Human brain tissue model for studying microglia phenotypes

Abstract

The present invention relates to a method of generating an in vitro modular 3D brain tissue model (3D BTM) comprising: a. mixing 50,000 to 1,000,000 freshly splitted, single neurons (NE) and astrocytes (AS) or their precursor cells in a ratio of NE:AS = 1:3 to 10:1, followed by spinning down to introduce self-aggregation in low attachment conditions, without addition of exogenous matrices; b. treating cells obtained in step a. with a NOTCH signaling inhibitor for driving terminal neuron differentiation for at least 2 days; c. treating the cells obtained in step b. with at least one mitotic inhibitor to remove residual dividing cells for at least 3 days, thereby yielding a post-mitotic aggregated 3D culture of neurons and astrocytes; d. adding differentiated microglia (MG) to the culture obtained in step c. in a ratio of NE:MG = 1:1 to 20:1, to initiate migration of MG into the culture for 3-10 days; e. culturing the culture obtained in step d. in neuron media with addition of TGFβ1 and cytokines selected from the group comprising CSF1-R agonists to support microglia maintenance and proliferation for at least 2 weeks, thereby inducing formation of a modular in vitro 3D BTM culture.

Description

[0001] PCT-Patent Application based on 24 201 798.6 Klinikum der Ludwig Maximilians Universitat Munchen; Klinikum der Ludwig-Maximilians-Universitat Munchen Anstalt des offentlichen Rechts

[0002] Vertreten durch den Arztlichen Direktor und den Kaufmannischen Direktor Vossius Ref.: AJ1544 PCT

[0003] Human brain tissue model for studying microglia phenotypes

[0004] The present invention relates to a method of generating an in vitro modular 3D brain tissue model (3D BTM) comprising: a. mixing 50,000 to 1 ,000,000 freshly splitted, single neurons (NE) and astrocytes (AS) or their precursor cells in a ratio of NE:AS = 1 :3 to 10:1 , followed by spinning down to introduce self-aggregation in low attachment conditions, without addition of exogenous matrices; b. treating cells obtained in step a. with a NOTCH signaling inhibitor for driving terminal neuron differentiation for at least 2 days; c. treating the cells obtained in step b. with at least one mitotic inhibitor to remove residual dividing cells for at least 3 days, thereby yielding a post-mitotic aggregated 3D culture of neurons and astrocytes; d. adding differentiated microglia (MG) to the culture obtained in step c. in a ratio of NE:MG = 1 :1 to 20:1 , to initiate migration of MG into the culture for 3-10 days; e. culturing the culture obtained in step d. in neuron media with addition of TGFpl and cytokines selected from the group comprising CSF1 -R agonists to support microglia maintenance and proliferation for at least 2 weeks, thereby inducing formation of a modular in vitro 3D BTM culture.

[0005] In this specification, a number of documents are cited. The disclosure of these documents, while not considered relevant for the patentability of this invention, is herewith incorporated by reference in its entirety. More specifically, all reference documents are incorporated by reference to the same extent as if each individual document was specifically and individually indicated to be incorporated by reference.

[0006] Investigating the physiological and disease-related interplay of human neurons and glia requires experimentally accessible and well-characterized model systems. Recent progress in the field of stem cell research now allows generating multiple brain cell types from induced pluripotent stem cells (iPSC) [1] and studying their interactions in co-cultures [2]. Furthermore, these human in vitro systems enable the investigation of pathomechanisms and pathogenesis in disease-relevant cell types [3], [4]. While two- dimensional (2D) models containing neurons and glia co-cultured in a dish allow relatively simple model generation [5], brain cells in 2D culture often lack physiological morphology, maturation, homeostatic behaviors, and intercellular crosstalk found in complex tissues. 3-dimensional (3D) in vitro models offer the possibility to study cell-cell and cell-matrix interactions in a more physiological environment compared to 2D cultures. In addition, they enable investigation of extracellular pathologies, such as the accumulation and aggregation of amyloid-p (Ap) peptides in Alzheimer’s disease (AD).

[0007] Age-related brain diseases belong to the most common causes of disability and death and among those disorders, AD is the most prevalent cause of dementia affecting around 57 million patients worldwide with 10 million new cases every year. Typical pathological hallmarks of AD in the brain are Ap-containing extracellular plaques, intracellular tangles made of Tau protein, brain inflammation indicated by gliosis, and degeneration of cortical neurons. While the neuropathology is well characterized, it is poorly understood how these late-stage hallmarks arise, and which cellular and genetic factors are involved in their formation. As patient brain cells are not directly accessible for experimental research, mechanistic studies rely on cell and animal models, which contributed a lot to our understanding of disease mechanisms.

[0008] However, these models also have drawbacks that limit research on disease pathways and drug development. Mouse models do not recapitulate all relevant disease aspects, such as the combined formation of plaques, tangles, and neurodegeneration in AD [6]. This is likely due to species-specific differences, as it has been shown that mice differ in central molecular and cellular disease pathways e.g., Tau isoform expression that is relevant for Tauopathy formation, diversity, function, and complexity of cell types e.g., astrocytes [7], expression signatures and activation states of microglia [8], [9] and neuronal vulnerability to AD pathology. Full-blown AD phenotypes can only be achieved in mice by massive non-physiological overexpression of mutant human APP and Tau genes, leading to models prone to artefacts, which restricts investigation of underlying mechanisms. These drawbacks may contribute to the fact that most drugs developed with current models failed in clinical trials. Based on these results, the inventors hypothesized that human-specific disease mechanisms contribute to the formation of complete AD pathology. Therefore, modelling of full pathology may require human brain cells in a physiological, brain-like environment, and the investigation of mechanisms causing and therapeutic strategies preventing AD would be greatly promoted by a human disease model.

[0009] Human models derived from iPSCs have great potential to test this hypothesis, as they allow studying affected cell types and the function and malfunction of disease-relevant proteins in their natural cellular environment. iPSC-based disease modelling has made major advances, e.g., by developing efficient differentiation protocols for major brain cell types, gene editing of disease-associated mutations with CRISPR / Cas (e.g., the inventors work

[0010] ,

[0011] ), 3-dimensional cell and organoid culture, and drug development approaches. Nevertheless, current iPSC-based disease models show low reproducibility and cell type diversity (e.g., lack of microglia and oligodendrocytes), lack physiological cell-cell and cellmatrix interactions (e.g., 2D cultures and Matrigel-embedded cells) and / or present an immature, proliferating environment as opposed to the largely postmitotic adult brain (e.g., brain organoids).

[0010] Current human iPSC-based 3D models include hydrogel-based systems, spheroid-like cultures, and brain organoids (BO). Hydrogel-based systems are generated by embedding iPSCs or neural precursor cells (NPCs) into synthetic or cell-derived matrices, such as Matrigel. However, the brain extracellular matrix (ECM) has a unique composition and properties

[0012] that cannot yet be replicated in vitro. As the brain ECM influences cell functionality, maturation, and likely also disease pathogenesis, reproduction of a physiological environment may be required to establish brain-like cell states, functions, and disease responses, thus limiting the applicability of hydrogel-based systems. Spheroid cultures are commonly derived by aggregation of NPCs

[0013] or NPCs and astrocytes

[0014] , followed by further maturation of the cells in 3D. While showing good reproducibility when generated in microwells, these models lack incorporation of homeostatic microglia and have been kept in culture only for a few weeks. Thereby, recapitulation of tissue characteristics is limited, including formation and maturation of ECM, and establishment of mature morphologies, functions, and interactions of the different cell types. BO develop in 3D directly from pluripotent stem cells and can thus uniquely model aspects of human brain development

[0015] , including important insights into human-specific mechanisms, e.g., regarding precursor cell biology and cell-type specifications

[0016] . However, BO cultured in vitro show limited suitability to study microglia, the brain-resident immune cells, which play major roles in both brain maintenance and neurodegenerative diseases

[0017] . Microglia arise from the extra-embryonic mesoderm and are therefore not observed in neuroectodermal organoids derived with guided protocols, while unguided protocols can yield few microglia if mesodermal cells are present

[0018] . Exogenously added iPSC-derived microglia rapidly decline in numbers and usually disappear within few weeks

[0019] ,

[0020] . Furthermore, these microglia do not show homeostatic signatures, including expression of P2RY12, or morphologies in vitro. These physiological features can currently only be achieved by transplantation of BO with microglia into mouse brain

[0019] , confirming the fundamental ability of iPSC-derived microglia to mature in environments provided by iPSC-derived models, but also illustrating their dependency on the correct environmental cues for adopting a homeostatic cell state.

[0011] Crucial extrinsic factors required for induction of a mature, homeostatic microglia state are thus not found in currently available human 2D and 3D in vitro models. These factors may include a post-mitotic, mature brain-parenchyma-like environment, low levels of cell death and associated stress, and reproducible and continuous presence of mature astrocytes. The inapt environment provided by current models, such as BO cultured in vitro, specifically the proliferative state with neural rosettes, the associated large-scale growth and subsequent necrotic core formation

[0021] ,

[0022] , or lack of mature astrocytes in trackable time frames

[0023] ,

[0024] , may thus prevent formation of a homeostatic microglia phenotype. As microglia are known to be highly sensitive to changes in their environment, this likely prevents adoption of a resting state with ramified morphologies and expression of typical markers such as P2RY12. Furthermore, the artificial activation observed in these non-homeostatic MG in vitro may impair studies on microglial responses to brain disease, such as AD.

[0012] Since current models are impaired by low reproducibility, maturity and cell-type diversity, there is still a need to provide a new method and means that allow studying physiological and disease phenotypes of human brain cells, especially microglia, in a mature, i.e., physiological, environment. This need is addressed by the present invention.

[0013] Accordingly, in a first aspect the present invention relates to a method of generating an in vitro modular 3D brain tissue model (3D BTM) comprising: a. mixing 50,000 to 1 ,000,000 such as 100,000 to 500,000 such as about 250,000 freshly splitted, single neurons (NE) and astrocytes (AS) or their precursor cells in a ratio of NE:AS = 1 :3 to 10:1 , followed by spinning down to introduce self-aggregation in low attachment conditions, without addition of exogenous matrices; b. treating cells obtained in step a. with a NOTCH signaling inhibitor for driving terminal neuron differentiation for at least 2 days; c. treating the cells obtained in step b. with at least one mitotic inhibitor to remove residual dividing cells for at least 3 days, thereby yielding a post-mitotic aggregated 3D culture of neurons and astrocytes; d. adding differentiated microglia (MG) to the culture obtained in step c. in a ratio of NE:MG = 1 :1 to 20:1 , to initiate migration of MG into the culture for 3-10 days; e. culturing the culture obtained in step d. in neuron media with addition of TGFpl and cytokines selected from the group comprising CSF1-R agonists to support microglia maintenance and proliferation for at least 2 weeks, thereby inducing formation of a modular in vitro 3D BTM culture.

[0014] The “brain tissue” in accordance with the present invention is to be understood to comprise tissue related to any part of the brain, for example cortex, mid-brain, and hippocampus, depending on the regional identity of the neurons and glia that have been differentiated from iPSCs.

[0015] A “3D brain tissue model” (3-dimensional brain tissue model, 3D BTM, interchangeably used herein as 3BTM) as used herein refers to a fully iPSC based, self-aggregated assembly of post-mitotic neurons, astrocytes and microglia in a network which mimics the physiological state of the brain tissue. The model is characterized by morphological, functional, and proteomic maturation of neurons, astrocytes and microglia highly similar to in vivo brain tissue without formation of a necrotic core. The “brain tissue” in accordance with the present invention is to be understood to comprise tissue that can be related to all parts of the brain, such as cortical, mid-brain, and hippocampus. In principle, the invention allows generating tissues for all brain regions. It is understood that the 3D brain tissue model thus consists of a 3D structure of actual cells.

[0016] The term “freshly split”, as used herein, refers to a condition where cells have undergone a detachment treatment to receive single cells. The attachments of the cells between each other and to the cell culture flask and / or plate are temporally nullified to achieve a singularization of the cells. In that regard, the cells can also be considered as “pure” or “purified”. The conditions for splitting of the cells are described in detail in the examples herein. It is to be understood that the numbers of 50,000 to 1 ,000,000 applies to the combination of NE and AS.

[0017] The term “astrocytes”, abbreviated with “AS” as used herein, refers to a neuroglial cell type of ectodermal origin and its progenitor cells, which is characterized by a round nucleus and a "star shaped" body and many long processes that can end on synapses or as vascular foot plates on the small blood vessels of the CNS. Astrocytes further provide homeostatic support for neurons and contribute to brain ECM generation.

[0018] The term “neurons”, abbreviated with “NE” as used herein generally refers to a cell type in the nervous system whose cells communicate with each other by the generation and spread of depolarization and in particular means cortical neurons characterized by their original location in the prefrontal cortex of the human brain as seen by expression of markers such as FoxG1 , Satb2 and Ctip2. The term “single” in connection with NE and AS is intended to mean that NE and AS are detached from any surface and any other cell.

[0019] As used herein, the term “neural precursor cell”, refers to a progenitor cell that has self-renewal and diverse differentiation potential, the ability to maintain proliferative capacity and can differentiate into neurons, oligodendrocytes and astrocytes under defined conditions, as known in the literature (Williams, L. A., Davis-Dusenbery, B. N. & Eggan, K. C. Snapshot: Directed Differentiation of Pluripotent Stem Cells. Cell 149, 1174-1174. e1 (2012). As used herein, the neural precursor cells are cultured in conditions for differentiation into neurons or astrocytes. The term “neural precursor cell” generally comprises late and early neural precursor cells. It is preferred to use precursor cells on day 20-90 of differentiation, more preferably on day 38-45 of differentiation for neurons and day 63-70 for astrocytes.

[0020] The term "microglia", abbreviated with “MG” as used herein, refers to a class of glial cells that are the resident immune cells in the brain. As such they are involved in the regulation of immune functions within the central nervous system, including surveillance of the tissue environment for danger and stress signals, pruning of excessive synapses during brain development, and reaction to tissue injuries, e.g., by starting an inflammatory response and removal of cell debris to restore tissue homeostasis.

[0021] As used herein, the term “spinning down”, refers to a step to apply g-force on the cells (such as by centrifugation) to press them together to introduce self-aggregation of the cells into an aggregate of cells according to conditions known in the art, preferably at a speed of about 200 g for about 2-4 min. Gravity alone will lead to formation of loose tissue chunks, which is not desired.

[0022] The term “low attachment conditions” as used herein comprises the use of low-attachment plates and wells and using anti-adherence rinsing solutions as described in the Examples herein below. Preferably, ultra-low attachment conditions using ultra-low attachment plates are applied. Low attachment conditions support self-aggregation of the cells into a tissue, rather than attachment to the plate.

[0023] The term “exogenous matrices”, as used herein, refers to synthetic or cell-derived matrices comprising at least one matrix protein in combination with at least one growth factor. A matrix protein is for example collagen or laminin. Cell-derived matrices are for example basement membrane extracellular matrices derived from a tumor extract. The extract is typically “Matrigel”, “Cultrex” or “EHS matrix”, which is often used in the art to embed cells and promote long-term stability. Since exogenous matrices, as used herein, are understood as matrices comprising a matrix protein in combination with at least one growth factor, application of those may result in unpredictable culture formation. Furthermore, exogenous matrices differ from the extracellular matrix (ECM) of in vivo brain tissue that the latter is mainly composed of hyaluronan, chondroitin-sulphate and heparane-sulphate proteoglycans, and linker proteins, enabling the physiological cell maturation and function in the brain.

[0024] A “NOTCH signaling inhibitor” as used herein, refers to biomolecules that inhibit the function of proteins of the notch signaling pathway especially proteins of the NOTCH family, i.e., notch receptors NOTCH1 , NOTCH2, NOTCH3, and NOTCH4. Notch signaling is involved in neurogenesis-promoting proliferation. Inhibition of notch signaling results in promotion of neuronal differentiation. Thus, a notch signaling inhibitor as used herein drives neural precursor cells into terminal neuron differentiation. Non-limiting examples of NOTCH signaling inhibitors within the context of the present invention comprise DAPT, MK- 0752 (CAS-No: 471905-41-6) and LY-411575 (CAS-No: 209984-57-6).

[0025] The function of NOTCH signaling inhibitors can be verified by treating cells with the NOTCH signaling inhibitor and subsequently extracting proteins and conducting a western blot for the Notch intracellular domain (NICD), which is normally released after the cleavage and whose levels will thus be reduced. Compounds which are not NOTCH signaling inhibitors will correspondingly not show reduced levels of NICD upon western blot.

[0026] As used herein, the term “mitotic inhibitor” means a reagent leading to the death of cells that undergo cell division and comprise nucleoside analogues, which inhibit nucleotide synthesis or are incorporated into the DNA during DNA replication inhibiting DNA synthesis, or inhibitors of mitosis and / or cell division comprising microtubule and tubulin inhibitors. Inhibition of microtubules and / or tubulin generally impacts the mitotic spindle assembly thereby inhibits chromosome pulling which is required for cell division. Application of at least one mitotic inhibitor thus induces formation of a post-mitotic culture. Non-limiting examples of mitotic inhibitors include pyrimidine analogs, such as 5’-Fluorouracil, cytosine arabinoside, FUdR (5-Fluoro-2'-deoxyuridine), or Aphidicolin. However, it is to be understood that any compound that leads to the death of dividing cells may serve as a mitotic inhibitor in the context of the application. The skilled person is aware of methods to determine whether a compound leads to the death of dividing cells, e.g., a simple experiment is to take dividing cells such as neural precursor cells (NPCs) and check if they die when treated with the compound, which can be checked by staining or FACS for example using dyes, such as FDA / PI, TUNEL or other cell death markers. One can also determine cell number at endpoint. Furthermore, absence of dividing cells after treatment of 3BTMs can be shown by reduction in KI67 staining, as is demonstrated in Fig. 1 D.

[0027] The term “dividing cell” is herein understood to relate to any cell that retains the capacity to divide, i.e., undergo mitosis. Dividing cells comprise stem cells and progenitor cells. A non-limiting example of dividing cells within the context of the nervous system are NPCs. In contrast, non-dividing cells do not retain the capacity to undergo mitosis and are typically terminally differentiated cells that no longer proliferate. A non-limiting example of non-dividing cells in the context of the nervous system are neurons.

[0028] The term “culturing 3D BTMs” refers to the maintenance of the generated 3D BTM cultures in vitro in media used in the art for culturing neuronal cells. If not described otherwise all cells were grown in a culture medium, such as 3D culture maintenance medium or Neurobasal with B27 supplement (without „PLUS“). For maintaining microglial function and survival as well as to support their maintenance, the media is additionally supplemented with at least one cytokine, such as displayed in the following Table.

[0029] Accordingly, the term “cytokine”, as used herein, preferentially refers to TGFpl to support microglial maintenance and proliferation. The term also comprises CSF1-R agonists, such as IL-34 (Interleukin- 34) and M-CSF (Macrophage colony-stimulating factor), which support microglial survival and proliferation. It is to be understood that the cells are primate cells. According to a preferred embodiment, the cells are human cells.

[0030] In a preferred embodiment of the first aspect of the present invention, the 3D culture maintenance medium comprises Neurobasal PLUS media, supplemented with B27 Plus supplement (B27+), Penicillin / Streptomycin (Pen / Strep), GlutaMax, and cytokines interleukin-34 (IL34), transforming growth factor beta 1 (TGFB1), macrophage colony-stimulating factor (M-CSF also known as CSF-1). The detailed composition of the 3D culture maintenance medium used in the steps generating and culturing a 3D brain tissue culture in the appended examples is summarized as follows:

[0031] In a preferred embodiment the 3D brain tissue model is a 3D cortical tissue model, wherein cortical neurons are employed in step a of the method in accordance with the first aspect.

[0032] Independent of the specific type of neurons, astrocytes or their precursor cells, the neurons and astrocytes or their precursor cells may also be split into single cells and frozen to be later used for mixing after thawing to assemble the culture in accordance with step a. of the first aspect of the present invention. Thus, in a preferred embodiment, 250.000 freshly split single cells of neurons and astrocytes or frozen and thawed single cells of neurons and astrocytes are used in step a. of the above-described first aspect for assembly of the culture.

[0033] In accordance with step a. of the method of the first aspect of the present invention, the preferred ratio for mixing the neurons and astrocytes is NE:AS = 3:1 . In another preferred embodiment, the addition of microglia in step d. the preferred ratio of neurons to microglia is NE:MG = 4:1. The microglia migrate into the culture for 3-10 days, preferably about 7 days thereby generating a 3D BTM culture. Culturing of the 3D BTM culture for at least 2 weeks in step e. in accordance with the method of the first aspect of the present invention allows maturation to a 3D BTM culture which is similar to physiological brain tissue characterized by ramified structure of microglia.

[0034] The inventors generated a modular model, allowing selective combination of different cell and genotypes for neurons, astrocytes, and microglia. This is not possible in prior art model systems in which these cell types, especially neurons and astrocytes, develop from the same stem or precursor cells, such as brain organoids, as cell types then inherently share the same genetic background. In contrast, by differentiating and combining all cell types separately, the model generated by the method according to the first aspect of the present invention facilitates the analysis of cell type interactions and cell-type- specific disease contributions and may be adjusted by modularization of the specific cell types.

[0035] The highly reproducible in vitro 3D brain tissue model is characterized by efficient and reproducible incorporation of neurons, astrocytes and microglia, which survive for many months in a postmitotic state without necrotic core formation. In contrast to current models, the present invention combines and aggregates cells that have achieved a stable cell fate but are still plastic enough to allow tissue integration. Therefore, the 3-dimensional structure formed in step e. by the co-cultured, differentiated cells feature several tissue characteristics, including formation and maturation of a brain-like ECM, as well as maturation of cell types on a morphological, functional, and protein expression level, and adoption of a homeostatic MG state with P2RY12 expression.

[0036] The stability of the 3D BTM generated by the method of the present invention in accordance with the first aspect allows long-term culturing, which leads to a more pronounced formation of a ramified microglia morphology and stronger expression of P2RY12. Prolongation of the model may in certain embodiments be prolonged by the addition of fresh MG.

[0037] As is shown in Figure 9 herein below, the inventors could confirm a more mature, homeostatic, and brain-like phenotype of all cell types if cultured in 3BTMs compared to 2D co-cultures, together with a more brain-like ECM and MG functionality, showcasing the improvement of the new model system compared to previously used 2D cultures. Furthermore, in Figure 10 the inventors could show a high degree of reproducibility of 3BTMs generated from three independent iPSC lines across many parameters including culture size, composition, cellular phenotypes, and proteome composition and maturation over time. This became especially apparent in comparison to NOs from two independent iPSC lines which show a much higher degree of variability in proteome composition and maturation. Finally, with regard to Figure 11 , it is to be noted that the modularity of 3BTMs (e.g., compared to brain organoids) enables the examination of the cell-type specific effect of astrocytes and modulation of their phenotype, thereby allowing the investigation of cell-type specific disease contributions.

[0038] Accordingly, in a preferred embodiment of the first aspect of the present invention, the present invention refers in step e to the method, wherein culturing the 3D BTM cultures in neuron media with addition of cytokines supporting microglia maintenance and proliferation is preferably performed for 1-6 months, thereby maintaining formation of a physiological, homeostatic microglial phenotype with ramified morphology and expression of P2RY12.

