Alterations in Schizophrenia-Associated Genes Can Lead to Increased Power in Delta Oscillations

Therapeutic dose prediction of α5-GABA receptor modulation from simulated EEG of depression severity

Treatment for major depressive disorder (depression) often has partial efficacy and a large portion of patients are treatment resistant. Recent studies implicate reduced somatostatin (SST) interneuron inhibition in depression, and new pharmacology boosting this inhibition via positive allosteric modulators of α5-GABAA receptors (α5-PAM) offers a promising effective treatment. However, testing the effect of α5-PAM on human brain activity is limited, meriting the use of detailed simulations. We utilized our previous detailed computational models of human depression microcircuits with reduced SST interneuron inhibition and α5-PAM effects, to simulate EEG of individual microcircuits across depression severity and α5-PAM doses. We developed machine learning models that predicted optimal dose from EEG with high accuracy and recovered microcircuit activity and EEG. This study provides dose prediction models for α5-PAM administration based on EEG biomarkers of depression severity. Given limitations in doing the above in the living human brain, the results and tools we developed will facilitate translation of α5-PAM treatment to clinical use.

Genetic mechanisms for impaired synaptic plasticity in schizophrenia revealed by computational modelling

Schizophrenia phenotypes are suggestive of impaired cortical plasticity in the disease, but the mechanisms of these deficits are unknown. Genomic association studies have implicated a large number of genes that regulate neuromodulation and plasticity, indicating that the plasticity deficits have a genetic origin. Here, we used biochemically detailed computational modelling of post-synaptic plasticity to investigate how schizophrenia-associated genes regulate long-term potentiation (LTP) and depression (LTD). We combined our model with data from post-mortem mRNA expression studies (CommonMind gene-expression datasets) to assess the consequences of altered expression of plasticity-regulating genes for the amplitude of LTP and LTD. Our results show that the expression alterations observed post mortem, especially those in anterior cingulate cortex, lead to impaired PKA-pathway-mediated LTP in synapses containing GluR1 receptors. We validated these findings using a genotyped EEG dataset where polygenic risk scores for synaptic and ion channel-encoding genes as well as modulation of visual evoked potentials (VEP) were determined for 286 healthy controls. Our results provide a possible genetic mechanism for plasticity impairments in schizophrenia, which can lead to improved understanding and, ultimately, treatment of the disorder.

Age-dependent increased sag amplitude in human pyramidal neurons dampens baseline cortical activity

Aging involves various neurobiological changes, although their effect on brain function in humans remains poorly understood. The growing availability of human neuronal and circuit data provides opportunities for uncovering age-dependent changes of brain networks and for constraining models to predict consequences on brain activity. Here we found increased sag voltage amplitude in human middle temporal gyrus layer 5 pyramidal neurons from older subjects and captured this effect in biophysical models of younger and older pyramidal neurons. We used these models to simulate detailed layer 5 microcircuits and found lower baseline firing in older pyramidal neuron microcircuits, with minimal effect on response. We then validated the predicted reduced baseline firing using extracellular multielectrode recordings from human brain slices of different ages. Our results thus report changes in human pyramidal neuron input integration properties and provide fundamental insights into the neuronal mechanisms of altered cortical excitability and resting-state activity in human aging.

Excitatory and inhibitory effects of HCN channel modulation on excitability of layer V pyramidal cells

Dendrites of cortical pyramidal cells are densely populated by hyperpolarization-activated cyclic nucleotide-gated (HCN) channels, a.k.a. Ih channels. Ih channels are targeted by multiple neuromodulatory pathways, and thus are one of the key ion-channel populations regulating the pyramidal cell activity. Previous observations and theories attribute opposing effects of the Ih channels on neuronal excitability due to their mildly hyperpolarized reversal potential. These effects are difficult to measure experimentally due to the fine spatiotemporal landscape of the Ih activity in the dendrites, but computational models provide an efficient tool for studying this question in a reduced but generalizable setting. In this work, we build upon existing biophysically detailed models of thick-tufted layer V pyramidal cells and model the effects of over- and under-expression of Ih channels as well as their neuromodulation. We show that Ih channels facilitate the action potentials of layer V pyramidal cells in response to proximal dendritic stimulus while they hinder the action potentials in response to distal dendritic stimulus at the apical dendrite. We also show that the inhibitory action of the Ih channels in layer V pyramidal cells is due to the interactions between Ih channels and a hot zone of low voltage-activated Ca2+ channels at the apical dendrite. Our simulations suggest that a combination of Ih-enhancing neuromodulation at the proximal part of the apical dendrite and Ih-inhibiting modulation at the distal part of the apical dendrite can increase the layer V pyramidal excitability more than either of the two alone. Our analyses uncover the effects of Ih-channel neuromodulation of layer V pyramidal cells at a single-cell level and shed light on how these neurons integrate information and enable higher-order functions of the brain.

