electrical field) with dominant-negative ErbB4 became largely unresponsive to the cortical attractant. (basically if you lose Erb4, you lose this migration) o They also created ErbB4 mouse mutants expressing ErbB4 under a cardiac-specific myosin promoter to avoid embryonic lethality due to failure of myocardial trabeculae formation. o IHC for calbindin, a marker of migrating GABAergic INs, revealed sig fewer migratinc cortical INs in those mice compared to controls” ErbB4 had previously been implicated in susceptibility to scz - this just shows that GABAergic IN migration may be affected by this
Steinecke et al 2012
electroporated MGE cells in embryonic mouse brain w anti-DISC1 miRNA to cause DISC1 RNAi, + Alexa 546 to visualise transfected cells (in control animals only the latter was transfected). RNAi of DISC1 sig reduced the no of neurons migrating tangentially to the cortex
Volk et al 2000
They performed ISH on tissue sections from 10 pairs of subjects (SCZ + healthy controls) w sulphur-35-labelled oligonucleotide probes. They found sig decreased density of GAD67 (GABAergic interneuron marker) labelled neurons in subjects w SCZ. However, a sig limitation of this study was the small sample size. Therefore, developmental defects in GABAergic IN migration may contribute to GABAergic PFC hypofunction, so increase risk of SCZ.
The Schizophrenia Working Group of the Psychiatric Genomics Consortium 2014
GWAS, 37,000 patients and 113,000 controls. Identified 600 genes associated with SCZ, 83 of which were newly implicated
Incl DRD2 many w NMDA
CACNA1A and CACNA2B
Some involved in immunity - immune enhancers (CD20 B lymphocytes e.g.)
MHC = strongest association
Sekar 2016
“C4 exists as two functionally distinct genes - C4A and C4B - these vary in structure and copy number. They assessed how C4 variation related to RNA expression of C4A and C4B in post mortem human adult brain samples. Found that each C4 allele was associated with schizophrenia risk in proportion to its effect on C4A expression. o Using IHC, and high-resolution structured illumination microscopy imaging, they determined that neuronal dendrites, axons and synapses expressed C4. For example, it co-localises w presynaptic markers VGLUT1/2 and postsynaptic marker PSD-95 o C4 was then shown to promote synapse elimination in mice during the developmentally timed maturation of a neuronal circuit”
Li et al 2007
Looked at the CA3-CA1 pathway in hippocampal slices - transduction of CA1 neurons with lentivirus expressing shRNAs, producing RNAi knockdown of ErbB4, or the addition of the ErbB4 antagonsit AG1478 significantly depressed NMDAR-mediated synaptic tranmission (whole cell postsynaptic patch clamp recordings) ErB4/NGR1 impairment leads to NMDAR dysfunction
Hikida 2007
Generated mice with truncated form of Disc1 under the control of promoter aCaMKII which drived gene expression in forebrain neurons Show enlarged lateral ventricles and decreased numbers of interneurons, hyperactivity, impaired PPI, increased immobililty in FST, social interactions, or spatial memory
Kim et al 2021
Found that phosphodiesterase 4 (PDE4) transcripts are elevated in neurons differentiated from iPSCs with the DISC mutation (RNA seq)
Pharma inhibition of PDE4 or or cAMP activation rescues synaptic deficits
Mouse knock in with same DISC1 mutation had elevated PDE4 and synaptic abnormalities and SCZ-like behaviour (was heterozygous KI so some DISC1 = normal)
Inhibition of PDE4 rescued synaptic abnormalities and behaviour
Could present a future drug target - however, PDE4 is ubiquitous - could we really target it?
Singh et al April 2022
- Described 10 robust rare varients in MA of whole exomes of 24,000 cases and 97,000 controls.
- Identified 10 ultra-rare coding variants in 10 genes conferring substantial risk for SCZ.
- These had diverse molecular functions including the formation, structure and function of the synapse.