[0039] The term “P2RY12”, as used herein, refers to a receptor from the family of purinergic receptors P2Y that is sensitive to adenosine and activated by ADP / ATP. P2RY12 is involved in microglial motility and migration and commonly used as homeostatic marker for microglia in the art, as its expression is rapidly lost after extraction of microglia from the brain.

[0040] In a further preferred embodiment of the first aspect of the present invention, the neurons and / or astrocytes, or their precursor cells, of step a. are obtained or obtainable by (i) differentiating iPSCs or (ii) differentiating embryonic stem cells by dual-SMAD inhibition or overexpression of transcription factors selected from the group including Ngn2, Ngn1 , ASCL1 , BRN2, MYT1 L, and NEUROD1 to directly differentiate neurons, or overexpression of transcription factors selected from the group including NFIA, NFIB, Sox9 to directly differentiate astrocytes. Generally, the term “iPSC” means induced pluripotent stem cell, wherein accordingly, the term “hiPSC” refers to human iPSCs. The term “pluripotent stem cell” refers to a cell which is capable of unlimited self-renewal and differentiation into cells of all three germ layers, i.e., endoderm, mesoderm and ectoderm. The ectoderm is the outermost layer of the three germ layers and differentiates to form epithelial and neural tissue (spinal cord, peripheral nerves and brain) and is thus of particular importance for the present invention. The term “induced pluripotent stem cell (iPSC)” as used herein refers to a type of pluripotent stem cell artificially derived from a non-pluripotent cell, typically a somatic cell, such as a skin cell or blood cell. iPSCs can be generated from somatic cells by epigenetic resetting of somatic cells, e.g., by retroviral transduction of somatic cells, such as fibroblasts, hepatocytes, or gastric epithelial cells, with transcription factors, such as Oct3 / 4, Sox2, Klf4, and optionally c-Myc (Takahashi and Yamanaka, 2006).

[0041] The term “embryonic stem cell” also referred to as “ES” is a cell with pluripotency, which is derived from the inner cell mass of the blastocyst, an early-stage embryo. In a preferred embodiment, the embryonic stem cells are obtained in a way that does not lead to the destruction of a human embryo.

[0042] Generally, “dual-SMAD inhibition” refers to a well-established method to derive neural progenitor cells from both human embryonic stem cells and iPSCs. The term “SMAD” is an abbreviation of the genes in Caenorhabditis elegans SMA ("small" worm phenotype) and MAD family ("Mothers Against Decapentaplegic") in Drosophila, wherein SMADs were first discovered. SMAD pathways mediate proliferation and transforming growth factor p (TGF-p) and bone morphogenic protein (BMP) signaling through SMAD members and have distinct effects on development and homeostasis. Dual-SMAD inhibition typically inhibits the undifferentiated state of iPSCs as well as mesodermal development through inhibition of TGFp and Activin / Nodal signaling using SB431542, respectively; and inhibition of non-neural ectoderm and trophectoderm development though inhibition of BMP signaling using Noggin or more commonly LDN-193189. Thus, inhibition of SMAD prevents differentiation of non-neural cells and thus increases the number of neural cells during differentiation from iPSCs. It is a well-established method to derive neural progenitor cells from both human embryonic stem cells and iPSCs. The term “dual-SMAD inhibition”, in accordance with the present invention, means the inhibition of the SMAD signaling pathway via two SMAD inhibitors selected from the non-exhaustive group of: SB431542, SB525334, LDN-193189, Noggin, DMH1 to derive neural progenitor cells from iPSCs and may be supported by WNT inhibition as described in the examples herein below. The protocol steps for differentiation with dual-SMAD inhibition is known to the skilled person and has been described for example by Dannert et al., 2023

[0025] .

[0043] As used herein, the term “transcription factor”, refers to a protein or polypeptide that binds specific DNA sequences associated with a genomic locus or gene of interest to control transcription. Transcription factors may promote (as an activator) or block (as a repressor) the recruitment of RNA polymerase to a gene of interest. Transcription factors may perform their function alone or as a part of a larger protein complex. With regard to information on transcriptional factors, mention is made, for example, of Latchman and DS (1997) Int. J. Biochem. Cell Biol. 29 (12): 1305-12; Lee Tl, Young RA (2000) Annu. Rev. Genet. 34: 77-137and Mitchell PJ, Tjian R (1989) Science 245 (4916): 371 -8, herein incorporated by reference in their entirety.

[0044] Transcription factors used for differentiation of hiPSC may be endogenously or exogenously expressed. For differentiation into neural cell lineages as described in the present disclosure, generally exogenous expression is used in the art, as it is technically much simpler to express from an exogenous source, such as a transgene, than inducing overexpression from the endogenous locus.

[0045] As described in the above preferred embodiment of the first aspect, an alternative to differentiation by dual-SMAD inhibition, is differentiation of iPSCs or embryonic stem cells (ES) by overexpression of transcription factors such as Ngn2, Ngn1 , ASCL1 , BRN2, MYT1 L, and NEUROD1 to directly differentiate neurons, or overexpression of transcription factors selected from the group including NFIA, NFIB and Sox9 to directly differentiate astrocytes.

[0046] In a preferred embodiment at least one, at least two, at least three, at least four, and at least five transcription factors used for neural differentiation by overexpression of transcription factors are selected from the group including Ngn2, Ngn1 , ASCL1 , BRN2, MYT1 L, and NEUROD1 . In another more preferred embodiment of the above-preferred embodiment, neural differentiation is induced by the expression of Ngn2.

[0047] In another preferred embodiment at least one or at least two transcription factors used for astrocyte differentiation by overexpression of transcription factors are selected from the group including NFIA, NFIB and Sox9. In another more preferred embodiment, astrocyte differentiation is induced by the expression of NFIA.

[0048] Generally, the term “NGN” means “neurogenin” and refers to the NGN family comprising NGN1 , NGN2 and NGN3, which are important throughout development of neurons. NGN2 is transiently expressed in multiple NPC types and downregulated during final differentiation stages to promote pro-neural pathways and subtype specification. NGN2 is preferably used for its ability to rapidly induce neuronal differentiation (Hulme AJ, Maksour S, St-Clair Glover M, Miellet S, Dottori M (2022) Making neurons, made easy: the use of neurogenin-2 in neuronal differentiation. Stem Cell Reports 17:14— 34. https: / / doi.Org / 10.1016 / j.stemcr.2021 .1 1 .015).

[0049] The term “Ngn2” (Uniprot ID: Q9H2A3; Gene ID: 63973) encoded by all sequences deposited under the NCBI Reference Sequence NP_076924, such as NP_076924.1) and “Ngn1 ” (Uniprot ID: Q92886; Gene ID: 4762) encoded by all sequences deposited under the NCBI Reference Sequence NP_006152, such as NP_006152.2) also refer to functional allelic variants thereof that are distinct from the hitherto described sequences by at least one base pair, but maintain with increasing preference at least 50%, 80%, 90% or 100% transcription factor activity of the described transcription factor. The forced expression of BRN2, ASCL1 and MYT1 L are known in the art to reprogram mouse fibroblasts into functional neurons and have been tested in human iPSCs.

[0050] The term “ASCL1 ”, as used herein above means the transcription factor “Achaete-Scute Family BHLH Transcription Factor 1 ” and is encoded by all sequences deposited underthe NCBI Reference Sequence NG_008950, such as NG_008950.1 ; Gene ID: 429, and Uniprot ID: P50553. The term “ASCL1 ” also refers to functional allelic variants thereof that are distinct from the hitherto described sequences by at least one base pair, but maintain with increasing preference at least 50%, 80%, 90% or 100% transcription factor activity of the described transcription factor. ASCL1 is sufficient to drive direct neuronal differentiation of both human and mouse PSCs.

[0051] The term “BRN2”, as used herein above refers to a POU domain transcription factor encoded by the POU3F2 gene and is encoded by all sequences deposited under the NCBI Reference sequence NP_005595, such as NP_005595.2; Gene ID: 5454; and Uniport ID: P20265. The term “BRN2” also refers to functional allelic variants thereof that are distinct from the hitherto described sequences by at least one base pair, but maintain with increasing preference at least 50%, 80%, 90% or 100% transcription factor activity of the described transcription factor. BRN2 likely contributes to neuronal maturation by increasing morphological complexity.

[0052] The term “MYT1 L”, as used herein above means the transcription factor “Myelin Transcription Factor 1 Like” and is encoded by all sequences deposited underthe NCBI Reference Sequence: NP_001289981 , such as NP_001289981.1 ; Gene ID: 23040, and Uniprot ID: Q9UL68. The term “MYT1 L” also refers to functional allelic variants thereof that are distinct from the hitherto described sequences by at least one base pair, but maintain with increasing preference at least 50%, 80%, 90% or 100% transcription factor activity of the described transcription factor. MYT1 L likely contributes to neuronal maturation by increasing morphological complexity. Forced expression of this gene in combination with the basic helix- loop-helix transcription factor NeuroDI and the transcription factors POU class 3 homeobox 2 and achaete-scute family basic helix-loop-helix transcription factor 1 can convert fetal and postnatal human fibroblasts into induced neuronal cells, which are able to generate action potentials.

[0053] The term “NEUROD1 ”, as used herein above means the transcription factor “Neurogenic Differentiation 1 ”. NEUROD1 is encoded by all sequences deposited under the NCBI Reference Sequence: NP_002491 , such as NP_002491.3; Gene ID: 4760, and Uniprot ID: Q13562. The term “NEUROD1 ” also refer to functional allelic variants thereof that are distinct from the hitherto described sequences by at least one base pair, but maintain with increasing preference at least 50%, 80%, 90% or 100% transcription factor activity of the described transcription factor. The forced expression of NEUROD1 is known in the art to differentiate hiPSCs into excitatory neurons.

[0054] The term “NFIA”, as used herein above, means the transcription factor “nuclear factor 1 A-type”. NFIA is encoded by all sequences deposited under the NCBI Reference Sequence NP_001128145, NP_005586, NP_001138983 and NP_001138984; Gene ID: 4774 and Uniprot: Q12857. The term “NFIA” also refer to functional allelic variants thereof that are distinct from the hitherto described sequences by at least one base pair, but maintain with increasing preference at least 50%, 80%, 90% or 100% transcription factor activity of the described transcription factor. The expression of NFIA is known in the art to differentiate stem cells into astrocytes.

[0055] The term “NFIB”, as used herein above, means the transcription factor “nuclear factor 1 B-type”. NFIB is encoded by all sequences deposited under the NCBI Reference Sequence NP_001177666, NP_005587, NP_001269716 and NP_001356(387-402); Gene ID: 4781 and Uniprot: 000712. The term “NFIB” also refer to functional allelic variants thereof that are distinct from the hitherto described sequences by at least one base pair, but maintain with increasing preference at least 50%, 80%, 90% or 100% transcription factor activity of the described transcription factor. The expression of NFIB is known in the art to differentiate stem cells into astrocytes.

[0056] The term “Sox9”, as used herein above, means the “SRY-box transcription factor 9” of the Sox family. Sox9 is encoded by all sequences deposited underthe NCBI Reference Sequence NP_000337.1 , Gene ID: 6662 and Uniprot P48436. The term “Sox9” also refer to functional allelic variants thereof that are distinct from the hitherto described sequences by at least one base pair, but maintain with increasing preference at least 50%, 80%, 90% or 100% transcription factor activity of the described transcription factor. The expression of Sox9 is known in the art to differentiate stem cells into astrocytes.

[0057] The differentiation using expression of transcription factors enables fast differentiation of neurons, which can easily be used for large scale drug and genetic screens. Models using transgenic transcription factor-differentiated neurons are also regularly used for iPSC-derived models of dementia. However, the purity and distinct neuronal fate of the resulting cultures is still under debate, making dual-SMAD differentiated neurons more preferable.

[0058] In a further preferred embodiment of the first aspect of the present invention, the neurons and / or astrocytes or their precursor cells (i.e., neural precursor cells) of step a., are obtained or obtainable by differentiating neural precursor cells, and / or wherein the microglia of step d. are obtained or obtainable by differentiating hematopoietic precursor cells (HPCs).

[0059] The term “neural precursor cell” as used herein refers to an undifferentiated cell which is capable of proliferation and has the potential to generate both neurons and astrocytes. Differentiation into neurons is achieved by application of a notch signaling inhibitor at a specific timepoint (see above), while differentiation into astrocytes is achieved by induction of the gliogenic switch using LIF and subsequent differentiation into astrocytes by addition of ciliary neurotrophic factor (CNTF) and extended culture periods.

[0060] The term “hematopoietic precursor cell” (HPC), as used herein, means a multipotent cell that can develop into all types of blood cells in a process called hematopoiesis but also all types of myeloid- lineage and lymphoid-lineage cells. Differentiation of HPCs into microglia is achieved by addition of cytokines IL34, TGF 1 , and M-CSF (https: / / doi.org / 10.15252 / embj.2021109108).

[0061] As described above, the 3D BTM culture generated in accordance with the method of the first aspect of the present invention is able to mimic physiological brain tissue, which also comprises physiology of diseased brain tissue. Thus, the modular 3D BTM model can mimic pathophysiology of neurological diseases when the neurons, astrocytes, their precursor cells and / or the microglia are genome edited to result in a diseased state.

[0062] Accordingly, in another preferred embodiment of the first aspect of the present invention, the neurons, astrocytes, their precursor cells and / or microglia are obtained or obtainable by differentiating iPSCs or stem cells that have undergone genome editing to insert mutations associated with neurological diseases selected from the group including Alzheimer’s Disease (AD), Frontotemporal lobar degeneration (FTLD), and Parkinson’s Disease (PD).

[0063] The skilled person is aware of methods for gene-editing, also referred to as genome editing using a genome-editing system, for example, for removing an intron between two exons, or introducing mutations into a gene of interest. Generally, the term “genome editing” refers to a technique capable of introducing a targeted alteration into the nucleotide sequence of a gene in animal, bacterial and plant cells, including human cells, and results in the knock-out of a specific gene, the knock-in of specific alterations in a gene, or introduction of a mutation into a non-coding DNA sequence that does not produce protein. In addition, genome editing enables deletion, duplication, inversion, replacement or rearrangement of genomic DNA. The “genome-editing system”, as used herein, refers to a system, such as CRISPR / Cas9, CRISPR / Cas12a, CRISPR base editing, CRISPR prime editing, CRISPRi, CRISPRa, Zinc Finger nucleases, or TALENs. CRISPR / Cas9 is a very versatile genome editing tool that allows cleaving genomes specifically and efficiently at most desired loci (Xiong, X., Chen, M., Lim, W.A. et al. (2016). CRISPR / Cas9 for human genome engineering and disease research. Annu. Rev. Genomics Hum. Genet. 17: 131-154. https: / / doi.org / 10.1 146 / annurev-genom-083115-022258).

[0064] Methods of genome editing, as used herein may be based on methods as described in Paquet et al., 2016; Yusa, K. Seamless genome editing in human pluripotent stem cells using custom endonucleasebased gene targeting and the piggyBac transposon. Nat Protoc 8, 2061-2078

[0065] (2013). https: / / doi.org / 10.1038 / nprot.2013.126, and are further described in the Examples herein below.

[0066] As described above, specific cell types with or without a mutation can be produced by techniques known in the art based on the multipotent ability of certain stem cells to differentiate into the desired cell type. In accordance with the present invention, it is also envisaged that neurons, astrocytes and microglia associated with disease conditions are derived from patients.

[0067] Accordingly, in another preferred embodiment of the first aspect of the present invention, cells employed in step a. and / or d. are derived from a patient having a mutation causing a neurological disease. The term "neurological disease", as used herein, refers also to neurodegenerative diseases, which are diseases associated with progressive loss of neuronal structure or function, including death of neurons. Examples of neurodegenerative diseases include, but are not limited to, multiple sclerosis, Alzheimer's disease, Parkinson's disease, Frontotemporal dementia, and amyotrophic lateral sclerosis.

[0068] In another preferred embodiment of the first aspect of the present invention, the low-attachment conditions comprise usage of ultra-low attachment material. Non-limiting examples are ultra-low- attachment, 96-well round-bottom or v-bottom plates or cell culture plates coated with anti-adherence rinsing solution. In a more preferred embodiment, low-attachment conditions may also be applied in step d., when microglia are added.

[0069] In another preferred embodiment of the first aspect of the present invention, the NOTCH inhibitor in step b. is DAPT, and the mitotic inhibitor in step c. is pyrimidine analog 5’-Fluorouracil, cytosine arabinoside, FUdR (5-Fluoro-2'-deoxyuridine), and / or Aphidicolin, preferably 5’-fluorouracil.

[0070] As used herein, the term “DAPT” means “N-[N-(3, 5-Difluorophenacetyl)-l-Alanyl]-s-Phenylglycine-T- butyl ester” (CAS-No: 208255-80-5) and is a chemical compound capable of inhibiting y-secretase directly and thus indirectly Notch - a substrate for y-secretase. DAPT addition drives terminal neuron differentiation when applied for at least 2 days as described above, preferably 7 days, in accordance with the first aspect of the present invention.

[0071] Addition of at least one mitotic inhibitor as defined above, such as pyrimidine analogs, yield a postmitotic cell culture when applied at least for 3 days, preferably 11 days, in accordance with the first aspect of the present invention. 5’-Fluorouracil (5’-FU) thereby kills residual dividing cells, while uridine minimizes the effect on RNA metabolism and thus toxicity in non-dividing cells. In a more preferred embodiment, 5’-FU may be used alone without uridine.

[0072] In another preferred embodiment of the first aspect of the present invention, the microglia of step d. are added on day 10 to 40 of differentiation, preferably on or around day 24 of differentiation.

[0073] As used herein, the “day of differentiation” refers to the number of days that have passed since differentiation has been initiated counted from the iPSC stage. The day the microglia are added to the 3D BTM culture is after aggregation of neurons and astrocytes and after 5’-FU and uridine treatment has ended. Accordingly, the earliest day for adding microglia to the culture of neurons and astrocytes is 5 days after culture generation, preferably the addition of microglia is on day 15 after culture generation. However, it is also possible to add microglia at a later stage to replenish the microglia population that slowly decreases over time. Accordingly, in a more preferred embodiment, microglia are added at least 3 months, at least 4 months, at least 5 months, or at least 6 months after culture generation. In accordance with a further preferred embodiment of the method of the first aspect of the present invention, other cell types apart from neurons, astrocytes and microglia may be added to form a 3D brain tissue model mimicking physiological brain tissue.

[0074] Thus, in a preferred embodiment of the first aspect of the present invention, the present invention refers to a method wherein oligodendrocytes (OLG), endothelial cells or pericytes are added after step c. or d.

[0075] The term “oligodendrocyte” (OLG), as used herein, means a glial cell which is exclusive to the central nervous system and is important for modulation of action potential propagation by myelination of axons as well as trophic and metabolic support of neurons. In a more preferred embodiment, oligodendrocytes are added to the culture in a ratio of neurons: oligodendrocytes (NE:OLG) of 5:1.

[0076] As used herein, the term “endothelial cell” refers to a cell that is found at the interior surface of blood vessels. Generally, endothelial cells are known in the art to line the innermost layer, cavities of the heart, blood vessels (including capillaries), and lymph vessels.

[0077] The term “pericyte”, as used herein, refers to mural cells of microcirculation that wrap around endothelial cells which line the capillaries throughout the body. They are spatially isolated, potentially contractile and have multiple functions and are important in the CNS for blood vessel formation, and maintenance of the blood-brain barrier.

[0078] In an additional preferred embodiment, the method further comprises adding exogenous Ap42 to the 3D brain tissue model.

[0079] A combination of the model in accordance with the first, second and third aspect (see below) of the present invention and exogenous Ap42 may thus be applied to study downstream pathogenic processes, such as neuritic dystrophies and associated tau pathology, as well as drug testing, e.g., for anti-amyloid antibodies. In another preferred embodiment of the second and third aspect, thus, the in vitro 3D brain tissue model of the second and third aspect is applied for drug testing of anti-amyloid antibodies. The in vitro 3D BTM in accordance with the second and third aspect of the present invention is thus also used to evaluate anti-amyloid-antibody-based therapies.

[0080] The inventors further show plaque formation upon addition of exogenous Ap42 (Example 6) and confirm suitability of the model to test anti-amyloid antibodies.

[0081] In a second aspect of the present invention, the present invention relates to a modular in vitro 3D brain tissue model produced or producible by the method of the first aspect of the present invention.

[0082] In a preferred embodiment of the second aspect of the present invention, the in vitro 3D brain tissue model of the second aspect is characterized by at least two features selected from the group of:

[0083] (i) ramified / branched morphology of neurons and astrocytes and increasing numbers of synapses over the first 3 months in culture;

[0084] (ii) ramified morphology of microglia; as defined by microglia extending 3 or more branched processes into the surrounding tissue with each process being longer than the average diameter of the cell body.

[0085] (iii) no necrotic or hypoxic core;

[0086] (iv) postmitotic state of neurons and astrocytes;

[0087] (v) homeostatic microglia expressing P2RY12;

[0088] (vi) expression of brain specific ECM-related proteins selected from the group comprising Neurocan, Brevican, CSPG5, and Tenascins;

[0089] (vii) surveillance of cellular environment by microglia by extending and retracting processes, and reaction of microglia to tissue damage by extending processes towards the injury site within 5-10 minutes while cell body stays in place;

[0090] In a more preferred embodiment of the second aspect, the in vitro 3D brain tissue model is characterized by, with increasing preference, at least three, at least four, at least five, at least six or all features selected from the group of (i) to (vii).

[0091] In a even more preferred embodiment of the second aspect, the in vitro 3D brain tissue model of the second aspect is characterized by, with increasing preference, at least two, at least three, or all features selected from the group of (ii), (iii), (vi) and (vii), wherein feature (vii) is particularly preferred.

[0092] In particular preferred embodiments of the second aspect, the in vitro 3D brain tissue model of the second aspect is characterized by the features (ii) and (iii); or by the features (ii) and (vi); or by the features (ii) and (vii); or by the features (iii) and (vi); or by the features (iii) and (vii); or by the features (vi) and (vii); or by the features (ii), (iii) and (vi); or by the features (ii), (iii) and (vii); or by the features (ii), (vi) and (vii); or by the features (iii), (vi) and (vii); or by the features (ii), (iii), (vi) and (vii).

[0093] In a more preferred embodiment, the in vitro 3D brain tissue model is characterized by at least one further feature selected from the group of: a. matured astrocytes expressing Aldhl 11 and / or Aqp4; b. expression of reactive astrocyte markers upon addition of inflammatory cytokines, selected from the group comprising NFKB, C3, SERPIN3A, and LCN2; c. expression of excitatory and inhibitory synaptic markers, including GRIA1 and GABRA1 , respectively; d. presence of spontaneous action potentials indicating neuronal activity; e. mature microglia having an immune-sensing state; f. Increased enrichment for a postnatal compared to a fetal human microglia signature upon maturation in tissue model; g. reaction of microglia to disease state, e.g., to pathologic Ap in an Alzheimer’s disease model induced by endogenous mutations or exogenous seeding. In a more preferred embodiment, the in vitro 3D brain tissue model is characterized by, with increasing preference, at least two, at least three, at least four, at least five, at least six or all features selected from the group of (a.) to (g.).

[0094] In an even more preferred embodiment, the in vitro 3D brain tissue model is characterized by (e.) or (f.), wherein feature (f.) is particularly preferred, or by both the features (e.) and (f.).

