Present and future modeling of human psychiatric connectopathies with brain organoids

Abstract

Brain organoids derived from human pluripotent stem cells are emerging as a powerful tool to model cellular aspects of neuropsychiatric disorders, including alterations in cell proliferation, differentiation, migration, and lineage trajectory. To date, most contributions in the field have focused on modeling cellular impairment of the cerebral cortex, with few studies probing dysfunction in local network connectivity. However, it is increasingly more apparent that these psychiatric disorders are connectopathies involving multiple brain structures and the connections between them. Therefore, the lack of reproducible anatomical features in these 3D cultures represents a major bottleneck for effectively modeling brain connectivity at the micro (cellular) and at the macroscale levels between brain regions. In this perspective, we review the use of current organoid protocols to model neuropsychiatric disorders with a specific emphasis on the potential and limitations of the current strategies to model impairments in functional connectivity. Finally, we discuss the importance of adopting interdisciplinary strategies to establish next generation, multiregional organoids that can model, with higher fidelity, the dysfunction in the development and functionality of long-range connections within the brain of patients affected by psychiatric disorders.

Introduction

The advanced cognitive processing power of the human brain results from a complex and protracted codevelopment of distinct but interconnected brain regions. The composition of connecting circuitry may range from a cluster of region-specific neurons that receive, modify, and relay information to nearby clusters (microcircuits), to long range intra-regional neuronal connections (macrocircuits). The groundwork for this circuitry is laid at early stages of human brain development and is refined throughout gestation and adolescence. Accumulating evidence indicates that neuropsychiatric disorders that share overlapping genetics including autism spectrum disorder, schizophrenia, bipolar disorder, and epilepsy may converge on dysregulation of neural connectivity (1) (as reviewed in (2). But how, when and in which cells do these dysfunctions originate? Genetic variants associated with these disorders predominantly converge on developmental networks spanning multiple stages, cell types, and brain regions (3). This would indicate that the pathophysiology of psychiatric disorders could initially arise during early human brain development, persists and is compounded throughout our lifespan. A blooming hypothesis among the neuropsychiatric field suggests that mutations associated with neuropsychiatric disorders largely contribute to micro-circuit/macro-circuit dysfunction and network activity imbalances during early neural development by altering neurogenesis rates, cell fate specification, spatiotemporal neuronal positioning, and abnormalities in spontaneous and synchronized activity (4-9) (Fig. 1). Currently, a large proportion of neuropsychiatric research has focused on the neocortex, as this structure has undergone significant evolutionary divergence (10) and is often associated with the convergence of several neuropsychiatric disorder mutations and phenotypes. However, it is increasingly clear that psychiatric disorders have the ability to alter crosstalk within and between multiple brain structures. As numerous studies have suggested, to comprehensively understand the pathogenesis of neuropsychiatric disorders, both local and distal circuit dysfunctions should holistically be considered as impairments can arise in the connectivity within and between these regions (11,12).

Figure 1.

Top Panels.

Brain organoids have become a useful model system to interrogate the genetic landscape and cellular architecture of the developing human brain. However, they remain a reductionist model system lacking the codevelopment of distinct but functionally interconnected regions, precise cytoarchitectural organization, accurate targeting of micro- and macro-circuit connections and experiential activity dependent maturation. Recent attempts to integrate distinct brain regions and improve macro-circuit formation have included multiregional organoid models with or without the assistance of a microfluidic device. After transplantation of organoids into the mouse brain the organoids are able to functionally integrate to receive thalamocortical and corticocortical inputs as well as produce sensory responses in human neurons (137) (VZ = Ventricular Zone, ISVZ = Inner Subventricular Zone, OSVZ = Outer Subventricular Zone, IZ = Intermediate Zone, SP = Subplate)

Bottom Panel.

The use of human patients to interrogate human brain development is largely limited to neuroimaging studies which provide snapshots of anatomy and activity. Although more in depth cellular and molecular characterization can be performed on post-mortem human brain tissue, its availability is limited, and the quality can vary depending on post-mortem interval (PMI). Alternatively, organoids allow for longitudinal monitoring of development and disease, are readily available, and yield consistently high-quality tissue as they have a negligible PMI.