- Protein truncating varients of GRIN2A, TRIO and CACNA1G are examples of genes implicated
- GRIN2A = NMDA receptor subunit = glutamatergic
- Showed for the first time that commonly regulatory variants from GWAs and ultra-rare coding variants disrupt an overlapping set of genes. Gives confidence that integration of results from sequencing consortia with GWAS efforts will have significant value for identifying specific genes beyond what any strategy can achieve on its own - GRIN2A was in this overlap
- There was also overlap in rare variant risk w autism, epilepsy and severe neurodevelopmental disorders - supports neurodevelopmental hypothesis
Schizophrenia Working Groups of the Pyschiatric Genomics Consortium 2017
meta-analysis incorporating over 21,000 SCZ cases and 20,000 controls
looking at CNVs
Generated definitive evidence for 8 loci - they suggest that these are carried by 1.4% of SCZ
Bernnand 2011
directly reprogramming SCZ pt fibroblasts into HiPSCs and then differentiated these into neurons
Using trans-neuronal spread of rabies virus (stronger synapse = better spread) as an assay of neuronal connectivity, they found this was decreased in these neyrons
Does not necessarily mean synaptic function is decreased (could be limitations fo synaptic activity assays)” “
Decreased number of neurites (in keeping with SCZ post-mortem and animal models), decreased PSD95 expression relative to MAP2AB but not to signifiance, and decreased glutamate R expression - Spontaneous neuronal activity unaffected” -
Almost 600 genes had expression altered more than 1.3-fold, 25% had been previously implicated in SCZ sig alterations in glutamate, cAMP and WNT signalling, needed for activity-dependent refinement of synaptic connections and LTP “
her APs did nto but could be dosing needs optimisation” - Data support idea of CZ model where many different combinations of gene dysfunction may disrupt the key pathways affected in SCZ. Increase number of hiPSC cases and see consistent genes affected, and essential pathways will come to light
Demhaja et al 2014
Treatment resistant SCZ shown to have normal 18F-DOPA uptake in PET studies (and normal ACC glutamate), as opposed to SCZ pts who respond to APs and have elevated 18F-DOPA uptake (and normal ACC glutamate)
Are the differences between glu and DA a cause or consequence of poor AP response?
Kim et al 2021
- Found that phosphodiesterase 4 (PDE4) transcripts are elevated in neurons differentiated from iPSCs with the DISC mutation (RNA seq)
- Pharma inhibition of PDE4 or or cAMP activation rescues synaptic deficits
- Mouse knock in with same DISC1 mutation had elevated PDE4 and synaptic abnormalities and SCZ-like behaviour (was heterozygous KI so some DISC1 = normal)
- Inhibition of PDE4 rescued synaptic abnormalities and behaviour
- Could present a future drug target - however, PDE4 is ubiquitous - could we really target it?
Schmack et al 2021
Designed behavioural sensory detection task in mice
- Given tones embedded in white noise. Would poke a port if signal was there and another if it wasnt. Given water reward
- Uncertainty was introduced by presenting auditory signals hardly distinguishable from the noise
- Their confidence in the choices determined by time willing to wait for the reward after a response - behavioural proxy for confidence
- Hallucination = high confidence false alarm (incorrect reports) (HALIP)
Results
- Increasing signal expectation by increasing proportion of signal to no signal trial = increased HALIPs
- Administration of ket increased HALIP rate in mice
They then devised a computational model
- Explains these mouse ‘hallucinations’ as a consequence of more weight being put on prior expectation than sensory evidence.
- Their model clarifies how these ‘hallucinations’ come from fluctuations in two different types of expection - reward expectations (DA fluctuation in v striatum) and perceptual expectations (DA fluctuation in tail of striatum)
Used genetically encoded DA responses with fibre photometry to monitor DA dynamics
- Found sustained increase of DA in the striatum in no-signal trials that preceded false alarm responses
- Optogenetic stimulation of DA release in striatum lead to increased HALIPs
- This was rescued with haloperidol
Supports idea hallucinations come from elevated DA makes bias favouring expectations over sensory evidence
Limitations
- Time investments dont mean confidence (could just be learning to keep nose in for londer)
- Didnt look at visual hallucinations etc
- Did not investigate in a model of psychosis
- Do not know if the neural processes underlying HALIPs overlap those that produce spontaneous hallucination sin scz
Future
- Human neuroimaging studies have shown how sound percepts are represented in the cortex during hallucinations. However, unclear how this auditory-related region of the striatum underpins this is unknown. Need deep-brain calcium imaging in large popultions of cortical and striatal neurons in animal models of hallucinations
- Opens door to potential treatment - one study showed DBS of the humans striatum in tinnitus = reduction in loudness of tinnitus. Although Schmack gave haloperidol to reduce HALIPs, it remained unanswered if silencing striatal DA neurons is enough to prevent hallucinations
o The chance observation that chlorpromazine induced a calming effect initiated a remarkable period of efficacious therapy
o In France, chlorpromazine was being tried as an anaesthesia adjunct that might prevent shock or have beneficial properties
o Chosen as the least sedative of the antihistamines available at the time, it was tested in a wide variety of patients
o When tried in patients with SCZ, Delay and Deniker reported their results at a psychiatric meeting in Luxemburg and then very quickly in a journal article in 1952 describing spectacular results in acutely psychotic SCZ pts
Delay and Deniker
Found a number of changes e.g. ventricular enlargement, decreased cortical, grey matter, hippocampal+thalamic volume + larger basal ganglia. Likewise, there is reduced blood flow to the globus pallidus and frontal lobes Overall, cerebral volume decreased by 3% but w careful sterotaxic counts it is seen that a higher density of neurons remains in this volume (no real change in number of neurons so brain is more densely populated - strange!)