[0095] The term “ramified morphology” in accordance with the present invention refers to the presence of highly branched microglia characterized by a small soma and fine ramifications, wherein the microglia are extending 3 or more branched processes into the surrounding tissue with each process being longer than the average diameter of the cell body. Ramified morphology is observable and quantifiable by methods known in the art such as immunofluorescence stainings.

[0096] The term “necrotic core” means a mass of dying and / or dead cells which is often caused by a lack of oxygen (hypoxia) and nutrients (starvation). Thus, a necrotic core in certain embodiments is also referred to as “hypoxic core”. In general, the presence of apoptotic and necroptotic cells can be visualized by TUNEL staining, wherein a variety of deoxynucleotides, such as d-UTP, are modified / tagged with fluorophores (or haptens such as biotin) and incorporated at the 3’OH-ends of fragmented DNA by the enzyme terminal deoxynucleotidyl transferase (TdT). The fluorophores / tags can be detected with antibodies by fluorescence cytometry and microscopy. Methods are known in the art, such as usage of commercially available TUNEL assay kits, for example the “In Situ Cell Death Detection Kit, TMR red” (Roche, Cat#12156792910), and are described herein below.

[0097] As used herein, the term “Aldh1 l1 ” refers to “cytosolic 10-formyltetrahydrofolate dehydrogenase” and is encoded by all sequences deposited under the NCBI Reference Sequence NP_001257293, NP_001257294, and NP_036322; Gene ID: 10840, and Uniprot ID: 075891. The term “Aldhl 11” also refers to functional allelic variants thereof that are distinct from the hitherto described sequences by at least one base pair, but maintain with increasing preference at least 50%, 80%, 90% or 100% transcription factor activity of the described transcription factor. Aldhl 11 catalyzes the NADP+- dependent conversion of 10-formyltetrahydrofolate to tetrahydrofolate and carbon dioxide.

[0098] The term “Aqp4”, as used herein, refers to Aquaporin 4, an integral membrane protein which forms a water-specific channel and plays an important rale in brain water homeostasis. It is encoded by all sequences deposited under the NCBI Reference Sequence NP_001641 , NP_004019, NP_001304316, NP_001304313, NP_001351215, NP_001351216, NP_001351218; Gene ID: 361 , and Uniprot ID: P55087. The term “Aqp4” also refers to functional allelic variants thereof that are distinct from the hitherto described sequences by at least one base pair, but maintain with increasing preference at least 50%, 80%, 90% or 100% transcription factor activity of the described transcription factor.

[0099] The term “neurocan”, as used herein, refers to the neurocan core protein, which is known to modulate neuronal adhesion and neurite growth during development by binding to neural cell adhesion molecules (NG-CAM and N-CAM). It is encoded by all sequences deposited under the NCBI Reference Sequence NP_004377, such as NP_004377.2; Gene ID: 1463, and Uniprot ID: 014594. The term “neurocan” also refers to functional allelic variants thereof that are distinct from the hitherto described sequences by at least one base pair, but maintain with increasing preference at least 50%, 80%, 90% or 100% transcription factor activity of the described transcription factor.

[0100] The term “brevican”, as used herein, refers to the brevican core protein, which plays a role in the terminal differentiation and the adult nervous system during postnatal development and is a member of the lectican family of chondroitin sulfate proteoglycans that is specifically expressed in the central nervous system. This protein is developmentally regulated and may function in the formation of the brain extracellular matrix, is highly expressed in gliomas, and may promote the growth and cell motility of brain tumor cells. It is encoded by all sequences deposited under the NCBI Reference Sequence NP_ NP_940819 and NP_068767; Gene ID: 63827, and Uniprot ID: Q96GW7. The term “brevican” also refers to functional allelic variants thereof that are distinct from the hitherto described sequences by at least one base pair, but maintain with increasing preference at least 50%, 80%, 90% or 100% transcription factor activity of the described transcription factor.

[0101] The term “CSPG5”, as used herein, refers to “chondroitin sulfate proteoglycan 5” and is a proteoglycan that may function as a neural growth and differentiation factor. It is encoded by all sequences deposited under the NCBI Reference Sequence NP_006565, NP_004019, NP_001193871 , NP_001193872, NP_001 193873, NP_001193874; Gene ID: 10675, and Uniprot ID: 095196. The term “CSPG5” also refers to functional allelic variants thereof that are distinct from the hitherto described sequences by at least one base pair, but maintain with increasing preference at least 50%, 80%, 90% or 100% transcription factor activity of the described transcription factor.

[0102] The term “tenascins”, as used herein, refers to a family of extracellular matrix proteins that are involved in neurite outgrowth and cell migration in hippocampal explants. Tenascins are referred to herein as linkage molecules of the ECM. Non-limiting examples are Tenascin-C (Uniprot: P24821) and Tenascin- R (Uniprot: Q92753). The term “tenascins” also refers to functional allelic variants thereof that are distinct from the hitherto described sequences by at least one base pair, but maintain with increasing preference at least 50%, 80%, 90% or 100% transcription factor activity of the described transcription factor.

[0103] As used herein, the term “inflammatory cytokine” refers to a type of cytokine (a signaling molecule) that is secreted from immune cells and certain other cell types that promotes systemic inflammation.

[0104] The term “GRIA1 ”, as used herein, refers to “glutamate ionotropic receptor AMPA type subunit 1 ”, generally also known as “glutamate receptor 1 ” and is ionotropic glutamate receptor that functions as a ligand-gated cation channel, gated by L-glutamate and glutamatergic agonists such as alpha-amino-3- hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA), quisqualic acid, and kainic acid. It is encoded by all sequences deposited under the NCBI Reference Sequence NP_000818, NP_001107655, NP_001244948, and NP_001244949; Gene ID: 2890, and Uniprot ID: P42261. The term “GRIA1 ” also refers to functional allelic variants thereof that are distinct from the hitherto described sequences by at least one base pair, but maintain with increasing preference at least 50%, 80%, 90% or 100% transcription factor activity of the described transcription factor.

[0105] The term “GABRA1 ”, as used herein, refers to “gamma-aminobutyric acid receptor subunit alpha-1 ” and is a gamma-aminobutyric acid (GABA) receptor. GABA is the major inhibitory neurotransmitter in the mammalian brain where it acts at GABA-A receptors, which are ligand-gated chloride channels. It is encoded by all sequences deposited under the NCBI Reference Sequence NP_001121116, NP_000797, NP_001121115, and NP_001 121117, and NP_001 121120; Gene ID: 2554, and Uniprot ID: P14867. The term “GABRA1 ” also refers to functional allelic variants thereof that are distinct from the hitherto described sequences by at least one base pair, but maintain with increasing preference at least 50%, 80%, 90% or 100% transcription factor activity of the described transcription factor.

[0106] In accordance with the present invention the term “neuronal activity” refers to the changes in membrane potential caused by ion influxes over an electrochemical gradient characterized by the presence of spontaneous action potentials and calcium activity as detected by calcium sensors, as, for example, Fluo-4.

[0107] As used herein, the term “immune-sensing state” in conjunction with microglia, refers to the state wherein microglia sense and respond rapidly to alternations in homeostasis and integrate the neural response to threats and stressors, as indicated by expression of receptors and molecules that are summarized in the “sensome” gene expression signature (Hickman, S., Kingery, N., Ohsumi, T. et al. The microglial sensome revealed by direct RNA sequencing. Nat Neurosci 16, 1896-1905 (2013). https: / / doi.Org / 10.1038 / nn.3554

[0108] The term “microglia signature”, as used herein, refers to the presence of certain markers which have previously been used to define gene-expression states in microglia associated with physiological, functioning, homeostatic microglia, but also disease-associated, activated microglia. Non-limiting examples to study microglia signature are qPCR, Nanostring, proteomics and scRNA sequencing.

[0109] The method of “scRNA” (single cell RNA) sequencing is described in detail in the examples below.

[0110] In a third aspect of the present invention, the in vitro 3D brain tissue model of the second aspect is applied for studying physiological and disease phenotypes of mature human brain.

[0111] The physiological tissue environment provided by the 3D BTM of the present invention enables modeling of neurological diseases by genome-editing of the iPSCs and precursor cells prior to differentiation into neuron, astrocytes and / or microglia. Using CRISPR / Cas9 genome editing, the inventors established a human AD model based on knock-in of synergistic APP mutations, which displayed disease-associated features including amyloid aggregation, increased phospho-tau levels, neuroinflammation and altered microglial expression signatures relevant for AD (see Example 6, Figures 6, 8 herein below). As shown in the corresponding mouse model, the mutations have synergistic effects on Ap aggregation, and they also induced accumulation and aggregation of Ap in the in vitro system using the 3DBTM model of the present invention.

[0112] Accordingly, in a more preferred embodiment, the in vitro 3D brain tissue model of the above preferred embodiment of the third aspect is applied for studying Alzheimer’s Disease phenotypes selected from the group of Ap accumulation, phospho-tau increase, aggregation of Ap, neuroinflammation, and AD- related changes of microglia signature.

[0113] “Alzheimer’s Disease phenotypes” are well-described in the art and comprise Ap aggregation and plaque formation, altered tau phosphorylation, neuroinflammation, and neurodegeneration. The Alzheimer’s Disease phenotypes in accordance with the present invention refer to Amyloid beta accumulation, phospho-tau increase, aggregation of amyloid-beta, neuroinflammation, and / or AD- related changes of microglia signature in scRNA sequencing.

[0114] The term “Amyloid-beta”, short Ap (Abeta), refers to a protein in the human brain that is naturally produced and decomposed in a normally developing and aging healthy brain, wherein its concentration increases with age. Ap is produced by sequentially cleaving amyloid precursor protein (APP) proteolytically by enzymes p- and y-secretase and the sequential cleavage of y-secretase causes formation of Ap peptide isoforms of different lengths, with longer isoforms, such as Ap42 and Ap43 being more aggregation-prone and toxic.

[0115] “Ap accumulation”, as used herein, indicates faulty decomposition of the protein, which leads to formation of Ap clumps, also called amyloid plaques. Ap can also aggregate and misfold to form oligomers and / or fibrils. The presence of accumulation of Ap, especially in form of plaques, is associated with Alzheimer’s Disease.

[0116] Phosphorylation of Tau (P-Tau) has been described as a major marker of pathologic Tau, especially if present at multiple, disease-relevant epitopes. Normal Tau is phosphorylated to a certain degree at specific epitopes to ensure the dynamic character of the microtubule network (see below). Tau in its aggregated form (as Paired Helical Filaments (PHFs) or Straight Filaments (SF)) contains 5-9 moles of phosphate / mole of the protein, defining it as hyperphosphorylated (Lippens, G., Sillen, A., Landrieu, L, Amniai, L., Sibille, N., Barbier, P., ... Wieruszeski, J. M. (2007). Tau Aggregation in Alzheimer’s Disease: What Role for Phosphorylation? Prion, 1 (1), 21-25. https: / / doi.Org / 10.4161 / pri.1.1.4055.; Kopke E, Tung YC, Shaikh S, Alonso AC, Iqbal K, Grundke-lqbal I. Microtubule-associated protein tau. Abnormal phosphorylation of a non-paired helical filament pool in Alzheimer disease. J Biol Chem. 1993 Nov 15;268(32):24374-84. PMID: 8226987.). The term “phospho-tau increase” may also be referred to as “abnormal phosphorylation”. Assays to detect / measure P-Tau and localization of phosphorylated Tau include but are not limited to immunostaining with an antibody, such as AT8, AT100, AT180, PHF-1 , PHF-13, and antibodies specific for pT181 , pS198, pS199, pT205, pS202+pT205, pT212, pS214, pT231 , pS238, pS262, E178, pS396, pS404, pS409, pS422. The term “neuroinflammation”, as used herein, refers to an inflammatory response within the brain or spinal cord, wherein cytokines, chemokines, reactive oxygen species and secondary messengers are produced and increased proliferation of glial cells, especially astrocytes and microglia, is observed. Neuroinflammation in general comprises immune, physiological, biochemical, and psychological responses mediated by microglia, astrocytes, endothelial cells and peripherally derived immune cells and may also comprise positive inflammatory responses (see DiSabato, D.J., Quan, N. and Godbout, J.P. (2016), Neuroinflammation: the devil is in the details. J. Neurochem., 139: 136-153. https: / / doi.org / 10.11 11 / jnc.13607). In accordance with the present invention, the term “neuroinflammation” mainly refers to the negative, pathology-associated inflammatory responses which are associated with Alzheimer’s Disease. Neuroinflammatory responses are seen in the model by activation of microglia as seen by increased expression of pro-inflammatory factors NFKB, CD68, C3 and C1 QC, and a general activation of microglia seen in single-cell RNA sequencing, which mimics the activation seen in microglia extracted from AD patient brains.

[0117] In 2D systems known in the art, the formation of dense-core, plaque-like structures as hallmark of AD has also been shown by addition of exogenous Ap. Those dense-core plaques comprise highly compacted Ap fibrils which can be stained with compounds binding to their characteristic p-sheet structure

[0051] .

[0118] In a fourth aspect of the present invention, the in vitro 3D brain tissue model of the second aspect is applied for drug profiling, testing and / or screening or toxicity testing, wherein the 3D brain tissue model is preferably upscaled.

[0119] The term “upscaled”, as used herein, refers to the generation of several hundreds to thousands of 3D BTM cultures at once to enable their use for drug profiling, testing and / or screening or toxicity testing.

[0120] In another preferred embodiment, the in vitro 3D brain tissue model, as far as obtained or obtainable by the method wherein oligodendrocytes (OLG), endothelial cells or pericytes are added to the culture after step c. or d. is for studying physiological and disease phenotypes of multiple sclerosis, optic neuritis, transverse myelitis, acute disseminated encephalomyelitis (ADEM), neuromyelitis optica (NMO), adrenoleukodystrophy or Alexander disease.

[0121] In the latter preferred embodiment, it is also preferred to add exogenous Ap42 to the 3D brain tissue model.

[0122] The term “a” or “an” can refer to one of or a plurality of the elements it modifies (e.g., “a reagent” can mean one or more reagents) unless it is contextually clear either one of the elements or more than one of the elements is described. The term “about” as used herein refers to a value within 10% of the underlying parameter (i.e., plus or minus 10%), and use of the term “about” at the beginning of a string of values modifies each of the values (i.e., “about 1 , 2 and 3” refers to about 1 , about 2 and about 3). For example, a weight of “about 100 grams” can include weights between 90 grams and 110 grams. Further, when a listing of values is described herein (e.g., about 50%, 60%, 70%, 80%, 85% or 86%) the listing includes all intermediate and fractional values thereof (e.g., 54%, 85.4%).

[0123] As regards the embodiments characterized in this specification, in particular in the claims, it is intended that each embodiment mentioned in a dependent claim is combined with each embodiment of each claim (independent or dependent) where said dependent claim depends from unless indicated otherwise. For example, in case of an independent claim 1 reciting 3 alternatives A, B and C, a dependent claim 2 reciting 3 alternatives D, E and F and a claim 3 depending from claims 1 and 2 and reciting 3 alternatives G, H and I, it is to be understood that the specification unambiguously discloses embodiments corresponding to combinations A, D, G; A, D, H; A, D, I; A, E, G; A, E, H; A, E, I; A, F, G; A, F, H; A, F, I; B, D, G; B, D, H; B, D, I; B, E, G; B, E, H; B, E, I; B, F, G; B, F, H; B, F, I; C, D, G; C, D, H; C, D, I; C, E, G; C, E, H; C, E, I; C, F, G; C, F, H; C, F, I, unless specifically mentioned otherwise.

[0124] Similarly, and also in those cases where independent and / or dependent claims do not recite alternatives, it is understood that if dependent claims refer back to a plurality of preceding claims, any combination of subject-matter covered thereby is considered to be explicitly disclosed. For example, in case of an independent claim 1 , a dependent claim 2 referring back to claim 1 , and a dependent claim 3 referring back to both claims 2 and 1 , it follows that the combination of the subject-matter of claims 3 and 1 is clearly and unambiguously disclosed as is the combination of the subject-matter of claims 3, 2 and 1 . In case a further dependent claim 4 is present which refers to any one of claims 1 to 3, it follows that the combination of the subject-matter of claims 4 and 1 , of claims 4, 2 and 1 , of claims 4, 3 and 1 , as well as of claims 4, 3, 2 and 1 is clearly and unambiguously disclosed.

[0125] The Figures show:

[0126] Figure 1 : Generation of a reproducible human 3D brain tissue model (3BTM). (A) Experimental approach to generate 3BTMs by self-aggregation of NE and AS in ultra-low attachment plates. (B) Overview of a representative 96-well plate with 3BTMs at 1 month of age, note uniform size and shape of the cultures. Right: Representative brightfield image of a 3D culture at 1 month, scale bar 100 pm. (C) Quantification of culture diameters in brightfield images over time. (D) Representative immunofluorescence (IF) stainings (left) and quantification (right) of 3D co-cultures over time showing Ki67-positive, proliferative cells at 3 days and absence of proliferating cells at later timepoints after 5FU treatment (n = 5, one-way ANOVA with Tukey’s multiple comparisons test), scale bar 100 pm. (E) Representative IF staining showing NeuN-positive NE and Sox9-positive AS in 3D co-cultures (scale bar 30 pm). Quantification of the NeuN:Sox9 ratio on the right. (F) Representative IF staining of a middle slice from a NE+AS culture at 3 months for MAP2 (NE) and GFAP (AS) showing dense and uniform neuritic network formation (scale bar 100 pm), zoom-in at the bottom (scale bar 20 pm). (G) Representative staining in 3-month-old NE+AS 3D co-cultures using Hypoxyprobe showing only minimal hypoxic areas under normal culture conditions (20% 02) and higher levels in a positive control under hypoxic conditions (5% 02). Quantification on the right, (n = 2-4, t-test), scale bar 100 pm. (H) Experimental approach to generate 3BTMs containing NE, AS, and MG. NE and AS are added into ultra-low-attachment wells and aggregate. MG are added 15 days later into the media and migrate into the cultures. (I) Representative brightfield images showing migration of MG towards and into 3D cultures over the course of one week, scale bar 100 pm. (K) Representative IF staining of NE+AS+MG 3D coculture at 1 month of age showing even distribution of microglia throughout the culture, scale bar 100 pm. (L) Quantification of MG numbers in 3D co-culture without and with MG at 1 month showing improved incorporation in the presence of MG (n = 3, t-test). (M) Representative IF staining of 3D cocultures at 6 months showing long-term survival of densely branched NE and AS networks and MG without necrotic core formation, scale bar 100 pm. (N) Representative IF staining of a middle slice from a 3-month-old NE+AS+MG culture showing the neuronal (MAP2) and astrocytic (GFAP) networks without necrotic core and the distribution of incorporated MG (Iba1) (scale bar 100 pm). Zoom-in showing physiological cell morphologies with fine and dense processes of NE and AS, and ramifications of MG (scale bar 20 pm). (O) Top left: Representative IF staining of MG at 6 months showing highly ramified morphology, scale bar 10 pm. Bottom left: quantification of ramification index and number of branches (#branches) per cell (n = 3, 61 -192 cells per timepoint, Kruskal-Wallis test with Dunn’s multiple comparisons test). Right: 3D reconstructions of representative MG morphologies at 1 , 3, and 6 months, scale bar 20pm. (P) IF stainings at 3 months for homeostatic membrane marker P2RY12 (scale bar 15 pm), quantification of P2RY12-positive cells on the right (n=3, one-way ANOVA with Tukey’s multiple comparisons test). (Q) Quantification of MG numbers over time. (R) Representative IF staining of 6- month-old NE+AS+MG cultures repopulated with MG at 5 months, scale bar 100 pm. (S) Representative IF staining of repopulated MG after 1 month in culture showing ramified morphologies, scale bar 20 pm. (T) Quantification showing percentage of P2RY12-positive cells at 6 months in cultures repopulated with MG at 5 months of age. (U) Representative IF stainings of 3D NE mono-cultures or NE+AS / NE+AS+MG co-cultures at 6 months, scale bar 100 pm. Quantification of TUNEL stainings on the right (n = 4, Repeated measures two-way ANOVA with Tukey’s multiple comparisons test comparing all cell type combinations between each timepoint). (V) Quantification of culture diameters at 1 , 3 and 6 months in an independent iPSC line, each dot represents one culture, n = 3. (W) Quantification of NE:AS ratio at 1 and 3 months in an independent iPSC line. (X) Quantification of Ki67-positive cells in an independent iPSC line at 1 month. (Y) IF staining of a 3-month-old NE+AS+MG culture of an independent iPSC line, scale bar 100 pm. Zoom-in on the right, scale bar 10 pm. (Z) Representative IF stainings of 2D hippocampal neuron differentiations (left) and of a hippocampal 3BTM culture at 3 months of age (right) for typical marker ZBTB20, scale bars 50 pm (left) and 100 pm (right). Data are represented as mean + SD.

[0127] Figure 2: 3BTM maturation increases similarity to human brain tissue. (A) Principal component analysis (PCA) of proteomic datasets based on all proteins after imputation showing increasing similarity of 3D cultures to juvenile human brain with increasing age of the cultures. Shown is the principal component 1 (PC1) which accounts for 43.8% of the variability in the dataset. Brain org = brain organoid sample (3 technical replicates), 3d NE+AS = 3-day-old 3BTM (6 biological replicates), fet brain = fetal brain (3 biological replicates), 3mo = 3-months-old 3BTM (6 biological replicates), 6mo = 6-months-old 3BTM (5 biological replicates), juv brain = juvenile brain (3 technical replicates). (B) Gene ontology (GO) analysis of top proteins influencing PC1 showing upregulated (red) or downregulated (blue) biological processes, sorted by false-discovery rate (FDR). (C) Electron micrographs of a NE+AS+MG culture at 3 months showing formation of synapses (left, scale bars 200 nm) with presynaptic terminals (arrows) and postsynaptic densities (arrowheads) and in some cases tripartite synapses (right, scale bar 200 nm) made of presynaptic axon (pseudo-colored in blue), postsynaptic dendrite (green), and adjacent astrocyte (purple). (D) Representative IF stainings of NE+AS+MG cultures analyzed over 6 months showing synapse formation by increasing amounts and colocalization of presynaptic Synapsin-1 (Syn1) and postsynaptic PSD-95 in synaptic puncta, scale bar 10 pm. Right: Quantification showing increasing synaptic puncta normalized to MAP2 over time, each colored line represents one experiment (n = 4, repeated-measures one-way ANOVA with Tukey’s multiple comparisons test). (E) Heatmap showing Iog2 fold changes (Iog2 FC) in protein levels of synapse-associated proteins found in proteomics datasets of brain organoids (BO), fetal brain (FB), 3-months-old NE+AS+MG cultures (3mo), 6-months-old NE+AS+MG cultures (6mo), and juvenile brain (JB) compared to 3-day-old NE+AS cultures (set to zero). Hierarchical clustering of the different samples on the right confirming similarity of 3-months and 6- months samples with fetal and juvenile brain. (F) IF stainings of NE+AS cultures at 3 days and 3 months showing deposition of Hyaluronan in 3D cultures over time, scale bars 100 pm. (G) Heatmap and clustering as in (E) showing Iog2 fold changes of extracellular matrix proteins in brain organoid, human brain samples, and 3- and 6-month-old 3BTMS, confirming increased similarity of physiological 3BTMs to fetal and juvenile brain samples compared to brain organoids and 3-day-old 3BTMs. (H) Pearson correlation of Iog2 transformed LFQ intensities of all detected proteins showing high reproducibility between 3D culture samples with similar correlation coefficients as technical replicates of brain organoid and juvenile brain samples. (I) IF staining showing mostly extracellular deposition of hyaluronan, as seen by exclusion of staining in DAPI-positive cell bodies, scale bar 10 pm. (J) IF staining showing increased levels of hyaluronan at 3 days in cultures containing AS (labelled by GFAP), scale bar 20 pm. (K) Heatmap and clustering as in (E) showing Iog2 fold changes in astrocyte maturation markers showing increased similarity of mature 3BTMs to juvenile brain. (L) Heatmap showing Iog2 fold changes (Iog2 FC) in protein levels of cell-cycle-associated proteins found in proteomics dataset in brain organoids (BO), fetal brain (FB), 3-month-old NE+AS+MG cultures (3mo NAM), 6-month-old NE+AS+MG cultures (6mo NAM), and juvenile brain (JB) compared to 3-day-old NE+AS cultures (set to zero). Hierarchical clustering of the different samples on the right confirming similarity of 3-months and 6-months samples with juvenile brain. Data are represented as mean + SD.

