1. Introduction
iInduced pluripotent stem cells (iIPSCs) may be derived from individuals with various monogenic and polygenic disorders providing a precious resource for understanding the underlying molecular and cellular mechanisms through personalized models of psychiatric diseases. Moreover, cells are reprogrammed to a very early stage of development; thus, they can provide information on developmental or differentiation defects, as well as the temporal sequence of events in the early stages of disease progression. The timeline for iPSC differentiation into central nervous system (CNS) cells follows the same trajectory as in the developing embryo; thus, these cells are handy tools to study psychiatric disorders, including schizophrenia (SZ) and neurodevelopmental disorders, such as autism spectrum disorder (ASD), where neurodevelopmental alterations are believed to feature. Moreover, brain organoid (BO) —which contains multiple organ-specific cell types with function and spatial organization similar to a human brain—has opened up a new avenue for investigating psychiatric disorders.
Finally, iPSCs could offer the possibility of overcoming the ethical concern of destroying a potential human life and the technical limitations of embryonic stem cells (ESC) use. Producing a large number of ESCs or deriving them from patients with diseases is impossible; thus, their use is limited to studies of normal cellular function or the introduction of known engineered genetic changes.
In the following paragraphs, overview of preliminary studies that used iPSCs in severe psychiatric disorders will be provided.
2. Schizophrenia
Schizophrenia (SZ) is a severe mental illness that is associated with subtle brain cortical structure changes and is characterized by the following symptoms: delusions, hallucinations, disorganized thinking, abnormal motor behavior, and negative symptoms.
In patients with SZ, iPSC studies are mainly devoted to investigating the pathogenesis of the illness.
The first human study by Brennand et al. found diminished neuronal connectivity, neurite number, and PSD95 synaptic protein levels, as well as altered gene expression profiles of the cAMP, wingless-related integration site (Wnt) signaling pathways, and glutamate receptors in iPSC-derived neurons of patients with SZ, compared to controls. The Wnts are secreted factors that regulate cell proliferation and differentiation during embryonic development and act by activating diverse signaling cascades inside the target cells. These alterations are consistent with those described in the postmortem brain of patients with SZ and animal models of SZ and could be ameliorated with the antipsychotic loxapine
[1]. More recent studies confirm the presence of early neurodevelopmental alterations. Migration capacity was impaired in induced neural stem cells (iNSCs) from patients with SZ compared to iNSCs from healthy controls and genetic high risk (GHR) individuals
[2]. Moreover, iNSCs from a GHR individual who later developed SZ showed migratory impairment similar to SZ-iNSCs.
Narla et al. identify the nFGFR1 signaling, which integrates signals from diverse pathways that are characterized by SZ-linked mutations, as a common altered mechanism and a potential therapeutic target in investigated patients
[3].
Hippocampal neurogenesis aberrations have been implicated in SZ pathogenesis. Thus, iPSC-based case/control studies investigated the early development in iPSC-derived hippocampal neurons of patients with SZ. The dentate gyrus (DG) of the hippocampus is one of the two areas of the brain where neurogenesis continues throughout life and its generated neurons play a key role in learning and memory. Yu
[4] described reduced neuronal activity and reduced spontaneous neurotransmitter levels released in the dentate granule neurons from SZ-iPSC-derived hippocampal NPCs. Moreover, hippocampal CA3 neurons that are derived from these patients had altered network connectivity when co-cultured with human dentate granule neurons
[5].
The mitochondrial tricarboxylic acid (TCA) cycle (also referred to as the Krebs cycle) is known to satisfy the energetic neuron demand broadly, thereby generating approximately 90% of cellular reactive oxygen species (ROS). Mitochondrial dysfunction enhances ROS formation, negatively impacting mitochondrial function and leading to oxidative damage that affects several cellular components, such as lipids, DNA, and proteins. The human fetal brain is particularly vulnerable to oxidative damage due to its high oxygen consumption, high ROS basal level production, and less developed antioxidant defense mechanisms relative to the adult brain. Collectively, the studies in iPSC-NPCs and neurons of patients with SZ showed perturbed mitochondrial respiration and morphology along with signs of increased oxidative damage compared to controls
[6].
BO studies in patients with SZ have revealed several phenotypes that may be associated with early neurodevelopmental defects, such as nFGFR1 signaling
[7] and immune response alterations
[8]. The BOs produced with iPSCs of eight patients with SZ and eight healthy controls were compared with a transcriptomic approach to identify disease-specific differences
[9]. Significantly, RNA sequencing demonstrated aberrant gene expression in pathways involved in synaptic biology, nervous system development, immune response, mitochondrial function, and excitatory and inhibitory neurotransmission modulation.