Psychiatric disorders are predominantly polygenic, meaning multiple genetic risk variants contribute towards risk, adding further difficulties in modeling and determining pathophysiology of these disorders (3). In addition, as these disorders largely affect cognitive capacities that are uniquely human and with limited access and interrogation tools of human fetal brain development, induced pluripotent stem-cell (iPSC) derived human brain organoids are an attractive model to elucidate developmental stages, processes and cell types affected during disease pathogenesis. In this review we will describe the current array of patterned 3D organoids and how they have been used to model neuropsychiatric disorders with a genetic etiology at the cellular and micro-circuit level. We will then describe recent clinical neuroimaging studies describing aberrant functional and morphological connectivity between brain regions, with particular emphasis on axes that may be modeled with currently available patterned brain organoids. Finally, we will discuss current and future efforts to model long-range connectivity using brain organoids.

Modeling human brain organogenesis with organoids

Benchmarking brain organoids against human brain development

Human derived iPSC 3D cultures largely take inspiration from in vivo patterning, which involve precise spatiotemporal signaling to generate regional identity. Two main approaches have been taken to generate cortical brain organoids, unpatterned and patterned protocols. These methods will be described in brief here. (Several reviews have been written which provide greater details) (13-15). In both approaches, dissociated pluripotent stem cells are allowed to self-assemble under suspension culture conditions into embryoid bodies (EBs). Unpatterned protocols utilize minimal extrinsic instructions allowing for intrinsic mechanisms to drive cells towards a neuroectoderm fate, ultimately giving rise to a range of brain regions. Patterned protocols take inspiration from known in vivo neurodevelopmental patterning strategies to direct EBs towards cell fates reminiscent of specific brain regions. Most common strategies to generate cortical brain organoids utilize either single or dual SMAD-inhibition with or without WNT inhibition to guide EBs toward neocortical development (16,17).

Cortical organoids are widely used because of their relevance to human neurodevelopmental disorders, as defects in higher order cognitive abilities are thought to predominately stem from dysregulation within this brain region. Cortical organoids have served as reliable models of early brain development regarding cellular and molecular features as analyzed via scRNA-seq and scATAC-seq (4,18-20) (reviewed in (21), and are able to achieve synchronized local network activity (22,23). Several new studies using multiomics, spatial transcriptomics, and lineage tracing approaches have demonstrated the high fidelity of cortical organoids with the developing human fetal brain (24-26). Despite patterning and culturing differences, cortical organoids across protocols can develop neural-tube-like neuroepithelial structures that follow a similar but accelerated temporal developmental trajectory as in vivo. These structures give rise to a diverse range of cortical cell types as seen in the endogenous human forebrain including, ventricular and outer radial glia, intermediate progenitors, deep- and upper-layer cortical neurons, astroglia, and oligodendroglia precursor cells. Although the cellular composition of the human fetal cortex is well recapitulated in cortical organoids, due to their reductionist nature, protocols have grappled with reliably generating stereotypical cortical layers as well as batch to batch reproducibility issues in some protocols (27,28). Proper anatomy of the cerebral cortex is required for the precise assembly of GABAergic interneurons and glutamatergic neurons into reproducible microcircuits as well as for the correct establishment of long-range circuit formation during fetal development (29).

Current protocols achieve patterning in vitro via bath application of morphogens, which does not wholly recapitulate the tightly regulated spatiotemporal gradients that are key for proper neurodevelopment. Therefore, directed differentiation approaches generate a single early brain region and do not recapitulate the simultaneous development of multiple brain regions or the long-range connectivity between them, which are necessary for refinement and specification of cells within a developing human brain (Fig. 2). While some cortical arealization spontaneously occurs in cortical organoids, this process remains stochastic. Indeed, while early intrinsic genetic programs (protomap) drive the establishment of an initial molecular map, locally secreted factors and activity-dependent mechanisms, which are not reproducible and currently uncharacterized in organoids, refine the formation of discrete functional areas within the human fetal cortex (30). Thus, although cortical organoids are the most well characterized and have a great degree of development homology with the developing fetal brain, there are still several limitations that will need to be addressed to accurately model neuropsychiatric disorders.

Figure 2.

A growing body of literature suggests that dysregulation of the neurogenesis program may result in impaired neuronal wiring. The establishment of proper neuronal circuits requires precise spatiotemporal control of neuronal positioning, neurogenesis, and afferent/efferent synaptic connectivity. Dysregulation of any one or combination of these processes can result in impaired micro- and macro-circuit formation in the form of hyper- or hypo-connectivity as well as general miswiring. Among other factors, major contributors to the proper development of neural circuitry are the timing with which neural cells progress through distinct phases of proliferation, differentiation and migration to their final positioning within the cortex. Similarly, informational factors such as spontaneous and synchronous neuronal activity influence cell-cell communication and ultimately, wiring. (WM = White matter, I= Inhibitory, E = Excitatory).