[0128] Figure 3: Microglia mature in 3BTMs and adopt transcriptomic signatures similarto human brain. (A) Schematic overview of the experimental design. MG from two independent experiments were isolated at 1 month or 3 months and processed for scRNAseq. (B) UMAP representation (res=0.3, dims = 1 :10) of 8827 cells and cluster annotation of MG at 1 and 3 months of age. (C) Feature and violin plots showing expression of key microglia genes TYROBP, AIF1, and HEXB at 1 and 3 months. (D) Top: Differential gene expression analysis (DEA) of MG at 1 or 3 months. Bottom: Gene set enrichment analysis of differentially expressed genes (DEGs) showing upregulation of immune functions and downregulation of metabolism-related pathways. (E) Comparison of MG signatures at 1 and 3 months with a previously published sensome signature

[0036] shows increased enrichment at 3 months by enrichment score (ES) as distribution (left) or projected onto UMAP (right). (F) Pseudo-time analysis showing trajectories from proliferative cells at 1 month towards resting and more activated states at 1 month and branching towards the 3 months clusters. Subset on the right showing trajectory for 1 -month MG alone. (G) UMAP plots showing integrative analysis of fetal and postnatal MG

[0037] with 3BTM MG at 1 and 3 months. (H) UMAPs of integrative analysis as in (C), showing enrichment for adult microglia signature by Galatro et al.

[0040] in the different datasets. (I) Violin plot reporting fetal vs. adult microglia signature (extracted from Yaqubi et al., DOI: 10.1186 / s12974-023-02809-7) ES in MG at 3 months. (K) In vitro tracker reporting ES of MG at 3 months across different model-specific signatures derived from the integrated dataset by Popova et al.

[0020] , including primary human fetal microglia cultured in 2D (2D, human primary fetal) or incorporated into cortical organoids (oMG, human primary fetal) as well as human iPSCs-derived microglia either cultured in 2D (2D, human iPSCs-derived) or xenotransplanted into mouse brain (xMG, human iPSCs-derived).

[0129] Figure 4: Neurons, astrocytes, and microglia in 3BTMs acquire functionalities typical for mature brain cells. (A) Patch-clamp measurement of single neurons in NE+AS+MG 3BTMs at 3 months showing spontaneous action potentials. (B) Patch clamp measurements as in (A) with current injection showing evoked action potentials. (C) Application of tetrodotoxin (TTX) inhibits action potential firing, quantification on the right (n = 4, paired t-test). (D) Resting membrane potential (EM) measurements showing values of --60-70 mV in most neurons analyzed. (E) qPCR analysis of NE+AS cultures at 1 month of age treated with PBS (ctrl) or TNFa+IL1a (Tl). Tl treatment induces strong upregulation of inflammation markers NFKB, C3, SERPINA3, and LCN2, suggesting astrocyte activation. (F) IF staining for DAPI-positive DNA and I ba1 -positive MG in 3-month-old 3BTM showing microglial process engulfing a dead cell’s nucleus (as seen by bright and fragmented DAPI staining, arrowhead), orthogonal projections in YZ and XZ confirm engulfment, scale bar 10 pm. (G) 3D surface rendering of IF staining for Synapsin-1 (Syn)-positive synaptic material and Iba1 -positive microglia at 1 month showing close association (left) and engulfment (right) of synaptic material by MG, scale bars 0.5 pm. (H) 2-Photon live-cell imaging of NE+AS+MG cultures at 3 months with MG expressing memYFP. Pictures showing different, color-coded timepoints (0, 6, 12 min), merged image on the right showing movement by colorcoding whereas stable structures appear white, maximum intensity projection of a 24 pm z-stack, scale bar 20 pm. (I) 2-Photon live-cell imaging as in E) showing reaction of microglia to a focal laser injury (red circle), maximum intensity projection of a 22 pm z-stack, scale bar 20 pm. (J) Spontaneous action potential (AP) frequency observed in patch-clamp measurements of single neurons in NE+AS+MG 3D cultures at 3 months showing spontaneous action potentials in 2 out of 7 neurons. (K) Evoked AP observed as in (J) showing heterogeneous firing patterns, suggesting neuronal heterogeneity. (L) Top: Examples images of live cell calcium imaging using Fluo-4 in NE+AS+MG 3BTMs at 3 months showing spontaneous and synchronized activity. Bottom: Quantification of Fluo-4 intensity over time, timeframe shown above is marked by lines. Data are represented as mean + SD.

[0130] Fig. 5: Oligodendrocytes in the 3D cultures (A) IF staining showing successful incorporation of the oligodendrocytes (labeled with an intracellular transgene, RFP) into 3D cultures also containing AS (labeled by GFAP), NE (labeled by MAP2), and MG (labeled by Iba1). (B) IF staining demonstrating increase of MBP levels over time (quantification in the right). (C) Electron microscopy (EM) showing successful axon myelination in the human in vitro 3D model.

[0131] Figure 6: APP mutant 3BTMs recapitulate central AD phenotypes. (A) MSD Immunoassay (IA) of 3BTM supernatants at 1 month to measure secretion of Ap38 / 40 / 42 showing increased total Ap levels (sum of all 3 species) in KI cultures (n = 3, unpaired t-test). (B) IA as in (A) showing increased Ap42:40 ratio in KI cultures (n = 3, unpaired t-test). (C) Representative IF stainings of 3BTMs over time showing accumulation of Ap (82E1 antibody) specifically in KI cultures, scale bar 20 pm. (D) Quantification of Ap IF stainings of KI and WT 3BTMs showing progressive accumulation in KI cultures (n = 4, two-way ANOVA with Sidak’s multiple comparisons test). (E) Quantification of Ap IF stainings in S6A-B after 6- months treatment with DMSO (as control) or 5 pM BACE-inhibitor C3 showing reduced Ap accumulation upon BACE inhibition (n = 4, unpaired t-test). (F) IA of insoluble fractions extracted from WT and KI 3BTMs showing insoluble Ap42 in KI cultures at 1 month and increased aggregation at 6 months (n = 3, two-way ANOVA followed by Tukey‘s multiple comparisons test). (G) Quantification of western blot (WB) of WT and KI 3BTMs (see Figure S6D) showing increased phospho-Tau (pTau, AT270 antibody - Thr 181 epitope) levels in KI cultures (n = 3, unpaired t-test). (H) WB analysis as in (G) showing no significant difference in pTau levels in 3BTMs containing only NE+AS (left) and significant difference in those containing NE+MG (right) (n = 3, unpaired t-tests). (I) WB analysis as in (G) showing significant increase in pTau levels (PHF-1 antibody, Ser396 / 404 epitope) in KI 3BTMs containing all 3 cell types at 3 months of age (see Figure S6E, n = 3, unpaired t-test). (J) Representative IF stainings of KI cultures at 6 months treated with DMSO (loading control), or 5 pM C3 (BACE-inhibitor IV) showing reduced Ap accumulation after C3 treatment, scale bar 20 pm. (K) Representative IF stainings of KI cultures with different cell type combinations at 6 months showing increased accumulation of Ap in the presence of AS and MG. Quantification on the bottom showing a time course for each combination (n = 4, two-way ANOVA with Tukey’s multiple comparisons test), scale bar 20 pm. (L) Immunoassay to detect Ap in supernatants of 1 -month-old KI cultures containing different cell type combinations showing no significant difference in total Ap production or Ap42:40 ratio. (M) Western blot (WB) analysis for phospho-tau (pTau, AT270 antibody - Thr181 epitope) and total Tau (tTau, K9JA antibody) of 3-months- old KI cultures containing different cell types, quantification in Fig. 6G / 6H. (N) WB analysis as in (L) for pTau using the PHF1 antibody (Ser396 / Ser404), quantification in Fig. 6I. Data are represented as mean + SD.

[0132] Figure 7: Microglia in AD-modeling 3BTMs recapitulate disease-related changes. (A) qPCR analysis of WT and KI 3BTMs at 3 months showing increased inflammatory markers NFKB and CD68 in KI cultures (n = 3, t-tests). (B) qPCR analysis as in (A) showing increased levels of C3 and C1 QC in KI cultures (n = 3, t-tests). (C) UMAP representation (res=0.4, dims = 1 :10) of the 3-month subset (subset 2, Fig. S1 B) of 10828 cells and cluster annotation of WT and KI MG. (D) UMAP plot split by genotype (E) Number of differentially expressed (DE) genes (p.adj<0.01 , log2FC>0.25) obtained by comparing “KI cluster 2” and “WT resting” clusters. (F) Gene set enrichment analysis (GSEA, gene ontology, biological process) of DEGs showing enrichment of upregulated genes in chemotaxis and immune activation and of downregulated genes in antigen presentation (q-value<0.2). A subset of up- and downregulated genes contributing to the load of GSEA functional terms is shown. (G) Violin plot reporting the AD microglia signature (by Zhou et al. [8]) enrichment score (ES) across “WT resting”, “KI cluster 1 ” and “KI cluster 2” cells (Wilcoxon signed-rank test). (H) Venn diagram showing overlap of KI MG signature from this invention with human AD microglia signature reported by Zhou et al. [8].

[0133] Figure 8: Addition of exogenous A 42 elicits plaque-like pathology and enables anti-amyloid antibody testing. (A) Ap IF staining of NE+AS+MG 3BTMs at 1 .5 months of age after addition of 2x 500 nM recombinant Ap42 on days 8 and 10, scale bar 100 pm. (B) Close-up of IF staining as in (A) showing formation of a plaque-like structure with aggregated core (stained by pFTAA) and fibrillar halo (positive for Ap), scale bar 10 pm. (C) IF staining as in (A) showing colocalization of Ap and I ba1 (arrow), scale bar 10 pm. (D) IF staining of NE+AS+MG 3BTMs at 1 month after addition of 2x 500 nM recombinant Ap42 on days 8 and 10, and treatment with 1 pg / ml control (ctrl, top left) or Aducanumab (Adu, top right) antibody on days 17, 21 , 24, 28 (scale bar 100 pm). Close-up on the bottom right showing boxed region, scale bar 10 pm. Quantification in top right showing percentage of Ap-positive (Ap+) area which is also antibody-positive (n=3, paired t-test). (E) IF staining as in (D) showing Ap accumulation in cultures treated for 1 month with control antibody (left) or Aducanumab (middle), scale bar 100 pm, quantification of Ap-positive area after 14 days or 1 month on the right (n=3, one-way ANOVA with Tukey’s multiple comparisons test). (F) IF staining showing co-localization of Ap and Aducanumab inside Iba1-positive microglia, arrows showing same position in each image, scale bar 10 pm. (G) Orthogonal views of IF stainings shown in (F). (H) IF staining for Ap and Iba1 in 3BTMs as described in (D), arrows pointing to colocalized signal. Scale bar 20 pm. Quantification on the right showing Ap load (Ap-positive area inside microglia) (n = 3, 11-22 microglia per replicate, Mann-Whitney test). (I) IF staining for CD68 and I ba1 in 3BTMs as described in (D), scale bar 20pm. Quantification on the right showing mean fluorescent intensity (MFI) of the CD68-positive area inside microglia (n = 3, 19-27 microglia per replicate, Mann-Whitney test).

[0134] Figure 9: Comparison of 3BTMs with 2D co-cultures of NE, AS, and MG. (A)-(C) Single-cell RNA sequencing analysis of NE extracted from 3BTMs (3D) or from 2D co-cultures at 3 months of age. (A) Differential gene expression analysis (DEA) of NE from 3BTMs or 2D co-cultures showing upregulation of neuron-function related pathways including synapse- and signal-transduction-associated terms in 3D. (B) Violin plots comparing signatures from NE in 2D or in 3BTMs with signatures containing genes involved in axon development (top) and synapse organization (bottom) by enrichment score (ES). Note the increased enrichment in NE from 3BTMs. (C) Heatmap showing expression of the top DE genes in the term “Synapse organization” in 2D vs. 3D showing higher expression in 3D cultures. (D)-(F) Singlecell RNA sequencing analysis of AS extracted from 3BTMs or from 2D co-cultures at 3 months of age. (D) DEA of AS from 3BTMs or 2D co-cultures showing upregulation of astrocyte-function related pathways including “cilium organization” and downregulation of pathways involved in AS reactivity such as “Antigen processing and presentation”. (E) Violin plots comparing expression of AS function-related genes in 2D or in 3BTMs. Note the increased expression in AS from 3BTMs. (F) Heatmap showing expression of the top DE genes in the term “Cilium organization” in 2D vs. 3D showing higher expression in 3D cultures. (G)-(K) Single-cell RNA sequencing analysis of MG extracted from 3BTMs or from 2D co-cultures at 3 months of age. (G) Differential gene expression analysis (DEA) of MG from 3BTMs or 2D co-cultures showing downregulation of reactivity-related processes including “activation of immune response”, “taxis”, and “phagocytosis” in 3D. (H) Violin plots comparing expression of homeostatic MG genes in 2D or in 3BTMs showing increased expression in MG from 3BTMs. (I) Heatmap showing expression of DE genes related to microglia activation confirming increased activation in 2D vs. 3D cultures. (K) Violin plots comparing signatures from MG in 2D or in 3BTMs with signatures containing genes involved in immune activation (top) and an adult human MG signature (bottom, extracted from Yaqubi et al., 2023 J Neuroinflammation), by enrichment score (ES). Note the decreased enrichment for immune activation and the increased enrichment for the adult MG signature in MG from 3BTMs. (L)- (M) Confirmation of scRNA sequencing results on protein level in proteomic dataset comparing 2D cocultures with 3BTMs. (L) Heatmap showing Iog2-fold changes (log2FC) in expression of proteins included in the term “synapse organization” (as in (C)) in 3BTMs compared to 2D co-cultures. Note the upregulation of most proteins in 3D cultures, confirming scRNA sequencing results. Not all genes included in (C) were detected on the protein level and only proteins detected in all replicates of both conditions are shown, (n = 5, unpaired t-test). (M) Heatmap as in (L) showing log2FC in expression of proteins involved in AS function (as in (E)) in 3D vs. 2D cultures, confirming higher expression in 3BTMs. (n = 5, unpaired t-test). (N) Heatmap as in (L) showing log2FC in expression of proteins involved in MG activation (as in (I)) in 3D vs. 2D cultures. Not all genes included in (I) were detected on the protein level and only proteins detected in all replicates of both conditions are shown (n = 5, unpaired t-test). (O) Heatmap showing log2FC at 3 months vs. 3 days in protein levels of astrocyte maturation markers (as in Fig. 2K) in 3BTMs and 2D co-cultures, suggesting overall increased AS maturation in 3D (n = 5, unpaired t-test). (P) Heatmap showing log2FC at 3 months vs. 3 days in levels of brain ECM-associated proteins in 3BTMs and 2D co-cultures, showing overall higher increases in 3BTMs. (Q) 2-Photon livecell imaging (as in Fig. 41) of 3BTMs (top) and 2D co-cultures (bottom) at 3 months with MG expressing memYFP. Pictures showing reaction of microglia to a focal laser injury (red circle) in 3BTMs, but not in 2D cultures, scale bars 50 pm. (R) Quantification of the percentage of cells reacting to the focal laser injury as in (Q) in 2D vs. 3D co-cultures showing stronger reaction in 3BTMs. Cells were counted as “reacting” if at least one process moved towards the lesion site, even if not reaching it in the time frame imaged or retracting half-way (n = 3-4, unpaired t-test). (S) Quantification of parameters of MG morphology as a proxy for homeostasis and maturity. Number of branches (#branches) and overall tree length (in pm) were quantified for MG in 2D co-cultures and 3BTMs at 3 months of age, note the increase in both parameters in MG from 3BTMs suggesting higher level of homeostasis / maturity. (T) Comparison of AS morphologies in 3BTMs (top) and 2D co-cultures (bottom) by incorporation of small percentages of eGFP-positive cells. Note the higher complexity of cells in 3BTMs. Scale bars 50pm.

[0135] Figure 10: Reproducibility of 3BTM properties across iPSC lines. (A) Quantification of 3BTM culture diameters in three independent iPSC lines and independent differentiations at 1 and 3 months showing consistent cultures sizes across all conditions. (B) Quantification of the percentage of Ki67-positive (Ki67+), proliferative cells in 3BTMs generated from three independent iPSC lines at 3 months of age showing almost complete absence of cell proliferation in all conditions. (C) Quantification of the percentage of TUNEL+ cells as a marker of cell death in 3BTMs generated from three independent iPSC lines at 3 months of age showing only minimal levels of cell death in all conditions. (D) Quantification of the fraction of NE and AS in 3BTMs generated from three independent iPSC lines at 3 months of age showing highly reproducible culture composition. (E)-(F) Quantification of MG morphology as measured by ramification index and total tree length (in pm) of the skeleton in 3BTMs generated from three independent iPSC lines ((E) - Line 1 , (F) Lines 2+3) showing consistent increases over time from 1 to 3 months in all conditions, suggesting reproducible increases in MG maturation / homeostatic state, (n = 3 independent experiments with 20-40 cells each, Mann-Whitney U test). (G) Principal component analysis (PCA) of proteomic dataset including 3BTMs generated from three independent iPSC lines (4- 5 biological replicates each) at 3 days and 3 months, and brain organoids (BO) generated from two independent iPSC lines at Day 90 (2 technical replicates each), based on all detected proteins after imputation. Note the dense clustering of 3BTM samples both within and between lines per timepoint, in contrast to the large distance between both BO lines. (H) Heatmap showing log2FC in protein levels of synapse-associated proteins found in proteomics dataset at 3 months normalized to 3-day-old cultures (set to zero). Shown are data for 3BTMs generated from three independent iPSC lines and for BOs generated from two independent iPSC lines. Note the consistent increase in protein levels across 3BTMs from different iPSC lines in contrast to the vast differences between BOs from different iPSC lines. (I) Heatmap as in (H) showing astrocyte maturation markers. Note the generally weaker increase in BOs suggesting lower maturity of AS. (K) Heatmap as in (H) showing brain ECM-related proteins. Note the consistent increase in 3BTMs from different lines in contrast to the differences in BOs from different lines, and the generally lower increase in BOs, suggesting higher variability and lower similarity of the ECM to human brain in BOs. (L) Heatmap as in (H) showing cell-cycle-associated proteins. Note again the consistent changes in 3BTMs in contrast to large variability across BOs from different lines.

[0136] Figure 11 : 3BTMs containing ApoE-KO AS replicate impairment in Ap clearance, showcasing modularity of the system. (A) Schematic overview showing treatment with synthetic Ap42 (sAp42) of 3BTMs containing WT NE, WT MG and either WT AS or APOE-KO AS. (B) Top: IF staining of 3BTMs with WT AS or ApoE-KO AS 24, 42 and 60 days after treatment with sAp42 for Ap (green), scale bars 100 pm. Bottom: quantification of Ap clearance over time (n = 3, t-tests).

[0137] The invention is illustrated by the examples.

[0138] 1. Materials and Methods

[0139] Human

[0140] We analyzed brain tissue samples from two fetal donors (GW 15, male, unknown cause of death, postmortem interval ~3 days; and GW 17, male, cause of death: amniotic infection syndrome, post-mortem interval ~5 days) and cortical tissue from the frontal pole of one juvenile donor (6 years, female, white / caucasian, post-mortem interval 11 hrs). Experiments were approved by the ethics committee of the University Hospital, LMU Munich, and informed consent for collection of tissue samples and use for research purposes was given by the legal guardians. iPSC lines iPSC experiments were performed in accordance with all relevant guidelines and regulations. Wildtype iPSC-lines used in this study: 1) 7889SA2, a single cell clone derived from the male 7889SA line described by Paquet et al.

[0010] (use was approved by Rockefeller University Institutional Review Board after informed consent was obtained from subjects by Coriell Institute), 2) female iPSC line A18944 (ThermoFisher #A18945), and 3) KOLF2.1J (The Jackson Laboratory, #JIPSC001000). iPSC maintenance iPSCs were maintained as described previously

[0025] on vitronectin (VTN)-coated plates in Essential 8 Flex medium (E8F, ThermoFisher #A2858501). Cells were kept as colonies and split every 3-4 days using PBS with 0.5 mM UltraPure EDTA (ThermoFisher #15575020).

[0141] CRISPR / Cas9 gene editing

[0142] The CRISPOR online tool (http: / / crispor.tefor.net / 68] was used to design sgRNAs and determine the most likely off-target loci. Gene editing was performed as described previously

[0069] ,

[0070] with minor modifications. Briefly, sgRNAs were cloned into plasmid MLM36

[0038] (Addgene #43860), and plasmid pSpCas9(BB)-2A-Puro (PX459) V2.0 (Addgene #62988) was used for Cas9 expression. Singlestranded oligodeoxynucleotides (ssODNs) were used as repair oligos, designed manually, and purchased from IDT. Sequences of sgRNAs and ssODNs can be found in Table S1 . For gene editing, single cells were resuspended in BTXpress electroporation solution (VWR #732-1285), electroporated with 2 pulses at 65 mV for 20 ms in a 1 mm cuvette (Fisher Scientific #15437270) and transferred to 10 cm plates in StemFlex Medium Gibco #A3349301) with Revitacell (RVC) supplement (1 :100, ThermoFisher #A2644501). Cells were selected with 350 ng / ml Puromycin (VWR #J593) for three days starting the day after electroporation. Edited single-cell clones were analyzed by RFLP assay, using enzymes HpyCH4IV (NEB #R0619S) for APP Arctic, and Ddel (NEB #R0175S) for APP Iberian, followed by Sanger sequencing as described

[0070] . Successfully edited clones were expanded and subjected to quality controls.

[0143] Quality controls of edited iPSC

[0144] Genomic integrity at the edited locus and the BCL2L1 gene was confirmed by quantitative genomic PCR as described previously

[0047] ,

[0070] . Undifferentiated state of iPSC was confirmed by immunostainings as described below using Oct4, Tra160, SSEA4 and Nanog antibodies. To verify the origin of each edited cell line, a microvariation (“fingerprint”) at the human D1 S80 locus was analyzed by PCR. Off-target effects at the most probable loci, as predicted by CRISPOR were excluded by PCR followed by Sanger sequencing. General genomic integrity was confirmed by molecular karyotyping using HumanOmni2.5Exome-8 BeadChip v1.4 (Life & Brain GmbH, Bonn).

