Human adherent cortical organoids in a multiwell format

In the growing diversity of human iPSC-derived models of brain development, we present here a novel method that exhibits 3D cortical layer formation in a highly reproducible topography of minimal dimensions. The resulting adherent cortical organoids develop by self-organization after seeding frontal cortex patterned iPSC-derived neural progenitor cells in 384-well plates during eight weeks of differentiation. The organoids have stereotypical dimensions of 3 × 3 × 0.2 mm, contain multiple neuronal subtypes, astrocytes and oligodendrocyte lineage cells, and are amenable to extended culture for at least 10 months. Longitudinal imaging revealed morphologically mature dendritic spines, axonal myelination, and robust neuronal activity. Moreover, adherent cortical organoids compare favorably to existing brain organoid models on the basis of robust reproducibility in obtaining topographically-standardized singular radial cortical structures and circumvent the internal necrosis that is common in free-floating cortical organoids. The adherent human cortical organoid platform holds considerable potential for high-throughput drug discovery applications, neurotoxicological screening, and mechanistic pathophysiological studies of brain disorders.

Existing 3D models suffer from considerable variability due to the complex and heterogeneous nature of the free-floating structures (Cederquist et al., 2019;Lancaster et al., 2013Lancaster et al., , 2017;;Paşca et al., 2015;Renner et al., 2017;Velasco et al., 2019;Yoon et al., 2019).A further challenge, in particular with free-floating organoids, is the necrotic core that emerges when tissue volumes exceed the limits of oxygen and nutrient diffusion beyond a radius of ~300-400μm (Lancaster et al., 2017;Qian et al., 2016b).Although recent progress has been made with slicing organoids prior to the emergence of necrosis followed by organotypic air-liquid interface culture (Giandomenico et al., 2019;Qian et al., 2020), even sliced organoids have to be repeatedly re-cut to prevent necrosis (Qian et al., 2020), which is both laborious and risks introducing another potential source of variability.The generation of vascularized organoids would be the ultimate solution to reduce the necrotic core and the inherent stress that is observed in cortical organoids (Bhaduri et al., 2020;Fan et al., 2022).It will be beneficial to have increased diffusion through microfluidics or vascularization (current efforts summarized in Matsui et al., 2021) and further study the mutually beneficial interaction of neural cells and vascular cells (Crouch et al., 2022;Mansour et al., 2018;Wang et al., 2023).
Here we propose a simplified approach to generating long-term hiPSC-derived adherent cortical organoids with reproducible dimensions and the potential for high-throughput screening in a 384-well format.The resulting adherent cortical organoids can be maintained in long-term culture and contain neurons with dendritic spines and robust activity, as well as several classes of glial cells including oligodendrocyte precursor cells, myelinating oligodendrocytes, and morphologically distinct sub-types of astrocytes.

Results
Self-organized topography of iPSC-derived adherent cortical organoids Adherent cortical organoids reproducibly self-organized into layered radial structures in 384-wells plates within 8 weeks of seeding with hiPSC-derived forebrain-patterned neural progenitor cells (NPCs).Three different hiPSC source cell lines were used in this study, including commercially available NPCs, and NPCs generated using a modified version of our previously described protocol (Gunhanlar et al., 2018) (Figure 1A).NPCs were capable of neural rosette formation and expressed SOX2, Nestin, and the frontal cortical NPC-marker FOXG1 (Figure 1B, C, D, Figure S1A).The initial 4 weeks after seeding the NPCs in Neural Differentiation Medium (ND) were characterized by proliferative expansion of NPCs and the emergence of early neural differentiation markers (Figure 1E, Figure S2).Between 4-and 8weeks post-seeding, neurons and glial cells emerged with a consistent spatial organization (Figure 1E, F, Figure S1B, Figure S2), in which the central region was densely packed with cell bodies while the periphery contained circumferentially and radially organized processes originating from cells in the centre.Typically, cortical organoids defined as a single radial structure per well were observed in ~80% of the wells seeded with NPCs after 60 days of differentiation.The structural integrity of single structure organoids remained quite stable and slowly diminished over time to about 50% of intact single structure organoids after 1 year in culture (Figure S1D).Organoid structure formation was highly dependent on the proliferation rate of NPCs, that can differ substantially between differentiation batches and hiPSC clones.Plotting the interaction between proliferation and the amount of NPCs required to be seeded for the successful generation of adherent cortical organoids, showed a significant correlation (r 2 =0.67) that can be used as a guideline for testing a range of NPC densities (Figure S1C).For each NPC line an optimal seeding density was estimated based on the proliferation rate of that NPC line.Multiple densities were seeded around the estimated optimal density and after 6 weeks it was visually determined which NPC density enabled adherent cortical organoid generation.Typically, too sparse seeding density generated neural networks lacking structure, while excessive densities resulted in overgrowth, leading to reduced structural organisation and compromised long-term survival.