Wright (2000)
SR of MRI studies looking at the role of antipsychotic medication onstructural changes found decreased grey matter and increased ventricular volume associated w antipsychotic treatment; limited changes in drug-naïve pts 14 of 26 longitudinal studies showed a decline in global brain or grey matter volume or an increase in ventricular or CSF volume during the course of drug treatment some evidence points towards the possibility that antipsychotic drugs reduce the volume of brain matter and increase ventricular fluid volume, ANTIPSYCHOTICS MAY CONTRIBUTE TO THE GENESIS OF SOME OF THE ABNORMALITIES USUALLY ATTRIBUTED TO SCHIZOPHRENIA
Moncrieff and Leo (2010)
published a series of case reports showing amphetamines, which cause DA release, induce psychotic symptoms akin to SCZ positive symptoms
APs (indeed all currently licensed) reduce DA transmission and inhibit positive symptoms of DA in ~70% of pts
Connel 1957
Looked at effects of amphetamine in 4 healthy subjects. Lead to hallucinations and delusions. Known to increase DA
amphetamine also increases NA and 5-HT, participants were experienced amphetamine users - might have neurophysiological/chemical changes in brain
Angrist et al 1970
clinical trial compared AP drug alpha-flupenthixol vs placebo and against beta-flupenthixol (optical isomer which does not bind to D2R), which was more efficacious than control and isomer, demonstrating necessity of DR2 blockage
Johnstone 1978
They used PET. Measured 3H haloperidol conc before/after addition of different APs. Plotted DAR affinity (ability to displace haloperidol) with clinical dose needed for efficacy, finding a strong correlation between efficacy of APs and affinity for D2 receptors
Was in vitro - does this apply in vivo?
Seeman 1976
18F-DOPA (radiolabelled DOPA) in PET of AP-free SCZ, increased DOPA uptake in striatal neurons, w extent of uptake correlating w the severity of symptoms (key finding), and elevated in pts w prodromal psychotic symptoms (before onset of disease), and further elevated w onset of acute psychosis
Howes 2013
Used 11[C]-raclopride which binds D2Rs to measure amphetamine-induced DA release, fewer radiotracer-bound Rs in SCZ, demonstrating greater DA release. DA release increment correlated to aggravated psychosis
11C raclopride has D3 binding. Amphetamine-induced DA release is not a physiologically similar situation to SCZ, and argued that effect could have been due to differences in amphetamine metabolism or its ability to enter the CNS - it is artificial - how much can you read into this? A
Volkow 1999
used SPECT with [123I]-IBZM to measure baseline dopamine occupancy of D2 receptors. o Schizophrenic patients and control subjects were administered alpha-methyl-para-tyrosine (TH inhibitor) to deplete endogenous dopamine. The decrease in receptor occupancy was significantly greater for schizophrenic patients (19% compared to 9%), indicating greater receptor occupancy at baseline, which supports the mesolimbic hyperdopaminergic theory of schizophrenia.