[0145] Cortical neuron differentiation from iPSC

[0146] Cortical neuron differentiation was performed as described recently

[0025] using a dual-SMAD inhibition approach and WNT inhibition using 5 pM XAV939 for the first 2 days. Late NPCs / early neurons were used for 3D cultures on day 38-45 of differentiation. 3D culture on days 20-70 of differentiation can be alternatively used depending on which cortical layer is desired.

[0147] Hippocampal neuron differentiation from iPSC

[0148] Hippocampal neuron differentiations followed the same basic steps as described for cortical neurons with additional compounds added to specify the cells toward a dorsomedial telencephalic fate and antagonize the formation of caudal and ventral tissue, For the first 10 days, media was additionally supplemented with 1 pM Cyclopamine and 5 pM XAV-939, and afterwards media was supplemented with 0.5 pM CHIR-99021 and 20 ng / ml BDNF to promote hippocampal identity, similar to a previously published protocol (Pomeshchik, DOI: 10.1016 / j.stemcr.2020.06.001). differentiation from iPSC

[0149] Differentiation of iPSC to astrocytes was performed following the protocol by Perriot et al.

[0026] ,

[0071] with modifications. Briefly, iPSCs were differentiated as in the cortical neuron differentiation protocol for 13- 15 days, when rosettes became visible. Media was then changed to glial precursor expansion medium as described by Perriot et al., and rosettes were manually isolated after 3 days as described in the neuron differentiation protocol. After growing confluent, rosettes were split to single cells using Accutase (Thermo Fisher #A1110501) after which the protocol by Perriot et al. was followed. Glial expansion medium was applied for a total of 6-9 days, astrocyte induction medium for 14 days, and astrocyte medium for 28 days. Astrocytes were then used for 3D cultures on day 64-69 of differentiation, but a usage from day 35 on is possible.

[0150] Microglia differentiation from iPSC

[0151] Differentiation of iPSC to microglia was performed following the protocol by McQuade et al.

[0072] with minor modifications as described by us recently

[0027] . Microglia were used for 3D cultures on day 24 of differentiation, but a usage from day 12 on is possible.

[0152] Generation of 3BTMs

[0153] To prepare 3BTMs only containing neurons late NPCs / early neurons were split into single cells using Accutase, plated into ultra-low-attachment, 96-well round-bottom plates (Sigma-Aldrich #CLS7007- 24EA) at 250.000 cells / well in 3D culture maintenance medium consisting of: Neurobasal Plus medium (ThermoFisher #A3582901) supplemented with B27 Plus supplement (1x, ThermoFisher #A3582801), GlutaMAX (0.5 mM, ThermoFisher #35050), and Penicillin / Streptomycin (100U / ml and 0.1 mg / ml, ThermoFisher #15140-122 ). Cells were spun down for 4 min at 200 g to improve cell aggregation and stored at 37 °C, 5% CO2, 80-90% humidity and ambient oxygen levels. 3D culture maintenance media was supplemented with 10 pM DAPT (Selleckchem #S2215) for at least 2 days, preferably 7 days after culture generation, and with 5 pM 5-Fluorouracil (5-FU, Sigma-Aldrich #F6627) and 5 pM Uridine (Sigma-Aldrich #U3750) from day 3 to day 15 after culture generation. At least 3 days of treatment are necessary to yield in a post-mitotic culture and driving terminal neuron differentiation. 7-10 days after generation, aggregated 3D cultures were transferred to wells of a 12-well plate coated with antiadherence rinsing solution (StemCell Technologies #07010) for long-term culture with half of the medium being replaced every 3-4 days. To prepare 3BTMs containing neurons and astrocytes, cells were split as described above, and plated at a neurons:astrocytes = 3:1 ratio with a total of 250.000 cells. Other ratios of neurons:astrocytes ranging from 1 :3 to 10:1 are possible. To prepare 3BTMs containing neurons, astrocytes, and microglia, neurons and astrocytes were mixed and aggregated as above. On day 15 after culture generation, when treatment with 5FU and uridine ended, cultures were transferred to ultra-low attachment 96-well V-bottom plates (Greiner Bio-One # 651970) in fresh 3D culture maintenance medium supplemented with 100 ng / ml IL43 (Peprotech #200-34), 50 ng / ml TGFpl (Peprotech #AF-100-21 C), and 25 ng / ml M-CSF (Peprotech #300-25). Microglia at day 24 of their differentiation were collected and added into the wells at 45.000 cells per 3D culture. The ratio vs. neurons at which microglia are added to the culture is a ratio of neurons:microglia of 4:1 , wherein other ratios ranging from 1 :1 to 20:1 are possible. When microglia had migrated into the cultures (7-10 days), cultures were transferred back to 12-well plates coated with anti-adherence rinsing solution for longterm culture as described above. The addition of microglia can also be done on day 20 after 3BTM generation. Later microglia addition to the culture however leads to accumulation of microglia on the edge of the culture yielding less distributed microglia across the culture and thus less microglia in the center of the culture. The resulting 3D BTM culture was cultured in neuron media with addition of cytokines, such as TGFp 1 for supporting development of microglia identity.

[0154] Generation of 2D co-cultures

[0155] To generate 2D co-cultures, cells were split as described above for 3BTMs and plated into 6-well plates coated with poly-L-Ornithine and Laminin as described previously

[0025] in the same media and cell ratios as described above. T reatments with DAPT and 5FU were performed as described above, and microglia were added as described above.

[0156] Seeding of 3BTMs with AB

[0157] For seeding experiments with recombinant A 42 (Holzel Biotech #AS-72216), lyophilized A 42 was prepared as previously described.5 In brief, A 42 was resuspended at 5 mM in DMSO and further diluted in PBS to 100 pM. Resuspended peptide was incubated at 4 °C for 24 h to allow oligomerization before being frozen at -80 °C. For experiments, recombinant A 42 was added into 3D culture maintenance media at 500nM on days 8 and 10.

[0158] Generation of brain organoids (BOs) iPSC to generate BOs were reprogrammed from healthy human fibroblasts (lines GM05399 - Coriell, and HPS0076 : 409B2 - RIKEN BioResource Research Center (BRC)) and cultured on Matrigel coated plates (Corning, 354234) in mTesRI basic medium (STEM Cell Technologies, 85857) at 37°C, 5% CO2, and ambient oxygen levels. Cells were dissociated for routine passaging using Gentle Cell Dissociation Reagent (STEM Cell Technologies, 100-1077).

[0159] Brain organoids were generated according to Lancaster et al.

[0073] with minor modifications. Briefly, iPSCs were dissociated into single cells using Accutase (Sigma-Aldrich, A6964), counted using automatic cell counter (LUNA-II™ Automated Cell Counter - Logos Biosystems) and 9000 cells per well were seeded to an ultralow-attachment 96-well plate (Corning) in human ES media with low concentration basic fibroblast growth factor (4 ng / ml1) and 50 pM Rho-associated protein kinase (ROCK) inhibitor (Calbiochem). After 6 days, media was replaced for neural induction media, and neuroepithelial tissues were fed every other day for 6 days. On day 12, embryoid bodies were transferred to droplets of Matrigel in differentiation media. After 4 days of stationary growth, organoids were moved to an orbital shaker at 57 rpm, at 37°C, 5% CO2, and ambient oxygen levels. Media was then replaced every 3-4 days. On Day 90, organoids were collected and washed with PBS. Once all the liquid was removed, organoids were snap-frozen on dry ice and kept at -80 °C.

[0160] Immunofluorescence stainings

[0161] For 2D immunofluorescence staining, cells were plated on acid-etched coverslips as described before

[0022] . At the day of fixation, cells were washed once with PBS and fixed with 4% PFA for 10 min at RT. Cells were then washed with PBS for 3x 5 min and then blocked for 1 hour at room temperature (RT) in blocking solution containing 3% donkey serum (Jackson ImmunoResearch #017-000-121), 0.1 % Triton X-100 (Thermo Fisher # 85111) and 0.02% NaN3 (Sigma-Aldrich #S2889) in PBS. Primary antibodies (Table S2) were diluted in blocking solution and incubation was performed overnight at 4 °C. Coverslips were washed with PBS for 3x5 min and then incubated with secondary antibodies (Table S2, 1 :500 dilution), and DAPI if required (ThermoFisher #D1306, stored as 5mg / ml stock solution, 1 :50.000 dilution), diluted in blocking solution for 2 hours at RT in the dark. Cells were washed again for 3x 5 min with PBS and mounted on glass slides (ThermoFisher #J1800AMNZ) in Fluoromount-G (ThermoFisher #00-4958-02). Coverslips were imaged with a fluorescent microscope (Zeiss Axio Observer Z.1) and images were processed in Fiji.

[0162] For 3D culture stainings, cultures were washed once with PBS, fixed with cold 4 % PFA for 30 min at RT, and washed again with PBS for 3x 30 min. Cultures were then transferred into 30 % sucrose in PBS over night at 4 °C, then transferred into OCT compound (Sakura Finetek Tissue-Tek O.C.T. Compound, Fisher Scientific #12351753), and cryosectioned into 30 pm slices. Slices were mounted onto glass slides, dried on a heat plate at 45°C, rehydrated with H2O, and transferred into a humidified box. All other steps were performed as described above, with 3x10 min washes in between incubations. TUNEL stainings were performed on slides as described above using the In Situ Cell Death Detection Kit, TMR red (Sigma-Aldrich #12156792910) according to manufacturer’s instructions. Hyaluronan (hyaluronic acid) was stained using biotin-coupled hyaluronic-acid-binding-protein followed by avidin-FITC-based detection following the protocol above. For hypoxia stainings, 3BTMs were incubated with 200 pM Pimonidazole HCI (Hypoxyprobe Omni Kit, Hypoxyprobe) for 2h at standard culture conditions (ambient 02 levels), or at 5% 02 for positive controls. Stainings were performed according to the manufacturer’s instructions, with primary antibody dilution 1 :50. Samples were imaged using a Zeiss LSM 880 confocal microscope using 10x, 40x and 100x objectives on ZEN black software (Version 10.0.4.910), and images were processed and analyzed with Fiji. TUNEL-positive cells, Ki67-positive cells, DAPI-positive cells and Ap puncta were quantified by thresholding the relevant channel and subsequent use of the “Analyze Particles” function. Synapses were quantified using the SynQuant plugin

[0074] . 3D surface renderings of synapses on / in microglia were done using Imaris Version 9.4 (Oxford Instruments). First, a surface was created based on the Iba1 staining to have a representation of the microglia, then only synapsin-1 -positive puncta touching this surface were displayed to remove all synapses not in contact with / not inside the microglia.

[0163] Amyloid-p (Ap) measurements in supernatant

[0164] Culture media was conditioned for 5 days, collected on ice, spun down at 300 g at 4 °C for 5 min, transferred into protein-low-binding tubes (Eppendorf #022431081), flash frozen in liquid nitrogen and stored at -80 °C. For analysis, supernatants were thawed on ice, centrifuged at 11 ,000 g for 10 min at 4 °C, and used to measure Ap38, Ap40 and Ap42 with the V-PLEX Plus Ap Peptide Panel 1 Kit (6E10, Meso Scale Discovery #K15200G) according to manufacturer’s instructions. “Total Ap” levels were calculated as the sum of the three different Ap isoforms. Measured Ap values were normalized to total protein or RNA levels of the cultures used for media conditioning. in extraction

[0165] Protein from 3D cultures was extracted in increasingly harsh buffers, namely 1) DEA buffer (0,2% DEA in 50mM NaCI, pH = 10), 2) RIPA buffer (20mM TRIS-HCI pH 7.5, 15mM NaCI, 1 mM Na2EDTA, 1 mM EGTA, 1 % NP-40, 1 % sodium deoxycholate, 2.5mM pyrophosphate), and 3) Guanidine-based buffer (Macherey-Nagel #740934) as “insoluble” fraction. To minimize protein degradation during the isolation, all buffers were supplemented with protease inhibitor (Sigma-Aldrich #P8340, 1 :200) and all steps up to addition of the Guanidine buffer were performed on ice or at 4 °C.

[0166] For the first step, 20-25 3D cultures per sample were pooled, washed once with PBS, transferred into 100 pl DEA buffer, and dissociated with ceramic beads (VWR # 432-0293) on a Precellys Evolution homogenizer at 6500 rpm for 30 sec at 4 °C. The resulting cell suspension was spun down for 10 min at 2500 g, the supernatant transferred to an ultracentrifuge tube (Beckman Coulter #357448) and centrifuged at 100.000 g for 30 min at 4 °C. The supernatant (“DEA fraction”) was neutralized by adding 0.5 M TRIS (pH = 6.75, 1 :10), and flash frozen. The pellet was resuspended in 20 pl RIPA buffer, homogenized together with the pellet formed after the initial centrifugation step, ultracentrifuged at 100.000 g for 30 min and the supernatant (“RIPA fraction”) flash frozen. The pellet was resuspended in 20 pl Guanidine buffer and spun down at 100.000 g for 30 min, and the resulting supernatant (“Guanidine fraction” = “insoluble fraction”) flash frozen. Protein concentrations were measured using Pierce BCA Protein-Assay (for DEA (undiluted) and Guanidine (1 :2 diluted) fractions) (ThermoFisher # 23227) and Pierce Detergent Compatible Bradford Assay Kit (for RIPA fraction, 1 :10 diluted, ThermoFisher #23246). Samples were analyzed with the MSD Human Ap V-PLEX kit (6E10) as above. RIPA and Guanidine fractions were diluted 1 :10 with Diluent 35 from the kit, and results normalized to total protein.

[0167] SDS-PAGE and Western Blotting

[0168] Total protein was extracted with RIPA buffer using a Precellys Evolution homogenizer as described above. For phospho-Tau detection, RIPA buffer was additionally supplemented with phosphatase inhibitors (Sigma-Aldrich #4906845001). Proteins were separated on 10% TRIS-Glycine gels hand- casted using the TGX Stain-Free FastCast Acrylamide Starter Kit, 10% (Bio-Rad #1610182), at 80 V for 25 min and at 120 V for 1-2 h, and transferred to nitrocellulose membranes (0.45 pm, VWR #10600002) at 100 V for 1 h on ice. Membranes were blocked in 3 % BSA (Sigma-Aldrich #A9418) in TBS with 0.1 % Tween-20 (TBS-T), incubated in primary antibody (AT270 1 :400, PHF-1 1 :200, K9JA 1 :10.000) in 1 % BSA in TBS-T overnight at 4 °C, and washed 3x 10 min in TBS-T. Blots were then incubated in HRP- labeled secondary antibodies (1 :10.000 in TBS-T) for 1 h at RT, washed 3x 10 min in TBS-T, and developed with Clarity Western ECL Substrate (BioRad #170-5060) on a Fusion Fx7 imager (Vilber). Band intensities were measured in Fiji.

[0169] Electrophysiological measurements

[0170] 3D cultures were embedded into 4% low-melting-point agarose (ThermoFisher #R0801) and cut into 250 pm sections. The agarose block was glued on a vibratome specimen holder and placed in 2-4°C cutting solution containing NMDG, KCI (2.5mM), MgCI2 (2mM), CaCI2 (2mM), NaH2PO4 (1.2mM), HEPES (10mM), NaHCO3 (21 mM), and Glucose (5mM), adjusted to pH 7.2 with HCI. Sections were recovered in 36°C cutting solution for 15min and 40-60 minutes at RT in aCSF containing NaCI (125 mM), KCI (2.5mM), MgCI2 (2mM), CaCI2 (2mM), NaH2PO4 (1 .2mM), NaHCO3 (21 mM), HEPES (10mM), and Glucose (5mM) adjusted to pH 7.2 with NaOH. Sections were continuously superfused with carbogenated aCSF while recording at a flow rate of ~2 ml / min at 30-32°C.

[0171] Current-clamp recordings were done in perforated patch clamp configuration. Cells were visualized with a fixed-stage microscope (BX51WI, Olympus, Hamburg, Germany) using a 60x water-immersion objective (LUMplan FL / N 60x, Olympus). Electrodes were prepared from borosilicate glass (Science Products #GB150-8P) with a vertical pipette puller (PC-10; Narishige) with tip resistances between 4 and 6 MQ. Recordings were performed with an EPC10 patch-clamp amplifier (HEKA, Lambrecht, Germany) controlled by PatchMaster (version 2.32; HEKA), and data acquired with a micro1410 data acquisition interface and Spike 2 (version 10) (both from CED) at 25 kHz and low-pass-filtered at 2 kHz with a four-pole Bessel filter. The calculated liquid junction potential of 14.6 mV was compensated or subtracted offline (calculated with Patcher’s Power Tools plug-in from https: / / www.mpibpc.mpg.de / groups / neher / index.php?page=softwarefor IGOR Pro 6 (Wavemetrics, Lake Oswego, OR, USA)). The pipette solution used for recordings contained K-gluconate (147mM), KCI (10mM), HEPES (10mM), EGTA (0.1 mM), and MgCI2 (2mM) adjusted to pH 7.2 with KOH. The patch pipette was tip filled and back filled with internal solution containing the ionophore amphotericin B (200 pg / ml, Sigma-Aldrich #A4888) and 0.02 % tetramethylrhodamine-dextran (3000MW, D3308, Invitrogen). Access resistance (Ra) was constantly monitored during the perforation process, and experiments were started after Ra had reached a steady state (-5-20 min) or the action potential amplitude was stable. Intrinsic characteristics were determined using a set of depolarizing and hyperpolarizing current injection either in perforated mode with stable action potential amplitude and / or membrane potential or after the membrane ruptured spontaneously.

[0172] 2-Photon live cell imaging

[0173] Experiments were performed using a Zeiss LSM7 MP multi-photon microscope with a W Plan- Apochromat 20x / 1.0 DIC D=0.17 M27 75mm objective. Cultures were immobilized in a 60 mm dish (VWR #734-2318) in 5 pl droplets of low-melting point agarose and 3D culture maintenance media was carefully added. Dishes were placed on a 37°C heat pad under the microscope and imaging was started at an excitation wavelength of 950 nm. Time lapse experiments of z-stacks were done at 1 pm axial step size, 512x512 pixel frame size, 3x zoom for 10-15 minutes with 40-50 seconds per z-stack and analyzed in Fiji.

[0174] Sinqle-cell RNA

[0175] To obtain a single cell suspension for scRNAseq, 4-5 cultures at 1 or 3 months of age were dissociated using the Neural Tissue Dissociation Kit P (Miltenyi Biotec #130-092-628) in a mix of 2.375 ml buffer X, 62.5 pl enzyme P, 10 pl enzyme A, 20 pl buffer Y and 11 .25 pM Actinomycin D (Sigma-Aldrich #A1410) in gentleMACS™ C Tubes (Miltenyi Biotec #130-093-237) on a gentleMACS™ Octo Dissociator (Miltenyi Biotec #130-095-937). A custom dissociation protocol was used (+ means counter-clockwise rotation, - means clockwise rotation): 4.5min +20rpm, 0.5min -100rpm, 4.5min +20rpm, 0.5min -250rpm, 4.5min +20rpm, 0.5min -300rpm, 4.5min +20rpm, 0.5min -300rpm, 2min +20rpm. The dissociated cell suspension was continuously kept on ice or at 4 °C, and plasticware was pre-treated with 1 % bovine serum albumin solution (BSA, ThermoScientific #15260-037) to prevent cell adhesion. Cells were filtered through a 70 pm cell strainer, spun down at 300 g for 5 min, resuspended in 50 pl FACS buffer (ThermoScientific #00-4222-26) and transferred to FACS tubes (Corning #352052). For multiplexing of biological replicates, cell suspensions were incubated for 20 min with Hashtag-coupled Antibodies (Biolegend #394631 or #394633, Table S2), washed once with FACS buffer, pelleted for 5 min at 300 g, and resuspended in 300 pl FACS buffer. Single cells were sorted on a FACSAria Fusion (BD Biosciences) into cooled protein low-binding tubes (Eppendorf #0030108132), DAPI (250 ng / ml) was added 5 min before sorting to exclude dead cells. Identical cell numbers of biological replicates (labelled with hashtag antibodies) were pooled into one collection tube, washed once in 0.04% BSA / PBS, and -16,500 cells per sample were loaded onto a 10x Chip G (10* Genomics). Single cell gene expression and cell hashing libraries were generated on the 10* Genomics Chromium platform using the Chromium Next GEM Single Cell 3’ Reagent Kits v3.1 according to the manufacturer’s protocol (CG000317 Rev C). Gene expression and hashtag libraries were sequenced on a NovaSeq 6000 S4 v1 .5 flow cell (Illumina).

[0176] CellRanger (v.7.1.0, 10X Genomics) was used to pre-process the Genomics GEX and TotalSeq data. The “cellranger multi” function was used to align raw sequencing reads against the default genome reference “refdata-gex-GRCh38-2020-A”. TotalSeq antibody tag sequences were provided to demultiplex the data by sample origin. All other CellRanger parameters were left as default. The Seurat pipeline (v4.3.0.1)

[0075] was employed for downstream analysis of preprocessed data. Cells were filtered for at least 20,000 read counts, 250 genes detected per cell and max. 7% mitochondrial reads, yielding a total of 14,462 cells.

[0177] Data were normalized and the 2000 most variable genes were selected for subsequent scaling and linear PCA dimensional reduction. The first 10 principal components were chosen as an input for nonlinear UMAP-based dimensionality reduction and clustering (resolutions.5), yielding 11 clusters of cells. Expression levels of key markers of microglia (AIF1 , TYROBP, CX3CR1 , and P2RY12), neurons (SOX11) and astrocytes (ALDH1 A1 , AQP4) were used to broadly annotate the cell types present in the whole dataset. For further investigation, the inventors focused on the analysis of microglia-marker- positive cells (13,946 cells) excluding two other mixed clusters (10 and 11) expressing both microglia (AIF1 , TYROBP) and neuronal (SOX11) genes (cluster 10 of 307 cells and cluster 11 consisting of 209 cells). Specific subsets of cells were generated in order to focus on either i) WT microglia cells including both 1- and 3-months’ time points, and ii) KI and WT microglia.

[0178] Clusters annotation was achieved by inspecting the top markers per cluster using the Seurat FindAIIMarkers function (only.pos = TRUE, min. pct = 0.25, logfc.threshold = 0.25). Targeted differential expression analyses between clusters of interest were conducted using the Seurat function FindMarkers (padj < 0.01 , logfc.threshold = 0.25, test.use=”MAST”).

[0179] Multiple gene signatures were used to assess the maturity level of microglia from 3D co-cultures across time points as well as to evaluate AD-like properties of microglia in the KI model. The sensome signature

[0036] , together with the adult microglia signature

[0040] and the late fetal microglia signature

[0039] were extrapolated as gene lists from the supplementary material published in literature

[0038] . An AD-associated microglia signature was also derived from supplementary table 4 published by Zhou et al. [8]. The signatures termed “early fetal” and “mid-fetal” were derived from genes obtained by contrasting two subpopulations within the Kracht et al. dataset

[0038] , namely microglia from gestational weeks (GW) 9 and 10 (“early fetal”) and GW 15 to 17 (“mid-fetal”). These were then compared with all other cells in the same dataset using the function FindMarkers (p_val_adj < 0.01 & avg_log2FC > 0). Likewise, the postnatal signature was obtained in a comparable fashion using the data from Han et al. (https: / / github.com / rzzli / FetalMicroglia) and applying FindMarkers (p_val_adj < 0.01 & avg_log2FC > 0, followed by selection of the top 400 genes) to find differentially expressed genes between postnatal and fetal cells

[0037] . These signatures were finally used for an enrichment test with the AddModuleScore and AddModuleScore_Ucell functions.