Brain signal predictions from multi-scale networks using a linearized framework

Simulations of neural activity at different levels of detail are ubiquitous in modern neurosciences, aiding the interpretation of experimental data and underlying neural mechanisms at the level of cells and circuits. Extracellular measurements of brain signals reflecting transmembrane currents throughout the neural tissue remain commonplace. The lower frequencies (≲ 300Hz) of measured signals generally stem from synaptic activity driven by recurrent interactions among neural populations and computational models should also incorporate accurate predictions of such signals. Due to limited computational resources, large-scale neuronal network models (≳ 10⁶ neurons or so) often require reducing the level of biophysical detail and account mainly for times of action potentials (‘spikes’) or spike rates. Corresponding extracellular signal predictions have thus poorly accounted for their biophysical origin.

Here we propose a computational framework for predicting spatiotemporal filter kernels for such extracellular signals stemming from synaptic activity, accounting for the biophysics of neurons, populations, and recurrent connections. Signals are obtained by convolving population spike rates by appropriate kernels for each connection pathway and summing the contributions. Our main results are that kernels derived via linearized synapse and membrane dynamics, distributions of cells, conduction delay, and volume conductor model allow for accurately capturing the spatiotemporal dynamics of ground truth extracellular signals from conductance-based multicompartment neuron networks. One particular observation is that changes in the effective membrane time constants caused by persistent synapse activation must be accounted for.

The work also constitutes a major advance in computational efficiency of accurate, biophysics-based signal predictions from large-scale spike and rate-based neuron network models drastically reducing signal prediction times compared to biophysically detailed network models. This work also provides insight into how experimentally recorded low-frequency extracellular signals of neuronal activity may be approximately linearly dependent on spiking activity. A new software tool LFPykernels serves as a reference implementation of the framework.

Understanding the brain’s function and activity in healthy and pathological states across spatial scales and times spanning entire lives is one of humanity’s great undertakings. In experimental and clinical work probing the brain’s activity, a variety of electric and magnetic measurement techniques are routinely applied. However interpreting the extracellularly measured signals remains arduous due to multiple factors, mainly the large number of neurons contributing to the signals and complex interactions occurring in recurrently connected neuronal circuits. To understand how neurons give rise to such signals, mechanistic modeling combined with forward models derived using volume conductor theory has proven to be successful, but this approach currently does not scale to the systems level (encompassing millions of neurons or more) where simplified or abstract neuron representations typically are used. Motivated by experimental findings implying approximately linear relationships between times of neuronal action potentials and extracellular population signals, we provide a biophysics-based method for computing causal filters relating spikes and extracellular signals that can be applied with spike times or rates of large-scale neuronal network models for predictions of population signals without relying on ad hoc approximations.

The effect of alterations of schizophrenia-associated genes on gamma band oscillations

Abnormalities in the synchronized oscillatory activity of neurons in general and, specifically in the gamma band, might play a crucial role in the pathophysiology of schizophrenia. While these changes in oscillatory activity have traditionally been linked to alterations at the synaptic level, we demonstrate here, using computational modeling, that common genetic variants of ion channels can contribute strongly to this effect. Our model of primary auditory cortex highlights multiple schizophrenia-associated genetic variants that reduce gamma power in an auditory steady-state response task. Furthermore, we show that combinations of several of these schizophrenia-associated variants can produce similar effects as the more traditionally considered synaptic changes. Overall, our study provides a mechanistic link between schizophrenia-associated common genetic variants, as identified by genome-wide association studies, and one of the most robust neurophysiological endophenotypes of schizophrenia.