Notaras et al. observed that in SZ BO, about 2.62% of global proteome was differentially regulated, with a depletion of factors that support neuronal development, differentiation, and/or function
[10]. Moreover, the differential regulation of two novel specific disease candidates identified through genome-wide association study (GWAS) arrays (namely, Pleiotrophin and Podocalyxin) was observed
[10]. The SZ organoids exhibited altered progenitor survival and disrupted neurogenesis, yielding fewer neurons within developing cortical areas
[11]. Single-cell sequencing revealed that SZ progenitors were specifically depleted of neuronal programming factors. It was suggested that multiple mechanisms in SZ-derived BO converge upon brain developmental pathways and contribute to raising the risk of developing SZ
[11].
In addition, CRISPR-mediated gene editing has contributed to the understanding of risk variants for SZ, which include common single-nucleotide variants (SNVs) with small effect sizes and copy-number variations (CNVs) with greater penetrance. By integrating GWAS and postmortem brain expression quantitative trait loci (eQTL) studies, common variants that affect SZ risk through regulation of gene expression have been identified. Schrode
[12] applied CRISPR-mediated gene editing, activation, and repression technologies to study one putative SZ-eQTL (
FURINrs4702) and four top-ranked SZ-eQTL genes (
FURIN,
SNAP91,
TSNARE1,
CLCN3), demonstrating a synergistic effect between SZ-eQTL genes that converge on synaptic function. The single and combinatorial CRISPRa/i manipulations of common variant loci and genes in human neurons suggest that synergy between risk variants may impact SZ risk.
It should be noted that some of the biological alterations that emerged from iPSC-based studies appeared not to be specific to SZ but to be shared by several disorders. Particularly, studies have identified in disrupted in schizophrenia 1 (DISC1) a gene associated with diverse mental disorders, following the finding that its coding sequence is interrupted by a balanced translocation in a Scottish family, in which the translocation co-segregates with SZ, bipolar disorder, and major depressive disorder
[13]. The variety of disorders in subjects that harbor the translocation supports the hypothesis that the translocation leads to a subtle disruption in neural development that predisposes to mental disorders by increasing vulnerability to other environmental and genetic risk factors. An isogenic iPSC model demonstrated that DISC1 gene disruption at the site of the balanced translocation causes loss of expression of longer DISC1 transcripts, which increases baseline Wnt signaling and alters the transcriptional profile of neural progenitor cells and neurons, resulting in neurodevelopmental disorders
[13].
Generated cerebral organoids by DISC1-disrupted iPSCs showed disorganized structural morphology and impaired proliferation, which is phenocopied by Wnt agonism and rescued by Wnt antagonism
[14]. The shared BO morphology and gene expression changes with DISC1 interruption and Wnt agonism highlight the link between DISC1 mutation, Wnt signaling abnormalities, and neuropsychiatric diseases.
Overall, the iPSC-based studies in patients with SZ suggest the presence of an early genetic dysregulation leading to altered neuronal and brain development.
3. Major Depressive Disorder
Major Depressive Disorder (MDD) is a common and severe mental disorder that is characterized by one or more major depressive episodes, defined as discrete periods of at least 2-week duration, but generally longer, in which at least five of the following nine symptoms are present: depressed mood, loss of pleasure or interest, significant appetite disturbance/body weight change, sleep disturbance, loss of energy, psychomotor changes, excessive guilt and/or worthlessness, decreased concentration, and recurring death and/or suicidal thoughts. Unsatisfactory response rates to currently approved antidepressant drugs, which are effective in approximately half of the treated patients, contribute to the heavy medical and economic burden of MDD
[15].
Therefore, iPSC-based studies on MDD have focused on treatment-resistant depression (TRD), which is defined as the failure to achieve a reduced baseline depressive symptomatology of at least 50% after at least two antidepressant treatment trials of adequate dosage and duration.
A study on patients with MDD showed that iPSCs derived from patients who are non-remitted with selective serotonin reuptake inhibitor (SSRI) antidepressant therapy displayed serotonin-induced hyperactivity downstream of upregulated excitatory serotonergic receptors (5-HT2A and 5-HT7) in contrast to what was seen in healthy and remitted patient-derived iPSCs. Lurasidone, which is a high-affinity 5-HT2A and 5-HT7 antagonist, partially rescued 5-HT-induced hyperactivity in non-remitted patient-derived iPSCs [
57].
