I'm pleased to share our review of the synaptic hypothesis of schizophrenia, out now in @molpsychiatry, with Oliver Howes,.@MRC_LMS @KingsIoPPN @QMUL_WIPH.We reflect on >4 decades of evidence, and use this to propose an updated synaptic hypothesis.

RT @sameerjauhar: Out in Translational Psychiatry , we examine the relationship between 2 of the major players thou….

RT @VernonLab_KCL: Curious about hiPSC to study psychiatric disorders? Check our consensus paper from the ECNP iPS….

Some unanswered questions:. Are synaptic alterations present in unmedicated first-episode, and clinical high-risk?. Are there post-synaptic deficits in vivo?. Does synaptic profile differ across distinct clinical sub-types?. Are synaptic and dopaminergic markers linked in vivo?.

➡️Genetic variants render synapses vulnerable to pruning. ➡️Stress/immune activation causes ⬆️pruning by glia. ➡️This causes EI imbalance, so cog+neg symptoms; and disinhibited projections to striatum, so hyperdopaminergia, so psychosis. ➡️Stress of psychosis causes more pruning.

With these lines of evidence, we can provide a further update to the synaptic hypothesis of schizophrenia (version III)

We do not find the normal association between synaptic density and glutamate in schizophrenia. This suggests a loss of glutamatergic synapses and/or lower glutamatergic-to-GABAergic synapse ratio, potentially contributing to observed EI imbalance.

In [11C]UCB-J PET, we now have a tool to measure a marker of synaptic density in the living human brain. We find evidence for reduced synaptic density in both cortical and subcortical regions in schizophrenia compared to controls:.

Synaptic changes may underlie imaging findings. In preclinical studies, stress is linked to reduced grey matter volume, dendritic volume and spine density. And dendritic volume loss and MRI-estimated grey matter volume loss are strongly associated (R2>0.9).

Meta-analyses of neuroimaging studies indicate lower cortical grey matter volumes and lower/unaltered subcortical volumes in schizophrenia, from first episode. Clinical high risk for schizophrenia subjects show similar patterns of lower grey matter volumes and cortical thinning.

Human iPSC-derived neurons show lower SV2 and synapsin I puncta density, vesicle release, PSD-95 protein levels and dendritic spine density in schizophrenia, and elevated, complement-dependent elimination of synaptic structures. Key work by @SellgrenLab ➡️

Post-mortem studies of schizophrenia show lower levels of markers of both pre- and post-synaptic elements in patients, particularly affecting cortical brain regions. See key meta-analysis of pre-synaptic markers by @eosimo.

In schizophrenia, genetic association studies point to alterations in genes involved in synaptic organisation, differentiation, transmission, and signalling, and complement expression 🧬. In preclinical studies, maternal infection 🫄🏿+ stress 😮‍💨 alter synaptic development.

A feature of neurodevelopment is the elimination of inactive synapses by glia. These are tagged by complement protein C1q, which triggers a cascade leading to the phagocytosis of synapses by glia

Human postmortem studies indicate that synaptic density peaks in early childhood, followed by intermediate levels in adolescence and early adulthood, and lower but stable levels in the remaining adulthood

Since versions I and II, methodological advances have produced a step-change in the available evidence. So now we can consider the evidence for a mechanism for faulty synaptic elimination in schizophrenia, and how this is linked to genetic or environmental risk factors.

Crucially, they added spatial specificity, proposing excessive cortical synaptic pruning and insufficient subcortical pruning in schizophrenia, given new neuroimaging evidence for altered brain volumes (CT, MRI) and frontal hypometabolism (PET, SPECT, 31P-MRS, 133Xe inhalation).

In 1994, Keshavan, Anderson and Pettegrew produced the 2nd major version of the synaptic hypothesis

So, maybe an error in developmental synaptic loss leads to schizophrenia. But much of the supporting evidence was indirect.

Feinberg noted that potential markers of synaptic function (e.g. EEG wave amplitude, brain metabolic rate) were altered in schizophrenia, and the illness typically emerges during a developmental time window involving extensive synaptic loss (late adolescence/early adulthood).

In 1982, Irwin Feinberg proposed arguably the 1st synaptic hypothesis of schizophrenia, arguing that the illness is caused by a loss of synaptic density