Reduced inhibition in depression impairs stimulus processing in human cortical microcircuits

Cortical processing depends on finely tuned excitatory and inhibitory connections in neuronal microcircuits. Reduced inhibition by somatostatin-expressing interneurons is a key component of altered inhibition associated with treatment-resistant major depressive disorder (depression), which is implicated in cognitive deficits and rumination, but the link remains to be better established mechanistically in humans. Here we test the effect of reduced somatostatin interneuron-mediated inhibition on cortical processing in human neuronal microcircuits using a data-driven computational approach. We integrate human cellular, circuit, and gene expression data to generate detailed models of human cortical microcircuits in health and depression. We simulate microcircuit baseline and response activity and find a reduced signal-to-noise ratio and increased false/failed detection of stimuli due to a higher baseline activity in depression. We thus apply models of human cortical microcircuits to demonstrate mechanistically how reduced inhibition impairs cortical processing in depression, providing quantitative links between altered inhibition and cognitive deficits.

Brain-specific heterozygous loss-of-function of ATP2A2, endoplasmic reticulum Ca2+ pump responsible for Darier's disease, causes behavioral abnormalities and a hyper-dopaminergic state

A report of a family of Darier's disease with mood disorders drew attention when the causative gene was identified as ATP2A2 (or SERCA2), which encodes a Ca2+ pump on the endoplasmic reticulum (ER) membrane and is important for intracellular Ca2+ signaling. Recently, it was found that loss-of-function mutations of ATP2A2 confer a risk of neuropsychiatric disorders including depression, bipolar disorder and schizophrenia. In addition, a genome-wide association study found an association between ATP2A2 and schizophrenia. However, the mechanism of how ATP2A2 contributes to vulnerability to these mental disorders is unknown. Here, we analyzed Atp2a2 heterozygous brain-specific conditional knockout (hetero cKO) mice. The ER membranes prepared from the hetero cKO mouse brain showed decreased Ca2+ uptake activity. In Atp2a2 heterozygous neurons, decays of cytosolic Ca2+ level were slower than control neurons after depolarization. The hetero cKO mice showed altered behavioral responses to novel environments and impairments in fear memory, suggestive of enhanced dopamine signaling. In vivo dialysis demonstrated that extracellular dopamine levels in the NAc were indeed higher in the hetero cKO mice. These results altogether indicate that the haploinsufficiency of Atp2a2 in the brain causes prolonged cytosolic Ca2+ transients, which possibly results in enhanced dopamine signaling, a common feature of mood disorders and schizophrenia. These findings elucidate how ATP2A2 mutations causing a dermatological disease may exert their pleiotropic effects on the brain and confer a risk for mental disorders.

Computation of the electroencephalogram (EEG) from network models of point neurons

The electroencephalogram (EEG) is a major tool for non-invasively studying brain function and dysfunction. Comparing experimentally recorded EEGs with neural network models is important to better interpret EEGs in terms of neural mechanisms. Most current neural network models use networks of simple point neurons. They capture important properties of cortical dynamics, and are numerically or analytically tractable. However, point neurons cannot generate an EEG, as EEG generation requires spatially separated transmembrane currents. Here, we explored how to compute an accurate approximation of a rodent's EEG with quantities defined in point-neuron network models. We constructed different approximations (or proxies) of the EEG signal that can be computed from networks of leaky integrate-and-fire (LIF) point neurons, such as firing rates, membrane potentials, and combinations of synaptic currents. We then evaluated how well each proxy reconstructed a ground-truth EEG obtained when the synaptic currents of the LIF model network were fed into a three-dimensional network model of multicompartmental neurons with realistic morphologies. Proxies based on linear combinations of AMPA and GABA currents performed better than proxies based on firing rates or membrane potentials. A new class of proxies, based on an optimized linear combination of time-shifted AMPA and GABA currents, provided the most accurate estimate of the EEG over a wide range of network states. The new linear proxies explained 85-95% of the variance of the ground-truth EEG for a wide range of network configurations including different cell morphologies, distributions of presynaptic inputs, positions of the recording electrode, and spatial extensions of the network. Non-linear EEG proxies using a convolutional neural network (CNN) on synaptic currents increased proxy performance by a further 2-8%. Our proxies can be used to easily calculate a biologically realistic EEG signal directly from point-neuron simulations thus facilitating a quantitative comparison between computational models and experimental EEG recordings.