Another study of the same research group revealed no significant differences in serotonin release/reuptake or genes related to serotonin pathways in iPSCs and serotonergic neurons of SSRI-treated patients with MDD
[16]. However, compared to healthy controls and remitted patients, non-remitted patient-derived serotonergic neurons exhibited altered neurite growth and morphology downstream of lowered key protocadherin-α gene expression
[16]. The protocadherin-α family includes cell adhesion molecules that are required for serotonergic projections to appropriately innervate target brain areas so that their loss from serotonergic neurons leads to unbalanced distributions of serotonergic axons
[17].
In MDD lymphoblastoid cell lines (LCLs) reprogrammed to iPSC and differentiated into cortical neurons, Avior
[18] identified bupropion response-specific biomarkers, including synaptic connectivity and morphology changes as well as specific gene expression. The LCLs utilized in this platform can be easily obtained from patients, promoting personalized treatment.
Ketamine, a glutamate N-methyl-d-aspartate (NMDA) receptor antagonist, is a promising treatment for patients with TRD, providing significant depressive symptom improvement within hours of intravenous administration. An iPSC-based study of ketamine’s antidepressant mechanisms of action supported the leading hypothesis that ketamine can enhance structural plasticity through the AMPA glutamate receptor-driven rise of Brain-Derived Neurotrophic Factor levels, leading to an increased synaptic number and function in the prefrontal cortex
[19].
The mechanisms underlying the prolonged antidepressant effects (1 week) after a single ketamine infusion are poorly understood. The triggering of synaptic function and plasticity was hypothesized because a ketamine half-life of approximately two h cannot explain the long-lasting antidepressant effects. A study that supported this view demonstrated that dopaminergic neurons that are differentiated from iPSCs obtained from healthy donors who are exposed for 6 h to ketamine metabolite (2R, 6R)-hydroxynorketamine (half-life 6–12 h) produced dendrite outgrowth when measured three days after exposure
[20].
4. Bipolar Disorders
Bipolar disorder (BD) is characterized by the recurring swing between the opposite mood states of major depressive episodes and mania (BD type I) or hypomania (BD type II). Mania is more severe than hypomania and causes markedly impaired social or occupational functioning or necessitates hospitalization. Both manic and hypomanic episodes are characterized by expansive or irritable mood, increased activity, self-confidence, talkativeness, distractibility, decreased need for sleep, racing thoughts, and poor judgment. First-line BD treatment includes mood stabilizers, mainly lithium. The IPSC-based studies on BD are focused on elucidating the pathogenetic mechanism also by examining the detailed effects of lithium and/or other BD drug exposure on the different signaling pathways.
The first study that investigated iPSC-derived neuronal cells from patients with BD revealed that the gene expression encoding membrane-bound receptors and ion channels, particularly transcripts involved in calcium signaling, was significantly increased in the neurons generated from three patients with BD compared to those obtained from three healthy controls
[21]. Additionally, in vitro lithium pretreatment significantly altered the calcium signaling and electrophysiological properties in BD neurons but not in controls
[21]. Notably, calcium signaling has a central role in controlling inappropriate neuronal responses and tonic excitability. A proteomic study in iPSC-derived dorsal anterior forebrain cortical neurons suggested that the molecular lithium-response pathway in patients with BD may function via collapsin response mediator protein-2, which acts to modify neuronal dendrite and dendritic spinal formation
[22].
The expression of multiple BD-linked genes that are involved in neuronal development, differentiation, and neuroplasticity are regulated by microRNAs (miRNAs), which are small, non-coding RNAs. In particular, miR-34a is predicted to target genes implicated as risk factors for BD and is reduced by lithium and valproic acid, two mood stabilizers that are widely prescribed to prevent depressive and manic recurrences
[23].
An increased miR-34a expression has been detected in postmortem brain tissue, directly induced neuronal cells, and iPSC-derived neuronal cells from patients with BD compared to healthy controls
[24]. The overexpression of miR-34a in vitro was reported to decrease CACNB3 and ANK3 gene expression, previously identified as BD risk genes, leading to impaired neuronal differentiation, synaptic protein expression, and neuronal morphology
[24].