Beyond Modeling of the Neocortex

In the previous section we have reviewed advances in the generation and characterization of cortical organoids, however, neuropsychiatric disorders do not solely affect cortical areas as clinical studies have described phenotypes across multiple brain structures (see below section on Connectivity Alterations in Psychiatric Disorders). Thus, the importance of generating organoids that overcome the default telencephalic programming is paramount towards accurately modeling mental illness. As the 3D brain organoid field has matured, protocols have been developed which provide additional patterning instructions to generate other brain regions. Several of these protocols have been developed which recapitulate the developing MGE, LGE, Retina, Thalamus, Hypothalamus, Hippocampus, Pituitary, Midbrain, Striatum, Hindbrain, Spinal Cord, and Choroid Plexus (31-42). As our knowledge of how local signaling drives fetal human brain regionalization expands, organoid protocols are becoming increasingly more precise and able to produce multiple subregions of the brain. For example, organoids of the arcuate nucleus of the hypothalamus have been generated (43). Additionally, cortical organoids with a higher proportion of prefrontal cortex identity have been reported through additional guidance with retinoic acid treatment (20). These advances in patterning and precisional generation of specific brain regions and subregions are indicators of a new frontier for organoid-based disorder modeling wherein brain regions that are highly implicated in psychiatric disorders can be reliably generated and used to elucidate how cellular, molecular, and micro-circuit dysfunction are altered. Currently, cortical organoids remain the most well characterized and, as we have described, several benchmarking studies have been performed addressing both reproducibility and fidelity to in vivo tissue. Similar assessments should be performed on regionally patterned brain organoids to ensure confidence in future modeling of early regional brain development both normally and in psychiatric disorder contexts. While long term culture of cortical and retinal organoids have been established, this effort has not been demonstrated for other region-specific organoids. Extended culture is essential to allow for maturation of functional features such as coordinated network activity (22,44).

Modeling Connectopathies with Brain Organoids

Modeling human brain connectivity at the micro and macroscale levels with brain organoids

Human pluripotent stem cell derived brain organoids have been utilized to successfully model early human neurodevelopmental processes, including progenitor dynamics, migration, emergence of cellular diversity, and characterizing spontaneous and network level activity (as reviewed by (13,14,45,46). For example, mutations in ASD- associated risk genes, including SUV420H1, ARID1B, CHD8, SYNGAP1, and FMR1, as well as idiopathic ASD have been found to confer asynchronous development of neurons and altered neuronal excitability within cortical organoids (4,5,47-52). Several of these studies have also reported morphological and functional, excitatory and inhibitory neuronal imbalances. Similarly, cortical organoid models of Rett syndrome (RTT) display irregular neuronal progenitor proliferation and differentiation dynamics, as well as abnormal neuronal activity (53-55). Asynchronous excitatory and inhibitory neuronal development due to altered proliferative dynamics as well as an increase in neuronal excitability in cortical organoids were also observed in schizophrenia (SCZ) models (56-59). Although several brain region specific organoids have been generated, cortical organoids have been predominantly used to model neurodevelopmental and psychiatric disorders. Therefore, questions about the effects of these mutations on developmental trajectories of other brain regions as well as their effect on long-range functional connectivity remain largely unanswered and constitute a priority for the field.

The most prominent approach to generate multiregional organoids has been through fusion of individually patterned organoids via co-culturing, termed assembloids (36,40-42,60-62). This approach has been utilized to study human relevant connections and, in some cases, to improve functional maturation in developing neurons (42). With the use of fused organoids, several studies have begun to investigate what defects may arise during development that contribute towards morphological and functional changes between multiple brain regions. Within cortical and subpallial organoids derived from patients with Timothy syndrome (TS), deficiencies in neuronal migration, saltation length and frequency, and aberrant cortical network activity was found (36,63). Also using fused cortical and subpallial organoids, hyperexcitability, abnormal neural oscillations, and impaired interneuron migration was found in organoids generated from patients with RTT (23,55). Similarly, in a pioneering macro-circuit study using Phelan-McDermid syndrome cortico-striatal fused organoids, functional alterations were found to be primarily in striatal neurons (42). These founding studies utilizing multiregional organoids have laid the groundwork for future research and modeling of macro-circuit dysconnectivity associated with psychiatric connectopathies. However, future efforts will be needed to address the lack of reproducible stereotypic macroscale anatomic structure and connectivity between organoids.