Assumes the magnitude of DA depletion is the same between SCZ and ctrl
Abi-Dargham 2000
meta-analysis of 44 in vivo studies with 600 pts and 600 controls looking at striatal dopamine with PET “Around a doubling of synaptic DA, and of D2 R occupancy in pts compared to controls They found that there was a highly sig elevation in presynaptic dopaminergic function in SCZ (dopamine synthesis capacity, basline DA levels and dopamine release). These are not targeted by current drug treatments, which primarily act to block D2 Rs - fail to target these abnormalities. Need future drugs to focus on the control ofpresynaptic DA synthesis and release capacity”
Howes et al. 2012
PET study measuring selective D1 receptor availability with labelled D1 antagonist. confirmed D1R increased throughout cortex in SCZ, only statistically sig difference was in dorsolateral PFC; thought to be increased as compensatory phenomenon due to DA deficit. May be that this upregulation is not functional due to lack of stimulation by endogenous DA and so PFC-mediated deficits (underlying negative + cognitive symptoms)
Abi-Dargham 2002
PET has only recently become sensitive enough to detect PFC DA alterations. This is the first in vivo example of this. They looked a 20 schizophrenics and 20 controls. They did PET imaging using 11[C]-GLB457 radiotracer which binds to both D2 and D3 receptors. They gave a bolus of amphetamine to each group (for DA release). Demonstrate that DA release is significantly reduced in DL PFC of SCZ.
Slifstein et al. 2015
They looked at COMT (enzyme involved in DA degradation) – specifically a common functional polymorphism, that accounts for 4-fold variation in enzyme activity and dopamine catabolism. Compared the presence of this to performance on Wisconsin Card Sorting test of executive function in 175 SCZ, 219 unaffected siblings, 55 controls. Found the genotype explained 4% of variance of WCS score (p=0.001); low activity Met allele predicted enhanced cognitive performance. Then examined COMT genotype on physiological PFC response using fMRI and Met allele predicted more efficient physiological response in PFC. They did a family-based association analysis and found that there was significant transmission of the Val allele to SCZ offspring (not the met allele) o So overall more COMT activity = more prefrontal DA catabolism = worse WCS score (impaired prefrontal cognition) and reduced PFC signalling (= means association with)
- Egan 2001
- generated transgenic mouse model in which D2R expression could be selectively upregulated in the striatum, mimicking DA hyperactivity
- Mice had decreased DA turnover in the PFC, and performed poorly in PFC-dependent cognitive tasks that persisted even long after the transgene had been turned off
- Also get D1R upregulation in these mice, as PET studies in humans SCZs show you do, but limited D1R antagonist trials show lack of efficacy or symptoms worsening
- Idea that excessive DA during development somehow causes cortical DA insufficiency and cognitive deficit, hence D2R antagonist in early adulthood have no effect on cognitive or negative symptoms
Limitations
- No direct link to cog deficits (could have used opto/chemgen to activate/inhibit). Also going from one transgenic mouse model to this is the core pathology of SCZ - bit of a leap.
Kellendonk et al 2006
Came up with DA deficit hypothesis
Carlsson et al 2006
Came up w aberrant salience theory
Kapur 2003
put forward first NMDAR dysfunction hypothesis of SCZ, showing that animals given NMDAR antagonists developed neurotoxic changes in cortical brain regions, which they suggested were similar to reduction in brain volume seen in pts w SCZ. AMPAR antagonists could block downstream effects of NMDAR antagonist on neurotoxicity, hypothesising that glutamate release might underlie neurotoxic effects
Olney and Farber 1995
confirmed this w microdialysis of healthy rats w subanaesthetic doses of ketamine, which they found increased extracellular glutamate in the PFC and found that this also impaired cognitive functions and produced motor sterotypy. This increase in extracellular glutamate had functional significance because blockade of AMPA receptors reduced these motoric and cognitive detriments of NMDAR blockade. Therefore, NMDA antagonsits appeared to increase glutamate release at some synapses which then abnormally increased glutamate neurotransmission at non-NMDAR, in particular AMPA receptors. These results suggests that the behavioural conseqeucnes of NMDAR deficiency is not due to a generalised glutamate hypofunction but dysregulation of glutamate neurotransmission that may involve NMDAR hypofunciton and excessive activity of non-NMDA receptors
Moghaddam et al. 1997
Group II mGluRs are autoreceptors that inhibit glutamate release. They gave group II agonist prior to PCP (NMDA antagonist). Microdialysis = increase glutamate in PFC and motor sterotypy and decrease in cog performance. All reversed by the agonist and confirmed to not be reduced by a reduction in DA
General animal model stuff (stereotypy not scz), furthermore, although in 2007 clinical trial mGlu2/3 agonists were shown to improve both positive and negative symtpoms, future trials have failed to show effect - maybe its to do with different subcategories of scz?