In addition, CXCR4 (CXC chemokine receptor-4) expressing NPCs were analyzed from two BD-affected brothers and their two unaffected parents in a family-based paradigm that has the advantage of controlling for the genetic background
[25]. The study revealed that patients with BD, compared with their unaffected parents, displayed multiple phenotypic differences at the neurogenesis level and gene expression critical for neuroplasticity, including Wnt pathway components and ion channel subunits.
Kim et al. studied iPSC-derived neurons obtained from patients with BD type I and their unaffected siblings from an Old Order Amish pedigree with a high BD incidence
[26]. The observed significant disease-associated differences in gene expression suggested that RNA biosynthesis and metabolism, protein trafficking, and receptor signaling pathway alterations may play a role in BD pathophysiology.
Accumulating evidence suggests a relationship between imbalanced inflammatory response and BD. Astrocytes participate in the brain inflammatory cascade after being activated by pro-inflammatory cytokines. Vadodaria
[27] demonstrated that BD astrocytes are transcriptionally different from controls and induced a reduction in neuronal activity when co-cultured with neurons, even without stimulation. Inflammatory stimulation produced an increased secretion of IL-6 from astrocytes.
Studies on iPSC-derived neuronal cells’ electrophysiological activity may help predict the lithium response and develop novel drugs for BD treatment. Differential hyperexcitability responses to in vitro lithium treatment in iPSC-derived hippocampal dentate gyrus-like neuronal cells from six patients with manic BD type I and four unaffected individuals were detected. Notably, the hyperexcitability phenotype in BD was selectively reversed by lithium treatment only in neurons that are derived from patients who responded to lithium treatment, suggesting that this model of iPSCs might help develop new drugs
[28]. Notably, the electrophysiological data obtained from iPSC-derived neuronal cells treated with lithium was used to train an algorithm that can predict the lithium responsiveness of a new patient with a success rate of over 92%
[28].
Osere
[29] explored the molecular effects of the three most used mood stabilizers (lithium, valproic acid, and lamotrigine) in iPSC-derived NPCs from healthy controls, Li responders, and Li non-treated BD patients. The study suggested that three mood stabilizers with different mechanisms of action affect a specific set of genes, possibly featuring high relevance for the drug mood stabilizing properties.
The first comprehensive study that compared BO generated from patients with BD type I and healthy individuals demonstrated transcriptomic differences with gene downregulation involved in cell adhesion, neurodevelopment, and synaptic biology, along with the gene upregulation involved in immune signaling in patients with BD type I
[30].
5. Autism Spectrum Disorder
Autism Spectrum Disorder (ASD) is a broad range of lifelong neurodevelopmental conditions characterized by impaired social abilities and cognitive functions. ASD can be of unknown polygenic etiology (idiopathic) or a specific syndromic disorder, such as Fragile X, caused by a single gene mutation
[31].
Since iPSCs recapitulate aspects of the neuronal development process while preserving the patient’s genetic background, they are frequently used to model idiopathic ASD, for clarifying pathogenetic mechanisms.
The increased proliferation and differentiation abnormalities of iPSC-derived NPCs from patients with idiopathic ASD with macrencephaly compared to non-ASD controls with normal brain size were suggested to result in early brain overgrowth
[32]. The treatment with the insulin-like growth factor 1(IGF-1), a compound currently in clinical trials, partially rescued the observed neuronal network abnormalities
[32]. Schafer et al. followed iPSCs from patients with idiopathic ASD with early brain overgrowth during their neuronal development to examine when and how the earliest ASD-specific abnormalities arise
[33]. Changes associated with ASD, involving temporal dysregulation of specific gene networks and morphological growth acceleration, tracked back even before the neuronal stage in neural stem cells (NSCs). Bypassing NSC-like stages by direct ASD iPSC conversion into neurons prevents the manifestation of the observed neuronal ASD-associated phenotypes phenotypes
[33].
Using gene expression analyses on patients with ASD iPSC-derived BO, Mariani et al. showed the overexpression of the gene FOXG1, which generates an overproduction of GABAergic neurons, and, in turn, an increased brain volume and imbalanced excitation and inhibition systems in the developing cortex
[34]. Unfortunately, studies on idiopathic ASD NPCs show conflicting results relating to the proportion of GABAergic inhibitory precursors compared to glutamatergic precursors
[35]. Recently, in idiopathic ASD without macrocephaly, impaired neural differentiation was demonstrated in the absence of altered proliferation, in contrast to previous studies
[36].