Connectivity Alterations in Psychiatric Disorders

An important advantage of using assembloids is the ability to combine individual organoids generated from both healthy and patient-derived iPSCs (mix and match approach)(46). This is a unique feature of this in vitro model system as neuroimaging studies have indicated that neuropsychiatric disorders affect multiple brain regions, thus the establishment of a system that allows for the interrogation of primary regional etiopathogenesis is highly needed. Here we will briefly review clinical neuroimaging findings which highlight the complex neural circuits spanning multiple brain regions implicated in psychiatric disorders. We will focus on primarily genetically driven psychiatric disorders with particular emphasis on circuits that may be modeled with the current array of patterned human brain organoids.

Cortico-cortical alterations remain the most significant driving force of effect size in clinical neuroimaging cohort studies. Here, we define cortico-cortical connections of the cerebral cortex as those in which long-range axons start in one cortical area and end in another distal cortical region. The end point of these axons is largely dependent upon the spatial orientation of their corresponding cell body within the cortical layers. Glutamatergic excitatory projection neurons (pyramidal neurons) account for a majority of the neuronal population within these neocortical layers. Projection neurons that originate in the upper cortical layers (L2-4) extend axonal projections exclusively to contralateral and intrahemispheric cortical regions, facilitating communication within and between the cerebral hemispheres. Among cortico-cortical circuits three major networks are associated with psychiatric disorders, the frontoparietal network (FPN), the Salience Network (SN) and the Default-mode Network (DMN) (64). Separately and in combination with one another, these networks have been highly implicated in neuropsychiatric and neurodevelopmental disorders (65-67). As previously described, current approaches have begun to generate subregion specific organoids such as prefrontal-dominant cortical organoids (20). With further improvements in cortical arealization strategies the contributions of individual cortical areas could be dissected. Rudimental attempts to establish an organoid model to study contralateral projections and connections have been made, however these models remain under characterized (see section below on next generation modeling of human connectopathies with brain organoids).

Pathological hallmarks of psychiatric disorders include complex behavioral deficits spanning sensory, motor, cognitive, and emotional domains involving multiple brain territories and connectivity loops. Accordingly, several clinical imaging studies have suggested morphological and functional alterations in various networks. For example, the visual system, consisting of the retina, the dorsal lateral geniculate nucleus of the thalamus, and primary visual cortex (V1), is known to be affected in a variety of neuropsychiatric disorders including SCZ, BD, and ASD. Impairments along this axis have primarily been focused on sensory processing and typically been attributed to morphological and/or functional abnormalities within the visual cortex. However, several studies have described clinical phenotypes in multiple structures along the primary visual axis including regions within the retina and thalamus. Using Optical Coherence Tomography (OCT) several studies have shown retinal morphological changes in patients with SCZ, BD, and ASD (68,69). Additionally, functional studies using electroretinography (ERG) have described attenuated a-wave and b-wave amplitudes as well as reductions in photopic negative response amplitude (70-72). Several of these morphological and functional anomalies within the retina have been observed in genetically high-risk children (73,74) suggesting the potential use of the retina as a biomarker for early detection of psychiatric disorders (75-80). Similarly, differences in thalamocortical connectivity have been described along the visual axis in psychiatric syndromes. Thalamic-occipital dysconnectivity has been shown via resting state functional connectivity MRI (rs-fcMRI) in patients with SCZ and BD (81). Interestingly, 6-week-old infants at high familial risk for ASD were found to have significant thalamo-occipital hyperconnectivity as well as aberrant thalamo-occipital white matter tracts compared to low-risk controls (82). Altogether, these findings suggest that with further characterization, retino-thalamic-cortical organoids can be helpful tools for interrogating the contribution of each structure within the visual system towards mental disorder pathogenesis.