[0180] Mass spectrometry

[0181] 3D cultures, brain organoids, or human fetal and juvenile brain samples were washed once with PBS and dissociated in RIPA buffer (50 mM HEPES pH 7.5, 500 mM LiCI, 1 mM EDTA pH 8, 1 % NP-40, 0.7 % sodium deoxycholate) supplemented with protease inhibitors (Sigma-Aldrich #4693159001) and phosphatase inhibitors (Sigma-Aldrich #4906845001) using a Precellys Evolution homogenizer as described above. Three 3BTMs were pooled for each sample to increase total protein yield, and from fetal brain sample 1 (GW 17), two pieces of tissue were extracted. 2D cultures were dissociated and lysed by direct addition of the same RIPA buffer and scraped off the plate with a cell scraper. Lysates were transferred to protein low-binding tubes and spun down for 20 min at 18.000 g and 4 °C. Protein concentrations of supernatants were measured, and samples were frozen at -80 °C until analysis. 15 pg per sample were subjected to tryptic digestion. To this end, MgCL (10mM) was added, and DNA was digested with 25 units Benzonase (Sigma-Aldrich) for 30 min at 37°C. Proteins were reduced using 15 mM dithiothreitol (DTT) for 30 min at 37°C, followed by cysteine alkylation with 60 mM iodoacetamide (IAA) for 30 min at 20 °C. Excess IAA was removed by adding DTT. Detergent removal and subsequent digestion with 0.2 pg LysC and 0.2 pg trypsin (Promega, Germany) was performed using the single-pot, solid-phase-enhanced sample preparation as previously described

[0076] . After vacuum centrifugation, peptides were dissolved in 20 pL 0.1 % formic acid (Sigma-Aldrich) and peptide concentrations estimated using the Qubit protein assay (ThermoFisher). LC-MS / MS analysis

[0182] 350 ng of peptides were separated on a nanoElute nanoHPLC system (Bruker, Germany) in an in-house packed C18 analytical column (15 cm x 75 pm ID, ReproSil-Pur 120 C18-AQ, 1.9 pm, Dr. Maisch GmbH). Peptides were separated with a binary gradient of water and acetonitrile (B) containing 0.1 % formic acid at flow rate of 300 nL / min (0 min, 2% B; 2 min, 5% B; 94 min, 24% B; 112 min, 35% B; 120 min, 60% B) and a column temperature of 50°C.

[0183] The nanoHPLC was online coupled to a TimsTOF pro mass spectrometer (Bruker, Germany) with a CaptiveSpray ion source (Bruker, Germany). For relative protein quantification, a Data Independent Acquisition (DIA) - Parallel Accumulation Serial Fragmentation (PASEF) method was used

[0077] . Each scan cycle included one full MS scan followed by 34 windows of 26 m / z width (1 m / z overlap) covering an m / z range of 350-1200 using 2 windows per PASEF ramp of 100 ms. This resulted in a cycle time of 1 .9 s. Brain organoid and juvenile brain samples were run in three technical replicates.

[0184] Data Analysis

[0185] The MS raw data was analyzed with DIA-NN software (Version 1.8)

[0078] . First, a spectral library was generated in DIA-NN using 24 samples. Trypsin was defined as protease and 2 missed cleavages were allowed. The data was searched against a canonical one protein per gene human protein database from UniProt (download: 2022-01-18, 20600 entries). Oxidation of methionines and acetylation of protein N- termini were defined as variable modifications, whereas carbamidomethylation of cysteines was defined as fixed modification. The precursor and fragment ion m / z ranges were limited from 350 to 1200 and 200 to 1700, respectively. Peptide and peptide fragment tolerances were optimized by DIA-NN. The match between runs option was enabled. An FDR threshold of 1 % was applied for peptide and protein identifications. The generated spectral library includes 12550 protein isoforms, 13783 protein groups and 187812 precursors in 149807 elution groups. Afterwards, all samples were searched against this spectral library with the same settings for protein label-free quantification (LFQ).

[0186] Keratins were deleted from the results tables because of likely skin particle contaminations. For statistical analysis, results from all iPSC derived cultures were compared with each other by means of cell composition, and culture time. Therefore, only proteins were considered which were quantified in all 6 replicates of at least one experimental group. After Iog2 transformation of LFQ intensities, missing Iog2 LFQ intensities were imputed from a normal distribution using a width of 0.3 standard deviations of the data with a downshift of 1 .8 using the software Perseus (Version 1 .6.14.00). To account for multiple comparison, first a one-way ANOVA test was used to determine if there are any statistically significant differences between the means of experimental groups. Afterwards, individual Student’s T-tests were applied to evaluate proteins with significantly different abundances between the experimental groups. Additionally, a permutation based false discovery rate estimation was used with an FDR of 5% at sO = 0.1 as threshold to take multiple comparisons into account.

[0187] Statistical

[0188] No statistical methods were used to predetermine sample size and the experiments were not randomized, “n” denotes the number of independent experiments, meaning cultures prepared from independent batches of cell differentiations were analyzed. Experimental data was analyzed for significance using GraphPad Prism v10.1. Multiplicity-adjusted p < 0.05 was considered statistically significant. Significance was analyzed either by 2-sided Student’s t-tests (paired or unpaired, as indicated), if only 2 groups were compared, one-way ANOVA if more than 2 groups were compared with 1 independent variable, two-way ANOVA if more than 2 groups were compared with 2 independent variables. Multiple-comparison post-testing with Tukey’s or Sidak’s method was performed as recommended by the software and indicated. All graphs are shown as mean + standard deviation (SD) if not indicated otherwise. *p < 0.05, **p < 0.01 , ***p < 0.001 , ****p<0.0001 .

[0189] Example 2: 3D brain tissue model in vitro

[0190] Generation and characterization of the 3D brain tissue model

[0191] Generation of a reproducible and controllable 3D brain tissue-like model (3BTM) can be achieved by combining and aggregating precursors that have achieved a stable cell fate but are still plastic enough to allow tissue integration. Differentiation of neurons, astrocytes and microglia from iPSCs has been described in Example 1 , wherein the methods enable equal access to growth and patterning factors and promote homogeneous differentiation and culture purity. Next, the differentiated cells were selfaggregated into 3D assemblies by plating into ultra-low attachment plates (Figure 1A). Exogenous ECMs, such as Matrigel, were avoided to minimize variability due to exogenous growth factors or other cues, and thus increase reproducibility and simplicity. NE and AS were combined in a 3:1 ratio, similar to human cortex

[0028] , and aggregated into uniform, sphere-like structures of about 750 pM diameter (Figure 1 B). To avoid culture growth and necrotic core formation and reach a stable, postmitotic state rapidly after tissue aggregation, cultures were treated with DAPT, a Notch-signaling inhibitor, to drive residual NPCs into terminal neuron differentiation, followed by 5-Fluorouracil (5FU), a pyrimidine analog, to remove residual dividing cells (Figure 1A). The combination of optimized cell differentiations, optimal time points for cell aggregation, and compound treatments yielded 3BTMs with stable diameters over several months in culture without relevant growth or shrinkage (Figure 1 C). Further confirming efficacy of the drug treatments and subsequent post-mitotic state, the inventors found that proliferating cells were abundant on Day 3 after culture generation, before 5FU treatment, but almost completely absent at all later timepoints (Figure 1 D). In addition, cultures maintained the initial NE:AS ratio of 3:1 within and between independent experiments and over time (Figure 1 E), again verifying reproducibility and stability of cell ratios. Both NE and AS formed dense and uniform networks as seen by immunostainings for marker proteins (Figure 1 F), indicating mature cell morphology and no necrotic core formation. Hypoxia, which usually contributes to necrosis, was largely absent in 3BTMs grown under normoxic conditions, whereas 3BTMs cultured under hypoxic conditions showed positive cells after treatment with Pimonidazole (“Hypoxyprobe”), which specifically stains proteins in hypoxic cells

[0029] (Figure 1 G). To increase cellular complexity and physiological relevance of the cultures and enable analysis of MG states and behavior, MG were added to 3BTMs (Figure 1 H). MG migrated into the 3D cultures autonomously within few days (Figure 11) and spread within the culture (Figure 1 K), resembling the tissue tiling seen in brain. MG incorporation was strongly improved in the presence vs. absence of AS (Figure 1 L). Triple co-cultures could be kept for many months without necrotic core formation, with NE and AS maintaining densely branched networks (Figure 1 M, 1 N). Strikingly, MG adopted an increasingly ramified morphology (Figure 10) and expressed the homeostatic membrane marker P2RY12 in a growing percentage of cells over time, suggesting maturation in vitro (Figure 1 P). Importantly, such homeostatic microglial phenotypes are absent in organoids cultured in vitro and have so far only been achieved by transplantation in vivo

[0016] . MG numbers slowly decreased over time (Figure 1 Q), possibly due to senescence

[0029] or migration out of the culture towards a growth factor gradient into the media. However, survival was much higher than in current in vitro models

[0016] ,

[0017] , enabling analysis of physiological and neuroinflammatory phenotypes at advanced ages of 3 to 6 months. In addition, the inventors found that MG can be incorporated into 3BTMs also at later timepoints, e.g., at 5 months, leading to a dense, ramified population at 6 months of age (Figure 1 R, 1 S). P2RY12 expression was similar to MG at 1 month (Figure 1T). Next, cell death was analyzed using TUNEL stainings, and consistently low levels were found, also in long-term cultures, especially when all three cell types were present (Figure 1 U), suggesting that microglia not only benefit from but also contribute to the maintenance of a physiological tissue environment, e.g., by phagocytosis of potentially toxic degradation products. To confirm reproducibility of 3BTM generation across cell lines, the inventors repeated analyses of size (Figure 1V), NE:AS ratio (Figure 1W), post-mitotic state (Figure 1X) and cell morphology (Figure 1Y) in an independent iPSC line, which produced very similar results. Finally, the inventors showed formation of 3BTMs representing a different brain region, by differentiation of iPSCs into hippocampal neurons, expressing typical marker ZBTB20 (Nielsen et al., DOI: 10.1093 / cercor / bhs400), and their aggregation into 3D cultures (Figure 1Z) as described for the cortical cultures. In summary, the inventors established a highly reproducible, controllable, mature, and stable and thus improved 3D brain tissue model with efficient and long-lasting incorporation of NE, AS and MGL in a post-mitotic state, which offers unique opportunities to investigate phenotypes of microglia showing homeostatic markers in vitro.

[0192] Example 3: Maturation state of the 3D brain tissue model

[0193] Mass spectrometry-based principal component analysis and Gene ontology analysis

[0194] To investigate maturation of the 3D brain tissue model, mass spectrometry (MS) analysis of the cultures generated according to Example 2 at 3 days, 3 months, and 6 months after culture generation was performed, and compared to brain organoid, fetal and juvenile human brain samples. The generation of brain organoids has been described above in Example 1. Principal component analysis (Figures 2A,) showed a dense clustering of samples of similar ages and replicates, confirming reproducibility of the cultures also regarding overall protein composition. This was confirmed by high correlation between sample replicates (Figure 2H), with biological replicates of the generated 3BTM cultures showing similarly high coefficients as technical replicates of the brain organoid and juvenile brain samples. Samples were aligned along PC1 in a putative maturation-dependent manner, with brain organoids and 3-day-old samples on one end, and 3- and 6-months-old 3BTM samples placed between the fetal and juvenile human brain samples (Figure 2A).

[0195] Gene ontology analysis indeed yielded maturation-dependent, synapse-associated biological processes such as chemical synaptic transmission and ion / neurotransmitter transport as being upregulated, and cell proliferation and DNA replication-related processes as being downregulated along the PC1 axis (Figure 2B). In addition, "Cell adhesion” was one of the significantly altered processes in the GO analysis. The inventors focused on PC1 , as PC2 explained much less of the variance in the dataset and was largely driven by “blood-dependent” processes not relevant for our differentiations, such as complement activation, carbon dioxide / oxygen transport, or B cell receptor signaling, as well as DNA replication. The corresponding proteins, such as hemoglobin subunits and immunoglobulins, as well as dividing cells, are not present in the cultures but to some degree in patient samples and immature culture systems, explaining the separation on PC2.

[0196] To confirm relevance of the identified pathways and maturation of 3BTMs over time, synapse formation as the main process driving PC1 was analyzed. Electron microscopy analysis showed abundant synapses (Figure 2C, left), and in some cases also brain-typical tripartite synapses (Figure 2C, right), suggesting not only functional connections between neurons but also neuron-glia interactions. In addition, immunostainings for presynaptic Synapsin-1 and postsynaptic PSD95 revealed increasing numbers of colocalized puncta at older ages (Figure 2D), suggesting synapse formation. To get a more global overview of the synaptic machinery, all synapse-associated proteins in our proteomics dataset were analyzed and found strong upregulation of most proteins in 3- and 6-month-old cultures (Figure 2E), making them more similar to the fetal and juvenile brain sample than to organoids or the 3-day-old cultures used for comparison. The downregulation of cell-cycle-associated proteins at older ages in the cultures was confirmed, again indicating higher maturity and similarity to the brain samples and confirming the post-mitotic state of the cultures (Figure 2L). The organoid sample did not differ considerably from fetal brain, or the 3-day-old cultures used for comparison, indicating lower maturation. As astrocytes are a main driver of brain tissue maturation

[0032] , the inventors analyzed the proteomics dataset to determine levels of mature human astrocyte markers described previously

[0033] . A strong upregulation of most astrocyte-specific maturation markers was found in 3- and 6-month-old cultures, including Aldh1 l1 , Aqp4, and excitatory amino acid transporters. This led to a higher similarity to the juvenile than to the fetal brain sample as seen by hierarchical clustering, showing the high maturity of AS inside 3BTMs. To further study AS maturity, a small percentage of membrane-YFP-expressing AS was incorporated into the cultures to investigate single cell morphologies. The inventors found several different types of AS with intricate morphologies, including highly ramified, “bushy” cells resembling protoplasmic astrocytes in human brain [7], less ramified cells extending several processes over 200- 300 pm distance, and very large cells extending few processes over the entire culture diameter. Similarly, single neuron morphology after transduction with AAVs expressing GFP under the control of a human Synapsin-1 promotor was investigated. Again, a physiological cell morphology with a branched dendritic tree and long, ramified axon is observable (Figure 2I), resembling pyramidal neurons in the brain. As astrocytes have been shown to promote microglia ramification and maturation

[0056] , this may further promote the improved microglial phenotypes compared to brain organoids, where immature astrocytes only develop after 2-3 months in culture and in varying numbers

[0023] ,

[0024] . ECM maturation

[0197] The brain ECM is essential for a variety of cell functions, such as network formation and signaling events

[0012] however, due to its distinct properties, it so far cannot be closely replicated in vitro

[0031] . Many current in vitro models use Matrigel, a mouse-tumor-derived matrix, to embed cells and promote long-term stability. However, as a basement-membrane matrix, Matrigel is mainly composed of Laminins and Collagens

[0057] , while the brain ECM mainly contains Hyaluronan, chondroitin-sulphate proteoglycans, and linker proteins

[0058] . As the brain ECM has been shown to influence cell functionality

[0012] , these differences likely impair physiological cell maturation and function. In addition, Matrigel contains various growth factors which additionally differ between lots, making it an unpredictable variable in 3D cultures. No exogenous ECM has been therefore added when generating the cultures as the inventors hypothesized, that the cells would produce their own brain-like matrix. To investigate the formation of a brain-like ECM in the 3D BTMs, maturation markers indicative of the establishment of tissue properties were analyzed. Strikingly, hyaluronan, a main component of brain ECM, was first present at low, disperse levels on day 3 and strongly increased at 3 months to a uniform staining (Figure 2F) sparing cell bodies (Figure 2I), suggesting deposition in an ECM-like structure. Hyaluronan levels were strongly increased in the presence of AS (Figure 2J), identifying them as the main producers. To get a global overview of ECM formation, ECM-related proteins in the proteomics dataset was analyzed. A variety of ECM molecules, including brain-specific or -enriched core proteins of the chondroitin sulphate proteoglycans (CSPGs), such as Neurocan, Brevican, and CSPG5, linker proteins Tenascin and HAPLNs, and other relevant molecules such as hyaluronan sulphate proteoglycans (HSPGs) were detected (Figure 2G). The inventors found a strong upregulation of many brain ECM proteins in the 3- months-old and 6-months-old samples, again making them more similar to the fetal and juvenile brain compared to the 3-day-old and brain organoid samples. Moreover, a specific downregulation of proteins was observed which are less abundant in juvenile vs. fetal brain, e.g., Glypican-2 and -4, especially at 6 months, suggesting not only build-up but also brain-like remodeling of the culture ECM over time. Together, the result of the analysis confirms that cortical neurons and astrocytes in 3BTMs acquire mature morphologies and protein composition, and cultures display ECM formation, and functional features similar to juvenile to adult brain cells in vivo. scRNA sequencing

[0198] Given that the ramified microglia morphology and expression of P2RY12 already suggested microglia maturation (Figure 1 K), single-cell RNA sequencing analysis (sc-RNAseq) was performed to characterize microglial gene expression more broadly upon long-term integration into 3BTMs (Figure 3A). To study maturation of the cells over time, cells from two different experiments each at 1 and 3 months of age were analyzed. A gentle dissociation protocol including Actinomycin D as a transcriptional inhibitor was applied to ensure retrieval of the cells without artificial activation

[0034] . The MG-focused extraction resulted in recovery of very few neurons and astrocytes, which due to their highly complex structure and incorporation in strongly connected networks likely require more harsh dissociation procedures for efficient retrieval. The inventors focused on the microglia subset and embedded them in nearest neighbor space using UMAP (Figure 3B). A total of 7 microglia states were found that were clearly separated by the age of the source cultures, with “resting” and chemokine-associated clusters present at both 1 and 3 months and additional proliferative and hypoxia-associated clusters present at 1 month. MG in all clusters expressed typical markers such as AIF1 , TYROBP, and HexB (Figure 3C). Differential gene expression analysis (DEA) between MG from 1 -month vs 3-months-old cultures showed large differences between the timepoints, with 784 up- and 619 downregulated genes (Figure 3D). Gene set enrichment analysis (GSEA) of DEGs pointed towards a metabolic shift over time, as well as an upregulation of immune functions including “MHC class II protein complex assembly” and “antigen processing and presentation” (Figure 3D). These terms indicate the acquisition of an immune-sensing state over time, a marker for microglia maturation, enabling the cells to surveil the environment and react to aversive cues

[0035] . The inventors therefore used a previously published “sensome” signature

[0036] and compared it to our MG signatures at 1 and 3 months. MG at 3 months had a significantly higher enrichment for the sensome signature, indicating increased maturity and ability to interact with their environment (Figure 3E). Pseudo-time analysis further validated the phenotypic progression of microglia from 1 to 3 months in culture, as highlighted by the inferred trajectory (Figure 3F). Starting from proliferative cells at 1 month, the trajectory extended towards the intermediate 1 month “resting” state and further to the chemokine and hypoxic states, and branched towards the 3 months states, as expected. Further supporting the cell state annotation suggested in Figure 1 B for 1 month MG, pseudotime analysis highlighted an unfolding of phenotypes starting from a proliferative state, passing through a resting state, and then branching into the activated and hypoxic clusters. (Figure 3F). To get more detailed information on MG maturation and compare the cells to MG from human brain, the dataset was integrated with a published dataset of postmortem human fetal (10-14 gestational weeks (GW)) and postnatal MG (14-23 yrs)

[0037] . Intriguingly, MG from 1-month-old cultures clustered more closely with fetal microglia, while MG from 3-months-old cultures clustered close to postnatal microglia (Figure 3G), suggesting a strong maturation of MG in the cultures over time, making them more similar to postnatal cells. Consistently, 1 -month MG showed enrichment for the fetal MG signature, while 3-month MG showed enrichment for the postnatal signature (Figure S3B). When comparing both datasets to the adult microglia signature published by Galatro et al.

[0040] , not only higher enrichment of postnatal compared to fetal microglia, as expected, but also higher enrichment of MG at 3 months compared to 1 month (Figure 3H) was observed. To confirm these results, 3D BTM-derived MG signatures were compared to fetal and adult human MG signatures from an independent study (DOI: 10.1 186 / s12974-023-02809-7 (Yaqubi et al., 2023 J Neuroinflammation), and again a strongly increased enrichment of 3-month MG for the adult compared to fetal signature was observed (Figure 3I). Finally, MG signature at 3 months were compared to different human “in vitro” model signatures

[0020] , comprising primary MG cultured in 2D or incorporated into brain organoids, and iPSC-derived MG cultured in 2D or xenotransplanted into mouse brain. The inventors found the strongest enrichment for xenotransplanted MG (Figure 3K), which closely resemble in vivo microglia [9],

[0041] ,

[0055] . Taken together, the scRNAseq analysis confirms that the tissue-like environment observed on the morphological, functional, and proteomic level also induces maturation of microglia globally on the gene expression level to a degree currently unmatched by existing in vitro systems. Example 4: Functional analysis of 3D brain tissue model

[0199] The inventors investigated if adoption of more mature states on multiple levels would also lead to a functional maturation of the tissues. As extensive synapse formation and expression of various synapse- related proteins and ion channels were observed, electrophysiological patch-clamp measurements to analyze neuron functionality were performed. Not only spontaneous action potentials in a subset of neurons were observed (Figure 4A, J), but action potentials in all neurons by current injection could be evoked (Figure 4B), which was blocked by application of tetrodotoxin (Figure 4C), showing dependance on voltage-gated sodium channels. The inventors further observed differential responses in different neurons, possibly indicating presence of various neuronal subtypes (Figure 4K), as described for the human cortex

[0042] . Finally, cells showed a resting membrane potential of around -65 mV (Figure 4D), as shown before for mature neurons in human cortex

[0042] . Altogether, the data suggest that 3BTMs contain mature neurons with functional synapses that express canonical ion channels. To analyze functionality and network formation on a larger scale, whole-mount calcium imaging using Fluo-4 and detected spontaneous and synchronized calcium waves were performed (Figure 4L), suggesting functional network formation.

[0200] To investigate astrocyte functionality, the inventors focused on the cells’ ability to react to an activating stimulus, such as inflammatory cytokines. 3BTMs of neurons and astrocytes were treated with TNFa and IL1 a as described previously

[0043] to promote a reactive astrocyte state. After 24 hours, a robust increase in typical inflammation and reactive astrocyte markers NFKB, C3, SERPIN3A, and LCN2 (Figure 4E) by qPCR, confirming the cells’ ability to switch to an activated state were observed.