Fragile X syndrome (FXS) is the most prevalent single-gene form of ASDs and is characterized by cognitive impairment, defective communication, hyperactivity, and impulsivity. In particular, FXS is caused by transcriptional FMR1 gene silencing on the X-chromosome during embryonic development with the consequent loss of Fragile X Mental Retardation Protein (FMRP) expression. This FMRP is a selective RNA-binding protein that regulates the translation of many genes at the synaptic sites. The lack of FMRP leads to aberrant differentiation in human iPSC-derived neural progenitor cells
[37].
Conventional bidimensional and 3D FXS models based on isogenic FMR1 knock-out mutant iPSCs display altered cortical neuron gene expression and impaired differentiation compared with the wild-type human iPSCs. Cortical BO models show an increased number of glial cells, such as astrocytes, and bigger organoid size, which suggests that FMRP is required to correctly support neuronal and glial cell proliferation and the correct excitation/inhibition ratio in human brain development
[38].
Nonsense-mediated RNA decay (NMD) is a cellular surveillance pathway that safeguards the quality and stability of mRNA transcripts by targeting them for degradation if altered. Deficiency in FMRP results in hyperactivated NMD in FXS fibroblast-derived iPSCs, with a negative consequence on iPSC maturation to neurons
[39].
6. Conclusions
The iPSC technology allows capturing the SZ genotype, including genetic risk factors, along with their effects on cellular and molecular endophenotypes during early neurodevelopment. This approach supports the neurodevelopmental hypothesis of SZ, stating that the disruption of early brain development increases the risk of later manifesting psychotic symptoms. Patient-derived iPSCs have proven to be a powerful tool in identifying SZ early neurodevelopmental defects, having revealed alterations in neuronal differentiation
[7][10][11], migration capacity
[2][40], neurite number and length
[1][41], synaptic biology
[9][11][42], connectivity
[1], and neuronal activity
[4][5][42].
Additionally, the studies showed perturbed mitochondrial respiration function and morphology associated with signs of increased oxidative damage in SCZ iPSC-NPCs and neurons relative to controls
[9][43]. These results agree with neuroimaging, postmortem brains, and patient-derived cells, which have implicated mitochondrial dysfunction in SZ pathogenesis. It is generally accepted that prenatal exposure of the developing brain to various environmental challenges increases susceptibility to later SZ, by interacting with a genetic predisposition. As the mechanisms underlying this process remain obscure, iPSC studies evaluated the role of possible factors such as the heat shock factor 1
[11] and the tumor necrosis factor
[11]. Further research should be devoted to finding biomarkers indicative of the early SZ stages to allow the delivery of treatment acting on the damaged neurons before neural impairment cannot be reversed.
In the development of MDD, environmental factors play a central role, even inducing epigenetic modifications, which are lost during the reprogramming process. This loss may hinder identifying the multiple pathways involved in MDD pathogenesis and may impact the study of antidepressant resistance, one of the main focuses of the iPSC approach to MDD. Environmental factors, particularly childhood maltreatment, are known to be associated with poor treatment response in MDD. Despite these limitations, iPSC studies provide important information on the factors involved in SSRI resistance
[44] and in response to bupropion and ketamine
[18][19][20]. It is essential to consider that iPSC experimentations have used developmentally-immature neurons; thus, it is unclear whether the observed iPSC-derived phenotypes persist into adulthood in humans.
Turning to BD, this is a complex illness with heterogeneous clinical presentation. To better understand its biological underpinnings, the iPSC approach focused on clinical phenotypes, such as lithium-responsive patient samples, or on populations with similar genetic profiles, such as BD individuals within a family or larger pedigree. As imaging and postmortem studies suggest a neurodevelopmental etiology of BD, with neuroanatomical abnormalities often present at the first episode, some iPSCs investigations addressed this. In this context, BD-derived iPSC studies support the presence of early pathologic changes, suggested by the differences between the BP and control iPSCs, in genes involved with proliferation, regulation, and differentiation processes
[21][24]. Moreover, BD iPSCs had a lower propensity for generating central neural progenitors, which, in turn, had lower proliferation rates
[25]. Additionally, several data points toward aberrant calcium signaling
[21][25] and electrical properties
[28] in BD neurons, some of which were reduced by lithium, although only in the neurons of lithium-responsive patients
[28][30]. As lithium is the first-line treatment of BD, reliably predicting lithium response using molecular markers would allow earlier initiation of effective therapy, an approach with demonstrated positive effects on the outcome.