Similarly, cortico-basal ganglia loops have also shown significant abnormalities in psychiatric syndromes. Three distinct basal ganglia neural circuits can be identified based on the input/output associated with one of the basal ganglia nuclei, including the striatum (caudate, putamen, and nucleus accumbens), substantia nigra, globus pallidus, ventral pallium, subthalamic nucleus, and ventral tegmental area. The three basal ganglia networks include the dorsolateral prefrontal circuit, which is involved in associative and cognitive functions, and orbitofrontal and anterior cingulate circuits, which are associated with limbic control and motivated behavior (83,84). Interestingly, the volumes of striatum and Globus Pallidus were found to be increased in several studies of SCZ patients (85-90). Additional works have pointed at a decrease in volume of these structures (91-95) suggesting that the patients’ treatment history may play a role in the observed structural alterations. A reduced cortico-striatal functional connectivity has been found in SCZ patients (96-100). One study did not observe significant baseline differences in functional connectivity of the striatum between healthy individuals and patients with psychosis (101), however patient selection criteria may explain this inconsistent finding. In addition, the main ASD symptoms of social, communication deficits and restricted and repetitive patterns of behavior can be associated with a dysfunction in the mesocorticolimbic system (102-107). Overall, these studies point to the importance of dissecting the contribution of cortico-basal ganglia circuits to disease pathogenesis. To date, cortico-striatal assembloids have been used to model Phelan-McDermid syndrome offering a proof of concept that brain organoids can provide insight into macro-circuit dysregulation (42).

Lastly, we will highlight connectivity alterations in cerebellar-thalamic-cortical circuits, as the cerebellum remains a clinically understudied brain region that has been heavily implicated in multiple psychiatric disorders with both motor and non-motor deficits (108-110). The cerebellum receives multiple visual, auditory, and somatosensory inputs from the spinal cord, inferior olive, and pontine nuclei; while the major output of the cerebellum is via the thalamus towards the sensory areas of the cerebral cortex, where the cerebral cortex can provide feedback to the cerebellum via the pontine nuclei (111). The multiple connections between the cerebellum and sensory regions of the cortex is known as the sensorimotor domain. Additionally, the cerebellum forms distinct connections with cortical regions associated with higher-order executive functions such as the prefrontal and parietal cortex (112). Using fMRI, a decrease in functional connectivity between the cortex and cerebellum was found in patients with ASD (113). Additionally, hyperconnectivity in the sensorimotor domain has been shown in children with ASD that displayed both cognitive impairments and repetitive motor behavior (114). Interestingly, these individuals also displayed reduced connectivity in the supramodal domain, suggesting that alterations in both tracts may have complementary roles in ASD pathogenesis. Similarly, in patients with schizophrenia, the medial dorsal nucleus of the thalamus showed reduced connectivity with the cerebellum, which was not seen in patients with BD (115). Altogether these findings highlight the importance of modeling and investigating this axis to improve our understanding of these psychiatric disorders. However, while thalamo-cortical assembloids (40) have been established, the lack of a protocol for long-term culture of cerebellar organoids with functional activity currently prevents the modeling of this circuit. In addition to date the main approach to obtaining functionally mature purkinje neurons, the major output of the cerebellum, is through the dissociation and plating of human cerebellar organoids onto mouse granule cells (32). Therefore the generation of organoids containing functional purkinje cells remains a high priority in the field.

Next generation modeling of human connectopathies with brain organoids

In previous sections we have discussed current methods to model micro- and macroscale connectivity with brain organoids. However, there are key features that need to be incorporated in current approaches including co-development of distinct central nervous system (CNS) structures with distinct anatomical organizations, establishment of reproducible and functional long-range connectivity between defined CNS structures, and activity-dependent maturation of each cell population within organoids (Fig. 2). To achieve co-development, approaches to generate multiregional brain organoids simultaneously from one initial aggregate have been developed by leveraging self and external patterning guidance cues. For example, human trunk neuromuscular organoids (116) have been developed starting with a multipotent progenitor population that is subsequently allowed to self-organize into two spatially distinct yet interconnected regions. While this is a promising approach relying on self-organization, the fidelity of long-range connectivity within these organoids remains unknown. Similarly, an inducible sonic hedgehog (SHH) expressing iPSC line was used to create a pseudo-SHH organizing center at one pole of a developing cortical organoid to simultaneously generate a subpallium domain (117). Currently, this approach has not been developed for additional brain regions, wherein additional characterization on the functional connectivity between brain regions will be needed. Several groups have reported the generation of neural tube-like organoids using either microfluidics-based or ECM scaffolding approaches, which are able to recapitulate both rostro-caudal and dorsal-ventral patterning of early human brain development (118-121). However, these neural tube organoids are not able to be kept in long term cultures, preventing inter-regional connectivity between neighboring or distal brain regions.