Cell type distribution and layer formation
The spatial organisation that evolved over the first 8 weeks after seeding was paralleled by a shift in cell type distribution.Tau+/MAP2-axons exhibited long extensions in a circular pattern, while MAP2+ dendrites exhibited orthogonally-oriented radial outgrowth (Figure 2A, Figure S3).
After 6-8 weeks following seeding, a self-organised rudimentary segregation of deep-and upper layer neurons emerged as shown by a clear macroscopic separation of deep-and upper layer neurons, although some neurons were spatially intermixed and some neurons were double-positive for CTIP2 and CUX1.Segregation was also observed between CUX1 and CUX2 positive cells as CUX2 is typically expressed over a wider range of upper cortical layers than CUX1 and also marks intermediate progenitors (Molyneaux et al., 2007) (Figure 2C, Figure S4).Analogous to the broad distribution of cortical cell subclasses, the majority of the neurons were glutamatergic, while GAD67+ interneurons were also present (Figure 2D), constituting ~10% of the NeuN-positive neuronal population consistently for all three source hiPSC lines (Figure S5).

Adherent cortical organoids contain multiple glial cell types
Within 8 weeks of seeding, a population of GFAP+/S100β+ astrocytes emerged.Many astrocytes had their soma located in the central region with process outgrowth radially (Figure 2E, Figure S6), while other astrocytes exhibited subtype-specific morphologies including fibrous astrocytes (Figure 2F), protoplasmic-like astrocytes (Figure 2G) and interlaminar astrocytes (Figure 2H).GFAP/PAX6 double-positive radial glia were present at the outskirts of the densely populated centre of the organoid with processes growing radially outwards (Figure 2I).Similar to free-floating organoids, adherent cortical organoids survived for longer periods compared to monolayer neural cultures grown on larger surfaces.The longevity allows for the development of cell types not usually seen in a monolayer culture that can typically be cultured up to a maximum of 2 to 3 months.By 6 weeks after seeding NPCs we observed the emergence of oligodendrocyte precursor cells (OPC), as shown by the expression of NG2 (Figure 3A).typological and temporal hierarchies in the developing human cortex (Bhaduri et al., 2021;Nowakowski et al., 2017Nowakowski et al., , 2016;;Uzquiano et al., 2022).
These functional adherent cortical organoids in a multi-well format should be amenable to high-throughput screening applications, mechanistic pathophysiological studies of neurodevelopmental and neuropsychiatric disorders, and pharmacological and phenotypic screening of disease phenotypes during early cortical development.Moreover, toxicological studies for novel therapeutic compounds also show an increasing need for testing specific effects in human neuronal models, whereas studies using rodent models have often shown poor predictive power for drug safety and efficacy in human central nervous system disorders (van Esbroeck et al., 2017).
Adherent cortical organoids also have some limitations, as the current iteration of adherent cortical organoids exhibits limited cortical layering and regional specification, that seem to be more advanced in floating whole brain organoids.However, these drawbacks are offset by significantly enhanced ease and higher-throughput possibilities of downstream analysis applications.
Taken together, we present a novel platform for cellular-level human brain modeling using adherent cortical organoids that exhibit high reproducibility and robust neuronal activity.
The ability to reliably generate human cortical organoids in multi-well plates combined with neural network functionality offers a unique potential for brain disease modeling and therapeutic screening applications.