Moghaddam 98
19pts. Given either placebo or 2 doses of K (0.1&0.5mg/kg). Produced behaviours similar to + and - symptoms (e.g. increases in hallucinations nand impaired in WCST so frontal dysfunction)
Recruitment through adverstising (homogenous population), not causal, ketamine is short-term but scz develops over a long period of time - different mechanisms?
Krystal 1994
Patil 2007 - phase II clinical trial - mGlu2/3 agonists were shown to improve both positive and negative symtpoms, future phase III trials have failed to show effect - maybe its to do with different subcategories of scz? Was only assessed for 4 weeks - AP drugs would need longer.Liu 2012 later showed that this agonist was only effective in pts with a specific SNP in the 5-HT2A receptor - although this second one was candidate gene approach (picked genes to investigate depending on their association with the pharmacological action) - maybe there is even more subcategorisation?
mGlu2/3 agonists
Generated mice expressing only 5% of the normal levels of NR1 subunit of NMDA - survive to adulthood and display behavioural abnormalities, including increased motor activity and deficits in social and sexual interactions, similar to those seen in pharmcologically-induced animal models. These can be ameliorated through treatment with APs (cloz and halo). No changes in DA with microdialysis (only glutamate and not DA) Supports model where reduced NMDAR activity causes SCZ-like behaviour
Maybe microdialysis at the time was not specific enough to pick up the D, could have used FSCV. Havent seen NR1 in GWAS. Used a homologus recombination technique which could lead to ectopic recombination in other places. Did global NDMA hypofuction, rather than in specific circuits. The mice also showed autism-related behaviooural deficits (maybe not representative of scz)
Mohn 1999
Built on previous studies (Mohn 99 in particular) .Transgenic mice using Cre-loxP, in which the essential NR1 subunit of the NMDAR was eliminated in 40-50% of cortical and hippocampal INs in early postnatal development.
Direct SCZ-related symptoms emerged after adolescence (including 1 novelty-induced hyperlocomotion, 2 mating and nest-building deficits, and 3 anhedonia-like and 4 anxiety-like behaviours). Also showed reduction in GAD67 - potential reduction in GABA synthesis - reduction in GABAergic inhibition - cortical disinhibition
not affected à neurodevelopmental hypothesis. Idea that early postnatal inhibition of NMDAR activity could contribute to pathophysiology of SCZ. Also no GAD67 reduction. Early postnatal NMDAR hypofunction in GABAergic IN is needed for SCZ-like behaviour
Limitations
1) Lack of consensus on altered subunit levels in post-mortem brain studies, so hard to ascertain construct validity (also in genetics), 2) based on expression of behavioural analogues for SCZ in humans, 3) which subtupe of GABA IN? 3) do both cortical and hippocampal INs play a role? (both were targetted), 4) only 50% of INs were even targetted
- Belforte et al. 2010
showed this using optogenetics (light-sensitive channelrhodopsin-2 cation channel) to study glutamatergic release in light-stimulated wild-type and DISC1 knockdown neurons in a rodent model. o They found a significantly decreased glutamate release probability in DISC1-/- neurons, indicating the importance of DISC1 in presynaptic glutamate release. o DISC1 also binds to the presynaptic enzyme serine racemase, stabilising it. This is responsible for the production of D-serine, which acts as a coactivator of the NMDAR at the glycine modulatory site. o Therefore, reduction of DISC1 activity would also lead to reduced D-serine production, further contributing to NMDAR hypofunction.