Using protein turnover to expand the applications of transcriptomics

RNA expression and protein abundance are often at odds when measured in parallel, raising questions about the functional implications of transcriptomics data. Here, we present the concept of persistence, which attempts to address this challenge by combining protein half-life data with RNA expression into a single metric that approximates protein abundance. The longer a protein's half-life, the more influence it can have on its surroundings. This data offers a valuable opportunity to gain deeper insight into the functional meaning of transcriptome changes. We demonstrate the application of persistence using schizophrenia (SCZ) datasets, where it greatly improved our ability to predict protein abundance from RNA expression. Furthermore, this approach successfully identified persistent genes and pathways known to have impactful changes in SCZ. These results suggest that persistence is a valuable metric for improving the functional insight offered by transcriptomics data, and extended application of this concept could advance numerous research fields.

Biophysically detailed forward modeling of the neural origin of EEG and MEG signals

Electroencephalography (EEG) and magnetoencephalography (MEG) are among the most important techniques for non-invasively studying cognition and disease in the human brain. These signals are known to originate from cortical neural activity, typically described in terms of current dipoles. While the link between cortical current dipoles and EEG/MEG signals is relatively well understood, surprisingly little is known about the link between different kinds of neural activity and the current dipoles themselves. Detailed biophysical modeling has played an important role in exploring the neural origin of intracranial electric signals, like extracellular spikes and local field potentials. However, this approach has not yet been taken full advantage of in the context of exploring the neural origin of the cortical current dipoles that are causing EEG/MEG signals. Here, we present a method for reducing arbitrary simulated neural activity to single current dipoles. We find that the method is applicable for calculating extracranial signals, but less suited for calculating intracranial electrocorticography (ECoG) signals. We demonstrate that this approach can serve as a powerful tool for investigating the neural origin of EEG/MEG signals. This is done through example studies of the single-neuron EEG contribution, the putative EEG contribution from calcium spikes, and from calculating EEG signals from large-scale neural network simulations. We also demonstrate how the simulated current dipoles can be used directly in combination with detailed head models, allowing for simulated EEG signals with an unprecedented level of biophysical details. In conclusion, this paper presents a framework for biophysically detailed modeling of EEG and MEG signals, which can be used to better our understanding of non-inasively measured neural activity in humans.

A unified computational model for cortical post-synaptic plasticity

Signalling pathways leading to post-synaptic plasticity have been examined in many types of experimental studies, but a unified picture on how multiple biochemical pathways collectively shape neocortical plasticity is missing. We built a biochemically detailed model of post-synaptic plasticity describing CaMKII, PKA, and PKC pathways and their contribution to synaptic potentiation or depression. We developed a statistical AMPA-receptor-tetramer model, which permits the estimation of the AMPA-receptor-mediated maximal synaptic conductance based on numbers of GluR1s and GluR2s predicted by the biochemical signalling model. We show that our model reproduces neuromodulator-gated spike-timing-dependent plasticity as observed in the visual cortex and can be fit to data from many cortical areas, uncovering the biochemical contributions of the pathways pinpointed by the underlying experimental studies. Our model explains the dependence of different forms of plasticity on the availability of different proteins and can be used for the study of mental disorder-associated impairments of cortical plasticity.

Disruption of gamma-delta relationship related to working memory deficits in first-episode psychosis

Working memory (WM) deficits constitute a core symptom of schizophrenia. Inadequacy of WM maintenance in schizophrenia has been reported to reflect abnormalities in the excitation/inhibition (E/I) balance between pyramidal neurons and parvalbumin basket cells, which may explain alterations of the dynamics of gamma and delta oscillations. To address this issue, we assessed event-related gamma (35-45 Hz) and delta (0.5-4 Hz) oscillatory responses in a visual n-back WM task in patients with first-episode psychosis (FEP) and healthy controls (HC). Periodicity analyses of oscillations were computed to explore the relationship between the psychiatric status and the WM load-related processes reflected by each frequency range. The correspondence between nested delta-gamma oscillations was estimated to assess the strength of the frontal E/I balance. In HC, gamma oscillations were synchronized by the stimulus in a 50-150 ms time range for all tasks, and periodicity of the delta cycle was comparable between the tasks. In addition, synchronization of gamma oscillations in HC occurred at the maximal descending phase of the delta cycle half-period, supporting the coexistence of delta-nested gamma oscillations. Compared with controls, FEP patients showed a lack of gamma synchronization independently of the nature of the task, and the period of delta oscillation increased significantly with the difficulty of the WM task. We thus demonstrated in FEP an inability to encode multiple items in short-term memory associated with abnormalities in the relationship between oscillations related to the difficulty of the WM task. These results argue in favor of a dysfunction of the E/I balance in psychosis.