Immunocytochemistry
For live-dead staining, living cultures were incubated with LIVE/DEAD™ Viability/Cytotoxicity Kit according to manufacturer's instructions (Thermo Fisher Scientific).For immunocytochemistry adherent cortical organoids were fixed for 20-30 minutes using 4% formaldehyde in phosphate-buffered saline (PBS), washed with PBS and blocked for 1 hour by pre-incubation in staining buffer containing 0.05UM Tris, 0.9% NaCl, 0.25% gelatin and 0.5% Triton-X-100 (pH 7.4).Primary antibodies were incubated for 48-72h at 4U°C in staining buffer, washed with PBS and incubated with the secondary antibodies in staining buffer for

Cyquant Proliferation Assay
CyQUANT™ Direct Cell Proliferation Assay, C35011 (Thermo Fisher Scientific) was used according to manufacturer specifications.
NG2+ cells remained present until at least 4 months (Figure3B/C).Staining for Myelin Basic Protein (MBP) revealed the emergence of MBP+ oligodendrocytes around 4 months after NPC seeding when the organoids were continuously grown in the presence of T3 (2ng/ml) (Figure3D).At 5 months, the oligodendrocytes showed increasingly mature morphologies (Figure3E-I) and exhibited MBP co-localization along NF200+ axons (Figure 3F/I).Adherent cortical organoids show synaptic connectivity and functional activity Neurons within adherent cortical organoids exhibited clear evidence of synaptogenesis (Figure 4A-D).Sparse labelling of excitatory neurons with AAV9.CamKII.eGFPrevealed the presence of Synapsin-positive (Syn+) mushroom-shaped dendritic spines (Figure 4E).To assess the functional activity of the cortical organoids, we used the genetically-encoded calcium indicator GCaMP6s under the control of the human Synapsin promoter (Figure 4F), allowing cell-type specific quantification of neuronal activity.Calcium imaging revealed robust synchronous network-level bursting (NB) (1.4 ± 0.07 NB/min) in which the vast majority of recorded neurons participated.In addition, substantial desynchronized activity was also observed (3.9 ± 0.5 events/min) during time periods outside of network-level bursting (Figure 4G-J, Figure 4 -Video 1).Discussion The study of early human brain development and related diseases has long been hampered by the inherent complexity of the human brain and the inaccessibility of living brain tissue at cellular resolution.Technological advances in induced pluripotent stem cell technology have now facilitated the opportunity to obtain living human neurons derived from specific individuals.We describe here a platform to model early human frontal cortical development with high reproducibility and simplified organization.While 3D floating organoids, or sliced organoids, nicely recapitulate layered cortex formation, they are subject to variation in the relative contribution of cortical tissue within the organoid, forming multiple cortical patches etal., 2019;Quadrato et al., 2017).Moreover, 3D floating organoids suffer from necrosis in the core of the organoid due to lack of oxygen and nutrient diffusion.Recently, other protocols have been published starting from rosette formation with a etal., 2023;Tidball et al., 2023).Our platform predefines a rosetteforming iPSC-derived cortical NPC population that self-organizes into adherent singular radial structures in a standard 384-well format.Other examples have highlighted the benefits of a multi-well format or systematic individual structure formation(Knight et al., 2018;Medda et al., 2016).Our platform now integrates these features to yield individual adherent layered cortical structures with robust functional synaptic connectivity of astrocytes.The small reproducible format of the organoids in a 384-well format has the