Maher 2012
meta-analysis of NMDAR-modifying drugs. Looked at about 800 subjects from 26 studies. Showed these significantly improves positive + negative and cognitive symptoms when administered in addition to existing AP treatment
Tsai and Lin 2010
Proton magnetic resonance spectroscopy of 10 SCZ and 10 controls – found sig increased glutamine levels in medial PFC of SCZ compared to controls
Bartha et al 1997
MRS. found that chronic SCZ pts have elevated glutamine levels in anterior cingulate cortex compared to healthy controls. The levels of glutamine also positively correlated with the severity of positive symptoms
Cav- 1H-MRS doesn’t distinguish IC vs EC - cant see if glutamate has actually been released. Cant tell if this is a cause or a consequence of the disease
Bustillo et al. 2014
“o In E13.5 medial ganglionic eminence (MGE)/cortex in vitro co-cultures, neurons from the MGE migrated towards the cortex (what happens in development) o However, MGE neurons electroporated (increase the permeability of the cell membrane with an electrical field) with dominant-negative ErbB4 became largely unresponsive to the cortical attractant. (basically if you lose Erb4, you lose this migration) o They also created ErbB4 mouse mutants expressing ErbB4 under a cardiac-specific myosin promoter to avoid embryonic lethality due to failure of myocardial trabeculae formation. o IHC for calbindin, a marker of migrating GABAergic INs, revealed sig fewer migratinc cortical INs in those mice compared to controls” ErbB4 had previously been implicated in susceptibility to scz - this just shows that GABAergic IN migration may be affected by this
Flames et al 2004
electroporated MGE cells in embryonic mouse brain w anti-DISC1 miRNA to cause DISC1 RNAi, + Alexa 546 to visualise transfected cells (in control animals only the latter was transfected). RNAi of DISC1 sig reduced the no of neurons migrating tangentially to the cortex
Steinecke et al 2012
They performed ISH on tissue sections from 10 pairs of subjects (SCZ + healthy controls) w sulphur-35-labelled oligonucleotide probes. They found sig decreased density of GAD67 (GABAergic interneuron marker) labelled neurons in subjects w SCZ. However, a sig limitation of this study was the small sample size. Therefore, developmental defects in GABAergic IN migration may contribute to GABAergic PFC hypofunction, so increase risk of SCZ.
Volk et al 2000
GWAS, 37,000 patients and 113,000 controls. Identified 600 genes associated with SCZ, 83 of which were newly implicated
Incl DRD2 many w NMDA
CACNA1A and CACNA2B
Some involved in immunity - immune enhancers (CD20 B lymphocytes e.g.)
MHC = strongest association
The Schizophrenia Working Group of the Psychiatric Genomics Consortium 2014
“C4 exists as two functionally distinct genes - C4A and C4B - these vary in structure and copy number. They assessed how C4 variation related to RNA expression of C4A and C4B in post mortem human adult brain samples. Found that each C4 allele was associated with schizophrenia risk in proportion to its effect on C4A expression. o Using IHC, and high-resolution structured illumination microscopy imaging, they determined that neuronal dendrites, axons and synapses expressed C4. For example, it co-localises w presynaptic markers VGLUT1/2 and postsynaptic marker PSD-95 o C4 was then shown to promote synapse elimination in mice during the developmentally timed maturation of a neuronal circuit”
Sekar 2016
Looked at the CA3-CA1 pathway in hippocampal slices - transduction of CA1 neurons with lentivirus expressing shRNAs, producing RNAi knockdown of ErbB4, or the addition of the ErbB4 antagonsit AG1478 significantly depressed NMDAR-mediated synaptic tranmission (whole cell postsynaptic patch clamp recordings) ErB4/NGR1 impairment leads to NMDAR dysfunction
Li et al 2007
Generated mice with truncated form of Disc1 under the control of promoter aCaMKII which drived gene expression in forebrain neurons Show enlarged lateral ventricles and decreased numbers of interneurons, hyperactivity, impaired PPI, increased immobililty in FST, social interactions, or spatial memory
Hikida 2007
Found that phosphodiesterase 4 (PDE4) transcripts are elevated in neurons differentiated from iPSCs with the DISC mutation (RNA seq)
Pharma inhibition of PDE4 or or cAMP activation rescues synaptic deficits
Mouse knock in with same DISC1 mutation had elevated PDE4 and synaptic abnormalities and SCZ-like behaviour (was heterozygous KI so some DISC1 = normal)
Inhibition of PDE4 rescued synaptic abnormalities and behaviour
Could present a future drug target - however, PDE4 is ubiquitous - could we really target it?
Kim et al 2021
- Described 10 robust rare varients in MA of whole exomes of 24,000 cases and 97,000 controls.
- Identified 10 ultra-rare coding variants in 10 genes conferring substantial risk for SCZ.
- These had diverse molecular functions including the formation, structure and function of the synapse.