The integration of tissue engineering approaches has led to the establishment of microfluidic platforms that have been used not only to pattern iPSCs into multiregional and neural-tube-like organoids, but also to facilitate discrete long-range functional connectivity between distinct organoids. Accordingly, several groups have developed long-range axonal tract models to connect cortical organoids(122-124) (Fig. 2). These approaches utilized either matrigel or collagen hydrogels to induce neurite extensions through microchannels connecting adjacent culturing wells under static conditions. Interestingly, robust reciprocal and/or uni-directional axonal fascicles were formed, although reproducibility and specificity of synaptic connections were not characterized.

Using one such long-range axonal tract model, a preprint by Osaki and Ikeuchi (2021)(125), has shown that fused cortical organoids have both less frequent burst-like activity and significantly higher lag of signal propagation when compared to long-range, axonally connected organoids. These findings suggest that the architecture of synaptic connections largely determines the level of neuronal activity within organoids and that axonally connected organoids form more complex neural circuits than fused organoids, although the mechanism and rationale for this improved spontaneous and evoked activity is not fully characterized. A drawback of these current microfluidic approaches stems from the static culture conditions utilized to allow axonal extension to occur. 3D brain organoid cultures require dynamic culture conditions that improve oxygenation, promote nutrient diffusion, and prevent hypoxia within the central region of organoids.

Due to this, it is important for future studies to focus on improving stereotypical anatomy within individual organoids and the connections between organoids while maintaining optimal culturing conditions necessary for proper neuronal development. Improvements in cortical lamina layering and extension of long-range axonal fascicles have been reported through sliced organoids culturing strategies (126,127). Similarly, the engineering of neuronally-relevant biomimetic cellular microenvironments to support maturation and anatomical structure has potential to improve modeling of later stages of functional connectivity development (13,128-130). Beyond improvements in stereotypic anatomy, approaches to instruct axonal pathfinding towards correct postsynaptic targets will need to be developed and incorporated into organoids. Possible strategies include the use of microcontact printed guidance cues and hydrogels, as well as optogenetic stimulation (131-133).

Finally, as more physiologically relevant culture systems are developed to model human connectopathies, improved readouts for unbiased, high throughput characterization and mapping of functional connectivity in next generation organoid models will need to be integrated. Only a handful of studies have evaluated the fidelity of synaptic connectivity between organoids to in vivo circuits using viral-based retrograde trans-synaptic tracers (41,42). Interestingly, initial characterization of synaptic connectivity has revealed that cortico-spinal and cortico-striatal fused organoids display a remarkable self-assembly capacity to form physiologically relevant functional connections wherein corticofugal projection neurons were found to largely interact with neurons over glial cells of the spinal or striatal organoids. Although an abundance of callosal projection neurons were present in these cortical organoids, only a small fraction of callosal projections were connecting with the adjacent organoid (41,42). Additional approaches to trace connectivity in both retrograde and anterograde directions are being developed for live-imaging and single-cell sequencing readouts which can be easily translated to brain organoids to further characterize the connectivity between organoids at a neuronal subtype specific level (134-136).

Conclusions

Accumulating evidence indicates that neuropsychiatric disorders converge on and emerge from impaired brain wiring during development. Therefore, the refinement of 3D human brain organoid models, which allow for the access and interrogation of early developmental processes, will be essential to faithfully model micro- and macro-circuit level dysfunctions between distinct brain regions. With the use of animal models, it is yet unclear which brain regions and cell types constitute the primary etiology of functional and morphological changes in human neuropsychiatric disorders. Next-generation brain organoids that allow for reproducible and physiologically relevant modeling of long-range connectivity offer the unique opportunity to “mix and match” distinct organoids with multiple genetic backgrounds to determine selective vulnerability of circuits in distinct brain regions. Finally, the establishment of functional and reproducible macro-circuit connectivity between distinct brain region-specific organoids does not guarantee improved maturation of circuits as these organoids will still lack experiential activity-dependent maturation (Fig.2), an essential in vivo process for refinement of circuit development and function that has recently been shown in a transplanted human cortical organoid to mouse model (137). Therefore, the budding effort in the field to improve functional maturation of human PSC derived neurons (138,139) are attractive and important methods that may be integrated into brain organoids to allow accurate modeling of connectivity alterations in psychiatric disorders.