[0201] The functionality of MG as innate immune cells which constantly surveil their environment with their processes, remove debris and excess synapses, and immediately react to adverse stimuli evoked for example by a tissue injury was investigated. Immunostainings for MG and either cell nuclei or pre- synaptic Synapsin-1 showed that MG engulfing both nuclei of dead cell, as seen by bright and fragmented DAPI staining (Figure 4F), as well as synaptic material (Figure 4G). The inventors analyzed membrane-YFP-expressing MG embedded in 3BTMs at 3 months using 2-photon live-cell imaging. At baseline conditions, MG constantly surveyed their environment using their ramified processes (Figure 4H). Strikingly, when a focal laser injury was inflicted into the culture, MG processes from all surrounding cells immediately extended towards the lesion, leading to a complete ensheathment of the lesion site after around 10-12 minutes (Figure 4I), while the cell bodies remained in their original position. This shows that the MG can launch a classical injury response similar to cells in vivo

[0019] ,

[0035] and confirms presence of functional P2RY12 on the cell surface, which is required for this rapid response

[0044] ,

[0045] . Altogether, all cell types gained brain-like functionalities indicating functional maturation in our 3D tissue model.

[0202] Example 5: Addition of Oligodendrocytes (OLG)

[0203] An optional feature to increase the complexity of our 3D model and further explore the role of this major glial cell type in AD is the addition of oligodendrocytes (OLGs) into the 3D cultures. After treatment of the 3D cultures with DAPT (around day 10) the OLGs were added in a ratio of 5:1 (NE:OLG) into the media, and let the cells migrate into the culture (Fig. 5A). Maturity of the OLG was assessed over time by quantifying the levels of MBP (Fig. 5B). By electron microscopy the presence of myelinated axons was observed in the 3D BTM cultures showing mature and functional OLG (Fig. 5C). The addition of this cell type into the 3D BTM cultures of the present invention allows not only to help understating their putative role in AD but could also be applied to several other applications, including but not limited to:

[0204] - modeling other CNS diseases with a putative role of OLGs, such as multiple sclerosis, optic neuritis, transverse myelitis, acute disseminated encephalomyelitis (ADEM), neuromyelitis optica (NMO), Adrenoleukodystrophy, Alexander disease etc.; and developing treatments,

[0205] - studying mechanisms of human OLG physiology and myelination,

[0206] - investigating pathways and treatment approaches for myelin repair,

[0207] - studying crosstalk between glial cell and neurons in physiological conditions and disease.

[0208] Example 6: Application of the 3D brain tissue model to Alzheimer’s Disease studies

[0209] APP mutation generation by CRISPR / Cas9

[0210] Since the 3D BTM model of the present invention recapitulated many aspects of physiological brain tissue, the inventors tested if it can be applied to induce and study aspects of Alzheimer’s disease, such as accumulation of amyloid beta (Ap) and neuroinflammation. A combination of synergistic, AD-causing mutations into wildtype iPSC was introduced using CRISPR / Cas9 gene editing to induce rapid pathogenesis while avoiding overexpression artefacts. This strategy has already been successfully applied in mouse models AD, and the resulting NL-G-F mice

[0046] are broadly used in the field. Starting from a previously published line carrying the “Swedish” mutation in the APP gene

[0010] , the inventors added the “Arctic” and “Iberian” mutations, which accelerate protofibril formation and increase Ap42:40 ratio, respectively, into the APP gene of the same iPSC line. The resulting iPSC line was quality controlled for successful editing of all mutations by Sanger sequencing, presence of both alleles of the edited locus

[0047] and absence of a trisomy at the Bcl2l1 locus, a common pro-survival aberration found in stem cells. The iPSC line was confirmed for an undifferentiated state, isogenic derivation from the parental line by fingerprinting, absence of chromosomal aberrations by molecular karyotyping, and absence of off-target effects at the most likely sites. In summary, the inventors generated iPSC with a triple knock-in (KI) of AD-causing APP mutations that can be used to study AD pathogenesis in an isogenic setting. I ntriguing ly , cells also displayed increased tau phosphorylation using AT270 and PHF- 1 antibodies, showcasing the ability to study the Ap-tau axis, one of the largest open questions in AD research.

[0211] To analyze AD pathologies in a customized 3D model of the present invention, WT and KI iPSC were differentiated into the different cell types, generated 3BTMs according to the method as described above and analyzed phenotype formation over time. As expected from the introduced APP mutations, the inventors observed about 4-fold increased total Ap levels (Figure 6A) as well as a 50-fold increased Ap42:40 ratio (Figure 6B) in KI compared to WT cultures. This increased Ap production lead to deposition in 3BTMs over time, shown by a progressive accumulation of Ap in puncta and partly neurite- like structures specifically in KI cultures (Figures 6C, D). This could be prevented by application of a BACE inhibitor, confirming phenotype specificity, and demonstrating suitability of the system for small molecule drug treatments (Figure 6E, Figure 6F). Interestingly, presence of AS and MG increased Ap deposition in 3BTMs (Figure 6G), despite unaltered total Ap levels and Ap42:40 ratio (Figure 6H). As these Ap accumulations may represent early stages of plaque formation, the inventors analyzed if also insoluble Ap, indicating aggregation of the peptide were detectable. Sequential protein extractions were performed to analyze the insoluble fraction and indeed found insoluble Ap42 already in 1 -month-old KI cultures, and increased amounts at 6 months of age (Figure 6F), suggesting progressive aggregation. Phospho-Tau (pTau) levels were analyzed as a downstream pathology marker of amyloid pathology, to check if aspects of the Ap-Tau axis could also be studied using the above-described customized 3D BTM model of the present invention. I ntriguingly , increased levels of pTau T181 , an early tauopathy marker

[0048] , were observed in KI cultures at 3 months of age (Figure 6G, Figure 6I). This effect was strongest in the presence of both glial cell types, weaker if AS were left out, and not visible if MG were not added (Figure 6H, Figure 6I), suggesting that MG and their crosstalk with AS may play an important role in the Ap-tau axis. Moreover, phospho-Tau stained with the PHF-1 antibody was increased, which binds to misfolded and phosphorylated Tau representing advanced stages of Tauopathy (Figure 6I, Figure 6J).

[0212] Microglia in AD-3D brain tissue models

[0213] Given the presence of AD-like neuropathology in the 3D BTM model of the present invention, the inventors next studied neuroinflammatory phenotypes, which is enabled by the unique long-term survival of mature microglia and astrocytes in our cultures. qPCR showed increased levels of NFKB, a central mediator of proinflammatory responses, and CD68, a microglia activation marker (Figure 7A), as well as increased levels of C3 and C1 QC (Figure 7B), which have recently been implicated in inflammatory cellular crosstalk between AS and MG in AD

[0048] . To get a more global overview of disease-associated microglial alterations in our AD model, scRNAseq on WT and KI cultures at 3 months of age was performed. Strikingly, a major shift in MG signatures was found, with KI microglia forming two clusters separate from the resting cluster observed in WT MG (Figure 7C, D). DEG analysis between clusters “KI cluster 2” and “WT resting” yielded 129 significantly upregulated genes, including TREM2, CD14, and PSAP, which were previously described as dysregulated in human AD patients [8],

[0050] , and 109 downregulated genes (Figure 7E, F). TREM2 is involved in chemotaxis towards and subsequent containment and removal of Ap plaques [59-61]. CD14 has previously been described as a receptor for Ap which induces microglia activation

[0062] ,

[0063] and modulates AD pathogenesis

[0064] . PSAP is a lysosomal protein whose levels have been shown to correlate with AD pathology in patients

[0065] . C1 QB is involved in synapse pruning

[0066] , a developmental process thought to be aberrantly re-activated in early stages of AD68. In addition, all four have been previously shown to be increased in MG in postmortem AD human brain [8],

[0051] . GO analysis of upregulated genes showed enrichment in synapse pruning, chemotaxis, and immune response, while antigen presentation and processing were downregulated (Figure 7F). The microglial expression signature derived from the AD model was compared to a published signature from post-mortem human AD brains [8]. I ntriguingly , a significant enrichment of the patient-derived signature in our KI cultures, and especially in the “KI cluster 2” (Figure 7G) was observed, suggesting a recapitulation of disease-specific neuroinflammatory processes found in post-mortem AD brains in the above-described AD model. The inventors analyzed which genes would be upregulated in both MG signatures and found a large and strongly significant overlap (Figure 7H). The modular experiments with glial cells using the 3D BTM model of the present invention suggested that microglia and their crosstalk with astrocytes are at the interface between Ap and tau pathologies. Due to their modularity and reproducibility, 3BTMs may allow further investigating such crosstalk and elucidating the underlying mechanisms and roles of each cell type in the pathological cascade. Together, these findings demonstrate the suitability of the model of the present invention to replicate and thus study potentially human-specific MG activation in disease and provide insights into early, AD-related changes in MG. Overall, these data showcase unique opportunities to study early AD-relevant mechanisms in our in vitro model, which can hardly be studied in human brain due to inaccessibility of brain tissue, and as the processes modeled here would take place years to decades before diagnosis.

[0214] Addition of

[0215] Given that the 3BTMs not only replicated features of human brain tissue, but also AD-like phenotypes that can be modulated by BACE-inhibition, it was investigated if formation of dense-core plaques could also be induced and therapeutically modulated by adding exogenous Ap, as shown in 2D systems [5]. Eliciting late-stage plaque pathology as a central hallmark of AD in 3BTMs could enable studying downstream pathogenic processes such as neuritic dystrophies and associated tau pathology, as well as drug testing, e.g., for anti-amyloid antibodies. Dense-core plaques are characterized by an aggregated core of highly compacted Ap fibrils which can be stained with compounds binding to their characteristic p-sheet structure

[0051] , such as pFTAA

[0052] , and a surrounding halo of smaller, more soluble fibrils and oligomers

[0053] . To induce plaque formation, wildtype 3BTMs were treated twice with 500 nM recombinant Ap42. The peptide was prepared as described previously [5] to obtain a heterogeneous mix of soluble oligomers and fibrils. The inventors analyzed the cultures 1 month after addition and found even distribution of the recombinant Ap in the cultures (Figure 8A), aggregation into plaque-like structures with a core positive for pFTAA surrounded by fibrillar Ap (Figure 8B), and microglia positive for the exogenous peptide (Figure 8C), suggesting clearance.

[0216] Ap aggregates are an important drug target, as shown by the recent development of anti-amyloid antibody therapies. To explore the potential of 3BTMs to evaluate such antibody-based therapies and their efficiency in removing Ap aggregates, seeded 3BTMs were treated with Aducanumab and Ap accumulation and phagocytosis by microglia was analyzed in comparison to cultures treated with a nontargeting control antibody. After only 2 weeks of treatment, a strong colocalization of Aducanumab but not control antibody with Ap was observed (Figure 8D), indicating successful penetration into the cultures and target engagement. Control-antibody treated cultures already showed a reduction in Ap load over time, suggesting that the peptide is cleared by microglia. Confirming efficacy of the antibody therapy, Aducanumab treatment further reduced Ap load after 2-weeks and even stronger after 1 month of treatment (Figure 8E). In addition, colocalization of Aducanumab and Ap inside microglia was found (Figure 8F, G), and higher Ap load in MG in Aducanumab- compared to control-antibody-treated cultures (Figure 8H), confirming the importance of microglia in the clearance process. Finally, increased levels of CD68 in MG from Aducanumab-treated cultures were observed, suggesting increased phagocytosis and activation of the cells in reaction to the treatment (Figure 8I). Taken together, the inventors induced dense-core Ap plaque-like pathology in 3BTMs by adding exogenous Ap and showed the suitability of the model to profile anti-amyloid antibodies and microglial reaction to antibody therapy.

[0217] In addition, due to the high reproducibility and simplicity of culture generation, the model can be upscaled easily and thus used for drug profiling, testing, and screening. This is especially important as the lack of models replicating central disease features and amenable for drug testing is a major hurdle for drug development against AD. To demonstrate the suitability of 3BTMs to improve on this problem, the inventors confirmed the efficiency of p-secretase inhibitor IV (C3) in preventing Ap accumulation, as well as anti-amyloid antibody Aducanumab in removing aggregated Ap.

[0218] References

[0219] 1 . Limone, F., Klim, J.R., and Mordes, D.A. (2022). Pluripotent stem cell strategies for rebuilding the human brain. Front Aging Neurosci 14. https: / / doi.org / 10.3389 / fnagi.2022.1017299.

[0220] 2. Liu, R„ Meng, X., Yu, X., Wang, G., Dong, Z„ Zhou, Z., Qi, M., Yu, X., Ji, T., and Wang, F. (2022). From 2D to 3D Co-Culture Systems: A Review of Co-Culture Models to Study the Neural Cells Interaction. Int J Mol Sci 23. https: / / doi.org / 10.3390 / ijms232113116.

[0221] 3. Vadodaria, K.C., Jones, J.R., Linker, S., and Gage, F.H. (2020). Modeling brain disorders using induced pluripotent stem cells. Cold Spring Harb Perspect Biol 12, 1-15. https: / / doi.Org / 10.1101 / cshperspect.a035659.

[0222] 4. Klimmt, J., Dannert, A., and Paquet, D. (2020). Neurodegeneration in a dish: advancing human stem-cell-based models of Alzheimer’s disease. Curr Opin Neurobiol 61, 96-104. https: / / d0i.0rg / l 0.1016 / j.conb.2020.01 .008.

[0223] 5. Bassil, R., Shields, K., Granger, K., Zein, L, Ng, S., and Chih, B. (2021). Improved modeling of human AD with an automated culturing platform for iPSC neurons, astrocytes and microglia. Nat Commun 12, 5220. https: / / doi.org / 10.1038 / s41467-021-25344-6.

[0224] 6. Gotz, J., Bodea, L.-G. & Goedert, M. Rodent models for Alzheimer disease. Nat Rev Neurosci 19, 583-598 (2018). PMID: 30194347

[0225] 7. Oberheim, N.A., Takano, T., Han, X., He, W., Lin, J.H.C., Wang, F., Xu, Q., Wyatt, J.D., Pilcher, W., Ojemann, J.G., et al. (2009). Uniquely hominid features of adult human astrocytes. Journal of Neuroscience 29, 3276-3287. https: / / doi.org / 10.1523 / JNEURQSCI.4707-08.2009.

[0226] 8. Zhou, Y., Song, W.M., Andhey, P.S., Swain, A., Levy, T., Miller, K.R., Poliani, P.L., Cominelli, M., Grover, S., Gilfillan, S., et al. (2020). Human and mouse single-nucleus transcriptomics reveal TREM2-dependent and TREM2-independent cellular responses in Alzheimer’s disease. Nat Med 26, 131-142. https: / / doi.org / 10.1038 / s41591-019-0695-9.

[0227] 9. Mancuso, R., Van Den Daele, J., Fattorelli, N., Wolfs, L., Balusu, S., Burton, O., Liston, A., Sierksma, A., Fourne, Y., Poovathingal, S., et al. (2019). Stem-cell-derived human microglia transplanted in mouse brain to study human disease. Nat Neurosci 22, 2111-2116. https: / / d0i.0rg / l 0.1038 / S41593-019-0525-x.

[0228] 10. Paquet, D., Kwart, D., Chen, A., Sproul, A., Jacob, S., Teo, S., Olsen, K.M., Gregg, A., Noggle,

[0229] S., and Tessier-Lavigne, M. (2016). Efficient introduction of specific homozygous and heterozygous mutations using CRISPR / Cas9. Nature 533, 125-129. https: / / d0i.0rg / l 0.1038 / nature17664.

[0230] 11. Kwart, D., Gregg, A., Scheckel, C., Murphy, E. A., Paquet, D., Duffield, M., Fak, J., Olsen, O., Darnell, R. B. & Tessier-Lavigne, M. A Large Panel of Isogenic APP and PSEN1 Mutant Human iPSC Neurons Reveals Shared Endosomal Abnormalities Mediated by APP p-CTFs, Not Ap. Neuron 104, 256-270. e5 (2019). PMID: 31416668

[0231] 12. Barros, C.S., Franco, S.J., and Muller, U. (2011). Extracellular Matrix: Functions in the nervous system. Cold Spring Harb Perspect Biol 3, 1-24. https: / / doi.org / 10.1101 / cshperspect.a005108.

[0232] 13. Simao, D., Silva, M.M., Terrasso, A.P., Arez, F., Sousa, M.F.Q., Mehrjardi, N.Z., Saric, T., Gomes- Alves, P., Raimundo, N., Alves, P.M., et al. (2018). Recapitulation of Human Neural Microenvironment Signatures in iPSC-Derived NPC 3D Differentiation. Stem Cell Reports 11, 552-564. https: / / d0i.0rg / l 0.1016 / j.stemcr.2O18.06.020.

[0233] 14. Krencik, R., Seo, K., van Asperen, J. V., Basu, N., Cvetkovic, C., Barias, S., Chen, R., Ludwig, C., Wang, C., Ward, M.E., et al. (2017). Systematic Three-Dimensional Coculture Rapidly Recapitulates Interactions between Human Neurons and Astrocytes. Stem Cell Reports 9, 1745- 1753. https: / / d0i.0rg / l 0.1016 / j.stemcr.2O17.10.026.

[0234] 15. Lancaster, M.A., Renner, M., Martin, C.A., Wenzel, D., Bicknell, L.S., Hurles, M.E., Homfray, T.,

[0235] Penninger, J.M., Jackson, A.P., and Knoblich, J.A. (2013). Cerebral organoids model human brain development and microcephaly. Nature 2013 501 :7467 501, 373-379. https: / / d0i.0rg / l 0.1038 / nature12517.

[0236] 16. Uzquiano, A., and Arlotta, P. (2022). Brain organoids: the quest to decipher human-specific features of brain development. Curr Opin Genet Dev 75. https: / / d0i.0rg / l 0.1016 / j.gde.2022.101955. 17. Gao, C., Jiang, J., Tan, Y., and Chen, S. (2023). Microglia in neurodegenerative diseases: mechanism and potential therapeutic targets. Signal Transduct Target Ther 8. https: / / doi.Org / 10.1038 / S41392-023-01588-0.

[0237] 18. Ormel, P.R., Vieira de Sa, R., van Bodegraven, E.J., Karst, H., Harschnitz, O., Sneeboer, M.A.M., Johansen, L.E., van Dijk, R.E., Scheefhals, N., Berdenis van Berlekom, A., et al. (2018). Microglia innately develop within cerebral organoids. Nat Commun 9. https: / / doi.org / 10.1038 / s41467-018- 06684-2.

[0238] 19. Schafer, S.T., Mansour, A.A., Schlachetzki, J.C.M., Pena, M., Ghassemzadeh, S., Mitchell, L.,

[0239] Mar, A., Quang, D., Stumpf, S., Ortiz, I.S., et al. (2023). An in vivo neuroimmune organoid model to study human microglia phenotypes. Cell 186, 2111-2126. e20. https: / / d0i.0rg / l 0.1016 / j.cell.2023.04.022.

[0240] 20. Popova, G., Soliman, S.S., Kim, C.N., Keefe, M.G., Hennick, K.M., Jain, S., Li, T., Tejera, D., Shin, D., Chhun, B.B., et al. (2021). Human microglia states are conserved across experimental models and regulate neural stem cell responses in chimeric organoids. Cell Stem Cell 28, 2153- 2166. e6. https: / / d0i.0rg / l 0.1016 / j.stem.2O21 .08.015.

[0241] 21. Giandomenico, S.L., Mierau, S.B., Gibbons, G.M., Wenger, L.M.D., Masullo, L., Sit, T., Sutcliffe, M., Boulanger, J., Tripodi, M., Derivery, E., et al. (2019). Cerebral organoids at the air-liquid interface generate diverse nerve tracts with functional output. Nat Neurosci 22, 669-679. https: / / d0i.0rg / l 0.1038 / S41593-019-0350-2.

[0242] 22. Qian, X., Su, Y., Adam, C.D., Deutschmann, A.U., Pather, S.R., Goldberg, E.M., Su, K., Li, S., Lu, L., Jacob, F., et al. (2020). Sliced Human Cortical Organoids for Modeling Distinct Cortical Layer Formation. Cell Stem Cell 26, 766-781 . e9. https: / / doi.Org / 10.1016 / j.stem.2020.02.002.

[0243] 23. Renner, M., Lancaster, M.A., Bian, S., Choi, H., Ku, T., Peer, A., Chung, K., and Knoblich, J.A. (2017). Self-organized developmental patterning and differentiation in cerebral organoids. EMBO J 36, 1316-1329. https: / / doi.org / 10.15252 / embj.201694700.

[0244] 24. Pasca, A.M., Sloan, S.A., Clarke, L.E., Tian, Y., Makinson, C.D., Huber, N., Kim, C.H., Park, J.Y., O’Rourke, N.A., Nguyen, K.D., et al. (2015). Functional cortical neurons and astrocytes from human pluripotent stem cells in 3D culture. Nature Methods 2015 12:7 12, 671-678. https: / / d0i.0rg / l 0.1038 / nmeth.3415.

[0245] 25. Dannert, A., Klimmt, J., Cardoso Gonpalves, C., Crusius, D., and Paquet, D. (2023). Reproducible and scalable differentiation of highly pure cortical neurons from human induced pluripotent stem cells. STAR Protoc 4. https: / / doi.Org / 10.1016 / j.xpro.2023.102266.

[0246] 26. Perriot, S., Canales, M., Mathias, A., and Du Pasquier, R. (2021). Differentiation of functional astrocytes from human-induced pluripotent stem cells in chemically defined media. STAR Protoc 2, 100902. https: / / doi.Org / 10.1016 / j.xpro.2021.100902.

[0247] 27. Reifschneider, A., Robinson, S., Van Lengerich, B., Gnorich, J., Logan, T., Heindl, S., Vogt, M.A., Weidinger, E., Riedl, L., Wind, K., et al. (2022). Loss of TREM2 rescues hyperactivation of microglia, but not lysosomal deficits and neurotoxicity in models of progranulin deficiency. EMBO J 41, e109108. https: / / doi.org / 10.15252 / EMBJ.2021109108.

[0248] 28. von Bartheld, C.S., Bahney, J., and Herculano-Houzel, S. (2016). The search for true numbers of neurons and glial cells in the human brain: A review of 150 years of cell counting. Journal of Comparative Neurology 524, 3865-3895. https: / / doi.org / 10.1002 / cne.24040.

[0249] 29. Varia, M.A., Ch, M.B., Calkins-Adams, D.P., Rinker, L.H., Kennedy, A.S., Novotny, D.B., Fowler, W.C., and Raleigh, J.A. (1998). Pimonidazole: A Novel Hypoxia Marker for Complementary Study of Tumor Hypoxia and Cell Proliferation in Cervical Carcinoma. Gynecol Oncol 71, 270-277.

[0250] 30. Askew, K., Li, K., Olmos-Alonso, A., Garcia-Moreno, F., Liang, Y., Richardson, P., Tipton, T., Chapman, M.A., Riecken, K., Beccari, S., et al. (2017). Coupled Proliferation and Apoptosis Maintain the Rapid Turnover of Microglia in the Adult Brain. Cell Rep 18, 391-405. https: / / doi.Org / 10.1016 / j.celrep.2016.12.041 .

[0251] 31 . Rauti, R., Renous, N., and Maoz, B.M. (2020). Mimicking the Brain Extracellular Matrix in Vitro: A Review of Current Methodologies and Challenges. Isr J Chem 60, 1141-1151. https: / / d0i.0rg / l 0.1002 / I JCH.201900052. 32. Clarke, L.E., and Barres, B.A. (2013). Emerging roles of astrocytes in neural circuit development. Nat Rev Neurosci 14, 311-321. https: / / doi.org / 10.1038 / nrn3484.