The Role of Parvalbumin-positive Interneurons in Auditory Steady-State Response Deficits in Schizophrenia

Despite an increasing body of evidence demonstrating subcellular alterations in parvalbumin-positive (PV⁺) interneurons in schizophrenia, their functional consequences remain elusive. Since PV⁺ interneurons are involved in the generation of fast cortical rhythms, these changes have been hypothesized to contribute to well-established alterations of beta and gamma range oscillations in patients suffering from schizophrenia. However, the precise role of these alterations and the role of different subtypes of PV⁺ interneurons is still unclear. Here we used a computational model of auditory steady-state response (ASSR) deficits in schizophrenia. We investigated the differential effects of decelerated synaptic dynamics, caused by subcellular alterations at two subtypes of PV⁺ interneurons: basket cells and chandelier cells. Our simulations suggest that subcellular alterations at basket cell synapses rather than chandelier cell synapses are the main contributor to these deficits. Particularly, basket cells might serve as target for innovative therapeutic interventions aiming at reversing the oscillatory deficits.

Computational Modeling of Genetic Contributions to Excitability and Neural Coding in Layer V Pyramidal Cells: Applications to Schizophrenia Pathology

Pyramidal cells in layer V of the neocortex are one of the most widely studied neuron types in the mammalian brain. Due to their role as integrators of feedforward and cortical feedback inputs, they are well-positioned to contribute to the symptoms and pathology in mental disorders-such as schizophrenia-that are characterized by a mismatch between the internal perception and external inputs. In this modeling study, we analyze the input/output properties of layer V pyramidal cells and their sensitivity to modeled genetic variants in schizophrenia-associated genes. We show that the excitability of layer V pyramidal cells and the way they integrate inputs in space and time are altered by many types of variants in ion-channel and Ca²⁺ transporter-encoding genes that have been identified as risk genes by recent genome-wide association studies. We also show that the variability in the output patterns of spiking and Ca²⁺ transients in layer V pyramidal cells is altered by these model variants. Importantly, we show that many of the predicted effects are robust to noise and qualitatively similar across different computational models of layer V pyramidal cells. Our modeling framework reveals several aspects of single-neuron excitability that can be linked to known schizophrenia-related phenotypes and existing hypotheses on disease mechanisms. In particular, our models predict that single-cell steady-state firing rate is positively correlated with the coding capacity of the neuron and negatively correlated with the amplitude of a prepulse-mediated adaptation and sensitivity to coincidence of stimuli in the apical dendrite and the perisomatic region of a layer V pyramidal cell. These results help to uncover the voltage-gated ion-channel and Ca²⁺ transporter-associated genetic underpinnings of schizophrenia phenotypes and biomarkers.

Biophysical Psychiatry-How Computational Neuroscience Can Help to Understand the Complex Mechanisms of Mental Disorders

The brain is the most complex of human organs, and the pathophysiology underlying abnormal brain function in psychiatric disorders is largely unknown. Despite the rapid development of diagnostic tools and treatments in most areas of medicine, our understanding of mental disorders and their treatment has made limited progress during the last decades. While recent advances in genetics and neuroscience have a large potential, the complexity and multidimensionality of the brain processes hinder the discovery of disease mechanisms that would link genetic findings to clinical symptoms and behavior. This applies also to schizophrenia, for which genome-wide association studies have identified a large number of genetic risk loci, spanning hundreds of genes with diverse functionalities. Importantly, the multitude of the associated variants and their prevalence in the healthy population limit the potential of a reductionist functional genetics approach as a stand-alone solution to discover the disease pathology. In this review, we outline the key concepts of a "biophysical psychiatry," an approach that employs large-scale mechanistic, biophysics-founded computational modelling to increase transdisciplinary understanding of the pathophysiology and strive toward robust predictions. We discuss recent scientific advances that allow a synthesis of previously disparate fields of psychiatry, neurophysiology, functional genomics, and computational modelling to tackle open questions regarding the pathophysiology of heritable mental disorders. We argue that the complexity of the increasing amount of genetic data exceeds the capabilities of classical experimental assays and requires computational approaches. Biophysical psychiatry, based on modelling diseased brain networks using existing and future knowledge of basic genetic, biochemical, and functional properties on a single neuron to a microcircuit level, may allow a leap forward in deriving interpretable biomarkers and move the field toward novel treatment options.