distinct advantage of being able to image entire organoids without slicing or clearing and to perform spatiotemporal functional analysis by fluorescence-based calcium imaging.We confirmed the self-organizing potential and reproducibility of adherent cortical organoids across multiple hiPSC lines and using different sources of NPCs, controlling the seeding density for the proliferation rate of the specific NPC batch.Seeding NPCs with frontal cortical identity in the defined geometry of a 384-well plate enabled the development of long-term functional neural networks in a complex radial structure resembling early human cortical development.Future studies aimed at single cell gene expression analysis and advanced image-based analysis solutions in this platform will be interesting in the context of the increasing knowledge on single cell topographical, al., 2011)  were collected on a FACSAria III Cell Sorter (BD Bioscience) and expanded in NPC medium consisting of: DMEM/F12, 1% N2 supplement, 2% B27-RA supplement (Thermo Fisher Scientific), 1 μg/ml laminin (L2020, Sigma-Aldrich), 20 ng/ml basic fibroblast growth factor (Merck-Millipore, Darmstadt, Germany) and 1% penicillin/streptomycin (Thermo Fisher Scientific).NPCs were differentiated to adherent cortical organoids between passage 3 and 7 after sorting.Neural differentiation384-well plates (M1937-32EA.Life Technologies) were coated with 50 µg/ml laminin in dH 2 0 (Sigma, L2020) for 30 minutes at 37U°C.The NPCs in NPC medium were dissociated with Accutase (Stem Cell Technologies), live cells were counted with Trypan Blue (Stem Cell Technologies) in a Burker counting chamber.The NPCs were seeded in the wells of a 384well plate at defined densities for each cell line.Specifically, the optimal seeding density was determined by visual inspection of the organoids between 28 to 42 days after seeding a range of cell densities in the 384-well plate wells.NPCs were seeded and differentiated in Neural Differentiation Medium: Neurobasal medium (Thermo Fisher Scientific), 1% N2 supplement (Thermo Fisher Scientific), 2% B27-RA supplement (Thermo Fisher Scientific), Bio), 1UμM dibutyryl cyclic adenosine monophosphate (Sigma-Aldrich), 200UμM ascorbic acid (Sigma-Aldrich), 2Uμg/ml laminin (Sigma-Aldrich) and 1% 2h at room temperature.The cultures were embedded in Mowiol 4-88 (Sigma-Aldrich), after which confocal imaging was performed with a Zeiss LSM700 and Zeiss LSM800 confocal microscope using ZEN software (Zeiss, Oberkochen, Germany).The following primary antibodies were used: SOX2 (Merck-Millipore AB5603, 1:200); Nestin (Merck-Millipore MAB5326, 1:200); MAP2 (Synaptic Systems 188004, 1:100); NeuN (Merck ABN78, 1:200); GFAP (Merck-Millipore AB5804, 1:300); FOXG1 (Abcam AB18259, 1:200); CUX1 (Abcam AB54583, 1:200); CTIP2 (Abcam AB18465, 1:100); Synapsin 1/2 (Synaptic Systems 106003, 1:200); and PSD95 (Thermo Fisher Scientific MA1-046, 1:100); Tau (Cell Signaling Technology 4019, 1:200); S100ß (Sigma-Aldrich S2532, 1:200); Pax6 (Santa Cruz sc-81649, 1:100); NG2 (Gift from W. Stallcup Lab, 1:100); NF200 (Sigma-Aldrich 083M4833 1:200); MBP (Abcam AB7349, 1:100); GFP (Abcam ab13970, 1:100); GAD67 (Merck-Millipore MAB5406, 1:100) DAPI (Thermo Fisher Scientific D1306

Figure 1
Figure Titles and Legends Figure 1 Adherent cortical organoid model.A Schematic representation of the differentiation protocol.B/C/D Representative NPCs from three different iPSC lines with markers SOX2, Nestin and FOXG1 (scale bar B, 50 µm; C/D, 20 µm).E Representative time

Figure 2
Figure 2 Adherent cortical organoids show an organized network of neuronal and astrocyte subtypes.A MAP2+ somas and dendrites alongside Tau+/MAP2-axons show