[0252] 33. Zhang, Y., Sloan, S.A., Clarke, L.E., Caneda, C., Plaza, C.A., Blumenthal, P.D., Vogel, H., Steinberg, G.K., Edwards, M.S.B., Li, G., et al. (2016). Purification and Characterization of Progenitor and Mature Human Astrocytes Reveals Transcriptional and Functional Differences with Mouse. Neuron 89, 37-53. https: / / doi.Org / 10.1016 / j.neuron.2015.11.013.

[0253] 34. Wu, Y.E., Pan, L., Zuo, Y., Li, X., and Hong, W. (2017). Detecting Activated Cell Populations

[0254] Using Single-Cell RNA-Seq. Neuron 96, 313-329. e6. https: / / doi.Org / 10.1016 / j. neuron.2017.09.026.

[0255] 35. Nimmerjahn, A., Kirchhoff, F., and Helmchen, F. (2005). Resting Microglial Cells Are Highly Dynamic Surveillants of Brain Parenchyma in Vivo. Science (1979) 308, 1314-1318. https: / / doi.Org / 10.1126 / science.11 10647.

[0256] 36. Hickman, S.E., Kingery, N.D., Ohsumi, T.K., Borowsky, M.L., Wang, L.C., Means, T.K., and El Khoury, J. (2013). The microglial sensome revealed by direct RNA sequencing. Nat Neurosci 16, 1896-1905. https: / / d0i.0rg / l 0.1038 / nn.3554.

[0257] 37. Han, C.Z., Li, R.Z., Hansen, E., Trescott, S., Fixsen, B.R., Nguyen, C.T., Mora, C.M., Spann, N.J.,

[0258] Bennett, H.R., Poirion, O., et al. (2023). Human microglia maturation is underpinned by specific gene regulatory networks. Immunity 56, 2152-2171 . https: / / doi.Org / 10.1016 / j.immuni.2023.07.016.

[0259] 38. Kracht, L., Borggrewe, M., Eskandar, S., Brouwer, N., Chuva de Sousa Lopes, S.M., Laman, J.D., Scherjon, S.A., Prins, J.R., Kooistra, S.M., and L Eggen, B.J. (2020). Human fetal microglia acquire homeostatic immune-sensing properties early in development. Science (1979) 369, 530- 537. https: / / doi.Org / 10.1126 / science.aba5906.

[0260] 39. Zhong, S., Zhang, S., Fan, X., Wu, Q., Yan, L., Dong, J., Zhang, H., Li, L., Sun, L., Pan, N., et al. (2018). A single-cell RNA-seq survey of the developmental landscape of the human prefrontal cortex. Nature 555, 524-528. https: / / doi.org / 10.1038 / nature25980.

[0261] 40. Galatro, T.F., Holtman, I.R., Lerario, A.M., Vainchtein, LD., Brouwer, N., Sola, P.R., Veras, M.M., Pereira, T.F., Leite, R.E.P., Moller, T., et al. (2017). Transcriptomic analysis of purified human cortical microglia reveals age-associated changes. Nat Neurosci 20, 1162-1171. https: / / d0i.0rg / l 0.1038 / nn.4597.

[0262] 41. Svoboda, D.S., Barrasa, M.I., Shu, J., Rietjens, R., Zhang, S., Mitalipova, M., Berube, P., Fu, D., Shultz, L.D., Bell, G.W., et al. (2019). Human iPSC-derived microglia assume a primary microglialike state after transplantation into the neonatal mouse brain. Proc Natl Acad Sci U S A 116, 25293-25303. https: / / d0i.0rg / l 0.1073 / pnas.1913541 116.

[0263] 42. Moradi Chameh, H., Rich, S., Wang, L., Chen, F. Der, Zhang, L., Carlen, P.L., Tripathy, S.J., and Valiante, T.A. (2021). Diversity amongst human cortical pyramidal neurons revealed via their sag currents and frequency preferences. Nat Commun 12. https: / / doi.org / 10.1038 / s41467-021 - 22741-9.

[0264] 43. Liddelow, S.A., Guttenplan, K.A., Clarke, L.E., Bennett, F.C., Bohlen, C.J., Schirmer, L., Bennett, M.L., Munch, A.E., Chung, W.-S., Peterson, T.C., et al. (2017). Neurotoxic reactive astrocytes are induced by activated microglia. Nature 541, 481-487. https: / / doi.org / 10.1038 / nature21029.

[0265] 44. Davalos, D., Grutzendler, J., Yang, G., Kim, J. V., Zuo, Y., Jung, S., Liftman, D.R., Dustin, M.L., and Gan, W.B. (2005). ATP mediates rapid microglial response to local brain injury in vivo. Nat Neurosci 8, 752-758. https: / / doi.org / 10.1038 / nn1472.

[0266] 45. Haynes, S.E., Hollopeter, G., Yang, G., Kurpius, D., Dailey, M.E., Gan, W.B., and Julius, D. (2006). The P2Y12 receptor regulates microglial activation by extracellular nucleotides. Nat Neurosci 9, 1512-1519. https: / / doi.org / 10.1038 / nn1805.

[0267] 46. Saito, T., Matsuba, Y., Mihira, N., Takano, J., Nilsson, P., Itohara, S., Iwata, N., and Saido, T.C. (2014). Single App knock-in mouse models of Alzheimer’s disease. Nat Neurosci 17, 661-663. https: / / d0i.0rg / l 0.1038 / nn.3697. 47. Weisheit, I., Kroeger, J.A., Malik, R., Wefers, B., Lichtner, P., Wurst, W., Dichgans, M., and Paquet, D. (2021). Simple and reliable detection of CRISPR-induced on-target effects by qgPCR and SNP genotyping. Nat Protoc. https: / / doi.org / 10.1038 / s41596-020-00481-2.

[0268] 48. Gonzalez-Ortiz, F., Kac, P.R., Brum, W.S., Zetterberg, H., Blennow, K., and Karikari, T.K. (2023). Plasma phospho-tau in Alzheimer’s disease: towards diagnostic and therapeutic trial applications. Mol Neurodegener 18. https: / / doi.org / 10.1186 / s13024-023-00605-8.

[0269] 49. Guttikonda, S., and Studer, L. (2021). Fully defined hiPSC-derived microglia and tri-culture system model C3 production in AD. Nat Neurosci, 343-354.

[0270] 50. Mathys, H., Davila-Velderrain, J., Peng, Z., Gao, F., Mohammadi, S., Young, J.Z., Menon, M., He, L., Abdurrob, F., Jiang, X., et al. (2019). Single-cell transcriptomic analysis of Alzheimer’s disease. Nature 570, 332-337. https: / / doi.org / 10.1038 / s41586-019-1195-2.

[0271] 51. Glenner, G.G. (1981). The Bases of the Staining of Amyloid Fibers: Their Physico-Chemical Nature and the Mechanism of their Dye-Substrate Interaction. Prog Histochem Cytochem 13, 1- 35.

[0272] 52. Brelstaff, J., Ossola, B., Neher, J.J., Klingstedt, T., Nilsson, K.P.R., Goedert, M., Spillantini, M.G., and Tolkovsky, A.M. (2015). The fluorescent pentameric oligothiophene pFTAA identifies filamentous tau in live neurons cultured from adult P301 S tau mice. Front Neurosci 9. https: / / doi.Org / 10.3389 / fnins.2015.00184.

[0273] 53. Querol-Vilaseca, M., Colom-Cadena, M., Pegueroles, J., Nunez-Llaves, R., Luque-Cabecerans, J., Munoz-Llahuna, L., Andilla, J., Belbin, O., Spires-Jones, T.L., Gelpi, E., et al. (2019). Nanoscale structure of amyloid-p plaques in Alzheimer’s disease. Scientific Reports 2019 9:1 9, 1-10. https: / / d0i.0rg / l 0.1038 / S41598-019-41443-3.

[0274] 54. Abud, E.M., Ramirez, R.N., Martinez, E.S., Healy, L.M., Nguyen, C.H.H., Newman, S.A., Yeromin,

[0275] A. V., Scarfone, V.M., Marsh, S.E., Fimbres, C., et al. (2017). iPSC-Derived Human Microglia-like Cells to Study Neurological Diseases. Neuron 94, 278-293. e9. https: / / d0i.0rg / l 0.1016 / j. neuron.2017.03.042.

[0276] 55. Hasselmann, J., Coburn, M.A., England, W., Figueroa Velez, D.X., Kiani Shabestari, S., Tu, C.H., McQuade, A., Kolahdouzan, M., Echeverria, K., Claes, C., et al. (2019). Development of a Chimeric Model to Study and Manipulate Human Microglia In Vivo. Neuron. https: / / d0i.0rg / l 0.1016 / j. neuron.2019.07.002.

[0277] 56. Rezaie, P., and Male, D. (2002). Differentiation, Ramification and Distribution of Microglia within the Central Nervous System Examined. Neuroembryology 1, 29-43.

[0278] 57. Kleinman, H.K., McGarvey, M.L., Liotta, L.A., Robey, P.G., Tryggvason, K., and Martin, G.R.

[0279] (1982). Isolation and characterization of type IV procollagen, laminin, and heparan sulfate proteoglycan from the EHS sarcoma. Biochemistry 21, 6188-6193. https: / / doi.org / 10.1021 / BI00267A025.

[0280] 58. Lau, L.W., Cua, R., Keough, M.B., Haylock-Jacobs, S., and Yong, V.W. (2013). Pathophysiology of the brain extracellular matrix: a new target for remyelination. Nat Rev Neurosci 14, 722-729. https: / / d0i.0rg / l 0.1038 / NRN3550.

[0281] 59. Wang, Y., Celia, M., Mallinson, K., Ulrich, J.D., Young, K.L., Robinette, M.L., Gilfillan, S.,

[0282] Krishnan, G.M., Sudhakar, S., Zinselmeyer, B.H., et al. (2015). TREM2 lipid sensing sustains the microglial response in an Alzheimer’s disease model. Cell 160, 1061-1071. https: / / d0i.0rg / l 0.1016 / j.cell.2O15.01 .049.

[0283] 60. Mazaheri, F., Snaidero, N., Kleinberger, G., Madore, C., Daria, A., Werner, G., Krasemann, S.,

[0284] Capell, A., Trumbach, D., Wurst, W., et al. (2017). TREM2 deficiency impairs chemotaxis and microglial responses to neuronal injury. EMBO Rep 18, 1186-1198. https: / / d0i.0rg / l 0.15252 / embr.2O1743922.

[0285] 61. Kleinberger, G., Yamanishi, Y., Suarez-Calvet, M., Czirr, E., Lohmann, E., Cuyvers, E., Struyfs, H., Pettkus, N., Wenninger-Weinzierl, A., Mazaheri, F., et al. (2014). TREM2 mutations implicated in neurodegeneration impair cell surface transport and phagocytosis. Sci Transl Med 6. https: / / doi.org / 10.1126 / SCITRANSLMED.3009093 / SUPPL_FILE / 6-243RA86_SM.PDF. 62. Reed-Geaghan, E.G., Savage, J.C., Hise, A.G., and Landreth, G.E. (2009). CD14 and toll-like receptors 2 and 4 are required for fibrillar Ap-stimulated microglial activation. Journal of Neuroscience 29, 11982-1 1992. https: / / doi.org / 10.1523 / JNEURGSCI.3158-09.2009.

[0286] 63. Liu, Y., Walter, S., Stagi, M., Cherny, D., Letiembre, M., Schulz-Schaeffer, W., Heine, H., Penke, B., Neumann, H., and Fassbender, K. (2005). LPS receptor (CD14): A receptor for phagocytosis of Alzheimer’s amyloid peptide. Brain 128, 1778-1789. https: / / doi.org / 10.1093 / brain / awh531.

[0287] 64. Reed-Geaghan, E.G., Reed, Q.W., Cramer, P.E., and Landreth, G.E. (2010). Deletion of CD14 attenuates Alzheimer’s disease pathology by influencing the brain’s inflammatory milieu. Journal of Neuroscience 30, 15369-15373. https: / / doi.org / 10.1523 / JNEURGSCI.2637-10.2010.

[0288] 65. Mendsaikhan, A., Tooyama, L, Bellier, J.P., Serrano, G.E., Sue, L.L, Lue, L.F., Beach, T.G., and Walker, D.G. (2019). Characterization of lysosomal proteins Progranulin and Prosaposin and their interactions in Alzheimer’s disease and aged brains: Increased levels correlate with neuropathology. Acta Neuropathol Commun 7. https: / / doi.org / 10.1186 / s40478-019-0862-8.

[0289] 66. Stevens, B., Allen, N.J., Vazquez, L.E., Howell, G.R., Christopherson, K.S., Nouri, N., Micheva,

[0290] K.D., Mehalow, A.K., Huberman, A.D., Stafford, B., et al. (2007). The Classical Complement Cascade Mediates CNS Synapse Elimination. Cell 131, 1164-1 178. https: / / doi.Org / 10.1016 / J.CELL.2007.10.036.

[0291] 67. Hong, S., Beja-Glasser, V.F., Nfonoyim, B.M., Frouin, A., Li, S., Ramakrishnan, S., Merry, K.M.,

[0292] Shi, Q., Rosenthal, A., Barres, B.A., et al. (2016). Complement and microglia mediate early synapse loss in Alzheimer mouse models. Science (1979) 352, 712-716. https: / / doi.org / 10.1126 / SCIENCE.AAD8373 / SUPPL_FILE / HONG.SM.PDF.

[0293] 68. Concordet, J.P., and Haeussler, M. (2018). CRISPOR: Intuitive guide selection for CRISPR / Cas9 genome editing experiments and screens. Nucleic Acids Res 46, \N2 2-\N2 5. https: / / d0i.0rg / l 0.1093 / nar / gky354.

[0294] 69. Kwart, D., Paquet, D., Teo, S., and Tessier-Lavigne, M. (2017). Precise and efficient scarless genome editing in stem cells using CORRECT. Nat Protoc 12, 329-334. https: / / d0i.0rg / l 0.1038 / nprot.2016.171 .

[0295] 70. Weisheit, I., Kroeger, J.A., Malik, R., Klimmt, J., Crusius, D., Dannert, A., Dichgans, M., and Paquet, D. (2020). Detection of Deleterious On-Target Effects after HDR-Mediated CRISPR Editing. Cell Rep 31, 107689. https: / / doi.Org / 10.1016 / j.celrep.2020.107689.

[0296] 71. Perriot, S., Mathias, A., Perriard, G., Canales, M., Jonkmans, N., Merienne, N., Meunier, C., El Kassar, L., Perrier, A.L., Laplaud, D.A., et al. (2018). Human Induced Pluripotent Stem Cell- Derived Astrocytes Are Differentially Activated by Multiple Sclerosis-Associated Cytokines. Stem Cell Reports 11, 1199-1210. https: / / doi.Org / 10.1016 / j.stemcr.2018.09.015.

[0297] 72. McQuade, A., Coburn, M., Tu, C.H., Hasselmann, J., Davtyan, H., and Blurton-Jones, M. (2018). Development and validation of a simplified method to generate human microglia from pluripotent stem cells. Mol Neurodegener 13, 1-13. https: / / doi.org / 10.1186 / s13024-018-0297-x.

[0298] 73. Lancaster, M.A., and Knoblich, J.A. (2014). Generation of cerebral organoids from human pluripotent stem cells. Nat Protoc 9, 2329-2340. https: / / doi.org / 10.1038 / nprot.2014.158.

[0299] 74. Wang, Y., Wang, C., Ranefall, P., Broussard, G.J., Wang, Y., Shi, G., Lyu, B., Wu, C.T., Wang, Y., Tian, L., et al. (2020). SynQuant: an automatic tool to quantify synapses from microscopy images. Bioinformatics 36, 1599-1606. https: / / doi.org / 10.1093 / BIQINFQRMATICS / BTZ760.

[0300] 75. Hao, Y., Hao, S., Andersen-Nissen, E., Mauck, W.M., Zheng, S., Butler, A., Lee, M.J., Wilk, A. J., Darby, C., Zager, M., et al. (2021). Integrated analysis of multimodal single-cell data. Cell 184, 3573-3587. e29. https: / / doi.Org / 10.1016 / j.cell.2021 .04.048.

[0301] 76. Sielaff, M., Kuharev, J., Bohn, T., Hahlbrock, J., Bopp, T., Tenzer, S., and Distler, U. (2017).

[0302] Evaluation of FASP, SP3, and iST Protocols for Proteomic Sample Preparation in the Low Microgram Range. J Proteome Res 16, 4060-4072. https: / / doi.org / 10.1021 / ACS.JPRQTEQME.7B00433.

[0303] 77. Meier, F., Brunner, A.D., Frank, M., Ha, A., Bludau, L, Voytik, E., Kaspar-Schoenefeld, S., Lubeck, M., Raether, O., Bache, N., et al. (2020). diaPASEF: parallel accumulation-serial fragmentation combined with data-independent acquisition. Nature Methods 2020 17:12 17, 1229-1236. https: / / d0i.0rg / l 0.1038 / S41592-020-00998-0. Demichev, V., Messner, C.B., Vernardis, S.L, Lilley, K.S., and Raiser, M. (2019). DIA-NN: neural networks and interference correction enable deep proteome coverage in high throughput. Nature Methods 2019 17:1 17, 41-44. https: / / doi.org / 10.1038 / s41592-019-0638-x.

Claims

CLAIMS1. A method of generating an in vitro modular 3D brain tissue model (3D BTM) comprising: a. mixing 50,000 to 1 ,000,000 freshly splitted, single neurons (NE) and astrocytes (AS) or their precursor cells in a ratio of NE:AS = 1 :3 to 10:1 , followed by spinning down to introduce self-aggregation in low attachment conditions, without addition of exogenous matrices; b. treating cells obtained in step a. with a NOTCH signaling inhibitor for driving terminal neuron differentiation for at least 2 days; c. treating the cells obtained in step b. with at least one mitotic inhibitor to remove residual dividing cells for at least 3 days, thereby yielding a post-mitotic aggregated 3D culture of neurons and astrocytes; d. adding differentiated microglia (MG) to the culture obtained in step c. in a ratio of NE:MG = 1 :1 to 20:1 , to initiate migration of MG into the culture for 3-10 days; e. culturing the culture obtained in step d. in neuron media with addition of TGFpl and cytokines selected from the group comprising CSF1-R agonists to support microglia maintenance and proliferation for at least 2 weeks, thereby inducing formation of a modular in vitro 3D BTM culture.

2. The method of claim 1 , wherein culturing the culture obtained in step e. in neuron media with addition of cytokines supporting microglia maintenance and proliferation is preferably performed for 1-6 months, thereby maintaining formation of a mature, homeostatic microglial phenotype with ramified morphology and expression of P2RY12.

3. The method of claim 1 or 2, wherein the neurons and / or astrocytes, or their precursor cells, of step a. are obtained or obtainable by (i) differentiating iPSCs or (ii) differentiating embryonic stem cells by dual-SMAD inhibition or overexpression of transcription factors selected from the group including Ngn2, Ngn1 , ASCL1 , BRN2, MYT1 L, and NEUROD1 to directly differentiate neurons, or overexpression of transcription factors selected from the group including NFIA, NFIB, and Sox9 to directly differentiate astrocytes.

4. The method of claim 1 or 2, wherein the neurons and / or astrocytes or their precursor cells of step a., are obtained or obtainable by differentiating neural stem cells, and / or wherein the microglia of step d. are obtained or obtainable by differentiating hematopoietic precursor cells.

5. The method of any one of claims 1-4, wherein the neurons, astrocytes, their precursor cells and / or microglia are obtained or obtainable by differentiating iPSCs or stem cells that have undergone genome editing to insert mutations associated with neurological diseases selected from the group including Alzheimer’s Disease (AD), Frontotemporal lobar degeneration (FTLD), and Parkinson’s Disease (PD).

6. The method of any one of claims 1-5, wherein cells employed in step a. and / or d. are derived from a patient having a mutation causing a neurological disease.

7. The method of any one of claims 1-6, wherein low-attachment conditions comprise usage of ultralow attachment material.

8. The method of any one of claims 1-7, wherein the NOTCH inhibitor in step b. is DAPT, and the mitotic inhibitor(s) in step c. are pyrimidine analogs 5’-fluorouracil, cytosine arabinoside, FUdR (5-Fluoro-2'-deoxyuridine), and / or Aphidicolin, preferably 5’-fluorouracil.

9. The method of any one of claims 1 -8, wherein the microglia of step d. are added on day 10 to 40 of differentiation, preferably on or around day 24 of differentiation.

10. The method of any one of claims 1-9, wherein oligodendrocytes (OLG), endothelial cells or pericytes are added to the culture after step c. or d.

11. The method of any of claims 1-10 further comprising adding exogenous Ap42 to the 3D brain tissue model.

12. A modular in vitro 3D brain tissue model produced or producible by the method of any one of claims 1-1 1.

13. The in vitro 3D brain tissue model of claim 12 characterized by at least two features selected from the group of:(i) ramified / branched morphology of neurons and astrocytes and increasing numbers of synapses over the first 3 months in culture;(ii) ramified morphology of microglia; as defined by microglia extending 3 or more branched processes into the surrounding tissue with each process being longer than the average diameter of the cell body.(iii) no necrotic or hypoxic core;(iv) postmitotic state of neurons and astrocytes;(v) homeostatic microglia expressing P2RY12;(vi) expression of brain specific ECM-related proteins selected from the group comprising Neurocan, Brevican, CSPG5, and Tenascins;(vii) surveillance of cellular environment by microglia extending and retracting processes, and reaction of microglia to tissue damage by extending processes towards injury site within 5- 10 minutes while cell body stays in place;14. The in vitro 3D brain tissue model of claim 13 characterized by at least one further feature selected from the group of: a. matured astrocytes expressing Aldhl 11 and / or Aqp4;b. expression of reactive astrocyte markers upon addition of inflammatory cytokines, selected from the group comprising NFKB, C3, SERPIN3A, and LCN2; c. expression of excitatory and inhibitory synaptic markers, including GRIA1 and GABRA1 , respectively; d. presence of spontaneous action potentials indicating neuronal activity; e. mature microglia having an immune-sensing state; f. Increased enrichment for a postnatal human microglia signature upon maturation in tissue model; g. reaction of microglia to disease state, e.g., to pathologic Abeta in an Alzheimer’s disease model induced by endogenous mutations or exogenous seeding.

15. The in vitro 3D brain tissue model of any one of claims 12 to 14 for studying physiological and disease phenotypes of mature human brain.

16. The in vitro 3D brain tissue model of any one of claims 12 to 14 for studying Alzheimer’s Disease phenotypes selected from the group of Ap accumulation, phospho-tau increase, aggregation of Ap, neuroinflammation, and AD-related changes of microglia signature.

17. In vitro 3D brain tissue model of any one of claims 12 to 14 for drug profiling, testing and / or screening or toxicity testing, wherein the 3D brain tissue model is preferably upscaled.

18. The in vitro 3D brain tissue model of any one of claims 12 to 14, as far as obtained or obtainable by the method of claim 10 or 11 for studying physiological and disease phenotypes of multiple sclerosis, optic neuritis, transverse myelitis, acute disseminated encephalomyelitis (ADEM), neuromyelitis optica (NMO), adrenoleukodystrophy or Alexander disease.

Images