- Protein truncating varients of GRIN2A, TRIO and CACNA1G are examples of genes implicated
- GRIN2A = NMDA receptor subunit = glutamatergic
- Showed for the first time that commonly regulatory variants from GWAs and ultra-rare coding variants disrupt an overlapping set of genes. Gives confidence that integration of results from sequencing consortia with GWAS efforts will have significant value for identifying specific genes beyond what any strategy can achieve on its own - GRIN2A was in this overlap
- There was also overlap in rare variant risk w autism, epilepsy and severe neurodevelopmental disorders - supports neurodevelopmental hypothesis
Singh et al April 2022
meta-analysis incorporating over 21,000 SCZ cases and 20,000 controls
looking at CNVs
Generated definitive evidence for 8 loci - they suggest that these are carried by 1.4% of SCZ
Schizophrenia Working Groups of the Pyschiatric Genomics Consortium 2017
directly reprogramming SCZ pt fibroblasts into HiPSCs and then differentiated these into neurons “Using trans-neuronal spread of rabies virus (stronger synapse = better spread) as an assay of neuronal connectivity, they found this was decreased in these neyrons o Does not necessarily mean synaptic function is decreased (could be limitations fo synaptic activity assays)” “Decreased number of neurites (in keeping with SCZ post-mortem and animal models), decreased PSD95 expression relative to MAP2AB but not to signifiance, and decreased glutamate R expression - Spontaneous neuronal activity unaffected” - Almost 600 genes had expression altered more than 1.3-fold, 25% had been previously implicated in SCZ sig alterations in glutamate, cAMP and WNT signalling, needed for activity-dependent refinement of synaptic connections and LTP “AP loxapine increased glutamate R expression and increased NRG1 expression and increased neuronal connectivity o Other APs did nto but could be dosing needs optimisation” - Data support idea of CZ model where many different combinations of gene dysfunction may disrupt the key pathways affected in SCZ. Increase number of hiPSC cases and see consistent genes affected, and essential pathways will come to light
Bernnand 2011
Treatment resistant SCZ shown to have normal 18F-DOPA uptake in PET studies (and normal ACC glutamate), as opposed to SCZ pts who respond to APs and have elevated 18F-DOPA uptake (and normal ACC glutamate)
Are the differences between glu and DA a cause or consequence of poor AP response?
Demhaja et al 2014
- Found that phosphodiesterase 4 (PDE4) transcripts are elevated in neurons differentiated from iPSCs with the DISC mutation (RNA seq)
- Pharma inhibition of PDE4 or or cAMP activation rescues synaptic deficits
- Mouse knock in with same DISC1 mutation had elevated PDE4 and synaptic abnormalities and SCZ-like behaviour (was heterozygous KI so some DISC1 = normal)
- Inhibition of PDE4 rescued synaptic abnormalities and behaviour
- Could present a future drug target - however, PDE4 is ubiquitous - could we really target it?
Kim et al 2021
- Designed behavioural sensory detection task in mice
- Given tones embedded in white noise. Would poke a port if signal was there and another if it wasnt. Given water reward
- Their confidence in the choices determined by time investments of nose in a given hole
- Hallucination = high confidence false alarm
- Hallucination-like percepts in mice increased with 1) ketas
Schmack et al 2021
Bitopertin
glycine reuptake inhibitor – 2014 RCT caused improvement of negative symtpoms. However, Roche reported failed endponts in 2 phase III trials in 2014 and it was discontinued
i.Yilmaz 2021
Yoshimizu et al 2015
i.used iPSCs to look at functional phenotype associated with an SNP identified in GWAS
Generated iPSCs with the allele and without
Found higher expression of CACNA1C in those carrying, confirmed by higher VGCC in those carrying by whole cell recording
Although iPSCs not same as in vivo
Lintunen 2021
longitudinal prospective study showed patients on calcium channel blockers had reduction in scz rehopsitalisation (although concomitant AP was allowed)
Karunakaran 2019 (Nature)
was all computational analysis
Identified the protein-protein interactions (interactome) of scz associated genes.
Looked at which drugs target the interactome and whether they show anticorrelation with the transcriptomic signature seen in the disease (transcriptomics)
Identified a shortlist of 12 potential drugs (top 3 were cromoglicic acid, acetazolamide and cinnarizine).
No actual screening to see if such drugs alter scz phenotypes – could have done animal model for each repurposed drugs
